Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Localization of Kv1.5 in native and heterologous cell systems Eldstrom, Jodene 2005

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2005-104864.pdf [ 16.96MB ]
Metadata
JSON: 831-1.0092336.json
JSON-LD: 831-1.0092336-ld.json
RDF/XML (Pretty): 831-1.0092336-rdf.xml
RDF/JSON: 831-1.0092336-rdf.json
Turtle: 831-1.0092336-turtle.txt
N-Triples: 831-1.0092336-rdf-ntriples.txt
Original Record: 831-1.0092336-source.json
Full Text
831-1.0092336-fulltext.txt
Citation
831-1.0092336.ris

Full Text

Localization of Kvl.5 in native and heterologous cell systems By Jodene Eldstrom B.Sc , University of British Columbia, 1994 A THESIS SUBMITTED IN PARTIAL F U L L F I L L M E N T OF THE REQUIREMENT FOR THE DEGREE OF DOCTOR OF PHILOSOPHY In THE F A C U L T Y OF G R A D U A T E STUDIES Physiology THE UNIVERSITY OF BRITISH C O L U M B I A June, 2005 © Jodene Eldstrom, 2005 ABSTRACT Ion channel synthesis, trafficking and localization within the plasma membrane are highly regulated processes. They involve convergence of signals at the level of mRNA synthesis, chaperone-mediated folding and trafficking, interplay of kinases and phosphatases and assembly of macromolecular complexes fixed in place via anchoring proteins connected to the cytoskeleton. Little is known regarding the mechanisms that ensure the efficient surface membrane expression and localization of the potassium ion channel, K v l . 5 , within cardiac myocytes or even heterologous cell systems. The work presented here is part of an ongoing attempt to understand these mechanisms. A common protein-protein binding motif has been studied that binds PDZ proteins found to be important for the localization of many membrane proteins. The PDZ protein, PSD-95 but not SAP97, efficiently binds the C-terminal PDZ binding domain of Kvl .5 in yeast two-hybrid assays, GST pull-down experiments, and in co-immunoprecipitations. Both of these PDZ proteins can also regulate channel expression through the N-terminus of the channel. While PSD-95 and a channel C-terminal truncation mutant were co-immunoprecipitated, no direct interaction was detected with SAP97 despite obvious increases in channel surface expression. In the process of developing efficient detection methods for Kvl .5 in native cardiac tissue, the expression and localization of the channel in mammalian cardiac myocytes was defined to understand potential targeting and retention mechanisms. A detailed characterization of antibodies and expression of Kvl .5 in canine cardiac tissue unambiguously demonstrated a physiological role for the channel expressed in the atria and contributing to the repolarization phase of the atrial action potential. A more detailed examination of channel localization failed to find evidence of targeting to specialized membrane domains such as caveolae and/or lipid rafts. At the resolution of fluorescence microscopy only minor colocalization was found with the caveolar protein, caveolin-3, and the channel was absent II from light-buoyant fractions along with raft markers in sucrose gradient fractionations. Overall, the work in this thesis has begun to provide insight into the multiple mechanisms regulating Kv channel surface expression and localization. Clearly, such mechanisms will prove to be of central importance in both the physiological and pharmacological control of channel activity in cardiac tissues. Ill TABLE OF CONTENTS ABSTRACT II TABLE OF CONTENTS IV LIST OF TABLES VII LIST OF FIGURES VII ABBREVIATIONS VIII ACKNOWLEDGEMENTS X CO-AUTHORSHIP STATEMENT XII CHAPTER 1: INTRODUCTION 1 Overview 2 General Structure of Kv Channels 4 Role of K v Channels 8 Biophysical properties of Kvl .5 8 Role of Kv l .5 in the heart 9 Modulators of I K UR /Kvl .5 11 Adrenergic Modulation 12 Androgens 14 Thyroid Hormone 15 Hypoxia 16 pH 17 Growth Factors 18 Roles for interacting proteins 19 Known channel interactions 23 Kvp Subunits 25 a-Actinin 28 Src Family Kinases 30 Scope of thesis Investigation 32 REFERENCES 34 CHAPTER 2: N-terminal PDZ Binding Domain in Kvl Potassium Channels 49 INTRODUCTION 50 M A T E R I A L S A N D METHODS 52 D N A constructs and site-directed mutagenesis 52 Preparation of GST- and T7-tagged proteins 53 In vitro binding assays 53 Yeast two-hybrid experiments 54 Co-immunoprecipitation 54 Deglycosylation experiments 55 Imaging 55 Electrophysiological procedures 56 RESULTS 57 PDZ domains bind both Kvl .5 C- and N-termini 57 Yeast two-hybrid assays confirm interaction 59 PSD95 co-immunoprecipitates with C-terminally truncated Kvl .5 60 PSD95 and Kvl .5 variants co-localize in cultured cells 62 Delineation of the N-terminal PDZ-binding region in hKvl.5 65 PSD95 influences Kv l .5 K + currents 67 DISCUSSION 69 IV A C K N O W L E D G E M E N T S 74 REFERENCES 74 CHAPTER 3: SAP97 Increases Kvl.5 Currents through an Indirect N-terminal Mechanism 78 INTRODUCTION 79 M A T E R I A L S A N D METHODS 80 D N A constructs 80 Electrophysiological Procedures 81 Myocyte Isolation, Immunolabeling and Imaging 81 Co-immunoprecipitations 82 Yeast two-hybrid experiments 82 RESULTS 83 Co-expression of SAP97 with hKvl.5 enhances hKvl.5 currents in H E K cells 83 hKvl.5 and SAP97 fail to co-localize in rat cardiac myocytes 85 hKvl.5 co-immunoprecipitations with SAP97 from cardiac myocytes 85 hKvl.5 fails also to co-immunoprecipitate with SAP97 from transfected H E K cells 85 SAP97 interacts with Kv l .4 but not with hKvl.5 in yeast two-hybrid experiments 87 SAP97 enhancement of hKvl.5 currents depends on an intact Kvl .5 N-terminus 87 a-Actinin2 co-immunoprecipitates with both hKvl.5 and SAP97 90 DISCUSSION 93 A C K N O W L E D G E M E N T S 96 REFERENCES 96 CHAPTER 4: Kvl.5 Is An Important Component Of Repolarizing K + Current In Canine Atrial Myocytes 99 INTRODUCTION 100 M A T E R I A L S A N D METHODS 101 RESULTS 103 Detection of Kv l .5 and Kv3.1 in dog atrium 103 Kv l .5 is expressed on the surface of canine atrial myocytes 107 Pharmacological studies suggest that Kvl .5 contributes to iKUR(d) 110 DISCUSSION 117 Detection of Kvl .5 in canine cardiac myocytes 117 Failure to detect Kv3.1 using molecular methods 118 Pharmacological properties of IicuR(d) and prolongation of the canine atrial action potential 118 Implications 120 Conclusions 120 A C K N O W L E D G E M E N T S 121 REFERENCES 121 CHAPTER 5: Kvl.5 and caveolin-3 do not interact or colocalize in rat or canine cardiac myocytes 124 INTRODUCTION 125 M A T E R I A L S A N D METHODS 127 Preparation of canine heart lysates 127 Myocyte preparation 128 Immunolabeling, and wide-field microscopy 128 Preparation of Tokuyasu sections and immunolabeling 129 Sucrose gradient raft fractionation 129 RESULTS 131 V Kvl .5 and Caveolin-3 are found in rat and canine cardiac tissue 131 Kvl .5 and Caveolin-3 do not co-immunoprecipitate with each other 131 Kvl .5 and Caveolin-3 show minimal co-localization in deconvolved fluorescent images 133 Kvl .5 and Caveolin-3 show little evidence of co-localization in immuno-EM samples ... 135 Kv l .5 does not co-sediment with Caveolin-3 in sucrose gradient fractionations 137 DISCUSSION 140 A C K N O W L E D G M E N T S 145 REFERENCES 145 C H A P T E R 6: DISCUSSION 149 SAP97 and PDZ domains: mechanisms and molecules that regulate Kvl .5 channel expression 150 Role of the amino terminus of the Kvl .5 protein in cell surface expression 151 Future studies that can shed more light on Kvl .5 surface regulation 155 Apparent cellular localization of Kv l .5 at the intercalated disc and in caveolae 156 Summary 158 REFERENCES 159 A P P E N D I C E S 163 APPENDIX A: Single Letter Amino Acid Sequence of human Kvl .5 164 APPENDIX B: Online Supplemental Material To Chapter 4 165 Channel expression in cell lines and electrophysiological recording conditions 165 Preparation of canine heart and brain membrane and cytosol fractions 167 Canine myocyte preparation 168 Canine atrial voltage / current clamp recordings 168 R N A isolation 169 RT-PCR 170 Fixing, immunolabeling, and imaging of myocytes and HEK293 cells 171 REFERENCES 174 VI LIST OF TABLES Table I: Results of a High Stringency Motif Scan of Human Kvl .5 24 LIST OF FIGURES Figure 1.1. Phylogenetic trees of identified Kv channels 3 Figure 1.2. K v channel structure 4 Figure 1.3. Cartoon of a voltage-dependent K + channel (red) with associated P-subunit (blue).. 25 Figure 2.1. PDZ domains bind N-termini of K v l channels 58 Figure 2.2. Yeast two-hybrid experiments demonstrate PDZ binding to Kv channel N - and C-termini 61 Figure 2.3. Co-immunoprecipitation of Kv l .5 variants with PSD95 from H E K and COS cells. 63 Figure 2.4. PSD95 colocalizes with Kvl .5 and K v l . 5 A E T D L in transfected HEK293 cells 64 Figure 2.5. Analysis of PDZ domain binding to Kvl .5 N-terminal deletion constructs 66 Figure 2.6. Effect of PSD95 expression on peak currents on WT K v l . 5 , and C- or N-terminal deletion mutants 68 Figure 3.1. Effect of SAP97 expression on hKvl.5 peak currents 84 Figure 3.2. SAP97 does not co-localize with Kvl .5 in rat ventricular myocytes 86 Figure 3.3. Attempted co-immunoprecipitation of SAP97 and Kv l .5 88 Figure 3.4. Yeast two-hybrid assay for the interaction of K v l .5 N - and C-terminal interactions with SAP97 89 Figure 3.5. A . SAP97 fails to increase Kvl.5AN209 peak currents 91 Figure 3.6. Both Kvl .5 and SAP97 interact with a-actinin2 in rat heart 92 Figure 4.1. Kv l .5 and Kv3.1 expression in HEK293 cells, canine heart, and brain 105 Figure 4.2. Attempted amplification of K v l . 5 , Kv3.1, and dystroglycan from canine atrium... 106 Figure 4.3. Detection of T7-tagged Kvl .5 and Kv3.1 in H E K cells by different antibodies 108 Figure 4.4. Kv l .5 is expressed at the surface of canine atrial myocytes 109 Figure 4.5. Effects of a Kvl.5-specific drug on cloned channels in HEK293 cells 111 Figure 4.6. Effects of 0.5 mmol/L T E A + and C9356 on canine atrial K + currents 113 Figure 4.7. Effects of 5 mmol/L T E A + and C9356 on K + currents in canine atrial myocytes... 114 Figure 4.8. Effects of C9356 and T E A + on canine atrial action potentials 116 Figure 5.1. Kv l .5 and Caveolin-3 are expressed in rat and canine atrial and ventricular tissues. 132 Figure 5.2. K v l . 5 and Caveolin-3 fail to co-immunoprecipitate from rat and canine atrial and ventricular homogenates 134 Figure 5.3. Minimal co-localization of Kvl .5 and Caveolin-3 in wide-field images of rat and canine cardiac myocytes 136 Figure 5.4. Kv l .5 and Caveolin-3 do not co-localize in immuno-EM images of rat cardiac myocytes 138 Figure 5.5. Kv l .5 does not localize to the lipid raft fraction in HEK293 cells or rat ventricular tissue 139 Figure 6.1. Single-letter amino acid sequence of human Kvl .5 164 Figure 6.2. Kv l .5 expression in canine atria and rat ventricle 173 VII ABBREVIATIONS Amino Acid Alanine Arginine Asparagine Aspartate Cysteine Glutamine Glutamate Glycine Histidine Isoleucine Leucine Lysine Methionine Phenylalanine Proline Serine Threonine Tryptophan Tyrosine Valine 3 Letter Code 1 Letter Code Ala A Arg R Asn N Asp D Cys C Gin Q Glu E Gly G His H He I Leu L Lys K Met M Phe F Pro P Ser S Thr T Try W Tyr Y Val V 4-AP - 4-aminopyridine AF - Atrial Fibrillation A K A P - A Kinase Associated Protein AP - Action Potential APD - Action Potential Duration B A P T A - bis-(o-aminophenoxy)ethane-N,N,N',N'-tetra-acetic acid CFTR - Cystic fibrosis transmembrane conductance regulator CHIP - C-terminus of Hsc70-Interacting Protein CHO - Chinese Hamster Ovary COPI - Coat Protein Complex type I COS-7 - African green monkey kidney cells CRE - cAMP Response Element CREB - cAMP Response Element Binding protein C R E M - cAMP Response Element Modulator protein D A G - Diacylglycerol ER - Endoplasmic Reticulum E R A D - Endoplasmic Reticulum Associated Degradation E R K - Extracellular-signal-related kinase FGF - Fibroblast Growth Factor H C N - Hyperpolarization-activated and cyclic-nucleotide gated HEK293 - Human Embryonic Kidney 293 hERG - human ether-a go-go channel HPV - Hypoxic Pulmonary Vasoconstriction Hsc40 - Heat shock protein 40 Hsc70 - Heat shock cognate 71 kDa protein IC50 - 50 % inhibitory concentration lea - calcium current VIII ID - Intercalated Disc IGF - Insulin-like Growth Factor IP3 - Inositol 1,4,5-tris-phosphate KATP - ATP-sensitive potassium channel K B F - K R E Binding Factor KChAP - Kv Channel Accessory Protein KChIP - K + channel Interacting Protein Kir - inwardly-rectifying potassium K R E - K v l .5 Repressor Element Kv - voltage-gated potassium LIZ - Leucine-Isoleucine Zipper M H C - Myosin Heavy Chain NGF - Nerve Growth Factor N M D A - N-methyl-d-aspartate PASMCs - Pulmonary Artery Smooth Muscle Cells PDGF - Platelet-Derived Growth Factor PDZ-PSD-95/DLG/ZO- l PI3K - Phosphoinositide 3,4,5"-triphosphate P K A - Protein Kinase A P K C - Protein Kinase C PLC - Phospholipase C PP1 - Protein Phosphatase 1 R E M - Rapid Eye Movement RMP - Resting Membrane Potential SAP97 - Synapse Associated Protein -97 SH3 - Src Homology 3 T l - Tetramerization T3 - triiodothyronine TdP - Torsades de pointes T E A + - Tetraethylammonium ion T H - Thyroid Hormone T M - Transmembrane V1/2 - Half activation voltage V m - Membrane Potential IX A C K N O W L E D G E M E N T S I would like to begin by thanking my supervisor, David Fedida, who has been extremely encouraging and supportive of me from the beginning, when even I thought this to be an impossible dream. He has created an excellent research environment, providing not just the right equipment and scientific support, but all my financial and a good deal of emotional support. He has also been very generous in allowing me to attend many scientific meetings and provided me with opportunities to collaborate, both from which I have derived a great deal of enjoyment. I also appreciate the latitude I have been given to pursue many things that I found of interest, but were never a sure thing for publication. David Mathers kindly gave me my start in graduate studies and then allowed me to follow my interests into a molecular and cellular laboratory, and I thank him for supporting me in that decision and his support along the way as well. I have had the privilege of learning from, and working with some other very talented scientists in the Fedida lab, Zhuren Wang and David Steele, who has also been a kindred spirit in my attempts to tread lightly upon this beautiful planet of ours. Other members of my supervisory committee Steve Kehl and Ross MacGillivray, have been nothing but supportive and encouraging and I thank them for reading my thesis and for their helpful input. David Van Wagoner has been a pleasure to collaborate with, always game to try an experiment for us, always positive and fun to chat with, whether of science or politics (maybe I shouldn't say always positive). There are many friendly faces in the lab and department, with whom I have shared many laughs and who have made a sometimes trying existence more bearable, including my dear friend Grace and especially Harley Kurata, who was there to sound out ideas with and who bore my psychological evaluations with the utmost of grace. The staff at the U B C Bio-imaging X facility have always been supportive and fun, and I appreciate their inclusion of me in the Cryo-E M course these last few years, which has allowed my to interact and learn from some very knowledgeable and interesting people. I would like to thank Marilyn and Rick Nichols, my kayaking buddies, who were my support team during the writing of this thesis. As they have always done, they went above and beyond normal human kindness and I will always miss the day-to-day contact we had during this time. Thank-you John and Odilia, friends who are there for a quick coffee or long walks, dinner, cards and a bottle of wine and temporary shelter during my many "transitions". Finally, there is my family, who has always been there for me and have done whatever they could to get me through the tough times and shared the joy of the good, however fleeting. I suppose I should also thank Telus for the $20/month plan that meant I could share with my mum, the day-to day joy and agony of writing a thesis. XI CO-AUTHORSHIP STATEMENT CHAPTER 2: Eldstrom J, Doerksen KW, Steele DF, Fedida D. (2002) N-terminal PDZ-binding domain in Kvl potassium channels. FEBS Lett. 531(3):529-37. Jodene Eldstrom was responsible for project management, experimental design, some of the cloning, mutagenesis, and in vitro bind assays, all of the co-immunoprecipitations, transfections, stable cell line generation, deglycosylation experiments, immunocytochemistry, confocal imaging, data analysis, some of the figure preparation and together with Dave Steele some of the writing and editing the manuscript. (Percent Contribution: 60%) Kyle Doerksen collected and analyzed the electrophysiology data. David Steele did the majority of the writing and along with David Fedida was an advisor to the project. David Fedida also assisted with writing and editing the manuscript. Chapter 3: Eldstrom J, Choi WS, Steele DF, Fedida D. (2003) SAP97 increases Kvl.5 currents through an indirect N-terminal mechanism. FEBS Lett. 547(l-3):205-ll. Jodene Eldstrom was responsible for project management, experimental design, immunoprecipitations and western blots, confocal imaging, cloning and transfections, data analysis, and together with Dave Steele, writing and editing the manuscript. (Percent Contribution: 60%) Woo Sung Choi collected and analyzed the electrophysiological data. David Steele did the majority of the writing and along with David Fedida was an advisor to the project. David Fedida also assisted with writing and editing the manuscript. CHAPTER 4: Fedida D, Eldstrom J, Hesketh JC, Lamorgese M, Castel L, Steele DF, Van Wagoner DR. (2003) Kvl.5 is an important component of repolarizing K+ current in canine atrial myocytes. Circ Res.93(8):744-51 Jodene Eldstrom was responsible for: i) all the immunocytochemistry and confocal imaging, including antibody testing and controls in HEK293 cells and canine myocytes; ii) sample preparation and testing of antibody cross reactivity using western blots and peptide blocking controls; iii) preparation of related figures, results and discussion and assisted in writing and XII editing each draft of the manuscript. (Percent contribution 25 %). David Fedida did the majority of the writing for the manuscript and supervised the overall project. Christian Hesketh tested the selectivity of C9356 for various cardiac ion channels expressed in HEK293 cells. Michelle Lamorgese, Laurie Castel and David Van Wagoner are responsible for the canine myocyte electrophysiology (EP) and isolation of myocytes for immunocytochemistry. David Van Wagoner was also responsible for most EP figure preparation and some writing and editing. David Steele was responsible for design of primers and oversight of RT-PCR experiments. C H A P T E R 5: Eldstrom J , Yu Q, Martens G , Van Wagoner DR, Moore E , Fedida D (in preparation) K v l . 5 and caveolin-3 do not interact or colocalize in rat or canine cardiac myocytes. Jodene Eldstrom was responsible for project conception, planning, tissue sample preparation, immunoprecipitations, E M tissue preparation and immunolabeling (with the exception of cryosectioning) and T E M imaging, sucrose fractionation and western blots, some immunocytochemistry (canine samples), writing of the manuscript, editing and figure preparation. (Percent contribution: 90%). Qixia Y u isolated, labeled and captured the wide field, deconvolved images of the rat myocytes, and imaged the canine myocytes also. Edwin Moore analyzed the deconvolved images and prepared raw figures from them. Garnet Martens did the cryosectioning of the Tokuyasu sections. David Van Wagoner provided the canine tissue and isolated myocytes as well as attempting the cholesterol depletion studies. David Fedida assisted in writing and editing the manuscript. XIII CHAPTER 1: INTRODUCTION 1 Overview Potassium channels are plasma membrane proteins that allow potassium ions (K + ) to pass down their electrochemical gradient between the cytoplasm and extracellular region of a cell. Voltage-gated potassium (Kv) channels are one family among several types of potassium channels whose activity is modulated by changes in potential across the cell membrane, with the various family members responding with different kinetics and sensitivity to the changes in membrane potential (V m ) . These different properties allow for channels suited to the myriad roles they play throughout the different cell types of tissues and organisms in which they are found. Voltage-gated potassium (Kv) channels comprise the largest sub-group of potassium channels, with approximately thirty-eight identified a-subunit genes divided among twelve subfamilies ( K v l - K v l 2 ; Gutman et al., 2003). Figure 1.1 shows a phylogenetic tree of the K v channels that have been identified to date. Classification is based on identity at the amino acid level and within subfamilies is around 65-70% compared to the approximately 40%> identity between channels of different subfamilies (Wei et al., 1990). Identity among channels is highest in what is described as the core and the greatest divergence is found in the extracellular and intracellular loops between transmembrane (TM) domains and is especially high in the cytoplasmic N - and C-termini. a-Subunits from the Kv5-9 families are electrically silent when expressed on their own (Drewe et al., 1992) but can form functional channels when expressed with select a-subunits, such as those from the Kv2 subfamily (Patel et al., 1997; Post et al., 1996; Salinas et al., 1997) or Kv4 subfamily (Jegla and Selkoff, 1997). Some of these heterotetrameric interactions appear to play a regulatory mechanism to either silence Kv2 and Kv3 channels (Hugnot et al., 1996) or alter their kinetics and responses to changes in V m (Jegla and Selkoff, 1997; Salinas et al., 1997). 2 • Kv5.1 [KCNF1, 2p25] r- Kv2.1 [KCNB1,20q13] I Kv2.2 [KCNB2, 8q12] Kv6.1 [KCNG1,20q13] Kv6.2[KCNG2, 18q22] K v3.4 [KCNC4, 1p21] - K V 3.2 [KCNC2, 19q13] Kv6.3 [KCNG3, 2p21] — Kv8.1 [KCNB3, 8q22] Kv9.3 [KCNS3, 2p24] Kv9.2 [KCNS2, 8q22] Kv9.1 [KCNS1, 20q12] Kv3.1 [KCNC1, 11p14] •3.3 [KCNC3, 19q13] 1(- Kv3.1 * — K v: Kv4.1 [KCND1,Xp11] Kv4.2 [KCND2, 7q31] Kv4.3[KCND3,1p13] K v1.7 [KCNA7, 19q13] | K V 1.4 [KCNA4, 11p14] |— K V 1.6 [KCNA6, 12p13] K v1.5 [KCNA5,12p13] Kv1.8 [Kv1.10, 19q13] | - Kv1.2 [KCNA2, 1p13] Kv1.1 [KCNA1, 12p13] Kv1.3 [KCNA3, 1p21] Kv7.1 [KVLQT1, KCNQ1,11p15] Kv7.2 [KCNQ2, 20q13] K v7.3 [KNCQ3, 8q24] K v7.5 [KCNQ5, 6q14] K v7.4 [KCNQ4,1p34] Kv12.2 [Elk-2, KCNH3,12q13] | I Kv12.1 [Elk-1,3] ' Kv12.3 [Elk-3, KCNH14, 17q21] | Kv10.1 [eag-1, KCNH1,1q32] ' Kv10.2 [eag-2, KCNH5, 14q24] C I Kv11.1 [erg-1, KCNH2, 7q35l ' Kv11.2 [erg-2,17] Kv11.3 [erg-3, 2] Figure 1.1. Phylogenetic trees of identified K v channels. a. K v l - K v 6 and Kv8-Kv9 families; b. Kv7 family; c, K v l 0 - K v l 2 families. Families were assigned using amino acid sequence alignments of core T M regions (CLUSTAL W) and then subjected to analysis by maximum parsimony (PAUP*). Figure from Gutman et al., 2003. 3 General Structure of Kv Channels Kv channels are formed from four identical a-subunits (homotetramers) or between two or more different a-subunits usually from the same subfamily (Figure 1.2; heterotetramers; Isacoff et al., 1990; Ruppersberg et al., 1990). Permissive interactions between a-subunits are dictated by the tetramerization (TI) domain found in the N-terminus of each a-subunit (Li et al., 1992; Shen and Pfaffinger, 1995). This TI domain is highly conserved among subfamily members and folds such that several key polar residues are exposed on surfaces directly involved in intersubunit interactions, which dictate subfamily specific interactions (Kreusch et al., 1998; Bixby et al., 1999). The tetrameric TI complex has been likened to a "hanging gondola" that hangs below the transmembrane domains of the channel, aligned with the central axis of the pore (Kobertz et al, 2000; Figure 1.3). Access to the permeation pathway is gained through gaps between the "cables" that link the two domains together. A B selectivity filter/inactivation gate activation gate Figure 1.2. Kv channel structure. (A) Membrane topology of a single Kv channel a-subunit. Each subunit has six transmembrane segments, S1-S6. S4 has a series of positive charges at every third position; S2 and S3 contain three conserved negative charges that interact electrostatically with S4; S5/P-loop/S6 form the pore domain which is homologous to the bacterial KcsA channel. An N-terminal inactivation ball is shown in this example (B) Cartoon of how the rest of T M segments (voltage-sensing domains) may wrap around a central KcsA-like pore domain. Front and back subunits are removed for clarity. (From Gandhi and Isacoff, 2002) 4 Each a-subunit is a six transmembrane domain protein with each T M domain being designated SI through S6 (Figure 1-2). Each a-subunit in the tetramer contributes to the ion conduction pathway, which is made up of the S5/P-loop/S6 region of each subunit (MacKinnon and Yellen, 1990; Yellen et al., 1991; Lopez et al., 1994; Doyle et al., 1998). The P-loop contains the potassium channel signature sequence TxxTxGYGD, also known as the selectivity filter (Hille, 2001). It is this sequence of amino acids at the narrowest point in the ion conduction pathway that is responsible for the 10,000:1 selectivity of potassium ions over sodium ions (Na+) as well as selectivity over other monovalent cations (Heginbotham et al., Q 1992). This ability of the channel to differentiate between ions even as it conducts up to 10 ions per second is due to the architecture of carbonyl oxygen atoms contributed by the backbone of pore T X G Y G residues that face the conduction pathway and mimics the coordination of a potassium ion by water. Sodium ions are smaller and the distances between the carbonyl oxygens do not quite match the hydration shell around Na + , making it energetically less favourable for N a + to dehydrate and enter the selectivity filter in the open state (Doyle et al., 1998). At the cytoplasmic end of the conduction pathway below the "bundle crossing" (where the S6 T M domains meet) is the S6 gate, which prevents ions entering the pore when the channel is closed. This gate is thought to be made up of protrusions formed by a proline-induced kink in the S6 helices, (Armstrong, 1971, Durrell et al., 1992; del Camino et al., 2000; Hackos et al., 2002; Jiang et al., 2002; Labro et al., 2003; Webster et al., 2004). The process by which a channel opens is known as activation, and when the channel is activated the S6 helices move relative to each other (Holmgren et al., 1998) perhaps by twisting so that the kink can no longer obstruct the conduction pathway thereby allowing ions to move through. The rest of the subunits (S1-S4) contribute to the voltage-sensor. Changes in the membrane potential are largely sensed by positively charged arginine or lysine residues in the 5 main voltage sensor, S4 the fourth T M domain. Outward movement of this S4 segment in response to A V m is somehow coupled to opening and closing of the activation gate. Proposed mechanisms include S4 movement-induced tension in the S4-S5 linker which pulls S5 out of the path of S6 (Pathak et al., 2004) or movement of S6 is mediated by direct interaction with the linker (Lu et al., 2002). There has been much debate over the type of movement the voltage sensor makes (Gandhi et. al., 2003; Horn, 2004; Jiang et al., 2003a; Jiang et al., 2003b; Lee et al., 2003) and how this movement is coupled to movements in S6 (Pathak et al., 2004; Labro et al., 2005; Jiang et al., 2003b; Lu et al., 2002). Changes in V m occur generally by changes in ionic conditions inside the cell, usually as a result of the opening of calcium or sodium channels to allow entry of cations into the cell (or opening of potassium channels to allow efflux of K + ) . This leads to an increase in the number of positive charges inside the normally negatively charged interior of the cell. Normal resting membrane potentials for mammalian excitable cells such as neurons, muscle cells (smooth, skeletal and cardiac), and endocrine cells are between -50 mV and -95 mV (Hille, 2001). Resting membrane potentials of non-excitable cells (e.g. fibroblasts and endothelial cells) are generally less negative than those measured in excitable cells. The different Kv channel family members respond by opening at different depolarized potentials (from resting potential) and with different time courses for their activation. The potential at which half of Kvl .3 channels will be open (V1/2 activation) is fairly negative at -35 mV (Wulff et al., 2004), where as Kv3.1 channels open at quite depolarized potentials, with an activation \ m of about +14 mV (Rudy et a l , 1999). Inactivation is another individual property of each of the Kv channels and can occur by two main mechanisms: fast N-type and slow C-type inactivation. N-type inactivation results from occlusion of the conduction pathway at the internal mouth of the pore by a globular amino terminal domain found in some Kv a-subunits (Hoshi et al., 1990) and in some accessory P-subunits (Rettig et al., 1994; Heinemann et al., 1996; Leicher et al., 1998). Based on mutagenesis studies, the ball receptor site is made up at least in part by the S4-S5 cytoplasmic linker domain (Isacoff et al., 1991). This type of inactivation occurs within the time frame of tens of milliseconds. C-type inactivation occurs through a conformation change or collapse of the selectivity filter. There is some evidence that two separable processes of inactivation are taking place; one involves rearrangement at the selectivity filter or domains closer to the external mouth of the pore that prevent ion permeation (P-type inactivation), and a second rearrangement stabilizes this closed conformation and is somehow coupled to the position of the S4 segment (C-type; Olcese et al., 1997; Yang et al., 1997; Loots and Isacoff, 1998). The rate of this type of inactivation can be increased by prior N-type inactivation of the channel (Hoshi et al., 1991) and inhibited by conditions that favour occupancy of one of the ion binding sites in the conduction pathway, such as high external potassium (Baukrowitz and Yellen, 1996). As a result, although C-type inactivation is often a slow process with a timescale of hundreds of milliseconds, it can be much faster in channels which also possess N-type inactivation, or be accelerated when the pore is unoccupied. Channels, once inactivated, must undergo a recovery process before they are able to undergo activation again. This process involves return of the voltage sensor to its resting state position and accompanying conformational changes in the gating mechanism and residues involved in C-type inactivation (not necessarily in this order). Channels that enter the C-type inactivated state take longer to recover than those that enter N-type alone, and are thus unavailable for re-activation during rapid repetitive stimuli. This is another property that makes some channels better suited for certain physiological roles over others. Role of K v Channels In excitable cells, Kv channels are responsible for terminating action potentials, setting the resting membrane potential, and determining the length and frequency of bursts of action 7 potentials (APs; Hille, 2001). In addition, Kv channels regulate the secretion of hormones from endocrine cells (Chen et al., 1990; MacDonald et al., 2001), and cause relaxation of smooth muscle cells by hyperpolarization of the resting V m (Tammaro et al., 2004). K v channels are also found in non-excitable cells and play cell-dependent roles in regulating many cellular processes such as cell development and proliferation (Chittajallu et al., 2002; Kotecha and Schlichter, 1999; reviewed in Pardo, 2004), migration (Rao et al., 2002), volume regulation (Deutsch and Chen, 1993; Felipe et al., 1993; Grinstein and Smith, 1990), and maintenance of membrane potential (Leonard et al., 1992). Kv channels are also important in the immune system, with their involvement in T-cell activation (Koo et al., 1997). Biophysical properties of K v l . 5 Kvl .5 homotetrameric channels show rapid, (time constant of ~12 ms at 0 mV) voltage-dependent activation at potentials positive to -30 mV. The slope factor of the Boltzmann fit of the activation curve is 6 mV, with a half-activation potential between -11 and -4 mV, when the channel is expressed in heterologous cells (Kurata et al., 2001; Fedida et al., 1993; Snyders et al., 1993; Uebele et al., 1996). The most defining biophysical characteristic of Kvl .5 channels is very slow inactivation during long voltage clamps at positive potentials. During 5 second depolarizing pulses to +50 mV, Kvl .5 whole-cell currents decrease only ~38 % when expressed in HEK293 cells (Uebele et al., 1996). When rapidly cycled between RMP and depolarized potentials, Kv l .5 shows little cumulative inactivation, unlike channels such as Kv2.1, that undergo significant inactivation from late closed states (Kurata et al., 2001). Kv l .5 is very sensitive to micromolar concentrations of 4-amino-pyridine (4-AP), quinidine and clofilium, and shows little sensitivity to T E A + and dendrotoxin (Chen and Fedida, 1998; Fedida et al., 1993; Snyders et al., 1993). These unique kinetic properties of homomeric Kvl .5 channels make them well suited as carriers of repolarizing current during repetitive depolarizations, as these channels 8 are rapidly activated and their availability remains high due to low rates of inactivation even at high rates of activity. Role of K v l . 5 in the heart Cardiac myocytes exhibit several different outward selective potassium currents including I t o i , I t 02, IKT, IK s , and IKUR, each with unique activation and inactivation kinetics, voltage dependency, pharmacology and distribution (reviewed in Nerbonne, 2000; Snyders, 1999). These currents are involved in the shaping of the cardiac action potential and in regulating its duration. The characterization of each of these currents has been the focus of a great deal of research, comparing regional, species, pathophysiological and underlying molecular correlates. Though great progress has been made in identifying many of the channels underlying these currents there is still some uncertainty due to conflicting results between laboratories and also due to built-in redundancy. Frequently, differences are attributed to species variation but there are also the issues of heterotetramerization, heteromultimerization with accessory subunits, upstream regulation of channels, and effects of myocyte isolation and culturing that still need to be worked out fully. Kv l .5 is widely recognized as the molecular correlate underlying the ultra-rapid delayed rectifier current (IKUR) in human atrial myocytes (Feng et al., 1997; Fedida et al., 1993). This current shows rapid voltage-dependent activation at positive potentials with little inactivation during 1000 ms depolarizations (Wang et al., 1993). IKUR is extremely sensitive to micromolar concentrations of 4-AP with an IC50 of 49 pM, but is relatively insensitive to T E A + , dendrotoxin, and extracellular Ba (Wang et al., 1993). These characteristics matched closely those of a channel cloned from human atrial myocytes and expressed in heterologous cells (Fedida et al., 1993; Syders et al., 1993; Tamkun et al., 1991). The pharmacological sensitivity of IKUR in human atrial myocytes essentially, ruled out involvement of other delayed rectifier-like K + channel subunits such as K v l . l and Kv l .2 (Wang et al., 1993), but it was anti-sense experiments 9 against Kvl .5 that finally firmly established the role of this channel in atrial repolarization in humans (Feng et al., 1997), rats (Bou-Abboud and Nerbonne, 1999) and in mice (Bou-Abboud et al., 2000). Despite the presence of Kvl .5 protein in ventricular samples (Fedida et al., 2003; Barry et al., 1995; Mays et al., 1995; Tamkun et al., 1991), its role and contribution to identified currents such as IKSUS have yet to be determined (Li et al., 1996; Fiset et al., 1997). In adult mouse ventricular myocytes, a sustained current that was sensitive to micromolar levels of 4-AP (IC5o of 14.5 uM) was identified (Fiset et al., 1997) and mRNA expression in mouse heart showed that of the cloned channels highly sensitive to 4-AP ( K v l . l , K v l . 5 , Kv3.1), only Kvl .5 mRNA was present at high levels (Grissmer et al., 1997). In another study in transgenic mice, expression of a dominant-negative mutant of K v l . l was associated with a longer QT interval, action potential prolongation and ventricular tachycardia (London et al., 1998). Ventricular myocytes from these animals showed decreased levels of Kvl .5 protein, supporting the argument that IKUR in mouse is at least in part made up by K v l .5, suggesting also that it may form heteromultimers with K v l . l in this tissue and present with different kinetic and pharmacological properties. However, given that this K v l . l mutant was over-expressed in these studies, which could result in promiscuous association with other K v l channels, further studies are needed to establish whether K v l . l and Kvl .5 heteromultimerize in mouse ventricle. A later study using a cardiac specific promoter to selectively express a non-conducting Kvl .5 resulted in the loss of the 4-AP sensitive component of IKSIOWI but had no effect on the sustained component, IKS  (Li et al., 2004). From these studies and others (Brouillette et al., 2005: Brouillette et al., 2003), there appears to be significant evidence for a role for Kvl .5 in mouse ventricle repolarization. In the rat ventricle, blockade of IKUR with S9947 displays an IC50 of ~1 uM which is similar to that determined for blockade of Kvl .5 channels in CHO cells and oocytes (0.417 uM and 0.646 uM respectively; Bachmann et al., 2001), leading those investigators to conclude that Kv l .5 plays a significant role in rat ventricular IKUR- However, others have determined that 10 while some rat neonatal ventricular myocytes show evidence of IKUR, this is absent in the adult rat myocytes (Guo et al., 1997). Blockade of IKUR with 4-AP (Wang et al., 1993) or the more specific Kvl .5 channel blocker C9356 (Fedida et al., 2003) increases the action potential duration in human and canine atrial myocytes, and the channel has therefore drawn a great deal of attention as a therapeutic target in the treatment of atrial fibrillation (AF; Van Wagoner et al., 2000). In isolated myocytes from patients with persistent AF it has been noted that the sustained current at the end of long depolarizations, carried by Kvl .5 channels, is reduced by approximately 50% and this is a result of a reduction in channel protein (Van Wagoner et al., 1997). Similar electrophysiological changes have been noted in goat AF models (van der Velden et al., 2000). As there is a parallel reduction in the L-type C a 2 + current (Van Wagoner et al., 1999), use of Kvl .5 channel blockers may very well be useful in prolonging the refractoriness of atrial myocytes and therefore reduce the rate of atrial contractions. AF is the most common sustained arrhythmia in humans, and occurs frequently as a post-operative complication of cardiac surgery (Brundel et al., 2002; Hogue and Hyder, 2000). While the arrhythmia itself is tolerated, patients are at higher risk for embolic events, which makes prevention of AF an important aspect of post-operative care. Modulators of IKUR /Kvl.5 Several studies have examined how various hormones and cellular factors affect Kvl .5 as well as IKUR in the heart. Understanding how the channel density and channel activity are regulated may lend some insight into potential binding partners of K v l . 5 , help determine whether the channel exists as a macromolecular complex, explain changes in pathophysiological states and in addition, upstream effectors of Kvl .5 may well also provide clues as to patterns of channel localization. 11 Adrenergic Modulation Catecholamines acting through adrenergic receptors in the heart influence the contractile state primarily through effects on intracellular calcium concentrations, influx, release and storage, as well as calcium sensitivity of the myofilaments and additional effects on a number of cardiac potassium currents that regulate AP repolarization and thus calcium influx (reviewed in Fedida et al., 1993; Kamp and Hell, 2000). Several studies have indicated that I K UR is subject to modulation by (3 and/or a-adrenergic stimulation in both human (Li et al., 1996), canine (Yue et al., 1999) and rat (Van Wagoner et al, 1996) atrial myocytes, although there is species variability in response to adrenergic agonists. Treatment of dissociated atrial myocytes with the p i agonist, isoproterenol, increased IKUR by 37% in human and -40 % in canine. This effect could be mimicked by affecting downstream components of the (3-adrenergic pathway, such as activating adenylate cyclase to increase intracellular cAMP or by blocking the downstream effects by using a P K A inhibitory peptide. In contrast, phenylephrine, an a-1 agonist, decreased IKUR in human 26%, increased IKUR in canine, and decreased IKss 28 % in rat. The isoproterenol- and phenylephrine-enhanced current in canine atrial myocytes in this particular study was sensitive to 5 m M T E A + (Yue et al., 1999), a concentration that would have minor effects on Kvl .5 channels (Snyders et a l , 1993; Fedida et al., 2003) and appears to represents a variable component of canine IKUR (Fedida et al., 2003). The decrease in human and increase in canine atrial IKUR could be blocked by PKC inhibition consistent with roles for PLC and PKC downstream of a-adrenergic activation (Fedida et al., 1993; Kamp and Hell, 2000). It is noteworthy that a 50% reduction in IKUR in human atrial myocytes achieved by treatment with 4-AP, results in a 60% increase in the action potential duration (APD; Wang et al., 1993), indicating that only fractional block of IKUR is required for significant effects on APD. 12 In smooth muscle cells from rabbit portal vein, a 4-AP sensitive delayed rectifier current is also regulated by P-adrenergic stimulation with isoproterenol and by P K A activity (Aiello et al., 1998). The a-subunits in this study were not directly identified but the authors suggested a role for Kv l .5 in this response. Later work from this group established that this current is carried by Kv l .5 /Kv l .2 heteromultimers (Kerr et al 2001). The sensitivity of Kv l .2 homotetramers to P-adrenergic modulation has also been demonstrated in oocytes (Huang et al., 1994) where P K A co-expression was shown to increase Kv l .2 currents. In contrast, studies of Kvl .5 channels expressed in oocytes have shown that treatments known to increase intracellular cAMP do not affect current amplitudes, but inhibition of P K A leads to a time dependent reduction in currents (Mason et al., 2002). The authors of that study suggested that in oocytes, Kv l .5 was already maximally activated by P K A under basal conditions. P K A modulation of Kvl .5 in oocytes did not require phosphorylation at either of two consensus P K A phosphorylation sites in the C-terminus (S538 and S545) nor did it require an intact PDZ binding domain. Disruption of cytoskeleton with cytochalasins or down regulation of actinin-2 with anti-sense oligonucleotides did however prevent decreases in Kvl .5 activity in response to P K A inhibitors. The non-specific protein tyrosine kinase inhibitor, genistein, also decreased Kvl .5 current amplitudes (Mason et al., 2002). In a study of adrenalectomized and reserpine-treated rats (and therefore catecholamine-depleted), Kv l .5 mRNA levels increased but no effect on current density or protein levels were detected (Bru-Mercier et al., 2003). Generally catecholamine-mediated increases in intracellular cAMP and P K A activation can lead to modulation of transcription of genes containing a cyclic A M P responsive element (CRE) in their promoter regions. The K v l .5 promoter has a CRE site and increasing cAMP levels results in an increase in Kvl .5 mRNA in primary neonatal atrial myocytes but a decrease in adult rat atrial myocytes and rat clonal pituitary (GH3) cells (Mori et al., 1993). The developmental and tissue specific differences in responses are likely due to tissue 13 dependent differences in expression of CRE binding proteins (CREBs and CREMs) and a protein that is found in cell types that express Kvl .5 endogenously and binds to a Kvl .5 Repressor Element (KRE). This protein, K R E binding factor (KBF) appears to bind the cis-acting silencer (KRE) in the Kvl .5 promoter and prevent a tertiary structure from forming which inhibits transcriptional activation (Mori et al., 1993; Valverde and Koren, 1999). Responsiveness to cAMP changes also depends on CREB and C R E M expression. Depending on the isoforms expressed, these proteins upon PKA-dependent phosphorylation can act as repressors or activators of transcription and the ratio between the activating and repressor forms dictates the cellular response to changes in cAMP levels (de Groot and Sassone-Corsi, 1993). A study that looked for circadian variation in rat cardiac K + channels found that Kvl .5 and IKUR were up-regulated during 12 hour dark periods and down-regulated during light periods (Yamashita et al., 2003), while Kv4.2 had the opposite pattern. The changes in Kvl .5 mRNA and protein were attenuated by propranolol added to the drinking water, indicating that at least some of the variation may be due to changes in sympathetic activation. In humans this would mean withdrawal of sympathetic tone, but rats, being nocturnal, may respond differently. Cardiac ion channel modulation by adrenergic activity is of additional interest given the suggestion that the incidence of several arrhythmias may be higher during changes of sympathetic tone such as transitions from non-REM to R E M sleep and from sleep to wakefulness (Gula et al., 2004). Androgens Differences in susceptibility to serious cardiac arrhythmia, such as torsades de pointes (TdP), have been noted between the sexes and has been related to differences in QT interval, which is longer in women. As the difference in QT interval is not significantly affected by the menstrual cycle, menopause or hormone replacement therapy, it is thought that testosterone plays a greater role in gender-based variation in ventricular repolarization (reviewed in Abi-Gerges et 14 al., 2004). While the main mechanism of drug-induced TdP appears to be as a result of hERG channel block, the molecular correlate of IKr in human ventricle, there is some evidence that Kv l .5 transcription is subject to hormonal regulation. The ventricular APD is shorter in male compared to female mice and generally these myocytes have a larger repolarizing K + current (Wu and Anderson, 2002; Trepanier-Boulay et al., 2001). In one study, the shorter APD in males was correlated with a higher expression of Kv l .5 protein and increased IKUR current density in ventricular myocytes (Brouillette et al., 2005). In orchidectomized mice, cardiac repolarization is prolonged (as shown by prolonged rate corrected Q-Tc interval and APD) and levels of Kv l .5 protein and IKUR density are decreased (Brouillette et al., 2003). These effects of androgens on channel expression obviously represent genomic effects of hormone activated androgen receptors on gene expression but little is known regarding the regulation of Kvl .5 expression at the level of transcription. Thyroid Hormone The effects of changes in plasma Thyroid Hormone (TH) levels on Kvl .5 expression are somewhat difficult to discern from the literature. In most cases, only the effects on mRNA levels were studied, not changes in protein or currents carried by the channel. Either K v l .5 gene expression was found to increase in rat cardiac tissue from both chambers in response to exogenous triiodothyronine (T3) treatment, used to simulate the hyperthyroid state, (Abe et al 1998; Watanabe et al., 2003; Ma et al., 2003) or there was no change (Ojamaa et al., 1999). Hyperthyroidism is associated with an increase in heart rate and a decrease in APD (Klein and Ojamaa, 2001; Watanabe et al., 2003; Ma et al., 2003) which would be consistent with an increase in repolarizing potassium current or a decrease in inward C a 2 + or N a + current. An increase in I K UR and a decrease in Ic a were noted in one study (Watanabe et al., 2003); however, there have been reports of an increase in Ic a in patients with latent hyperthyroidism (Kreuzberg et al., 2000) perhaps reflecting differences in species, degree of TH elevation or duration of disease 15 state. AF is a complicating factor of hyperthyroidism in 5-15% of patients, and is likely a result of the decrease in atrial effective refractory period associated with hyperthyroidism, which promotes intra-atrial re-entry (Klein and Ojamaa, 2001). As changes in Ic a are also noted in AF , with initiation associated with calcium overload followed by a decrease in Ic a (Van Wagoner et al., 1999), the increase in Ic a in latent hyperthyroidism may be indicative of early stages of electrical remodeling. Hypothyroidism is associated with a reduced heart rate, prolongation of APD and delayed ventricular repolarization in guinea pig (Bosch et al., 1999) and humans (Klein and Ojamaa, 2001). When hypothyroidism was induced in rats, Kv l .5 mRNA was reduced in ventricle but no change was noted in atrial myocytes (Ojamaa et al., 1999; Ma et al., 2003), with some exceptions (Abe et al., 1998). The unresponsiveness of atrial Kv l .5 mirrors the pattern of T3 modulation of myosin heavy chain (MHC) genes in the adult rat, where the ventricle is responsive to changes in T3 and up regulates MHCcc transcription, whereas the atria are not responsive (Brent, 1994). This may reflect differences in thyroid hormone receptor levels in the different cell types (Banerjee et al., 1988) or subtypes in the two. Hypoxia A role for Kvl .5 in vascular response to hypoxia has also been established, specifically hypoxic pulmonary vasoconstriction (HPV) an important mechanism for ventilation-perfusion matching in the lungs. In pulmonary artery smooth muscle cells (PASMCs), hypoxia causes vasoconstriction due to an inhibition of a 4-AP-sensitive outward potassium current (Archer et al., 1998) and increases in cytoplasmic calcium (Gelbrand and Gelbrand, 1997). Application of anti-Kvl.5 antibody inhibited the hypoxia-induced reduction in IK in denuded rat pulmonary artery rings, in isolated lungs (Archer et al., 1998) and isolated myocytes (Gelbrand and Gelbrand, 1997). Targeted replacement of Kvl .5 by K v l . l (which failed to express K v l . l protein in the lungs) in transgenic mice, resulted in a reduction in 4-AP sensitive currents and 16 impaired HPV in homozygous mutant mice (Archer et al., 2001). Kv l .5 homomeric channels expressed in mouse L-cells however, failed to show sensitivity to hypoxia unless co-expressed with K v l . 2 (Hulme et al., 1999). Heteromultimerization of the two ct-subunits created a channel that activated in the physiological range of membrane potential for PASMCs, and showed significant decreases in current density in response to hypoxia. It should be noted also that heteromeric channels made up of Kv2.1 and Kv9.3 also were sensitive to hypoxia at relevant membrane potentials and both have been detected in PASMCs (Patel et al., 1997). Kv l .5 is also found in systemic vascular beds and despite exhibiting an opposite response to hypoxia (vasodilatation) both Kvl .5 and Kv2.1 have been implicated in the calcium-dependent hypoxic response of these vessels (reviewed in Thorne et al., 2004). pH A few studies have examined the effects of extracellular acidification on Kvl .5 and have shown that rKvl .5 (Steidl and Yool, 1999) and hKvl.5 (Kehl et al., 2002) current amplitudes decrease with an increase in extracellular protons. Decreases in pH also result in a slight depolarizing shift in the half activation voltage indicating that protonation of the channel at some unidentified site(s) may impede or screen the voltage sensor from changes in membrane potential. The proton-mediated decrease in current amplitude was attributed to an enhancement of P-type inactivation an affect mediated in part by a protonatable histidine residue in the outer rim of the channel pore. Studies in cardiac myocytes have shown that acidosis can either prolong or shorten the APD (reviewed in Orchard and Kentish, 1990) with the changes being attributed at least in part to changes in Ic a and outward potassium currents. In rat ventricular myocytes acidosis prolonged the APD and while the current affected was a non-inactivating delayed rectifier, it is thought that Kv l .2 or Kv2.1 carries this current in this tissue (Komukai et al., 2002a). Kv l . 2 , however, appears to be insensitive to pH, which leaves either Kv2.1 or the possibility of Kv l .5 /Kv l .2 heteromultimers. Protonation of Histidine 463 on 1-2 subunits of the 17 tetrameric channel is required for significant acid affects on hKvl.5 (Kehl et al., 2002) indicating that this effect would predominate in heterotetramers. Acidosis shortens the rat atrial APD, which is inconsistent with observed affects on K v l . 5 , but again the role of Kvl .5 in this tissue is uncertain (Komukai et al., 2002b). Growth Factors Kvl .5 may also be regulated by activation of several growth factor receptors. In oocytes co-expressing Kvl .5 and several different receptors that are known to activate phospholipase Cy (PLCy), including platelet-derived growth factor (PDGF), fibroblast growth factor (FGF), 5-hydroxytryptamine (5-HT; serotonin), and thrombin receptors, treatment with ligand caused a slow decay in Kvl .5 current amplitude, resulting in a 70% decrease by 30 minutes (Timpe and Fantle, 1994). Treatment with IP3 and D A G , downstream products of PLC activation resulted in a similar decline with a shorter lag time consistent with a requirement for receptor-mediated PLC activation in this decline of Kvl .5 current density. No effect of activation of the endogenous insulin-like growth factor receptor (IGFR) was observed in oocytes, a receptor that does not couple to PLCy. In contrast to the lack of acute effects of IGF activation on Kvl .5 current amplitudes, 72-hour treatment resulted in an upregulation of Kvl .5 protein and associated currents in cultured rat neonatal ventricular myocytes as well as an increase in myocyte size (Guo et al., 1997), consistent with an hypertrophic effect of this factor. Both the hypertrophic effect and effects on channel density were blocked by the protein tyrosine kinase inhibitor genistein but not by the PKC inhibitor, staurosporine (Guo et al., 1998). Similarly, longer treatment of oligodendroglial progenitor cells with PDGF and FGF also increased Kvl .5 protein and IK. In these cells, this effect could be blocked by a Src kinase inhibitor, a sphingosine kinase inhibitor and the C a 2 + chelator, BAPTA, but not PI3K, E R K pathway, PLCy nor PKC inhibition (Soliven et al., 2003). 18 In cultured rat neonatal skeletal muscle, Kv l .5 protein is undetectable upon dissociation, appears at day two in culture and increases dramatically at time of myoblast fusion on day four (Vigdor et al., 1999). Treatment of cultured myotubes with nerve growth factor (NGF) causes Kvl .5 to appear sooner in the cultured cells with expression increasing until no further increase is observed at 72 hours, calcium-dependent K + currents were decreased on NGF treatment in this same preparation. The increase in channel protein was correlated with an increase in the rate of repolarization of spontaneously occurring APs. Roles for interacting proteins Several protein-protein interactions are possible during the life cycle of a membrane glycoprotein. These interactions may begin with catalysts for proper folding of the nascent polypeptide. Kv channel ct-subunits acquire their secondary structure as they are synthesized (Kosolapov et al., 2004) and their tertiary structure while the nascent peptide is still bound to the translational machinery, (Tu et al., 2000). T l domains of individual subunits begin to associate with each other at this point, even before synthesis of SI (Lu et al., 2001), a mechanism thought to increase the likelihood of homomultimerization (Kosolapov and Deutsch, 2003). Signal sequences involved in directing insertion of the protein into the ER membrane are found mainly in the S2 segment, which is unlike most membrane proteins, that generally contain a cleavable N-terminal signal sequence upstream of any T M segment (Tu et al., 2000). This could potentially mean that the entire hydrophobic S1 domain can be found in the cytoplasm near other SI domains during synthesis and any perturbation of the translational efficiency or translocation could lead to cytoplasmic aggregation of the channel protein via hydrophobic interactions. For other proteins, this potential problem is frequently avoided by efficient biogenesis. In K v channels this may include, perhaps, coupling tertiary folding and tetramerization of the T l domains (Robinson and Deutsch, 2005) and/or by the use of chaperone proteins that detect the hydrophobic domains and mask them before aggregation can take place (reviewed in Fink, 19 1999). Chaperones (e.g. Hsp70 and Hsp40) can also detect improperly folded segments (generally by detecting hydrophobic domains) in the ER and either tag them for degradation through ubiquitination and ER associated degradation (ERAD) or assist in their refolding. Resident ER proteins, Hsc70 and CHIP, are involved in detecting misfolded CFTR and promoting its ubiquitination. This results in translocation of the channel polypeptide out of the ER and degradation via the 26S proteasome (Meacham et al., 2001). Current research into treatment of cystic fibrosis is aimed at finding a chemical chaperone that can assist in the refolding of CFTR A F508 (Cohen and Kelly, 2003). This truncation mutation is the most common cause of the disease (Zielenski, 2000) and studies have shown it can form functional channels i f folding intermediates are stabilized so that it can exit the ER instead of being quickly degraded (Sato et al., 1996). The amino acid sequence of a protein contains intrinsic signals that regulate intracellular transport such as ER retention or surface membrane expression. Many examples of transport signals have been found in various potassium channels including the diacidic (E/D)X(E/D) ER export motif found in the inwardly rectifying potassium channel Kir 1.1, the V X X S L sequence implicated in efficient trafficking of the voltage-gated potassium channel Kv l .4 to the cell surface, and the arginine-based ER retention signal (RXR) found in the Kir6.2/Surl KATP channel (reviewed in Griffith, 2001). Proteins are exported or stabilized at the membrane by several mechanisms including masking of retention or uptake signals through 1) proper folding and multimerization (KATP channel; Zerangue et al., 1999); 2) binding of auxiliary subunits (Kvl.2 + p-subunit; Shi et al., 1996), or 3) binding of anchoring proteins (Kir4.1/PSD-95/SAP90; Horio et al., 1997). As potassium channels are only functional as tetramers one possible mechanism for preventing the expression of monomeric channel proteins at the surface has been proposed for the R X R containing Kir6.2. Somehow, resident ER and Golgi proteins "sense" the properly folded and assembled state of the channel and allow it to move forward (be 20 recruited) to vesicles headed for the Golgi. Monomeric Kir6.2 protein will have an exposed R X R motif in its C-terminus that can be bound by the coat protein complex I (COPI) and this mechanism may mediate retrieval of monomers from the ER-Golgi intermediate compartment or retention in the ER (Yuan et al., 2003). 14-3-3 proteins can also bind R X R motifs and may compete with COPI for binding, to allow forward trafficking. As mentioned previously, protein-protein interactions may be required to tag channels for sorting into secretory vesicles bound for the surface membrane. Binding of proteins such as the Kvl .3 and Kv4.2 channel associated protein, KChAP, improves surface expression of the channel without affecting its properties, but KChAP does not stay bound once the complex arrives at the cell surface (Wible et al., 1998). This may mean that KChAP either masks retention signals, contains forward trafficking signals, or improves the stability of the channels in the ER so that assembly is greater than E R A D or serves all of these functions. Once at the surface, channels require anchorage to ensure that they stay within functional domains of the plasma membrane. This is usually achieved by binding cytoskeletal proteins or via intermediary scaffold proteins. Intercalated disc (ID) localization of Nav1.5 has been linked to its ability to bind the cytoskeletal protein ankyrin (Mohler et al., 2004) and the phosphorylation state of its [31 -subunit (Malhotra et al., 2004). It has been suggested that this localization plays a role in initiation and conduction of the cardiac action potential (Kucera et al., 2002; Maier et al., 2002). Failure of Nav1.5 to bind ankyrin-G has been linked to Brugada syndrome, a disease which is characterized by incomplete right bundle branch block and increased risk of sudden cardiac death as a result of ventricular fibrillation (Mohler et al., 2004). Similar to Navl.5, Kv l .5 is also highly enriched at intercalated discs of cardiac myocytes but the mechanism behind this localization is undetermined. Proteins, such as kinases and phosphatases, which regulate channel activity, may also be linked via scaffold proteins to channel complexes. This linking of regulatory enzymes ensures 21 that when upstream signals lead to enzyme activation they are in close proximity to their target for efficient transduction. This has been demonstrated for the cardiac ryanodine receptor, the L -type Ca channel and Kv7.1 (reviewed in Marx, 2003). For each of these channels, interactions with adaptor proteins from the A kinase associated protein (AKAP) family, via leucine-isoleucine zipper (LIZ) domains found in both the channel and the adaptors, link kinases and phosphatases to the channel complex. Mutation of the LIZ domain in Kv7.1, the a-subunit that underlies IKs in the heart, uncouples the channel from sympathetic regulation due to a loss of associated PICA and PP1 (Marx et al., 2002). Submission of the Kvl .5 sequence to 2ZIP, a program that predicts LIZ domains (http://2zip.molgen.mpg.de/index.html; Bornberg et al., 1998), however, fails to detect a LIZ domain in the channel indicating that if the channel exists as a macromolecular complex it must, therefore, utilize alternate protein-protein interactions modules. Preliminary dissection of the role of a candidate interaction module in the C-terminus of the channel, a PDZ binding domain is the focus of some of the work carried out in this thesis and is presented in later chapters. Finally, membrane proteins undergo internalization (endocytosis) and are either sorted to lysosomes for degradation or are recycled for reinsertion into the surface membrane. Protein-protein interactions likely underlie the selection process for internalization. This may occur via an initial site-specific phosphorylation of the channel (cw-acting signal) or an accessory subunit (trans-acting signal) which targets the channel for polyubiquitination. Polyubiquitination acts as a signal that causes recruitment to endosomes of many different T M proteins or acts as a post-internalization sorting signal that targets endocytosed proteins for degradation rather than for recycling (Katzmann et al., 2002). Plasma membrane proteins can be internalized via clathrin-mediated or clathrin-independent mechanisms, both appearing to result in delivery to a common endosomal system (Naslavsky et al., 2003) except perhaps proteins internalized as part of detergent-resistant microdomains which may traffic through a separate endosomal system 22 (Nichols, 2002). Initial studies of the predominant cardiac sodium channel, Navl.5, and a cardiac expressed E3 ligase, Nedd4-2, indicate that this enzyme can increase the proportion of Navl.5 that is ubiquitinated and decreases IN 3 in co-transfected HEK293 cells (Bemmelen et al., 2004). Whether this occurs in cardiac tissue remains to be investigated. Nedd4-2 interacts via WW domains with target proteins containing a xPPxY motif (Rotin et al., 2000) which is not found in hKvl.5 (appendix) and no studies to date have shown Kvl .5 to be the target of ubiquitinating ligases. The proline-rich domain in the N-terminus of hKvl.5 has two group II WW binding domains (consensus sequence PPLP) which overlap the SH3 domains, and may potentially mediate interactions with as yet unidentified ligases. Several examples of membrane proteins that are ubiquitinated that are without a PY domain exist, and it is possible, as has been found for the insulin-like growth factor receptor, that an adaptor protein could mediate ubiquitinization of the channel (reviewed in Rotin et al., 2000). In addition, several proteins are recruited for endocytosis by the adaptor protein, AP2. This interaction is mediated by di-leucine or tyrosine based motifs in membrane protein cytoplasmic domains (reviewed in Sorkin et al., 2004). The mechanism for selection of Kvl .5 for endocytosis has not yet been determined. Known channel interactions To date, many of the protein-protein interactions with Kvl .5 have been localized to the N-terminus. This includes channel oligomerization to form a tetramer, association of P-subunits or KChAP (Wible et al., 1998) and interactions with the cytoskeletal protein a-actinin-2 (Maruoka et al., 2000; Cukovic et al., 2001; Figure 1-2) and PSD-95 (Eldstrom et al., 2002). In addition, there have been reports that SAP97 (Folco et al., 2004; Godreau et al., 2003; Murata et al., 2001) and Caveolin-3 (Folco et al., 2004) also bind to Kvl .5 in transiently transfected cells. Submission of the human Kvl .5 protein sequence (Accession number P22460) to a motif scanner ("http://scansite.mit.edu; Obenauer et al., 2003) for a high stringency scan, results in several 23 predictions for protein interactions (Table I). This program uses information from binding assays, peptide screens and phage experiments in conjunction with biochemical data to derive algorithms that predict protein-protein interactions involving the eighteen most common interaction modules such as SH3, SH2, WW, PTB, PDZ and 14-3-3 domains (Obenauer et al., 2003). The results include several SH3 protein-protein binding domains in a proline rich-region of the N-terminus of the channel. To date the only protein that has been shown to bind to this proline rich domain are the Src family of tyrosine kinases (Nitabach et al., 2001). While the program detects the consensus sequence for binding by the catalytic subunit of PI3K, p85, at least one study has concluded that this protein does not interact with Kvl .5 (Holmes et al., 1996). Highlighting some of the limitations of this sort of analysis, one of the predicted binding sites is found in an extracellular domain of the protein (shown in grey in table) and at high stringency the C-terminal PDZ binding domain that mediates PSD95 binding (Eldstrom et al., 2002) is not Table I: Results of a High Stringency Motif Scan of Human Kvl.5 (Accession number: P22460; http://scansite.mil ..edu; Obenauer et al., 2003) Motif Site Score Percentile Sequence SA SH3 1 P68 0.1777 0.001 % DSGVRPLPPLPDPGV 0.891 Src SH3 P79 0.2373 0.004 % DPGVRPLPPLPEELP 0.891 Grb2 SH3 P68 0.3502 0.008 % DSGVRPLPPLPDPGV 0.891 Grb2 SH3 P79 0.3755 0.017 % DPGVRPLPPLPEELP 0.891 p85 SH3 model/PIK3Rl P79 0.4387 0.041 % DPGVRPLPPLPEELP 0.891 Amphiphysin SH3 P88 0.4447 0.056 % LPEELPRPRRPPPED 6.366 Cortactin SH3 P79 0.4850 0.062 % DPGVRPLPPLPEELP 0.891 p85 SH3 model P68 0.4960 0.138% DSGVRPLPPLPDPGV 0.891 PLCg SI 13 P314 0.5104 0.150% PSGPTVAPLLPRTLA 0.279 PKCA (Baso ST kin) 2 S562 0.3862 0.119% K V S G S R G S F C K A G G T 0.553 PDK1 (Kinbind) 3 E241 0.3282 0.052 % RQVWLIFEYPESSGS 1.011 Erk D-domain/ M A P K 1 V191 0.4869 0.034 % R L R R P V N V S L D V F A D 0.389 Erk D-domain/ M A P K 1 L193 0.3411 0.001 % R R P V N V S L D V F A D E I 0.202 She P T B 4 Y521 0.5060 0.188% V I V S N F N Y F Y H R E T D 0.959 1) Src homology 3 group (SH3) 2) Basophilic serine/threonine kinase group (Baso_ST_kin) 3) Kinase binding site group (Kin bind) 4) Phosphotyrosine binding group (PTB) 24 detected. Of course these are only predictions and proper verification using biochemical and functional analysis will be required to identify true and physiologically relevant binding partners o f K v l . 5 . Kvp Subunits Three families of KvP subunits have been cloned so far from mammalian cells, KvP 1-3 (Chouinard et al., 1995; Rettig et a l , 1994; Scott et al., 1994; England et al., 1995; Majumder et al., 1995; Morales et al., 1995) and a fourth has been isolated from embryonic Xenopus nervous system (Lazaroff et al., 1999). Additional K v p l subunits are encoded by mRNA splice variants (England et al., 1995; McCormack et al., 1995). These accessory subunits form and associate with the pore forming a-subunits in the ER (Shi et al., 1996; Nagaya and Papazian, 1997) in a 1:1 stoichiometry (Parcej et al., 1992; Gulbis et al., 1999), through an interaction with the cytoplasmic T l domain (Figure 1-3; Sewing et al., 1996; Gulbis et al., 2000). Not all K v Figure 1.3. Cartoon of a voltage-dependent K channel (red) with associated P-subunit (blue). The pore region is based on the KcsA K + channel crystal structure (Doyle et al., 1998). The structures of the voltage-sensing and linking regions of mammalian K v channel are unclear and are therefore depicted schematically. An NH2-terminal inactivation peptide is shown entering a lateral opening to gain access to the pore. (Figure taken from Gulbis et al., 2000) 25 channels can associate with the different (3-subunits; in fact it appears that only the K v l family can associate with the Kv pis (Sewing et al., 1996; Wang et al., 1996; Y u et al., 1996). K v p i and Kvp3 family members can confer fast N-type inactivation on some Kv channels and enhance N-type inactivation in channels that already inactivate by this mechanism (Majumder et al., 1995; Rettig et al., 1994). In addition, there is some evidence that KvP subunits may enhance surface expression of some K v channels in cultured cells (Shi et al., 1996; Campomanes et al., 2002). However, this may depend on the expression system or on the efficiency of the a-subunit processing in the absence of the P-subunit (Nagaya and Papazian, 1997), as a chaperone effect was not evident in Kvp2 null mice (McCormack et al., 2002). Despite phenotypic evidence of hyperexcitability, shortened life span and cold-swim tremors, these Kvp2 knockout mice showed normal levels of mature K v l . l and K1.2 channel subunits in the brain and normal localization in the cell types examined. Kvl .5 has been shown to interact with K v p i . l , Kvpi .2 , Kvpi .3 and KvP2.1 and as a result, with the exception of K v p i . l (Sewing et al., 1996), opens at more hyperpolarized potentials when associated with these subunits, as indicated by more negative midpoints of activation (England et al., 1995a; England et al., 1995b; Majumder et al., 1995; Uebele et al., 1996; De Biasi et al., 1997). In addition, co-expression with the pi-subunits imparted partial fast N-type inactivation at depolarized potentials between +30 and + 90 mV to Kvl .5 channels, and slowed deactivation (England et al., 1995a; England et al., 1995b; Majumder et al., 1995; Sewing et al., 1996; De Biasi et al., 1997). Several laboratories have also reported that Kvl .5 channel current amplitude was decreased by co-expression with KvP 1.2- and KvP2-subunits (Majumder et al., 1995; Accili et al., 1997; De Biasi et al., 1997), while co-expression with Kvpi .3 appeared to increase expression of Kvl .5 (Williams et al., 2002). 26 The N-type inactivation conferred by Kvpl.3 subunits can be inhibited by either P K A activation (Kwak et al., 1999a) or protein kinase C inhibition with calphostin C (Kwak et al., 1999b). The latter effect is also true for Kv(31.2. This indicates that either Kvl .5 and/or the Kv(31 subunits may require de-phosphorylation of a P K A site(s) or phosphorylation of P K C site(s) to induce N-type inactivation or both. The researchers were able to narrow the P K A site to Ser24 in the N-terminus of Kv(31.3 and rule out the involvement of one potential PKC site on Kvpi.3 and all Kv l .5 C-terminal P K C sites in this phenomenon. Further mutational analysis indicated that it was an electrostatic effect of phosphorylation that accounted for loss of N-type inactivation on P K A stimulation (Kwak et al., 1999a). A follow-up study by this same group showed that stimulation of PKC with P M A decreased Kvl .5 current density when the channel was assembled with Kvpi .2 but not Kvpi.3 (Williams et al., 2002). This data shows that a whole spectrum of channel activity can be achieved by co-assembly of the channel with KvP 1 subunits and further modulation by P K C and P K A pathways. Co-expression of hKvl.5 with hKvp3.1 resulted in channels that showed very fast and complete N-type inactivation without changing activation kinetics (Liecher et al., 1998). In contrast, rKvP3, which shows 92% sequence identity with hKvP3.1 (Liecher et al., 1998), has no effect on Kvl .5 (Heinemann et ah, 1995). Another P-subunit found in mouse, mKvP4 was shown to immunoprecipitate with Kvl .5 but had no detected effects on the channel current amplitude or kinetics (Fink et al., 1996). This is in contrast to the subunits effects of increasing Kv2.2 but not Kv2.1 current density via a C-terminal interaction. Another K v channel accessory subunit, KChAP, acts as a chaperone in the trafficking of Kv4.3 and Kv2.1 channels; this association increases channel expression without affecting channel kinetics (Wible et al., 1998). While a yeast two-hybrid assay indicated that KChAP could bind the N-terminus of K v l . 5 , it appears to have no effect on channel expression in 27 transfected cells and the two proteins do not co-immunoprecipitate from cardiac tissue. In co-injected oocytes, however, KChAP can prevent Kv(31.2 mediated decreases in Kvl .5 currents, due likely to competitive interactions between KChAP and Kvpi .2 (Kuryshev et al., 2001). In general, Kv(3 subunit interactions with Kvl .5 tend to limit the amount of current the channel conducts at depolarized potentials, which would further limit the channel's contributions during the plateau and early re-polarization phases of an action potential. Together with the channel's slow rates of recovery, contributions made by Kvl .5 :KvP complexes would be limited to situations of low stimulus frequency and/or activity in the sub-threshold voltage range. Channel contributions could be increased by P K A stimulation through such pathways as P-adrenergic stimulation (Li et al., 1996) in the heart or by an increase in phosphatase activity. With the coassembly of the channel with p-subunits comes a whole new layer of regulation and phenotypic behavior. a-Actinin a-Actinin-2 was identified in a yeast two-hybrid screen of a human cardiac library as a Kvl .5 interacting protein (Maruoka et al, 2000). Actinins are cytoskeletal interacting proteins which bind actin as well as several surface proteins and receptors such as the N M D A receptor (Wyszynski et al 1997) and integrins (Otey et al., 1990). There are four isoforms of actinin found in humans, actinin-1 is ubiquitously expressed (Pavalko and Burridge, 1991), actinin-2 and -3 are described as muscle-specific (Beggs et al, 1992) and actinin-4, which is also widely expressed but at lower levels than actinin-1 (Honda et al., 1998). Actinins are found in a variety of junctions including cell-cell contacts, cell-matrix contacts, cellular protrusions, lamellipodia and stress fibre dense regions, including focal adhesions, adherens junctions and hemidesmosomes, where they provide structural stability and elasticity during mechanical strain (reviewed in Otey and Carpen, 2004). Actinins exist as anti-parallel dimers with each subunit 28 consisting of an N-terminal actin binding domain, several spectrin repeats and a C-terminal calmodulin domain, made up of 2 EF-hand repeats. Due to mutations acquired over evolutionary time in the muscle specific forms of actinin (2 and 3), this calmodulin domain has lost its ability to bind calcium at physiological concentrations (Beggs et al., 1992). When calcium-bound, the non-muscle isoforms have a lower affinity for actin, breaking the link between stress fibres and focal adhesions, as part of the detachment step of cell migration. Actinins are also regulated by phosphatidylinositol intermediaries, binding by phosphatidylinositol (3,4,5) phosphate (PIP3) decreases actinins affinity for integrin as well as its ability to bundle actin filaments (Greenwood et al., 2000), resulting in a remodeling of focal adhesions during retraction of cell surface membrane domains. Anti-sense knockdown of a-actinin-2 in HEK293 cells expressing Kvl .5 resulted in a large increase in channel current density after 2 hours, similar to that obtained by treatment of the cells with the actin depolymerization agent cytochalasin-D (Maruoka et al., 2000). A similar effect of cytochalasin was observed in cells expressing Kv4.2 (Wang et al., 2004). These studies highlight the importance of the actin cytoskeleton in the regulation of channel expression but the exact mechanism involved in the increase in channel density remains to be discovered. Increases in Kv4.2 current were near maximal at 1 hour of cytochalasin treatment and had no effect on channel kinetics nor single channel conductance (Wang et al., 2004), ruling out an increase in gene transcription and, likely also, new synthesis of protein. Increases in channel number within this time frame indicate a decrease in E R A D and/or a decrease in channel turnover, where the normal half-life of Kv channels is around 4 hours (Takimoto et al., 1993). Research on other multimeric channel complexes have shown ER associated channel assembly to be quite inefficient with only 20-40% of synthesized subunits making it to the plasma membrane as part of mature complexes (Merlie and Lindstrom, 1983; Shmidt and Catterall, 1986; Christianson and Green, 2004). It appears likely that the majority of synthesized subunits are ubiquitinated and re-29 translocated out of the ER for degradation by the 26S proteasome (Christianson and Green, 2004), and therefore the level of surface expression is a result of competition between assembly and degradation. Src Family Kinases Several studies have shown that Src family tyrosine kinases can regulate Kvl .5 channel activity by simply binding as well as by phosphorylating the channel in a variety of cell types. Human Kvl .5 and Src kinase co-localize at adhesion zones in human myocardium and can be co-immunoprecipitated from ventricular tissue as well (Holmes et al., 1996). The same study found that when expressed together with the constitutively active v-Src in heterologous cells, Kv l .5 is phosphorylated at an undetermined tyrosine and current amplitudes are suppressed without a detectable difference in protein expression. In oocytes, Src showed two modes of modulation, one phosphorylation-dependent and one phosphorylation-independent but binding dependent. Both lead to a decrease in Kvl .5 currents (Nitabach et al., 2002). In Schwann cells isolated from mouse sciatic nerve (4 days postnatal), Kv l .5 could be co-immunoprecipitated with another Src family kinase, Fyn, both from acutely dissociated cells and those cultured for several days (Sobko et al., 1998). Treatment of these cells with a membrane permeant tyrosine kinase inhibitor, Herbamycin A, led to a decrease in whole-cell sustained current but the phosphorylation state of Kv l .5 did not appear to change. In addition, when Fyn and ATP were added to the patch pipette solution, whole-cell currents increased, which is inconsistent with the affects of Src on Kvl .5 observed in heterologous cells (Holmes et al., 1996) and oocytes (Nitabach et al., 2001) but has been noted in primary culture of other cell types and may be related to phosphatase activity in these cells. A loss of Protein Tyrosine Phosphatase s (PTPe) activity as a result of targeted knock-out of this gene in mice led to higher voltage-gated K + currents and hyper-phosphorylation of Kvl .5 and Kv2.1 in Schwann cells from these transgenic animals (Peretz et al., 2000). Loss of PTPs was associated with delayed 30 myelination in knockout mice, perhaps by preventing exit from the cell cycle. However, sciatic nerve axons from adult mice were normal, indicating that the delay was not necessarily fatal. Kv l .5 has also been co-immunoprecipitated with Fyn from rat hippocampal neurons (Nitabach et al.,2001). In astrocytes (MacFarlane and Sontheimer, 2000) and microglia (Kotecha and Schlichter, 1999), Kv l .5 expression is associated with proliferating phenotypes. However, once confluencey/differentiation is achieved, Kvl .5 is downregulated, and at least in cultured astrocytes this is associated with a loss in the tyrosine phosphorylation state of the channel. Treatment of these cells with a Src family tyrosine kinase inhibitor, PP2, did not prevent the interaction between a 55 KDa Src family protein and Kv l .5 , though Kvl .5 was no longer phosphorylated and the delayed rectifier component of whole-cell recording was reduced by 44% (MacFarlane and Sontheimer, 2000). This same study noted that Kvl .5 remained tyrosine phosphorylated throughout the cell cycle in proliferating cells and thus associated changes in channel activity must be modulated by a mechanism other than tyrosine kinases, or other K + channels found in these cells predominate during the transient hyperpolarization required for cell-cycle progression (reviewed in Pardo, 2004). Treatment of oligodendritic progenitor cells with platelet-derived growth factor leads to an increase in Kvl .5 protein and associated currents Chittajallu et al., 2002), an effect that can be blocked by treatment with the Src-kinase inhibitor PP2 (Soliven et al., 2003). When assembled as a part of an heterotetramer, Kvl .5 can also act as an adaptor protein for tyrosine phosphorylation of Kvl .4 , which does not contain a proline-rich domain and is normally unaffected by co-expression with the constitutively active v-Src (Nitabach et al., 2002). If the proline-rich domain from Kvl .5 is inserted into the cytoplasmic N - or C-terminus of Kv l .4 , the resulting mutants become sensitive to v-Src and similar to Kvl .5 currents are suppressed by binding dependent and phosphorylation-dependent mechanisms. 31 Heteromultimerization likely has physiological relevance since immunoprecipitates of Kvl .5 from hippocampal tissue contain Kv l . 2 , Kv l .4 and Fyn (Nitabach et al., 2001). As the authors of that study noted, adaptor activity of this nature would generate channels sensitive to a larger array of signaling pathways and perhaps this explains the difficulty in determining the identity of channels that underlie specific currents in various tissues. Scope of thesis Investigation Many of the studies detailed in this thesis were undertaken to better understand Kvl .5 localization in both the expression system used in the lab, cultured H E K cells, and in cardiac myocytes. The driving hypothesis was that localization was dictated by protein-protein interactions with scaffold and/or cytoskeletal elements, proteins that would be involved in anchoring the channel at functional sites in the membrane. As a starting point for examining the scope of possibilities for anchoring proteins, we began with an archetypal PDZ domain containing protein PSD-95 to determine if the C-terminal consensus PDZ binding site in this channel was functional. These studies were carried out solely in cultured cells stably expressing channel and transiently transfected PSD-95. We found, rather surprisingly, that PSD-95 could bind to the C- and N-termini of the channels, and that depending whether we deleted the N - or C-terminus, co-expression had opposite effects on current density (Eldstrom et al., 2002). While this was of great interest in terms of the potential for regulation of channel expression, we wanted to focus on PDZ proteins that had been identified in cardiac tissue and so in the second series of experiments we chose to study another PDZ protein, SAP97. Others had shown that this protein, when co-expressed with K v l . 5 , more than doubled current density in CHO cells (Godreau et a l , 2003), and that same group was able to co-immunoprecipitate Kvl .5 using an anti-PSD-95 family antibody from atrial membrane fractions. Based on the molecular weight of the protein immunoprecipitated, they determined that SAP97 interacted with Kvl .5 channels in the heart. Our own studies confirmed the effect of co-expression on current density but we were 32 unable to show that the two proteins interacted in either HEK, CHO, COS-7 or rat cardiac tissue, or in a yeast two-hybrid screen of the PDZ domains (Eldstrom et al., 2003). Our results were significant in that we showed that the SAP97-mediated increase in Kvl .5 was dependent on the presence of the N-terminus of the channel. The third set of experiments were a departure from the other two in that we were taking a step back to characterize IKUR in canine atrial myocytes so that we could continue looking at Kv l .5 in cardiac tissue rather than in artificial expression systems. There has been some debate over the molecular correlate of IKUR in canine with some reports that the current is a result of Kv3.1 expression in this species, (Yu et al., 2000). Our results demonstrate clearly that, at least in the samples used in our study, Kvl .5 is an important contributor to canine atrial repolarization (Fedida et al., 2003). This study also allowed us to determine the gross cellular distribution of the channel in canine atrial myocytes, which was most apparent at the intercalated discs and to a lesser extent in lateral membranes with little T-tubule staining. The fourth set of experiments were undertaken to further our knowledge of channel distribution within the plasma membrane, with the hopes that we could narrow distribution to discrete membrane domains. This work stemmed from a publication that suggested that Kvl .5 was associated with lipid rafts and in particular caveolae, in a transiently transfected mouse fibroblasts cell line (Martens et al., 2001). Early experiments during our study failed to show a strong relationship between caveolin-3 and Kvl .5 in terms of cellular distribution and a complete inability to co-immunoprecipitate the two proteins (Eldstrom et al., in preparation). In addition, Kvl .5 expressed in H E K cells and in rat cardiac tissue did not co-sediment in light buoyant fractions along with flotillin or caveolin-3, both of which are considered raft marker proteins, but instead Kvl .5 was found in denser fractions along with non-raft associated transferrin receptors. 33 REFERENCES Abe A , Yamamoto T, Isome M , Ma M , Yaoita E, Kawasaki K, Kihara I, Aizawa Y . (1998) Thyroid hormone regulates expression of shaker-related potassium channel mRNA in rat heart. Biochem Biophys Res Commun. 245(l):226-30. Abi-Gerges N , Philp K, Pollard C, Wakefield I, Hammond TG, Valentin JP. (2004) Sex differences in ventricular repolarization: from cardiac electrophysiology to Torsades de Pointes. Fundam Clin Pharmacol. 18(2): 139-51. Accil i EA, Kiehn J, Yang Q, Wang Z, Brown A M , Wible B A . (1997) Separable Kvbeta subunit domains alter expression and gating of potassium channels. J Biol Chem. 272(41):25824-31. Archer SL, London B, Hampl V , Wu X , Nsair A , Puttagunta L , Hashimoto K , Waite RE, Michelakis ED. (2001) Impairment of hypoxic pulmonary vasoconstriction in mice lacking the voltage-gated potassium channel Kv l .5 . FASEB J. 15(10):1801-3. Archer SL, Souil E, Dinh-Xuan AT, Schremmer B, Mercier JC, E l Yaagoubi A , Nguyen-Huu L , Reeve HL, Hampl V . (1998) Molecular identification of the role of voltage-gated K + channels, Kv l .5 and Kv2.1, in hypoxic pulmonary vasoconstriction and control of resting membrane potential in rat pulmonary artery myocytes. J Clin Invest. 101(11):2319-30. Armstrong, C. M . (1971) Interaction of tetraethylammonium ion derivatives with the potassium channels of giant axon. J. Gen. Physiol. 58, 413-437 Bachmann A, Gutcher I, Kopp K , Brendel J, Bosch RF, Busch A E , Gogelein H. (2001) Characterization of a novel Kvl .5 channel blocker in Xenopus oocytes, CHO cells, human and rat cardiomyocytes. Naunyn Schmiedebergs Arch Pharmacol. 364(5):472-8. Banerjee SK, Ulrich JM, Kaldor GJ. (1988) Nuclear thyroid hormone receptors in rabbit heart: reduced triiodothyronine binding in atrium compared with ventricle. Circ Res. 63(1):267-71 Barry D M , Trimmer JS, Merlie JP, Nerbonne JM. (1995) Differential expression of voltage-gated K + channel subunits in adult rat heart. Relation to functional K + channels? Circ Res. 77(2):361-9. Baukrowitz T, Yellen G. (1996) Use-dependent blockers and exit rate of the last ion from the multi-ion pore of a K + channel. Science. 271(5249):653-6. Beggs A H , Byers TJ, Knoll JH, Boyce F M , Bruns GA, Kunkel L M . (1992) Cloning and characterization of two human skeletal muscle alpha-actinin genes located on chromosomes 1 and 11. J Biol Chem. 267(13):9281-8. Bixby K A , Nanao M H , Shen N V , Kreusch A, Bellamy H, Pfaffinger PJ, Choe (1999) Zn2+-binding and molecular determinants of tetramerization in voltage-gated K + channels. Nat Struct Biol. 6(l):38-43. 34 Bosch RF, Wang Z, L i GR, Nattel S. (1999) Electrophysiological mechanisms by which hypothyroidism delays repolarization in guinea pig hearts. Am J Physiol. 277(1 Pt 2):H211-20. Bou-Abboud E, Nerbonne JM. (1999) Molecular correlates of the calcium-independent, depolarization-activated K + currents in rat atrial myocytes. J Physiol. 517 (Pt 2):407-20. Bou-Abboud E, L i H , Nerbonne JM. (2000) Molecular diversity of the repolarizing voltage-gated K + currents in mouse atrial cells. J Physiol. 529 Pt 2:345-58. Brent GA. (1994) The molecular basis of thyroid hormone action. N Engl J Med. 331(13): 847-53. Brouillette J, Rivard K, Lizotte E, Fiset C. (2005) Sex and strain differences in adult mouse cardiac repolarization: importance of androgens. Cardiovasc Res. 65(1): 148-57. Brouillette J, Trepanier-Boulay V, Fiset C. (2003) Effect of androgen deficiency on mouse ventricular repolarization. J Physiol. 546(Pt 2):403-13. Bru-Mercier G, Deroubaix E, Capuano V, Ruchon Y , Rucker-Martin C, Coulombe A, Renaud JF. (2003) Expression of heart K + channels in adrenalectomized and catecholamine-depleted reserpine-treated rats. J Mol Cell Cardiol. 35(2): 153-63. Campomanes CR, Carroll KI, Manganas L N , Hershberger M E , Gong B, Antonucci DE, Rhodes KJ , Trimmer JS. (2002) Kv beta subunit oxidoreductase activity and K v l potassium channel trafficking. J Biol Chem. 277(10):8298-305. Chen C, Zhang J, Vincent JD, Israel JM (1990) Somatostatin increases voltage-dependent potassium currents in rat somatotrophs. Am J Physiol. 259(6 Pt 1):C854-61. Chen FS, Fedida D. (1998) On the mechanism by which 4-Aminopyridine occludes quinidine block of the cardiac K + channel, hKvl .5. J Gen Physiol. 111(4):539-54. Chittajallu R, Chen Y , Wang H, Yuan X , Ghiani CA, Heckman T, McBain CJ, Gallo V. (2002) Regulation of K v l subunit expression in oligodendrocyte progenitor cells and their role in G l /S phase progression of the cell cycle. Proc Natl Acad Sci U S A . 99(4):2350-5. Chouinard SW, Wilson GF, Schlimgen A K , Ganetzky B. (1995) A potassium channel beta subunit related to the aldo-keto reductase superfamily is encoded by the Drosophila hyperkinetic locus. Proc Natl Acad Sci U S A . 92(15):6763-7. Christianson JC, Green WN. (2004) Regulation of nicotinic receptor expression by the ubiquitin-proteasome system. E M B O J. 23(21):4156-65 Cohen FE, Kelly JW. (2003) Therapeutic approaches to protein-misfolding diseases. Nature. 426(6968):905-9. De Biasi M , Wang Z, Accili E, Wible B, Fedida D. (1997) Open channel block of human heart hKvl.5 by the beta-subunit hKv beta 1.2. Am J Physiol. 272(6 Pt 2):H2932-41. 35 de Groot RP, Sassone-Corsi P. (1993) Hormonal control of gene expression: multiplicity and versatility of cyclic adenosine 3',5'-monophosphate-responsive nuclear regulators. Mol Endocrinol. 7(2): 145-53. del Camino, D., Holmgren, M . , Liu, Y . & Yellen, G. (2000) Blocker protection in the pore of a voltage-gated K + channel and its structural implications. Nature 403, 321-325 Deutsch C, Chen LQ. (1993) Heterologous expression of specific K + channels in T lymphocytes: functional consequences for volume regulation. Proc Natl Acad Sci U S A . 90(21): 10036-40. Doyle DA, Morais Cabral J, Pfuetzner RA, Kuo A, Gulbis JM, Cohen SL, Chait BT, MacKinnon R. (1998) The structure of the potassium channel: molecular basis of K+ conduction and selectivity. Science. 280(5360):69-77 Drewe JA, Verma S, Freeh G, Joho RH. (1992) Distinct spatial and temporal expression patterns of K + channel mRNAs from different subfamilies. J Neurosci. 12(2):538-48. Durell, S. R. & Guy, H. R. (1992) Atomic scale structure and functional models of voltage-gated potassium channels. Biophys. J. 62, 238-247 Eldstrom J, Choi WS, Steele DF, Fedida D. (2003) SAP97 increases Kvl .5 currents through an indirect N-terminal mechanism. FEBS Lett. 17;547(l-3):205-ll. Eldstrom J, Doerksen K W , Steele DF, Fedida D. (2002) N-terminal PDZ-binding domain in K v l potassium channels. FEBS Lett. 2002 531(3):529-37. England SK, Uebele V N , Kodali J, Bennett PB, Tamkun M M . (1995) A novel K + channel beta-subunit (hKv beta 1.3) is produced via alternative mRNA splicing. J Biol Chem. 270(48):28531-4. England SK, Uebele V N , Shear H, Kodali J, Bennett PB, Tamkun M M . (1995b) Characterization of a voltage-gated K + channel beta subunit expressed in human heart. Proc Natl Acad Sci U S A . 92(14):6309-13. Fedida D, Eldstrom J, Hesketh JC, Lamorgese M , Castel L, Steele DF, Van Wagoner DR. (2003) Kvl .5 is an important component of repolarizing K + current in canine atrial myocytes. Circ Res. 93(8):744-51. Fedida D, Wible B, Wang Z, Fermini B, Faust F, Nattel S, Brown A M . (1993) Identity of a novel delayed rectifier current from human heart with a cloned K + channel current. Circ Res. 1993 Jul;73(l):210-6. Felipe A , Snyders DJ, Deal K K , Tamkun M M . (1993) Influence of cloned voltage-gated K + channel expression on alanine transport, Rb+ uptake, and cell volume. A m J Physiol. 265(5 Pt l):C1230-8. 36 Feng J, Wible B, L i GR, Wang Z, Nattel S. (1997). Antisense oligodeoxynucleotides directed against Kvl .5 mRNA specifically inhibit ultrarapid delayed rectifier K + current in cultured adult human atrial myocytes. Circ Res. 80(4):572-9. Fink A L . (1999). Chaperone-mediated protein folding. Physiol Rev. 79(2):425-49. Fink M , Duprat F, Lesage F, Heurteaux C, Romey G, Barhanin J, Lazdunski M . (1996). A new K + channel beta subunit to specifically enhance Kv2.2 (CDRK) expression. J Biol Chem. 271(42):26341-8. Folco EJ, Liu G X , Koren G. (2004). Caveolin-3 and SAP97 form a scaffolding protein complex that regulates the voltage-gated potassium channel Kvl .5 A m J Physiol Heart Circ Physiol. 287(2):H681-90. Gandhi, C.S., E. Clark, E. Loots, A . Pralle, and E.Y. Isacoff. (2003). The orientation and molecular movement of a K + channel voltage-sensing domain. Neuron. 40:515-525 Gandhi CS, Isacoff E Y . (2002). Molecular models of voltage sensing. J Gen Physiol. 120(4):455-63. Greenwood JA, Theibert A B , Prestwich GD, Murphy-Ullrich JE. (2000). Restructuring of focal adhesion plaques by PI 3-kinase. Regulation by Ptdlns (3,4,5)-p(3) binding to alpha-actinin. J Cell Biol. 150(3):627-42. Grinstein S, Smith JD. (1990). Calcium-independent cell volume regulation in human lymphocytes. Inhibition by charybdotoxin. J Gen Physiol. 95(1):97-120. Godreau D, Vranckx R, Maguy A , Goyenvalle C, Hatem SN. (2003). Different isoforms of synapse-associated protein, SAP97, are expressed in the heart and have distinct effects on the voltage-gated K + channel Kv l .5 . J Biol Chem. 278(47):47046-52. Griffith L C . (2001). Potassium channels: the importance of transport signals. Curr Biol. ll(6):R226-8. Gula LJ , Krahn A D , Skanes A C , Yee R, Klein GJ. (2004). Clinical relevance of arrhythmias during sleep: guidance for clinicians. Heart. 90(3):347-52. Gulbis JM, Mann S, MacKinnon R. (1999). Structure of a voltage-dependent K + channel beta subunit. Cell. 97(7):943-52. Gulbis JM, Zhou M , Mann S, MacKinnon R. (2000). Structure of the cytoplasmic beta subunit-T l assembly of voltage-dependent K + channels. Science. 289(5476): 123-7. Guo W, Kada K, Kamiya K, Toyama J. (1997). IGF-I regulates K(+)-channel expression of cultured neonatal rat ventricular myocytes. Am J Physiol. 272(6 Pt 2):H2599-606. Guo W, Kamiya K, Yasui K, Kodama I, Toyama J. (1998). alpha 1-adrenoceptor agonists and IGF-1, myocardial hypertrophic factors, regulate the Kvl .5 K + channel expression differentially in cultured newborn rat ventricular cells. Pflugers Arch. 436(1 ):26-32. 37 Gutman G A , Chandy K G , Adelman JP, Aiyar J, Bayliss DA, Clapham DE, Covarriubias M , Desir GV, Furuichi K , Ganetzky B, Garcia M L , Grissmer S, Jan L Y , Karschin A , Kim D, Kuperschmidt S, Kurachi Y , Lazdunski M , Lesage F, Lester HA, McKinnon D, Nichols CG, O'Kelly I, Robbins J, Robertson GA, Rudy B, Sanguinetti M , Seino S, Stuehmer W, Tamkun M M , Vandenberg CA, Wei A, Wulff H , Wymore RS; International Union of Pharmacology. (2003). International Union of Pharmacology. X L I . Compendium of voltage-gated ion channels: potassium channels. Pharmacol Rev. 55(4):583-6. Hackos, D.A., T.-H. Chang, and K.J . Swartz. (2002). Scanning the intracellular S6 activation gate in the Shaker K + channel. J. Gen. Physiol. 119:521-532 Heginbotham L, Abramson T, MacKinnon R. (1992). A functional connection between the pores of distantly related ion channels as revealed by mutant K + channels. Science. 258(5085): 1152-5. Heinemann SH, Rettig J, Graack HR, Pongs O. (1996). Functional characterization of K v channel beta-subunits from rat brain. J Physiol. 493 ( Pt 3):625-633. Heinemann SH, Rettig J, Wunder F, Pongs O. (1995). Molecular and functional characterization of a rat brain Kv beta 3 potassium channel subunit. FEBS Lett. 377(3):383-9. Hille, B, (2001). Ion channels of excitable membranes, 3rd Edition Sunderland, M A : Sinauer Assoc. Holmgren, M . , K.S. Shin, and G. Yellen. (1998). The activation gate of a voltage-gated K + channel can be trapped in the open state by an intersubunit metal bridge. Neuron. 21:617-621. Holmes TC, Fadool DA, Ren R, Levitan IB. (1996). Association of Src tyrosine kinase with a human potassium channel mediated by SH3 domain. Science. 274(5295):2089-91. Honda K, Yamada T, Endo R, Ino Y , Gotoh M , Tsuda H, Yamada Y , Chiba H, Hirohashi S. (1998). Actinin-4, a novel actin-bundling protein associated with cell motility and cancer invasion. J Cell Biol. 140(6): 1383-93. Horio Y , Hibino H , Inanobe A, Yamada M , Ishii M , Tada Y , Satoh E, Hata Y , Takai Y , Kurachi Y . (1997). Clustering and enhanced activity of an inwardly rectifying potassium channel, Kir4.1, by an anchoring protein, PSD-95/SAP90. J Biol Chem. 272(20): 12885-8. Horn R. (2004). How S4 segments move charge. Let me count the ways. J Gen Physiol. 123(l):l-4. Hoshi T, Zagotta W N , Aldrich RW. (1990). Biophysical and molecular mechanisms of Shaker potassium channel inactivation. Science. 250(4980):533-8. Hoshi T, Zagotta WN, Aldrich RW. (1991). Two types of inactivation in Shaker K + channels: effects of alterations in the carboxy-terminal region. Neuron. 7(4):547-56. Huang X Y , Morielli A D , Peralta EG. (1994). Molecular basis of cardiac potassium channel stimulation by protein kinase A. Proc Natl Acad Sci U S A . 91(2):624-8. 38 Hugnot JP, Salinas M , Lesage F, Guillemare E, de Weille J, Heurteaux C, Mattei M G , Lazdunski M . (1996) Kv8.1, a new neuronal potassium channel subunit with specific inhibitory properties towards Shab and Shaw channels. E M B O J. 15(13):3322-31. Hulme JT, Coppock EA, Felipe A , Martens JR., Tamkun M M . (1999) Oxygen sensitivity of cloned voltage-gated K(+) channels expressed in the pulmonary vasculature. Circ Res. 85(6):489-97. Isacoff E Y , Jan Y N , Jan L Y . (1990) Evidence for the formation of heteromultimeric potassium channels in Xenopus oocytes. Nature. 345(6275):530-4. Isacoff E Y , Jan Y N , Jan L Y . (1991) Putative receptor for the cytoplasmic inactivation gate in the Shaker K + channel. Nature. 353(6339):86-90. Jegla T, Salkoff L . (1997) A novel subunit for shal K + channels radically alters activation and inactivation. JNeurosci. 17(l):32-44. Jiang Y , Ruta V , Chen J, Lee A, MacKinnon R. (2003a) The principle of gating charge movement in a voltage-dependent K + channel. Nature. 423(6935):42-8. Jiang Y , Lee A, Chen J, Ruta V , Cadene M , Chait BT, MacKinnon R. (2003b) X-ray structure of a voltage-dependent K + channel. Nature. 423(6935):33-41. Jiang, Y . , A . Lee, J. Chen, M . Cadene, B.T. Chait, and R. MacKinnon. (2002a) Crystal structure and mechanism of a calcium-gated potassium channel. Nature. 417:515-522 Jiang, Y , Lee, A , Chen, J, Cadene, M , Chait, BT, MacKinnon, R. (2002) The open pore conformation of potassium channels. Nature 417, 523-526 Kamp TJ, Hell JW. (2000) Regulation of cardiac L-type calcium channels by protein kinase A and protein kinase C. Circ Res. 87(12): 1095-102. Kehl SJ, Eduljee C, Kwan DC, Zhang S, Fedida D. (2002) Molecular determinants of the inhibition of human Kv l .5 potassium currents by external protons and Zn(2+). J Physiol. 2002 541(Pt l):9-24. Klein I, Ojamaa K. (2001) Thyroid hormone and the cardiovascular system. N Engl J Med. 344(7):501-9. Kobertz WR, Williams C, Miller C. (2000) Hanging gondola structure of the T l domain in a voltage-gated K(+) channel. Biochemistry. 39(34): 10347-52. Koo GC, Blake JT, Talento A , Nguyen M , Lin S, Sirotina A , Shah K, Mulvany K , Hora D Jr, Cunningham P, Wunderler DL, McManus OB, Slaughter R, Bugianesi R, Felix J, Garcia M , Williamson J, Kaczorowski G, Sigal N H , Springer MS, Feeney W. (1997) Blockade of the voltage-gated potassium channel Kvl .3 inhibits immune responses in vivo. J Immunol. 158(11):5120-8. 39 Kosolapov A , Deutsch C. (2003) Folding of the voltage-gated K + channel T l recognition domain. J Biol Chem. 278(6):4305-13. Kosolapov A, Tu L, Wang J, Deutsch C. (2004) Structure acquisition of the T l domain of Kvl .3 during biogenesis. Neuron. 44(2):295-307. Kotecha SA, Schlichter L C . (1999) Kvl .5 to Kvl .3 switch in endogenous hippocampal microglia and a role in proliferation. J Neurosci. 19(24): 10680-93. Kreusch A , Pfaffinger PJ, Stevens CF, Choe S. (1998) Crystal structure of the tetramerization domain of the Shaker potassium channel. Nature. 392(6679):945-8. Kreuzberg U , Theissen P, Schicha H, Schroder F, Mehlhorn U , de Vivie ER, Boknik P, Neumann J, Grohe C, Herzig S. (2000) Single-channel activity and expression of atrial L-type Ca(2+) channels in patients with latent hyperthyroidism. A m J Physiol Heart Circ Physiol. 278(3):H723-30. Kucera JP, Rohr S, Rudy Y . (2002) Localization of sodium channels in intercalated disks modulates cardiac conduction. Circ Res. 91(12): 1176-82. Kurata HT, Soon GS, Fedida D. (2001) Altered state dependence of c-type inactivation in the long and short forms of human K v l . 5 . J Gen Physiol. 118(3):315-32. Kuryshev Y A , Wible B A , Gudz T l , Ramirez A N , Brown A M . (2001) KChAP/Kvbetal.2 interactions and their effects on cardiac Kv channel expression. A m J Physiol Cell Physiol. 281(l):C290-9. Komukai K , Brette F, Pascarel C, Orchard CH. (2002a) Electrophysiological response of rat ventricular myocytes to acidosis. A m J Physiol Heart Circ Physiol. 283(l):H412-22. Komukai K , Brette F, Orchard CH. (2002b) Electrophysiological response of rat atrial myocytes to acidosis. Am J Physiol Heart Circ Physiol. 283(2):H715-24. Kwak Y G , Hu N , Wei J, George A L Jr, Grobaski TD, Tamkun M M , Murray KT. (1999a) Protein kinase A phosphorylation alters Kvbetal.3 subunit-mediated inactivation of the Kvl .5 potassium channel. J Biol Chem. 274(20): 13928-32. Kwak Y G , Navarro-Polanco RA, Grobaski T, Gallagher DJ, Tamkun M M . (1999b) Phosphorylation is required for alteration of kvl.5 K(+) channel function by the Kvbetal.3 subunit. J Biol Chem. 274(36):25355-61. Labro AJ , Raes A L , Snyders DJ. (2005) Coupling of voltage sensing to channel opening reflects intrasubunit interactions in Kv channels. J Gen Physiol. 125(l):71-80. Labro, A.J. , A . L . Raes, I. Bellens, N . Ottschytsch, and D.J. Snyders. (2003) Gating of Shaker-type channels requires the flexibility of S6 caused by prolines. J. Biol. Chem. 278:50724-50731. Lazaroff M A , Hofmann A D , Ribera A B . (1999) Xenopus embryonic spinal neurons express potassium channel Kvbeta subunits. J Neurosci. 19(24): 10706-15. 40 Lee, H.C., J .M. Wang, and K.J . Swartz. (2003). Interaction between extracellular Hanatoxin and the resting conformation of the voltage-sensor paddle in Kv channels. Neuron. 40:527-536 Leicher T, Bahring R, Isbrandt D, Pongs O. (1998) Coexpression of the K C N A 3 B gene product with Kv l .5 leads to a novel A-type potassium channel. J Biol Chem.273(52):35095-101. L i GR, Feng J, Wang Z, Fermini B, Nattel S. (1996) Adrenergic modulation of ultrarapid delayed rectifier K + current in human atrial myocytes. Circ Res. 78(5):903-15. L i GR, Feng J, Yue L, Carrier M , Nattel S. (1996) Evidence for two components of delayed rectifier K + current in human ventricular myocytes Circ Res. 1996 Apr;78(4):689-96. L i H , Guo W, Yamada K A , Nerbonne JM. (2004) Selective elimination of I(K,slowl) in mouse ventricular myocytes expressing a dominant negative Kvl.5alpha subunit. Am J Physiol Heart Circ Physiol. 286(1 ):H319-28. L i M , Jan Y N , Jan L Y . (1992) Specification of subunit assembly by the hydrophilic amino-terminal domain of the Shaker potassium channel. Science. 257(5074): 1225-30. London B, Jeron A , Zhou J, Buckett P, Han X , Mitchell GF, Koren G. (1998) Long QT and ventricular arrhythmias in transgenic mice expressing the N terminus and first transmembrane segment of a voltage-gated potassium channel. Proc Natl Acad Sci U S A . 95(6):2926-31. Loots E, Isacoff E Y . (1998) Protein rearrangements underlying slow inactivation of the Shaker K + channel. J Gen Physiol. 112(4):377-89. Lopez GA, Jan Y N , Jan L Y . (1994) Evidence that the S6 segment of the Shaker voltage-gated K + channel comprises part of the pore. Nature. 367(6459): 179-82. Lu, Z., A . M . Klem, and Y . Ramu. (2002). Coupling between voltage sensors and activation gate in voltage-gated K + channels. J. Gen. Physiol. 120:663-676 Lu J, Robinson JM, Edwards D, Deutsch C. (2001) T l - T l interactions occur in ER membranes while nascent K v peptides are still attached to ribosomes. Biochemistry. 40(37): 10934-46. Ma M L , Watanabe K, Watanabe H, Hosaka Y , Komura S, Aizawa Y , Yamamoto T. (2003) Different gene expression of potassium channels by thyroid hormone and an antithyroid drug between the atrium and ventricle of rats. Jpn Heart J. 44(1): 101-10. Majumder K, De Biasi M , Wang Z, Wible B A (1995) Molecular cloning and functional expression of a novel potassium channel beta-subunit from human atrium. FEBS Lett. 361(1): 13-6. MacDonald, P. E., Ha, X . F., Wang, J., Smukler, S. R., Sun, A . M . , Gaisano, H. Y . , Salapatek, A . M . F., Backx, P. H. & Wheeler, M . B. (2001) Members of the K v l and Kv2 Voltage-Dependent K + Channel Families Regulate Insulin Secretion. Mol . Endocrinol. 15, 1423-1435 41 MacFarlane SN, Sontheimer H. (2000) Modulation of Kvl .5 currents by Src tyrosine phosphorylation: potential role in the differentiation of astrocytes. J Neurosci. 20(14):5245-53. MacKinnon R, Yellen G. (1990) Mutations affecting TEA blockade and ion permeation in voltage-activated K + channels. Science. 250(4978):276-9. Maier SK, Westenbroek RE, Schenkman K A , Feigl EO, Scheuer T, Catterall WA. (2002) An unexpected role for brain-type sodium channels in coupling of cell surface depolarization to contraction in the heart. Proc Natl Acad Sci U S A . 99(6):4073-8. Malhotra JD, Thyagarajan V , Chen C, Isom L L . (2004) Tyrosine-phosphorylated and nonphosphorylated sodium channel betal subunits are differentially localized in cardiac myocytes. J Biol Chem. 279(39):40748-54. Maruoka ND, Steele DF, Au BP, Dan P, Zhang X , Moore ED, Fedida D. (2000) alpha-actinin-2 couples to cardiac Kvl .5 channels, regulating current density and channel localization in H E K cells. FEBS Lett. 12;473(2): 188-94. Mason HS, Latten MJ, Godoy L D , Horowitz B, Kenyon JL. (2002) Modulation of Kvl .5 currents by protein kinase A , tyrosine kinase, and protein tyrosine phosphatase requires an intact cytoskeleton. Mol Pharmacol. 61(2):285-93. Mays DJ, Foose JM, Philipson L H , Tamkun M M . (1995) Localization of the Kvl .5 K + channel protein in explanted cardiac tissue. J Clin Invest. 96(l):282-92. McCormack K, McCormack T, Tanouye M , Rudy B, Stuhmer W. (1995) Alternative splicing of the human Shaker K + channel beta 1 gene and functional expression of the beta 2 gene product. FEBS Lett. 370(l-2):32-6. McCormack K , Connor JX, Zhou L, Ho L L , Ganetzky B, Chiu SY, Messing A . (2002) Genetic analysis of the mammalian K + channel beta subunit Kvbeta 2 (Kcnab2). J Biol Chem. 277(15):13219-28. Meacham GC, Patterson C, Zhang W, Younger JM, Cyr D M . (2001) The Hsc70 co-chaperone CHIP targets immature CFTR for proteasomal degradation. Nat Cell Biol. 3(l):100-5. Merlie JP, Lindstrom J (1983) Assembly in vivo of mouse muscle acetylcholine receptor: identification of an alpha subunit species that may be an assembly intermediate. Cell 34: 747-757 Mohler PJ, Rivolta I, Napolitano C, LeMaillet G, Lambert S, Priori SG, Bennett V . (2004) Navl.5 E1053K mutation causing Brugada syndrome blocks binding to ankyrin-G and expression of Navl.5 on the surface of cardiomyocytes. Proc Natl Acad Sci U S A . 101(50):17533-8. Mori Y , Folco E, Koren G. (1995) GH3 cell-specific expression of Kvl .5 gene. Regulation by a silencer containing a dinucleotide repetitive element. J Biol Chem. 270(46):27788-96. 42 Mori Y , Matsubara H , Folco E, Siegel A , Koren G. (1993) The transcription of a mammalian voltage-gated potassium channel is regulated by cAMP in a cell-specific manner. J Biol Chem. 268(35):26482-93. Morales MJ , Castellino RC, Crews A L , Rasmusson RL, Strauss HC. (1995) A novel beta subunit increases rate of inactivation of specific voltage-gated potassium channel alpha subunits. J Biol Chem. 270(11):6272-7. Murata M , Buckett PD, Zhou J, Brunner M , Folco E, Koren G. (2001) SAP97 interacts with Kv l .5 in heterologous expression systems. Am J Physiol Heart Circ Physiol. 2001 Dec;281(6):H2575-84. Nagaya N , Papazian D M . (1997) Potassium channel alpha and beta subunits assemble in the endoplasmic reticulum. J Biol Chem. 272(5):3022-7. Naslavsky N , Weigert R, Donaldson JG. (2003) Convergence of non-clathrin- and clathrin-derived endosomes involves Arf6 inactivation and changes in phosphoinositides. Mol Biol Cell. 14(2):417-31. Nerbonne JM. (2000) Molecular basis of functional voltage-gated K + channel diversity in the mammalian myocardium. J Physiol. 525 Pt 2:285-98. Nichols BJ. (2002) A distinct class of endosome mediates clathrin-independent endocytosis to the Golgi complex. Nat Cell Biol. 4(5):374-8. Nitabach M N , Llamas DA, Thompson IJ, Collins K A , Holmes TC. (2002) Phosphorylation-dependent and phosphorylation-independent modes of modulation of shaker family voltage-gated potassium channels by SRC family protein tyrosine kinases. J Neurosci. 22(18):7913-22. Nitabach M N , Llamas DA, Araneda RC, Intile JL, Thompson IJ, Zhou YI , Holmes TC. (2001) A mechanism for combinatorial regulation of electrical activity: Potassium channel subunits capable of functioning as Src homology 3-dependent adaptors. Proc Natl Acad Sci U S A . 98(2):705-10. Obenauer JC, Cantley L C , Yaffe M B . (2003) Scansite 2.0: Proteome-wide prediction of cell signaling interactions using short sequence motifs. Nucleic Acids Res. 31(13):3635-41. Ojamaa K, Sabet A , Kenessey A, Shenoy R, Klein I. (1999) Regulation of rat cardiac Kvl .5 gene expression by thyroid hormone is rapid and chamber specific. Endocrinology. 140(7):3170-6. Olcese R, Latorre R, Toro L , Bezanilla F, Stefani E. (1997) Correlation between charge movement and ionic current during slow inactivation in Shaker K + channels. J Gen Physiol. 110(5):579-89. Orchard C H , Kentish JC. (1990) Effects of changes of pH on the contractile function of cardiac muscle. Am J Physiol. 258(6 Pt 1):C967-81. Otey CA, Carpen O. (2004) Alpha-actinin revisited: a fresh look at an old player. Cell Motil Cytoskeleton. 58(2): 104-11. 43 Otey C A , Pavalko F M , Burridge K. (1990) An interaction between alpha-actinin and the beta 1 integrin subunit in vitro. J Cell Biol. 111(2):721-9. Parcej D N , Scott V E , Dolly JO. (1992) Oligomeric properties of alpha-dendrotoxin-sensitive potassium ion channels purified from bovine brain. Biochemistry. 31(45): 11084-8. Pardo L A . (2004) Voltage-gated potassium channels in cell proliferation. Physiology (Bethesda). 19:285-92. Patel AJ , Lazdunski M , Honore E. (1997) Kv2.1/Kv9.3, a novel ATP-dependent delayed-rectifier K + channel in oxygen-sensitive pulmonary artery myocytes. E M B O J. 16(22):6615-25. Pathak M , Kurtz L , Tombola F, Isacoff E. (2005) The cooperative voltage sensor motion that gates a potassium channel. J Gen Physiol. 125(l):57-69. Pavalko F M , Burridge K. (1991) Disruption of the actin cytoskeleton after microinjection of proteolytic fragments of alpha-actinin. J Cell Biol. 114(3):481-91. Peretz A, Gil-Henn H, Sobko A , Shinder V , Attali B, Elson A . (2002) Hypomyelination and increased activity of voltage-gated K(+) channels in mice lacking protein tyrosine phosphatase epsilon. E M B O J. 2000 19(15):4036-45. Post M A , Kirsch GE, Brown A M . (1996) Kv2.1 and electrically silent Kv6.1 potassium channel subunits combine and express a novel current. FEBS Lett. 399(1-2): 177-82. Rao JN, Platoshyn O, L i L, Guo X , Golovina V A , Yuan JX, Wang JY. (2002) Activation of K(+) channels and increased migration of differentiated intestinal epithelial cells after wounding. Am J Physiol Cell Physiol. 282(4):C885-98. Rettig J, Heinemann SH, Wunder F, Lorra C, Parcej DN, Dolly JO, Pongs O. (1994) Inactivation properties of voltage-gated K + channels altered by presence of beta-subunit. Nature. 369(6478):289-94. Robinson JM, Deutsch C. (2005) Coupled tertiary folding and oligomerization of the T l domain of Kv channels. Neuron. 45(2):223-32. Rotin D, Staub O, Haguenauer-Tsapis R. (2000) Ubiquitination and endocytosis of plasma membrane proteins: role of Nedd4/Rsp5p family of ubiquitin-protein ligases. J Membr Biol. 2000 Jul 1;176(1):1-17. Rudy B, Chow A , Lau D, Amarillo Y , Ozaita A , Saganich M , Moreno H , Nadal MS, Hernandez-Pineda R, Hernandez-Cruz A , Erisir A , Leonard C, Vega-Saenz de Miera E. (1999) Contributions of Kv3 channels to neuronal excitability. Ann N Y Acad Sci. 868:304-43. Ruppersberg JP, Schroter K H , Sakmann B, Stocker M , Sewing S, Pongs O. (1990) Heteromultimeric channels formed by rat brain potassium-channel proteins. Nature. 345(6275):535-7. 44 Salinas M , de Weille J, Guillemare E, Lazdunski M , Hugnot JP. (1997 Modes of regulation of shab K + channel activity by the Kv8.1 subunit. J Biol Chem. 272(13):8774-80. Sato S, Ward C L , Krouse M E , Wine JJ, Kopito RR. (1996) Glycerol reverses the misfolding phenotype of the most common cystic fibrosis mutation. J B i o l Chem. 271(2):635-8. Schmidt JW, Catterall W A (1986) Biosynthesis and processing of the alpha subunit of the voltage-sensitive sodium channel in rat brain neurons. Cell 46: 437-444 Scott V E , Rettig J, Parcej D N , Keen JN, Findlay JB, Pongs O, Dolly JO. (1994) Primary structure of a beta subunit of alpha-dendrotoxin-sensitive K + channels from bovine brain. Proc Natl Acad Sci U S A . 91(5): 1637-41. Sewing, S, Roeper, J, Pongs, O, (1996) Kv (3-Subunit Binding Specific for Shaker-Related Potassium Channel a-Subunits. Neuron 16(2):455-463 Shen N V , Pfaffinger PJ. (1995) Molecular recognition and assembly sequences involved in the subfamily-specific assembly of voltage-gated K + channel subunit proteins. Neuron. 14(3):625-33. Shi G, Nakahira K, Hammond S, Rhodes KJ , Schechter L E , Trimmer JS. (1996) Beta subunits promote K + channel surface expression through effects early in biosynthesis. Neuron. 16(4):843-52. Snyders DJ. (1999) Structure and function of cardiac potassium channels. Cardiovasc Res. 42(2):377-90. Snyders DJ, Tamkun M M , Bennett PB. (1993) A rapidly activating and slowly inactivating potassium channel cloned from human heart. Functional analysis after stable mammalian cell culture expression. J Gen Physiol. 101(4):513-43. Sobko A, Peretz A, Attali B. (1998) Constitutive activation of delayed-rectifier potassium channels by a src family tyrosine kinase in Schwann cells. E M B O J. 17(16):4723-34. Soliven B, Ma L, Bae H, Attali B, Sobko A, Iwase T. (2003) PDGF upregulates delayed rectifier via Src family kinases and sphingosine kinase in oligodendroglial progenitors. A m J Physiol Cell Physiol. 284(l):C85-93. Sorkin A.(2004) Cargo recognition during clathrin-mediated endocytosis: a team effort. Curr Opin Cell Biol. 16(4):392-9. Steidl JV, Yool AJ . (1999) Differential sensitivity of voltage-gated potassium channels Kvl .5 and K v l . 2 to acidic pH and molecular identification of pH sensor. Mol Pharmacol. 55(5):812-20. Takimoto K, Fomina AF, Gealy R, Trimmer JS, Levitan ES. (1993) Dexamethasone rapidly induces Kvl .5 K + channel gene transcription and expression in clonal pituitary cells. Neuron. ll(2):359-69. 45 Tamkun M M , Knoth K M , Walbridge JA, Kroemer H , Roden D M , Glover D M . (1991) Molecular cloning and characterization of two voltage-gated K + channel cDNAs from human ventricle. FASEB J. 1991 Mar l;5(3):331-7. Tammaro P., Smith, A L , Hutchings, SR, Smirnov, SV (2004) Pharmacological evidence for a key role of voltage-gated K + channels in the function of rat aortic smooth muscle cells. Br. J. of Pharm. 143:303-317. Timpe L C , Fantl WJ. (1994) Modulation of a voltage-activated potassium channel by peptide growth factor receptors. J Neurosci. 14(3 Pt 1): 1195-201. Thorne GD, Ishida Y , Paul RJ. (2004) Hypoxic vasorelaxation: Ca2+-dependent and Ca 2 + -independent mechanisms. Cell Calcium. 36(3-4):201-8. Trepanier-Boulay V , St-Michel C, Tremblay A , Fiset C. (2001) Gender-based differences in cardiac repolarization in mouse ventricle. Circ Res. 89(5):437-44. Tu L, Wang J, Helm A , Skach WR, Deutsch C. (2000) Transmembrane biogenesis of K v l . 3 . Biochemistry. 39(4):824-36. Uebele V N , England SK, Chaudhary A, Tamkun M M , Snyders DJ. (1996) Functional differences in Kvl .5 currents expressed in mammalian cell lines are due to the presence of endogenous Kv beta 2.1 subunits. J Biol Chem. 271(5):2406-12. Valverde P, Koren G. (1999). Purification and preliminary characterization of a cardiac Kvl .5 repressor element binding factor. Circ Res. 84(8):937-44. van der Velden H M W , van der Zee L, Wijffels M C , van Leuven C, Dorland R, Vos M A , Jongsma HJ, Allessie M A . (2000) Atrial fibrillation in the goat induces changes in monophasic action potential and mRNA expression of ion channels involved in repolarization. J Cardiovasc Electrophysiol. 11(11): 1262-9. Van Wagoner DR. (2000) Pharmacologic relevance of K(+)Channel remodeling in atrial fibrillation. J Mol Cell Cardiol. 32(10): 1763-6. Van Wagoner DR, Pond A L , Lamorgese M , Rossie SS, McCarthy P M , Nerbonne JM. (1999) Atrial L-type C a 2 + currents and human atrial fibrillation. Circ Res. 85(5):428-36. Van Wagoner DR, Pond A L , McCarthy P M , Trimmer JS, Nerbonne JM. (1997) Outward K + current densities and Kvl .5 expression are reduced in chronic human atrial fibrillation. Circ Res. 80(6):772-81. Vigdor-Alboim S, Rothman C, Braiman L, Bak A, Langzam L, Yosef O, Sterengarz B B , Nawrath H , Brodie C, Sampson SR. (1999) Discoordinate regulation of different K channels in cultured rat skeletal muscle by nerve growth factor. J Neurosci Res. 56(3):275-83. Wang Z, Eldstrom JR, Jantzi J, Moore ED, Fedida D. (2004) Increased focal Kv4.2 channel expression at the plasma membrane is the result of actin depolymerization. A m J Physiol Heart Circ Physiol. 286(2):H749-59. 46 Wang Z, Fermini B, Nattel S. (1993) Sustained depolarization-induced outward current in human atrial myocytes. Evidence for a novel delayed rectifier K + current similar to Kvl .5 cloned channel currents. Circ Res. 73(6): 1061-76. Wang Z, Kiehn J, Yang Q, Brown A M , Wible B A . (1996) Comparison of binding and block produced by alternatively spliced Kvbetal subunits. J Biol Chem. 271(45):28311-7. Watanabe H , Ma M , Washizuka T, Komura S, Yoshida T, Hosaka Y , Hatada K , Chinushi M , Yamamoto T, Watanabe K, Aizawa Y . (2003) Thyroid hormone regulates mRNA expression and currents of ion channels in rat atrium. Biochem Biophys Res Commun. 308(3):439-44. Webster SM, Del Camino D, Dekker JP, Yellen G. (2004) Intracellular gate opening in Shaker K + channels defined by high-affinity metal bridges. Nature. 428(6985):864-8. Wei A , Covarrubias M , Butler A , Baker K, Pak M , Salkoff L . (1990) K + current diversity is produced by an extended gene family conserved in Drosophila and mouse. Science. 1990 248(4955):599-603. Wible B A , Yang Q, Kuryshev Y A , Accili EA, Brown A M . (1998) Cloning and expression of a novel K + channel regulatory protein, KChAP. J Biol Chem. 273(19):11745-51. Williams CP, Hu N , Shen W, Mashburn A B , Murray KT. (2002) Modulation of the human Kvl .5 channel by protein kinase C activation: role of the Kvbetal.2 subunit. J Pharmacol Exp Ther. 302(2):545-50. Wu Y , Anderson M E . (2002) Reduced repolarization reserve in ventricular myocytes from female mice. Cardiovasc Res. 53(3):763-9. Wulff H , Knaus HG, Pennington M , Chandy K G . (2004) K + channel expression during B cell differentiation: implications for immunomodulation and autoimmunity. J Immunol. 173(2):776-86. Wyszynski M , Lin J, Rao A, Nigh E, Beggs A H , Craig A M , Sheng M . (1997) Competitive binding of alpha-actinin and calmodulin to the N M D A receptor. Nature. 385(6615):439-42. Yamashita T, Sekiguchi A , Iwasaki Y K , Sagara K , Iinuma H , Hatano S, Fu LT, Watanabe H. (2003) Circadian variation of cardiac K + channel gene expression. Circulation. 107(14): 1917-22. Yellen G, Jurman M E , Abramson T, MacKinnon R. (1991) Mutations affecting internal T E A blockade identify the probable pore-forming region of a K + channel. Science. 251(4996):939-42. Y u W, Xu J, L i M (1996) N A B domain is essential for the subunit assembly of both alpha-alpha and alpha-beta complexes of shaker-like potassium channels. Neuron. 16(2):441-53. Yuan H, Michelsen K , Schwappach B. (2003) 14-3-3 dimers probe the assembly status of multimeric membrane proteins. Curr Biol. 13(8):638-46. 47 Yue L, Feng J, Wang Z, Nattel S. (1999) Adrenergic control of the ultrarapid delayed rectifier current in canine atrial myocytes. J Physiol. 516 (Pt 2):385-98. Yue L , Wang Z, Rindt H , Nattel S. (2002) Molecular evidence for a role of Shaw (Kv3) potassium channel subunits in potassium currents of dog atrium. J Physiol. 527 Pt 3:467-78. Zerangue N , Schwappach B, Jan Y N , Jan L Y . (1999) A new ER trafficking signal regulates the subunit stoichiometry of plasma membrane K(ATP) channels. Neuron. 22(3):537-48. Zielenski J. (2000) Genotype and phenotype in cystic fibrosis. Respiration. 67(2): 117-33. 48 C H A P T E R 2: N-terminal P D Z Binding Domain in K v l Potassium Channels Jodene Eldstrom*, Kyle W. Doerksen, David F. Steele and David Fedida Department of Physiology, University of British Columbia, FEBS Letters 531 (2002) 529-537 Contributions by Jodene Eldstrom: - Responsible for project management, experimental design, some of the cloning, mutagenesis, and in vitro bind assays, all of the co-immunoprecipitations, transfections, stable cell line generation, deglycosylation experiments, immunocytochemistry, confocal imaging, data analysis, some of the figure preparation together with Dave Steele, writing and editing the manuscript. (Percent Contribution: 60%) 49 INTRODUCTION Ion channels are not distributed randomly across cellular surfaces. In fact, their specialized localization is frequently essential to cell function. In neurons, for example, Kv2.1 localizes to high-density clusters on the soma and proximal dendrites in the hippocampus (Lim et al., 2000), voltage-gated K + channels co-localize with Na + channels at the nodes of Ranvier (Black et al., 1990), and the N M D A receptor is found almost exclusively at the post-synaptic density (Cho et a l , 1992a). Implicated in the clustering of N M D A receptors (Kornau et al., 1995) and Shaker-type K + channels (Kim et al., 1995) is PSD95. PSD95, the prototypical mammalian PDZ-containing protein (Cho et al., 1992b) contains three of the domains in tandem. Originally identified as a 95 kDa protein highly enriched in the post-synaptic density, PSD95 binds N M D A receptors via their NR2 subunits (Kornau et al., 1995; Muller et al., 1996; Niethammer et al., 1996) and this binding is most likely responsible for the channels' post-synaptic clustering. Other channels localized by interaction with PDZ proteins include muscle N a + channels (Gee et al., 1998), inwardly rectifying K + channels (Cohen et al., 1996; Horio et al., 1997; Nehring et al., 2000) and Shaker-type channels such as Kvl .4 , the clustering of which in neurons may also be dependent on linkage to PSD95 (Arnold and Clapham, 1999). PDZ domains are present in a variety of dissimilar proteins and are among the most common protein motifs (Ponting et al., 1997). Consisting of 80-120 amino acids, they are involved in protein-protein interactions, mostly with components of the cytoskeleton and associated structures. Proteins as diverse as syntrophins, PSD95 and neuronal nitric oxide synthase all contain the motifs (Sheng and Sala, 2001). Most interactions with PDZ domains involve the C-terminal 4 amino acids of the interacting protein and a peptide-binding groove in the PDZ motif which ends with a conserved carboxylate binding loop (Doyle et al., 1996). The NR2 subunit of the N M D A receptor and Kv l .4 both bind PSD95 in this way. But other binding mechanisms have been described. A few proteins bind via PDZ-PDZ interactions. The neuronal 50 and muscle isoform of nitric oxide synthase, nNOS, binds the PDZ domains of PSD95 and oti-syntrophin (Brenman et al., 1996) by a mechanism that involves a P-finger that essentially mimics a canonical C-terminal PDZ-binding motif (Hillier et a l , 1999). Other proteins bind PDZs by as yet undetermined mechanisms. For example, the PDZ motif of the actinin-associated L I M protein, A L P , binds to the internal spectrin repeats of a-actinins (Xia et al., 1997; Vallenius et al., 2000). To further our understanding of PDZ domain interactions with ion channels, we have utilized archetypal PDZ domains from the protein PSD95, to investigate potential interactions with the C- and N-termini of a Shaker-type K + channel, K v l . 5 . We have found that, as with Kv l .4 , these domains bind a C-terminal fragment of the channel, very probably the C-terminal sequence ETDL. Significantly, we have also uncovered an interaction between the PDZ domains and the N-terminus of K v l . 5 . This second binding site localizes to the Kvl .5 TI domain and appears to be independent of P-finger involvement. This N-terminal binding is not restricted to Kv l .5 but occurs also with other Kvl-type channels. 51 M A T E R I A L S A N D M E T H O D S D N A constructs and site-directed mutagenesis D N A encoding the 240 amino acid Kvl .5 N-terminus and internal deletions were cloned in frame with the GST-tag into pET42 (Novagen). Sequence-confirmed PCR-derived segments encoding N-terminal fragments of K v l . l (aa 1-167), K v l . 2 (aa 1-64), Kvl .3 (a 1-182), Kv l .4 (aa 90-305) and Kv4.2 (aa 1-183) were similarly cloned into pET42. Deletion mutations in the Kvl .5 N-terminus were made by restriction digests or internal digestion followed by incubation with nuclease Bal31 for varying times. Digestions were stopped by addition of 20 mM E G T A and the D N A was ligated and recovered after transformation into Escherichia coli. A 412 amino acid N-terminal portion of PSD95, containing the protein's 3 PDZ domains was cloned in frame with the T7 tag into pET28a. ct-Actinin2, minus amino acids 1 through 11, was similarly cloned into pET42. For immunocytochemical detection, Kvl .5 and its truncation mutants were N -terminally T7 tagged in pcDNA3. Briefly, Kv l .5 was subcloned as a Hindlll-NotI fragment into pET28-a. The tagged channel was then recovered by digesting the resultant plasmid with Ndel plus NotI and cloning the T7-tag-Kvl.5 fragment into EcoRV-Notl-digested pcDNA3. In all cases, the presence of in-frame fusions with the glutathione-S-transferase (GST) tag or T7 tag was confirmed by D N A sequencing. Site-directed mutagenesis was performed using Stratagene's Quikchange kit with appropriate primers. Deletion mutations were produced using PCR-based strategies. The presence of the targeted mutations was confirmed by D N A sequencing. GFP-tagged PSD95 was made by subcloning the 2.1 kb PSD95 SacII-EcoRI fragment from a clone in pGEX-2T into Hindlll-EcoRI digested pGFP (Grabner et al., 1998). Kv l .5 deletion mutants for expression in HEK293 cells were made by deletion of the Ncol-Ncol fragment of Kvl .5 (Kvl.5AN209) or by deletion of the coding sequence downstream of the internal BamHI site of Kvl .5 (Kvl.5AC51). The former lacks the first 209 amino acids of the channel; the latter lacks 52 amino acids 563-613. K v l . 5 A E T D L was produced by P C R using primers such that the C -terminal E T D L sequence was specifically deleted. Preparation of GST- and T7-tagged proteins Recombinant proteins were expressed in E. coli strain BL21(DE3) and purified using BugBuster Extraction Reagent (Novagen). Proteins detected in the soluble fraction by Coomassie staining of S D S - P A G E gels were purified using a T7-tag affinity purification kit (Novagen) or a G S T bind resin and buffer kit (Novagen), as appropriate. Proteins expressed and appearing in the insoluble fraction were washed and solubilized using Novagen's solubilization buffer containing 0.3% N-lauroylsarcosine according to the company's recommendations. The solubilized proteins were then dialyzed against 20 m M Tr i sHCl to remove residual detergent. In vitro binding assays Approximately 2 mg (normalized by comparison to standards on Coomassie-stained S D S - P A G E gels) of GST, G S T N-terminus of K v l . 5 G S T - C-terminus of K v l . 5 , other G S T K v l . 5 N-terminal and actinin fragments were individually combined with 2 mg of aT7 PSD95 P D Z domain construct comprising amino acids 1-412 of PSD95 in binding buffer (2mM T r i s H C l , p H 7.5, 150mM N a C l , 1 m M E D T A , 0.5 m M dithiothreitol, 0.1% Triton X-100) (Galliano et al., 2000). The mixtures were incubated at room temperature for 1 hour with periodic mixing. Glutathione-sepharose beads 4B (Amersham Pharmacia) pre-washed in binding buffer were added to each tube and incubated with mixing for 30 minutes at room temperature. The mixtures were spun for 5 minutes at 1000 rpm in a microcentrifuge and the pelleted beads were washed and repelleted four times in wash buffer (25 m M T r i s H C l , p H 7.5, 1 m M E D T A , 0.5 m M dithiothreitol) (Galliano et al., 2000). The pelleted bead-protein complexes were boiled in SDS sample buffer for five minutes and aliquots containing 0.2 mg of the G S T fusion were then resolved by S D S - P A G E . The proteins were transferred to P V D F membranes and probed as Western blots with horseradish peroxidase-labeled monoclonal anti-T7 antibodies (Novagen). 53 Antibody binding was detected using a chemiluminescent reagent (Renaissance, New England Nuclear). To check for equivalent loading of constructs, identical quantities of GST fusion proteins were subjected to PAGE on a separate gel and stained with Coomassie blue. Yeast two-hybrid experiments The K v l channel N-terminal fragments and the PSD95 fragment were cloned into pGAD424 and p G B D - C l , respectively. yff-Galactosidase assays were conducted in Y190. P-Galactosidase activity was measured spectrophotometrically at OD420, were 1 P-gal unit = 1000 X OD420/[OD600 X time (h) X volume of initial culture used (ml)]. A l l assays were performed on two to three different yeast transformants on several different experimental days. Co-immunoprecipitation HEK293 cells were co-transfected with individual T7-tagged Kvl .5 clones and a GFP-tagged PSD95 clone. Supernatants from cell extracts were made as previously reported (Maruoka et al., 2000) with the exception that 0.5 % OGEPAL CA-630 (Sigma) was used in place of Triton X-100 in the Lysis buffer. The supernatants were precleared with twenty ml of pre-swelled Sepharose CL4B in bead buffer (20 m M HEPES, pH 7.4, 5% glycerol, 100 mM NaCl, 0.1 m M EDTA), then transferred to fresh tubes. Five to ten ml of the appropriate antibody (anti-T7, Novagen, or anti-GFP, Torrey Pines Biolabs; buffer in control) was added to each tube and incubated on ice for 1 h with mixing. 15 ml pre-swelled protein A-sepharose (Sigma) was added to each tube in bead buffer and incubated on ice for one hour with mixing. The mixtures were spun 15 seconds in the microfuge and the pelleted beads were washed two times each with wash buffer (10 mM TrisHCl, 140 mM NaCl, 0.1% Triton X-100), once with Tris-saline (10 mM TrisHCl, 140 mM NaCl), then once with 50 mM TrisHCl, pH 6.8). The pellets were then boiled in 20 ml SDS-PAGE loading buffer and run on SDS-PAGE. After transfer to PVDF, membranes were probed with either HRP conjugated mouse anti-T7 antibody (Novagen) or with rabbit anti-GFP primary antibody (Torrey Pines Biolabs) and HRP-conjugated goat anti-rabbit IgG (Jackson 54 Laboratories). Antibody binding was detected using a chemiluminescent reagent (Renaissance, New England Nuclear). Deglycosylation experiments T7-tagged Kvl .5 was immunoprecipitated as above. After the final wash with 50 mM Tris, samples were resuspended in Denaturing Buffer (5% SDS. 10% P-mercaptoethanol) and boiled for 10 minutes. To one aliquot, 1/10 Volume of EndoH buffer (0.5M sodium citrate, pH5.5) and 1500 U Endoglycosidase H (New England Biolabs) was added. 1/10 volume PNGaseF buffer (0.5M sodium phosphate, pH7.5), 1/10 volume of 10% NP-40 and 1500U Peptide-N-glycosidase F (New England Biolabs) was added to a second aliquot. A third aliquot was left untreated. Samples were incubated for 1 hour at 37° C then subjected to Western Blot analysis. Imaging Stable cell lines were generated from HEK293 cells transfected with appropriate T7-tagged Kv l .5 constructs in pcDNA3 using LipofectAMINE™ 2000 (Invitrogen, Carlsbad, CA). Three days after transfection, 0.5 mg/mL geneticin was added to the growth media and after 10 days the cells were tested for expression of the tagged full length and truncated channel proteins. For co-expression studies, the stable Kvl .5 lines were transiently transfected with full length PSD95 in pcDNA3 and incubated for 24 hours prior to fixation. The cells were rinsed and fixed with 4% paraformaldehyde for 12 minutes at room temperature (RT). After three 5-minute washes with I X phosphate buffered saline (PBS; 137 mM NaCl, 2.7 mM KC1, 4.3 mM Na 2 HP0 4 , 1.4 mM K H 2 P 0 4 ; pH 7.2), cells were incubated in a Blocking Solution (PBS containing 2% B S A and 0.2% Triton X-100) for 30 minutes at RT. A mouse monoclonal antibody to the T7 Tag (1:1000; Novagen) or a rabbit polyclonal anti-PSD95 (1:500; Zymed Laboratories) was diluted in Blocking Solution and incubated at 4°C overnight or for 2 hours at RT. Cells were then washed three times, 5 minutes in PBS on a rotator before incubation with 55 secondary antibodies, Alexa 594 conjugated goat anti-mouse IgG antibody and Alexa 488 conjugated goat anti-rabbit IgG antibody (1:1000; Molecular Probes) for 1 hour on the rotator at RT. Cover slips were once again washed three times with PBS prior to mounting with 10 ul of a 90% glycerol, 2.5% w/v DABCO-PBS solution. Images of labeled cells were taken using a Bio-Rad Radiance Plus on an inverted Zeiss Axiovert microscope using BioRad LaserSharp 2000 software. Images were later viewed and prepared using NIH Image and PhotoShop software packages. Electrophysiological procedures Stable line HEK-293 cells were transfected with either pGFP or PSD95:pGFP and the experimenter was blinded to transfection cDNA group. Coverslips containing cells were removed from the incubator before experiments and placed in a superfusion chamber (volume 250 ul) containing the control bath solution at ambient temperature (22-23 °C), and perfused with bathing solution throughout the experiments. Transfected cells were selected using fluorescence microscopy. Whole-cell current recording and data analysis were done using an Axopatch 200A amplifier and pClamp 8 software (Axon Instruments, Foster City, CA). Patch electrodes were fabricated using thin-walled borosilicate glass (World Precision Instruments; FL, USA). Electrodes had resistances of 1-3 M Q when filled with control filling solution. Capacity compensation and 80% series resistance compensation were used in all whole-cell recordings. No leak subtraction was used when recording currents, and zero current levels are denoted by the dotted lines in the current tracings in Figure 2.6. Data were sampled at 10-20 kHz and filtered at 5-10 kHz. Membrane potentials have not been corrected for small junctional potentials between bath and pipette solutions. Patch pipettes contained (in mM): NaCl, 5; KC1, 135; Na2ATP, 4; GTP, 0.1; M g C l 2 , 1; EGTA, 5; HEPES, 10; and was adjusted to pH 7.2 with K O H . The bath solution contained (in mM): NaCl, 135; KC1, 5; HEPES, 10; sodium acetate, 2.8; M g C l 2 , 1; CaCl 2 , 1; and was adjusted to pH 7.4 with NaOH. A l l chemicals were from Sigma Aldrich 56 Chemical Co. (Mississauga, Ont.). Data are presented as mean ± SEM. Statistical significance was determined using a two-tailed Student's T-test. R E S U L T S P D Z domains bind both Kv l .5 C- and N-termini The C-termini of many Shaker-type K + channels are known to bind to the PDZ-domains of proteins like PSD95 and SAP97. Although hKvl.5 includes a canonical sequence for PDZ-binding at its extreme C-terminus, an interaction between the two proteins has never been directly tested. To confirm the interaction specifically with K v l . 5 , we assayed the binding of Kv l .5 fragments with the PDZ-containing core of PSD95. GST-tagged fragments of Kvl .5 and a T7-tagged N-terminal fragment of PSD95 that contained all three of the protein's PDZ-domains, were expressed in E. coli using the pET system. After purification, the C-terminal (amino acids 535-613) and N-terminal (amino acids 1-240) fragments of Kv l .5 were tested for in vitro binding to the PSD95 fragment. Negative controls were included to ensure that the fragments were binding specifically to the tested partners and not to GST or to glutathione-sepharose. GST-tagged a-actinin2, previously shown to bind the PSD95 fragment (Grace Lu, unpublished observations), was included as a positive control. Glutathione-sepharose beads were added and, after extensive washing, the protein complexes were pelleted and run on SDS-PAGE for Western analysis. As expected, a-actinin2 bound the PDZ domains, as did the Kvl .5 C-terminus (Figure 2.1 A). GST or glutathione-sepharose alone failed to bring down the fragment. Surprisingly, however, the Kvl .5 N-terminus also bound the PDZ construct, and it did so with apparently similar avidity to that detected with the positive controls. This finding of N-terminal binding to the PDZ construct was in contrast to the many published reports that deletion or mutation of the extreme C-termini of Kv channels eliminates PDZ-binding or its effects (Kim et al., 1995; Kim and Sheng, 1996; Arnold et al., 57 A T 7 - P D Z 1,2,3 Figure 2.1. PDZ domains bind N-termini of K v l channels. A . PDZ Domains of PSD95 Bind Kv l .5 C- and N-termini. Purified T7-tagged PSD95 N -terminus was incubated alone or with GST-tagged a-actinin2, Kv l .5 N-terminus, Kv l .5 C-terminus, or GST-tag alone. Glutathione-sepharose beads were added and following further incubation, the beads were pelleted, washed extensively, and subjected to SDS-PAGE. Western blots were performed using anti-T7 antibody to detect the PSD95 fragment. GS refers to PSD95 incubated with glutathione sepharose in the absence of a GST-fusion protein. B. PSD95 Binds Other K v l Channels but not Kv4.2. Experimental procedures were similar to A . T7-tagged PSD95 N-terminus was incubated with GST-tagged N-termini of K v l . l , K v l . 2 , K v l . 3 , K v l . 4 , K v l . 5 and Kv4.2 prior to pelleting with glutathione sepharose and visualization by Western blot. 58 1999; Murata et al., 2001). Most of these reports concern Kvl.4-PSD95 interactions, however, and only two (Kim and Sheng, 1996; Murata et al., 2001) included N-termini in their test systems. As far as we are aware, there have been no specific tests of Kvl.5-PSD95 interactions. To ascertain whether this N-terminal binding property of Kvl .5 to PSD95 was shared by other Shaker-type channels, the N-termini of additional Kv channels were tested for binding to the PDZ construct. GST-tagged N-terminal fragments of K v l . l , K v l . 2 , K v l . 3 , Kv l .4 (Cukovic et al., 2001) and Kv4.2 were expressed, purified and tested in our assay. Because of difficulty expressing the full N-terminus of Kvl .4 , the fragment expressed of this channel lacked the first 89 amino acids. This deletion leaves the great majority of the N-terminus intact, including the entirety of its N-terminal homology with the other tested Kv channels. As shown in Figure 2.IB, all K v l N-termini were found to interact with the PDZ domains except K v l . 4 which bound very weakly, i f at all. This latter finding is consistent with the previous reports that failed to detect a Kvl.4-PSD95 interaction outside the C-terminus (Kim et al., 1995; Kim and Sheng, 1996; Arnold et al., 1999; Murata et al., 2001). The N-terminus of Kv4.2 also did not bind the PDZ construct. Thus, at least as measured by our in vitro assay, N -terminal interactions with PDZ domains are common in K v l channels but are not general to all voltage-gated K + channels. That an interaction occurs between two proteins in an in vitro system does not guarantee that the binding reflects an in vivo process. Misfolding of one or both of the tested proteins, (effectively) inappropriate compartmentalization or a lack of alternative binding partners, for example, could all allow binding where no interaction occurs in vivo. We therefore chose to test the putative interactions in a variety of living systems. Yeast two-hybrid assays confirm interaction While the yeast two-hybrid system is subject to many of the same caveats as the in vitro binding assays, the method is commonly used in protein-protein interaction studies and it 59 provides a system in which the proteins are surrounded by a natural cellular milieu. It also provides a simple method to roughly compare the strengths of interactions between proteins (Brown et al., 1994). We therefore used this system to further investigate the interactions of Kv l .5 N-terminus with our PDZ construct and to test also whether other K v channel N-termini can also bind that construct. With the exception of Kvl .4 , the same fragments used in the in vitro binding assay were employed in vectors appropriate to the 2-hybrid system (pGAD424 for the K v channel N -termini, pGBD-Cl for the PDZ construct); for Kvl .4 , the full N-terminus was used rather than the truncated piece necessary in the in vitro experiments. Yeast strain Y190 was co-transformed with the constructs and the P-galactosidase activities of the transformants were measured. Y190 co-transformed with the PDZ construct and the C-terminus of Kv l .4 served as positive control. Confirming the results of the in vitro studies, these experiments provided strong evidence of an interaction between all of the K v l N-termini and the PDZ construct. In every case, the K v channel N-terminus was found to interact with the PDZ domains (Figure 2.2). Strikingly, the apparent strengths of the PDZ interactions with the each of K v l channel N-terminal fragments were comparable to that with the Kv l .4 C-terminus. This was surprising in the case of K v l . 4 since it bound poorly to the PDZ domains in our in vitro assay. Further studies with this channel will be necessary to resolve this discrepancy. PSD95 co-immunoprecipitates with C-terminally truncated Kvl.5 To determine whether PSD95 binds the N - and C-termini of Kvl .5 in vivo, co-immunoprecipitation experiments were performed in transfected HEK293 and COS cells. Cells were transfected with GFP-tagged PSD95 and one of two T7-tagged Kvl .5 constructs. One, the Kvl .5 wild-type, included the C-terminal ETDL sequence reportedly necessary for Kv channel binding to PDZ domains (Kim et al., 1995; Kim and Sheng, 1996; Arnold et al., 1999; Murata et al., 2001). In the other, Kv l . 5AETDL , these four amino acids were specifically deleted. 60 Figure 2.2. Yeast two-hybrid experiments demonstrate PDZ binding to Kv channel N- and C-termini. P-Galactosidase activities from yeast strain Y190 initiated by the interaction of K v l channel N -or C-termini with the PSD domains of PSD95. The K v channel fragments were expressed from the pGAD424 vector; PSD95 PDZ domains 1-3 were expressed from p G B D - C l . Representative examples of controls expressing K v channel or PDZ channel constructs alone are shown at right. 1 p-galactosidase unit = 1000 X OD420/[OD600 X time (h) X volume of initial culture used (ml)]. A l l data represent means +/- S E M for three experiments conducted with fresh transformants on separate experimental days. "Alone" refers to yeast co-transformed with the reported construct plus its empty vector partner for the yeast two-hybrid assay. 61 A s shown in Figures 2.3A and B , antibody to G F P (which specifically immunoprecipitates the tagged P D Z protein) pulled down comparable amounts of both the wi ld-type and A E T D L versions of K v l . 5 in H E K cells co-expressing PSD95 and the channel. Similarly, anti-T7 brought down PSD95 along with the K v l . 5 constructs to which the antibody was directed (data not shown). The K v l . 5 proteins were not pulled down by anti-GFP from cells expressing the channels and the G F P tag alone. Thus, the P D Z protein interacts with K v l . 5 irrespective of the presence of the canonical PDZ-binding motif at the channel's C-terminus. The N-terminal binding region identified in vitro very likely underlies this phenomenon. Interestingly, the antibody directed to the K v l . 5 protein consistently immunoprecipitated two isoforms of the channel, whether the channel expressed was the full length K v l . 5 or K v l . 5 A E T D L (Figures. 2.3C, D). The antibody directed to the PSD95 fusion protein also pulled down two K v l . 5 isoforms when the wild-type K v l . 5 channel was expressed with it. However, when the A E T D L version of the channel was co-expressed with the PSD95 construct, the PSD95-specific G F P antibody brought down only the smaller K v l . 5 band (Figure 2.3B). Similar results were observed in transfected COS-7 cells (data not shown), showing that this interaction is not a cell-specific phenomenon. Deglycosylation experiments using transfected H E K cells demonstrated that the migration difference between the two bands was due to differences in their degrees of glycosylation (Figure 2.3E). PNGase F treatment of the proteins collapses the two bands to a single lower band, suggesting that the upper band represents a mature glycoprotein. Consistent with this interpretation, EndoH treatment had no effect. It is likely that the glycosylation differences reflect differences between cell surface and internal channel pools (L i et al., 2000). PSD95 and K v l . 5 variants co-localize in cultured cells In order to further confirm the interactions in vivo, confocal imaging experiments were conducted. A s shown in Figure 2.4A, when transfected into H E K 2 9 3 cells, both wild-type K v l . 5 62 E Kv1.5 Figure 2.3. Co-immunoprecipitation of K v l . 5 variants with PSD95 from H E K and COS cells. A . Co-immunoprecipitation of K v l . 5 with PSD95. Extracts from T7-tagged K v l . 5 , GFP-tagged PSD95 double-expressing H E K cells were mixed without antibody (Lane 1) or with anti-GFP (Lane 2). The protein-antibody complexes were precipitated with Protein A-sepharose and subjected to Western analysis and probed with anti-T7 antibody. B . Co-immunoprecipitation o f K v l . 5 A E T D L with PSD95. The blot is identical to that in (A) except that the H E K cells expressed K v l . 5 A E T D L rather than wild-type K v l . 5 . The quantity of protein loaded, antibody used and exposure times were the same for both blots. C , D . Anti-T7 pulls down two forms of both K v l . 5 and K v l . 5 A E T D L . Extracts from T7-tagged K v l . 5 , GFP-tagged PSD95 (Lanes 1 and 2) or T7-tagged K v l . 5 A E T D L , GFP-tagged PSD95 (Lanes 3 and 4) double-expressing H E K cells were mixed without antibody (Lanes 1 and 3) or with anti-T7 (Lanes 2 and 4). E . Glycosylation differences underlie K v l . 5 doublet. K v l . 5 was immunoprecipitated with anti-T7 and then denatured by boiling. Test aliquots were then treated with either PNGase F or Endo H glycosidase and subjected to Western analysis. The blot was probed with anti-T7 antibody. 63 Channel P S D 9 5 Merged Figure 2.4. PSD95 colocalizes with K v l . 5 and K v l . 5 A E T D L in transfected HEK293 cells. Confocal images of single optical slices of HEK293 cells stably expressing either T7-tagged hKvl .5 (A) or T7 tagged Kv l .5AETDL (B) also transfected with PSD95. K v channels were detected with anti-T7 (red); PSD95 was detected with anti-PSD95 (green). Yellow indicates co-localization of the channel with PSD95 in the merged image. 64 and PSD95 localize to the cell surface, although a substantial portion of the Kvl .5 signal is found in the cytoplasm. When doubly transfected, the co-localization of the two proteins is substantial and restricted to the cell surface. Imaging experiments of PSD95 with K v l . 5 A E T D L in which the channel's C-terminal PDZ-binding motif is deleted, yielded nearly identical results (Figure 2.4B). As it did with the wild- type K v l . 5 , PSD95 co-localized with Kv l .5AETDL at the membrane in doubly transfected H E K cells. Thus, the extreme C-terminal amino acids of Kvl .5 are not essential to its association with PSD95. Delineation of the N-terminal PDZ-binding region in hKvl .5 Binding of PDZ domains to internal sites in target proteins is not common (Sheng et al., 2001). Thus, delineation of the PSD95 binding site in the Kvl .5 N-terminus could add to our understanding of these rare interactions. To roughly locate the region in the Kvl .5 N-terminus where the PDZ domains are binding, a number of deletion mutants were constructed. As illustrated in Figure 2.5A, GST-tagged deletion constructs lacking amino acids 2-92, 2-162, 85-209, 135-240 and 150-208 of the Kvl .5 N-terminus were produced and tested for binding with the T7-tagged PSD95 PDZ construct. Figure 2.5B illustrates that the PDZ domains bound separate Kvl .5 N-terminal deletion constructs lacking amino acids 2-92, and 150-208 but did not bind deletions of amino acids 2-161, 85-208, nor 135-240. This places the binding site in the Kvl .5 between amino acids 92 and 149, immediately adjacent to or within the T l domain of the channel. There are a number of superficially canonical PDZ-binding motifs within the T l domain of Shaker-type channels. Were the sequences located at the C-terminus of the channel, ETQL at positions 132-135 and ISGL spanning residues 126-129 would be excellent candidates for Type I PDZ-binding sites. An INI sequence at 124-126 would be a similarly good candidate binding site for Type II PDZ-domains (Sheng et al., 2001). A l l three reside in a P-finger-like structure folded tightly into the T l domain of the channel (Kreusch et al., 1998). Hillier et al., 65 Binds PDZ? 220 240 N-termHius 1-134 1-149 93-240 163-240 A85-209 +++ A150-209 120 130 140 150 160 QR V H INTSGJ R FE TQ1 XrV,. A Q F F N T 1 I G D P A K R l . R Y F D F l R K F Y F F D R N R P 170 180 190 200 210 220 P S F D C H L Y Y Y Q S C X J R I J I ^ T7-PDZ 1,2,3 Figure 2.5. Analysis of PDZ domain binding to K v l . 5 N-terminal deletion constructs. A , Schematic diagram of the various Kvl .5 constructs. The binding activity of the constructs to PSD95 is indicated in the column on the right. B, Representative Western blots showing PDZ domain-binding by the various constructs. GST-tagged Kv l .5 N-terminal deletion constructs were incubated separately with PSD95 N-terminus and glutathione sepharose beads were added. Following incubation, the beads were pelleted, washed extensively, then subjected to SDS-P A G E . Western blots were performed using monoclonal anti-T7 (Novagen, Madison, WI) to visualize the PSD95 fragment. 66 (Hillier et al., 1999) have reported PDZ-binding to a pseudocanonical domain at the end of a P-finger in nNOS. Only the INI sequence is located analogously to the nNOS motif in the T I P-finger-like structure, but it is oriented towards the interior of the domain. E T Q L is oriented externally but is located at the base of the finger-like domain. The S G L sequence traverses the tip of the P-finger and, without significant divergence from the published crystal structure, could not interact with the P D Z groove. On the slight possibility that INI or E T Q L might be involved in K v l . 5 - P S D 9 5 binding, they were modified by site-directed mutagenesis. Mutation of INI to I N D and E T Q L to A A Q A both failed, however, to affect the binding of the PSD95 fragment to the K v l . 5 N-terminus (data not shown). PSD95 influences K v l . 5 K + currents The effects of PSD95 co-expression with K v l . 5 on potassium current density were investigated in H E K cells. A PSD95-GFP fusion construct was transfected into HEK293 cell lines stably expressing K v l . 5 or one of two K v l . 5 deletion mutants in which either the N - or C -terminus of the channel was removed. A s illustrated in Figure 2.6 (A, D) , PSD95 had no effect on K v l . 5 currents when the intact channel was expressed. However, PSD95 had profound effects on currents carried by the K v l . 5 deletion mutants. Currents in H E K cells expressing both PSD95 and an N-terminal deletion mutant of K v l . 5 , K v l . 5 A N 2 0 9 , were increased at least three-fold over those in cells expressing the K v l . 5 mutant alone (Figure 2.6 B,E) . Currents from some cells co-expressing PSD95 and K v l . 5 A N 2 0 9 were so large that they could not be clamped. Thus, the average values obtained for the peak currents of these cells are underestimates of the actual peak values, because such large currents were omitted from the analysis. N o cells expressing K v l . 5 A N 2 0 9 without PSD95 exhibited currents that could not be clamped. Similar experiments in which a K v l . 5 C-terminally deleted channel (lacking the C-terminal 51 amino acids) was co-expressed with PSD95 yielded very different results. Instead of 67 Kv1.5 pGFP PSD95 B Kv1.5AN209 pGFP PSD95 DpGFP 10nA | P S D 9 5 5nA OnA 1 DpGFP 10nA | P S D 9 5 5nA OnA f -— r jC Kv1.5AC51 pGFP PSD95 ILL F 15nA DpGFP 10nA | PSD95 ~ h -60 -10 +20 +60 Voltage Step (mV) -60 -20 +20 +60 Voltage Step (mV) -60 -10 0 +20 Voltage Step (mV) Figure 2.6. Effect of PSD95 expression on peak currents on W T K v l . 5 , and C - or N -terminal deletion mutants. H E K - 2 9 3 cells stably expressing K v l . 5 , K v l . 5 A C 5 1 , and K v l . 5 A N 2 0 9 were used to test the effect of PSD95 on peak currents when the N - and C-terminal P D Z binding domains were removed. Cells were transiently transfected with either pGFP or PSD95:pGFP and peak currents at several voltages were measured under voltage-clamp in the whole-cell configuration. The protocol consisted of four 75 ms pulses (sufficient to allow activating currents to reach steady-state) with an inter-sweep interval of 5 s. The voltage-steps were as follows: the first below the activation threshold (-60 m V ) , the second near the Vy 2-activation of each channel (-10 m V ) , and the third and forth above the maximum activation where the I-V relationship is linear (+20, +60 m V ) . Representative current traces and mean data are shown for W T K v l . 5 (A, D) , K v l . 5 A N 2 0 9 (B, E) , and K v l . 5 A C 5 1 (C, F) . Bar graph data are presented as mean data ± S E M from n=6-8 cells. The asterisks denote statistical significance with p<0.1 (*) or p<0.05 (**) . (Data collected and analyzed by K . Doerksen) 68 increasing potassium currents, PSD95 substantially reduced those currents (Figure 2.6 C, F). Potassium currents in the Kvl.5AC51 mutant cell line are unusually large and PSD95 attenuates those currents. Peak current measurements above -20 mV could not be made since 67% of the Kvl.5AC51/PSD95 co-expressing cells and 100% of the cells expressing the Kvl.5AC51 mutant alone exhibited currents too large to clamp. Attempts to perform the same experiment using H E K cells transiently expressing the shorter C-terminal deletion mutant Kv l .5AETDL +/- the PSD95-GFP fusion similarly yielded large currents although the general trend was similar (data not shown). DISCUSSION The present study demonstrates that the prototypical PDZ domains of PSD95 bind to both C- and N-terminal fragments of Kvl .5 and other K v l channels and that these interactions affect Kvl .5 potassium currents. While, based on the sequence of the extreme C-terminus of the channel, the interaction of the two proteins was expected, the involvement of the N-terminus was wholly unpredicted. Nevertheless, the evidence for the N-terminal interaction is strong. The interaction is readily detected with in vitro binding assays, by yeast 2-hybrid analysis, and by co-immunoprecipitation (Figures 2.1-2.3). Imaging analysis shows that both wild-type Kvl .5 and the C-terminal mutant co-localize with PSD95 at or very near the cell surface (Figure 2.4) and the relative amounts co-immunoprecipitated and the yeast 2-hybrid results further show that the N - and C-terminal interactions occur with similar avidity. Finally, PSD95 has differential effects on Kvl .5 potassium currents depending on the individual presence of the channel's two PDZ-binding sites (Figure 2.6). While PDZ-domain binding to internal motifs is uncommon, it is not unprecedented. Internal PDZ binding sites have been demonstrated in phospholipase C (Xian-Zhong et al., 1998; Van Huizen et al., 1998), protein kinase C (Xian-Zhong et al., 1998; Van Huizen et al., 1998), 69 and nNOS, among others (Hillier et al., 1999). The best characterized of these interactions occurs between a P-finger of nNOS and the PDZ domains of al-syntrophin (Hillier et al., 1999). In this interaction, the first strand of the p-finger mimics a canonical COOH-peptide ligand. The sharp turn of the p-fmger allows an internal ETTF sequence at the turn to slip into the binding groove of the syntrophin PDZ domain. Interestingly, this P-fmger itself is part of a PDZ domain in nNOS. There are no PDZ domains in Shaker-type, channels. And mutation of consensus PDZ-binding sequences within the one potential P-finger in T l did not affect PSD95 binding. The interaction described herein must, therefore, occur via some other mechanism. The mechanisms by which the Drosophila PDZ-protein InaD binds internally to protein kinase C and to phospholipase C-P are also unknown (Xian-Zhong et al., 1998; Van Huizen et al., 1998) as is that between the mouse proteins PTP-BL and RIL (Cuppen et al., 1998). The latter, like the Kvl.5-N-terminal-PSD95 interaction, does not involve PDZ-PDZ interactions. Neither does it appear to involve consensus PDZ-binding sequences in P- or zinc-fingers (Cuppen et al., 1998). Previous work has established a role in subcellular localization for the C-terminal binding motif, and shown that the C-terminal site is in itself necessary and sufficient for that role. Deletion or mutation of the canonical C-terminal motifs abolishes PSD95-mediated clustering interactions with K v l channels (Kim et al., 1995; K im and Sheng, 1996; Arnold et al., 1999; Tiffany et al., 2000). A similar deletion in Kv4.2 eliminates clustering of that channel and prevents an apparent PSD95-mediated increase in its surface expression (Wong et al., 2002). We were unable to detect any focal clustering of Kvl .5 by PSD95 either for the wild-type or AETDL forms of the channel (Figure 2.4). This was true both in the HEK293 cells shown herein and in COS-7 cells (data not shown). While clustering was not evident, co-expression with PSD95 greatly increased Kvl .5 potassium currents in transfected H E K cells - but only when the N-terminus of the channel was deleted. Kvl.5AN209 is normally difficult to express 70 and currents tend to be small with this mutant. Consequently, the increase in current magnitudes to greater than typically seen for the wild-type channel is quite striking. One possible explanation for this phenomenon is that the mutant traffics to the cell surface in much the same manner as does the wild-type channel. Lacking an N-terminus, however, it is not stable there. Perhaps N -terminal interactions with actinin (Maruoka et al., 2000) or other cellular constituents are necessary to maintain the channel at the surface of the H E K cells. Addition of PSD95 to the system might allow its interaction with the Kvl .5 C-terminus to perform a similar function, stabilizing an otherwise temporary surface expression of the channel. This scenario would be consistent with the finding that PSD95 can stabilize but not promote cell surface expression of other K v l channels (Tiffany et al., 2000). That the wild-type channel is not affected by co-expression with PSD95 suggests that wild-type surface retention is maximized via the N-terminal interactions, i f this model is correct. The discovery of a PDZ-binding region in the N-termini of K v l channels is more difficult to reconcile with previous work. It is highly unlikely that the PDZ-Kv channel N-terminal interaction can be important to clustering or to the promotion of cell surface expression. Too much data exists that the C-terminus alone is responsible for these effects (Kim et al., 1995; Kim and Sheng, 1996; Arnold et al., 1999; Tiffany et al., 2000; Wong et al., 2002). The role of the N -terminal interaction must be quite different. We have two pieces of evidence for such a divergent role. Co-immunoprecipitation experiments give different results in the presence and absence of the Kv channel ETDL C-terminal binding motif (Figure 2.3) and PSD95 co-expression has opposing effects on Kvl .5 currents depending on whether the expressed channel retains the C- or the N-terminal PDZ-binding domain (Figure 2.6). Two isoforms of Kvl .5 are routinely seen in Western blots of transfected H E K cell extracts and antibody to the Kvl .5 fusion consistently pulls down both of these isoforms (Figure 2.3). This is true whether the transfected channel is wild-type or AETDL. If the expressed channel is wild-type, the same two isoforms consistently 71 coimmunoprecipitate with PSD95 when antibody against the PDZ protein is used. However, only the faster migrating isoform of Kvl .5 coimmunoprecipitates with PSD95 if the expressed channel lacks the C-terminal ETDL sequence. Deglycosylation experiments showed that this faster migrating Kvl .5 is a less glycosylated, possibly internal form of the channel (Figure 2.3E). The differential binding of PSD95 to the two isoforms definitely suggests functional differences between N - and C-terminal PDZ-binding. Perhaps PSD95 and/or other PDZ proteins might retain misfolded channels via the N-terminal interaction or influence the trafficking of the channels. Effects on Kvl .5 potassium currents provide more evidence that PSD95-binding to the N-terminus of the channel has a very different role than does C-terminal binding. Whereas co-expression of PSD95 with an N-terminally deleted Kvl .5 mutant dramatically increased potassium currents, the effect of coexpression with a C-terminally deleted Kvl .5 mutant was quite different. Current levels were significantly reduced by this experimental manipulation (Figure 2.6). It would seem that PSD95 interaction with the N-terminus somehow indeed interferes with surface expression of the Kvl .5 channel. How could one protein, normally expressed at the cell membrane, have such divergent effects on the expression of another membrane protein? In this artificial system, both PSD95 and Kvl .5 are over-expressed. Both can be detected in the cytoplasm as well as at the cell surface (Figure 2.4). Thus, there is at least an opportunity for internal interactions that may or may not occur in neurons or other constitutively expressing cells. The N-termini of K v channels interact with a large number of proteins. P-Subunits, KChIP, KChAP and a-actinin2 are among the proteins known to bind at or near the PDZ binding site uncovered here. In such a busy area, it is quite possible that the PDZ-binding site is normally masked from PSD95 at the cell surface. It is conceivable even that PSD95 in our artificial system is taking on the role of some other PDZ 72 protein(s) and interacting with an N-terminal Kv channel domain that PSD95 normally fails to see. A plethora of PDZ proteins have been described (see Bezprozvanny and Maximov, 2001, for a partial list) and, thus, there are many potential binding partners for these channel regions. Some obvious candidates in neurons include SAP 102 and PSD93 (Craven and Bredt, 1998). ZASP (Faulkner et al., 1999) and E N H (Nakagawa et al., 2000) are among many potential interactors in the heart. SAP97 is also expressed in the heart and has been shown to interact with various K v l channels, where it clusters the channels in the interior of the cell (Kim and Sheng., 1996; Tiffany et al., 2000; Murata et al., 2001). Mutations in the K v l . 4 C-terminal PDZ binding motif eliminate this co-clustering phenotype, however (Kim and Sheng, 1996; Tiffany et al., 2000; Murata et al., 2001). The K v channel N-terminal PDZ binding site is within or very near the TI domain, a region first identified as necessary for the assembly of the channel subunits themselves (Shen et al., 1993). Binding to the TI domain could thus affect channel assembly, intersubunit interactions, or even channel gating (Cushman et al., 2000). Alternatively, PDZ binding might compete with the binding of other cellular constituents, such as the P-subunits or ot-actinin2, perhaps affecting channel assembly, folding, trafficking or inactivation. It will be of great interest to learn which, if any, of these roles PDZ proteins may play. 73 ACKNOWLEDGEMENTS We are grateful to Dr. Terry Hebert (Universite de Montreal) and Dr. Barry Johnson (Cardiome Pharma, Inc.) for the PSD95 clones. We thank Grace Lu and Daniela Cukovic for excellent technical assistance. REFERENCES Arnold, D.B., and Clapham, D.E. (1999). Molecular determinants for subcellular localization of PSD-95 with an interacting K + channel. Neuron 23, 149-157. Bezprozvanny, I., and Maximov, A . (2001). Classification of PDZ domains. FEBS Lett 509, 457-462. Black, J.A., Kocsis, J.D., and Waxman, S.G. (1990). Ion channel organization of the myelinated fiber. Trends Neurosci 13, 48-54. Brenman, J.E., Chao, D.S., Gee, S.H., McGee, A.W., Craven, S.E., Santillano, D.R., Wu, Z., Huang, F., Xia, H. , Peters, M.F., Froehner, S.C., and Bredt, D.S. (1996). Interaction of nitric oxide synthase with the postsynaptic density protein PSD-95 and alpha 1-syntrophin mediated by PDZ domains. Cell 84,151-161. Brown, N.G. , Costanzo, M.C. , and Fox, T.D. (1994). Interactions among three proteins that specifically activate translation of the mitochondrial COX3 mRNA in Saccharomyces cerevisiae. Mol Cell Biol 14, 1045-1053. Cho, K.O. , Hunt, C.A., and Kennedy, M.B. (1992). The rat brain postsynaptic density fraction contains a homolog of the Drosophila discs-large tumor suppressor protein. Neuron 9, 929-942. Cohen, N.A. , Brenman, J.E., Snyder, S.H., and Bredt, D.S. (1996). Binding of the inward rectifier K + channel Kir 2.3 to PSD-95 is regulated by protein kinase A phosphorylation. Neuron 17, 759-767. Craven, S.E., and Bredt, D.S. (1998). PDZ proteins organize synaptic signaling pathways. Cell 93, 495-498. Cukovic, D., Lu, G.W., Wible, B. , Steele, D.F., and Fedida, D. (2001). A discrete amino terminal domain of Kvl .5 and Kv l .4 potassium channels interacts with the spectrin repeats of alpha-actinin-2. FEBS Lett 498, 87-92. Cuppen, E., Gerrits, H. , Pepers, B. , Wieringa, B. , and Hendriks, W. (1998). PDZ motifs in PTP-B L and RIL bind to internal protein segments in the L I M domain protein RIL. Mol Biol Cell 9, 671-683. Cushman, S.J., Nanao, M.H. , Jahng, A.W., DeRubeis, D., Choe, S., and Pfaffinger, P.J. (2000). Voltage dependent activation of potassium channels is coupled to T l domain structure. Nat Struct Biol 7, 403-407. 74 Doyle, D.A., Lee, A. , Lewis, J., Kim, E., Sheng, M . , and MacKinnon, R. (1996). Crystal structures of a complexed and peptide-free membrane protein-binding domain: molecular basis of peptide recognition by PDZ. Cell 85, 1067-1076. Faulkner, G., Pallavicini, A. , Formentin, E., Comelli, A. , Ievolella, C , Trevisan, S., Bortoletto, G., Scannapieco, P., Salamon, M . , Mouly, V. , Valle, G., and Lanfranchi, G. (1999). ZASP: a new Z-band alternatively spliced PDZ-motif protein. J Cell Biol 146, 465-475. Galliano, M.F., Huet, C , Frygelius, J., Polgren, A. , Wewer, U . M . , and Engvall, E. (2000). Binding of A D A M 12, a marker of skeletal muscle regeneration, to the muscle-specific actin-binding protein, alpha -actinin-2, is required for myoblast fusion. J Biol Chem 275, 13933-13939. Gee, S.H., Madhavan, R., Levinson, S.R., Caldwell, J.H., Sealock, R., and Froehner, S.C. (1998). Interaction of muscle and brain sodium channels with multiple members of the syntrophin family of dystrophin-associated proteins. JNeurosci 18, 128-137. Grabner, M . , Dirksen, R.T., and Beam, K . G . (1998). Tagging with green fluorescent protein reveals a distinct subcellular distribution of L-type and non-L-type C a 2 + channels expressed in dysgenic myotubes. Proc Natl Acad Sci U S A 95, 1903-1908. Hillier, B.J., Christopherson, K.S., Prehoda, K.E . , Bredt, D.S., and Lim, W.A. (1999). Unexpected modes of PDZ domain scaffolding revealed by structure of nNOS-syntrophin complex. Science 284, 812-815. Horio, Y . , Hibino, FL, Inanobe, A. , Yamada, M . , Ishii, M . , Tada, Y . , Satoh, E., Hata, Y . , Takai, Y . , and Kurachi, Y . (1997). Clustering and enhanced activity of an inwardly rectifying potassium channel, Kir4.1, by an anchoring protein, PSD-95/SAP90. J Biol Chem 272, 12885-12888. Kim, E., Niethammer, M . , Rothschild, A. , Jan, Y . N . , and Sheng, M . (1995). Clustering of Shaker-type K + channels by interaction with a family of membrane-associated guanylate kinases. Nature 378, 85-88. Kim, E., and Sheng, M . (1996). Differential K+ channel clustering activity of PSD-95 and SAP97, two related membrane-associated putative guanylate kinases. Neuropharmacology 35, 993-1000. Kornau, H.C., Schenker, L.T., Kennedy, M.B. , and Seeburg, P.H. (1995). Domain interaction between N M D A receptor subunits and the postsynaptic density protein PSD-95. Science 269, 1737-1740. Kreusch, A. , Pfaffinger, P.J., Stevens, C.F., and Choe, S. (1998). Crystal structure of the tetramerization domain of the Shaker potassium channel. Nature 392, 945-948. L i , D., Takimoto, K. , and Levitan, E.S. (2000). Surface expression of K v l channels is governed by a C-terminal motif. J Biol Chem 275, 11597-11602. 75 Lim, S.T., Antonucci, D.E., Scannevin, R.H., and Trimmer, J.S. (2000). A novel targeting signal for proximal clustering of the Kv2.1 K + channel in hippocampal neurons. Neuron 25, 385-397. Maruoka, N.D., Steele, D.F., Au, B.P., Dan, P., Zhang, X . , Moore, E.D., and Fedida, D. (2000). alpha-actinin-2 couples to cardiac Kvl .5 channels, regulating current density and channel localization in H E K cells. FEBS Lett 473, 188-194. Muller, B . M . , Kistner, U . , Kindler, S., Chung, W.J., Kuhlendahl, S., Fenster, S.D., Lau, L.F., Veh, R.W., Huganir, R.L., Gundelfinger, E.D., and Garner, C.C. (1996). SAP102, a novel postsynaptic protein that interacts with N M D A receptor complexes in vivo. Neuron 17, 255-265. Murata, M . , Buckett, P.D., Zhou, J., Brunner, M . , Folco, E., and Koren, G. (2001). SAP97 interacts with Kvl .5 in heterologous expression systems. A m J Physiol Heart Circ Physiol 281, H2575-2584. Nakagawa, N . , Hoshijima, M . , Oyasu, M . , Saito, N . , Tanizawa, K. , and Kuroda, S. (2000). ENH, containing PDZ and L I M domains, heart/skeletal muscle-specific protein, associates with cytoskeletal proteins through the PDZ domain. Biochem Biophys Res Commun 272, 505-512. Nehring, R.B., Wischmeyer, E., Doring, F., Veh, R.W., Sheng, M . , and Karschin, A. (2000). Neuronal inwardly rectifying K(+) channels differentially couple to PDZ proteins of the PSD-95/SAP90 family. J Neurosci 20, 156-162. Niethammer, M . , Kim, E., and Sheng, M . (1996). Interaction between the C terminus of N M D A receptor subunits and multiple members of the PSD-95 family of membrane-associated guanylate kinases. J Neurosci 16, 2157-2163. Ponting, C P . , Phillips, C , Davies, K.E . , and Blake, D.J. (1997). PDZ domains: targeting signalling molecules to sub-membranous sites. Bioessays 19, 469-479. Shen, N.V. , Chen, X . , Boyer, M . M . , and Pfaffrnger, P.J. (1993). Deletion analysis of K + channel assembly. Neuron 11, 67-76. Sheng, M . , and Sala, C. (2001). PDZ domains and the organization of supramolecular complexes. Annu Rev Neurosci 24, 1-29. Tiffany, A . M . , Manganas, L .N . , Kim, E., Hsueh, Y.P. , Sheng, M . , and Trimmer, J.S. (2000). PSD-95 and SAP97 exhibit distinct mechanisms for regulating K(+) channel surface expression and clustering. J Cell Biol 148, 147-158. Vallenius, T., Luukko, K. , and Makela, T.P. (2000). CLP-36 PDZ-LIM protein associates with nonmuscle alpha-actinin-1 and alpha-actinin-4. J Biol Chem 275, 11100-11105. van Huizen, R., Miller, K. , Chen, D .M. , L i , Y . , Lai, Z . C , Raab, R.W., Stark, W.S., Shortridge, R.D., and L i , M . (1998). Two distantly positioned PDZ domains mediate multivalent INAD-phospholipase C interactions essential for G protein-coupled signaling. Embo J 17, 2285-2297. 76 Wong, W., Newell, E.W., Jugloff, D.G., Jones, O.T., and Schlichter, L .C. (2002). Cell surface targeting and clustering interactions between heterologously expressed PSD-95 and the Shal voltage-gated potassium channel, Kv4.2. J Biol Chem 277, 20423-20430. Xia , H. , Winokur, S.T., Kuo, W.L., Altherr, M.R., and Bredt, D.S. (1997). Actinin-associated L I M protein: identification of a domain interaction between PDZ and spectrin-like repeat motifs. JCel l Biol 139,507-515. Xu, X .Z . , Choudhury, A. , L i , X . , and Montell, C. (1998). Coordination of an array of signaling proteins through homo- and heteromeric interactions between PDZ domains and target proteins. J Cell Biol 142, 545-555. 77 CHAPTER 3: SAP97 Increases Kvl.5 Currents through an Indirect N-terminal Mechanism Jodene Eldstrom*, Woo Sung Choi, David F. Steele and David Fedida FEBS Letters 547 (2003) 205-211 * Contributions by Jodene Eldstrom: Responsible for project management, experimental design, immunoprecipitations and western blots, confocal imaging, cloning and transfections, data analysis, and together with Dave Steele, writing and editing the manuscript. (Percent Contribution: 60%) 78 I N T R O D U C T I O N The targeting of ion channels to their intended cellular locations is a complex process (Sheng and Wyszynski, 1997). Folding and assembly occur concurrently with the synthesis of the channel proteins in the endoplasmic reticulum (Papazian, 1999; Sheng and Deutsch, 1998; Schulteis et al., 1998) and the assembled channels are then trafficked to their final destination on the cell surface or retained in intracellular pools (reviewed by (Griffith, 2001)). This transport or internal retention of the assembled channels is influenced by signal sequences that are undoubtedly recognized by ancillary proteins whose job it is to target the channels to their programmed locations in the cell (Li et al., 2000; Ma et al., 2001; Zhu et al., 2001). For example, a V X X S L motif present in the Kvl .4 C-terminus allows efficient trafficking of that channel to the cell surface whereas mutation to V X X S N (as exists in wild-type Kvl .2) yields channels that traffic poorly (Li et al., 2000). Pore sequences have also been implicated in channel trafficking (Zhu et al., 2001). The anchoring of the channels at their final destinations requires still other signal sequences. PSD-95, for example, binds an extreme C-terminal motif present in many ion channels (Kornau et al., 1995; Kim et al., 1995; Kim et al., 1996; Tiffany et al., 2000) and has been shown to anchor voltage-gated potassium channels at the membrane (Tiffany et al., 2000). SAP97, another PDZ-protein closely related to PSD95, was found by Tiffany et al. (Tiffany et al., 2000) to retain heterologously-expressed K v channels in the cell interior and to prevent their trafficking to the surface. Probably hundreds of PDZ proteins like PSD95 and SAP97 are expressed in human tissues. A SMART search of human genome project data (Schultz et al., 2000) turns up 259 potential genes encoding PDZ domains. Recently, we have found a second PDZ-binding domain in Kvl-type potassium channels (Eldstrom et al., 2002). In addition to C-terminal binding motifs, we found that PSD95 binds an N-terminal sequence within or very near the T l assembly domain. Deletion of either binding 79 region profoundly influenced the expression of hKvl.5 in HEK293 cells expressing both the channel and PSD95. PSD95 is abundant in neurons where it localizes to the post-synaptic density and associates with a number of proteins including ion channels (Reviewed in (Sheng et al., 2001; Wheal et al., 1998)). PSD95 is reportedly not present in the heart, however (Kistner et al., 1993). Instead, a number of related PDZ proteins are present in that tissue. One or more of these proteins might well perform a similar function in influencing Kv channel expression in that organ. Among the many possible cardiac PSD95 analogs are ZASP, E N H and SAP97 (Faulkner et al., 1999; Nakagawa et al., 2000; Leonoudakis et al., 2001). SAP97 is highly homologous to PSD95 and is abundant in the heart (Leonoudakis et al., 2001). It has also been recently reported to co-localize with Kvl .5 in cardiac myocytes and to co-immunoprecipitate with the channel when the two are co-expressed in COS-7 cells (Murata et al., 2001). Singularly among K v l channels, the expression of which is normally down-regulated by SAP97 (Tiffany et al., 2000; Kim et al., 1996), co-expression of Kvl .5 with SAP97 causes a large increase in cellular current levels (Murata et al., 2001). We have further investigated this interaction in an attempt to learn whether this PDZ protein might also bind the N-terminus of hKvl.5 in a manner analogous to the hKvl.5-PSD95 interaction. Indeed, we found that SAP97 co-expression significantly increased Kvl .5 currents in transfected H E K cells and that the N-terminus is essential for this effect. Surprisingly, we could find no evidence of binding of the two proteins in either transfected HEK293 cells or via yeast 2-hybrid assays and scant evidence for any interaction in heart. Thus, SAP97 modulation of Kvl .5 currents is likely through an indirect mechanism. MATERIALS AND METHODS DNA constructs For expression in HEK293 cells, SAP97 was cloned as an EcoRl fragment into pcDNA3 (Invitrogen, Carlsbad, CA). For the yeast two-hybrid assays an EcoRI-BstEII fragment of SAP97 80 containing the three PDZ domains and half of the SH3 domain was cloned into pGAD424 and sequenced to confirm the correct reading frame. Other constructs were as previously described (Eldstrom et al., 2002). Electrophysiological Procedures Procedures were as previously described (Eldstrom et al., 2002), except that HEK293 cells stably expressing Kvl .5 were transfected with pGFP and pcDNA3 or SAP97:pcDNA and pGFP. Myocyte Isolation, Immunolabeling and Imaging Myocytes were isolated and prepared using the method of Scriven et al. (2000). Isolated myocytes were fixed with 2% paraformaldehyde for 10 minutes followed by neutralization with glycine buffer for 10 minutes. They were then washed with phosphate buffered saline (PBS) for 10 minutes, made permeable with lu.L/mL PBS Triton-X (10 minutes), washed again for 10 minutes with PBS, and finally stored in PBS-azide solution. Cells were plated onto poly-L-lysine coated coverslips for a minimum of 3 hours at room temperature or overnight at 4° C. Myocytes were labeled for 3 hours at room temperature or overnight at 4° C with primary mouse monoclonal antibodies against SAP97 (1:300, BD Transduction, San Jose, CA), and rabbit polyclonal antibodies developed in our laboratory against Kvl .5 (1:300). Secondary Alexa 594 conjugated goat anti-mouse and Alexa 488 conjugated goat anti-rabbit (Molecular Probes, Eugene, OR) antibodies were incubated for 2 hours at room temperature. Cells were washed three times with PBS prior to mounting with a 90% glycerol, 2.5% w/v DABCO-PBS solution. Images of labeled cells were taken using a Bio-Rad Radiance Plus on an inverted Zeiss Axiovert microscope, using BioRad LaserSharp 2000 software. Images were prepared using NIH Image and Photoshop Software Packages. 81 Co-immunoprecipitations Co-immunoprecipitations from transfected HEK293 cell lines were as previously described (Eldstrom et al., 2002) except that a mouse monoclonal antibody to SAP97 was employed as appropriate. In cardiac experiments, excised ventricular or atrial tissue was added to ice cold non-denaturing lysis buffer (20 mM HEPES, 150 mM NaCl, 2 mM EDTA, 10% glycerol, 0.5% IGEPAL CA-630, 1% hemoglobin, lOmM iodoacetamide, 1 mM PMSF and Aprotinin (0.2 TIU/ml)), and underwent homogenization, two pulses of 30 seconds using an Ultra-Turrax T25 homogenizer (Janke & Kunkel, Staufen, Germany). The lysate was transferred to 1.5 ml Eppendorf tubes and nuclei and debris pelleted by spinning at lOOOg for 10 minutes, the supernatant was transferred to a fresh tube and pre-cleared twice with pre-washed CL4B sepharose beads. The pre-cleared supernatant was mixed with 3-7 (al of antibody and incubated on ice for 1 hour. Pre-swelled and washed Protein A sepharose beads were added to the lysate/antibody mixture and incubated for 1 hour with frequent mixing. The beads were pelleted and the supernatant removed. Beads were then washed three times with TSA-T (10 mM Tris pH 8.0, 140 mM NaCl, 0.1% Triton X-100), once with TSA (10 mM Tris pH 8.0, 140 mM NaCl) and once with 50 mM Tris pH 6.8. Beads were then suspended in SDS sample buffer and heated to 95°C for 5 minutes before being run on a 12 % SDS polyacrylamide gel. Proteins were then transferred to PVDF membranes and probed with either rabbit anti-Kvl.5 (1:10,000; developed in our lab), mouse anti-SAP97 (1:5,000) or mouse anti-actinin (1:10,000, Sigma, St. Louis, MO) and HRP conjugated goat anti-rabbit IgG or sheep anti-mouse (1:10,000; Jackson Laboratories). Antibody binding was detected using a chemiluminescent reagent (Renaissance, New England Nuclear). Yeast two-hybrid experiments Kvl .5 and Kv l .4 N - and C-terminal fragments were cloned into appropriate pGBD vectors; SAP97 and a-actinin2 were cloned in pGAD vectors (James et al., 1996). For two-82 hybrid growth assays, yeast strain PJ69-4a was transformed with appropriate pairs of clones. Growth on media lacking adenine and histidine was monitored to assay for interaction. (3-Galactosidase assays were conducted in the Y190 strain as previously described (Eldstrom et al., 2002). RESULTS Co-expression of SAP97 with hKvl.5 enhances hKvl.5 currents in HEK cells As a first step towards understanding the role of SAP97 in Kvl .5 expression, we co-expressed hKvl.5 and SAP97 in H E K cells. To accomplish this, a SAP97 construct was transfected into HEK293 cells that stably express hKvl.5. Experiments were carried out so that the electrophysiologist was blinded to the transfection group (SAP97 or vector alone). As illustrated in Figure 3.1, hKvl.5 currents were significantly enhanced in the presence of the SAP97 construct 24 hours after transfection. There was no apparent change in the activation or inactivation time courses of the current. It was found that peak currents were approximately doubled (from 7.9±1.4 nA to 18.6+2.3 nA at +60 mV, n=9) 24 hours after SAP97 transfection compared to control cells that were transfected with empty vector (pGFP, n=8). This effect was independent of the test potential used (Figure 3.1). Thus, SAP97 has a significant effect on hKvl.5 expression levels in transfected H E K cells, a result confirming in mammalian cells that obtained by Murata et al. (2001) who found that SAP97 co-expression substantially enhanced Kv l .5 currents in doubly-injected Xenopus oocytes. These electrophysiological data demonstrate a role for SAP97 in the regulation of hKvl.5 currents at least in heterologous cell systems. The enhancement of hKvl.5 currents could be due to a direct interaction between SAP97 or to an indirect action, either through a bridging molecule or through an effect on another molecule or molecules that somehow regulates hKvl.5 activity. 83 pGFP SAP97 Figure 3.1. Effect of SAP97 expression on hKvl .5 peak currents. H E K 2 9 3 cells stably expressing h K v l . 5 were used to test the effect o f SAP97 co-expression on ionic currents. Cells were transiently transfected with either empty vector (pcDNA3) + pGFP , or SAP97 + pGFP and peak currents at several voltages were measured under voltage clamp in the whole-cell configuration using a step voltage clamp protocol illustrated under the current tracings. Representative current traces at potentials of -20, -10, 20 and 60 m V are shown, and mean data ± S . E . M . (B) for 8 (SAP97) or 9 (pGFP) cells are presented. The asterisks indicate statistical significance with P<0.01. (Data collected and analyzed by W S Choi) 84 To begin to distinguish between these possibilities, we conducted co-localization, co-immunoprecipitation and yeast two-hybrid experiments. hKvl.5 and SAP97 fail to co-localize in rat cardiac myocytes Co-localization experiments were performed to test whether hKvl.5 and SAP97 interact in vivo. Confocal images were obtained of isolated rat ventricular myocytes cross-reacted with antibodies to hKvl.5 and SAP97. As shown in Figure 3.2, both Kvl .5 and SAP97 were widely distributed in the myocytes. The two molecules co-localized very poorly - i f at all - in these cells, however. Kv l .5 was enriched at the intercalated disks and SAP97 was generally internal. SAP97 appeared to be largely distributed along Z-lines hKvl.5 co-immunoprecipitations with SAP97 from cardiac myocytes In order to further test whether or not Kvl .5 and SAP97 directly interact in atrial and/or ventricular myocytes, we conducted co-immunoprecipitation experiments. We found that while Kvl .5 and SAP97 were efficiently immunoprecipitated by their respective antibodies, SAP97 failed to co-immunoprecipitate with Kvl .5 in five of six experiments from ventricular extracts, although in one experiment out of six, SAP97 did co-immunoprecipitate very poorly with hKvl.5 (Figure 3.3A). Whether this represents a real interaction or is a one-time artifact is unclear. We could not detect any co-immunoprecipitation of Kvl .5 by antibody to SAP97, nor of SAP97 with Kvl .5 from atrial extracts (data not shown). These results are consistent with the lack of co-localization seen in imaging experiments. hKvl.5 fails also to co-immunoprecipitate with SAP97 from transfected HEK cells While Murata et al. (2001) were unable to consistently co-immunoprecipitate the proteins from the cardiac myocytes, they were able to co-immunoprecipitate SAP97 and Kvl .5 from transfected COS-7 cells. We therefore decided to test whether the two proteins could be co-immunoprecipitated from a highly over-expressed system using transfected H E K cells. 85 SAP97 Kv1.5 Merged Figure 3.2. S A P 9 7 does not co-localize with K v l . 5 in rat ventr icular myocytes. Confocal images of single optical slices of ventricular myocytes. SAP97 was detected with Alexa594-labeled secondary antibody (red); Kvl.5 was detected with Alexa488-labeled secondary antibody (green). Yellow indicates co-localization of the channel with SAP97 in the merged image. 86 HEK293 cells stably expressing a T7-tagged hKvl.5 construct were transfected with a SAP97 clone in pcDNA3 and co-immunoprecipitation experiments were conducted. As shown in Figure 3.3B, both hKvl.5 and SAP97 could be readily immunoprecipitated from these cells. But antibody to one never brought down the other. In control experiments in which PSD95 was co-expressed in place of SAP97, PSD95 readily co-immunoprecipitated with hKvl.5 (Figure. 3.3C). This was a surprising result given that the influence of SAP97 on hKvl.5 currents in this cell line is marked. SAP97 interacts with Kv l .4 but not with hKvl .5 in yeast two-hybrid experiments In a final attempt to determine whether we could detect an interaction under quite different conditions, we conducted yeast two-hybrid studies. Yeast strain PJ69-4a (James et al., 1996) was co-transformed with C- and N - terminal fragments of hKvl.5 or the C-terminus of Kv l .4 in pGBD and the PDZ domains of SAP97 in pGAD. As shown in Figure 3.4, yeast harboring both the K v l . 4 C-terminus and the SAP97 construct grew well on media lacking adenine and histidine as did another control in which the hKvl.5 N-terminus was tested against a-actinin2. Neither the hKvl.5 C- nor N-terminal fragments allowed growth on the test media. Similar results were obtained when the constructs were co-transformed into yeast strain Y190 and (3-galactosidase assays were conducted. Only the Kv l .4 C-terminus/SAP97 and hKvl.5 N -terminus/a-actinin2 combinations yielded P-galactosidase activities above control (data not shown). SAP97 enhancement of hKvl .5 currents depends on an intact K v l . 5 N-terminus The results of all of the experiments described above are inconsistent with a typical PDZ-extreme C-terminal interaction between SAP97 and K v l . 5 . Instead, SAP97 must exert its effect by some different, most likely indirect, mechanism. To gain further insight into this issue, experiments with hKvl.5 C- and N-terminal deletion mutants were conducted. SAP97 was transfected into HEK293 cells stably expressing either hKvl .5AETDL or hKvl.5AN209 87 Ab: None Kv1.5 SAP97 175 k D _ 83 kD * B Kv1.5 Ab: None T7 SAP97 83 kD 63 kD SAP97 Ab: None T7 SAP97 (Kv1.5) 175 kD 83 kD Kv1.5 Ab: None T7 GFP (Kv1.5) (PSD95) 175 kD 83 kD Figure 3.3. Attempted co-immunoprecipitation of SAP97 and Kvl.5. A . The single example (1 of 6) of co-immunoprecipitation of SAP97 with K v l . 5 from rat ventricular extracts. Extracts from ventricular tissue were mixed with no antibody, anti-Kvl.5 or anti-SAP97. The protein-antibody complexes were precipitated with Protein A-sepharose and subjected to Western analysis probed with anti-SAP97. B. Kv l .5 and SAP97 fail to co-immunoprecipitate from transfected HEK293 cells. Extracts of SAP97, T7-tagged K v l . 5 double-expressing H E K cells were mixed with no antibody, anti-T7 or anti-SAP97. The complexes were precipitated with Protein A-sepharose and subjected to Western analysis probing with anti-T7 (left panel) or anti-SAP97 (right panel). C. PSD95 co-immunoprecipitates with K v l . 5 from transfected H E K cells. Control experiments conducted as in (B) except that the H E K cells co-expressed T7-tagged Kvl .5 and GFP-tagged PSD95. PSD95 was immunoprecipitated with anti-GFP. The blot was probed with anti-T7. 88 -ade-his -leu-trp N-Kv1.5/SAP97 C-Kv1.5/SAP97 C-Kv1.4/SAP97 Kv1.5/a-actinin2 N-Kv1.5 alone C-Kv1.5 alone C-Kv1.4 alone SAP97 alone Figure 3.4. Yeast two-hybrid assay for the interaction of Kv l .5 N - and C-terminal interactions with SAP97. The K v l channel fragments were expressed as fusions to the DNA-binding domain of the yeast G A M protein. SAP97 and a-actinin2 clones were expressed as fusions to the G A L 4 transcription activating domain. Growth on -ade, -his media indicates a detectable interaction. Growth on -leu, -trp confirms that both fusion vectors are present in the yeast. " A l o n e " refers to yeast co-transformed with the reported construct plus its empty vector partner for the yeast two-hybrid assay. 89 (Maruoka et al., 2000) and potassium currents were measured as before. The hKvl .5AETDL mutant lacks the C-terminal four amino acids required for interaction with PDZ proteins and hKvl.5AN209 lacks the N-terminal 209 amino acids of the channel. Strikingly, it was the N -terminus and not the C-terminus that proved essential to the SAP97-enhancement of Kvl .5 potassium currents. hKvl.5AN209 K + currents were not affected at all by SAP97 co-expression (Figure 3.5A) but, as with the wild-type channel, SAP97 co-expression significantly increased potassium currents in the AETDL line (Figure 2.5B). Again, this effect of SAP97 on the channel was potential-independent. This lack of dependence of the SAP97 effect on the C-terminal acids of Kvl .5 is in stark contrast to the results with other K v l channels (Tiffany et al., 2000; Kim et al., 1996). A novel regulatory mechanism must underlie the SAP97-mediated increase in Kvl .5 potassium currents. a-Actinin2 co-immunoprecipitates with both hKvl .5 and SAP97 The pattern of SAP97 staining in the confocal imaging experiments (Figure 3.2) suggests that the protein might be in large measure localized to T-tubules or to Z-disks. Occasionally, a similar but weaker Z-band-type staining pattern can be seen in Kvl .5 imaging experiments as well (data not shown). One of the major proteins at the Z-line is ot-actinin2 (Takada et al., 2001), a protein we have previously demonstrated to interact with hKvl.5 in heterologous cells (Maruoka et al., 2000). To test whether there is any potential for a role for actinin in the SAP97-Kvl.5 interplay, co-immunoprecipitation of Kvl .5 with a-actinin2, and of SAP97 with a-actinin2 was attempted from cardiac myocyte lysates. As shown in Figure 3.6, both SAP97 (Figure 3.6A) and Kvl .5 (Figure 3.6B) co-immunoprecipitated well with a-actinin2. While actinin may or may not be involved in the SAP97 enhancement of hKvl.5 currents, it could well underlie the weak co-immunoprecipitation of SAP97 with Kvl .5 in cardiac cells. Limited co-90 A AN209 pGFP SAP97 75 ms -60 mV -20 mV 20 mV 60 mV B AETDL pGFP SAP97 20 Figure 3.5. A . SAP97 fails to increase Kvl.5AN209 peak currents. H E K 2 9 3 cells stably expressing K v l . 5 A N 2 0 9 were used to test the effect o f SAP97 co-expression on peak currents. Representative current traces are shown on the left. The bar graph illustrates the mean currents + S . E . M for 4 (SAP97) and 7 (pGFP) cells. B . SAP97 increases K v l . 5 A E T D L peak currents. H E K 2 9 3 cells stably expressing K v l . 5 A E T D L were used to test the effect o f SAP97 co-expression on peak currents. Representative current traces (A) and mean data + S . E . M . (B) for 9 (SAP97) and 9 (pGFP) cells are presented. In both A and B , cells were transiently transfected with either empty vector (pcDNA3) + p G F P or SAP97 + p G F P and peak currents at several voltages were measured under voltage clamp in whole-cell configuration. The asterisks indicate statistical significance with P<0.01. (Data collected and analyzed by W S Choi) 91 B a-Actinin Atria Ventricle Ab: None Kv1.5 a-Act None Kv1.5 a-Act Figure 3.6. Both K v l . 5 and SAP97 interact with ct-aetinin2 in rat heart. A . Co-immunoprecipitation of SAP97 with a-actinin2 from rat ventricular extracts. Extracts from ventricular tissue were mixed with no antibody, anti-a-actinin2 or anti-SAP97. The protein-antibody complexes were precipitated with Protein A-sepharose and subjected to Western analysis probed with anti-SAP97 (left panel) or anti-a-actinin2 (right panel). B. K v l . 5 and a-actinin2 co-immunoprecipitate from rat cardiac extracts. Extracts from atrial or ventricular tissue were mixed with no antibody, anti-a-actinin2 or anti-Kvl.5. The protein-antibody complexes were precipitated with Protein A-sepharose and subjected to Western analysis. Blots were probed with anti-cc-actinin2. 92 association of the two proteins with Z-band ct-actinin2, for example, could be more than sufficient to yield the results shown in Figure 3.3A. DISCUSSION We have recently shown that PSD95, a protein closely related to SAP97, binds both the C- and N-termini of hKvl.5 via its PDZ domains (Eldstrom et al., 2002). PSD95 is reportedly not expressed in heart (Kistner et al., 1993) and we wondered whether the N-terminal PDZ-binding domain in Kvl .5 had any relevance to the channel's regulation in that organ. For this reason, we have examined the interaction between Kvl .5 and SAP97, a PDZ protein that is widely expressed in mammalian heart (Leonoudakis et al., 2001). Without a doubt, SAP97 significantly increases hKvl.5 currents in heterologous cells. In this regard, our results are similar to those reported by Murata et al. (2001) using Xenopus oocytes. At first glance, it seems surprising, therefore, that we could find very little evidence for any direct interaction between the two proteins. But the effect of SAP97 co-expression on Kvl .5 currents is exceptional and dramatically different from its known effects on other K v l channels. K v l . l , K v l . 2 , 1.3 and Kv l .4 channels are all clustered internally by SAP97 and their surface expression is substantially reduced (Tiffany et al., 2000; Kim et al., 1996). Indeed, given this stark difference in effects of SAP97 on most K v l channels versus its effect on Kv l .5 , it would be surprising i f SAP97 did in fact bind Kv l .5 . While Murata et al. (2001), report that the two proteins do interact, they, like us, had difficulty detecting these interactions in the heart. They were unable to co-immunoprecipitate the two proteins from that tissue and, importantly, their co-localization experiments employed a fairly non-specific antibody directed against PSD95. The most likely explanation for the co-localization they report in myocytes is that they detected an overlap between Kvl .5 and a PDZ protein other than SAP97. The SAP97 distribution they report in their myocytes is unusual in that it is concentrated at the cell surface and at the intercalated 93 disk. In our hands, and those of Leonoudakis et al. (2001), cardiac myocyte SAP97 staining was mainly intracellular along the Z-disks or perhaps at T-tubules. K v l . 5 , on the other hand, was most prevalent at the intercalated disks. How can SAP97 enhance Kvl .5 currents yet have no physical interaction with the channel? It remains formally possible that Kvl .5 and SAP97 do directly interact but that the interaction is fleeting. We cannot exclude the possibility that phosphorylation or some other modification(s) of either the channel or SAP97 influence binding. Phosphorylation at a C-terminal serine residue is known to disrupt the interaction of PSD95 with Kir2.3 (Cohen et al., 1996). Similar phosphorylation inhibits A M P A receptor subunit GluR2 binding to the PDZ domains of GRIP but not of PICK1 (Matsuda et al., 1999; Chung et al., 2000). Conceivably differences in intrinsic modification rates in various cell types could explain the markedly different results we obtained in co-immunoprecipitations from transfected H E K cells and those obtained by Murata et al. (2001) in COS-7 cells. Perhaps in COS-7 cells the state of the channel is permissive of a stable SAP97-Kvl.5 interaction that is not seen in other cell types (including heart). But even i f SAP97 does bind the Kvl .5 C-terminus under specific conditions, that fact cannot explain the lack of a dependence of the SAP97-mediated current enhancement on the extreme Kvl .5 C-terminus. Indeed, it would appear that the Kvl .5 N-terminus, not the C-terminus, is necessary for the SAP97 effect. This must be mediated by a different mechanism than is dominant with the other K v l channels. One possibility is that SAP97 frees Kvl .5 to traffic to the surface by titrating away another binding protein(s). Similarly, it could free another Kvl .5 binding protein to traffic to the surface, and then stabilize the channel there. SAP97 might even affect Kvl .5 processing or its transcript levels by an as yet unidentified cascade. Any of these scenarios would be consistent with our near-complete failure to detect any molecular interaction Kvl .5 and SAP97. 94 We found that both Kvl .5 and SAP97 interact strongly with a-actinin2. Thus, it is conceivable that actinin could be involved in the SAP97 enhancement of Kvl .5 currents. For example, a-actinin2 might serve as a bridge between the two molecules. Indeed, SAP97 cardiac myocyte staining suggests that SAP97 localizes to the Z-bands, structures very rich in a-actinin2 (Beggs et al., 1992), and, in occasional cells, some Kvl .5 appears similarly distributed. It is hard to reconcile this actinin bridge hypothesis with the failure of SAP97 and Kvl .5 to co-localize in the majority of myocytes, however. The length of actinin is too small to account for a complete lack of Kvl.5-SAP97 overlap. If actinin is involved, perhaps it is more likely that it is titrated away from Kvl .5 by SAP97. If, for example, actinin binding to Kvl .5 anchors the channel in the cell interior, SAP97 binding to an overlapping site in actinin could free the channel to traffic to the cell surface. Such a possibility would be consistent, at least, with the known effects of disruption of the cytoskeleton and a-actinin2 anti-sense experiments. Kv l .5 currents are greatly increased by these treatments (Maruoka et al., 2000). In summary, we have found that SAP97 has a profound influence on Kvl .5 currents in heterologous cells. This effect is dependent on an intact Kvl .5 N-terminus, yet little evidence for a direct interaction between the two proteins could be found. Further research will be required to determine the exact mechanism by which SAP97 increases Kvl .5 K + currents. 95 A C K N O W L E D G E M E N T S We thank Grace Lu and Daniela Cukovic for excellent technical assistance and Brook Glanville for assistance with myocyte preparations. R E F E R E N C E S Beggs, A . H . , Byers, T.J., Knoll, J.H., Boyce, F .M. , Bruns, G.A., and Kunkel, L . M . (1992). Cloning and characterization of two human skeletal muscle alpha-actinin genes located on chromosomes 1 and 11. J Biol Chem 267, 9281-9288. Chung, H J . , Xia, J., Scannevin, R.H., Zhang, X . , and Huganir, R.L. (2000). Phosphorylation of the A M P A receptor subunit GluR2 differentially regulates its interaction with PDZ domain-containing proteins. J Neurosci 20, 7258-7267. Cohen, N.A. , Brenman, J.E., Snyder, S.H., and Bredt, D.S. (1996). Binding of the inward rectifier K + channel Kir 2.3 to PSD-95 is regulated by protein kinase A phosphorylation. Neuron 77,759-767. Eldstrom, J., Doerksen, K.W., Steele, D.F., and Fedida, D. (2002). N-terminal PDZ-binding domain in K v l potassium channels. FEBS Lett 531, 529-537. Faulkner, G., Pallavicini, A. , Formentin, E., Comelli, A. , Ievolella, C , Trevisan, S., Bortoletto, G., Scannapieco, P., Salamon, M . , Mouly, V. , Valle, G., and Lanfranchi, G. (1999). ZASP: a new Z-band alternatively spliced PDZ-motif protein. J Cell Biol 146, 465-475. Griffith, L .C. (2001). Potassium channels: the importance of transport signals. Curr Biol 11, R226-228. James, P., Halladay, J., and Craig, E.A. (1996). Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast. Genetics 144, 1425-1436. Kim, E., Niethammer, M . , Rothschild, A. , Jan, Y . N . , and Sheng, M . (1995). Clustering of Shaker-type K + channels by interaction with a family of membrane-associated guanylate kinases. Nature 378, 85-88. Kim, E., and Sheng, M . (1996). Differential K + channel clustering activity of PSD-95 and SAP97, two related membrane-associated putative guanylate kinases. Neuropharmacology 35, 993-1000. Kistner, U . , Wenzel, B . M . , Veh, R.W., Cases-Langhoff, C , Garner, A . M . , Appeltauer, U . , Voss, B. , Gundelfinger, E.D., and Garner, C.C. (1993). SAP90, a rat presynaptic protein related to the product of the Drosophila tumor suppressor gene dlg-A. J Biol Chem 268, 4580-4583. 96 Kornau, H.C., Schenker, L.T., Kennedy, M.B. , and Seeburg, P.H. (1995). Domain interaction between N M D A receptor subunits and the postsynaptic density protein PSD-95. Science 269, 1737-1740. Leonoudakis, D., Mailliard, W., Wingerd, K. , Clegg, D., and Vandenberg, C. (2001). Inward rectifier potassium channel Kir2.2 is associated with synapse-associated protein SAP97. J Cell Sci 114, 987-998. L i , D., Takimoto, K. , and Levitan, E.S. (2000). Surface expression of K v l channels is governed by a C-terminal motif. J Biol Chem 275, 11597-11602. Ma, D., Zerangue, N . , Lin, Y.F . , Collins, A. , Yu , M . , Jan, Y . N . , and Jan, L . Y . (2001). Role of ER export signals in controlling surface potassium channel numbers. Science 291, 316-319. Maruoka, N.D., Steele, D.F., Au, B.P., Dan, P., Zhang, X . , Moore, E.D., and Fedida, D. (2000). alpha-actinin-2 couples to cardiac Kvl .5 channels, regulating current density and channel localization in H E K cells. FEBS Lett 473, 188-194. Matsuda, S., Mikawa, S., and Hirai, H . (1999). Phosphorylation of serine-880 in GluR2 by protein kinase C prevents its C terminus from binding with glutamate receptor-interacting protein. JNeurochem 73, 1765-1768. Murata, M . , Buckett, P.D., Zhou, J., Brunner, M . , Folco, E., and Koren, G. (2001). SAP97 interacts with Kvl .5 in heterologous expression systems. A m J Physiol Heart Circ Physiol 281, H2575-2584. Nakagawa, N . , Hoshijima, M . , Oyasu, M . , Saito, N . , Tanizawa, K. , and Kuroda, S. (2000). ENH, containing PDZ and L I M domains, heart/skeletal muscle-specific protein, associates with cytoskeletal proteins through the PDZ domain. Biochem Biophys Res Commun 272, 505-512. Papazian, D . M . (1999). Potassium channels: some assembly required. Neuron 23, 7-10. Schulteis, C.T., Nagaya, N . , and Papazian, D . M . (1998). Subunit folding and assembly steps are interspersed during Shaker potassium channel biogenesis. J Biol Chem 273, 26210-26217. Schultz, J., Copley, R.R., Doerks, T., Ponting, C P . , and Bork, P. (2000). SMART: a web-based tool for the study of genetically mobile domains. Nucleic Acids Res 28, 231-234. Scriven, D.R., Dan, P., and Moore, E.D. (2000). Distribution of proteins implicated in excitation-contraction coupling in rat ventricular myocytes. Biophys J 79, 2682-2691. Sheng, Z., and Deutsch, C. (1998). Assembly of ion channels. Methods Enzymol 293, 17-32. Sheng, M . , and Sala, C. (2001). PDZ domains and the organization of supramolecular complexes. Annu Rev Neurosci 24, 1-29. Sheng, M . , and Wyszynski, M . (1997). Ion channel targeting in neurons. Bioessays 19, 847-853. 97 Takada, F., Vander Woude, D.L., Tong, H.Q., Thompson, T.G., Watkins, S.C., Kunkel, L.M., and Beggs, A.H. (2001). Myozenin: an alpha-actinin- and gamma-filamin-binding protein of skeletal muscle Z lines. Proc Natl Acad Sci U S A 98, 1595-1600. Tiffany, A.M. , Manganas, L.N., Kim, E., Hsueh, Y.P., Sheng, M. , and Trimmer, J.S. (2000). PSD-95 and SAP97 exhibit distinct mechanisms for regulating K(+) channel surface expression and clustering. J Cell Biol 148, 147-158. Wheal, H.V., Chen, Y., Mitchell, J., Schachner, M. , Maerz, W., Wieland, H., Van Rossum, D., and Kirsch, J. (1998). Molecular mechanisms that underlie structural and functional changes at the postsynaptic membrane during synaptic plasticity. Prog Neurobiol 55, 611-640. Zhu, J., Watanabe, I., Gomez, B., and Thornhill, W.B. (2001). Determinants involved in K v l potassium channel folding in the endoplasmic reticulum, glycosylation in the Golgi, and cell surface expression. J Biol Chem 276, 39419-39427. 98 C H A P T E R 4: K v l . 5 Is A n Important Component Of Repolarizing K + Current In Canine Atrial Myocytes David Fedida 1 ' 2, Jodene Eldstrom*1, J. Christian Hesketh2, Michelle Lamorgese3, Laurie Castel3, David F. Steele1, and David R. Van Wagoner3 'Department of Physiology, University of British Columbia, 2146 Health Sciences Mall, Vancouver, British Columbia, Canada V6T 1Z3; 2Cardiome Pharma Corp., 3650 Wesbrook Mall, Vancouver, BC, Canada V6S 2L2; 3Department of Cardiology, Cleveland Clinic Foundation, Cleveland, OH 44195, USA. Circulation Research 93(8) (2003) 744-51 * Contributions made by Jodene Eldstrom: Responsible for i) all the immunocytochemistry and confocal imaging, including antibody testing and controls in HEK293 cells and canine myocytes; ii) sample preparation and testing of antibody cross reactivity using western blots and peptide blocking controls; iii) prepared related figures, results and discussion and assisted in writing and editing each draft of the manuscript. (Percent contribution 25 %) 99 I N T R O D U C T I O N In the heart, opening of voltage-activated K + channels, after the initial Na + and C a 2 + current driven depolarizations, governs the amplitude and duration of the cardiac action potential. In human atria, as well as in other species, the atrial action potential is abbreviated with respect to the ventricular action potential, and the plateau is at a lower potential, because of a greater density of repolarizing K + currents compared with inward currents. Progress has been made in understanding the molecular correlates of the rapid early repolarization current, I t 0, which is observed in both atrial and ventricular myocytes, (Hume et al., 1990) and is now thought to be attributable to differential expression across species of Kv4.2/4.3 and Kv l .4 subfamily a subunits (Barry and Nerbonne, 1996; Nerbonne, 2000). An additional rapidly activating K + current, IKUR, is important in speeding the repolarization of atrial myocytes in many species (Wang et al., 1993; Boyle and Nerbonne, 1992). Although the Kvl .5 channel that underlies IKUR is expressed at both the mRNA (Gidh-Jain et al., 1996; Snyders et al., 1993) and protein levels (Mays et al., 1995) in human ventricular tissue, albeit at low levels, (Kaab et al., 1998) its functional activity appears to be specific to the atrium (Feng et al., 1997). Thus, block of IKUR may produce an atrial-selective increase in action potential prolongation, and thus refractoriness.(Knobloch et al., 2002) Drugs that modulate IKUR may have useful selectivity and safety advantages over compounds that block ionic currents like IKF present in both atrial and ventricular myocytes. IKUR blockade may thus represent a promising new approach to treating atrial fibrillation (AF).(Van Wagoner, 2000). Like the transient outward current, there seems to be species variation in the molecular correlates of IKUR- Kv l .5 is thought to underlie this current in human and rat atrium, (Barry et al., 1996; Deal et al., 1996; Feng et al., 1997; Fedida et al., 1993; Van Wagoner et al., 1997) whereas Kv l .2 has an important role in rat lK ) Siow(Bou-Abboud and Nerbonne, 1999). Studies in canine atrium have suggested 100 that Kv l .5 is absent in that species (Yue et al., 2000), and that Kv3.1 is the primary potassium channel subunit contributing to the so-called IKURM in dog (Yue et al., 2000; Yue et al., 1996a). Because the canine model has been widely used in A F studies, it is important to be certain whether K v l . 5 , Kv3.1, or both underlie this current in canine atria. In this study we have tested the hypothesis that Kvl .5 is present in canine atrium and have found Kvl .5 to be expressed at high levels in both canine atrium and ventricle. Surprisingly, we were unable to find biochemical evidence of Kv3.1 expression in these tissues, although current sensitive to low concentrations of tetraethylammonium ion (TEA + ) was present in some canine atrial myocytes. Based on immunofluorescence and pharmacological data, Kv l .5 would seem to contribute significantly also to canine IKUR, at least in the dogs tested here. These results have important implications for the use of the dog model in the study of the atrial action potential and for the treatment of AF. The differences between our results and those reported by Yue, et al. (2000), raise the possibility that there is significant variation in channel expression from animal to animal. M A T E R I A L S A N D M E T H O D S hKvl.5 channels were studied in a human embryonic kidney cell line (HEK293) as reported previously (Fedida et al., 1993). Rat Kv4.2, rat Kv2.1, rat Kv3.1 and human HI Na + channels were also stably expressed in HEK293 cells. Atrial myocytes were dissociated from canine atrial specimens using a chunk dissociation technique (Van Wagoner et al., 1999). In paired blots shown in Figure 4.IB and D, for comparison of membrane and cytosol fractions, the protein concentration in the membrane and cytosol was measured using the bicinchoninic acid method (Pierce) and then equal amounts were loaded on the gels (30 pg for Kv3.1 and Kvl .5 blots). Other Western blots were performed as described in Online Supplement (Chapter4; Appendix; available at http://www.circresaha.org; Maruoka et al., 2000). Blots were probed with a 101 polyclonal anti-Kvl.5 antibody (Upstate Biotechnologies, Lake Placid N Y ; 1:200 dilution) or with a Kv l .5 antibody generated in our laboratory, a rabbit antibody against the C-terminus of hKvl.5 (aa 537-553 EQGTQSQGPGLDRGVQR; 1:10,000). This antibody was chosen for its unique sequence region at the C-terminus that is not shared with other Kv channels. It shares 12 of 17 residues with canine Kvl .5 and detected canine Kvl .5 under several experimental conditions (see below Figure. 4.1 and 4.4). Blots for Kv3.1 were probed with anti-Kv3.1 (Alomone Labs, Jerusalem Israel; 1:5000). Detection was by HRP-labeled goat anti-rabbit antibody (Jackson Immuno Research, PA) and Renaissance Western Blot Chemiluminescence Reagent Plus (NEN). R N A was isolated from heart or brain samples using the RNeasy Midiprep kit (Qiagen, Mississauga ON, Canada) according to the manufacturer's instructions. Invitrogen's Superscript One-Step reverse transcriptase-polymerase chain reaction (RT-PCR) with Platinum Taq kit was used for most RT-PCR experiments. An expanded Material and Methods section can be found in the online data supplement available at http://www.circresaha.org. 102 R E S U L T S Detection of Kvl.5 and Kv3.1 in dog atrium Figure 4.1 A shows representative Western blots of untransfected HEK293 cells and H E K cells stably expressing K v l . 5 , Kv3.1, Kv l . 2 , and Kvl .4 , as well as canine atrial, ventricular and brain tissue preparations, all probed with our C-terminal Kvl .5 antibody. A band migrating at almost the same position as Kvl .5 in transfected HEK293 cells was abundantly expressed in all of the canine crude membrane fractions (20,000-g pellet). Identical results were obtained using cardiac tissue from each of four canine atrial preparations tested. These bands from canine atrium, and also from rat ventricle were prevented after incubation with the blocking peptide used to generate the antibody (Figure 6.2, Appendix 2, also available at http://www.circresaha.org as an online data supplement). Using the Upstate polyclonal antibody, channel protein was found in the crude membrane fraction of paired samples of canine atrium and canine ventricle (Figure 4. IB). Although a strong band was detected in each sample used, it was conceivable that some of the Kvl .5 originated from non-cardiac muscle tissue. Therefore, we used protein extracts from isolated atrial myocytes in additional Western blots to confirm Kvl.5-specific expression in myocytes (Figure 4.1C). These blots showed weaker bands because of the relatively low yield of atrial myocytes, but an identically sized band is clearly detected from the hKvl.5 HEK293 stable line and from the canine atrial myocytes. Given that Yue et al. (2000) have attributed canine rapid delayed rectifier, IKUR(CI), to the presence of Kv3.1 in that tissue, we sought to confirm the biochemical presence of Kv3.1 in canine heart, and expected a 97-kDa Kv3.1 band, as detected by Yue et al (2000). However, paired Western blots from the same protein preparations used for the Kvl .5 experiments consistently failed to uncover a band of the expected size for Kv3.1 in either atrial or ventricular extracts (Figure 4.ID), although an 85-kDa band was evident in rat brain (RB) membrane 103 preparations (RB). Bands of -70, 60, and -50 kDa were present in cytosolic and membrane fractions, respectively of the canine heart preparations. It seems unlikely that these smaller bands were immature or less glycosylated Kv3.1 protein, because they were still present (marked by arrows) after incubation with the blocking peptide (Figure 4.IE, right). In addition, although Kv3.1 could be immunoprecipitated from dog brain preparations, no band was pulled down in several attempts to immunoprecipitate the protein from cardiac preparations (data not shown). To further characterize the expression of Kvl .5 and Kv3.1 in canine atria, RT-PCR experiments were conducted (Figure 4.2). For K v l . 5 , an oligonucleotide pair designed to amplify a 743-bp portion of that gene's transcript was used. For Kv3.1, oligonucleotide pairs identical to those used to clone the channel were used (Yue et al., 2000), as were other oligonucleotides, (Appendix 2: Online data supplement). Amplified Kvl .5 transcript was readily detectable from DNasel-treated dog atrial R N A preparations as shown by the bright band adjacent to the ladder in Figure 4.2. No band was detectable if the R N A preparations were treated with RNAase A prior to RT-PCR (data not shown). Kv3.1 from brain was also readily detected using primers 1 and 7 or 4 and 10 (Appendix 2: Online data supplement), but products could not be amplified from atrium in >10 attempts using any of the primer pairs tested. In Figure 4.2, for Kv3.1, primer pairs 4 and 10 were used and a 725-bp transcript was expected. As a control for the quality of the heart RNA, we amplified a 588-bp segment of dystroglycan message across a splice junction (Figure 4.2, far right lane). Although this is expected for amplification from mRNA, the primers are more than 17 kbp apart in the dog genome (Leeb et al., 2003). 104 Figure 4.1. K v l . 5 and Kv3.1 expression in HEK293 cells, canine heart, and brain. A , C-terminal antibody detects Kvl .5 in H E K cells expressing K v l . 5 , and in canine tissues. Extracts from untransfected HEK293 cells, those expressing cloned K v l . 5 , 3.1, 1.2, and 1.4, and homogenized canine atrial (A), ventricular (V) and brain (B) tissues were subjected to Western analysis. B , Western blot of canine atrial and ventricular fractions. Membrane (M) and cytosolic (C) fractions of homogenized atria (At) and left ventricles (LV) were probed with anti-Kvl.5 (Upstate Biotechnologies). A larger band in the cytosolic fraction probably represents cross-reacting protein. C, Upstate Antibody to Kv l .5 detects a band of the expected size in a membrane extract from isolated canine atrial myocytes (labeled C A M ) , and from H E K cells (HEK/Kvl .5) heterologously expressing K v l . 5 . D, Western blot of canine atrial (At), left ventricular (LV) and rat brain (RB) extracts probed with anti-Kv3.1 (Alomone Labs). These were paired samples with those in panel C. E, Kv3.1 antibody-blocking peptide does not prevent detection of bands in canine atrium. Paired Western blots probed with anti-Kv3.1 alone in the left panel (Alomone); and with anti-Kv3.1 plus blocking peptide (1 ug/ug antibody, right). Lanes are (left to right), in each case, Kv3.1 stably expressed in H E K cells, R B extract and canine atrial extracts (CA). 105 LO —^• CO > > Q At At C B At Figure 4.2. Attempted amplification of K v l . 5 , Kv3.1, and dystroglycan from canine atrium. RT-PCR was conducted on R N A extracts from canine atria (At) and brain (CB) using oligonucleotides complementary to K v l . 5 , Kv3.1 and dystroglycan coding sequences (see online data supplement). Arrows indicate the expected positions of Kvl .5 (1), Kv3.1 (2) and dystroglycan (3) amplification products. 106 Kvl.5 is expressed on the surface of canine atrial myocytes Immunocytochemistry and confocal imaging experiments using an hKvl.5 C-terminal antibody confirmed Kvl.5 expression at the cell surface of canine atrial myocytes. This antibody was raised to a unique sequence in Kvl.5 not shared by other Kv channels. We also tested commercial Kvl.5 antibodies from Alomone Labs and the polyclonal antibody from Upstate Biotechnology. Control experiments (Figure 4.3A) showed that the C-terminal antibody did not react with untransfected HEK293 cells, but localized to the surface of dual-labeled cells expressing a T7-tagged Kvl.5 protein, as did the T7 antibody, revealing extensive co-localization (Figure 4.3A, far right). The Alomone antibody showed diffuse cytosolic staining and some faint membrane localization in these highly expressing cells (Figure 4.3C, far right). The Upstate monoclonal antibody failed to detect T7-tagged hKvl.5 at the cell surface (Figure 4.3C, far right). The Kv3.1 antibody did not stain untransfected HEK cells but detected cells transfected with Kv3.1 (Figure 4.3B far left, and left respectively). Neither Kv3.1 nor Kvl.5 antibodies stained the HEK cells transfected with the other channel (Figure 4.3B, right), and staining by both antibodies in their concordant transfected cell lines was blocked by pre-incubation with the immunogenic peptides (Figure 4.3C, left). In some cases in Figure 4.3, DAPI has been used to identify cell nuclei and render cells visible. Strong membrane staining of canine atrial myocytes with the C-terminal Kvl.5 antibody is apparent from single optical slices of two different myocytes (Figures 4.4A and 4.4C), and projection images demonstrate increased staining at cell ends near intercalated discs (Figures 4.4B and 4.4D). A myocyte cross-section confirms little detection of intracellular protein (Figure 4.4B, inset). Kvl.5 labeling was specifically inhibited by pre-incubation with the antigenic peptide (Figure 4.4E). In contrast to these bright images, the Upstate polyclonal Kvl.5 107 Figure 4.3. Detection of T7-tagged K v l . 5 and Kv3.1 in H E K cells by different antibodies. A . HEK293 cell line stably expressing T7-tagged Kvl .5 (Kvl.5 T7s) was used to validate the C-terminal (C-Term) antibody to Kv l .5 . Each panel is a single slice obtained with the confocal microscope, representative of the antibody labeling obtained, described above each panel. Size bars all represent 6.7 pm. DAPI was used on the untransfected cells (left panel) to identify cell nuclei for visibility. The merged panel denotes co-localization between the two antibodies. B. Left two panels: testing of untransfected and an HEK293 cell line stably expressing Kv3.1 (Kv3.1s) with Alomone Kv3.1 antibody. Right, Cross-reactivity of Kv l .5 C-terminal antibody in the Kv3.1s cell line. Far right, Control for cross-reactivity of the Kv3.1 antibody in the Kv l .5 cell line. C. Left two panels, Preincubation of Kvl .5 C-terminal antibody and Kv3.1 antibody with their respective antigenic peptides (1 pg/pg antibody) eliminates labeling of the stable cell lines. DAPI staining was used to identify cells in the Kv3.1 experiment. Right two panels, Testing of Upstate and Alomone antibodies to K v l . 5 . Note that the Upstate monoclonal antibody failed to reveal specific Kvl .5 staining. 108 6.67 um Figure 4.4. Kvl.5 is expressed at the surface of canine atrial myocytes. A through D, Two examples of canine myocytes labeled with the C-terminal antibody to Kvl.5. Each case (A and C) shows single confocal slices through the centre of each myocyte, whereas B and D show the full projection of the labeling across all layers of the same myocytes. Scale bars in A and C are valid for images in B and D, respectively. The inset cross section across the plane depicted in B shows plasma membrane labeling by the antibody. E, Example of canine myocytes where the antibody was preincubated with the Kvl.5 peptide before labeling. F, Canine atrial myocytes labeled with the Upstate polyclonal anti-Kvl.5. G, Canine myocytes labeled with the Alomone antibody to Kv3.1. H, Canine myocyte labeled after preincubation of antibody with the Kv3.1 peptide. 109 antibody showed only patchy nuclear staining (Figure 4.4F). The Alomone Kv3.1 antibody showed faint membrane and intracellular staining of myocytes (Figure 4.4G), but this was still present after pre-incubation with the Kv3.1 peptide (Figure 4.4H). Control experiments with only secondary antibodies did not reveal any non-specific staining of myocytes (data not shown). Pharmacological studies suggest that Kvl.5 contributes to IKUR (CI) Although biochemical and imaging experiments demonstrate that Kvl .5 is widely expressed at the surface of isolated canine atrial myocytes, they give little insight into the role of Kv l .5 in IicuR(d) current and during canine atrial action potentials. A pharmacological approach was therefore taken to understand this role. Kv l .5 and Kv3.1 differ markedly in sensitivity to extracellularly applied T E A + . Kv3.1, expressed in mammalian cells, has reported IC50S of 130 umol/L (Yue et al., 2000) and 200 umol/L (Grissmer et al., 1994), whereas blockade of Kvl .5 requires ~500-fold higher concentrations of T E A + (Grissmer et al., 1994). Thus, i f Kv3.1 is the only component of IicuR(d), the current should be blocked by 0.5 mmol/L T E A + . We also tested the effects of a new drug that blocks Kvl .5 selectively. C9356 was synthesized at Cardiome Pharma, Corp in Vancouver. It is a potent Kvl .5 blocker ((IC50 4.4 umol/L), and very selective for Kv l .5 over other channels like Kv2.1, which is quite TEA+-sensitive, Kv4.2 and Kv3.1 (Figure 4.5). Concentration-response relationships predict that 10 umol/L C9356 will block 75% to 80%o of Kvl .5 with no effect on Kv3.1 and minor effects on Kv2.1 and 4.2. 110 1 10 100 1000 C9356 concentrat ion (pM) Figure 4.5. Effects of a Kvl.5-specific drug on cloned channels in HEK293 cells. Concentration-response relationships for C9356 against representative channels from different K v classes known to be expressed in cardiac myocytes. The channels used were K v l . 5 , Kv2.1, Kv3.1, Kv4.2, hHl N a + channel, and the hERG channel. The voltage protocol used to record currents is described in the online supplement. Three to four points were used to construct the relationships and all points are the mean ± S E M of 3-13 complete determinations. Curves were fit to a standard Hi l l equation with IC50 values as stated. Hi l l coefficients were 1.2, 1.3, 1.1, 0.9, 1.1 and 1.8 for Kv l .5 , Kv2.1, Kv3.1, Kv4.2, hERG, and hHl respectively. Ill Data in Figure 4.6 shows the effects of C9356 on canine atrial K + . Both 0.5 mmol/L T E A + and C9356 reduced the sustained outward currents (Figure 4.6A), as shown by difference currents in the bottom panels, with only minor inhibition of the transient outward current. However, C9356 had a more prominent action on tail currents than did T E A + (Figure 4.7A). These effects were readily reversible on washout. Importantly, with respect to TEA+-sensitive current, two populations of cells were present, as shown in Figure 4.6B. Most cells had little TEA+-sensitive current, whereas the amount of C9356-sensitive current was more normally distributed. Overall, mean drug-sensitive difference currents were similar for the two compounds, although mean TEA+-sensitive currents were subject to greater variation (Figure 4.6C). Higher concentrations of T E A + had essentially similar effects (Figure 4.7). Control currents during clamp steps are shown in Figure 4.7A along with a rather modest action of 5 mmol/L T E A + . Addition of 10 umol/L C9356 in the presence of T E A + resulted in substantial inhibition of the sustained current in this example, as shown by drug-sensitive currents in the bottom panel. Overall, difference currents sensitive to the two drugs are similar (Figure 4.7B), but again seem not to be fully additive in individual myocytes. This finding was confirmed when the drugs were added in a different order (Figure 4.7C and 4.7D). Addition of a second compound has a lesser effect on the sustained outward current than that produced by the first compound. One obvious difference between the two drugs is that the TEA+-sensitive currents are activated at ~20 mV more positive potentials than the C9356-sensitie current (Figure 4.7B). 112 Control 10 uM C9356 u. 5 -B I 4H | 3^ o § 2\ 1 | 1 "5 A A C9356 + 500nM TEA Wash C9356 -(C9356+TEA) Wash-(C9356+TEA) •3210-canm* a pp D • A O n „ A A A • u • A • A A A A TEA C9356 0.5 mM 10 uM 2.5n f 2.0 Q. c" 1.5-1 I 1.0-1 n o § 0.5 I i o.o^ - ° - 1 0 n M C9356 (n=15) - o - 0 . 5 m M TEA (n=14) - A - C9356 + 0.5mM TEA (n=12 ) -40 -20 0 20 40 60 test potential, mV 80 Figure 4.6. Effects of 0.5 mmoi/L T E A + and C9356 on canine atrial currents. A , Representative currents during 250-ms clamp steps recorded in presence of 10 pmol/L C9356, C9356+0.5 mmol/L T E A + , and upon washout. Difference currents showing effects of each compound are in the lower row, as is the voltage-clamp protocol. B, Scatter plot illustrates heterogeneity of currents sensitive to T E A + , versus normally distributed C9356-sensitive current densities. C, Difference currents (final 50 ms of voltage steps) in response to 10 pmol/L C9356, 0.5 mmol/L T E A + , or the combination. Numbers of myocytes in each group are indicated in the legend. (Data collected and analyzed by M . Lamorgese, L. Castel and D. Van Wagoner) 113 Control g. 2.0 a * 1.5 i 1.0 £ 0.5 0.0-^  - • -10 M M C9356(n=15) - A - 5 mM TEA (n=9) - A - C9356+5 mM TEA (n=13) | D t 8 2 s ra TEA TEA + C9356 wash treatment -40 -20 0 20 40 test potential 60 80 Last control C9356 C9356+TEA Wash treatment Figure 4.7. Effects of 5 mmol/L T E A + and C9356 on K + currents in canine atrial myocytes. A , Representative sequential traces showing the effect of 5 mmol/L T E A + , T E A + plus C9356 (10 umol/L) and washout in the upper series. Difference currents (TEA+-sensitive, C9356-sensitive, and TEA ++C9356 sensitive) are shown in the bottom row. Dotted lines in all panels represent zero current. B, Difference current-voltage relations measured during the final 50 ms of the 250 ms voltage step for myocytes exposed first to either 10 umol/L C9356 or 5.0 mmol/L T E A + , or the combination of these agents. C, Late step (final 50 ms) current remaining following exposure to C9356, C9356 plus 5 mmol/L T E A + , and following washout (n=7 myocytes). D, Late step (final 50 ms) current remaining following exposure to 5 mmol/L T E A + , 5 mmol/L T E A + plus 10 umol/L C9356, and after washout (n=6 myocytes). Asterisks represent normalized values significantly different from control: * P<0.05. (Data collected and analyzed by M . Lamorgese, L. Castel and D. Van Wagoner) 114 C9356 has substantial effects on the canine atrial action potential (Figure 4.8). Control APs are shown at 30, 60 and 120 bpm (Figure 4.8A). C9356 effects alone and in combination with 0.5 mmol/L T E A + are shown at these stimulation rates in Figures 4.8B through 4.8D. In every case, C9356 markedly prolonged the plateau (rather than late repolarization) of the canine AP, whereas further addition of 0.5 mmol/L T E A + was without effect. A l l drug effects were reversible on washout, as shown. Rate-dependent effects of C9356 (in the absence of TEA) on the canine AP are summarized in Figure 4.8E. Action potential duration was measured at 20%, 50% and 90%> repolarization, as shown in the three rows of the graph. APD50 is the parameter most affected by the compound, consistent with a block of sustained delayed rectifier K + current in these cells. 115 2sbcl 1s 0.5s Cycle length Figure 4.8. Effects of C9356 and TEA + on canine atrial action potentials. Action potentials in each panel are shown at three cycle lengths, 2, 1, and 0.5 seconds. A , Control action potentials at each cycle length. B, After a control action potential at the 2-second basic cycle length (2 s bcl), 10 pmol/L C9356 was applied. At steady state, 500 pmol/L T E A + was added in the continued presence of C9356. After 4 minutes in the presence of both compounds, they were washed off and a postcontrol action potential was recorded. C and D, Same experiment as in B except that action potentials were recorded at a 1-second cycle length in C and 0.5-second cycle length in D. Similar results are seen in 9 cells. Dotted lines in all panels represent zero current. E, Steady-state analysis of action potentials recorded at 0.5-, 1- or 2-second basic cycle lengths. Action potential duration was determined at 20% (APD20), 50% (APD50) and 90% (APD90) repolarization (n=9 in each case). Asterisks reflect values significantly different from control: * PO.05; ** PO.01. (Data collected and analyzed by M . Lamorgese, L. Castel and D. Van Wagoner) 116 DISCUSSION Detection of Kvl.5 in canine cardiac myocytes. There is little doubt from the data presented in the Figures 4.1 through 4.4 that Kvl.5 was amply expressed in our samples of canine heart. Kvl.5 message was readily detectable in canine atrium (Figure 4.2), and protein was found in membrane fractions of homogenized whole atrial or ventricular heart tissue from a variety of different sources and in isolated canine atrial myocytes (Figure 4.1) using two different antibodies. Immunofluorescence experiments using our own C-terminal antibody confirmed the presence of Kvl.5 on the cell membrane and concentrated at the intercalated discs (Figure 4.4). This antibody was independently validated against T7-tagged Kvl.5 expressed in HEK293 cells (Figure 4.3) and also detected Kvl.5 in Western blots (Figure 4.1 A). Others have shown Kvl.5 is present at the intercalated discs in human atrial myocytes and is also expressed in a punctuate fashion at the cell surface in newborn tissue (Mays et al., 1995). This increases our confidence that the C-terminal antibody correctly labels Kvl.5 in canine atrium. Several commercial antibodies were variably successful in detecting Kvl.5 at the surface of HEK cells and canine cardiac myocytes. The Upstate Biotechnologies' polyclonal antibody was successful in detecting Kvl.5 in Western blots, but stained primary nuclei in canine myocytes (Figure 4.4F). The Alomone antibody was unsuccessful in Western blots as well but showed faint staining in transfected HEK cells that could be attributed to Kvl.5 in these cells expressing Kvl.5 highly (Figure 4.3C). Because of its poor activity against Kvl.5, no staining was attempted in canine myocytes. This low activity may provide an explanation for the failure of Yue et al. (2000) to detect Kvl.5 in canine myocytes, because they used the Alomone antibody for their Western blot experiments. It is also possible that this antibody, although weakly detecting human Kvl.5, does not detect canine Kvl.5. 117 Failure to detect Kv3.1 using molecular methods Previous reports on the presence of Kv3 channels in canine heart have been conflicting. Dixon et al. (1996) reported that Kv3.1, Kv3.2 and Kv3.3 gene transcripts were all expressed at negligibly low levels in canine hearts, with the exception of Kv3.4 which was weakly detectable in canine ventricle (Dixon et al., 1996). On the other hand, the channel was clearly present in the cardiac tissues tested by Yue et al., (2000) and a range of pharmacological experiments, as well as actual molecular cloning of the channel supported its presence in canine atrium (Yue et al., 1996a; Yue et al., 2000). However, in agreement with Dixon et al. (1996) we could find no definitive trace of it in any of the dog hearts we tested. Although the Kv3 subfamily consists of only four genes (Kv3.1, Kv3.2, Kv3.3, and Kv3.4), each gene may encode multiple products by alternative splicing of 3' ends (Vega-Saenz de Miera et al., 1994; Weiser et al., 1994; Rudy et al., 1999). Several alternative Kv3 transcripts are known to be expressed in neuronal populations within the central nervous system, all of which can have different sizes (Ponce et al., 1997; Gelband et al., 1999). For Kv3.1, these seem to be between 85 and 97 kDa. In heart, the band detected by Yue et al. (2000) was 97 kDa, but the short cytosolic/microsomal bands detected in our experiments (Figure 4.ID and 4.IE) in canine atrium were <75 kDa and remained after incubation with the antigenic peptide (Figure 4.IE), and so are unlikely to be related to the presence of Kv3.1. That we were unable to immunoprecipitate either band from myocytes strengthens the case that these are not Kv3.1 isoforms. Finally, in atrial myocytes, immunofluorescent labeling of Kv3.1 was equivocal at best (Figure 4.4G and 4.4H). Pharmacological properties of IKUR((I) and prolongation of the canine atrial action potential Our interpretation of the molecular nature of the sustained atrial delayed rectifier current that we have recorded from canine atrial myocytes is more complicated. The current and also 118 the APD50 were sensitive to C9356, which selectively blocks Kvl .5 (Figures 4.5 through 4.8). We conclude that Kvl .5 mRNA and protein are not only present our canine atrial myocytes but that Kvl .5 contributes to membrane currents and action potential repolarization. A recent study has also shown that a large component of IKURW in adult canine atrium is resistant to 5 mmol/L T E A + (Dun et al., 2003), and our data suggest that this component may well be K v l . 5 . The IKUR(CI) in some atrial myocytes was also partially sensitive to 0.5 mmol/L T E A + , a concentration that will block most Kv3.1 channels, in agreement with reports of such a current in canine atrium (Yue et al., 1996a). Higher concentrations of T E A + (5 mmol/L, Figure 4.7) had little additional effect on outward current, which supports a role for Kv3.1 in these myocytes. Another candidate channel that is relatively sensitive to T E A + is Kv2.1. Although Kv2.1 has not been reported to be expressed in canine heart, only ventricle has been tested (Dixon et al., 1996). It is difficult to definitively identify the TEA+-sensitive current that we have observed as Kv3.1 because we were unable to detect the channel using molecular methods in several different canine heart preparations in two separate laboratories. However, there was considerable variation in the presence of TEA+-sensitive current (Yue et al., 1996b), so the tissues that we used for biochemical analysis may have lacked Kv3.1. Cells used for patch-clamp experiments were used for up to 6 to 8 hours after isolation, whereas the tissue for biochemistry was immediately flash frozen. It is possible that the expression of the two channels could change over time or be modulated by factors such as the age of the animals (Dun et al., 2003). Both our subject animals and those of Yue, et al. (2000) were mongrels of indefinite age and breed differences in ion channel expression may exist. Alternatively, presently unknown differences in channel distribution in various regions of the atrium might explain the differences between results from the two groups. 119 Implications Kv l .5 is a prominent repolarizing current in human atria. In atrial myocytes from patients with valvular disease and persistent AF , we observed that the sustained K + current was reduced by ~50%, and there was a parallel reduction in the expression of Kvl .5 protein (Van Wagoner et al., 1997). A similar conclusion was reached by Brandt and colleagues (2000). Although the Kv l .5 protein and functional K + currents are reduced in A F , this does not imply that drugs that suppress this current would not be effective in prolonging atrial refractoriness, because there is typically an even greater reduction (60% to 70%>) in the density of the L-type C a 2 + current (Van Wagoner et al., 1999). Thus, it is possible that drugs that suppress Kvl .5 currents might be more, rather than less, effective in AF patients than in controls. Studies designed to evaluate this hypothesis need to be performed in an appropriate model system. This study suggests that the canine model, already well characterized from an electrophysiological perspective, may be useful in evaluating the efficacy and utility of this new therapeutic approach. It should be noted that the electrophysiological role of the Kvl .5 expressed in canine ventricle (Figure 4.1 A) remains to be clarified. Conclusions Kvl .5 protein and messenger RNA can be clearly detected in the atria of dogs, using Western blotting, RT-PCR, and specific Kvl .5 immunofluorescence. In addition, a sustained delayed rectifier current [Ii<UR(d)] is detectable in canine atria that can be pharmacologically attributed, at least in part, to K v l . 5 . This current seems to have a functional role in canine atrial repolarization, as blockade leads to significant action potential prolongation. Our results suggest that canine models of atrial disease may prove valuable in the study of the corresponding human pathologies. 120 A C K N O W L E D G E M E N T S The authors acknowledge project support to D.F. from the Canadian Institutes of Health Research and the Heart and Stroke Foundation of British Columbia and Yukon, and to D.V.W. from the NIH (National Heart, Lung, and Blood Institute, 1R01-HL-65412). They thank Paul Murray, for access to canine tissues, and Dan Minor and Bruce Tempel for the Kv l .2 clones. Jie Liu provided molecular biology support, Shunping Lin provided HEK293 electrophysiology, and Anu Khurana prepared the cells. R E F E R E N C E S Barry D M , Nerbonne JM. (1996). Myocardial potassium channels: Electrophysiological and molecular diversity. Annu Rev Physiol 58:363-394. Bou-Abboud E, Nerbonne JM. (1999). Molecular correlates of the calcium-independent, depolarization-activated K + currents in rat atrial myocytes. J Physiol (Camb ) 517:407-420. Boyle WA, Nerbonne JM. (1992). Two functionally distinct 4-aminopyridine-sensitive outward K + currents in rat atrial myocytes. J Gen Physiol 100:1041-1067. Brandt M C , Priebe L , B6hle T, Sudkamp M , Beuckelmann DJ. (2000). The ultrarapid and the transient outward K + current in human atrial fibrillation. Their possible role in postoperative atrial fibrillation. J Mol Cell Cardiol 32:1885-1896. Deal K K , England SK, Tamkun M M . (1996). Molecular physiology of cardiac potassium channels. Physiol Rev 1996;76:49-67. Dixon JE, Shi W M , Wang HS, McDonald C, Y u H , Wymore RS, Cohen IS, McKinnon D. (1996). Role of the Kv4.3 K + channel in ventricular muscle - A molecular correlate for the transient outward current. Circ Res 79:659-668. Dun W, Yagi T, Rosen MR, Boyden PA. (2003). Calcium and potassium currents in cells from adult and aged canine right atria. Cardiovasc Res 58:526-534. Fedida D, Wible B, Wang Z, Fermini B, Faust F, Nattel S, Brown A M . (1993). Identity of a novel delayed rectifier current from human heart with a cloned K + channel current. Circ Res. 73:210-216. Feng JL, Wible B, L i GR, Wang ZG, Nattel S. (1997). Antisense oligodeoxynucleotides directed against Kvl .5 mRNA specifically inhibit ultrarapid delayed rectifier K + current in cultured adult human atrial myocytes. Circ Res 80:572-579. 121 Gelband C H , Warth JD, Mason HS, Zhu M , Moore JM, Kenyon JL, Horowitz B, Sumners C. (1999). Angiotensin II type 1 receptor-mediated inhibition of K + channel subunit Kv2.2 in brain stem and hypothalamic neurons. Circ Res 84:352-359. Gidh-Jain M , Huang B, Jain P, El-Sherif N . (1996). Differential expression of voltage-gated K + channel genes in left ventricular remodeled myocardium after experimental myocardial infarction. Circ Res 79:669-675. Grissmer S, Nguyen A N , Aiyar J, Hanson DC, Mather RJ, Gutman GA, Karmilowicz MJ, Auperin DD, Chandy K G . (1994). Pharmacological characterization of five cloned voltage-gated K + channels, types K v l . l , 1.2, 1.3, 1.5, and 3.1, stably expressed in mammalian cell lines. Molec Pharmacol 45:1227-1234. Hume JR, Uehara A , Hadley RW, Harvey RD. (1990). Comparison of K + channels in mammalian atrial and ventricular myocytes. In: Potassium Channels: Basic function and Therapeutic Aspects. Progress in Clinical and Biological Research, vol. 334. Colatsky TJ, ed. Wiley-Liss, New York. Kaab S, Dixon J, Due J, Ashen D, Nabauer M , Beuckelmann DJ, Steinbeck G, McKinnon D, Tomaselli GF. (1998). Molecular basis of transient outward potassium current downregulation in human heart failure: a decrease in Kv4.3 mRNA correlates with a reduction in current density. Circulation 98:1383-1393. Knobloch K , Brendel J, Peukert S, Rosenstein B, Busch A E , Wirth KJ . (2002). Electrophysiological and antiarrhythmic effects of the novel Ticur channel blockers, S9947 and S20951, on left vs. right pig atrium in vivo in comparison with the IKT blockers dofetilide, azimilide, d,I-sotalol and ibutilide. Naunyn-Schmiedebergs Archives of Pharmacology 366:482-487. Leeb T, Neumann S, Deppeg A, Breen M , Brenig B. (2003). Genomic organization of the dog dystroglycan gene DGA1 locus on chromosome 20ql5.1-ql5.2. Genome Res 10:296-301. Maruoka ND, Steele DF, Au BPY, Dan P, Zhang X , Moore EDW, Fedida D. (2000). a-Actinin-2 couples to cardiac Kvl .5 channels, regulating current density and channel localization in H E K cells. FEBS Letts 473:188-194. Mays DJ, Foose JM, Philipson L H , Tamkun M M . (1995). Localization of the Kvl .5 K + channel protein in explanted cardiac tissue. J Clin Invest 96:282-292. Nerbonne JM. (2000). Molecular basis of functional voltage-gated K + channel diversity in the mammalian myocardium. J Physiol (Camb ) 525:285-298. Ponce A, De Miera EVS, Kentros C, Moreno H , Thornhill B, Rudy B. (1997). K + channel subunit isoforms with divergent carboxy-terminal sequences carry distinct membrane targeting signals. J Membrane Biol 159:149-159. Rudy B, Chow A , Lau D, Amarillo Y , Ozaita A , Saganich M , Moreno H , Nadal MS, Hernandez-Pineda R, Hernandez-Cruz A , Erisir A , Leonard C, De Miera EVS. (1993). Contributions of Kv3 channels to neuronal excitability. Ann NY Acad Sci 868:304-343. 122 Snyders DJ, Tamkun M M , Bennett PB. (1993). A rapidly activating and slowly inactivating potassium channel cloned from human heart. Functional analysis after stable mammalian cell culture expression. J Gen Physiol 101:513-543. Van Wagoner DR. (2000). Pharmacologic relevance of K + channel remodeling in atrial fibrillation. J Mol Cell Cardiol 32:1763-1766. Van Wagoner DR, Pond A L , Lamorgese M , Rossie SS, McCarthy P M , Nerbonne JM. (1999). Atrial L-type C a 2 + currents and human atrial fibrillation. Circ Res 85:428-436. Van Wagoner DR, Pond A L , McCarthy P M , Trimmer JS, Nerbonne JM. (1997). Outward K + current densities and Kvl .5 expression are reduced in chronic human atrial fibrillation. Circ Res 80:772-781. Vega-Saenz de Miera E, Weiser M , Kentros C, Lau D, Moreno H, Serodio P, Rudy B. (1994). Shaw-related K + channels in mammals. In: Membrane channels: molecular and cellular physiology. Peracchia C, ed. Academic Press, San Diego. Wang Z, Fermini B, Nattel S. (1993). Sustained depolarization-induced outward current in human atrial myocytes: Evidence for a novel delayed rectifier K + current similar to Kvl .5 cloned channel currents. Circ Res 73:1061-1076. Weiser M , Vega-Saenz de Miera E, Kentros C, Moreno H, Franzen L, Hillman D, Baker H, Rudy B. (1994). Differential expression of S/zaw-related K + channels in the rat central nervous system. J Neurosci 14:949-972. Yue L X , Feng JL, L i GR, Nattel S. (1996). Characterization of an ultrarapid delayed rectifier potassium channel involved in canine atrial repolarization. J Physiol (Camb ) 496:647-662. Yue L X , Feng JL, L i GR, Nattel S. (1996). Transient outward and delayed rectifier currents in canine atrium: Properties and role of isolation methods. Am J Physiol Heart Circ Physiol 270:H2157-H2168. Yue L, Wang ZG, Rindt H , Nattel S. (2000). Molecular evidence for a role of Shaw (Kv3) potassium channel subunits in potassium currents of dog atrium. J Physiol (Camb ) 527:467-478. 123 C H A P T E R 5: K v l . 5 and caveolin-3 do not interact or colocalize in rat or canine cardiac myocytes. Jodene Eldstrom1*, Qixia Y u 1 , Garnet Martens2, David Van Wagoner3, Edwin Moore 1, David Fedida1 'Department of Physiology, University of British Columbia, 2146 Health Sciences Mall, Vancouver, British Columbia, Canada V6T1Z3;2UBC Bio-Imaging Facility, University of British Columbia, 6270 University Boulevard, Vancouver, British Columbia, Canada V6T1Z4 3Department of Cardiology, Cleveland Clinic Foundation, Cleveland, OH 44195, USA. * Contributions made by Jodene Eldstrom: Responsible for project conception, planning, tissue sample preparation, immunoprecipitations, E M tissue preparation and immunolabeling (with the exception of cryosectioning) and T E M imaging, sucrose fractionation and western blots, some immunocytochemistry, all the writing of the manuscript, editing and figure preparation. (Percent contribution: 90%) 124 I N T R O D U C T I O N Studies in cultured mouse Itk- cells (Martens et al., 2000; Martens et al., 2001), tsA201 cells, mouse brain (Wong and Schlichter, 2004), pancreatic |3-cells (Xia et al., 2004) and Jurkat T-lymphocytes (Bock et al., 2003) have shown that several subtypes of voltage-gated potassium (Kv) channels may be sorted to specialized lipid domains called lipid rafts. The significance of this sorting of specific subtypes of K v channels to lipid rafts is that it provides a potential mechanism for the differential regulation of targeted channels, as lipid rafts are thought to locally concentrate signaling molecules for more efficient transduction. One particular subgroup of lipid rafts of particular interest in myocytes is caveolae, which are flask-like invaginations of the surface membrane containing the muscle-specific caveolin-3. Here, we have studied K v l . 5 , a channel that is highly expressed in human atrial myocytes (Fedida et al., 1993) and is the molecular correlate of IKUR (Fedida et al., 1993; Feng et al., 1997). IKUR is an important contributor to the repolarization phase of the atrial action potential, but appears to have little electrophysiological role in the ventricle. Therefore, Kv l .5 has generated significant interest as an atrial-specific target for the treatment of atrial fibrillation (Van Wagoner, 2000; Fedida et al., 2003). Obviously, it would be of great interest to know whether Kvl .5 localizes to lipid rafts, and more specifically caveolae in association with caveolin-3, in normal mammalian cardiac myocytes where the proteins are expressed at physiological levels. Caveolae are characterized by a coat of caveolin-1 or caveolin-3, and a third type of caveolin, caveolin-2, although studies of the subtypes have indicated that caveolin-2 requires co-expression of caveolin-1 or -3 in order to express in the plasma membrane (Li et al., 1998). Mice lacking caveolin-3 show a defective T-tubule system and changes in the dystrophin-glycoprotein complex distribution associated with the loss of caveolae in skeletal muscle (Galbiati et al., 2001) and a progressive hypertrophic cardiomyopathy (Woodman et al., 2002). 125 Over-expression of caveolin-3 leads to severe cardiac tissue damage and fibrosis and a prolonged QRS duration in transgenic mice (Aravamudan et al., 2003). Several proteins are associated with these domains including c-Src, G-protein ot-subunits (Li et al., 1996b), H-Ras (Song et al., 1996) and eNOS (Feron et al., 1996) as well as various receptors (Lasley et al., 2000; Fujita et al., 2001) including the P2-Adrenergic receptor (Xiang et al., 2002). Caveolae have been implicated in the regulation of targeted signal cascades (reviewed in Lasley and Smart, 2001; van Deurs et al., 2003), in clathrin-independent endocytosis (reviewed in Nabi and Le, 2003), calcium signaling (reviewed in Isshiki and Anderson, 2003), and in cholesterol trafficking (reviewed in van Deurs et al., 2003). Caveolins may also stabilize the caveolar invagination and slow the dynamin-dependent internalization process (Nabi et al., 2003). Significantly, many of the same signaling molecules known to localize in caveolae affect K v l . 5 . There is evidence that Kvl .5 can be modulated by protein kinases such as Src (Nitabach et al., 2002), and P K A (Mason et al., 2002) and I K UR is influenced by P2-adrenergic stimulation (Li et al., 1996a). Localization to caveolae appears to be a dynamic process with several receptors, such as the P2-adrenergic receptor (Dupree et al., 1993) and the muscarinic acetylcholine receptor (Feron et al., 1997) entering caveolae upon agonist stimulation; receptors such as the adenosine A i receptor (Lasley et al., 2000) translocate out of caveolae upon activation. It is conceivable that regulation of Kvl .5 could similarly involve regulated localization of the channel or localization of molecules that regulate the channel. Intriguingly, Kvl .5 may be inactivated by phosphorylation (Nitabach et al., 2002), and Src kinase, which has been shown to regulate Kvl .5 activity, appears to translocate out of caveolae upon activation (Smythe et al., 2003). We have used biochemical methods, wide-field microscopy combined with deconvolution techniques and immunoelectron microscopy to investigate the spatial localization 126 of Kvl .5 and caveolins in rat and canine cardiac myocytes. We show that Kvl .5 does not directly interact with the major caveolar coat protein, caveolin-3, nor does the channel co-localize appreciably with this marker for caveolae in myocytes, and we were unable to detect the channel in lipid raft fractions from cardiac or cultured cell membranes. M A T E R I A L S A N D M E T H O D S Preparation of canine heart lysates Canine atrial and ventricular tissue was obtained from heartworm-negative mongrel dogs (20-30 kg) as previously described (Fedida et al., 2003). Dogs were anesthetized with intravenous pentobarbital sodium (30 mg/kg) and fentanyl citrate (15 pg/kg) and placed on positive-pressure ventilation. A catheter was inserted into the right femoral artery, and the dogs were exsanguinated. A left lateral thoracotomy was performed through the fifth intercostal space, and the heart was arrested. A l l surgical procedures and experimental protocols were approved by the Cleveland Clinic Foundation Institutional Animal Care and Use Committee (Cleveland, OH). For rat samples, adult male Wistar rats weighing between 250-300 grams were sacrificed with an overdose of sodium pentobarbital (65 mg/kg body weight) and 2 units of heparin sodium USP. These procedures were approved by the U B C Committee on Animal Care (Vancouver, BC). The freshly excised ventricular or atrial tissue was added to ice cold non-denaturing lysis buffer and homogenized as described in Fedida et al., (2003) and lysates were processed as described previously (Eldstrom et al., 2003). Following transfer to PVDF membranes the samples were probed with either rabbit anti-Kvl.5 (1:10,000; developed in our lab; Fedida et al., 2003), mouse anti-Caveolin-3, rabbit anti-Caveolin-1, mouse anti-(3-dystroglycan (1:10,000; BD Bioscience), mouse anti-Caveolin-2 (1:5000;BD Bioscience), rabbit anti-eNOS (1:5000; Chemicon Int.), or mouse anti-actinin (1:10,000, Sigma). HRP conjugated goat anti-rabbit IgG or 127 sheep anti-mouse (1:10,000; Jackson Laboratories) were used as secondary antibodies for detection using a chemiluminescent reagent (Western Lightning, Perkin Elmer). Myocyte preparation Atrial myocytes were dissociated from canine atrial specimens using a chunk dissociation technique (Fedida et al., 2003), and rat myocytes using methods described by Scriven et al (2000) except that when the hearts began to soften, atria were removed separately, and triturated in physiological saline. Canine ventricular myocytes were isolated using the segmental coronary artery perfusion technique (Hohl and Altschuld, 1991). Immunolabeling, and wide-field microscopy Myocytes, stored in PBS-azide solution after fixation (Fedida et al., 2003), were plated onto poly-L-lysine coated coverslips. For rat myocytes, after blocking, the rabbit polyclonal Kv l .5 primary antibody, diluted in antibody buffer (150mM NaCl, 15mM Na3 Citrate, 3mM NaN 2 with 1% BSA, 2% (v/v) goat serum and 0.05% (v/v) Triton-X), was incubated with the cells first, followed by secondary antibody conjugated with fluorescein isothiocyanate (FITC); then the mouse anti-caveolin-3 monoclonal antibody, followed by a Texas Red secondary antibody. A l l primary antibodies were incubated overnight; the secondary antibodies were incubated for 1.5 hours. Immunolabeling of canine samples was as previously reported (Fedida et al., 2003). Images were acquired using a Nikon Diaphot 200 inverted microscope. 2D images (magnification, x600) acquired from a thermoelectrically-cooled CCD camera through the whole cell at 0.25 um intervals were stacked into 3D images after subtraction of dark current and background, and deconvolved using an empirically determined point-spread-function (Scriven et al., 2000). Dual fluorescence beads were added to mounting media and used to guide the overlapping of two images acquired at different emission wavelengths. The threshold intensity 128 was applied to remove the noise on images before analyzing the data. Data were presented as mean ± s. e. mean. N refers to the number of cells collected from 3-4 animals. Preparation of Tokuyasu sections and immunolabeling After overdose with Pentobarbitol, rats were perfused via the right ventricle with 10 ml of cold PBS using a peristaltic pump before a 10 minute perfusion with 4% paraformaldehyde in 0.1 M phosphate buffer (PB). After removal of the heart, the tissue was cut into smaller sections under fixative. After a 50 mM buffered glycine aldehyde block the small tissue pieces were infiltrated (37 °C with constant rotation) with increasing concentrations of buffered gelatin up to 12% gelatin. The gelatin was allowed to harden at 4° C and the tissue pieces were cut out and placed into 2.3 M sucrose in 0.1 M PB and allowed to rotate overnight at 4°C. Cryoprotected tissue was mounted onto cryo-pins and frozen in liquid nitrogen. Thin sections were cut using a Leica Ultracut T (Leica Microsystems, Canada) and a 35° cryo-diamond knife (Diatome). Ribbons were picked up in a drop of 1% methyl cellulose and 1.15 M sucrose, allowed to thaw on copper grids and immediately placed section side down onto a plate of hardened 2% buffered gelatin for short-term storage. For immunolabeling of Tokuyasu sections (adapted from (Slot et al., 1991) grids were floated on drops of a) 0.1 M PB (20 min); b) 1% glycine (5 min); c) PB (3X 3min); d) PBS/3% B S A (10 min); e) mouse anti-Caveolin-3 and rabbit anti Kv l .5 antibody diluted 1/10 and 1/120 in PBS/3% B S A (30 min); f) PB (5 X 3 min); g) PBS/3% B S A (5 min) h) anti-mouse IgG-10 nm gold (Sigma) and anti-rabbit IgG-5 nm gold (20 min; Ted Pella Inc.); i) PB (5X3 min); j) drl^O (10 X 3 min); k) 2:8 ratio of 2% uranyl acetate (UA) to 4% methyl cellulose (MC) on ice (5 min), excess U A / M C was wicked away and samples air dried. Images were acquired using a Hitachi H7600 Transmission Electron Microscope. Sucrose gradient raft fractionation HEK293 cells, stably expressing K v l . 5 , from five ~95 % confluent T75 Culture Flasks 129 were suspended in 5 ml of ice-cold Homogenization Buffer (150 mM NaCl, 20 mM Na2PO"4, 2 mM NafbPC^, 20% (w/v) Glycerol) and homogenized in a Dounce Homogenizer on ice. The homogenate was then sonicated (Vibra-Cell Ultrasonic Processor, Sonics and Materials Inc) 3-30 second pulses with 1 minute cooling periods between. The homogenate was then centrifuged at 10,000 rpm for 11 minutes at 4°C (SW55 Rotor) to pre-clear the sample of nuclei and debris. The supernatant was then centrifuged at 55,000 rpm for 33 minutes at 4°C to isolate the membrane fraction. The supernatant from this spin was saved as the cytosolic fraction. The membrane pellet was resuspended and incubated 15 minutes in ice-cold solubilization buffer (25 mM Mes pH 6.5, 150 mM NaCl, 1% TritonX-100 plus protease inhibitors). Solubilized membrane (2.0 ml) was transferred to a 12 ml ultracentrifuge tube and 2.0 ml of ice-cold 80% sucrose in Mes-Buffered Saline (MBS; 25 mM Mes pH 6.5, 150 mM NaCl) was added and mixed thoroughly. 4.0 ml of 30%> sucrose in MBS was layered gently on top of the first layer and then 4 mis of 5%> sucrose in MBS above the 30%>. The samples were then spun at 35,000 rpm for 12.5 hours at 4°C in a SW41 Ti Rotor. 12-1 ml fractions were collected from the top of the gradient and the pellet was washed extensively with PBS before being solubilized in MBS with 2% SDS. The re-suspended pellet was centrifuged at 21,000 X g and the supernatant run with the rest of the gradient fractions on a 10 % SDS-PAGE gel as the fractionation pellet (FP). The fractions were probed with mouse anti-human transferrin receptor (Zymed Laboratories), mouse anti-Flotillin-1 (BD Transduction) and rabbit anti-Kvl.5 antibodies. 130 R E S U L T S Kvl.5 and Caveolin-3 are found in rat and canine cardiac tissue We have used several approaches to investigate the putative interaction of Kv l .5 and caveolins in cardiac myocytes. Since caveolin-3 is found predominantly in muscle tissue (Scherer et al., 1997), we initially attempted Western blotting to determine the presence of K v l . 5 , caveolin-2, and caveolin-3 in samples prepared from rat and canine atrium and ventricle (Figure 5.1). The rat atrium and ventricular samples are shown to the left of the canine samples and Kvl .5 is apparent as faint single bands at around 83 kDa in all lanes. In contrast, caveolin-3 in the middle panels was highly expressed, also in all samples. Caveolin-2 (Figure 5.1, bottom) could only be detected in rat atrial and ventricular samples. In order to ensure that the caveolin-2 antibody could detect protein in other cell systems, we used mouse Itk- cells as a positive control ( L M band in Figure 5.1, bottom panels), as fibroblasts are known to express caveolin-2 (Scherer et al., 1997). Kvl.5 and Caveolin-3 do not co-immunoprecipitate with each other Having established the presence of Kvl .5 and caveolins in the tissue lysates, co-immunoprecipitations were carried out to determine i f Kv l .5 directly interacts with caveolin-3 in these tissues (Figure 5.2). Experiments were carried out in rat and canine atrium and ventricle and the results are presented in the same manner as in Figure 5.1, with the rat samples shown to the left (Figure 5.2). In Figure 5.2A (upper panels), antibody to Kvl .5 efficiently immunoprecipitated Kvl .5 in both rat and canine atrial and ventricular tissue, as indicated by the dark bands on the blot probed with the Kvl .5 antibody. However, antibody to caveolin-3 failed to co-immunoprecipitate K v l . 5 , as did control reactions lacking a precipitating antibody (-Ab lanes). A similar result was obtained with the reverse experiment (Figure 5.2A, lower panels). 131 RA RV CA cy 16.5-1 16.5-1 LM RA RV CACV Figure 5.1. Kvl.5 and Caveolin-3 are expressed in rat and canine atrial and ventricular tissues. Homogenates were prepared from atrial or ventricular tissue and immunoblotted with anti-Kvl.5 (upper panels), anti-caveolin-3 (middle panels), or anti-caveolin-2 (lower panel) as indicated to the left of the blots (see Methods). (RA-Rat Atrial Lysate; RV-Rat Ventricular Lysate; C A -Canine Atrial Lysate; C V - Canine Ventricular Lysate; LM-Mouse Itk- Cells). 132 Here, antibody to caveolin-3 clearly immunoprecipitated caveolin as shown in lanes 4 and 6 (left panels) and lanes 3 and 6 (right panels), but failed to bring down with them detectable levels of Kvl .5 in the rat or canine samples, as indicated by the lack appropriate bands in the lanes probed with the anti-Kvl.5. Since we were unable to demonstrate co-immunoprecipitation of the Kvl .5 and caveolin-3, we carried out a number of control experiments with other proteins to ensure that it was possible to coimmunoprecipitate known interacting proteins with Kvl .5 and caveolin-3 using our protocols and tissue (Feron et al., 1996; Maruoka et al., 2000; Sotgia et al., 2000). In the first set of controls (Figure 5.2B), immunoprecipitations of Kvl .5 were probed with an antibody to sarcomeric a-actinin and a band corresponding to actinin was found in each of the samples from rat and canine heart (Figure 5.2B, upper panels). Antibody to caveolin-3 co-immunoprecipitated eNOS in the rat ventricle (left lower panel) and co-immunoprecipitated P-dystroglycan from canine atrial lysates (right lower panel). None of these proteins were detected in the antibody-free controls (Figure 5.2). K v l . 5 and Caveolin-3 show minimal co-localization in deconvolved fluorescent images The co-immunoprecipitation experiments above show that Kvl .5 does not directly interact with caveolin-3. Having established this, we went on to examine whether the channel is present in the caveolar domain, as the channel could localize to caveolae without directly interacting with caveolins. To test this possibility, we looked at the distribution of caveolin-3 and Kvl .5 in myocytes using wide-field microscopy and deconvolution techniques. Data in Figure 5.3 shows representative samples of Kvl .5 and caveolin-3 immunolabeled rat (Figure 5.3A) and canine (Figure 5.3B) myocytes. In these images caveolin-3 is visualized with Alexa594 (red) and Kvl .5 with Alexa499 (green). In the rat atrial (top) and ventricular samples (bottom) it can be seen that caveolin-3 is highly expressed across the surface of the myocytes, as 133 Rat Cardiac Tissue -Ab RA RV Kv1.5 Probed 175— 8 3 -RA RV Kv1.5 Cav3 Kv1.5Cav3 6 2 -RA RV Kv1.5Cav3 Kv1.5 Cav3 32.5- i Cav-3 Probed B RA RA -AJ? KY.1.5 A • 1 7 5 _ (t-Actmin Probed 8 3 — 62-eNOS -Ab Cav3 eNOS 175-Probed 83-Canine Cardiac Tissue CA CV 175-83--Ab Kvl.5 Cav3 -AbKy1.5Cay3 • • • • • • • B l 175-1 83--Ab Kv1.5 Cav3 -At K y l 5 Qav3 32.5-1 CA cv Ab Kvl.5 -Ab Kvl.5 175- 175-83- 83-CA 47.5-P-Dystroglycan Probed -Ab elMDS Cav3 Figure 5.2. Kv l . 5 and Caveolin-3 fail to co-immunoprecipitate from rat and canine atrial and ventricular homogenates. (A) Western blots of Kvl.5 and caveolin-3 irnmunoprecipitations from rat (left side of figure) and canine (right side of figure) cardiac tissue. The precipitating antibodies are indicated above each lane and the antibodies used to probe the blots are indicated to the left of the blots. The upper panels show immunoprecipitation of Kvl.5 but not co-immunoprecipitation with caveolin-3, while the lower panels show immunoprecipitation if caveolin-3 but not co-immunoprecipitation of K v l .5. (B) Western blots showing positive controls of known co-immunoprecipitating proteins in heart. These include actinin with K v l .5 (upper panels), eNOS with caveolin-3 (left lower panel), and P-dystroglycan with caveolin-3 (right lower panel). Other labels are as for Figure 5.1. 134 clearly confirmed in the cross-sectional images. In contrast, Kv l .5 levels are much lower, except at the intercalated disc, where Kvl .5 staining is very intense. In the ventricle, excluding the intercalated disc, Kv l .5 staining is most intense at the Z-line, adjacent to the striated pattern of caveolin-3 staining. Kvl .5 shows weaker staining at the surface and interior of the cell. Colocalization analysis revealed that 11% ± 4%> of the voxels specific for Kvl .5 also contained caveolin-3 (n=4). In rat atrial myocytes (Figure 5.3A) Kvl .5 weakly labeled the surface membrane and the interior of the cell, and only 8% ± 1%> of voxels that contained Kvl .5 also contained caveolin-3 (n=5). In canine myocytes (Figure 5.3B) caveolin-3 immunolabeling was also marked at the cell surface in atrial cells (upper panel), and at the surface and along Z-lines in canine ventricular myocytes (lower panel). Kv l .5 was also well-expressed in both canine atrial and ventricular myocytes with strong surface and Z-line expression in both types of cell. Despite the widespread distribution of both proteins, colocalization was minor in the atrial myocytes at 12 ± 2%. It was significantly larger in the ventricle at 20 ± 3%> (P<0.02). Values are mean ± SEM, with an n of 5 for each cell type. Kvl.5 and Caveolin-3 show little evidence of co-localization in immuno-EM samples To examine the distribution of Kvl .5 and caveolin-3 in cardiac tissue at a higher resolution, we immunogold labeled -70 nm thick Tokuyasu sections (see Methods) of rat atrial and ventricular tissue (Figure 5.4). In these samples caveolin-3 was immediately detectable as 10 nm gold particles distributed both at the cell surface and immediately below in flask-like structures (Figure 5.4B). In contrast, Kv l .5 was labeled with 5 nm gold particles and more rarely seen (identified by arrows in panels A-D). Despite the low density of Kvl .5 staining in both atrial (Figure 5.4A-C) and ventricular samples (Figure 5.4D), Kv l .5 was well separated from caveolin-3 labeling. In none of the E M sections illustrated was coincident labeling of the two proteins found. Overall, in less than 1%> of Kvl .5 labeling, 5 and 10 nm gold particles were found within 10 nm of each other (not shown), suggesting that this was a very rare occurrence. 135 Figure 5.3. Minimal co-localization of Kvl.5 and Caveolin-3 in wide-field images of rat and canine cardiac myocytes. Rat (A) and canine (B) atrial (above) and ventricular (below) myocytes labeled with antibodies specific for K v l .5 (green) and caveolin-3 (red), colocalized voxels are white. A l l of the images are 3D reconstructions, showing both X - Y and X - Z views of individual myocytes. (A) X Y views are 6 pm thick, and their corresponding X Z views are 1 pm thick. Scale bar is 5 pm. The first X - Z view of the rat ventricular myocyte is at the level of the intercalated disc and highlights the intricate topology of this region as well as the higher levels of Kvl .5 staining (B) X Y view of the atrial myocyte is 7pm thick, and the X Y view of the ventricular myocyte is 3 pm thick. Their corresponding X Z views are 1 pm thick. Scale bar is 5 pm. The double-headed arrow points to the intercalated disc, the single arrow points to the Z line. (Imaging and analysis carried out by E. Moore and Q. Yu) 136 Panel A of Figure 5.4 shows a T E M image of an atrial section (Figure 5.4A) and illustrates a lateral membrane region rich in caveolae with 2-5nm gold particles indicated by the arrow in the blow up of the boxed region (A). Figure 5.4B is a segment of surface membrane showing at least one clear caveolus, a segment of a Z-line, surface and internal Kvl .5 staining possibly corresponding to SR localization or as vesicular cargo. Panel C shows the membrane interface region of two myocytes. The ventricular image (Figure 5.4D) illustrates an intercalated disc region. The labeling of Kvl .5 found along the Z-lines and at the intercalated discs confirms the observations made in the fluorescent images (Figure 5.3). Kvl.5 does not co-sediment with Caveolin-3 in sucrose gradient fractionations In order to examine whether the channel could be biochemically localized to rafts, proteins in the membrane fractions from both HEK293 cells stably expressing Kvl .5 and from rat ventricular tissue were solubilized in 1% Triton X-100 at 4°C and then fractionated in a sucrose step gradient. Twelve-one milliliter fractions were removed from the top of the gradient and samples from each fraction were run on a denaturing gel. The light buoyant density fraction (Fraction 5) was identified using the raft marker flotillin (Figure 5.5A lower panel) as our HEK293 cells express very little caveolin-1 (not shown). The transferrin receptor was used as a control for a non-raft associated protein (Figure 5.5A and B upper panels), and was found in fractions 9-12 as well as the pellet from the fractionation (FP). Kv l .5 fractionated into the same fractions as the transferrin receptor, fractions 9-12, and appears to be absent from fraction 5 in both the cultured cells (Figure 5.5A middle panel) and the cardiac tissue sample (Figure 5.5B, lower panel). We also attempted to assess the effects of lipid raft disruption on the sustained potassium current (IKsus) in canine atrial myocytes by extracting cholesterol from the membrane with 137 Figure 5.4. Kvl.5 and Caveolin-3 do not co-localize in immuno-EM images of rat cardiac myocytes. Images are of rat Atrial (A-C) and ventricular (D) tissue. 70 nm cryosections were immunogold labeled for K v l .5 (5 nm gold; indicated by black arrows) and Caveolin-3 (10 nm gold). Boxed region in (A) is shown at higher magnification in the inset. Image in (D) shows a region through the intercalated disc. Other than in the inset scale bars represent 100 nm. (In B: PM= Plasma Membrane; Z=Z-line). 138 Figure 5.5. Kvl.5 does not localize to the lipid raft fraction in HEK293 cells or rat ventricular tissue. (A) HEK293 cells stably expressing Kvl .5 and (B) Rat ventricular tissue were mechanically disrupted and the membrane fractions were separated by discontinuous sucrose gradient after solubilization in 1% Triton X-100 at 4°C. Twelve fractions were collected starting from the top of the gradient and analyzed by Western Blot. Flotillin-1 was used to delineate the raft fraction and the transferrin receptor the non-raft fractions. C is the supernatant from a 128,000 x g spin that pelleted the membranes and represents the cytosolic fraction. P is the soluble protein extracted from the sucrose gradient pellet with 2% SDS. 139 methyl-P-cyclodextrin. Treatment of the myocytes made them difficult to perforate and increased their leakiness, making it difficult to interpret the results. In general, there was a decrease in the instantaneous current (ITO) but no obvious effects on the sustained current (not shown), which is thought to be made up at least in part, by Kvl .5 in canine atrial myocytes (Fedida et al., 2003) DISCUSSION In this study we have examined the direct physical interaction of Kvl .5 and the major caveolar coat protein, caveolin-3 using biochemical techniques, and the spatial distribution using wide field deconvolved and E M images of labeled myocytes. In addition, we have determined whether the channel is found in lipid rafts, membrane fractions characterized by a light-buoyant density and insensitivity to extraction by TritonX-100 at 4°C. In these studies in isolated myocytes and cardiac tissue, we were unable to detect significant interactions (Figure 5.2A) or colocalization between the two proteins (Figures 5.3 and 5.4), although interactions with other proteins known to associate with Kvl .5 or caveolin could be detected (Figure 5.2B). These observations were supported by fractionation results in which Kvl .5 was detected only in non-raft fractions in both HEK293 cells and in cardiac tissue under our assay conditions (Figure 5.5). Previous work with caveolin-1 in cultured cells (Martens et al., 2001), also showed no direct interaction between Kvl .5 and caveolin using co-immunoprecipitation experiments; however, more recent literature examining the association of Kv l .5 and caveolin-3 did detect an interaction in double, transiently transfected cultured cells (Folco et al., 2004). Our experimental conditions are quite different from these previous studies in that we used tissue in which the proteins are endogenously expressed. In addition to the caveolin-3 studies, we looked for an interaction of Kvl .5 with caveolin-2. However, while we were able to detect caveolin-2 in rat atrial and ventricular lysates (Figure 5.1, bottom panel), we could not detect caveolin-2 in 140 isolated rat cardiac myocytes despite repeated immunolabeling attempts under varied conditions (data not shown). Woodman et al. (2002) similarly failed to find caveolin-2 in myocytes and reported that caveolin-2 expression is limited to the endothelium and endocardium. Our findings are consistent with their data and in similar experiments using canine tissue we were unable to detect caveolin-2 at all (Figure 5.1). This may have been due to the greater sequence divergence between human and canine caveolin-2, than between human and rat. An association between Kvl .5 and immuno-isolated caveolae was described by Martens et al. (2000), and they reported that post-translationally modified Kvl .5 channel was absent from immuno-isolated caveolae in their mouse Itk- cells. This may explain our results showing that Kvl .5 and caveolin-3 did not coimmunoprecipitate or colocalize in cardiac myocytes (Figures. 5.2-5.4). In our experience, cardiac myocytes lack significant amounts of immature protein (as indicated by a single band in Western blots of cardiac lysates; Figure 5.1), suggesting that in native cells the channel is more efficiently processed. Our previous work with canine myocytes indicated that results of Kvl .5 localization studies are very antibody dependent (Fedida et al., 2003). The development of the Kvl .5 antibody used by our lab has enabled us to compare the distribution of Kvl .5 in rat and canine myocytes to that observed by others in rat (Barry et al., 1995), mouse (Trepanier-Boulay et al., 2001) and human (Mays et al., 1995). The most consistent observation is that Kvl .5 staining is low on lateral membranes and Z-lines and enriched at the intercalated disc in rat, canine and human while in mouse, the labeling appears to be more generally surface and Z-line in the ventricular cells examined (Trepanier-Boulay et al., 2001). Other channels have shown strong intercalated disc staining including Kv4.2 (Barry et al., 1995) and Nav1.5 (Maier et al., 2002; Kucera et al., 2002) and at least for the sodium channel this localization was suggested to play a role in initiation and conduction of the action potential (Kucera et al., 2002; Maier et al., 2002). A sodium-hydrogen exchanger has also been localized to this domain (Petrecca et al., 1999). 141 Fluorescent imaging experiments were very useful for getting a sense of Kvl .5 distribution throughout the myocyte and indicated that very little of the Kvl .5 channel staining overlaps with the abundantly expressed caveolin-3 protein (Figure 5.3A, B). The 8-20 % apparent colocalization under the light microscope appears to be an overestimate when protein localization is examined in the higher resolution E M images. However, when the resolution of this florescence imaging system is calculated (122 nm in the X - Y plane and 250 nm in the Z plane), this voxel size together with the occasional relative proximity of Kvl .5 and caveolin immunogold staining would result in voxels positive for both proteins. These results highlight how complimentary these two imaging techniques are. While the T E M experiments greatly improved the resolution and enabled us look more carefully at the channel distribution in relation to caveolin-3 and several ultra-structural elements, the fluorescent images gave a more general sense of staining that is impossible to appreciate when tightly focused at specific regions of the cell. The T E M images showed sparse Kvl .5 staining and prompted us to calculate the expected channel density in rat atrium. Assuming a specific capacitance of 0.01 pF/pm 2 (average atrial myocyte capacitance of 40.7 pF and therefore a cell surface area of about 4070 pm2), a maximal slope conductance of 103 pS/pF (Van Wagoner et al., 1996), a maximal open probability of 0.80, and a single channel conductance of 13 pS for K v l . 5 , then we would expect to see ~1 channel per 10 pm 2. If we decrease the open probability to 0.5 then we would expect a channel in every 6.3 pm 2. These calculations would indicate that in each 6-10 pm length of surface membrane examined in the 70 nm thick E M sections, we would expect to see less than one channel. Depending on how much of the channel partitioned into each given section there is the possibility to label each subunit in the channel, which would result in up to four labels per channel, depending on the degree of steric hindrance. If the average thickness of a myocyte were 12 pm then we would expect that each myocyte would produce ~170 sections. From these 142 numbers it would seem that the occurrence of functional channels in the lateral membranes is quite low especially given the apparent concentration at the intercalated disc region shown in the fluorescent images (Figure 5.3). It is therefore, not surprising that in the T E M images of immunolabeled myocytes, channel abundance was quite low. We have no way of calculating how many non-functional channels there may be, both at the surface and those in intracellular compartments, such as in early endosomes, which would be detected in the immunolabeling experiments but would not contribute to current density. Given that Kvl .5 did not localize to caveolae, it was important to establish whether the channel would partition into non-caveolar lipid rafts in our expression system (HEK293 cells) and in cardiac tissue. Previous results showing raft localization were acquired using transfected mouse Itk- cells (Martens et al., 2001). Again our results differ from this earlier report in that we did not see any Kvl .5 immunoreactivity in the raft fraction from HEK293 cells or the rat ventricular sample, the majority of the channel protein clearly fractionated with the non-raft marker, the Transferrin Receptor (Figure 5.5). However, our levels of expressed protein appear to be significantly less, as well the fraction of mature to immature protein, than in the fibroblasts used in previous work (Martens et al., 2001). This group was able only to demonstrate the presence of immature Kvl .5 in the raft fraction. Fibroblasts also express caveolin-2 as well as caveolin-1 whereas HEK293 cells appear to express little caveolin-1 and no caveolin-2 (data not shown). In addition, both clathrin-mediated and caveolar-mediated internalization of proteins can result in re-localization of surface proteins to the early endosome (Sharma et al., 2003), which could account for some of the apparent co-localization of caveolin and Kvl .5 in both the Itk cells (Martens et al., 2001) and the cardiac myocytes (Figure 5.3). It is also possible that only a small fraction of K v l . 5 , below the level of detection by indirect chemiluminescence, is raft associated in the samples we examined. 143 Further indirect evidence that Kvl .5 is not associated with caveolae is provided by experiments that disrupt the actin cytoskeleton. We have shown that pretreatment of HEK293 cells with the actin-depolymerizing agent, cytochalasin D leads to a large increase in Kvl .5 surface expression (Maruoka et al., 2000). In contrast, it is known that surface caveolin-1 levels are decreased by disruption of the actin cytoskeleton in CHO cells, which, like H E K cells express caveolin-1 (Mundy et al., 2002). In summary, we have shown that Kvl .5 and caveolin-3 do not directly associate in rat and canine cardiac myocytes. Given the low degree of co-localization of the channel and caveolin-3 and the failure to detect the channel in raft fractions, it seems likely that caveolae and lipid raft domains play little or no role in the localization of Kv l .5 in native cardiac myocytes. 144 ACKNOWLEDGMENTS This work has been supported by grants to D.F. and E . M . from the Heart and Stroke Foundations for British Columbia and Yukon and from the Canadian Institute for Health Research, and by a grant to D.V.W. from the National Institute of Health (NHLBI, 1R01-HL-65412). We thank Paul Murray, Ph.D. for providing access to canine tissues, Changiz Taghibiglou, Ph.D. for his assistance and the raft isolation protocol, George Posthuma, Ph.D. for Tokuyasu protocols and E M advice, and Dave Steele, Ph.D. for editorial comments. REFERENCES Aravamudan B, Volonte D, Ramani R, Gursoy E, Lisanti MP, London B, Galbiati F (2003), Transgenic overexpression of caveolin-3 in the heart induces a cardiomyopathic phenotype, Hum.Mol.Genet. 12: 2777-2788 Barry D M , Trimmer JS, Merlie JP, Nerbonne J M (1995), Differential expression of voltage-gated K + channel subunits in adult rat heart: Relation to functional K + channels, Circ.Res. 11: 361-369 Bock J, Szabo I, Gamper N , Adams C, Gulbins E (2003), Ceramide inhibits the potassium channel Kvl .3 by the formation of membrane platforms, Biochem.Biophys.Res.Commun. 305: 890-897 Dupree P, Parton RG, Raposo G, Kurzchalia TV, Simons K (1993), Caveolae and sorting in the trans-Golgi network of epithelial cells, EMBO J. 12: 1597-1605 Eldstrom J, Choi WS, Steele DF, Fedida D (2003), SAP97 increases Kvl .5 currents through an indirect N-terminal mechanism, FEBS Lett. 547: 205-211 Fedida D, Eldstrom J, Hesketh JC, Lamorgese M , Castel L , Steele DF, Van Wagoner DR (2003), Kv l .5 is an important component of repolarizing K+ current in canine atrial myocytes, Circ.Res 93: 744-751 Fedida D, Wible B, Wang Z, Fermini B, Faust F, Nattel S, Brown A M (1993), Identity of a novel delayed rectifier current from human heart with a cloned K + channel current, Circ.Res. 73: 210-216 Feng JL, Wible B, L i GR, Wang ZG, Nattel S (1997), Antisense oligodeoxynucleotides directed against Kv l .5 mRNA specifically inhibit ultrarapid delayed rectifier K + current in cultured adult human atrial myocytes, Circ.Res. 80: 572-579 145 Feron O, Belhassen L, Kobzik L , Smith TW, Kelly RA, Michel T (1996), Endothelial nitric oxide synthase targeting to caveolae. Specific interactions with caveolin isoforms in cardiac myocytes and endothelial cells, J.Biol.Chem. 271: 22810-22814 Feron O, Smith TW, Michel T, Kelly R A (1997), Dynamic targeting of the agonist-stimulated m2 muscarinic acetylcholine receptor to caveolae in cardiac myocytes, J.Biol.Chem. 272: 17744-17748 Folco EJ, Liu G X , Koren G (2004), Caveolin-3 and SAP97 form a scaffolding protein complex that regulates the voltage-gated potassium channel K v l .5, Am.J.Physiol Heart Circ.Physiol 287: H681-H690 Fujita T, Toya Y , Iwatsubo K, Onda T, Kimura K , Umemura S, Ishikawa Y (2001), Accumulation of molecules involved in alpha 1-adrenergic signal within caveolae: caveolin expression and the development of cardiac hypertrophy, Cardiovasc.Res. 51: 709-716 Galbiati F, Engelman JA, Volonte D, Zhang X L , Minetti C, L i M , Hou H , Jr., Kneitz B, Edelmann W, Lisanti MP (2001), Caveolin-3 null mice show a loss of caveolae, changes in the microdomain distribution of the dystrophin-glycoprotein complex, and t-tubule abnormalities, J.Biol.Chem. 276: 21425-21433 Hohl C M , Altschuld R A (1991), Response of isolated adult canine cardiac myocytes to prolonged hypoxia and reoxygenation, Am.J.Physiol 260: C383-C391 Isshiki M , Anderson RG (2003), Function of caveolae in Ca2+ entry and Ca2+-dependent signal transduction, Traffic. 4: 717-723 Kucera JP, Rohr S, Rudy Y (2002), Localization of sodium channels in intercalated disks modulates cardiac conduction, Circ.Res. 91: 1176-1182 Lasley RD, Narayan P, Uittenbogaard A , Smart EJ (2000), Activated cardiac adenosine A( l ) receptors translocate out of caveolae, J.Biol.Chem. 275: 4417-4421 Lasley RD, Smart EJ (2001), Cardiac myocyte adenosine receptors and caveolae, Trends Cardiovasc.Med. 11: 259-263 L i GR, Feng JL, Wang ZG, Fermini B, Nattel S (1996a), Adrenergic modulation of ultrarapid delayed rectifier K + current in human atrial myocytes, Circ.Res. 78: 903-915 L i S, Couet J, Lisanti MP (1996b), Src tyrosine kinases, Galpha subunits, and H-Ras share a common membrane-anchored scaffolding protein, caveolin. Caveolin binding negatively regulates the auto-activation of Src tyrosine kinases, J.Biol.Chem. 271: 29182-29190 L i S, Galbiati F, Volonte D, Sargiacomo M , Engelman JA, Das K, Scherer PE, Lisanti MP (1998), Mutational analysis of caveolin-induced vesicle formation. Expression of caveolin-1 recruits caveolin-2 to caveolae membranes, FEBS Lett. 434: 127-134 Maier SK, Westenbroek RE, Schenkman K A , Feigl EO, Scheuer T, Catterall W A (2002), An unexpected role for brain-type sodium channels in coupling of cell surface depolarization to contraction in the heart, Proc.Natl.Acad.Sci.U.S.A 99: 4073-4078 146 Martens JR, Navarro-Polanco R, Coppock EA, Nishiyama A, Parshley L , Grobaski T, Tamkun M M (2000), Differential targeting of Shaker-like potassium channels to lipid rafts, JBiol Chem 275: 7443-7446 Martens JR, Sakamoto N , Sullivan SA, Grobaski TD, Tamkun M M (2001), Isoform-specific localization of voltage-gated K+ channels to distinct lipid raft populations. Targeting of Kv l .5 to caveolae, J.Biol.Chem. 276: 8409-8414 Maruoka ND, Steele DF, Au BPY, Dan P, Zhang X , Moore EDW, Fedida D (2000), a-Actinin-2 couples to cardiac Kvl .5 channels, regulating current density and channel localization in H E K cells, FEBS Letts. 473: 188-194 Mason HS, Latten MJ , Godoy L D , Horowitz B, Kenyon JL (2002), Modulation of Kvl .5 currents by protein kinase A , tyrosine kinase, and protein tyrosine phosphatase requires an intact cytoskeleton, Mol.Pharmacol. 61: 285-293 Mays DJ, Foose JM, Philipson L H , Tamkun M M (1995), Localization of the Kvl .5 K + channel protein in explanted cardiac tissue, J.Clin.Invest. 96: 282-292 Mundy DI, Machleidt T, Ying Y S , Anderson RG, Bloom GS (2002), Dual control of caveolar membrane traffic by microtubules and the actin cytoskeleton, J.Cell Sci. 115: 4327-4339 Nabi IR, Le PU (2003), Caveolae/raft-dependent endocytosis, J. Cell Biol. 161: 673-677 Nitabach M N , Llamas DA, Thompson IJ, Collins K A , Holmes TC (2002), Phosphorylation-dependent and phosphorylation-independent modes of modulation of Shaker family voltage-gated potassium channels by Src family protein tyrosine kinases, J.Neurosci. 22: 7913-7922 Petrecca K, Atanasiu R, Grinstein S, Orlowski J, Shrier A (1999), Subcellular localization of the Na+/H+ exchanger NHE1 in rat myocardium, Am.J.Physiol 276: H709-H717 Scherer PE, Lewis R Y , Volonte D, Engelman JA, Galbiati F, Couet J, Kohtz DS, van Donselaar E, Peters P, Lisanti MP (1997), Cell-type and tissue-specific expression of caveolin-2. Caveolins 1 and 2 co-localize and form a stable hetero-oligomeric complex in vivo, J.Biol.Chem. 272: 29337-29346 Scriven DR, Dan P, Moore ED (2000), Distribution of proteins implicated in excitation-contraction coupling in rat ventricular myocytes, Biophys.J. 79: 2682-2691 Sharma DK, Choudhury A, Singh RD, Wheatley CL, Marks DL, Pagano RE (2003), Glycosphingolipids internalized via caveolar-related endocytosis rapidly merge with the clathrin pathway in early endosomes and form microdomains for recycling, J.Biol.Chem. 278: 7564-7572 Slot JW, Geuze HJ, Gigengack S, James DE, Lienhard GE (1991), Translocation of the glucose transporter GLUT4 in cardiac myocytes of the rat, Proc.Natl.Acad.Sci.U.S.A 88: 7815-7819 Smythe G M , Eby JC, Disatnik M H , Rando T A (2003), A caveolin-3 mutant that causes limb girdle muscular dystrophy type 1C disrupts Src localization and activity and induces apoptosis in skeletal myotubes, J.Cell Sci. 116: 4739-4749 147 Song KS, L i S, Okamoto T, Quilliam L A , Sargiacomo M , Lisanti MP (1996), Co-purification and direct interaction of Ras with caveolin, an integral membrane protein of caveolae microdomains. Detergent-free purification of caveolae microdomains, J.Biol.Chem. Ill: 9690-9697 Sotgia F, Lee JK, Das K , Bedford M , Petrucci TC, Macioce P, Sargiacomo M , Bricarelli FD, Minetti C, Sudol M , Lisanti MP (2000), Caveolin-3 directly interacts with the C-terminal tail of beta -dystroglycan. Identification of a central WW-like domain within caveolin family members, J.Biol.Chem. 275: 38048-38058 Trepanier-Boulay V, St Michel C, Tremblay A , Fiset C (2001), Gender-based differences in cardiac repolarization in mouse ventricle, Circ.Res. 89: 437-444 van Deurs B, Roepstorff K , Hommelgaard A M , Sandvig K (2003), Caveolae: anchored, multifunctional platforms in the lipid ocean, Trends Cell Biol. 13: 92-100 Van Wagoner DR (2000), Pharmacologic relevance of K + channel remodeling in atrial fibrillation, J.Mol.Cell.Cardiol. 32: 1763-1766 Van Wagoner DR, Kirian M , Lamorgese M (1996), Phenylephrine suppresses outward K + currents in rat atrial myocytes, Am.J.Physiol.Heart Circ.Physiol. 271: H937-H946 Wong W, Schlichter L C (2004), Differential recruitment of K v l . 4 and Kv4.2 to lipid rafts by PSD-95, J.Biol.Chem. 279: 444-452 Woodman SE, Park DS, Cohen A W , Cheung M W , Chandra M , Shirani J, Tang B, Jelicks L A , Kitsis RN, Christ GJ, Factor SM, Tanowitz HB, Lisanti MP (2002), Caveolin-3 knock-out mice develop a progressive cardiomyopathy and show hyperactivation of the p42/44 M A P K cascade, J.Biol.Chem. 277: 38988-38997 Xia F, Gao X , Kwan E, Lam PP, Chan L, Sy K , Sheu L, Wheeler M B , Gaisano H Y , Tsushima R G (2004), Disruption of pancreatic beta-cell lipid rafts modifies Kv2.1 channel gating and insulin exocytosis, J.Biol.Chem. 279: 24685-24691 Xiang Y , Rybin VO, Steinberg SF, Kobilka B (2002), Caveolar localization dictates physiologic signaling of beta 2-adrenoceptors in neonatal cardiac myocytes, J.Biol.Chem. 277: 34280-34286 148 CHAPTER 6: DISCUSSION 149 SAP97 and PDZ domains: mechanisms and molecules that regulate Kvl.5 channel expression Since the publication of our paper showing that the effects of SAP97 on K v l .5 expression are not due to direct binding of the PDZ protein with the channel (Eldstrom et al., 2003a), two more publications have appeared showing direct interaction between these two proteins (Folco et al., 2004; Godreau et al. 2003). In each of these studies, the expression system was of double-transient transfection into cultured cells, which may be an important difference between our work and theirs. In our studies we have used either a cell line stably expressing the channel into which we transfected SAP97, or cardiac tissue lysates in which endogenously expressed proteins are found. More recent data from our lab has shown that a small fraction of immature K v l .5 channel protein can be co-immunoprecipitated with SAP97 from HEK293 cell lysates, only i f both proteins are the result of transient transfection (Mathur et al., unpublished data). It is possible that this reflects temporal overlap in overexpression of the two proteins that allows promiscuous interaction, co-segregation into protein aggregates, or amplification of a normally fleeting intracellular interaction. In addition, we have found that the SAP97-mediated increase in Kv l .5 is dependent on the presence of a threonine residue within a P K C consensus site in the N -terminus of the channel. This could mean that SAP97 is actually affecting localization or activity of a P K C isoform that is in turn regulating channel surface density through a phosphorylation event. Regardless of the nature of the interaction between Kvl .5 and SAP97, even recent data fails to make a strong case for SAP97 as a meaningful binding partner of Kvl .5 in the heart. This leaves a PDZ binding domain lacking a defined physiological role. Given the accumulating evidence that membrane proteins in general and ion channels more specifically exist at the cell surface as part of macromolecular complexes (reviewed in Marx, 2003), it seems likely that this PDZ binding domain will play an important role in trafficking Kvl .5 to the cell 150 surface or in anchoring the channels once at the plasma membrane. It is also possible that this domain may act as a binding site for adaptor proteins that in turn bind to kinases or phosphatases involved in regulating the channel activity. At the present time, in the absence of further experimental data, the role of this domain remains speculative. However, it is interesting to consider the roles demonstrated for other PDZ domains found in the intracellular domains of K v channels. The PDZ domain in the C-terminus of K v l . l is somehow involved in regulating P-subunit conferred inactivation. The P-subunit mediated inactivation can be enhanced by two mechanisms that dissociate the C-terminus from the cytoskeleton including phosphorylation on a serine residue in the C-terminus or deletion of the PDZ binding domain (Jing et al., 1997). Phosphorylation of the channel is mediated by P K A , an event that also appears to enhance K v l . l channel surface expression in oocytes (Levin et al., 1996). Kv l .5 is also sensitive to levels of P K A activity in oocytes, as inhibition of P K A leads to a slow decline in current (Mason et al., 2002), and while sensitive to cytoskeletal disruption, the decline is still observed with a C-terminally truncated mutant. This fits with an important role of the N-terminus in mediating interactions with the actin cytoskeleton through a-actinin (Maruoka et al., 2000) and frees the C-terminus to perhaps play a role in scaffolding rather than in anchoring. Role of the amino terminus of the Kvl .5 protein in cell surface expression As a further note on the importance of cytoskeletal interactions in regulating Kvl .5 expression, it is of some interest that treatment of HEK293 cells, expressing K v l . 5 , with antisense oligonucleotides directed towards a-actinin results in an increase in Kvl .5 expression of similar magnitude as that observed upon treatment of cells with cytoskeletal disrupting agents (Maruoka et al., 2000), SAP97 co-expression (Fig3.1; Eldstrom et al., 2003a) or inhibition of dynein mediated vesicular trafficking (Choi et al., manuscript in preparation). The latter two effects at least are dependent on an intact channel N-terminus. These results all point to 151 regulation at the level of protein stability at the membrane, and perhaps could be explained by the recently identified role of SAP97 in endocytosis (Hasson, 2003; Osterweil et al., 2005; Wu et al., 2002). Overexpression of SAP97 in this scenario would look a lot like disruption of the dynein motor complex by overexpression of p50 (Burkhardt et al., 1997), which prevents retrograde transport of endosomes leading to a backlog of vesicles at the P M and accumulation of membrane proteins reliant on this mechanism for internalization. While it is difficult to understand why the highly related Kv l .4 is actually decreased by SAP97 overexpression (Tiffany et al., 2000), it is at least consistent with the decrease observed with p50 overexpression (Choi et al., unpublished observation) and leaves room for the suggestion of an alternative mechanism for trafficking of this channel. In neurons, Myosin VI interacts with endocytic vesicles carrying the a-Amino-3-hydroxy-5-methylisoxazole-4-propionate receptor (AMPAR) but not the N-methyl-D-aspartate receptor (NMDAR; Osterweil et al., 2005) and immunoprecipitates of myosin VI contain both the endocytic adaptor protein AP-2 and SAP97. If, as has been suggested, lipid rafts constitute a separate endocytic pathway, this could be a potential explanation for the opposite response seen between the probable non-raft associated Kvl .5 (Fig5.5) and partly raft-associated Kv l .4 (Wong and Schlichter, 2004). While it has been shown that several factors can change the expression of K v l . 5 , what has not been worked out is how exactly the channel dwell time at the cell surface is regulated. The re-uptake of membrane proteins frequently involves a phosphorylation event that subsequently allows binding of proteins involved in the endocytic process, but cursory analysis of the sequence of Kvl .5 fails to uncover many of the canonical motifs identified to date and reviewed in the introduction. Regulation of Kv l .2 channel expression has been linked to channel tyrosine kinase phosphorylation which leads to channel endocytosis (Nesti et al., 2004). Despite evidence that tyrosine phosphorylation suppresses Kvl .5 currents also (Holmes et al., 1996), Kv l .2 is clearly regulated by distinct mechanisms from those of K v l . 5 , as disruption of the actin 152 cytoskeleton also leads to endocytosis of Kv l .2 , whereas Kvl .5 accumulates at the cell surface (Maruoka et al., 2001). The occasional observation of a faster migrating band in SDS-PAGE gels of canine and rat lysates, that is recognized by a C-terminal Kvl .5 antibody (Kurata et al., submitted), and reported also by one group for K v l . l as a Tl-less fragment not detected by their N-terminal antibody (Strang et al., 2001), allows speculation of protease-mediated turnover of the channels. In the case of K v l . l , it was found that while the S1-S2 extracellular linker was more susceptible to trypsin cleavage, the T l - S l was also a site of proteolysis. Together with data on the response of the channel to growth factor receptor stimulation in oocytes, in which a down regulation of the channel is dependent on PLCy activation and an increase in intracellular calcium (Timpe and Fantle, 1994), this potentially implicates calcium-dependent proteases in this process. In support of this idea is the observation that upon washout of growth factor, Kv l .5 current amplitudes recover very slowly. This may reflect a recovery process that is dependent upon insertion of new channels into the membrane rather than involving a phosphorylation-dependent event. Cleavage of the N-terminus could lead to internalization of Kv l .5 by leaving an N-terminal amino acid that fits into the N-end pathway for ubiquitination (reviewed in Glickman and Ciechanover, 2002) or it could release the channel from stable membrane domains and allow it to diffuse to less stable membrane domains undergoing internalization. To extend this scenario further, the fact that P K A activity is already maximally affecting Kvl .5 expression in oocytes and that down regulation is achieved by P K A inhibition falls in line with three important additional observations. First, P K A regulates calpastatins, which are specific inhibitors of calpains, the calcium-dependent proteases. Phosphorylation of the inhibitor decreases its effectiveness in inhibiting the proteases, and leads to increased proteolytic activity. The second observation relevant to this mechanism is that (3-adrenergic stimulation of cardiac myocytes leads to both an increase in P K A activity and a large increase in calpastatin synthesis (reviewed in Goll et al., 153 2003), and as mentioned earlier, IKUR/KV1.5 is upregulated in response to P-adrenergic stimulation. The third observation is that calpain activity is higher in patients with AP, consistent with observed increases in intracellular calcium levels (Brundel et al., 2002) and where also the levels of Kvl .5 are reduced (Brundel et al., 2001; Van Wagoner et al., 1997). The only known study examining the effects of rapid pacing on calpain activity and Kvl .5 expression utilized a mouse atrial cell line, in which no change in Kvl .5 was noted despite elevated calpain activity and decreased L-type calcium channel levels (Brundel et al., 2004). Interestingly, the faster migrating band of Kvl .5 has not been observed in mouse cardiac lysates using the same antibody that was used for the canine and rat samples (J. Nerbonne, personal communication). It has been noted elsewhere that Kvl .5 has a short half-life (~4 hours; Takimoto et al., 1993) which means that the channel density can be changed relatively rapidly if re-uptake is either enhanced or prevented. It will be of great interest to see if N-terminal deletions such as Kvl.5A209, frequently used in our lab to study the role of the N-terminus in diverse mechanisms such as inactivation (Kurata et al., 2001), cytoskeletal interactions (Eldstrom et al., 2002, 2003a) and dynein-mediated trafficking (Choi et al, manuscript in preparation) has a shorter half-life than the wild-type channel. Given the immediacy of increases in surface Kv4.2 protein noted (within 1 hour; Wang et al., 2004) upon treatment of cells with an actin depolymerizing agent, and the similarity in response of Kvl .5 (Maruoka et al., 2000), it is tempting to speculate that there is a pool of channels that are in a recycling compartment, similar to that observed for connexin 43, another protein with a short half-life (reviewed in Saffitz et al., 2000). We have previously shown that deletions involving short segments of the extreme N -terminus (AN2-19, AN2-91 and AN2-119) and those involving much more of the N-terminus including a large portion of the TI domain (AN2-188 and AN2-209), are well tolerated and allow expression of functional channels, though the latter is associated with some interesting changes in channel inactivation properties (Kurata et al., 2002). Among the deletions tested were two 154 that involved intermediate lengths of deletions that did not express functional channels (AN2-135 and AN2-162) despite appropriately sized proteins based on the amino acid sequence being detected in Western blots. The protein appears to be immature, lacking extensive glycosylation, and upon examination of gross cellular distribution using confocal microscopy appears cytosolic and aggregated (Eldstrom et al., 2003b). The amino acid sequence upstream of the AN2-162 deletion displays several potential arginine based ER retention motifs (RxR) that have been identified in other potassium channels (reviewed in Griffith, 2001). This region of the channel is complicated by the secondary structure that each a-subunit adopts and comparison with the crystal structure of the Kv l .2 tetrameric TI domain (Minor et al., 2000) shows that only the R L R sequence at positions 184-186 may be exposed in a properly folded monomer and probably is masked in the tetrameric complex. When N-terminal portions of the TI domain are deleted though, it is expected that folding of the entire domain will be affected and may lead to aggregation and accumulation of protein via hydrophobic interaction. Our preliminary data involving mutagenesis of arginines at both position 167 and 184 to disrupt potential retention signals in AN2-162 have resulted in only small currents in transiently transfected cells. The cellular distribution of the double arginine mutant appears more reticulate, as apposed to the aggregated appearance of the single R167A mutant, indicating that this protein is perhaps more efficiently inserted into the ER. These studies are still quite preliminary and the role of the potential retention motifs in preventing wild-type monomeric subunits from escaping the ER has yet to be assessed. Future studies that can shed more light on Kvl.5 surface regulation Re-uptake and/or mechanisms behind regulation of channel expression are not the only aspects of Kvl .5 not fully understood. A l l levels of regulation of channel transcription, synthesis, modification and expression are still only in the early stages of investigation. Most 155 studies to date have not gone beyond observations of increase or decreases at the level of mRNA, protein and current amplitudes. One approach to discover Kvl .5 interacting proteins is using immunoprecipitates of the channel, in-gel trypsinization of excised SDS-PAGE gel bands and carrying out MALDI-TOF analysis of cleavage products. This proteomics-based approach has proven successful for others looking to identify endogenous ion channel binding proteins involved in channel trafficking and anchoring in various tissues (Husi et al., 2000; Leonoudakis et al., 2004). By identifying interacting proteins, mechanisms for Kvl .5 trafficking out of the ER and to the cell surface can be established, as well as potential ways the channel can be up- or down-regulated. Apparent cellular localization of Kv l .5 at the intercalated disc and in caveolae One of the perplexing aspects of Kvl .5 is its apparent enrichment at the intercalated disc (ID; Figure 4-4 and 5-3; Barry et a l , 1994; Mays et al., 1995; Fedida et al., 2003). While some attention has been paid to a similar enrichment of Nav1.5 (Kucera et al., 2002; Maier et al., 2002; Mohler et al., 2004), little has been said or done to understand Kvl .5 localization here. The fact that predominant N a + / K + ATPase staining was also noted at the ID (McDonough et al., 1994), suggests that there is functional significance to the localization. In models of AP propagation between cardiac myocytes it has been suggested that activation of ID localized sodium channels in one cell would lead to a more negative cleft space potential, this would be "sensed" by the second cell across the cleft and decrease the threshold for activation and thus lead to improved impulse propagation (reviewed in Sperelakis, 2002). However, it appears there is some disagreement as to the significance, i f any, of this type of mechanism given the presence and contribution of gap junctions to AP propagation (reviewed in Rohr, 2004). Given the limitations on the resolution of fluorescent microscopy combined with additional factors associated with indirect immunocytochemistry, such as the existence of 156 reliable antibodies, labeling efficiency, (which is related to antibody specificity and steric hindrances) and the radial uncertainty associated with using two antibodies (length from epitope to fluorescent tag -30 nm) it is not surprising that there is still a lot to be learned about the cellular distribution of K v l . 5 . Biochemical approaches to identifying protein association with specialized lipid domains are also fraught with isolation artifacts and discrepancy between laboratories are commonplace, the controversy on Kv channel association with lipid rafts being a case in point. It is possible that differences are due to the cell types used, and several hypotheses exist to explain why membrane proteins can exist in both detergent-resistant membranes/caveolae and the bulk plasma membrane. These include suggestions that these membrane domains are for storage of inactivated proteins awaiting a stimulus for recruitment into active membrane domains; or alternatively, this may represent a population of channels in the process of being downregulated and are part of a clathrin-independent internalization pathway. Association may depend on the expression of an adaptor protein and different populations of the channel may have different roles depending on the macromolecular complex with which they assemble. These would all require association of the channel with a protein in some cell types that conferred association with lipid rafts or prevented raft association, that was not present in others; to be otherwise would mean that raft association was an innate property of the amino acid sequence of Kvl .5 and thus why would it not always associate with rafts. Even in reports that have established lipid raft association of K v l . 5 , it has constituted a very small fraction of the total protein, and apparently of an immature product (Martens et al., 2001). Notably, though, fractional raft association of proteins has been noted for other channels including the epithelial sodium channel (ENaC; Hi l l et al., 2002), Kv4.2 and K v l . 4 (Wong and Schlichter, 2004) and may be cell type dependent (Martens et al., 2000). As Kv l .5 frequently associates with other K v l channels to form heterotetramers in various tissues, especially with Kv l .2 and Kvl .4 , it is possible that this is a mechanism for Kvl .5 to achieve raft association. 157 Especially, as previously mentioned, Kvl.4 has been shown to associate with rafts in brain tissue (Wong and Schlichter, 2004). Summary The work described in this thesis has furthered the understanding of Kvl.5 distribution in both heterologous expression systems (HEK293 cells) and in cardiac myocytes. We have established that Kvl.5 can interact with a PDZ protein, PSD95, in heterologous cells via both the N- and C-termini and that this binding can have significant effects on the channel expression. In contrast, SAP97, which also has significant effects on expression, does so via an indirect mechanism that is dependent on the N-terminus of the channel. Neither of these two PDZ proteins appear to be of significance to channel expression in cardiac tissue and efforts are ongoing to establish whether there exists a partner or partners for the PDZ binding domains in Kvl.5 in native myocytes. We have also established that Kvl.5 is a significant contributor to the repolarization phase of the canine atrial action potential and have determined that the channel likely is not appreciably associated with lipid rafts or caveolae in cardiac cells. 158 REFERENCES Barry D M , Trimmer JS, Merlie JP, Nerbonne JM. (1995) Differential expression of voltage-gated K + channel subunits in adult rat heart. Relation to functional K + channels? Circ Res. 77(2):361-9. Burkhardt JK, Echeverri CJ, Nilsson T, Vallee RB. (1997) Overexpression of the dynamitin (p50) subunit of the dynactin complex disrupts dynein-dependent maintenance of membrane organelle distribution. J Cell Biol. 139(2):469-84. Brundel BJ, Ausma J, van Gelder IC, Van der Want JJ, van Gilst WH, Crijns HJ, Henning RH. (2002) Activation of proteolysis by calpains and structural changes in human paroxysmal and persistent atrial fibrillation. Cardiovasc Res. 54(2):380-9. Brundel BJ, Kampinga H H , Henning RH. (2004) Calpain inhibition prevents pacing-induced cellular remodeling in a HL-1 myocyte model for atrial fibrillation. Cardiovasc Res. 62(3):521-8. Brundel BJ, Van Gelder IC, Henning RH, Tuinenburg A E , Wietses M , Grandjean JG, Wilde A A, Van Gilst WH, Crijns HJ. (2001) Alterations in potassium channel gene expression in atria of patients with persistent and paroxysmal atrial fibrillation: differential regulation of protein and mRNA levels for K + channels. J Am Coll Cardiol. 37(3):926-32. Eldstrom J, Choi WS, Steele DF, Fedida D. (2003a) SAP97 increases Kvl .5 currents through an indirect N-terminal mechanism. FEBS Lett. 17;547(l-3):205-ll. Eldstrom J, Doerksen K W , Steele DF, Fedida D. (2002) N-terminal PDZ-binding domain in K v l potassium channels. FEBS Lett. 2002 531(3):529-37. J. Eldstrom, H.T. Kurata, D.F. Steele, D. Fedida (2003b) Exploring ion channel retention signals in the amino terminus of hKvl.5. Mol . Biol. Cell. 14:S228a Fedida D, Eldstrom J, Hesketh JC, Lamorgese M , Castel L, Steele DF, Van Wagoner DR. (2003) Kvl .5 is an important component of repolarizing K + current in canine atrial myocytes. Circ Res. 93(8):744-51. Folco EJ, Liu G X , Koren G. (2004) Caveolin-3 and SAP97 form a scaffolding protein complex that regulates the voltage-gated potassium channel K v l . 5 . A m J Physiol Heart Circ Physiol. 287(2):H681-90 Glickman M H , Ciechanover A. (2002) The ubiquitin-proteasome proteolytic pathway: destruction for the sake of construction. Physiol Rev. 82(2):373-428. Godreau D, Vranckx R, Maguy A, Goyenvalle C, Hatem SN (2003) Different isoforms of synapse-associated protein, SAP97, are expressed in the heart and have distinct effects on the voltage-gated K + channel K v l . 5 . J Biol Chem. 278(47):47046-52. 159 Goll DE, Thompson VF, L i H , Wei W, Cong J. (2003) The calpain system. Physiol Rev. 2003 Jul;83(3):731-801 Griffith L C . (2001) Potassium channels: the importance of transport signals. Curr Biol. ll(6):R226-8. Hasson T. (2003) Myosin VI: two distinct roles in endocytosis. J Cell Sci. 116(Pt 17):3453-61. Hi l l WG, An B, Johnson JP. (2002) Endogenously expressed epithelial sodium channel is present in lipid rafts in A6 cells. J Biol Chem. 277(37):33541-4. Holmes TC, Fadool DA, Ren R, Levitan IB. (1996) Association of Src tyrosine kinase with a human potassium channel mediated by SH3 domain. Science. 274(5295):2089-91. Husi H , Ward M A , Choudhary JS, Blackstock WP, Grant SG. (2000) Proteomic analysis of N M D A receptor-adhesion protein signaling complexes. Nat Neurosci. 3(7):661-9. Jing J, Peretz T, Singer-Lahat D, Chikvashvili D, Thornhill WB, Lotan I. (1997) Inactivation of a voltage-dependent K + channel by beta subunit. Modulation by a phosphorylation-dependent interaction between the distal C terminus of alpha subunit and cytoskeleton. J Biol Chem. 272(22): 14021-4. Kucera JP, Rohr S, Rudy Y . (2002) Localization of sodium channels in intercalated disks modulates cardiac conduction. Circ Res. 91(12): 1176-82. Kurata HT, Soon GS, Eldstrom JR, Lu GW, Steele DF, Fedida D. (2002) Amino-terminal determinants of U-type inactivation of voltage-gated K + channels. J Biol Chem. 277(32):29045-53. Kurata HT, Soon GS, Fedida D. (2001) Altered state dependence of c-type inactivation in the long and short forms of human K v l . 5 . J Gen Physiol. 2001 Sep;l 18(3):315-32. Leonoudakis D, Conti LR, Anderson S, Radeke C M , McGuire L M , Adams M E , Froehner SC, Yates JR 3rd, Vandenberg CA. (2004) Protein trafficking and anchoring complexes revealed by proteomic analysis of inward rectifier potassium channel (Kir2.x)-associated proteins. J Biol Chem. 279(21):22331-46 Levin G, Chikvashvili D, Singer-Lahat D, Peretz T, Thornhill WB, Lotan I. (1996) Phosphorylation of a K + channel alpha subunit modulates the inactivation conferred by a beta subunit. Involvement of cytoskeleton. J Biol Chem. 271(46):29321-8. Maier SK, Westenbroek RE, Schenkman K A , Feigl EO, Scheuer T, Catterall WA. (2002) An unexpected role for brain-type sodium channels in coupling of cell surface depolarization to contraction in the heart. Proc Natl Acad Sci U S A . 99(6):4073-8. Martens JR, Navarro-Polanco R, Coppock EA, Nishiyama A , Parshley L, Grobaski TD, Tamkun M M . (2000) Differential targeting of Shaker-like potassium channels to lipid rafts. J Biol Chem. 275(11):7443-6. 160 Maruoka ND, Steele DF, Au BP, Dan P, Zhang X , Moore ED, Fedida D. (2000) alpha-actinin-2 couples to cardiac Kvl .5 channels, regulating current density and channel localization in H E K cells. FEBS Lett. 473(2): 188-94. Marx S. (2003) Ion channel macromolecular complexes in the heart. J Mol Cell Cardiol. 35(l):37-44. Mason HS, Latten MJ, Godoy L D , Horowitz B, Kenyon JL. (2002) Modulation of Kvl .5 currents by protein kinase A , tyrosine kinase, and protein tyrosine phosphatase requires an intact cytoskeleton. Mol Pharmacol. 61(2):285-93. Mays DJ, Foose JM, Philipson L H , Tamkun M M . (1995) Localization of the Kvl .5 K + channel protein in explanted cardiac tissue. J Clin Invest. 96(l):282-92. McDonough A A , Zhang Y , Shin V , Frank JS. (1996) Subcellular distribution of sodium pump isoform subunits in mammalian cardiac myocytes. Am J Physiol. 270(4 Pt l):C1221-7. Minor DL, Lin Y F , Mobley BC, Avelar A , Jan Y N , Jan L Y , Berger JM. (2000) The polar TI interface is linked to conformational changes that open the voltage-gated potassium channel. Cell. 102(5):657-70. Mohler PJ, Rivolta I, Napolitano C, LeMaillet G, Lambert S, Priori SG, Bennett V (2004) Navl.5 E1053K mutation causing Brugada syndrome blocks binding to ankyrin-G and expression of Navl.5 on the surface of cardiomyocytes. Proc Natl Acad Sci. 101(50): 17533-8. Nesti E, Everill B, Morielli A D . (2004) Endocytosis as a mechanism for tyrosine kinase-dependent suppression of a voltage-gated potassium channel. Mol Biol Cell. 15(9):4073-88. Osterweil E, Wells DG, Mooseker MS. (2005) A role for myosin VI in postsynaptic structure and glutamate receptor endocytosis. J Cell Biol. 168(2):329-38. Rohr S. (2004) Role of gap junctions in the propagation of the cardiac action potential. Cardiovasc Res. 62(2):309-22. Saffitz JE, Laing JG, Yamada K A . (2000) Connexin expression and turnover : implications for cardiac excitability. Circ Res. 86(7):723-8. Sperelakis N . (2002) An electric field mechanism for transmission of excitation between myocardial cells. Circ Res. 91(11):985-7. Strang C, Cushman SJ, DeRubeis D, Peterson D, Pfaffinger PJ. (2001) A central role for the TI domain in voltage-gated potassium channel formation and function. J Biol Chem. 2001 Jul 27;276(30):28493-502. Takimoto K, Fomina AF, Gealy R, Trimmer JS, Levitan ES. (1993) Dexamethasone rapidly induces Kvl .5 K + channel gene transcription and expression in clonal pituitary cells. Neuron. ll(2):359-69. 161 Tiffany A M , Manganas L N , Kim E, Hsueh YP, Sheng M , Trimmer JS. (2000) PSD-95 and SAP97 exhibit distinct mechanisms for regulating K(+) channel surface expression and clustering. J Cell Biol. 148(1): 147-58. Timpe L C , Fantl WJ. (1994) Modulation of a voltage-activated potassium channel by peptide growth factor receptors. JNeurosci. 14(3 Pt 1):1195-201. Van Wagoner DR, Pond A L , McCarthy P M , Trimmer JS, Nerbonne JM. (1997) Outward K + current densities and Kvl .5 expression are reduced in chronic human atrial fibrillation. Circ Res. 80(6):772-81. Wang Z, Eldstrom JR, Jantzi J, Moore ED, Fedida D. (2004) Increased focal Kv4.2 channel expression at the plasma membrane is the result of actin depolymerization. Am J Physiol Heart Circ Physiol. 286(2):H749-59. Wu H , Nash JE, Zamorano P, Garner CC. (2002) Interaction of SAP97 with minus-end-directed actin motor myosin VI. Implications for A M P A receptor trafficking. J Biol Chem. 277(34):30928-34 Wong W, Schlichter LC . (2004) Differential recruitment of K v l . 4 and Kv4.2 to lipid rafts by PSD-95. J Biol Chem. 279(l):444-52. 162 APPENDICES 163 A P P E N D I X A : Single Letter Amino Acid Sequence of human K v l . 5 MEIALVPLEN GGAMTVRGGD EARAGCGQAT GGELQCPPTA GLSDGPKEPA PKGRGAQRDA DSGV|RPLMMDPGV|RPLPP LIEELPRPRR PPPEDEEEEG DPGLGTVEDQ ALGTASLHH 4Q RVHINISGLR FETQLGTLAQ FPNTLLGDPA KRLPYFDPLR NEYFFDRNRP SFDGILYYYQ SGGRLRRPVN VSLDVFADEI RFYQLGDEAM ERFREDEGFI KEEEKPLPRN EFQRQVWLIF siEYPESSGSAR AIAIVSVLVI ySJTFCLE TLPEFRDERE LLRHPPAPHQ P P A P A P G A B B@fflKf.PSGP J_VAPLLPRJL_ADPs2FFIVETT CVIWFTFELL VRFFACP^KAjGf^RNisaMNII DWAIFPYFITLGTELAE&Q PGGGGGGQNGQQAM_SLAS4lLR VIRLVRVFRI FKLSRHSKGL C^LGjm.QASMREssLGLLIFF LFIGVILFSS AVYFAEADNQ GTHFSSIPDA FWWAVMTMTT VGYGPDMRPIT VGGKselVGSLC AIAGVLTIAL PVPVIVSNFN YFYHRETDHE EPAVLKEEQG T Q S Q G P G L D R G V Q R K V S G S R G S F C K A G G T L ENADSAgRGS^CP^LEKCNVKA KSNVD |LRRSL 7 YALCLDTSRE TDW Figure 6.1. Single-letter amino acid sequence of human K v l . 5 . Key:_ExfracgHular_lpp^ loops; Transmembrane Domains S1-S6; Turret Region; 1 .Overlapping WW and SH3 proline-rich binding domains; 2. Start of the tetramerization domain; 3. Glycosylation site; 4. Potassium channel pore signature sequence; 5. P K A consensus phosphorylation site; 6. Potential 14-3-3 binding site; 7. Trafficking determinant; 8. C-terminal PDZ binding domain 164 APPENDIX B: Online Supplemental Material To Chapter 4 This material represents supplemental data and methods for Chapter 4: Fedida D, Eldstrom J, Hesketh JC, Lamorgese M , Castel L , Steele DF, Van Wagoner DR. (2003) Kvl .5 is an important component of repolarizing K + current in canine atrial myocytes. Circ Res. 93(8):744-51. This material is also available online at http://www.circresaha.org. Channel expression in cell lines and electrophysiological recording conditions hKvl.5 channels were studied in a human embryonic kidney cell line (HEK293) as reported previously (Fedida et al., 1993). Rat Kv4.2, rat Kv2.1, rat Kv3.1, human Kvl .4 , Kv l .2 and human HI N a + channels were also stably expressed in HEK293 cells. The levels of channel expression for the Kvl .4 , Kv l . 2 , and Kv3.1 cell lines used in Figure 4.1A were comparable at 15.9 ± 2.7 nA/cell (+60 mV, n = 17), 13.6 ± 1.4 nA (+60 mV, n = 5), and 0.1 ± 2.2 nA (+10 mV, n = 4) respectively. Cells were dissociated for passage by using trypsin-EDTA and were maintained in minimum essential medium (MEM), 10% fetal bovine serum, penicillin-streptomycin and 0.5 mg ml"1 geneticin at 37 °C in an atmosphere of 5% CO2 in air. A l l tissue culture supplies were obtained from Invitrogen (Burlington, ON, Canada). For recording K + currents, the bath solution contained, in mM, 135 NaCl, 5 KC1, 2.8 Na-acetate, 2 CaCb, 1 MgCb, its pH was adjusted to 7.4 with NaOH, and 10 mM HEPES was used as buffer. The pipette solution contained, in mM, 130 KC1, 5 EGTA, 10 HEPES, 4 Na2ATP, and 1 MgChand was adjusted to pH 7.2 with K O H . Chemicals were from Sigma Aldrich Chemical Co. (Oakville, ON, Canada). Coverslips containing cells were removed from the incubator before the experiments and placed in a superfusion chamber containing the control bath solution at 22 - 23° C. The bath solution was constantly flowing through the chamber and the solution was exchanged by switching the perfusate at the inlet of the chamber, with complete bath solution changes taking -10 s. Whole-cell current recording and data analysis were done using an Axopatch 200B 165 amplifier and pClamp8 software (Axon Instruments, Foster City, CA). Patch electrodes were fabricated using thin-walled borosilicate glass (WPI, Sarasota FL) with electrode resistances of 1-2 MQ,. Capacitance and series resistance compensation, typically 80%, were routinely used. Data were filtered at 10 kHz and sampled at 50 kHz for all protocols. Leak subtraction was not used. Data are shown as mean ± s.e.m. To record currents from K v l . 5 , Kv2.1, Kv3.1, and Kv4.2 channels, cells were depolarized at 1 Hz from the holding potential of -80 mV to a voltage of +50 mV for 200 ms to fully open the channel. After 200 pulses of control recording, when a stable current level is reached, one concentration of drug was washed in. In the presence of drug, the recording was maintained for 100 pulses or until the current level was stable. After the concentration-response curve was complete, the solution was returned to control to observe washout. For hERG channels, cells were depolarized from the holding potential of -80 mV to a voltage of+20 mV for 4 s to fully open and inactivate the channel, then stepped back to -50 mV for 6 s to record the tail current. The inter-pulse interval between each trace was 15 s. After control recordings (usually 10 traces) when the current was stable, the drug was washed in. In the presence of drug, current was recorded until it reached a steady state (depending on the drug, usually 5-10 traces). After the concentration-response curve was complete, the solution was returned to control to observe washout. For hHl Na + channels, cells were depolarized from the holding potential of -100 mV to a voltage of-30 mV for 10 ms to fully open and inactivate the channel. The inter-pulse interval between each trace was 1 s (1 Hz). After control current was recorded, there was a one minute rest while the drug was washed in (stimulation at 0.25 Hz), then current in the presence of the drug was recorded at 1 Hz. In control, 10-20 traces were usually collected; in the presence of a drug, sometimes more than 100 traces were collected to make sure that the current reached steady-state. When the drug effect was stable, the concentration was increased in a half-log manner. Again, after the concentration-response curve 166 was complete, the solution was returned to control to observe washout. Preparation of canine heart and brain membrane and cytosol fractions Heartworm-negative mongrel dogs (20-30 kg) were anesthetized with intravenous pentobarbital sodium (30 mg/kg) and fentanyl citrate (15 fig/kg) and placed on positive-pressure ventilation. A catheter was inserted into the right femoral artery, and the dogs were exsanguinated. A left lateral thoracotomy was performed through the fifth intercostal space, and the heart was arrested. The heart and lungs were removed from the thorax en bloc and perfused with cardioplegic solution. The atria were dissected from the ventricles and either snap frozen in liquid N 2 for the biochemical studies, or utilized for the preparation of isolated atrial myocytes (see below). Brain tissue was obtained from the same animals, following removal of the heart and lungs. A l l surgical procedures and experimental protocols were approved by the Cleveland Clinic Foundation Institutional Animal Care and Use Committee (Cleveland, OH). Approximately 0.2 g of frozen heart, isolated atrial myocytes, or brain tissue was minced on dry ice and homogenized immediately in 5 mL lysis buffer (lOrnM TrisHCl, pH 8.0, 140 mM NaCl, 1% hemoglobin (Sigma), 1 mM iodoacetamide, 0.2 trypsin inhibitor units/mL aprotinin, 1 mM PMSF) using an Ultra-Turrax T25 homogenizer (IKA Laboratechnik, Germany). Nuclei and cell debris were pelleted at lOOOg for 10 minutes. The supernatant was centrifuged at 20,000g for 1 hour. The supernatant ("cytosol"), which also contains SR in a microsomal fraction, was removed from the membrane pellet and the fractions run on SDS-PAGE gels for Western analysis. A l l homogenization and fractionations were carried out at 4°C. In most Western blots, due to the presence of 1% hemoglobin in the lysis buffer direct quantitation of sample protein content was not possible. In order to load roughly equal amounts of protein, PVDF membranes were stained with Coomassie blue after transfer, and loading sample volumes adjusted as necessary based on visual assessment. 167 Canine myocyte preparation Atrial myocytes were dissociated from canine atrial specimens using a chunk dissociation technique (Van Wagoner et al., 1999). Briefly, tissue was minced in a low Ca2 +dissection buffer (DB) containing (in mmol/L): 134 sucrose, 35 NaCl, 25 NaHCCb, 16 Na 2HP04, 4.75 KC1, 1.2 KH2PO4, 10 HEPES, 10 glucose, pH 7.40 with NaOH. The tissue was first exposed to a solution containing protease (Sigma type X X I V ) and collagenase (type II, Worthington Biochemical, Lakewood NJ) for 45 minutes at 31°-33°C. Supernatant from this enzyme exposure was aspirated and discarded, and fresh enzyme solution containing only collagenase was added to the tissue for six sequential exposures of 10 min each. The supernatant of each of these digestions was collected by aspiration, and myocytes were concentrated by low speed centrifugation. Myocytes were kept oxygenated at room temperature until used, within 8 hours of isolation. Yields were in the range of 20-50% for viable, calcium tolerant myocytes. Only well striated, rod-shaped myocytes were used in the electrophysiological studies. Canine atrial voltage / current clamp recordings Myocytes were allowed to adhere to laminin-coated 35 mm culture dishes mounted in a thermal stage controller (Bioptech AT system, Butler, PA), maintained at 35°C and gassed with 100%) 0 2 . Solutions were changed via a six-port gravity flow system. Recordings were performed under continuous flow conditions. Patch pipettes were prepared from Corning 8161 glass (WPI), and the shanks were covered with Sylgard. Tips were fire-polished immediately before use to an access resistance of 2-3 M f l , when filled with the pipette solution. Action potentials were recorded in current clamp mode using conventional whole-cell recording conditions, with a pipette solution containing (in mmol/L): 140 KC1, 4 mM MgATP, 4 mM creatine phosphate, 3 mM Na-pyruvate, 1 mM MgCb, and 50 p M EGTA, pH 7.20 with 168 K O H . Recordings were begun after 4-5 minutes dialysis, when the membrane potential and access resistance had stabilized. The bath solution for these experiments contained (in mmol/L): 140 NaCl, 3 Na-acetate, 5 KC1, 5 Hepes, 5 glucose, 1 M g C l 2 , 2 CaCl 2 , pH 7.40 with NaOH. Action potentials were recorded at cycle lengths of 2, 1, and 0.5 seconds, under constant flow conditions, at a temperature of 35°C. At least 12 steady-state action potentials were recorded at each cycle length. These were averaged, and action potential amplitude and duration (20%, 50% and 90% repolarization) were measured from the averaged traces. Canine atrial potassium currents were recorded using the perforated patch recording technique, also at 35°C. Bath solutions were identical to those above, except for the addition of 0.2 mmol/L CdCl 2(to suppress Ica). The pipette solution contained (in mmol/L): 100 K-MES, 40 KC1, 10 Tris, 10 Hepes, 0.05 K 2 E G T A , 2 M g C l 2 , pH 7.2. Nystatin was added to the pipette solution at a final concentration of 100 pg/ml, from a stock solution made fresh daily. Once nystatin was added to the buffer, the pipette solution was sonicated (30 sec) and used within 3 hours. Nystatin-free pipette solution was placed in the tip of the pipette by capillary action (3-4 sec), and then nystatin-containing solution was backfilled in the pipette immediately prior to use. Junction potentials were nulled immediately before seal formation. After seal formation, increases in the capacitative response to a -10 mV step pulse (from a -60 mV holding potential) occur as nystatin perforates the patch. Cell capacitance and access resistance were checked throughout the experiment by tuning the patch clamp amplifier with small square wave voltage steps. R N A isolation R N A was isolated from heart or brain samples using the RNeasy Midiprep kit (Qiagen, Mississauga ON, Canada) according to the manufacturer's instructions. Briefly, 140 mg frozen tissue samples were minced on dry ice then homogenized in RLT buffer (provided in the kit) at 169 maximum speed. After Proteinase K digestion, R N A was purified from the sample using RNeasy spin columns and eluted in 150 uL diethylpyrocarbonate (DEPC) -treated water and treated with DNAase I in accordance with the RNeasy kit instructions (Invitrogen). For some experiments, polyA R N A was further purified on oligo(dT)-cellulose spin columns (Amersham Biosciences, Baie d'Urfe, Quebec, Canada). R T - P C R Invitrogen's Superscript One-Step RT-PCR with Platinum Taq kit was used for most RT-PCR experiments. Whole-cell RNA fractions were used for all Kv l .5 amplifications and for some Kv3.1 amplifications. Purified polyA RNA was used in the remainder of the Kv3.1 experiments. For Kvl .5 experiments, reverse transcription was primed with a D N A oligonucleotide of sequence CTTCGGGCACTGTCTGCATTC (corresponding to nucleotides 1727 to 1747 in our Kvl .5 clone), obtained from the sequence of hKvl.5. The same oligonucleotide plus another of sequence T G G A G A C C A C G T G C G T G A T C T (corresponding to nucleotides 994-1014) were used for PCR amplification of the reverse transcript. Kv3.1 reverse transcription was attempted with five separate oligonucleotides: 1. C G G A A T T C T G G T C A A G T C A C T C T C A C A G C C T C (nucleotides 1738-1760) 2. G C T C T A G A G G T C A A G T C A C T C T C A C A G C C T C (nucleotides 1738-1760) 3. C G G A A T T C CTTTGGTAGTTTCTGCTTAGC (nucleotides 1342-1362) 4. G G T C A A G T C A C T C T C A C A G C C T C (nucleotides 1738-1760) 5. CTTTGGTAGTTTCTGCTTAGC (nucleotides 1342-1362) For PCR amplification of Kv3.1 reverse transcripts, the oligonucleotides of sequence: 6. C C C A A G C T T A T G G G C C A A G G G G A C G A G A G (nucleotides 1-18) 7. C C C A A G C T T CTGCTGCTTATCATCTTCCTG (nucleotides 1036-1056) 8. C C C A A G C T T A T G G G C C A A G G G G A C G A G , (nucleotides 1-18) 170 9. A T G G G C C A A G G G G A C G A G A G (nucleotides 1-18) 10. CTGCTGCTTATCATCTTCCTG (nucleotides 1036-1056) respectively, were paired with those used for RT. These should amplify -0.7 and 1.4 kb fragments of the Kv3.1 transcript. An additional set of oligonucleotide pairs, designed to amplify the entire 1.8 kb coding sequence of the channel in a single reaction was used in some experiments. Amplification conditions were as recommended by the manufacturer. An annealing temperature of 59°C was employed for Kvl .5 PCR; a wide variety of annealing temperatures, ranging from 45 to 61°C were attempted in Kv3.1 experiments. For some experiments with Kv3.1, separate reverse transcription and amplification steps were performed. In these experiments, a variety of annealing temperatures, ranging from 45° to 60°, and magnesium concentrations, ranging from 2 to 4 mM, were employed. For amplification of canine dystroglycan from the same heart RNA preparations, reverse transcription was primed with an oligonucleotide of sequence C A T T G T T C T G G T T C A G G G A G , corresponding to nucleotides 214-233 in the coding sequence and 130927-46 in the published genomic clone (Leeb et al., 2000). The same oligonucleotide plus another of sequence G G G C G C T C A T T T C G A G T G A C (corresponding to nucleotides 783-802 in the coding sequence and 130927-46 in the genomic sequence) were used for PCR amplification of the reverse transcript. Fixing, immunolabeling, and imaging of myocytes and HEK293 cells For comparison of Kvl .5 antibodies with an anti-T7 antibody using immunocytochemistry the T7 Tag sequence was inserted upstream of the Kvl .5 start codon, in frame such that the 11 amino acids of the T7 Tag (Bold) plus a linker sequence precedes the first methionine of the channel (underlined), M A S M T G G Q Q M G R G S E F E L R R Q A S R A M . A stable cell line was generated from transfected HEK293 cells growing on sterile cover slips in 35 mm 171 petri dishes. Cells were transfected with 1.5 ug of T7-Kvl.5 in pcDNA3 plasmid D N A using the polycationic lipid preparation, LipofectAMINE™ 2000 Reagent (Invitrogen) as recommended by the supplier. Two days after transfection 0.5 mg/mL geneticin was added to the growth media to select for transfected cells. After 10 days cells were tested for expression of the tagged protein. Cells were prepared according to previously published methods (Scriven et al., 2000). Briefly, HEK293 cells seeded onto coverslips or myocytes in suspension were fixed with 2 or 4% paraformaldehyde for 10 min followed by neutralization with glycine buffer for 10 min. They were then washed with phosphate buffered saline (PBS; 137 mM NaCl, 2.7 mM KC1, 4.3 mM Na 2 HP0 4 , 1.4 mM K H 2 P 0 4 ) for 10 min, made permeable with 1 uL/mL Triton-X in PBS (10 min), washed again for 10 min with PBS, and finally stored in PBS-azide solution. Myocytes were plated onto poly-L-lysine coated coverslips and were blocked, then labeled for 3 hours at room temperature or overnight at 4°C with one of the following primary antibodies: mouse monoclonal anti-T7 (1:1000, Novagen, Madison,WI); mouse monoclonal anti-Kvl.5 (1:200; Upstate); rabbit polyclonal anti-Kvl.5 (1:200; Alomone and Upstate); or a rabbit antibody developed in our laboratory against the C-terminus of hKvl.5 (aa 537-553 EQGTQSQGPGLDRGVQR; 1:500). Secondary Alexa 594 conjugated goat anti-mouse, Alexa 488, or Texas red-conjugated goat anti-rabbit (Molecular Probes, Eugene, OR, 1:1000 dilution; Jackson Immunoresearch, West Grove, PA, USA, 1:200 dilution), antibodies were incubated for 2 hours at room temperature. Cells were washed three times with PBS and then incubated in 300 nM 4',6-diamidino-2-phenylindole, dihydrochloride (DAPI; Molecular Probes) in PBS for 5 min before 3-5 min washes with PBS. Cells were once again washed three times prior to mounting with a 90% glycerol, 2.5% w/v DABCO-PBS solution. Images of labeled cells were taken using a Bio-Rad Radiance Plus on an inverted Zeiss Axiovert microscope, equipped with a Mai Ti Sapphire laser (Spectra Physics, Mountain View, CA) using BioRad LaserSharp 2000 software. Images were prepared using NIH Image and Photoshop Software Packages. 172 C-term Ab C-term Ab + peptide Figure 6.2. K v l . 5 expression in canine atria and rat ventricle. Extracts from homogenized canine atrial (CA) and rat ventricular (RV) tissues were subjected to Western analysis. Paired western blots were probed with our C-Term anti-Kvl.5 antibody with (right) and without (left) pre-incubation of the antibody with blocking peptide. 173 R E F E R E N C E S Fedida D, Wible B, Wang Z, Fermini B, Faust F, Nattel S, Brown A M . (1993). Identity of a novel delayed rectifier current from human heart with a cloned K + channel current. Circ Res. 73:210-216. Leeb T, Neumann S, Deppe A, Breen M , Brenig B. (2000). Genomic organization of the dog dystroglycan gene DAG1 locus on chromosome 20-ql5.1-ql5.2. Genome Research. 10:295-301. Scriven DR, Dan P, Moore ED. (2000). Distribution of proteins implicated in excitation-contraction coupling in rat ventricular myocytes. Biophys J. 79:2682-2691. Van Wagoner DR, Pond A L , Lamorgese M , Rossie SS, McCarthy P M , Nerbonne JM. (1999). Atrial L-type Ca2+currents and human atrial fibrillation. Circ Res. 85:428-436. 174 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092336/manifest

Comment

Related Items