UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The structure of excitation-contraction coupling in atrial cardiomyocytes Schulson, Meredith Nicole 2009

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2009_spring_schulson_meredith.pdf [ 2.56MB ]
Metadata
JSON: 24-1.0066907.json
JSON-LD: 24-1.0066907-ld.json
RDF/XML (Pretty): 24-1.0066907-rdf.xml
RDF/JSON: 24-1.0066907-rdf.json
Turtle: 24-1.0066907-turtle.txt
N-Triples: 24-1.0066907-rdf-ntriples.txt
Original Record: 24-1.0066907-source.json
Full Text
24-1.0066907-fulltext.txt
Citation
24-1.0066907.ris

Full Text

THE STRUCTURE OF EXCITATION- CONTRACTION COUPLING IN ATRIAL CARDIOMYOCYTES  by  Meredith Nicole Schulson  B.Sc., McGill University, 2005    A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE  in  The Faculty of Graduate Studies  (Physiology)    THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  January 2009    © Meredith Nicole Schulson, 2009  ABSTRACT  Standard local control theory, which describes Ca2+ release during excitation- contraction coupling (ECC), assumes that all Ryanodine Receptor (RyR) complexes are equivalent.  Recent data from our laboratory has called this assumption into question. Specifically, we have shown that RyR complexes in ventricular myocytes differ depending on their location within the cell.  This, and other data, has led us to hypothesize that similar differences occur within the rat atrial cell. To test this hypothesis, we have triple-labeled enzymatically-isolated, fixed myocytes to examine the distribution and colocalization of RyR, calsequestrin (CSQ), voltage-gated Ca2+ channels (Cav1.2), sodium-calcium exchangers (NCX), and caveolin- 3 (cav-3).  All images were acquired on a wide-field microscope, deconvolved, and subject to extensive analysis, including a novel method of measuring statistical significance of the recorded colocalization values. Overall, eight surface RyR populations were identified, depending on its binding partners.  One of these groups, in which RyR, Cav1.2, and NCX colocalize, may provide the structural basis for ‘eager’ sites of Ca2+ release in atria, while other groups were defined based on their association with cav-3, and are therefore highly likely to be under the influence of other signaling molecules located within caveolae.  Importantly, although a small portion of the surface RyR in atria do colocalize with NCX alone, the majority are tightly linked to Cav1.2 alone or Cav1.2 and NCX together.  Therefore, it appears likely that Cav1.2-mediated calcium-induced calcium release (CICR) is the primary method of initiating Ca2+ release from the SR during EC coupling.     ii TABLE OF CONTENTS ABSTRACT....................................................................................................................... ii TABLE OF CONTENTS ................................................................................................ iii LIST OF TABLES ............................................................................................................ v LIST OF FIGURES ......................................................................................................... vi LIST OF ABBREVIATIONS ........................................................................................ vii ACKNOWLEDGEMENTS ............................................................................................. x CO-AUTHORSHIP STATEMENT ............................................................................... xi I INTRODUCTION.......................................................................................................... 1 I .I Physiological role of atria......................................................................................... 1 I.2 Ultrastructure of atrial myocytes .............................................................................. 2 I.3 Excitation-contraction coupling................................................................................ 4 I.4 Molecular components of EC coupling ..................................................................... 6 I.4.1 RyR..................................................................................................................... 6 I.4.2 Calsequestrin (CSQ), Triadin (trd), and Junctin (jcn) ........................................ 7 I.4.3 L-type calcium channel ...................................................................................... 8 I.4.4 NCX.................................................................................................................... 9 I.4.5 SERCA ............................................................................................................. 11 I.4.6 Caveolae ........................................................................................................... 12 I.5 EC coupling in atrial cardiomyocytes ..................................................................... 14 I.5.I Ca2+ signaling patterns ...................................................................................... 14 I.5.2 Junctional vs. non-junctional RyR.................................................................... 14 I.5.3 Modulation of atrial Ca2+ transients ................................................................. 15 I.5.4 Ca2+ sparks in atrial myocytes .......................................................................... 17 I.5.5 ‘Blocking’ of Ca2+ transients ............................................................................ 19 I.6 Hypothesis and specific aims................................................................................... 21 I.6.1 Hypothesis ........................................................................................................ 21 I.6.2 Specific aims..................................................................................................... 21 I.7 References................................................................................................................ 22 2 MULTIPLE SURFACE RYANODINE RECEPTOR POPULATIONS EXIST IN RAT ATRIAL MYOCYTES ......................................................................................... 27 2.I Introduction ............................................................................................................. 27 2.2 Results ..................................................................................................................... 29 2.2.I Controls ............................................................................................................ 29 2.2.2 Triple-labeling experiments ............................................................................. 32   iii 2.3 Discussion ............................................................................................................... 37 2.3.1 Imaging ............................................................................................................ 37 2.3.2 Structural organization of atrial myocytes....................................................... 38 2.3.3 Functional significance of atrial RyR subpopulations ..................................... 40 2.3.4 Conclusion ....................................................................................................... 43 2.4 Materials and Methods ........................................................................................... 45 2.4.I Cell isolation and preparation........................................................................... 45 2.4.2 Immunolabeling ............................................................................................... 46 2.4.3 Imaging ............................................................................................................ 47 2.4.4 Image deconvolution and analysis ................................................................... 47 2.5 Tables...................................................................................................................... 52 2.6 Figures .................................................................................................................... 55 2.7 References ............................................................................................................... 67 3 DISCUSSION ............................................................................................................... 69 3.I Imaging.................................................................................................................... 69 3.2 Data Analysis .......................................................................................................... 71 3.3 Structural organization of atrial myocytes ............................................................. 72 3.4 Functional significance of atrial RyR subpopulations............................................ 75 3.5 Directions of future study........................................................................................ 78 3.6 Summary ................................................................................................................. 79 3.7 References ............................................................................................................... 81 APPENDICES................................................................................................................. 83 Appendix 1.  Composition of Solutions ......................................................................... 83 Appendix 2.  List of antibodies used ............................................................................. 90 Appendix 3. List of antibodies that do not work for immunofluorescence of rat cardiomyocytes ............................................................................................................. 91 Appendix 4. Protocols and control experiments for dual- and triple-labeling of atrial myocytes........................................................................................................................ 92 Appendix 5.  Figures from control experiments............................................................ 94 Appendix 6. Cav1.2 antibody production ...................................................................... 97 A6.I Rationale ........................................................................................................... 97 A6.2 Protocol ............................................................................................................ 97 A6.3 Results.............................................................................................................. 99 A6.4 Discussion ...................................................................................................... 103 A6.5 Figures............................................................................................................ 104 A6.6 References...................................................................................................... 108   iv LIST OF TABLES  TABLE2.1 Colocalization of RyR with NCX and Cav1.2 …………………………...55 TABLE 2.2 Colocalization of RyR with CSQ and NCX …………………………… 55 TABLE 2.3 Colocalization of RyR with Cav-3 and Cav1.2 …………………………56 TABLE 2.4 Colocalization of RyR with NCX and Cav-3 …………………………...56 TABLE 2.5 Colocalization in Atria vs. Ventricle ……………………………………57                           v LIST OF FIGURES  Fig. 2.1 Single labeling of atrial myocytes …………………………………………….58 Fig. 2.2 Dual labeling of RyR and NPC……………………………………………….59 Fig. 2.3 Density of antibody labeling across experiments ……………………………60 Fig. 2.4 Triple labeling of RyR, Cav1.2, and NCX……………………………………61 Fig. 2.5 Colocalization of RyR, Cav1.2, and NCX ……………………………………62 Fig. 2.6 Triple labeling of RyR, CSQ, and NCX……………………………………...63 Fig. 2.7 Colocalization of RyR, CSQ, and NCX………………………………………64 Fig. 2.8 Triple labeling of RyR, Cav-3, and Cav1.2. ………………………………….65 Fig. 2.9 Colocalization of RyR, Cav-3, and Cav1.2……………………………………66 Fig. 2.10 Triple labeling of RyR, NCX, and Cav-3…………………………………...67 Fig. 2.11 Colocalization of RyR, NCX, and Cav-3……………………………………68 Fig. 2.12 Method of measuring colocalization. ……………………………………….69 Fig.A5.1 Images from control experiments ……………………………………….….98 Fig.A5.2 Double colocalization values across experiments ……………………….…99 Fig. A6.1 Pre-immune serum western blots…………………………………………107 Fig. A6.2 Western blots from animal M1……………………………………………108 Fig. A6.3 Western blots from animal M2……………………………………………109 Fig. A6.4 Western blots from antibody purification column….…………………...110        vi LIST OF ABBREVIATIONS Ab   Antibody AF   Atrial Fibrillation AP   Action Potential β-AR   Beta-Adrenergic Receptor [Ca] i   Intracellular calcium concentration [Ca] SR  Intra-SR calcium concentration Cav1.2  Voltage-gated calcium channel; α1C subunit Cav-3   Caveolin-3 CFA   Complete Freund’s Adjuvant CICR   Calcium-Induced Calcium Release CPA   Cyclopiazonic acid CSQ   Calsequestrin DABCO  1,4-Diazabicyclo [2.2.2] Octane Em   Membrane potential ECC   Excitation-Contraction Coupling EM   Electron Microscopy EGTA  Ethylene Glycol-bis (2-aminoethylether) – N,N,N’,N’-                                Tetraacetic Acid ET-1   Endothelin-1   vii FWHM  Full Width Half Maximum HRP   Horseradish Peroxidase ICa, L   Current through voltage-gated calcium channels IF   Immunofluorescence IFA   Incomplete Freund’s Adjuvant ISO   Isoproterenol jcn   Junctin jSR   Junctional Sarcoplasmic Reticulum kDa   Kilodalton LTCC  L-type Calcium Channel MeOH  Methanol Mono   Monoclonal MW   Molecular Weight [Na] i   Intracellular sodium concentration NCX   Sodium-Calcium Exchanger ND   Not determined NKA   Na+/K+ ATPase NPC   Nuclear Pore Complex proteins O/N   Overnight Po   Open probability   viii PKA   Protein Kinase A Poly   Polyclonal PBS   Phosphate Buffered Saline PIS   Pre-Immune Serum PLB   Phospholamban PSF   Point Spread Function PSS   Physiological Saline Solution PVDF  Polyvinylidene fluoride RT   Room Temperature RyR   Ryanodine Receptor SERCA  Sarco/Endoplasmic Reticulum ATP-ase SR   Sarcoplasmic Reticulum trd   Triadin T-tubule  Transverse Tubule TATS  Transverse-Axial Tubular System TBS   Tris-Buffered Saline TBS-T  Tris-Buffered Saline with 10% Tween-20 (v/v)   volume/volume (w/v)   weight/volume WB   Western Blot   ix ACKNOWLEDGEMENTS  I would like to take this opportunity to thank the many people who have helped me throughout the course of this degree.  Without them, this simply would not have been possible.  My sincere gratitude must first go to my supervisor, Dr. Edwin Moore.  His constant support, understanding, and generosity has kept me going through the rough patches, and his tremendous teaching ability has been invaluable. Second, I would like to thank the other members of my supervisory committee – Drs. Steve Kehl, Eric Accili, and Wayne Vogl – for the time and energy they have committed, and the support they have given to myself and this project. Third, I must extend many thanks to my fellow lab members, who have committed valuable time and effort in helping me finish this project.  I must especially thank Dr. David Scriven, for his tireless work and endless patience.  His intellect and technical skills have helped me tremendously along my path, but most of all David has kept me smiling.  My particular thanks also go to Patrick, who put in many extra hours of hard work on my behalf, and to Parisa, who has taught me so much.  She has been an inspiration to me, but most of all, an extraordinary friend. Finally, I would like to thank the most important people in my life, my parents. Without their unconditional love and support I would not be where or who I am today. They are my guiding light along this journey called life.        x  CO-AUTHORSHIP STATEMENT Chapter 2 of this manuscript (Multiple Surface RyR Populations Exist in Rat Atrial Myocytes) was co-authored by P.Fletcher, D.R.L. Scriven, and E.D.W. Moore. My contribution was to the design of the experiments, the execution of all experiments and data analysis, and the writing of the manuscript. E.D.W. Moore was involved in the identification of the research problem, and the design of the experiments.  D.R.L. Scriven and P.Fletcher designed the statistical program used to analyze the data.    xi I INTRODUCTION I .I Physiological role of atria In the heart, the thin-walled atria are critically linked to the normal circulatory function of the body.  During a cardiac cycle, most of the blood flows passively from atria to ventricles.  Under conditions where cardiac output must increase however, the atria contract more forcefully and actively contribute to ventricular refilling.  Highly adaptable, these low-pressure chambers can contribute anywhere from 10-50% of the total stroke volume (Godtfredsen, 1999).  This ‘atrial kick’ becomes even more significant in individuals with decreased ventricular compliance (Alpert et al., 1988). Perhaps because of the enormous adaptability of the atria, they quite easily fall into a state of cardiac dysrhythmia known as atrial fibrillation (AF) (Godtfredsen, 1999). During AF, electrical impulses arise from multiple sites inside the atria at very high frequency (~ 5 X higher than normal) and as a result, the atria do not contract coordinately (Bootman et al., 2006).  Failure of the atria to contract normally, though unlikely to immediately influence hemodynamics, will over time become highly detrimental.  In both physiological (e.g. ageing) and pathological (e.g. mitral stenosis) states, loss of atrial contribution to stroke volume because of AF can be deadly.  In particular, a loss of pumping capacity of the heart or pooling of blood in the atria and potential clotting may have lethal consequences (Bootman et al., 2006) (Godtfredsen, 1999).  Although the particular trigger of AF is often unknown, it likely involves mechanical and/or electrical abnormalities at the cellular level, or the linking of the two, within atrial cardiomyocytes (Waktare and Camm, 1999).  Like any pathology, it is vital to ascertain the normal physiology of the system in order to effectively understand the abnormal condition.  In the case of AF, a disease that currently affects approximately 2.5   1 million American adults and is the most common arrhythmia in elderly people (Go et al., 2001), elucidating the precise physiological mechanism by which atrial myocytes undergo excitation-contraction coupling is of utmost importance if we wish to make progress in understanding this disease. I.2 Ultrastructure of atrial myocytes Characterizing the basic structure of cardiomyocytes largely came through a number of electron microscopy (EM) studies on these cells from about the 1960s until the late 1990s.  Until this time, little attention had been paid to distinguishing the differences between atrial and ventricular myocytes.  In 1969, Hibbs and Ferrans observed a number of different types of rat atrial myocytes ranging from ‘clear, Purkinje-like cells to compact, electron-dense cells resembling ventricular muscle cells’.  Though this observation was consistent with researchers before them, they felt that the morphological differences in these cells were too great for a rigid classification system.  This particular study also demonstrated that atrial myocytes had no deep invaginations of the plasma membrane, or transverse tubules (t-tubules), like the ones observed in ventricular cells. They did however, have a number of flattened projections of the sarcoplasmic reticulum (SR) that lay just underneath the sarcolemmal membrane, much like in the ventricle.  A study the following year (Forssmann and Girardier, 1970) found that while most atrial cells did in fact lack t-tubules, a small portion actually contained a sophisticated network of tubules similar to that in ventricular cells.  In fact, these authors also discovered that both atrial and ventricular cardiomyocytes actually had a much more complex system that consisted of both transverse and longitudinal tubular elements (later called transverse- axial tubular system or TATS; (Forbes et al., 1984)).  At this time, although the   2 mechanism was unclear, it was suggested that the role of the tubular system was to synchronize contraction within cardiac cells.  The ramifications of not having a sophisticated T system in atrial myocytes were not brought up in this particular study, but it had been previously suggested that because of the small diameter of these cells, an extensive tubular system that propagates electrical conduction quickly through the cell might be unnecessary (Nelson and Benson, 1963).   It is now well known that the SR organizes itself into four major structural components: junctional, corbular, network, and cisternal (Bootman et al., 2006).  Areas of the SR that contain RyR channels and closely associate with either plasmalemma or t- tubules are termed junctional (jSR).  Corbular SR appears as small swellings of SR within the cell that also bear RyR but are not associated with surface or t-tubular membranes.  Running alongside myofibrils is an extensive meshwork of SR termed network SR, while flattened sacs of SR are called cisternal SR (Yamasaki et al., 1997) (Jorgensen et al., 1993) (Franzini-Armstrong et al., 2005).  Each of these SR components are present in both ventricular and atrial muscle cells, but to a varying extent.  For example, atrial myocytes appear to have more corbular SR components, but a less sophisticated network SR arrangement (Yamasaki et al., 1997).  Interestingly, calsequestrin, the main SR Ca2+ buffering protein, has been shown to specifically localize to the lumen of only junctional and corbular SR components and it has been hypothesized that corbular SR represents an additional site of SR Ca2+ release during EC coupling, away from the dyads (Jorgensen et al., 1985). Both atrial and ventricular myocytes have numerous couplings between peripheral elements of the SR and the plasma membrane (Leeson, 1980).  In atrial cells however, a   3 distinct component of the SR seems to form transverse striations along the length of the cell.  Originally it was proposed that these regions of SR were actually dense clusters of corbular SR lining up along Z-lines, and they were called ‘Z-tubules’ (Forbes et al., 1990).  Eventually it was shown that Z-tubules are actually single transverse tubules formed from SR membrane that line up between individual sarcomeres (Yamasaki et al., 1997), and are present in every atrial cell. I.3 Excitation-contraction coupling Excitation-contraction coupling (ECC) refers to the linking of electrical stimuli to the mechanical contraction of a cell.  In a cardiomyocyte, the electrical stimulus is generally provided in the form of an action potential (AP) and mechanical contraction occurs as a result of calcium release into the myoplasm and subsequent physical movement of myofilaments within the cell (Bers, 2002).  Upon electrical activation of the sarcolemmal membrane, Ca2+ ions enter the cell via the opening of voltage-gated calcium channels (Cav1.2) and cause a much larger release of calcium from ryanodine receptors (RyR) located in the closely apposed ‘junctional’ SR (jSR); a process known as calcium- induced calcium release (CICR) (Fabiato, 1983). CICR has been modeled through a number of theoretical mechanisms, including a series of ‘common pool’ models in which one large universal pool of cytoplasmic Ca2+ exists (e.g. (Wong, 1981) (Hilgemann and Noble, 1987)).  For each of these models, the universal Ca2+ pool provides both the trigger for SR Ca2+ release, and is also the pool which calcium from both Cav1.2 and RyR enter into.  Here, CICR is a positive feedback system in which Ca2+ release continues until completion; in other words, an ‘all-or-none’ response.  None of these models however, can provide an explanation for experimental   4 evidence showing a graded SR Ca2+ release in response to varying levels of trigger Ca2+. In fact, at very high concentrations of trigger Ca2+, release from the SR is actually inhibited or inactivated (Fabiato, 1983).  This ‘Ca2+ paradox’, has been theoretically solved through the formation of a local control theory. In order to circumvent the common pool model, it is necessary to divide cellular Ca2+ pools in some way, and this was in part accomplished by theoretically placing Cav1.2 and RyR in very close proximity to one another (Stern, 1992).  Structurally coupling these channels allowed for the theoretical development of a very high local Ca2+ concentration ([Ca2+]), which would trigger SR Ca2+ release from RyR located only in that dyad.  As Ca2+ diffused away, it would be diluted as it entered the much larger intracellular Ca2+ pool, and propagation of CICR to neighbouring RyR would not occur (Stern, 1992).  This meant that each RyR cluster would be activated independently only through its association with closely apposed Cav1.2 channels and a local [Ca2+].  This theory was plausible, but it is the large amount of experimental evidence that has made it the most widely accepted theory of EC coupling today. Structural support for the local control theory originally came from EM and immunofluorescence studies showing that Cav1.2 and RyR channels are indeed coupled into dyads, separated by a space of approximately 10-15nm (Carl et al., 1995) (Sun et al., 1995).  Additional functional studies also provided support.  For example, in 1993, using a laser-scanning confocal microscope and an intracellular Ca2+ indicator dye, Cheng et al reported the presence of spontaneous local increases in intracellular Ca2+ concentration ([Ca]i).  These Ca2+ ‘sparks’ appeared to be a result of spontaneous openings of a single or small cluster of RyR, and were restricted to a small volume of the cell approximately   5 1.5 µm in radius (Cheng et al., 1993) Further investigation revealed that Ca2+ sparks could also be electrically invoked during an action potential, and that whole-cell Ca2+ transients appeared to be the temporal and spatial summation of individual, depolarization-induced sparks (Cannell et al., 1995) (Lopez-Lopez et al., 1995).   It appeared from these studies that the amplitude of the current from a single voltage-gated, or ‘L-type’ Ca2+ channel (LTCC), rather than whole cell current amplitude, was a significant factor in determining the probability of Ca2+ release from the SR, and that the most efficient trigger Ca2+ may be a high local concentration of Ca2+ within the dyad (Lopez-Lopez et al., 1995), which would form a Ca2+ signaling ‘microdomain’.  Although this theory can largely explain why CICR does not propagate between separate dyads, the discovery that the SR releases only ~50-60% of its Ca2+ content (Bassani et al., 1995) implied that there still had to be an inherent shut-off mechanism for RyR channels in the dyad.  Currently, this shut-off mechanism remains unsolved, but two leading theories have emerged.  The first is one of RyR adaptation or inactivation that depends on local [Ca2+], and the second states that closure of RyR channels depends on the [Ca2+] within the lumen of the SR ([Ca2+]SR) (Bers, 2008). I.4 Molecular components of EC coupling      I.4.1 RyR The main SR Ca2+ release channels, RyR2, are grouped together in a cluster closely apposed to a number of LTCCs at each dyad in a cardiac myocyte.  The ratio of RyR: LTCC at each couplon ranges from 4-10, depending on the species (Bers, 2001). An individual RyR is activated by sub-micromolar Ca2+ concentrations, reaches maximal activity level at ~100μM Ca2+ and decreases at very high, non-physiological (Bers, 2008) Ca2+ concentrations (~5-10mM) (Xu et al., 1998) (Rousseau and Meissner, 1989).  Other   6 substances modulating the release of Ca2+ from the SR include ATP and caffeine (enhance release), and Mg2+ and calmodulin (CaM) (inhibit release) (for a review see Bers, 2001). Localized to jSR components, RyR in ventricular myocytes appear to line up directly underneath the sarcolemmal membrane and within the interior of the cell at the level of the Z-lines.  Here, they are closely apposed to LTCCs in the t-tubular membrane (Scriven et al., 2000).  In atrial myocytes, the overall distribution of RyR is largely the same, with one important difference.  Atrial RyR are separated structurally and functionally into: i) a small subsarcolemmal population coupled to LTCCs termed ‘junctional’ RyR, and ii) ‘non-junctional’ RyR; a larger population in the interior of the myocyte not apposed to any sarcolemmal membrane component (Carl et al., 1995; Mackenzie et al., 2001) (Hatem et al., 1997).  These two populations appear to be separated by an ~2 μm gap around the periphery of the cell directly underneath the sarcolemmal membrane (Mackenzie et al., 2001).  The functional characteristics of the two populations will be discussed in more detail in section I.5.2.      I.4.2 Calsequestrin (CSQ), Triadin (trd), and Junctin (jcn) Many of the regulatory proteins involved in the modification of RyR function are localized within a large macromolecular complex surrounding the channel.  These proteins include, but are not limited to, CSQ, trd, and jcn (Bers, 2004). Calsequestrin is the major intra-SR calcium buffering protein; each molecule can bind up to 20 Ca2+ ions with relatively low affinity (Mitchell et al., 1988).  CSQ is localized to the junctional face of the SR membrane (Franzini-Armstrong et al., 1987) and is physically coupled to RyR in a Ca2+ sensitive manner through the actions of   7 transmembrane proteins trd and jcn (Zhang et al., 1997).  A multitude of studies has shown that CSQ likely acts as a luminal SR Ca2+ sensor for RyR and directly regulates the open probability (Po) of the channel (Gyorke et al., 2004) (Terentyev et al., 2007).  A potential mechanism for regulating Po of RyR is that CSQ directly interacts with the luminal domain of trd under conditions of low [Ca2+]SR and inhibits the release of Ca2+ through RyR (Terentyev et al., 2007).  Junctin also appears to help regulate the Ca2+ sensitivity of RyR as an overexpression of this protein impairs Ca2+ homeostasis (Gergs et al., 2007). In rat ventricular myocytes, immunolabeling shows CSQ distribution mainly along Z-lines largely colocalized with RyR (Scriven et al., 2000), which corresponds with its positioning at the jSR membrane.  Triadin also appears to be distributed along Z-lines in transverse bands spaced approximately every 2µm along the cell, very closely related to RyR, in rabbit ventricle and atrium (Carl et al., 1995) though specific colocalization values between these two proteins, and detailed analysis of the subcellular distribution of junctin, have yet to be reported.      I.4.3 L-type calcium channel One of two types of calcium channels present in the heart, the L-type channel is the most dominant, and is present in all cardiomyocytes (Bers, 2001).  A T-type Ca2+ channel is also present, but to a variable degree amongst the different types of cardiac cells.  For example, purkinje fibers contain a large T-type Ca2+ current (ICa,T) whereas it is either very small or completely absent in adult ventricular myocytes (Bean, 1989).  In rabbit ventricular myocytes, there are approximately 300,000 LTCCs per cell with only ~ 3% of them open at the peak of the Ca2+ current during an AP (Bers, 2001).  The L-type   8 Ca2+ channel is a sarcolemmal membrane protein composed of five subunits (α1, α2, β, γ, and δ), of which α1 is the pore-forming subunit (Bodi et al., 2005).  Immunolocalization studies using primary antibodies directed against both the α1 (Scriven et al., 2002) and α2 (Carl et al., 1995) subunits of the LTCC have shown that, in adult ventricular myocytes, this channel localizes to the periphery and along t-tubules in close proximity to RyR.  In fact, in rat ventricle, approximately 56% of LTCCs colocalize with RyR, while 36% of RyR colocalizes with LTCCs (Scriven et al., 2000).  In both atrial (Carl et al., 1995) and neonatal (Sedarat et al., 2000) cardiomyocytes however, LTCCs are restricted to the periphery of the cell, presumably colocalized with RyR in the dyads.      I.4.4 NCX During the relaxation phase of a cardiac cycle, the sarcolemmal membrane sodium-calcium exchanger (NCX) plays an integral role in removing Ca2+ ions from the myoplasm (Bers, 2001).  The stoichiometry of this reversible counter-transport system is largely accepted to be at or near to 1 Ca2+ ion out: 3 Na+ ions in in ‘forward’ mode (Bers and Ginsburg, 2007).  During the course of an AP, NCX operates in both forward and reverse modes, as its direction of flux depends on the membrane potential (Em) as well as the intra- and extracellular concentrations of both Na+ and Ca2+ ions (Bers, 2001). Because of its ability to bring Ca2+ ions into the myoplasm, CICR via reverse-mode NCX current has long been considered a plausible mechanism.  In 1988 Bers et al. (Bers et al., 1988) showed that a sufficiently elevated intracellular Na+ concentration ([Na+]i) could indeed cause Ca2+ entry via NCX and induce contraction either directly or indirectly (via SR Ca2+ release through RyR).  Voltage-gated Na+ channels in particular could raise   9 [Na] i high enough to cause reverse-mode conductance of NCX, especially if very closely associated to NCX molecules. Later studies both supported (e.g. (Levesque et al., 1994)) and disputed (e.g. (Sipido et al., 1995)) the notion of NCX-induced CICR.  In 1995, Lopez-Lopez et al. found that Ca2+ entry via NCX produced a gradual uniform rise in intracellular calcium levels, rather than inducing Ca2+ sparks which are the individual events that cause whole cell transients during contraction (see section I.3).  Further studies indicated that the efficiency of reverse-mode NCX in triggering SR Ca2+ release was ~4 times less than that of current through LTCCs (ICa,L)  and that if both triggers are present, as in an intact myocyte, that CICR is almost entirely controlled by ICa,L (Sipido et al., 1997).  It now seems clear that although NCX is not the main physiological trigger in CICR, it likely plays an important role in modulation of the EC coupling process, in particular by potentially elevating local Ca2+ concentration in the dyad and sensitizing RyR prior to LTCC activation (Bers, 2008).  Furthermore, NCX may also act as a ‘back-up’ mechanism to L-type Ca2+ channels if they fail to open at any given dyad, eventually raising local Ca2+ concentrations high enough to trigger SR Ca2+ release through RyR (Bers, 2008).  Structural data characterizing the distribution of NCX in rat ventricular myocytes also supports the idea that NCX largely plays a supporting role.  Although NCX is distributed along the surface and in the t-tubular membrane much like RyR and Cav1.2, detailed colocalization analysis between these 3 proteins and voltage-gated Na+ channels showed that neither NCX nor Na+ channels are located in the dyads and they are also not colocalized with one another (Scriven et al., 2000).  Therefore, in these particular cells, it again seems unlikely that NCX is triggering CICR, but it may be modulating the   10 local [Ca2+] depending on the extent of the ‘fuzzy space’ (the area of restricted diffusion underneath the sarcolemmal membrane) (Scriven et al., 2000).  It is important to note at this point that the relative contribution of NCX to EC coupling is likely significantly different between various species and cell types, and also at different stages of development.  For example, both the current density and relative activity level of NCX varied in a study comparing isolated rat, guinea pig, hamster ventricles, and human atria. In the four species studied, the density of inward NCX current was greatest in the hamster and least in the rat (Sham et al., 1995).  Also, during development of rabbit ventricular myocytes, the distribution of NCX changes dramatically.  Immunofluorescence studies show that the majority of NCX re-distributes itself from primarily surface to both surface and interior (in T-tubular membranes) as the animal ages (Dan et al., 2007).  The colocalization of RyR with NCX also changes; the highest level of colocalization at the surface appears in myocytes 3 days old, dropping gradually with age.  Conversely, interior colocalization increases with age, peaking once the animal is matured (Dan et al., 2007).  What this suggests, and what later studies confirm (Huang et al., 2008), is that NCX-mediated CICR is likely the predominant mode of CICR in neonates and that, during development, the more efficient LTCC-triggered CICR gradually takes over. Although atrial myocytes are similar to neonatal in many ways, it is unclear whether or not NCX provides the trigger Ca2+ for CICR in these cells.      I.4.5 SERCA The Sarco/Endoplasmic Reticulum Ca2+-ATPase (SERCA) is the other main mechanism (along with NCX) by which the cell removes Ca2+ ions from the cytoplasm during the relaxation phase of the cardiac cycle (Bers, 2001).  A family of 3 genes   11 encodes the SERCA pump, and over 10 different isoforms that exhibit tissue and developmental specificity have been detected, including the cardiac isoform SERCA2a (Periasamy and Kalyanasundaram, 2007).  For every one molecule of ATP consumed, SERCA transports 2 Ca2+ ions against their concentration gradient back into the SR lumen (Bers, 2001).  ImmunoEM studies have shown that SERCA and its main regulatory protein phospholamban (PLB) are uniformly distributed in network SR of rat and canine cardiomyocytes (Jorgensen and Jones, 1987; Jorgensen et al., 1982).  PLB is an endogenous inhibitor of SERCA, decreasing Ca2+ transport particularly at low [Ca]i (Bers, 2001).  However, during β-adrenergic stimulation for example, PLB is reversibly phosphorylated and the activity of SERCA increases (Lindemann et al., 1983). Interestingly, when comparing atria and ventricles of rat, mouse and guinea pig, SERCA and PLB protein levels were higher and lower, respectively, in atria.  This was put forth as a potential explanation for shortened contraction times in atrial tissues of these animals (Luss et al., 1999).      I.4.6 Caveolae Caveolae, flask-shaped invaginations of the plasma membrane, are involved in a variety of cellular functions such as vesicular transport, cholesterol homeostasis, and signal transduction (Cohen et al., 2004).  Caveolin (cav) proteins 1,2, and 3 (cav-3 being the muscle-specific and most predominant isoform in cardiomyocytes) are the structural components of caveolae and can also function as scaffolding proteins, recruiting and regulating the activity of signaling molecules within the cell (Cohen et al., 2004). Caveolae have been postulated to play an important role in EC coupling in a variety of mammalian cell types, including the adult (Calaghan and White, 2006) and neonatal   12 (Lohn et al., 2000) rat, cow (Bossuyt et al., 2002), and mouse (Balijepalli et al., 2006).  In the adult rat ventricular myocyte, Calaghan and White (2006) reported that disrupting the caveolae using methyl-β-cyclodextrin (MβC) reduced both the amplitude of the [Ca2+]i transient and contraction of the cell.  They hypothesized that this was due to a reduction in coupling efficiency between Ca2+ entry and release.  Similarly, the use of MβC in rat arterial smooth muscle and neonatal cardiomyocytes, which both lack a defined t-tubule network, caused a reduction in frequency, width, and amplitude of Ca2+ sparks without modulation of ICa,L (Lohn et al., 2000).  This data again suggests a role for caveolae in coupling trigger Ca2+ with release of Ca2+ from the SR, and implies the localization of RyR and/or LTCCs within or near caveolae. A great deal of effort has been put forth recently into determining the constituents and neighbours of caveolae in cardiac myocytes.  Immunoprecipitation (IP) studies suggest that NCX1, the cardiac specific isoform of NCX, associates specifically with cav- 3 in bovine sarcolemma (Bossuyt et al., 2002).  The major proteins involved in β- Adrenergic receptor (βAR) activation of protein-kinase A (PKA), which phosphorylates key proteins including LTCCs and RyR and alters EC coupling, as well as Cav1.2, have also been localized to caveolar-enriched membrane fractions (Balijepalli et al., 2006). Interestingly, in the same study, immunogold labeling showed that not all LTCCs localized to caveolae; some appeared to localize to areas of the sarcolemma that did not contain caveolae.  These authors also demonstrated that β2-AR (and not β1) regulation of LTCCs required the presence of cav-3 in mouse ventricular cells (Balijepalli et al., 2006). Using immunofluorescence and sophisticated colocalization analysis, Scriven et al. also attempted to determine some of the occupants and neighbours of caveolae (Scriven et al.,   13 2005).  Using a primary antibody against caveolin-3 in isolated rat ventricular myocytes, these authors observed strong labeling along the surface membrane and in transverse stripes at approximately the level of the Z-lines, presumably in t-tubular membranes.  At the surface of the myocyte, cav-3 colocalized with RyR, NCX, Na+ channel, and Cav1.2, apparently in distinct subpopulations.  In the interior of the myocyte, only a sub- population of RyR that have no adjacent LTCC and a small portion of LTCCs are within or near caveolae (Scriven et al., 2005).  Overall, there is overwhelming evidence that caveolae are intimately involved in regulation of Ca2+ signaling during ECC in cardiomyocytes. I.5 EC coupling in atrial cardiomyocytes      I.5.I Ca2+ signaling patterns During an action potential, electrical depolarization passes along the sarcolemmal membrane, opening voltage-gated Ca2+ channels.  In a ventricular myocyte, t-tubules conduct the electrical current quickly throughout the cell, resulting in a nearly synchronous global Ca2+ transient (Cheng et al., 1994) that initiates as transverse striations composed of multiple focal Ca2+ release sites (Ca2+ sparks) near the level of the t-tubule (Cheng et al., 1994).  Without a well-developed t-tubule network, atrial myocytes exhibit a unique spatiotemporal pattern of Ca2+ release that initiates only at the periphery and variably propagates into the central bulk of the myocyte (Kockskamper et al., 2001; Mackenzie et al., 2001).      I.5.2 Junctional vs. non-junctional RyR As previously mentioned, RyR distribution in atria is similar to that in ventricle; distributed along the sarcolemmal membrane and with regular transverse striations at   14 approximately the level of the Z-line.  During an AP, a small portion of the junctional RyR are activated through CICR by the stochastic opening of one or more LTCCs directly apposing them (Kockskamper et al., 2001).  Recording cytoplasmic Ca2+ increases with a Ca2+ indicator dye, early signals (sparks) are seen at discrete locations in the subsarcolemmal space following depolarization (Mackenzie et al., 2001).  Eventually, at the peak of the signal, a subsarcolemmal ring of elevated Ca2+ is observed, which propagates only weakly into the central bulk of the myocyte where recorded Ca2+ signals are smaller in amplitude and rate of rise (Mackenzie et al., 2004).  This results in a spatially inhomogeneous whole-cell Ca2+ transient.  It has been verified however, that non-junctional RyR are indeed functional.  During caffeine application, these release sites respond with a Ca2+ signal similar to those from junctional sites, producing a ‘stereotypical’ homogeneous whole cell transient (Mackenzie et al., 2001).      I.5.3 Modulation of atrial Ca2+ transients It has been shown that a variety of other conditions will lead to recruitment of central RyR and a spatially homogenous Ca2+ signal.  Acutely raising the extracellular [Ca2+] for example results in a more substantial response from central RyR.  By increasing extracellular Ca2+ from 1mM to 10mM and exposing cells to 10 subsequent electrical stimulations, Mackenzie et al. (2004) showed that the Ca2+ transient in the center of the cell increased by ~500%!  This consequently resulted in a sevenfold increase in the magnitude of contraction of the cell.  They determined that the short exposure to increased extracellular Ca2+ did not significantly increase SR load and therefore, the increase in central RyR response was due to greater Ca2+ influx via LTCCs (Mackenzie et al., 2004).  These authors produced the same response from central RyR   15 by: i) reversibly sensitizing RyR with submillimolar levels of caffeine (not sufficient to activate Ca2+ release alone) prior to stimulation, ii) loading the SR Ca2+ stores (thereby sensitizing RyR), and iii) incubating cells with β-adrenergic agonist isoproterenol (ISO) or cardioactive hormone endothelin-1 (ET-1) prior to stimulation (Mackenzie et al., 2004).  It is clear therefore, that activation of both populations of RyR and a spatially homogenous Ca2+ transient does occur under certain circumstances, which corresponds with the highly adaptable role of the atria.  Furthermore, the observed patterns of Ca2+ release in atrial myocytes provide additional experimental evidence for the local control theory.  In this case, the increase in Ca2+ signal in a small subsarcolemmal ring following depolarization is the result of the summation of electrically evoked Ca2+ sparks from junctional RyR only.  Central, or non-junctional, RyR are not in close proximity to Cav1.2 channels, nor are they influenced by Ca2+ signals from junctional RyR. Therefore, the small rise in Ca2+ signal towards the central bulk of the myocyte is a result of simple diffusion.  Independent activation of each RyR cluster does seem to depend on the physiological state of the myocyte however, as under certain conditions (e.g. β- adrenergic stimulation), Ca2+ release from junctional RyR can influence non-junctional RyR and propagate the Ca2+ signal between these neighbouring release sites, causing an overall homogenous Ca2+ transient in response to depolarization. Still, under basal conditions, local control theory does appear to hold true for atrial myocytes.  However, one of the inherent assumptions of this theory is that each RyR Ca2+ release unit is functionally equivalent.  Not only does there appear to be some substantial evidence indicating otherwise, but there may also be other mechanisms at play in the cell preventing propagation of the Ca2+ signal inwards during contraction.   16      I.5.4 Ca2+ sparks in atrial myocytes The different functional characteristics of the two sub-populations of atrial RyR have largely been observed through Ca2+ spark properties, both spontaneous and evoked, in each region.  Although both subsets of RyR exhibit spontaneous Ca2+ sparks, they occur at different rates, and have different amplitudes and time courses than one another. In rat atrial myocytes, approximately 2.5 sparks per second were observed in both peripheral and central regions of the cell (Woo et al., 2003).  When normalized for myocyte area however, spontaneous sparks were ~5 times more frequent in peripheral regions. Ca2+ transients from respective regions of the cell were not different when exposed to caffeine, and therefore, differential SR Ca2+ load was eliminated as a potential reason for the observed variation (Woo et al., 2003).  Interestingly, when examining spontaneous Ca2+ sparks, blockade of ICa,L did not alter the frequency of either peripheral or central local Ca2+ releases, indicating that variation in frequency does not occur as a result of association with LTCCs (Woo et al., 2003).  The variation in size of electrically- evoked Ca2+ sparks at the periphery and center of the cell is still disputable, as some groups have shown no difference in the amplitudes of the two (Woo et al., 2002), and some have shown higher amplitudes of peripheral sparks vs. central (Mackenzie et al., 2004) (Kockskamper et al., 2001).  There is a consensus however that sparks in the periphery have both a higher rate of rise and a faster dissipation rate than those in the center of the myocyte (Kockskamper et al., 2001; Mackenzie et al., 2004; Woo et al., 2002).  These local Ca2+ releases also appear larger than electrically evoked Ca2+ sparks in ventricular myocytes (Woo et al., 2003). Not only does there appear to be a difference in the characteristics of peripheral and central Ca2+ release sites, but there is also significant data showing differences   17 between peripheral sites themselves.  Upon depolarization of rat atrial myocytes, peripheral Ca2+ release sites have been observed to differ in their amplitude and timing of release.  Some regions, termed ‘eager sites’, are the fastest to respond to electrical stimulation and produce large Ca2+ transients as measured by an intracellular Ca2+ dye. Immediately adjacent to such sites are areas that produce slowly developing Ca2+ signals with ~25% of the amplitude and are termed ‘failure sites’ (Mackenzie et al., 2001). Furthermore, in a plot of 23 eager sites, these authors showed that activation occurs in a reproducible sequence, and that sites activated earlier had a faster time course to the peak of their response.  The same relationship held for frequency of spontaneous sparks at each site.  That is, the further down the eager site was in the activation order, the longer it took to reach its maximal amplitude following depolarization, and the less frequently it produced spontaneous Ca2+ sparks (Mackenzie et al., 2001).  The activation order of these eager sites was consistent even upon loading of the SR or increased ICa,L.  What this suggests, is that depolarization-induced Ca2+ sparks in rat atrial cells are not simply activated by the stochastic opening of closely apposed LTCCs, but that LTCCs, and consequently junctional RyR, are activated in a pre-determined manner.  It should be noted that there may be species variation in this matter.  In feline atrial myocytes for example, ‘eager’ and ‘failure’ Ca2+ release sites do appear at the periphery upon depolarization of the cell, but the activation order of these sites appears to be random, unlike the pre-determined order in rat atria (Kockskamper et al., 2001).  Feline atrial cells also appear to have a robust central response, and therefore a relatively homogeneous Ca2+ response under basal conditions, unlike the rat (Kockskamper et al., 2001).  Overall however, a wealth of evidence supports the fact that there are intrinsic functional   18 differences between peripheral and central RyR in atria.   The variations in Ca2+ transients produced from different peripheral release sites suggest either non-uniformity in microarchitecture of the dyads or a disparity in the functional characteristics of junctional RyR.      I.5.5 ‘Blocking’ of Ca2+ transients Other than the independent control of each RyR Ca2+ release unit, another proposed hypothesis for the lack of inward propagation of Ca2+ transient under basal conditions in atrial myocytes is the presence of a molecular ‘blockade’ between junctional and non-junctional RyR. In order to examine this hypothesis, Mackenzie et al. (2004) applied a small amount of cyclopiazonic acid (CPA), which partially inhibited SERCA pump activity while retaining levels of intracellular and SR Ca2+ consistent with control conditions. Before CPA application, depolarization of atrial myocytes produced a typical heterogeneous Ca2+ transient while post-application, electrical stimulation resulted in a nearly homogeneous response with an increase in Ca2+ transients at both peripheral and central release sites (Mackenzie et al., 2004).  These results suggest that under basal conditions, SERCA activity may be involved in the inhibition of Ca2+ transient propagation into the center of the myocyte.  The presence of SERCA pumps throughout the entirety of the cell (Jorgensen et al., 1982; Mackenzie et al., 2004) present the problem however, that if they are preventing propagation of Ca2+ signals in to the center of the cell, why are they not preventing CICR in the subsarcolemmal region? In a variety of cell types it has been shown that mitochondria can alter the spatiotemporal pattern of intracellular Ca2+ signals and regulate local [Ca2+] within a   19 microdomain.  A specific example of this is in rat cortical astrocytes where mitochondria appear to be positioned in close proximity to endoplasmic reticulum (ER) Ca2+ release sites, lowering local [Ca2+] and preventing CICR at these sites (Duchen, 2000).  In cardiomyocytes, upon relaxation of the cell, mitochondria take up a very small amount of intracellular Ca2+ via their Ca2+ uniporter (Bers, 2001).  In atrial cells, where a large volume of the cell is occupied by these organelles, it is reasonable to consider that they may play a more significant role in shaping intracellular Ca2+ transients.  To test this theory, researchers loaded rat atrial myocytes with oligomycin and antimycin, which depolarize the mitochondrial membrane and eliminate the energy source mitochondria need to uptake Ca2+ (Mackenzie et al., 2004).  In a protocol that prevented mitochondria from taking up Ca2+ but did not decrease cellular ATP concentration, these authors showed that the amplitude of central Ca2+ release during contraction increased to ~95% of peripheral values.  Further investigation revealed that although all mitochondria in the cells were capable of taking up Ca2+, under basal conditions those at the periphery were taking in ~10 X the amount of central organelles.  It was concluded that, in these particular cells, mitochondria at the periphery of the myocyte are buffering Ca2+ signals released from peripheral sites and preventing the inward propagation of transients (Mackenzie et al., 2004).  This conclusion makes assumptions however that: i) mitochondria are positioned close enough to Ca2+ release sites to efficiently sequester enough Ca2+ to prevent further CICR in to the center of the myocyte, ii) if positioned correctly, mitochondria are able to take up sufficient levels of Ca2+ on a beat-to-beat basis to affect local [Ca2+], and iii) that by playing a major role in regulation of local [Ca2+], the   20 energy expended would not compromise the mitochondria’s main role of ATP production, all of which have been contended (Huser et al., 2000). Overall, the specific mechanisms by which atrial myocytes regulate intracellular Ca2+ transients, and therefore their contraction level, remain unclear.  Whether or not RyR in these cells are under local control, it is undoubtedly clear that elucidating the molecular architecture of the cell is of utmost importance.  To date however, the localization of proteins integral to EC coupling and their spatial relationships between one another has been greatly understudied. I.6 Hypothesis and specific aims      I.6.1 Hypothesis Given the nature of excitation-contraction coupling in atria – in particular, the inhomogeneous calcium transients, as well as the differences in calcium spark characteristics between the periphery and interior and between peripheral sites themselves – we hypothesized that atrial RyR complexes are heterogeneous, and that particular groups may be targeted to specific intracellular domains.      I.6.2 Specific aims In order to examine this hypothesis, our specific aims were to: 1) determine the molecular architecture of atrial dyads and 2) determine the distribution of molecules that may regulate RyR open probability and/or calcium levels within the cell.       21  I.7 References  Alpert, J. S., Petersen, P. and Godtfredsen, J. (1988). Atrial fibrillation: natural history, complications, and management. Annu Rev Med 39, 41-52. Balijepalli, R. C., Foell, J. D., Hall, D. D., Hell, J. W. and Kamp, T. J. (2006). Localization of cardiac L-type Ca(2+) channels to a caveolar macromolecular signaling complex is required for beta(2)-adrenergic regulation. Proc Natl Acad Sci U S A 103, 7500-5. Bassani, J. W., Yuan, W. and Bers, D. M. (1995). Fractional SR Ca release is regulated by trigger Ca and SR Ca content in cardiac myocytes. Am J Physiol 268, C1313-9.  Bean, B. P. (1989). Classes of calcium channels in vertebrate cells. Annu Rev Physiol 51, 367-84.  Bers, D. M. (2001). Excitation-Contraction Coupling and Cardiac Contractile Force: Kluwer Academic Publishers.  Bers, D. M. (2002). Cardiac excitation-contraction coupling. Nature 415, 198- 205.  Bers, D. M. (2004). Macromolecular complexes regulating cardiac ryanodine receptor function. J Mol Cell Cardiol 37, 417-29.  Bers, D. M. (2008). Calcium Cycling and Signaling in Cardiac Myocytes. Annu. Rev. Physiol. 70, 23-49.  Bers, D. M., Christensen, D. M. and Nguyen, T. X. (1988). Can Ca entry via Na-Ca exchange directly activate cardiac muscle contraction? J Mol Cell Cardiol 20, 405-14.  Bers, D. M. and Ginsburg, K. S. (2007). Na:Ca stoichiometry and cytosolic Ca- dependent activation of NCX in intact cardiomyocytes. Ann N Y Acad Sci 1099, 326-38.  Bodi, I., Mikala, G., Koch, S. E., Akhter, S. A. and Schwartz, A. (2005). The L-type calcium channel in the heart: the beat goes on. J Clin Invest 115, 3306-17.  Bootman, M. D., Higazi, D. R., Coombes, S. and Roderick, H. L. (2006). Calcium signalling during excitation-contraction coupling in mammalian atrial myocytes. J Cell Sci 119, 3915-25.  Bossuyt, J., Taylor, B. E., James-Kracke, M. and Hale, C. C. (2002). The cardiac sodium-calcium exchanger associates with caveolin-3. Ann N Y Acad Sci 976, 197-204.  Calaghan, S. and White, E. (2006). Caveolae modulate excitation-contraction coupling and beta2-adrenergic signalling in adult rat ventricular myocytes. Cardiovasc Res 69, 816-24.  Cannell, M. B., Cheng, H. and Lederer, W. J. (1995). The control of calcium release in heart muscle. Science 268, 1045-9.  Carl, S. L., Felix, K., Caswell, A. H., Brandt, N. R., Ball, W. J., Jr., Vaghy, P. L., Meissner, G. and Ferguson, D. G. (1995). Immunolocalization of sarcolemmal dihydropyridine receptor and sarcoplasmic reticular triadin and ryanodine receptor in rabbit ventricle and atrium. J Cell Biol 129, 672-82.   22  Cheng, H., Cannell, M. B. and Lederer, W. J. (1994). Propagation of excitation-contraction coupling into ventricular myocytes. Pflugers Arch 428, 415-7.  Cheng, H., Lederer, W. J. and Cannell, M. B. (1993). Calcium sparks: elementary events underlying excitation-contraction coupling in heart muscle. Science 262, 740-4.  Cohen, A. W., Hnasko, R., Schubert, W. and Lisanti, M. P. (2004). Role of caveolae and caveolins in health and disease. Physiol Rev 84, 1341-79.  Dan, P., Lin, E., Huang, J., Biln, P. and Tibbits, G. F. (2007). Three- dimensional distribution of cardiac Na+-Ca2+ exchanger and ryanodine receptor during development. Biophys J 93, 2504-18.  Duchen, M. R. (2000). Mitochondria and calcium: from cell signalling to cell death. J Physiol 529 Pt 1, 57-68.  Fabiato, A. (1983). Calcium-induced release of calcium from the cardiac sarcoplasmic reticulum. Am J Physiol 245, C1-14.  Forbes, M. S., Hawkey, L. A. and Sperelakis, N. (1984). The transverse-axial tubular system (TATS) of mouse myocardium: its morphology in the developing and adult animal. Am J Anat 170, 143-62.  Forbes, M. S., Van Niel, E. E. and Purdy-Ramos, S. I. (1990). The atrial myocardial cells of mouse heart: a structural and stereological study. J Struct Biol 103, 266-79.  Forssmann, W. G. and Girardier, L. (1970). A study of the T system in rat heart. J Cell Biol 44, 1-19.  Franzini-Armstrong, C., Kenney, L. J. and Varriano-Marston, E. (1987). The structure of calsequestrin in triads of vertebrate skeletal muscle: a deep-etch study. J Cell Biol 105, 49-56.  Franzini-Armstrong, C., Protasi, F. and Tijskens, P. (2005). The assembly of calcium release units in cardiac muscle. Ann N Y Acad Sci 1047, 76-85.  Gergs, U., Berndt, T., Buskase, J., Jones, L. R., Kirchhefer, U., Muller, F. U., Schluter, K. D., Schmitz, W. and Neumann, J. (2007). On the role of junctin in cardiac Ca2+ handling, contractility, and heart failure. Am J Physiol Heart Circ Physiol 293, H728-34.  Go, A. S., Hylek, E. M., Phillips, K. A., Chang, Y., Henault, L. E., Selby, J. V. and Singer, D. E. (2001). Prevalence of diagnosed atrial fibrillation in adults: national implications for rhythm management and stroke prevention: the AnTicoagulation and Risk Factors in Atrial Fibrillation (ATRIA) Study. Jama 285, 2370-5.  Godtfredsen, J. (1999). Physiology and pathophysiology of the atria: its role in atrial fibrillation. J Thromb Thrombolysis 7, 13-9.  Gyorke, I., Hester, N., Jones, L. R. and Gyorke, S. (2004). The role of calsequestrin, triadin, and junctin in conferring cardiac ryanodine receptor responsiveness to luminal calcium. Biophys J 86, 2121-8.  Hatem, S. N., Benardeau, A., Rucker-Martin, C., Marty, I., de Chamisso, P., Villaz, M. and Mercadier, J. J. (1997). Different compartments of sarcoplasmic reticulum participate in the excitation-contraction coupling process in human atrial myocytes. Circ Res 80, 345-53.  Hibbs, R. G. and Ferrans, V. J. (1969). An ultrastructural and histochemical study of rat atrial myocardium. Am J Anat 124, 251-70.   23  Hilgemann, D. W. and Noble, D. (1987). Excitation-contraction coupling and extracellular calcium transients in rabbit atrium: reconstruction of basic cellular mechanisms. Proc R Soc Lond B Biol Sci 230, 163-205.  Huang, J., Hove-Madsen, L. and Tibbits, G. F. (2008). Ontogeny of Ca2+- induced Ca2+ release in rabbit ventricular myocytes. Am J Physiol Cell Physiol 294, C516-25.  Huser, J., Blatter, L. A. and Sheu, S. S. (2000). Mitochondrial calcium in heart cells: beat-to-beat oscillations or slow integration of cytosolic transients? J Bioenerg Biomembr 32, 27-33.  Jorgensen, A. O. and Jones, L. R. (1987). Immunoelectron microscopical localization of phospholamban in adult canine ventricular muscle. J Cell Biol 104, 1343- 52.  Jorgensen, A. O., Shen, A. C., Arnold, W., McPherson, P. S. and Campbell, K. P. (1993). The Ca2+-release channel/ryanodine receptor is localized in junctional and corbular sarcoplasmic reticulum in cardiac muscle. J Cell Biol 120, 969-80.  Jorgensen, A. O., Shen, A. C. and Campbell, K. P. (1985). Ultrastructural localization of calsequestrin in adult rat atrial and ventricular muscle cells. J Cell Biol 101, 257-68.  Jorgensen, A. O., Shen, A. C., Daly, P. and MacLennan, D. H. (1982). Localization of Ca2+ + Mg2+-ATPase of the sarcoplasmic reticulum in adult rat papillary muscle. J Cell Biol 93, 883-92.  Kockskamper, J., Sheehan, K. A., Bare, D. J., Lipsius, S. L., Mignery, G. A. and Blatter, L. A. (2001). Activation and propagation of Ca(2+) release during excitation-contraction coupling in atrial myocytes. Biophys J 81, 2590-605.  Leeson, T. S. (1980). T-Tubules, couplings and myofibrillar arrangements in rat atrial myocardium. Acta anat. 108, 374-388.  Levesque, P. C., Leblanc, N. and Hume, J. R. (1994). Release of calcium from guinea pig cardiac sarcoplasmic reticulum induced by sodium-calcium exchange. Cardiovasc Res 28, 370-8.  Lindemann, J. P., Jones, L. R., Hathaway, D. R., Henry, B. G. and Watanabe, A. M. (1983). beta-Adrenergic stimulation of phospholamban phosphorylation and Ca2+-ATPase activity in guinea pig ventricles. J Biol Chem 258, 464-71.  Lohn, M., Furstenau, M., Sagach, V., Elger, M., Schulze, W., Luft, F. C., Haller, H. and Gollasch, M. (2000). Ignition of calcium sparks in arterial and cardiac muscle through caveolae. Circ Res 87, 1034-9.  Lopez-Lopez, J. R., Shacklock, P. S., Balke, C. W. and Wier, W. G. (1995). Local calcium transients triggered by single L-type calcium channel currents in cardiac cells. Science 268, 1042-5.  Luss, I., Boknik, P., Jones, L. R., Kirchhefer, U., Knapp, J., Linck, B., Luss, H., Meissner, A., Muller, F. U., Schmitz, W. et al. (1999). Expression of cardiac calcium regulatory proteins in atrium v ventricle in different species. J Mol Cell Cardiol 31, 1299-314.  Mackenzie, L., Bootman, M. D., Berridge, M. J. and Lipp, P. (2001). Predetermined recruitment of calcium release sites underlies excitation-contraction coupling in rat atrial myocytes. J Physiol 530, 417-29.   24  Mackenzie, L., Roderick, H. L., Berridge, M. J., Conway, S. J. and Bootman, M. D. (2004). The spatial pattern of atrial cardiomyocyte calcium signalling modulates contraction. J Cell Sci 117, 6327-37.  Mitchell, R. D., Simmerman, H. K. and Jones, L. R. (1988). Ca2+ binding effects on protein conformation and protein interactions of canine cardiac calsequestrin. J Biol Chem 263, 1376-81.  Nelson, D. A. and Benson, E. S. (1963). On the structural continuities of the transverse tubular system of rabbit and human myocardial cells. J Cell Biol 16, 297-313.  Periasamy, M. and Kalyanasundaram, A. (2007). SERCA pump isoforms: their role in calcium transport and disease. Muscle Nerve 35, 430-42.  Rousseau, E. and Meissner, G. (1989). Single cardiac sarcoplasmic reticulum Ca2+-release channel: activation by caffeine. Am J Physiol 256, H328-33.  Scriven, D. R., Klimek, A., Lee, K. L. and Moore, E. D. (2002). The molecular architecture of calcium microdomains in rat cardiomyocytes. Ann N Y Acad Sci 976, 488- 99.  Scriven, D. R. L., Dan, P. and Moore, E. D. W. (2000). Distribution of Proteins Implicated in Excitation-Contraction Coupling in Rat Ventricular Myocytes. Biophysical Journal 79, 2682-2691.  Scriven, D. R. L., Klimek, A., Asghari, P., Bellve, K. and Moore, E. D. W. (2005). Caveolin-3 is Adjacent to a group of Extradyadic Ryanodine Receptors. Biophysical Journal 89, 1893-1901.  Sedarat, F., Xu, L., Moore, E. D. and Tibbits, G. F. (2000). Colocalization of dihydropyridine and ryanodine receptors in neonate rabbit heart using confocal microscopy. Am J Physiol Heart Circ Physiol 279, H202-9.  Sham, J. S., Hatem, S. N. and Morad, M. (1995). Species differences in the activity of the Na(+)-Ca2+ exchanger in mammalian cardiac myocytes. J Physiol 488 ( Pt 3), 623-31.  Sipido, K. R., Carmeliet, E. and Pappano, A. (1995). Na+ current and Ca2+ release from the sarcoplasmic reticulum during action potentials in guinea-pig ventricular myocytes. J Physiol 489 ( Pt 1), 1-17.  Sipido, K. R., Maes, M. and Van de Werf, F. (1997). Low efficiency of Ca2+ entry through the Na(+)-Ca2+ exchanger as trigger for Ca2+ release from the sarcoplasmic reticulum. A comparison between L-type Ca2+ current and reverse-mode Na(+)-Ca2+ exchange. Circ Res 81, 1034-44.  Stern, M. D. (1992). Theory of excitation-contraction coupling in cardiac muscle. Biophys J 63, 497-517.  Sun, X. H., Protasi, F., Takahashi, M., Takeshima, H., Ferguson, D. G. and Franzini-Armstrong, C. (1995). Molecular architecture of membranes involved in excitation-contraction coupling of cardiac muscle. J Cell Biol 129, 659-71.  Terentyev, D., Viatchenko-Karpinski, S., Vedamoorthyrao, S., Oduru, S., Gyorke, I., Williams, S. C. and Gyorke, S. (2007). Protein protein interactions between triadin and calsequestrin are involved in modulation of sarcoplasmic reticulum calcium release in cardiac myocytes. J Physiol 583, 71-80.  Waktare, J. and Camm, J. A. (1999). In Atrial Fibrillation, pp. 4-6: Informa Health Care.   25  Wong, A. Y. (1981). A model of excitation-contraction coupling of mammalian cardiac muscle. J Theor Biol 90, 37-61.  Woo, S. H., Cleemann, L. and Morad, M. (2002). Ca2+ current-gated focal and local Ca2+ release in rat atrial myocytes: evidence from rapid 2-D confocal imaging. J Physiol 543, 439-53.  Woo, S. H., Cleemann, L. and Morad, M. (2003). Spatiotemporal characteristics of junctional and nonjunctional focal Ca2+ release in rat atrial myocytes. Circ Res 92, e1-11.  Xu, L., Tripathy, A., Pasek, D. A. and Meissner, G. (1998). Potential for pharmacology of ryanodine receptor/calcium release channels. Ann N Y Acad Sci 853, 130-48.  Yamasaki, Y., Furuya, Y., Araki, K., Matsuura, K., Kobayashi, M. and Ogata, T. (1997). Ultra-high-resolution scanning electron microscopy of the sarcoplasmic reticulum of the rat atrial myocardial cells. Anat Rec 248, 70-5.  Zhang, L., Kelley, J., Schmeisser, G., Kobayashi, Y. M. and Jones, L. R. (1997). Complex formation between junctin, triadin, calsequestrin, and the ryanodine receptor. Proteins of the cardiac junctional sarcoplasmic reticulum membrane. J Biol Chem 272, 23389-97.                              26 2 MULTIPLE SURFACE RYANODINE RECEPTOR POPULATIONS EXIST IN RAT ATRIAL MYOCYTES1  2.I Introduction Excitation-contraction coupling (ECC) in both atrial and ventricular myocytes depends on calcium-induced calcium release (CICR).  During CICR, a small influx of calcium (Ca2+) through voltage-gated calcium channels (Cav1.2) initiates a much larger release of Ca2+ from ryanodine receptors (RyR) located in the adjacent junctional sarcoplasmic reticulum (jSR) (Fabiato, 1983). In ventricular cells, a well-developed transverse and axial tubular system (TATS) spreads depolarization in to the myocyte, resulting in a nearly synchronous SR Ca2+ release throughout the entire cell (Wier et al., 1995).  Most atrial cells however, lack such tubules and show a unique spatiotemporal pattern of calcium release during EC coupling, characterized by a subsarcolemmal ring of elevated calcium concentration which variably propagates into the central bulk of the myocyte in direct proportion to cellular Ca2+ load and Ca2+ influx (Berlin, 1995; Huser et al., 1996; Mackenzie et al., 2001).  Mackenzie, Bootman et al. (2001) also showed the presence of so-called ‘eager sites’ (those sites that respond very quickly to rises in Ca2+ concentration) along the periphery of atrial myocytes separated by one or more ‘failure sites’, which exhibit weak regenerative Ca2+ responses with about one-quarter the magnitude of the ‘eager’ sites.  Carl, Felix et al. (1995) found that Cav1.2 channels were primarily distributed along the periphery of atrial myocytes and highly colocalized with RyR there.  Although this colocalization could potentially explain the increased responsiveness of peripheral sites to electrical depolarization, it does not explain why the  1 A version of this chapter will be submitted for publication.  Schulson, M., Fletcher, P., Scriven, D.R.L., and Moore, E.D.W. Multiple Surface Ryanodine Receptor Populations Exist in Rat Atrial Myocytes.   27 so-called eager and failure sites, both classified as junctional RyR, respond so differently to electrical stimulation. One of the many proteins associated with RyR in a macromolecular complex is calsequestrin (CSQ) (for a review, see Bers, 2004).  Multiple studies have shown CSQ’s role as a luminal Ca2+ sensor (Beard et al., 2005) and a regulatory protein that alters RyR open probability (e.g. (Gyorke et al., 2004; Terentyev et al., 2007)).  This protein is distributed largely along Z-lines and is highly colocalized with RyR throughout the ventricular myocyte (Scriven et al., 2000), but its localization in atrial myocytes, particularly with respect to RyR, has yet to be elucidated. Caveolae, flask-shaped invaginations of the plasma membrane, are involved in a variety of cellular functions such as vesicular transport, cholesterol homeostasis, and signal transduction (Cohen et al., 2004).  They have also been postulated to play an important role in EC coupling; in the adult rat ventricular myocyte, Calaghan and White (2006) reported that disrupting the caveolae using methyl-β-cyclodextrin (MβC) reduced both the amplitude of the [Ca2+]i transient and the contraction.  They hypothesized that this was due to a reduction in coupling efficiency between Ca2+ entry and release, which supports colocalization studies that place RyR and Cav1.2 channels in close proximity to caveolin-3 on the surface of isolated rat ventricular myocytes (Scriven et al., 2005).  The use of MβC in rat arterial smooth muscle and neonatal cardiomyocytes, which lack a defined TATS much like atrial myocytes, caused a reduction in frequency, width, and amplitude of Ca2+ sparks without modulation of the current through Cav1.2 channels (ICa,L) (Lohn et al., 2000).  Their data indicate a role for caveolae in regulating local SR Ca2+ release, and may indicate the presence of Cav1.2 channels within or near caveolae.   28 The cardiac sodium-calcium exchanger (NCX1) has also been shown to associate with caveolin-3 (Bossuyt et al., 2002) and has long been implicated in the regulation of EC coupling.  First viewed as a possible alternate mechanism by which to activate CICR, it is now largely agreed upon that NCX simply acts as a modulator of the process in mature cardiac myocytes (Bers, 2008).  Structural data from Scriven et al. (2005) demonstrates that while couplons (Cav1.2 and RyR) and NCX molecules are both closely associated with caveolin-3 in rat ventricular myocytes, they are not colocalized with one another.  Due to a lack of a well-developed TATS, NCX is largely distributed along the cell periphery in atria (Bootman et al., 2006), but may colocalize with proteins such as RyR and Cav1.2 channels there.  If so, it could act as an important modulator of EC coupling in this region. We have used immunofluorescence and wide-field microscopy, coupled with deconvolution and digital image analysis to examine the distribution and colocalization of RyR, Cav1.2, CSQ, NCX, and Caveolin-3 in adult rat atrial myocytes. 2.2 Results      2.2.I Controls Prior to beginning all triple-labeling experiments, a series of controls needed to be performed in order to ensure the validity of our results.  First, rat atrial myocytes were labeled with each antibody individually: RyR, Cav1.2, CSQ, NCX, or Caveolin-3.  This was done in order to examine not only the distribution of each protein, but also to ensure that the labeling characteristics did not change when cells were incubated with multiple antibodies simultaneously.  Deconvolved images from the surface and interior (left and right panels respectively, for each pair of images) of the labeled cells are shown in Fig. 2.1.  At the surface, antibodies against both RyR and CSQ (Figs.2.1A and B, left panel)   29 display highly punctate labeling in medium sized clusters spread quite evenly across the membrane.  In the interior of the myocyte (Figs.2.1A and B right panel), the labeling pattern of these two proteins is also very similar.  Discrete clusters of both RyR and CSQ line up largely along the Z-lines, but also appear in some longitudinal elements between Z-lines (arrows).  We also confirm the presence of a small subsarcolemmal gap (arrowheads) in RyR labeling, and show that this phenomenon exists in CSQ distribution in atrial cells as well.  From these images it seems that, due to their similar labeling patterns, RyR and CSQ are highly associated in these myocytes.  Detailed analysis of their colocalization is discussed later on. Figs.2.1C, D, and E show the distribution of Cav1.2, NCX, and Cav-3 proteins respectively.  At the surface (Fig.2.1C, left panel), Cav1.2 appears to label in larger clusters than either RyR or CSQ, and is spread fairly evenly across the membrane.  NCX labeling on the surface (Fig.2.1D, left panel) on the other hand, is dense but somewhat uneven, and appears as even larger aggregates than Cav1.2.  The very bright staining at one side of the membrane is an artifact created by imaging the sarcolemma in parallel to the optical axis of the microscope (edge effect).  An antibody directed against the caveolin-3 protein also collects in large clusters on the sarcolemmal membrane (Fig.2.1E, left panel) that appear to be oriented in longitudinal striations.  Cav1.2, NCX, and Cav-3 proteins are only expressed to a very small degree in the interior of atrial myocytes (Figs.2.1 C, D, and E, right panel) and the antibodies against these three proteins label only faintly in this region.  This labeling most likely represents either synthesis/degradation of the proteins or the presence of rudimentary t-tubules in some atrial cells.   30 The ability of the Zenon kits to eliminate cross-reaction between the secondary antibodies of two monoclonal primary antibodies was tested using primary antibodies that are known to localize to distinct regions of the cell: RyR and anti-nuclear pore complex antibody (NPC).  We labeled myocytes simultaneously with these two antibodies using the manufacturer’s protocol for Zenon labeling kits.  In Fig. 2.2, we show the results from this experiment.  RyR and its corresponding secondary antibody (Alexa 594; shown in red) and NPC with Alexa 488 (shown in green) do indeed stay localized to separate regions of the cell – the Z lines (arrow) and nuclear pore region (arrowhead), respectively.  There is negligible red labeling in the nuclear region and/or green labeling at the Z-lines, indicating no cross-reaction between the secondary antibodies.  Colocalization analysis of these two proteins confirms what we see in Fig. 2.2 (0.32% RyR colocalizes with NPC, 6.88% NPC colocalizes with RyR; whole cell values).  The higher percentage of NPC with RyR is a result of a much lower density of NPC as compared to RyR, and also of colocalization along the Z-axis where RyR lies on top of, and underneath, the nuclear membrane.  We also performed this experiment by labeling RyR overnight at 4°C, then adding an Alexa 594 goat anti-mouse secondary antibody for 1.5 hours at room temperature.  This was followed by the NPC antibody conjugated to Alexa 488 goat anti-mouse with a Zenon labeling kit.  The results (not shown) were identical to the image in Fig.2.2 and therefore confirmed that this protocol could also be used to label cells with two monoclonal antibodies without cross-reaction. Furthermore, the results of these two control experiments confirmed that we could label three monoclonal antibodies simultaneously without fear of cross-reaction by incubating   31 the first overnight at 4°C, followed by its secondary antibody for 1.5 hours at room temperature, then the remaining two monoclonal antibodies with a Zenon labeling kit. The final step in confirming the validity of our triple-labeling experiments was to compare 1) the densities of individual proteins that were labeled in more than one experiment, and 2) the colocalization values between pairs of proteins that occurred in more than one experiment to see if they were significantly different from one another. Fig.2.3 displays the density values for RyR (A), CSQ (B), Cav1.2 (C), NCX (D), and Cav-3 (E).  Each triple-labeling experiment that the protein was involved in is included, and grouped according to surface and interior values.  For CSQ, which was only included in one triple-labeling experiment, an independent RyR/CSQ double-labeling experiment was performed for comparison.  Within each group, values were compared to one another using a student’s t-test or ANOVA, and no significant differences were found.  Anytime a combination of two proteins together was found in more than one experiment (e.g. RyR with NCX), the ‘double’ colocalization values (calculated as stated in materials and methods) were compared, and no significant differences were found among these values either (results not shown).  With the completion of these control experiments and calculations, we are confident that the labeling characteristics of each protein involved in our triple-labeling experiments are independent of one another, and that our results are a true indication of the structural organization of atrial myocytes and the association of these proteins within the cells.      2.2.2 Triple-labeling experiments The first of four triple-labeling experiments reported here examines the potential association between RyR, NCX, and Cav1.2 in rat atria.  In Fig.2.4, a representative atrial   32 cell from this experiment is shown.  Each image (A-E) is the same 1 μm thick layer from the surface of the myocyte; Fig.2.4A shows the entire field of view and Figs.2.4B-E are smaller sections of the cell surface.  A significant edge effect appears in Fig.2.4A as an overabundance of NCX (green) labeling.  In order to avoid any errors in our calculations, the cell segment in Fig.2.4B was chosen to measure densities and colocalization from, and these are shown in Figs.2.3 and 2.5 respectively.  When the cell is viewed at the level of individual protein clusters (Figs.2.4 C, D, and E), it appears that the primary association with respect to these 3 proteins is the ryanodine receptor with the voltage- gated Ca2+ channel (pink/purple).  The values in Fig.2.5 and Table 2.1 reflect this, as the highest colocalization percentage in this experiment is between these two proteins (Figs.2.5 A and B; 34.0% RyR with Cav1.2, 49.5% Cav1.2 with RyR).  NCX on the other hand seems to be mainly organized into large clusters that are either: 1) completely separate from the RyR/Cav1.2 unit and do not associate with either protein (e.g. Fig.2.4C), 2) integral within the RyR/Cav1.2 unit (and therefore triple-colocalized with them), but also extend outside of the main colocalized area where they weakly associate with one of the other two proteins (e.g. Fig.2.4D, arrows), or 3) very tightly linked in a relatively small area of triple colocalization but also surround the colocalized area in ‘islands’ of labeling that may or may not associate with either RyR or NCX (Fig.2.4E, arrowheads).  Again, the values in Fig.2.5 reflect this organization.  Approximately 10- 15% of the total number of voxels contain all 3 proteins (RyR, Cav1.2, and NCX), which we see as large clusters of white voxels in Figs.2.4 D and E.  A small number of voxels that contain NCX also contain either RyR or Cav1.2 (Fig.2.5C; 13.6% and 8.5% respectively), which is representative of those clusters of NCX that loosely associate with   33 either RyR or Cav1.2 at their edges (e.g. Fig.2.4D, arrows and Fig.2.4E, arrowheads). With further statistical analysis, we see that all of the triple-colocalization values, as well as the colocalization between RyR and Cav1.2 and RyR and NCX, are significantly greater (p<0.001) than the colocalization values we would expect through random distribution of the molecules, i.e. by chance.  Also, compared to rat ventricular myocytes, the colocalization between RyR and NCX is significantly higher in atria while colocalization between RyR and Cav1.2 is almost the same (Table 2.5). The association between RyR, CSQ, and NCX was examined in a triple-labeling experiment, and the results are shown in Figs. 2.6 and 2.7.  Fig.2.6 shows a representative atrial cell from this experiment at different magnification levels.  Each panel (A – E) is a 1μm thick segment from the cell surface; B is an expanded version (at 1.1X magnification) of the boxed region in A, and C, D, and E are images from the enclosed regions in B (at 5.2X the magnification of B).  Reminiscent of the RyR/NCX/Cav1.2 experiment, the most dominant form of association between these proteins is RyR with CSQ alone (shown in pink/purple).  The colocalization values for this experiment (Fig.2.7 and Table 2.2) confirm this observation; 35.5% of the voxels that contain RyR also include CSQ (Fig.2.7A), and 37.5% of the voxels with CSQ also contain RyR (Fig.2.7B), the highest of any value in this experiment.  These values are comparable to rat ventricular myocytes (Table 2.5).  Only the interior colocalization of CSQ with RyR is significantly lower in the atria (45.6 ± 2.6% as compared to 61.6 ± 7.2% in the ventricle; (Scriven et al., 2000)).  It should be noted however, that the published values for the ventricle are for the whole cell, but the interior and surface numbers are identical (E.Moore, unpublished observation).  There are also a moderate number of triply   34 colocalized voxels (~8%; Fig.2.7 and Table 2.2) in this experiment.  These seem to occur in clusters at the center of a region where all three proteins are very tightly associated (e.g. Fig.2.6C and E, arrowheads).  In other regions of the cell, there are areas of triple colocalization that sit directly within an island of NCX labeling (e.g. Fig.2.6E, arrow). This NCX labeling extends beyond the region of triple colocalization however, where it does not associate with either RyR or CSQ.  This organization is reflected in the large percentage of NCX that is not colocalized with either RyR or CSQ (Fig.2.7C).  In Fig.2.6D in fact, we see a large cluster of NCX coming very near to a RyR/CSQ group, but very little double, let alone triple, colocalization occurs.  Also similar to the RyR/NCX/Cav1.2 experiment, all triple-colocalization values for this experiment, and the colocalization between RyR and CSQ and RyR and NCX, are significantly greater than expected by chance (p<0.001). Figs. 2.8 and 2.9 summarize an experiment that labeled atrial myocytes with antibodies against RyR, Cav1.2, and Cav-3.  In Fig.2.8, a 1μm thick layer from the cell surface is shown in its entirety (A), in a segment that excludes all edges (B), and at the level of individual protein clusters (C).  The highest colocalization values for this experiment are RyR with Cav1.2 (39.2%) and Cav1.2 with RyR (39.4%), as expected (Figs.2.9 A and C, and Table 2.3).  Less than 10% of the voxels that contain Cav-3 on the other hand, also contain either RyR or Cav1.2, or contain all 3 proteins together (Fig.2.9 and Table 2.3).  This may partially result from the high density of Cav-3 labeling at the surface (see Fig.2.3E).  Compared to rat ventricular myocytes (Table 2.5), the percentage of RyR with Cav-3 and Cav1.2 with Cav-3 are significantly lower in the atria (26.8% and 19.7% in the atria vs. 37.6% and 39.0% in the ventricle respectively).  The primary   35 arrangements of RyR, Cav1.2, and Cav-3 that make up the colocalization values are shown in Figs.2.8 C, D, and E.  The high colocalization between RyR and Cav1.2 is reflected in the large number of tightly coupled blue and red clusters (pink/purple; Figs.2.8 C and D, arrows).  Occasionally, a small portion of a Cav-3 cluster is integral to a triply colocalized area (e.g. Fig.2.8D, white voxels), but Cav-3 largely appears to either sit adjacent to RyR/Cav1.2 clusters (e.g. Fig.2.8C, arrowheads) or intersperse amongst individual RyR clusters in a long, continuous band (e.g. Fig. 2.8E).  This pattern of Cav-3 distribution, along with its high labeling density, results in the small percentage of colocalized voxels with the 2 other proteins.  However small these percentages, statistical testing shows that all triple-colocalization for this experiment, as well as RyR and Cav1.2 and RyR and Cav-3 colocalization values, are significantly greater than expected by chance (p<0.001). Finally, in order to examine the association between NCX and Cav-3 relative to RyR, we triple labeled atrial myocytes with antibodies against these 3 proteins.  Figs.2.10 A-E visually represent the results from this experiment.  In Fig.2.10A, the edge effects and artificially high level of colocalization are extreme, and the difference between this image and the segment taken in Fig.2.10B (excluding cell edges) is especially noticeable. In Fig.2.11 and Table 2.4, we see that there is a small amount of triple colocalization between RyR, NCX, and Cav-3, and a consistently low to moderate level of colocalization in all two-protein combinations in this experiment.  This contrasts with the three previous triple-labeling experiments in which one double combination of proteins dominates (e.g. RyR/Cav1.2 and RyR/CSQ).  The images in Figs.2.10 C-E provide an explanation for these values.  The dominating pattern of organization between these 3   36 proteins is shown in Fig.2.10C.  Here we see large clusters of each of the three proteins overlapping to various extents (therefore creating a spectrum of colors) and just a handful of triply colocalized voxels.  This type of organization results in a moderate number of doubly colocalized voxels for all combinations of proteins.  The areas of triple colocalization that do exist seem to be in small regions at the edges of protein clusters, with no specific pattern (e.g. Fig.2.10D).  Occasionally, there are areas of triple colocalization that fall in the center of a well-defined area that contains small, even clusters of each protein (e.g. Fig.2.10E).  Overall, the images shown here give the impression of a somewhat random structural distribution and a relatively loose association between RyR, NCX, and Cav-3.  Interestingly however, all colocalization values between the three proteins are much higher than would be expected by chance (p<0.001). 2.3 Discussion      2.3.1 Imaging We have used established imaging protocols, adapted to allow for triple-labeling of cells, to investigate the distribution of proteins implicated in EC coupling in rat atrial myocytes.  Prior to beginning these triple-labeling experiments, a series of controls were performed to ensure the validity of our results.  These included the single-labeling of myocytes with each antibody that was to be used in our experiments.  We compared the labeling characteristics of each antibody in single- and triple-labeling experiments and made certain that these characteristics did not change with the addition of multiple antibodies simultaneously.  In addition, we compared the calculated densities and double colocalization values for proteins included in more than one triple-labeling experiment and importantly, observed no significant differences.  Finally, using monoclonal   37 antibodies that localized to very specific and separate regions of the myocyte, we determined that our protocols for triple-labeling experiments did indeed prevent the cross-reaction of anti-mouse secondary antibodies, and that we could confidently label myocytes with multiple monoclonal antibodies.  By completing this wide range of controls, we are certain that the results presented here are an accurate representation of the actual cellular structure we are examining.      2.3.2 Structural organization of atrial myocytes Through four triple-labeling experiments, we have identified eight distinct surface populations of RyR, depending on its binding partners.  These are: 1) RyR with Cav1.2, 2) RyR with CSQ, 3) RyR with NCX, 4) RyR with Cav1.2 and NCX, 5) RyR with CSQ and NCX, 6) RyR with Cav-3, 7) RyR with Cav1.2 and Cav-3, and 8) RyR with NCX and Cav-3.  Importantly, through sophisticated statistical analysis we have shown that the colocalization values we recorded for each of these 8 groups are significantly higher than values that would occur purely by chance. In contrast to atrial cells, while rat ventricular myocytes have a very high colocalization of RyR with both Cav1.2 and CSQ, its association with NCX in these cells is negligible (Scriven et al., 2000).  Similarities between the two cell types are seen however, in the association of Cav-3 proteins with RyR and Cav1.2.  Our observation that a portion of the total Cav1.2 in atrial myocytes colocalizes with Cav-3 agrees with previous studies in both rat and mouse ventricular myocytes (Balijepalli et al., 2006; Scriven et al., 2005).  Interestingly, at the surface of ventricular myocytes specifically, Cav-3 seems to associate equally with RyR and Cav1.2, presumably at the dyads.  In the interior however, Cav-3 colocalizes with a distinct subpopulation of RyR alone (Scriven   38 et al., 2005).  We have shown here that at the surface of atrial myocytes, although there does appear to be a small population of RyR, Cav1.2, and Cav-3 together (~10% of the total surface RyR), there are actually more RyR with Cav-3 alone (~15%).  In order to fully evaluate the differences between atria and ventricle however, the identical triple- labeling experiments that were performed here would have to be done in the ventricle. Then, colocalization values could be compared fairly between the two cell types. In atria, because of the high colocalization of RyR with Cav1.2 and RyR with CSQ, we hypothesize that the 4th and 5th surface populations of RyR listed above (RyR/Cav1.2/NCX and RyR/CSQ/NCX) are likely the same.  This is further supported by the images showing virtually identical distribution patterns of these groups – that is, primarily RyR tightly linked with either Cav1.2 or CSQ and surrounded by large clusters of NCX that occasionally join at their edges to form a triply colocalized group (Figs. 2.4 and 2.6).  The possibility that the 7th RyR surface population listed above (RyR/Cav1.2/Cav-3) is also not a distinct subpopulation, but instead that RyR/Cav1.2/CSQ/NCX/Cav-3 are all contained within one surface group, cannot be ruled out.  This is due to the presence of a significant portion of triply colocalized voxels of RyR, Cav-3, and NCX together (the 8th surface RyR population listed above).  However, we hypothesize that the RyR/Cav1.2/Cav-3, and therefore the RyR/NCX/Cav-3, populations are indeed structurally distinct.  This hypothesis was formed based on the fact that the organization of RyR, NCX, and Cav-3 is clearly different from that of RyR with Cav1.2 or CSQ, and NCX.  Instead of distinct overlapping between the centers of protein clusters, we see large clusters of NCX and Cav-3 coming together primarily at their edges and occasionally coming into close contact with RyR.  Overall, the triply   39 colocalized voxels that do exist between these 3 proteins do not seem to be formed in the same fashion as those in the RyR/Cav1.2/NCX and RyR/CSQ/NCX groups.  However, using two new techniques of statistical analysis, we have shown that all of the triple- colocalization values reported here are highly significant.  The probability of finding three of these proteins within the same voxel purely by chance is very small, and therefore it is highly likely that the triply-colocalized groups we have identified here have come together as a result of specific trafficking to the same subcellular space.  For all further discussion then, the surface RyR populations that will be considered to be structurally and functionally distinct are: 1) RyR with Cav1.2, 2) RyR with CSQ, 3) RyR with NCX, 4) RyR with CSQ, Cav1.2, and NCX, 5) RyR with Cav-3, 6) RyR with Cav1.2 and Cav-3, and 7) RyR with NCX and Cav-3.      2.3.3 Functional significance of atrial RyR subpopulations One of the biggest differences between atrial and ventricular myocytes that we report here is the variation in the level of interior colocalization of CSQ with RyR.  In atrial cells, this colocalization is significantly less than in ventricles (~45% and 61% respectively), indicating many more central RyR in atria that are not tightly coupled to CSQ.  Under the assumption that CSQ acts as a luminal Ca2+ sensor for, and Po inhibitor of, RyR channels when it is bound (Gyorke et al., 2004; Terentyev et al., 2007), the central RyR in atrial myocytes would be more likely to release a greater amount of Ca2+ than their ventricular counterparts.  Indeed, it has been reported that Ca2+ sparks in atrial cells release up to three times the number of Ca2+ ions as sparks in the ventricle (Woo et al., 2003).  However, reports have also shown differences in the amplitude and frequency of Ca2+ sparks between the periphery and center of atrial myocytes, and we observe here   40 that no such variation in the CSQ with RyR colocalization exists between these two regions.  Therefore, if CSQ is indeed acting as a potent inhibitor of RyR Po, it may be (at least partially) responsible for the difference between atria and ventricle in the number of Ca2+ ions released during a spark, but it is highly unlikely that it is responsible for the variation in Ca2+ spark characteristics from the periphery to the center of atrial myocytes. We have also concluded from these experiments that there is a distinct RyR subpopulation at the surface of atrial myocytes that, of the four other proteins tested for, only associates with CSQ.  Although this population of RyR would not be able to initiate (electrically-invoked) Ca2+ release, it could act as a potential amplifier of the Ca2+ response following initiation from other surface RyR. We have shown here that approximately 35% of all surface RyR colocalize only with Cav1.2 channels.  Compared to the ~10% that colocalize solely with NCX, it would appear that Cav1.2-mediated CICR is the primary method of initiating Ca2+ release from RyR during EC coupling.  Interestingly, although atrial myocytes are similar to neonatal ventricular cells in their lack of a well-developed TATS, they do not share the same heightened level of RyR/NCX colocalization (Dan et al., 2007).  It seems therefore that atrial cells do not rely on NCX-mediated CICR as neonates do (Huang et al., 2008), which was unclear to this point.  If however, the RyR that colocalize with NCX also closely associate with voltage-gated Na+ channels, then the possibility that these sites function by NCX-mediated CICR becomes much more likely.  Whether or not this is the case has yet to be determined.  Because these sites make up only ~10% of the total surface RyR however, Cav1.2-mediated CICR would still be the dominating form.  In the cases where RyR are tightly linked to both Cav1.2 and NCX, although the trigger for   41 CICR is likely to come almost entirely from ICa,L (Sipido et al., 1997), NCX may play an important role in modulating the functionality of these specific groups. Previous published works suggest that if NCX molecules are localized in just the right cellular space, then they are highly likely to alter the EC coupling process (Bers, 2008).  The very close association of RyR, Cav1.2, and NCX in particular surface domains then, implies that NCX would have the capacity to pre-elevate local [Ca2+]i in these dyads and sensitize RyR prior to activation by ICa,L.  What this further implies, is that these sites would potentially be more likely to release Ca2+, and would do so more quickly than surface RyR that are associated with Cav1.2 alone.  By this reasoning, we hypothesize that the sites of triple colocalization between RyR, Cav1.2, and NCX that we have shown form the structural basis for ‘eager sites’ of Ca2+ release at the surface of atrial myocytes. All additional surface RyR populations that we have identified here include caveolin-3 proteins – specifically RyR with Cav-3 alone, RyR with Cav1.2 and Cav-3, and RyR with NCX and Cav-3.  Prior studies have shown that the efficiency of coupling between Ca2+ entry and release is significantly reduced when caveolae are disrupted in rat ventricular myocytes (Calaghan and White, 2006).  This agrees with our data that show a significant portion of RyR and Cav1.2 within very close proximity to Cav-3 – the assumption of course being that the disruption of caveolae damages the functional linkage between RyR and Cav1.2.   Furthermore, disruption of caveolae in rat arterial smooth muscle and neonatal ventricular myocytes – both lacking a well-developed TATS much like atrial cells – has been shown to cause a reduction in the frequency, width, and amplitude of Ca2+ sparks (Lohn et al., 2000).  Most interestingly, this alteration in Ca2+   42 spark characteristics occurred without modulation of ICa,L and therefore provides additional evidence for the population of RyR colocalized only with Cav-3 that we have identified.  Greater functional implications due to the close association of RyR with caveolae, with or without Cav1.2, are inevitable due to the scaffolding activity of, and regulation of signaling molecules by, caveolin-3 (Cohen et al., 2004).  For example, in the case of the RyR/NCX/Cav-3 subpopulation, we must assume that there are multiple influences on the actions of RyR and NCX from other molecules localized to the caveolae.  However, without a comprehensive listing of the proteins within these particular caveolae (because as we and others have shown, (Balijepalli et al., 2006; Scriven et al., 2005) the inhabitants and/or neighbours of individual caveolae can vary greatly), we are unable to accurately predict the functional consequences of this, and the RyR/Cav-3 and RyR/Cav1.2/Cav-3, populations.  We can say with some certainty however, that RyR alone or with either NCX or Cav1.2, are likely being specifically directed either near to or within specific populations of the caveolae in atrial myocytes. Furthermore, because of increasing evidence on the role of caveolae in the regulation of EC coupling (Balijepalli et al., 2006; Bossuyt et al., 2002; Calaghan and White, 2006; Scriven et al., 2005), these associations are unlikely to be trivial.      2.3.4 Conclusion The distribution of some of the major proteins involved in excitation-contraction coupling of rat atrial myocytes was elucidated using novel triple-labeling immunofluorescence techniques.  We have provided abundant evidence to support our hypothesis that atrial RyR complexes are heterogeneous and may be targeted to specific intracellular domains.  In particular, we have identified seven surface populations of   43 RyR, based on its binding partners, that we believe are structurally and functionally distinct.  These are: • RyR with Cav1.2 • RyR with CSQ • RyR with NCX • RyR with CSQ, Cav1.2, and NCX • RyR with Cav-3 • RyR with Cav1.2 and Cav-3 • RyR with NCX and Cav-3 From the identification of these groups, we have come to a number of conclusions.  First, Cav1.2-mediated CICR is the dominant form in atrial myocytes.  The structural basis for NCX-mediated CICR is present in these cells, but the localization of voltage-gated Na+ channels with respect to RyR and NCX needs to be determined before any concrete conclusions can be drawn.  Second, the discovery of a distinct population of RyR, CSQ, Cav1.2, and NCX tightly coupled together provides a very likely candidate for the structural basis of eager sites present at the periphery of atrial myocytes.  Third and finally, the characteristics of atrial Ca2+ sparks and the efficiency of EC coupling in certain regions of atrial myocytes may be controlled by caveolae that contain or are neighboring to RyR with or without Cav1.2 or NCX. One of the main principles of the local control theory of EC coupling – that all RyR are functionally equivalent – is significantly challenged by the discovery of the array of distinct RyR complexes shown here.  While each individual RyR complex may   44 be independently controlled, their molecular constituents and binding partners are highly unlikely to be identical. 2.4 Materials and Methods All chemicals were purchased from Sigma-Aldrich (Oakville, ON) unless otherwise stated.  Animal handling was done in accordance with the guidelines of the Canadian Council on Animal Care.      2.4.I Cell isolation and preparation Atrial myocytes were isolated from freshly excised hearts using the method of Rodrigues and Severson (Rodrigues, 1997).  Briefly, adult male Wistar rats weighing 200-250g were given 200 units of heparin (Organon Canada, Toronto, ON) intra- peritoneally 15-20 min. prior to sacrifice with sodium pentobarbital (20mg/100g; MTC Pharmaceuticals, Cambridge, ON). The heart was immediately excised and hung for retrograde Langendorff perfusion with warm (37°C) Joklik MEM (M0518) supplemented with, in mM, 23 NaHCO3, 1.2 MgSO4, and 1 DL-carnitine that had been equilibrated with 95% O2/5% CO2. Perfusion was adjusted to give a flow of 7ml/min and maintained for 5 min. to drain the heart of blood. It was then switched to a Joklik solution containing 162 U/mL type II collagenase (Worthington Biochemical, Lakewood, NJ). Once the heart softened, the atria were separated from the ventricles, minced, and filtered through a Nitex nylon mesh (200 µm).  Preparations contained approximately 30-50% rod-shaped quiescent cells that were incubated for 20 min. at 37°C in M-199 (M4530) supplemented with, in mM, 25 hepes, 1 DL-carnitine, 0.1 Insulin, 0.56 penicillin, 0.14 streptomycin sulfate, 2 EGTA, and 0.01g/mL fatty acid-free BSA (pH7.4).  Cells were then fixed for 10 min. in freshly made 2% paraformaldehyde.  Fixation was quenched with 100mM   45 glycine (pH 7.4) for 10 min. after which cells were washed 3 x 10 min. in phosphate buffered saline (PBS; in mM: 137 NaCl, 8 NaH2PO4, 2.7 KCl, 1.5 KH2PO4. pH 7.4). Cells were then permeabilized with 0.1% Triton X-100 for 10 min followed by 3 x 10 min. washes in PBS.      2.4.2 Immunolabeling Primary antibodies were:  monoclonal anti-RyR2 (MA3-916; Affinity BioReagents, Golden, CO); polyclonal anti-calsequestrin (PA1-913; Affinity BioReagents); monoclonal anti-caveolin-3 (610420; BD Biosciences, Mississauga, ON); monoclonal anti-NCX (R3F1; Swant, Bellinzona, Switzerland); and polyclonal anti- Cav1.2 (CNC1, gift of Dr. W Catterall (Hell et al., 1993)). Secondary antibodies were affinity purified and highly cross-adsorbed to minimize species cross-reactivity and were either goat anti-rabbit or goat anti-mouse conjugated to one of Alexa Fluor 350, 488, or 594 (Molecular Probes, Eugene, OR). Immunolabeling is as previously described (Scriven et al., 2000) with the following differences:  Incubations involving one monoclonal and one polyclonal primary antibody were performed sequentially overnight at 4°C.  All labeling experiments involving two monoclonal primary antibodies were done using mouse IgG1 Zenon labeling kits (Molecular Probes).  Immunolabeling with three monoclonal primary antibodies was done by incubating cells with one primary antibody overnight at 4°C, followed by a goat anti-mouse secondary antibody conjugated to an appropriate fluorophore for 1.5 hours at room temperature.  Cells were then incubated simultaneously with the other two primary antibodies using a mouse IgG1 Zenon labeling kit. Experiments involving one polyclonal and two monoclonal antibodies were done one of   46 the two following ways: 1) cells were incubated with one monoclonal antibody overnight at 4°C, and then with an anti-mouse secondary antibody for 1.5 hours at room temperature.  The second monoclonal primary was conjugated to its fluorophore using a Zenon labeling kit and then added to cells for 1 hour at room temperature.  The third primary antibody, a polyclonal, was added to cells overnight at 4°C and then an anti- rabbit secondary antibody conjugated to a chosen fluorophore was added for 1.5 hours at room temperature.  2) cells were incubated with the polyclonal antibody overnight at 4°C, and then with an anti-rabbit secondary antibody for 1.5 hours at room temperature.  The myocytes were then incubated simultaneously with the two monoclonal antibodies that had been conjugated to their appropriate fluorophore using a mouse IgG1 Zenon labeling kit.  For each triple labeling experiment, an appropriate control experiment was performed in order to ensure no cross-reaction between the monoclonal antibodies.      2.4.3 Imaging Images were captured using a Zeiss AxioObserver inverted microscope with a 63x/1.4 oil immersion objective and then passed through a narrow bandpass filter (Semrock, Rochester, NY) specific for the chosen fluorophore.  Images were then captured on a thermoelectrically cooled CCD camera with an 80% quantum efficiency (SITe SI502AB chip). The Z position of the sample was controlled by a PZM2000 piezo stage (Applied Scientific Instrumentation, Eugene, OR).  Pixel size of these images was 95nm in X and Y and were acquired at 0.25 μm intervals in Z.      2.4.4 Image deconvolution and analysis A detailed description of these processes can be found in Scriven et al. (2000). Briefly, wide-field images were deconvolved using the algorithm developed by   47 Carrington (Carrington et al., 1995).  All images were dark-current and background subtracted and flat-field corrected to allow for non-uniformity in illumination and camera sensitivity across the field of view.  After deconvolution, the images were aligned using fiduciary markers that emitted in all wavelengths. Finally, images were peeled one layer of voxels at a time and numbered from 0 (surface) to 12 (center) (Scriven et al., 2005). Because of some uncertainty in the exact position of the surface, layer –1 (immediately outside the surface) was also included.  Colocalization and labeling density (# of lit voxels/total # of voxels) were measured as a function of distance into the cell; ‘surface’ and ‘interior’ representing the mean values obtained from layers –1 to +2 and +4 to +10 respectively.  Only voxels with identical x, y, and z coordinates were recorded as colocalized, and all values were obtained from cell segments chosen to eliminate possible edge effects (Scriven et al., 2008).  Colocalization was calculated in one of two ways (Fig. 2.12).  First, the voxels encompassing all 3 antibodies were counted as triply colocalized (e.g. A with B and C), but not as part of the ‘double’ colocalized values (e.g. A with B, B with C, and C with A).  This method was used for measuring atrial colocalization values (Tables 2.1-2.4).  Second, voxels that included all 3 antibodies (e.g. A + B + C) were included in all 3 ‘double’ colocalization values for that set (e.g. A with B, B with C, and C with A).  This method was used to compare the molecular architecture of the atria to the ventricle (Table 2.5), since the latter were double and not triple label experiments.  In addition, this method was vital as a control to ensure that dual colocalization values were not significantly different across experiments that used some of the same antibodies (data not shown).   48 Simply determining that colocalization has occurred between two or more proteins however, is not enough.  It is imperative to determine if the measured colocalization represents a significant event or if it could have occurred purely by chance. In 2005, Scriven et al. used a computer-based Monte Carlo randomization technique whereby individual pixels in an image were randomly placed in a volume equivalent to that being studied. This randomization was repeated 1000 times and the derived colocalizations were ranked.  If the rank of the observed colocalization fell in the top 50 of the random distribution (p < 0.05), the colocalization was regarded as significant.  An underlying assumption of this technique however is that the pixels in an image are independent of one another, which is not the case.  A single point source, even when deconvolved, contains between 35 and 50 voxels, depending on point intensity and the image threshold.  For this reason, our lab (D.Scriven and P.Fletcher) developed a new technique based on the block-bootstrapping method (Carlstein et al., 1998) and used it to analyze the significance of colocalization values obtained in the experiments reported here. For this technique, two approaches were used.  In the first, each object in the original image (defined as a group of face-connected (6-connected) lit voxels) was isolated, its centre of mass determined, and then placed in memory.  Also, the mask of each image, which determines the areas of the image to be analyzed, had its voxel coordinates (x, y, z) placed in a linear array.  This facilitated a single-step random pick of a coordinate within the mask.  One object was then selected at random from memory, and its center was placed at one of the random coordinates from the mask.  Object pixels that fell outside the mask were excluded.  The process was stopped when the number of lit   49 pixels in the generated image matched those within the mask of the original image.  This process was repeated for each wavelength and the colocalizations between the lit voxels of each wavelength were measured. The simulation was repeated 1000 times, the generated data ranked, and the position of the observed data in this ranking was used to determine its probability of occurrence by chance. The second approach to the technique is a modification of the first.  Once the objects are isolated, the local maxima are found.  If there is more than one maximum, the objects are cut along the minima lying between the local maxima to produce multiple ‘blobs’ from that object.  Blobs are then put in to memory and selected at random as before, with the maximum of each blob being placed at random positions within the mask of the image.  The stop criterion and number of simulations are the same as for the first method. The bootstrapping technique requires that selection of each object and its position be made at random.  This means that a single object or blob can be chosen multiple times or not at all, and that they can be placed on top of one another. Randomization was achieved using the Mersenne twister (Matsumoto and Nishimura, 1998), a pseudo- random generator that has an equal probability of choosing any element and has a period of 219937 − 1. The two techniques outlined here were applied to all images obtained in these experiments, and gave virtually identical results; the only difference being that the second approach was 8X faster than the first.  Furthermore, increasing the number of iterations to 2000 for each generated data point did not alter the significance of the colocalization values.   50 Overall, data are presented as means + standard errors. Groups of data were compared using a student’s t-test or one-way ANOVA, and a probability of P<0.05 was taken as significant.   51 2.5 Tables  TABLE 2.1 Colocalization of RyR with NCX and Cav1.2  Protein Surface Colocalization (%) RyR with NCX RyR with Cav1.2  RyR with NCX and Cav1.2        RyR 10.2 ± 2.5 34.0 ± 3.8 10.4 ± 1.8 NCX with RyR NCX with Cav1.2 NCX with RyR and Cav1.2  NCX 13.6 ± 2.6 8.5 ± 2.6 13.8 ± 2.2 Cav1.2 with RyR Cav1.2 with NCX Cav1.2 with RyR and NCX  Cav1.2 49.5 ± 3.5 9.9 ± 2.8 14.2 ± 2.1  Values are mean ± s.e.m. n=10  TABLE 2.2 Colocalization of RyR with CSQ and NCX   Protein Surface Colocalization (%) RyR with CSQ RyR with NCX RyR with CSQ and NCX  RyR 35.5 ± 3.4 9.8 ± 1.8 8.5 ± 1.1 CSQ with RyR CSQ with NCX CSQ with RyR and NCX  CSQ 37.5 ± 3.4 9.7 ± 2.1 9.1 ± 1.2 NCX with RyR NCX with CSQ NCX with RyR and CSQ  NCX 9.0 ± 0.9 8.4 ± 1.2 8.2 ± 0.8  Values are mean ± s.e.m. n=10   52  TABLE 2.3 Colocalization of RyR with Cav-3 and Cav1.2  Protein Surface Colocalization (%) RyR with Cav-3 RyR with Cav1.2  RyR with Cav-3 and Cav1.2  RyR 15.8 ± 5.0 39.2 ± 3.0 9.6 ± 2.0 Cav-3 with RyR Cav-3 with Cav1.2 Cav-3 with RyR and Cav1.2  Cav-3 9.7 ± 4.4 7.4 ± 2.2 5.6 ± 1.6 Cav1.2 with RyR Cav1.2 with Cav-3 Cav1.2 with RyR and Cav-3  Cav1.2 39.4 ± 7.5 11.5 ± 2.8 9.8 ± 2.8  Values are mean ± s.e.m. n=5   TABLE 2.4 Colocalization of RyR with NCX and Caveolin-3   Protein Surface Colocalization (%) RyR with NCX RyR with Cav-3  RyR with NCX and Cav-3  RyR 15.9 ± 3.2 18.6 ± 2.2 8.4 ± 2.2 NCX with RyR NCX with Cav-3 NCX with RyR and Cav-3  NCX 20.4 ± 2.2 23.8 ± 3.6 11.4 ± 2.3 Cav-3 with RyR Cav-3 with NCX Cav-3 with RyR and NCX  Cav-3 14.8 ± 2.1 13.2 ± 1.4 5.7 ± 0.4  Values are mean ± s.e.m. n=6      53  TABLE 2.5: Colocalization in Atria vs. Ventricle  Atrial Colocalization Colocalized Proteins Surface (%) Interior (%) n Ventricular Colocalization* (%) n RyR with CSQ 46.1 ± 4.1 51.1 ± 3.6 8 55.8 ± 6.2 (wc) 5 CSQ with RyR 49.6 ± 4.4 45.6 ± 2.6** 8 61.6 ± 7.2 (wc) 5 RyR with NCX 27.4 ± 2.2** N/A 10 7.8 ± 2.3 (wc) 5 NCX with RyR 24.6 ± 2.0** N/A 10 5.8 ± 1.9 (wc) 5 RyR with Cav1.2 44.5 ± 4.9 N/A 10 36.7 ± 4.8 (wc) 6 Cav1.2 with RyR 58.5 ± 4.0 N/A 10 56.7 ± 5.1 (wc) 6 RyR with Cav-3 26.8 ± 3.7** N/A 5 37.6 ± 4.0 (s) 11 Cav-3 with RyR 15.6 ± 3.4 N/A 5 22.2 ± 1.9 (s) 11 Cav-3 with Cav1.2 11.3 ± 1.8 N/A 5 10.95 ± 1.42 (s) 9 Cav1.2 with Cav-3 19.7 ± 1.6** N/A 5 39.05 ± 3.44 (s) 9 * Ventricular values are either whole-cell (wc) or surface (s), and were taken from Scriven et al. (2000, 2005). ** Denotes significant difference (p<0.05) from ventricular value Note: Atrial colocalization values came from the following sources: 1) RyR/CSQ from independent RyR/CSQ experiment 2) RyR/NCX and RyR/Cav1.2 from RyR/NCX/Cav1.2 experiment 3) RyR/Cav-3and Cav-3/Cav1.2 from RyR/Cav-3/Cav1.2 experiment   54  2.6 Figures  Fig. 2.1 Single labeling of atrial myocytes Atrial myocyte segment single-labeled with antibodies against RyR (A), CSQ (B), Cav1.2 (C), NCX (D), or Cav- 3 (E).  For each antibody, the image shown is either a 1.5 μm thick section taken from the cell surface (left panel), or a single 250nm plane taken from approximately the middle of the cell (right panel).  For the interior segments of RyR and CSQ (A and B respectively, right panel), some longitudinal elements of labeling are present (arrows), as well as a discrete subsarcolemmal ‘gap’ with no labeling of either protein (arrowheads). Scale bar is 5 μm.  N = nucleus.   55  Fig. 2.2 Dual labeling of RyR and NPC Deconvolved image of an atrial myocyte labeled with two monoclonal antibodies (RyR-red, NPC-green) using a Zenon labeling kit. The image is a 250nm thick slice taken from approximately the center of the cell.  RyR labels at the Z-lines (arrow), while NPC localizes to the nuclear membrane (arrowhead).  Scale bar is 5 μm.  N=nucleus.                            56  Fig. 2.3 Density graphs for RyR (A), CSQ (B), Cav1.2 (C), NCX (D), and Cav-3 (E).  N values are the same as for each triple labeling experiment (RyR/NCX/Cav1.2 = 10; RyR/NCX/CSQ = 10; RyR/CSQ = 8; RyR/Cav-3/Cav1.2 = 5; RyR/NCX/Cav-3 = 6).     0% 5% 10% 15% 2 0% 25% 3 0% Surface Interio r %  I l l u m i n a t i o n  o f  R y R RyR/NCX/Cav1.2 RyR/NCX/CSQ RyR/Cav-3/Cav1.2 RyR/NCX/Cav-3 A 0 % 5% 10 % 15% 20 % 2 5% 30 % Surface Inte rio r %  I l l u m i n a t i o n  o f  C S Q RyR/NCX/CSQ RyR/CSQ B     0 % 5% 10 % 15% 20 % 2 5% 30 % Surface Interio r %  I l l l u m i n a t i o n  o f  C a v 1 . 2 RyR/NCX/Cav1.2 RyR/Cav-3/Cav1.2 C 0% 5% 10% 15% 2 0% 25% 3 0% Surface Interio r %  I l l u m i n a t i o n  o f  N C X RyR/NCX/Cav1.2 RyR/NCX/CSQ RyR/NCX/Cav-3 D 0 % 5% 10 % 15% 20 % 2 5% 30 % Surface Inte rio r%  I l l u m i n a t i o n  o f  C a v - 3 RyR/Cav-3/Cav1.2 RyR/NCX/Cav-3 E    57  Fig. 2.4 Atrial myocyte segments labeled with antibodies against RyR (blue), NCX (green), and Cav1.2 (red).  All images are 1μm thick slices from the cell surface.  Voxels containing all 3 proteins (triply colocalized) are shown in white.  RyR/Cav1.2 double colocalization is shown in pink/purple, NCX/RyR in light blue, and NCX/Cav1.2 in yellow.  Scale bar is 5μm. N=10. A View of the entire cell showing surface distribution and colocalization of RyR, NCX, and Cav1.2 B Segment of the cell surface from enclosed area in A (1.2X magnification). C-E Segments from topmost, middle, and lowermost boxed regions in B respectively (4.1X).  Arrows indicate clusters of NCX that are separate from the main triple colocalized area.  Arrowheads show NCX extending beyond triple colocalized region.      58  Fig. 2.5 Colocalization of RyR, NCX, and Cav1.2.  Mean ± SE.  N =10.  * denotes significant difference (p<0.05). A Colocalization of RyR with NCX and Cav1.2  B Colocalization of Cav1.2 with RyR and NCX  C Colocalization of NCX with RyR and Cav1.2     0 10 20 30 40 50 60 NCX with RyR and  Cav1.2 NCX with RyR NCX with Cav1.2 %  C o l o c a l i z a t i o n C 0 10 20 30 40 50 60 RyR with NCX and  Cav1.2 RyR with NCX RyR with Cav1.2 %  C o l o c a l i z a t i o n ** A 0 10 20 30 40 50 60 Cav1.2  with RyR and  NCX Cav1.2  with RyR Cav1.2  with NCX %  C o l o c a l i z a t i o n * * B *                  59  Fig. 2.6 Atrial myocyte segments labeled with antibodies against RyR (red), NCX (green), and CSQ (blue).  All images are 1μm thick slices from the cell surface.  Voxels containing all 3 proteins are shown in white.  RyR/CSQ double colocalization is shown in pink/purple. Scale bar is 5μm. N=10. A View of whole cell showing surface distribution and colocalization of RyR, NCX, and CSQ B Cell segment from enclosed area in A (1.1X). C-E Cell segments from topmost, middle, and lowermost boxed regions in B (5.2X) respectively.  Arrowheads indicate large clusters of triple colocalization, while arrows point to triply colocalized voxels in the center of a large NCX cluster.        60  Fig. 2.7 Colocalization of RyR, CSQ, and NCX.  Mean ± SE.  N =10.  * denotes significant difference (p<0.05). A Colocalization of RyR with CSQ and NCX  B Colocalization of CSQ with NCX and RyR  C Colocalization of NCX with CSQ and RyR      0 5 10 15 2 0 25 3 0 35 4 0 45 NCX with CSQ and RyR NCX with CSQ NCX with RyR %  C o l o c a l i z a t i o n C 0 5 10 15 20 25 30 35 40 45 RyR with CSQ and NCX RyR with CSQ RyR with NCX %  C o l o c a l i z a t i o n * A * 0 5 10 15 20 25 30 35 40 45 CSQ with NCX and RyR CSQ with NCX CSQ with RyR %  C o l o c a l i z a t i o n ** B           61  Fig. 2.8 Atrial myocyte segments labeled with antibodies against RyR (blue), Cav-3 (green), and Cav1.2 (red).  All images are 1μm thick slices from the cell surface.  Voxels containing all 3 p (triply colocalize shown in white. RyR/Cav1.2 double colocalization is shown in pink/purple.  Scale bar is 5μm. N=5. roteins d) are A View of the entire cell showing surface distribution and colocalization of RyR, Cav-3, and Cav1.2 B Segment of the cell surface from enclosed area in A (1.2X magnification). C-E Segments from topmost, middle, and lowermost boxed regions in B respectively (8.0X). Arrows indicate the tight association between RyR and Cav1.2 while arrowheads show the neighboring Cav-3 clusters.   62  Fig. 2.9 Colocalization of RyR, Cav-3, and Cav1.2.  Mean ± SE.  N =5.  * denotes significant difference (p<0.05).  A Colocalization of RyR with Cav-3 and Cav1.2  B Colocalization of Cav-3 with RyR and Cav1.2  C Colocalization of Cav1.2 with  RyR and Cav-3        0 10 2 0 3 0 4 0 50 RyR with Cav- 3 and Cav1.2 Ryr with Cav-3 RyR with Cav1.2 %  C o l o c a l i z a t i o n * * A 0 10 2 0 3 0 4 0 50 Cav-3 with RyR and Cav1.2 Cav-3 with RyR Cav-3 with Cav1.2 %  C o l o c a l i z a t i o n B 0 10 20 30 40 50 Cav1.2 with RyR and Cav- 3 Cav1.2 with RyR Cav1.2 with Cav-3 %  C o l o c a l i z a t i o n ** C                63  Fig. 2.10 Atrial myocyte segments labeled with antibodies against RyR (red), NCX (green), and Cav-3 (blue).  All images are 1μm thick slices from the cell surface.  Voxels containing all 3 proteins (triply colocalized) are shown in white.  RyR/NCX double colocalization is shown in yellow, RyR/Cav-3 in pink/purple, and NCX/Cav-3 in light blue.  Scale bar is 5μm. N=6. A View of the entire cell showing surface distribution and colocalization of RyR, NCX, and Cav-3. B Segment of the cell surface from enclosed area in A (1.1X magnification). C-E Segments from topmost, middle, and lowermost boxed regions in B respectively (7.9X).                  64  Fig. 2.11 Colocalization of RyR, NCX, and Cav-3.  Mean ± SE.  N =6.  * denotes significant difference (p<0.05).  A Colocalization of RyR with Cav-3 and NCX  B Colocalization of NCX with Cav-3 and RyR  C Colocalization of Cav-3 with NCX and RyR   0 5 10 15 2 0 25 3 0 35 RyR with Cav- 3 and NCX RyR with Cav- 3 RyR with NCX %  C o l o c a l i z a t i o n * A 0 5 10 15 20 2 5 30 3 5 Cav-3 with NCX and RyR Cav-3 with NCX Cav-3 with RyR %  C o l o c a l i z a t i o n * C * 0 5 10 15 20 2 5 30 3 5 NCX with Cav- 3 and RyR NCX with Cav- 3 NCX with RyR %  C o l o c a l i z a t i o n * B            65  Figure 2.12 Calculating colocalization for triple-labeled images.  A, B, and C represent specific labeling produced by 3 antibodies.  The region that overlaps in all 3 colors (blue, red, and green) is shown in grey, and would be counted as a ‘triple’ colocalization (e.g. A with B and C, B with A and C, and C with A and B).  The striped areas (asterisks) are regions of overlap between 2 of the antibodies and would count as ‘double’ colocalization (e.g. A with B, B with C, or C with A).  Areas designated as triply colocalized are excluded from double colocalization values (Tables 2.1-2.4). For comparison with ventricular colocalization values however, the grey region is included in all ‘double’ colocalization values in order to perform a fair comparison between colocalizations in the two cell types (Table 2.5).    * * *  66  2.7 References  Balijepalli, R. C., Foell, J. D., Hall, D. D., Hell, J. W. and Kamp, T. J. (2006). Localization of cardiac L-type Ca(2+) channels to a caveolar macromolecular signaling complex is required for beta(2)-adrenergic regulation. Proc Natl Acad Sci U S A 103, 7500-5.  Beard, N. A., Casarotto, M. G., Wei, L., Varsanyi, M., Laver, D. R. and Dulhunty, A. F. (2005). Regulation of ryanodine receptors by calsequestrin: effect of high luminal Ca2+ and phosphorylation. Biophys J 88, 3444-54.  Berlin, J. R. (1995). Spatiotemporal changes of Ca2+ during electrically evoked contractions in atrial and ventricular cells. Am J Physiol 269, H1165-70.  Bers, D. M. (2008). Calcium Cycling and Signaling in Cardiac Myocytes. Annu. Rev. Physiol. 70, 23-49.  Bootman, M. D., Higazi, D. R., Coombes, S. and Roderick, H. L. (2006). Calcium signalling during excitation-contraction coupling in mammalian atrial myocytes. J Cell Sci 119, 3915-25.  Bossuyt, J., Taylor, B. E., James-Kracke, M. and Hale, C. C. (2002). The cardiac sodium-calcium exchanger associates with caveolin-3. Ann N Y Acad Sci 976, 197-204.  Calaghan, S. and White, E. (2006). Caveolae modulate excitation-contraction coupling and beta2-adrenergic signalling in adult rat ventricular myocytes. Cardiovasc Res 69, 816-24.  Carlstein, E., Do, K. A., Hall, P., Hesterberg, T. and Kunsch, H. R. (1998). Matched-block bootstrap for dependent data. Bernoulli 4, 305-328.  Carrington, W. A., Lynch, R. M., Moore, E. D., Isenberg, G., Fogarty, K. E. and Fay, F. S. (1995). Superresolution three-dimensional images of fluorescence in cells with minimal light exposure. Science 268, 1483-7.  Cohen, A. W., Hnasko, R., Schubert, W. and Lisanti, M. P. (2004). Role of caveolae and caveolins in health and disease. Physiol Rev 84, 1341-79.  Dan, P., Lin, E., Huang, J., Biln, P. and Tibbits, G. F. (2007). Three- dimensional distribution of cardiac Na+-Ca2+ exchanger and ryanodine receptor during development. Biophys J 93, 2504-18.  Fabiato, A. (1983). Calcium-induced release of calcium from the cardiac sarcoplasmic reticulum. Am J Physiol 245, C1-14.  Gyorke, I., Hester, N., Jones, L. R. and Gyorke, S. (2004). The role of calsequestrin, triadin, and junctin in conferring cardiac ryanodine receptor responsiveness to luminal calcium. Biophys J 86, 2121-8.  Hell, J. W., Yokoyama, C. T., Wong, S. T., Warner, C., Snutch, T. P. and Catterall, W. A. (1993). Differential phosphorylation of two size forms of the neuronal class C L-type calcium channel alpha 1 subunit. J Biol Chem 268, 19451-7.  Huang, J., Hove-Madsen, L. and Tibbits, G. F. (2008). Ontogeny of Ca2+- induced Ca2+ release in rabbit ventricular myocytes. Am J Physiol Cell Physiol 294, C516-25.  Huser, J., Lipsius, S. L. and Blatter, L. A. (1996). Calcium gradients during excitation-contraction coupling in cat atrial myocytes. J Physiol 494 ( Pt 3), 641-51.   67   Lohn, M., Furstenau, M., Sagach, V., Elger, M., Schulze, W., Luft, F. C., Haller, H. and Gollasch, M. (2000). Ignition of calcium sparks in arterial and cardiac muscle through caveolae. Circ Res 87, 1034-9.  Mackenzie, L., Bootman, M. D., Berridge, M. J. and Lipp, P. (2001). Predetermined recruitment of calcium release sites underlies excitation-contraction coupling in rat atrial myocytes. J Physiol 530, 417-29.  Matsumoto, M. and Nishimura, T. (1998). Mersenne Twister: A 623- Dimensionally Equidistributed  Uniform Pseudo-Random Number Generator. ACM Transactions on Modeling and Computer Simulation 8, 3-30.  Rodrigues, B. a. S., D.L. (1997). Preparation of Cardiomyocytes. In Biochemical Techniques in the Heart,  (ed. J. H. McNeil), pp. 101-115. Boca Raton, FL: CRC Press.  Scriven, D. R., Lynch, R. M. and Moore, E. D. (2008). Image acquisition for colocalization using optical microscopy. Am J Physiol Cell Physiol 294, C1119-22.  Scriven, D. R. L., Dan, P. and Moore, E. D. W. (2000). Distribution of Proteins Implicated in Excitation-Contraction Coupling in Rat Ventricular Myocytes. Biophysical Journal 79, 2682-2691.  Scriven, D. R. L., Klimek, A., Asghari, P., Bellve, K. and Moore, E. D. W. (2005). Caveolin-3 is Adjacent to a group of Extradyadic Ryanodine Receptors. Biophysical Journal 89, 1893-1901.  Sipido, K. R., Maes, M. and Van de Werf, F. (1997). Low efficiency of Ca2+ entry through the Na(+)-Ca2+ exchanger as trigger for Ca2+ release from the sarcoplasmic reticulum. A comparison between L-type Ca2+ current and reverse-mode Na(+)-Ca2+ exchange. Circ Res 81, 1034-44.  Terentyev, D., Viatchenko-Karpinski, S., Vedamoorthyrao, S., Oduru, S., Gyorke, I., Williams, S. C. and Gyorke, S. (2007). Protein protein interactions between triadin and calsequestrin are involved in modulation of sarcoplasmic reticulum calcium release in cardiac myocytes. J Physiol 583, 71-80.  Wier, W. G., Lopez-Lopez, J. R., Shacklock, P. S. and Balke, C. W. (1995). Calcium signalling in cardiac muscle cells. Ciba Found Symp 188, 146-60; discussion 160-4.  Woo, S. H., Cleemann, L. and Morad, M. (2003). Spatiotemporal characteristics of junctional and nonjunctional focal Ca2+ release in rat atrial myocytes. Circ Res 92, e1-11.    68  3 DISCUSSION      3.I Imaging We have used established imaging protocols, adapted to allow for triple-labeling of cells, to investigate the distribution of proteins implicated in EC coupling in rat atrial myocytes.  Prior to beginning these triple-labeling experiments, a series of controls were performed to ensure the validity of our results.  These included the single-labeling of myocytes with each antibody that was to be used in our experiments.  We compared the labeling characteristics of each antibody in single- and triple-labeling experiments and made certain that these characteristics did not change with the addition of multiple antibodies simultaneously.  In addition, we compared the calculated densities and double colocalization values for proteins included in more than one triple-labeling experiment and importantly, observed no significant differences.  We also incubated myocytes with secondary antibodies alone, and established that these antibodies were incapable of producing specific labeling by themselves.  Furthermore, extensive testing was performed on the potential cross-reaction of anti-mouse secondary antibodies in protocols that involved labeling myocytes with more than one monoclonal antibody.  Using antibodies that localized to very specific and separate regions of the myocyte, we determined that our protocols for triple-labeling experiments did indeed prevent the cross-reaction of anti- mouse secondary antibodies, and that we could confidently label myocytes with multiple monoclonal antibodies.  By completing this wide range of controls, we are certain that the results presented here are an accurate representation of the actual cellular structure we are examining.  That being said however, there are certain limitations to these experiments that should be acknowledged.   69  All antibodies used in our experiments are highly specific for their given epitopes and do not cross react with other cardiac myocyte proteins.  They were also affinity purified to minimize background labeling.  Even with a high affinity antibody however, it is unlikely that all epitopes within the cell are bound (Harlow and Lane, 1988).  It is also possible that these epitopes may be hidden or masked in some way by subcellular structures.  Therefore, although our antibodies likely label a very high fraction of the available epitopes, this number is not 100 percent.  These antibodies may also bind to epitopes on non-functional proteins, clouding the interpretation of our results.  This is also however, likely to represent a small proportion of the labeling within the cell. The settings used to capture images on our microscope are designed to produce the most accurate results possible.  For example, our voxel sizes (100 nm in X and Y and 250 nm in Z) follow nyquist criteria, which ensure that our images accurately represent the objects they are based on.  There are limitations however in the resolving power of our microscope.  For two objects labeled in the same wavelength, the resolving power was determined by measuring the full width half maximum (FWHM) of a point-spread function (PSF) for that wavelength, and is equal to approximately 200nm in the X- and Y-axes.  A colocalization table was used to measure the resolving power of two objects labeled in different wavelengths, and this is approximately 50nm in X and Y. Finally, another limitation to our experiments is the wide range of atrial cell types that exist.  As previously mentioned, the morphology of atrial myocytes can range from clear, purkinje-like cells, to cells with rudimentary TATS, to electron-dense cells that resemble ventricular myocytes.  The cell preparations for our triple-labeling experiments came from both left and right atria, and contained all cell types.  Because of the limited yield of   70  live myocytes and the fact that no rigid morphological classification system exists for atrial cells, labeled myocytes of all morphologies were used for final colocalization analysis.  Therefore, the data presented here is an average of all cell types, and may not accurately represent colocalization values for each individual cell class.      3.2 Data Analysis Significant consideration was given to preparing images and measuring colocalization values throughout the course of these experiments.  All images used for density and colocalization measurements were segmented to exclude any cell edges.  This was done to eliminate the possibility of unusually high density and colocalization values in areas where the sarcolemma was imaged in parallel to the optical axis of the microscope (and therefore was subject to edge effects).  As a result, we are confident that our images are free from biases created by including cell edges in our calculations. During the actual data analysis, two methods of measuring colocalization between 3 proteins were used.  The first method was used to derive all final colocalization values that we believe represent the actual structural organization of atrial myocytes while the second method was used as a control to compare double colocalization values across experiments and to compare atrial and ventricular colocalization values.  In the first method, we calculated what we call ‘true’ double colocalizations, which does not count any triply colocalized voxels that would be included in three separate areas of double colocalization.  In the second method, these triply colocalized voxels are counted towards the three double colocalization numbers as well as being counted as (triple) colocalized on their own.  Using these two distinct methods of calculating colocalization, we were able to accurately report on the structure of the atrial cells we were examining, as well as   71  compare values across experiments and with ventricular myocytes.  Finally, using a novel method to measure the statistical significance of colocalization, we determined what the probability of obtaining each of our colocalization values was by chance, and compared this to our actual recorded values.  This gave us an invaluable tool to analyze the significance of each of the structural groups we identified.      3.3 Structural organization of atrial myocytes The four triple-labeling experiments that have been outlined here examine the localization of RyR, Cav1.2, NCX, CSQ, and Cav-3 and their association with one another in rat atrial myocytes.  Our original hypothesis was that atrial RyR complexes are heterogeneous, and that the different groups may be targeted to specific intracellular domains.  Here, we have presented very convincing data in support of this hypothesis. Overall, we report the presence of eight distinct surface populations of RyR, depending on its binding partners.  These are: 1) RyR with Cav1.2, 2) RyR with CSQ, 3) RyR with NCX, 4) RyR with Cav1.2 and NCX, 5) RyR with CSQ and NCX, 6) RyR with Cav-3, 7) RyR with Cav1.2 and Cav-3, and 8) RyR with NCX and Cav-3.  Importantly, through sophisticated statistical analysis we have shown that the colocalization values we recorded for each of these 8 groups are significantly higher than values that would occur purely by chance. In contrast to atrial cells, rat ventricular myocytes have a very high colocalization of RyR with both Cav1.2 and CSQ, but its association with NCX in these cells is negligible (Scriven et al., 2000).  Similarities between the two cell types are seen however, in the association of Cav-3 proteins with RyR and Cav1.2.  Our observation that a portion of the total Cav1.2 in atrial myocytes colocalizes with Cav-3 agrees with   72  previous studies in both rat and mouse ventricular myocytes (Balijepalli et al., 2006; Scriven et al., 2005).  Interestingly, at the surface of ventricular myocytes specifically, Cav-3 seems to associate equally with RyR and Cav1.2, presumably at the dyads.  In the interior however, Cav-3 colocalizes with a distinct subpopulation of RyR alone (Scriven et al., 2005).  We have shown here that at the surface of atrial myocytes, although there does appear to be a small population of RyR, Cav1.2, and Cav-3 together (~10% of the total surface RyR), there are actually more RyR with Cav-3 alone (~15%).  In order to fully evaluate the differences between atria and ventricle however, the identical triple- labeling experiments that were performed here would have to be done in the ventricle. Then, colocalization values could be compared fairly between the two cell types. In atria, because of the high colocalization of RyR with Cav1.2 and RyR with CSQ, we hypothesize that the 4th and 5th surface populations of RyR listed above (RyR/Cav1.2/NCX and RyR/CSQ/NCX) are likely the same.  This is further supported by the images showing virtually identical distribution patterns of these groups – that is, primarily RyR tightly linked with either Cav1.2 or CSQ and surrounded by large clusters of NCX that occasionally join at their edges to form a triply colocalized group (Figs. 2.4 and 2.6).  The possibility that the 7th RyR surface population listed above (RyR/Cav1.2/Cav-3) is also not a distinct subpopulation, but instead that RyR/Cav1.2/CSQ/NCX/Cav-3 are all contained within one surface group, cannot be ruled out.  This is due to the presence of a significant portion of triply colocalized voxels of RyR, Cav-3, and NCX together (the 8th surface RyR population listed above).  However, we hypothesize that the RyR/Cav1.2/Cav-3, and therefore the RyR/NCX/Cav-3, populations are indeed structurally distinct.  This hypothesis was formed based on the   73  fact that the organization of RyR, NCX, and Cav-3 is clearly different from that of RyR with Cav1.2 or CSQ, and NCX.  Instead of distinct overlapping between the centers of protein clusters, we see large clusters of NCX and Cav-3 coming together primarily at their edges and occasionally coming into close contact with RyR.  Overall, the triply colocalized voxels that do exist between these 3 proteins do not seem to be formed in the same fashion as those in the RyR/Cav1.2/NCX and RyR/CSQ/NCX groups.  However, using two new techniques of statistical analysis, we have shown that all of the triple- colocalization values reported here are highly significant.  The probability of finding three of these proteins within the same voxel purely by chance is very small, and therefore it is highly likely that the triply-colocalized groups we have identified here have come together as a result of specific trafficking to the same subcellular space. Through these experiments, we have also identified additional surface populations of these proteins that do not contain RyR.  These are: 1) Cav1.2 with CSQ, 2) Cav1.2 with NCX, 3) NCX with CSQ, 4) NCX with Cav-3, and 5) Cav-3 with Cav1.2.  However, without further increasing our ‘n’ value, at this point we can only say that the NCX/Cav-3 population is statistically significant with any certainty.  This finding is in agreement with a study showing that NCX1 coprecipitates with Cav-3 in bovine cardiac sarcolemmal vesicles (Bossuyt et al., 2002). For all further discussion, the surface RyR populations that will be considered to be structurally and functionally distinct are: 1) RyR with Cav1.2, 2) RyR with CSQ, 3) RyR with NCX, 4) RyR with CSQ, Cav1.2, and NCX, 5) RyR with Cav-3, 6) RyR with Cav1.2 and Cav-3, and 7) RyR with NCX and Cav-3.   74       3.4 Functional significance of atrial RyR subpopulations One of the biggest differences between atrial and ventricular myocytes that we report here is the variation in the level of interior colocalization of CSQ with RyR.  In atrial cells, this colocalization is significantly less than in ventricles (~45% and 61% respectively), indicating many more central RyR in atria that are not tightly coupled to CSQ.  Under the assumption that CSQ acts as a luminal Ca2+ sensor for, and Po inhibitor of, RyR channels when it is bound (Gyorke et al., 2004; Terentyev et al., 2007), the central RyR in atrial myocytes would be more likely to release a greater amount of Ca2+ than their ventricular counterparts.  Indeed, it has been reported that Ca2+ sparks in atrial cells release up to three times the number of Ca2+ ions as sparks in the ventricle (Woo et al., 2003).  However, reports have also shown differences in the amplitude and frequency of Ca2+ sparks between the periphery and center of atrial myocytes, and we observe here that no such variation in the CSQ with RyR colocalization exists between these two regions.  Therefore, if CSQ is indeed acting as a potent inhibitor of RyR Po, it may be (at least partially) responsible for the difference between atria and ventricle in the number of Ca2+ ions released during a spark, but it is highly unlikely that it is responsible for the variation in Ca2+ spark characteristics from the periphery to the center of atrial myocytes. We have also concluded from these experiments that there is a distinct RyR subpopulation at the surface of atrial myocytes that, of the four other proteins tested for, only associates with CSQ.  Although this population of RyR would not be able to initiate (electrically-invoked) Ca2+ release, it could act as a potential amplifier of the Ca2+ response following initiation from other surface RyR. We have also identified here that approximately 35% of all surface RyR colocalize only with Cav1.2 channels.  Compared to the ~10% that colocalize solely with   75  NCX, it would appear that Cav1.2-mediated CICR is the primary method of initiating Ca2+ release from RyR during EC coupling.  Interestingly, although atrial myocytes are similar to neonatal ventricular cells in their lack of a well-developed TATS, they do not share the same heightened level of RyR/NCX colocalization (Dan et al., 2007).  It seems therefore that atrial cells do not rely on NCX-mediated CICR as neonates do (Huang et al., 2008), which was unclear to this point.  If however, the RyR that colocalize with NCX also closely associate with voltage-gated Na+ channels, then the possibility that these sites function by NCX-mediated CICR becomes much more likely.  Whether or not this is the case has yet to be determined.  Because these sites make up only ~10% of the total surface RyR however, Cav1.2-mediated CICR would still be the dominating form. In the cases where RyR are tightly linked to both Cav1.2 and NCX, although the trigger for CICR is likely to come almost entirely from ICa,L (Sipido et al., 1997), NCX may play an important role in modulating the functionality of these specific groups. Previous published works suggest that if NCX molecules are localized in just the right cellular space, then they are highly likely to alter the EC coupling process (Bers, 2008).  The very close association of RyR, Cav1.2, and NCX in particular surface domains then, implies that NCX would have the capacity to pre-elevate local [Ca2+]i in these dyads and sensitize RyR prior to activation by ICa,L.  What this further implies, is that these sites would potentially be more likely to release Ca2+, and would do so more quickly than surface RyR that are associated with Cav1.2 alone.  By this reasoning, we hypothesize that the sites of triple colocalization between RyR, Cav1.2, and NCX that we have shown form the structural basis for ‘eager sites’ of Ca2+ release at the surface of atrial myocytes.   76  All additional surface RyR populations that we have identified here include caveolin-3 proteins – specifically RyR with Cav-3 alone, RyR with Cav1.2 and Cav-3, and RyR with NCX and Cav-3.  Prior studies have shown that the efficiency of coupling between Ca2+ entry and release is significantly reduced when caveolae are disrupted in rat ventricular myocytes (Calaghan and White, 2006).  This agrees with our data that show a significant portion of RyR and Cav1.2 within very close proximity to Cav-3 – the assumption of course being that the disruption of caveolae damages the functional linkage between RyR and Cav1.2.   Furthermore, disruption of caveolae in rat arterial smooth muscle and neonatal ventricular myocytes – both lacking a well-developed TATS much like atrial cells – has been shown to cause a reduction in the frequency, width, and amplitude of Ca2+ sparks (Lohn et al., 2000).  Most interestingly, this alteration in Ca2+ spark characteristics occurred without modulation of ICa,L and therefore provides additional evidence for the population of RyR colocalized only with Cav-3 that we have identified.  Greater functional implications due to the close association of RyR with caveolae, with or without Cav1.2, are inevitable due to the scaffolding activity of, and regulation of signaling molecules by, caveolin-3 (Cohen et al., 2004).  For example, in the case of the RyR/NCX/Cav-3 subpopulation, we must assume that there are multiple influences on the actions of RyR and NCX from other molecules localized to the caveolae.  However, without a comprehensive listing of the proteins within these particular caveolae (because as we and others have shown, (Balijepalli et al., 2006; Scriven et al., 2005) the inhabitants and/or neighbours of individual caveolae can vary greatly), we are unable to accurately predict the functional consequences of this, and the RyR/Cav-3 and RyR/Cav1.2/Cav-3, populations.  We can say with some certainty   77  however, that RyR alone or with either NCX or Cav1.2, are likely being specifically directed either near to or within specific populations of the caveolae in atrial myocytes. Furthermore, because of increasing evidence on the role of caveolae in the regulation of EC coupling (Balijepalli et al., 2006; Bossuyt et al., 2002; Calaghan and White, 2006; Scriven et al., 2005), these associations are unlikely to be trivial.      3.5 Directions of future study Currently, we are unable to say with confidence whether or not the surface RyR colocalized only with NCX are likely subject to NCX-mediated CICR.  In order to come to a conclusion, we need to analyze the distribution of voltage-gated Na+ channels with respect to NCX and RyR using the triple-labeling protocols that we have outlined here.  If these channels colocalize with NCX and RyR, it is much more likely that reverse-mode conductance of NCX is causing CICR from RyR in regions where they associate closely. At this time, although we are confident that one of the atrial subpopulations that do not contain RyR (NCX/Cav-3) has a colocalization value significantly greater than expected by chance, we cannot yet say with confidence whether or not the Cav1.2/CSQ, Cav1.2/NCX, NCX/CSQ, and Cav-3/Cav1.2 subpopulations are statistically significant. In order to determine if the associations between these proteins are likely to have occurred purely by chance, we will need to increase our ‘n’ value for each experiment, and re-conduct statistical testing on the recorded colocalization values. Finally, understanding the changes in molecular architecture that occur in certain pathological states is of utmost importance.  In atrial fibrillation for example, the particular trigger is often unknown, although it likely involves mechanical abnormalities of contraction, or the linking between electrical impulses and contraction of the cell (i.e.   78  EC coupling) (Waktare and Camm, 1999).  A multitude of animal models exist for this disease in particular (for a review see Nattel, 2005), and performing the same triple- labeling experiments that we have done here in one of the pathological models could give us great insight into the molecular mechanisms behind the disease.  Ultimately, the goal would be to understand the differences between the normal and pathological states so that more effective progress could be made in treating the pathology.      3.6 Summary The distribution of some of the major proteins involved in excitation-contraction coupling of rat atrial myocytes was elucidated using novel triple-labeling immunofluorescence techniques.  We have provided abundant evidence to support our hypothesis that atrial RyR complexes are heterogeneous and may be targeted to specific intracellular domains.  In particular, we have identified seven surface populations of RyR, based on its binding partners, that we believe are structurally and functionally distinct.  These are: • RyR with Cav1.2 • RyR with CSQ • RyR with NCX • RyR with CSQ, Cav1.2, and NCX • RyR with Cav-3 • RyR with Cav1.2 and Cav-3 • RyR with NCX and Cav-3 From the identification of these groups, we have come to a number of conclusions.  First, Cav1.2-mediated CICR is the dominant form in atrial myocytes.  The   79  structural basis for NCX-mediated CICR is present in these cells, but the localization of voltage-gated Na+ channels with respect to RyR and NCX needs to be determined before any concrete conclusions can be drawn.  Second, the discovery of a distinct population of RyR, CSQ, Cav1.2, and NCX tightly coupled together provides a very likely candidate for the structural basis of eager sites present at the periphery of atrial myocytes.  Third and finally, the characteristics of atrial Ca2+ sparks and the efficiency of EC coupling in certain regions of atrial myocytes may be controlled by caveolae that contain or are neighboring to RyR with or without Cav1.2 or NCX. One of the main principles of the local control theory of EC coupling – that all RyR are functionally equivalent – is significantly challenged by the discovery of the array of distinct RyR complexes shown here.  While each individual RyR complex may be independently controlled, their molecular constituents and binding partners are unlikely to be identical.                   80  3.7 References  Balijepalli, R. C., Foell, J. D., Hall, D. D., Hell, J. W. and Kamp, T. J. (2006). Localization of cardiac L-type Ca(2+) channels to a caveolar macromolecular signaling complex is required for beta(2)-adrenergic regulation. Proc Natl Acad Sci U S A 103, 7500-5.  Bers, D. M. (2008). Calcium Cycling and Signaling in Cardiac Myocytes. Annu. Rev. Physiol. 70, 23-49.  Bossuyt, J., Taylor, B. E., James-Kracke, M. and Hale, C. C. (2002). The cardiac sodium-calcium exchanger associates with caveolin-3. Ann N Y Acad Sci 976, 197-204.  Calaghan, S. and White, E. (2006). Caveolae modulate excitation-contraction coupling and beta2-adrenergic signalling in adult rat ventricular myocytes. Cardiovasc Res 69, 816-24.  Cohen, A. W., Hnasko, R., Schubert, W. and Lisanti, M. P. (2004). Role of caveolae and caveolins in health and disease. Physiol Rev 84, 1341-79.  Dan, P., Lin, E., Huang, J., Biln, P. and Tibbits, G. F. (2007). Three- dimensional distribution of cardiac Na+-Ca2+ exchanger and ryanodine receptor during development. Biophys J 93, 2504-18.  Gyorke, I., Hester, N., Jones, L. R. and Gyorke, S. (2004). The role of calsequestrin, triadin, and junctin in conferring cardiac ryanodine receptor responsiveness to luminal calcium. Biophys J 86, 2121-8.  Harlow, E. and Lane, D. (1988). Antibody-Antigen Interactions. In Antibodies: a laboratory manual. Cold Spring Harbour, NY: Cold Spring Harbour Laboratory.  Huang, J., Hove-Madsen, L. and Tibbits, G. F. (2008). Ontogeny of Ca2+- induced Ca2+ release in rabbit ventricular myocytes. Am J Physiol Cell Physiol 294, C516-25.  Lohn, M., Furstenau, M., Sagach, V., Elger, M., Schulze, W., Luft, F. C., Haller, H. and Gollasch, M. (2000). Ignition of calcium sparks in arterial and cardiac muscle through caveolae. Circ Res 87, 1034-9.  Scriven, D. R. L., Dan, P. and Moore, E. D. W. (2000). Distribution of Proteins Implicated in Excitation-Contraction Coupling in Rat Ventricular Myocytes. Biophysical Journal 79, 2682-2691.  Scriven, D. R. L., Klimek, A., Asghari, P., Bellve, K. and Moore, E. D. W. (2005). Caveolin-3 is Adjacent to a group of Extradyadic Ryanodine Receptors. Biophysical Journal 89, 1893-1901.  Sipido, K. R., Maes, M. and Van de Werf, F. (1997). Low efficiency of Ca2+ entry through the Na(+)-Ca2+ exchanger as trigger for Ca2+ release from the sarcoplasmic reticulum. A comparison between L-type Ca2+ current and reverse-mode Na(+)-Ca2+ exchange. Circ Res 81, 1034-44.  Terentyev, D., Viatchenko-Karpinski, S., Vedamoorthyrao, S., Oduru, S., Gyorke, I., Williams, S. C. and Gyorke, S. (2007). Protein protein interactions between triadin and calsequestrin are involved in modulation of sarcoplasmic reticulum calcium release in cardiac myocytes. J Physiol 583, 71-80.  Waktare, J. and Camm, J. A. (1999). In Atrial Fibrillation, pp. 4-6: Informa Health Care.   81   Woo, S. H., Cleemann, L. and Morad, M. (2003). Spatiotemporal characteristics of junctional and nonjunctional focal Ca2+ release in rat atrial myocytes. Circ Res 92, e1-11.                                            82  APPENDICES Appendix 1.  Composition of Solutions  A. Live myocyte isolation solutions   Chemical Amount  Joklik MEM 5.62 g NaHCO3 1 g MgSO4 72 mg DL-Carnitine 99 mg   Joklik Solution dH2O to 500mL Gas for 30 min. with 95% O2-5% CO2 at 37οC.  pH 7.4 (37οC).  Chemical  Amount Fatty-acid free BSA  50 mg Collagenase Type II (270units/mg) 50 mg Joklik Solution  50 mL    Enzyme Solution 100mM CaCl2  12.5 μL           83  Chemical  Amount Hepes  0.298 g Fatty-acid free BSA  0.5 g Insulin (30μg/mL M199)  1.0 mL DL-Carnitine  10 mg Penicillin G  10 mg Streptomycin Sulfate  10 mg EGTA  38 mg     Media-199 Complete      M-199  50 mL Add media in laminar flow hood.  pH 7.4 (4οC).  Sterile filter into tissue culture flask and keep at 37οC.     B. Solutions for fixing and labeling cells  Chemical  Amount (mM) KCl  27 KH2PO4  15 NaCl  1370  PBS 10 X      Na2HPO4•7H2O  80 pH 7.4 (after dilution to 1 X)       84  Chemical  Amount (mM) NaCl  3014  SSC 20 X Na3Citrate  350 pH 7.2  Chemical Amount  BSA 0.5 g  Na Azide 0.01 g  Normal Goat Serum 1.0 mL  Triton X-100  25 μL  SSC 20 X  2.5 mL   Antibody Buffer dH2O to 50 mL          Chemical        Amount  SSC 20 X 5 mL  Triton X-100 50 μL   Antibody Wash dH2O to 100 mL    Chemical  Amount (mM) Glycine Buffer Glycine 100 pH to 7.4    85  Chemical Amount  PBS 10 X 10 mL Triton X-100 100 μL   Permeabilization Solution dH2O to 100 mL   Chemical  Amount Paraformaldehyde 2 g   Paraformaldehyde Fixative PBS 1 X to 100 mL Heat to 65οC.  When solution clears, cool to room temp. and filter.  pH 7.4  Chemical Amount  1,4-Diazabicyclo[2.2.2]Octane (DABCO) 2.5 g PBS 10 X 10 mL NaN3 0.02 g     DABCO Mounting Medium Glycerol to 100 mL Dissolve DABCO in PBS 10 X.  Add NaN3 and Glycerol. Stir until clear. Store away from light.                86  C. SDS-PAGE/Western blot solutions    Chemical Amount  Tris Base 30.3 g Glycine 144.1 g  10 X Transfer Buffer dH2O to 1 L pH 8.3   Chemical Amount  10 X transfer buffer 100 mL Methanol (MeOH) 100 mL  1 X Transfer Buffer dH2O to 1 L    Chemical Amount  Tris Base 24.2 g NaCl 80.0 g  10 X TBS dH2O to 1 L pH 7.6           87  Chemical Amount 10 X TBS 50 mL Tween 20 5 mL  TBST dH2O to 500 mL   Chemical Amount  Tris Base 30.3 g Glycine 144 g Sodium dodecyl sulfate (SDS) 10 g   10 X Running/Tank Buffer dH2O to 1 L    Chemical Amount  30% Acrylamide: Bis Acrylamide (37.5:1) 6 mL dH2O  7.5 mL 4X running gel buffer 4.5 mL 10% ammonium persulfate (APS) 150 μL   10% Running Gel Temed 10 μL            88   Chemical Amount  30% Acrylamide: Bis Acrylamide (37.5:1) 0.75 mL dH2O 3 mL 4X stacking gel buffer 1.25 mL 10% APS 50 μL   4% Stacking Gel Temed 10 μL    Chemical Amount  Tris Base 91 g (1.5 M) SDS 2 g  4 X Running gel buffer dH2O to 500 mL pH 8.8   Chemical Amount  Tris Base 30.25 g (0.5 M) SDS 2 g  4 X Stacking gel buffer dH2O to 500 mL pH 6.8    89  Appendix 2.  List of antibodies used   NAME COMPANY CATALOG # MONO- OR POLY- CLONAL CLASS    [ ] (if spec- ified) IF DILUTION  WB DILUTION  WORKS WITH ZENON KIT? ZENON DILUTION (in 20 μL Ab buffer) NOTES α- actinin SIGMA A 7811 Mono IgG1 21 mg/mL 1in1800 ND ND ND None Cav-3 BD Biosciences 610420/ 610421 Mono IgG1 250 μg/mL 1in7.5 1in5000 Y 2 μL None CNC1 Gift of Dr. William Catterall  Poly  0.541 mg/mL 1in1000 ND ND ND Block cells for 1h with 4% BSA prior to staining CSQ ABR PA1-913 Poly IgG 0.2 mg/mL 1in40 1in2500 ND ND None NCX Swant R3F1 Mono IgG1  1in1000 ND Y 0.25 μL None RyR ABR MA3-916 Mono IgG1 1 μg/μL 1in100 ND Y 1.5 μL None SERCA 2 ABR MA3-910 Mono IgG1  1in10  Y 4 μL Stain O/N at room temp. on shaker Vinculin Sigma V4505 Mono IgG1 2 mg/mL 1in15 1in75 ND ND None  90  Appendix 3. List of antibodies that do not work for immunofluorescence of rat cardiomyocytes     NAME COMPANY CATALOG # LTCC (alpha 1 subunit)  EMD/Calbiochem 684507 LTCC (alpha 1 subunit)  SIGMA C1603 LTCC (alpha 1 subunit)  Millipore AB5412 LTCC (alpha 1 subunit)  NeuroMab 75-053 LTCC (alpha 2 subunit)  Abcam Ab2864 LTCC (Beta 2 subunit)  Millipore AB5787 CSQ  Millipore 06-382 CSQ  SIGMA C2491 CSQ  Abcam Ab32801 NCX  RDI RDI-NACAXABM  91  Appendix 4. Protocols and control experiments for dual- and triple-labeling of atrial myocytes   Types of Antibody involved in experiment: Protocol: Control Experiment: 1 polyclonal and 1 monoclonal *  • 1st 1° Ab - O/N at 4°C • 1st 2° Ab - 1.5h at RT • 2nd 1° Ab - O/N at 4°C • 2nd 2° Ab - 1.5h at RT 1) Cells are labeled with 2° antibody alone 2) Cells are incubated with a polyclonal antibody and anti-mouse 2° antibody and vice versa. 2 monoclonals  Both monoclonals incubated with Zenon labeling kit components separately as per Molecular Probes protocol. Labeled antibodies combined and incubated with cells for 1h at room temp. Perform experiment with 2 monoclonals that are not localized to same area in cell. Ex. 1 monoclonal is nuclear pore complex antibody conjugated to Alexa488 and other is RyR conjugated to Alexa594 - both using Zenon labeling kit. Should not see nuclear pore labeling in 594 wavelength or Z-line staining in 488 wavelength. 1 polyclonal and 2 monoclonals  • Polyclonal antibody- O/N at 4°C • 2° antibody- 1.5h at room temp. • 2 monoclonals prepared using Zenon labeling kit and incubated with cells for 1h at room temp. 1) Same control as for 2 monoclonals 2) Add unblocked anti-mouse Fab fragment conjugated to a fluorophore (Zenon kit component ‘A’) to unlabeled polyclonal 1° antibody.  Should see no labeling.  1 polyclonal and 2 monoclonals • 1st monoclonal Ab - O/N at 4°C • Anti-mouse 2°Ab -1.5h at RT • 2nd monoclonal Ab- prepared with Zenon labeling kit and incubated with cells for 1h at RT • Polyclonal Ab- O/N at 4°C • Anti-goat 2° Ab- 1.5h at RT  1) Perform experiment with 2 monoclonals that are not localized to same area in cell. Ex. For 1st monoclonal Ab, use RyR and Alexa 594 goat anti-mouse 2° Ab.  For 2nd monoclonal Ab, conjugate nuclear pore complex Ab to Alexa 488 using a Zenon labeling kit.  Should not see nuclear pore labeling in 594 wavelength or Z-line staining in 488 wavelength. 2) Add unblocked anti-mouse Fab fragment conjugated to a fluorophore (Zenon kit component ‘A’) to unlabeled polyclonal 1° Ab. Should see no labeling.     92  3 monoclonals* • 1st monoclonal Ab- O/N at 4°C • Anti-mouse 2° Ab - 1.5h at RT • 2 monoclonals prepared using Zenon labeling kit and incubated with cells for 1h at room temp. 1) Same as with 2 monoclonals prepared with Zenon labeling kits 2) Same as control 1) for 1 polyclonal and 2 monoclonals protocol. *order in which antibodies are added must be determined by experimenter                                       93  Appendix 5.  Figures from control experiments  In order to ensure that the secondary antibodies we used were not capable of producing specific labeling on their own, we incubated atrial myocytes with a secondary antibody alone for 1.5 hours at room temperature at the same concentrations used in all other experiments.  In Fig.A5.1 A, an atrial cell labeled with a goat anti-mouse Alexa350 secondary antibody is shown.  The diffuse, non-specific labeling in this raw (not deconvolved) image shows that the secondary antibody alone is not capable of producing specific labeling in these cells.  The same protocol was used to test all other secondary antibodies used in our experiments, and yielded the same result (images not shown).  As a comparison, a raw image of an atrial myocyte single-labeled with RyR is shown in B. The bright, specific labeling along the Z-lines of the cell provides significant contrast to the faint, non-specific labeling of the secondary antibody alone in A.  An additional control was performed following our triple-labeling experiments to help ensure the validity of our results.  In Fig.A5.2 we show all combinations of two proteins together that occur in multiple experiments (e.g. RyR and NCX are used together in 3 out of 4 experiments).  The double colocalization values for these proteins are measured and compared across experiments to make sure no significant differences are present.  The same control was performed measuring the densities of individual proteins across experiments and the results are displayed in the manuscript.      94   Fig. A5.1 A Atrial myocyte labeled with secondary antibody only (goat anti-mouse Alexa350).  B RyR single-labeling of an atrial myocyte.  Both A and B are raw (not deconvolved) images.  N= nucleus. Scale b is 5 μm.  ar                           95    0 5 10 15 20 25 30 35 40 45 50 RyR with NCX NCX with RyR %  C ol oc al iz at io n A  Fig. A5.2 A The double colocalization values for RyR with NCX and NCX with RyR are compared across triple-labeling experiments.  The same is shown for RyR and Cav1.2 (B), and RyR and Cav-3 (C). No significant differences are observed within each group (p<0.05).   0 10 20 30 40 50 60 70 RyR with Cav1.2 Cav1.2 with RyR %  C ol oc al iz at io n B  0 10 20 30 40 50 60 RyR with Cav-3 Cav-3 with RyR %  C ol oc al iz at io n C     96  Appendix 6. Cav1.2 antibody production  A6.I Rationale A multitude of commercially produced calcium channel antibodies have been tried in our laboratory but have failed to consistently produce high-quality results (those tried are listed in appendix 3).  One effective antibody had been used previously, but was a gift, and at the onset of my research was no longer available.  As this antibody was essential to completion of my project, we designed a protocol to produce our own polyclonal antibody directed against the pore-forming α-1 subunit of the Cav1.2 calcium channel. A6.2 Protocol The specific protocol for preparation of this antibody can be found in Gordon et al. (1987).  Briefly, a synthetic peptide against residues 821-838 (KTTKINMDDLQPSENEDKS) was ordered from the Peptide Synthesis Facility at the Brain Research Centre, UBC.  This was the sequence used to produce the antibody given to us as a gift.  In addition, a lysine residue was added at position 1 of the sequence for coupling purposes.  A portion of the synthetic peptide was coupled to BSA using the manufacturer’s protocol for a HOOK peptide coupling kit (G Biosciences, Maryland Heights, MO).  Two female New Zealand white rabbits were immunized: one with peptide alone (M1), and one with the BSA-coupled peptide (M2).  The primary injection was done on Aug.1st, 2007.  Each animal was injected with 100 μg of peptide in a mixture of 1/3 PBS 1X, 2/3 complete freund’s adjuvant (CFA) to a total of 1mL.  An emulsion was made by vortexing each mixture for several minutes, after which each solution was transferred to a glass syringe in preparation for injection.  Prior to the 1st   97  injection, a blood sample was taken from each animal for analysis of their ‘pre-immune’ serum (PIS).  Subsequent to the 1st injection, a booster shot was given to each animal two weeks later.  The booster injection contained 100 μg of peptide in a mixture of 1/3 PBS 1X, 2/3 incomplete freund’s adjuvant (IFA) to a total of 1mL.  10 days post-injection, a small sample (~15mL) of each animal’s blood was collected.  Following this, a booster shot was given every 4-6 weeks, and a blood sample taken 10 days post-injection.  Each blood sample was centrifuged repeatedly at 2500 RPM for 5 minutes and the serum collected following each spin, until maximum yield was reached. For each serum sample, an SDS-PAGE was run with homogenized whole-heart (rat) tissue lysate (~25 μg of protein per well) on a 10% running gel at 130V for ~1 hour or until proteins reached within 1cm of bottom of gel.  Two lanes per gel were loaded with a protein ladder standard (Bio-Rad, Hercules, CA) as a control, and for molecular weight (MW) measurements during analysis.  Gels were then transferred to a Polyvinylidene fluoride (PVDF) membrane using a wet transfer system overnight (O/N) at 23V on ice.  Following transfer, membranes were incubated for ~3-5 min. with a 0.1%(w/v) ponceau-S in 5%(v/v) glacial acetic acid solution to view proteins and ensure complete transfer.  Membranes were rinsed repeatedly with dH2O to remove ponceau-S stain, and then washed 3 X 15 min. with tris-buffered saline containing 10% tween-20 (TBS-T).  After washing, membranes were blocked with 5% milk containing 0.025% NaN3 for 1 hour at room temperature (RT) on a shaker.  Following milk block, membranes were split and incubated with either raw rabbit serum or a control antibody, each diluted in 5% milk containing 0.025% NaN3 for 2.5 hours at RT on a shaker. Membranes were washed 3 X15 min. with TBS-T and subsequently incubated with either   98  a goat anti-rabbit or goat anti-mouse secondary antibody conjugated to horseradish peroxidase (HRP) diluted to an appropriate concentration in 5% milk solution.  A final wash was conducted 3 X 15 min. in TBS-T and membranes were then incubated with chemiluminescent substrate (Pierce, Rockford, IL) for detection of HRP.  Finally, membranes were exposed for varying lengths of time and images captured on Kodak scientific imaging film (Kodak, Rochester, NY). Once an appropriate level of Cav1.2 was detected in the raw rabbit serum by western blot, we tested the serum using immunofluorescence on rat ventricular myocytes (for detailed protocol, see materials and methods).  In order to purify the antibody, we used the manufacturer’s protocol for an AminoLink plus immobilization kit (Pierce). Briefly, the peptide used to immunize the animals was covalently attached to a beaded agarose support within an affinity purification column.  The raw rabbit serum was allowed to mix with the column components for approximately 1 hour.  Multiple washing and eluting steps were then performed on the column using a spin-purification method. A sample was collected from the column following each wash and elution step, and each sample tested by western blot using the same protocol as on the raw serum. A6.3 Results      A6.3.I Pre-Immune Serum (PIS) In order to see the potential development of our antibody over time, we compared all post-immunization results with each animal’s PIS.  To view any pre-existing antibodies in the raw serum of animals M1 and M2, we probed whole rat heart tissue lysate with various dilutions of the raw PIS of each animal. Fig.A6.1A is an example of this.  Here, a membrane has been split into 3, with each lane incubated with M1 PIS at different dilutions: 1in10, 1in50, and 1in100, and exposed overnight.  For each dilution,   99  we see no specific labeling, indicating that the PIS is clear of any antibodies against rat heart lysate.  For dilutions 1in10 and 1in50 however, a lot of non-specific labeling is seen, and therefore, a dilution of 1in100 was chosen to perform all subsequent experiments with M1 post-immune serum.  Fig.A6.1 B is a verification of the dilution test, with only a 1in100 dilution tested (left panel).  On the same membrane, one lane of heart lysate has been incubated with a primary antibody against calsequestrin (~55 kilodaltons (kDa)) as a control (right panel).  A similar experiment was performed testing the M2 PIS at various dilutions, and a dilution of 1in100 was also chosen to perform all post-immune serum tests for this animal.  Fig. A6.1 C shows a membrane that has been split and two lanes of whole heart lysate incubated with PIS from animal M2 (left panel) and two lanes with a primary antibody against vinculin (cytoskeletal protein; ~117 kDa) as a control (right panel).  There is also no specific binding of the PIS from this animal, indicating that this serum is also free of antibodies against rat heart.  Note especially the lack of binding in the range of 150kDa and above, which is where we would expect to see specific binding against a Cav1.2 Ab (165-195 kDa; (Perez-Reyes and Schneider, 1995)). It is important to note that there are no reports in the literature of any breakdown products of the α-1 subunit that produce specific bands at other molecular weights under reducing or non-reducing conditions.     A6.3.2 M1 The results of the 1st and 2nd bleeds from animal M1 are shown in Fig. A6.2 A and B respectively.  Each membrane has been split in two, and both sides have two lanes of rat heart lysate probed with either an anti-vinculin antibody (control; right panels), or the raw rabbit serum diluted 1in100 (left panels).  For each bleed, it appears that some specific antibodies against the rat heart lysate may have developed, and although they   100  appear to have grown stronger in bleed 2, there is no binding in the 165-195 kDa range. The appearance of some specific antibodies that were not present in the PIS is likely a result of a general increase in the immune system activity of the animal as a result of immunization with adjuvant.  Animal M1 was injected with 6 booster shots total, and a blood sample was taken and tested for any sign of specific Cav1.2 antibodies each time. None of the subsequent bleeds showed any sign of a band in the 165-195 kDa range, and all further research on M1 was ended at that point.      A6.3.3 M2 The results from the 1st and 3rd bleeds from animal M2 are shown in Fig. A6.3 A and B respectively.  Each membrane in A and B has been split in two, and both sides have two lanes of rat heart lysate probed with either an anti-vinculin antibody (control; right panels), or the raw rabbit serum diluted 1in100 (left panels).  In each case, many specific bands are present, but most important is the presence of a specific band at ~200kDa.  The film in A was exposed overnight, while in B it was only exposed for 30 minutes.  Therefore, it seems that with more booster shots to animal M2, the band at ~165-195kDa got stronger in intensity.  Because of its molecular weight, and the fact that it appeared following immunization of the animal, it seems likely that this band represents an anti-Cav1.2 antibody that is present in the raw serum of the rabbit.  This experiment was repeated for bleeds 2,4,5,and 6.  The serum from each of these bleeds showed the presence of this same specific band, growing in intensity with later bleeds. As with animal M1, later bleeds that were tested also showed the presence of specific bands at other MWs with very high intensity.  This is likely as a result of a general increase in the immune system activity of the animal following immunization.  Once we detected a specific band at MW 165-195 kDa, the raw serum was tested using   101  immunofluorescence on rat ventricular myocytes.  Various dilutions were tested from different bleeds, but labeling was highly non-specific due to the presence of other antibodies in the raw serum (data not shown).  Therefore, an attempt to purify the antibody was made.     A6.3.4 Antibody purification An example of the samples produced from the affinity purification column is shown in Fig.A6.4 A and B.  In A6.4 A, one membrane was split and incubated either vinculin (Vin) as a control or with an undiluted portion of the 1st elution fraction (EF I) produced from the column.  In EFI, we see that a band is still present at the desired MW (165-195kDa).  However, we also see bands at other MWs, just like in the raw serum.  In Fig. A6.4 B, elution fractions II and III (EF II and III) are shown.  Each of these elutions produced no specific bands at our desired MW, but did express one very intense band at a much lower MW.  As mentioned previously, no breakdown products have been reported of the α-1 subunit of this channel, and therefore we believe this band to be completely unrelated to our antibody.  EF I was tested using immunofluorescence on rat ventricular myocytes, and again, very non-specific binding of the cells was observed, indicating that other antibodies were still present in the purified serum.  A number of adjustments were made to the antibody purification protocol, such as coupling the peptide to the column using a coupling buffer with a more basic pH, and using a less vigorous gravity-flow method to collect the column samples instead of a spin-purification method.  Samples produced using these protocols and others gave similar results from our original experiment (results not shown), and therefore at this point, the experiment was stopped.   102  A6.4 Discussion After more than one year of work in attempting to produce this antibody, we believe that we collected raw serum from one animal (M2) that contained a fairly high Cav1.2 antibody concentration due to the presence of a very intense specific band at MW 165-195 kDa.  However, the clear presence of other non-specific antibodies in the raw serum, and the very non-specific staining we observed using immunofluorescence to test this serum on rat ventricular myocytes indicated the importance of affinity purifying the antibody to produce good results.  Multiple protocols were attempted and fellow researchers consulted in the hopes of improving our yield.  Due to the presence of very intense specific bands at other MWs when testing the ‘purified’ serum by western blot, we have determined that either i) the specific band at ~200 kDa in our raw serum was not in fact a Cav1.2 antibody, but another antibody binding to a rat heart protein at this MW, ii) the Cav1.2 antibody in our raw serum samples did not effectively bind to the peptide in the column, or did not elute properly, or iii) other antibodies in the raw rabbit serum also bound to and eluted from our affinity purification column with our Cav1.2 antibody.  Due to time constraints, further purification of the raw rabbit serum was not attempted. However, we believe that the raw serum we have stored does contain Cav1.2 antibodies and with the proper purification protocol, they could be isolated from the serum.        103  A6.5 Figures     Fig. A6.1 Pre-immune serum from animals M1 (A and B) and M2 (C).  In A, the pre-immune serum from M1 was tested at multiple dilutions (from left to right: 1in10, 1in50, 1in100) to determine which dilution produced a clear background.  Beginning with a clear background was a control to ensure that any bands produced in the post-immune serum were strictly as a result of the animal’s immunization.  B A second test of M1 pre- immune serum at a dilution of 1in100 (left panel) on the same membrane as a control (CSQ, right panel; ~55kDa).  C M2 pre-immune serum (left panel) tested at a dilution of 1in100 on the same membrane as a control (Vinculin, right panel; ~117 kDa).  Arrows represent MW markers from protein ladder standard.                   104   Fig. A6.2 Raw serum from the 1st (A) and 2nd (B) bleeds from M1.  In A and B, the membranes have been split and probed with either the raw serum of M1 at a dilution of 1in100 (left panels) or vinculin as a control (right panels; ~117kDa).  Neither bleed shows the presence of a specific band between 165-195 kDa where we would expect to see an anti- Cav1.2 antibody bind.                          105      Figure A6.3 Raw serum from the 1st (A) and 3rd (B) bleeds from M2.  In A and B, the membranes have been split and probed with either the raw serum of M2 at a dilution of 1in100 (left panels) or vinculin (right panels; ~117kDa).  Both bleeds show the presence of a specific band at ~165-195 kDa which may indicate the presence of an anti-Cav1.2 antibody in the raw serum of M2.                    106         Figure A6.4 Example of elution fractions from an antibody purification column.  In A, the membrane was split and probed with either an anti-vinculin antibody (Vin; left panel) or the 1st elution fraction (EFI) from an affinity purification column.  EF I still shows a faint band at ~165-195 kDa, but also shows very strong, specific bands at other MWs.  In B, the 2nd and 3rd elutions fractions (EF II and III respectively) from the same column are shown. These EFs only exhibit one very strong band at ~35 kDa.  Therefore, the protein that appeared at ~165-195 kDa was likely not purified by the column to any significant degree.                107  A6.6 References Gordon, D., Merrick, D., Auld, V., Dunn, R., Goldin, A.L., Davidson, N., and Catterall, W.A. (1987). Tissue-specific expression of the RI and RII sodium channel subtypes. Proc Natl Acad Sci U S A 84, 8682-6.  Perez-Reyes, E. and Schneider, T. (1995). Molecular biology of calcium  channels. Kidney Int 48, 1111-24.      108

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0066907/manifest

Comment

Related Items