UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structural and biochemical insights into the cardiac and skeletal muscle excitation-contraction coupling… Wong King Yuen, Siobhan Meagan 2018

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_2018_september_wong_king_yuen_siobhan.pdf [ 11.39MB ]
Metadata
JSON: 24-1.0368619.json
JSON-LD: 24-1.0368619-ld.json
RDF/XML (Pretty): 24-1.0368619-rdf.xml
RDF/JSON: 24-1.0368619-rdf.json
Turtle: 24-1.0368619-turtle.txt
N-Triples: 24-1.0368619-rdf-ntriples.txt
Original Record: 24-1.0368619-source.json
Full Text
24-1.0368619-fulltext.txt
Citation
24-1.0368619.ris

Full Text

STRUCTURAL AND BIOCHEMICAL INSIGHTS INTO THE CARDIAC AND SKELETAL MUSCLE EXCITATION-CONTRACTION COUPLING MACHINERY by  Siobhan Meagan Wong King Yuen  B.Sc. Hons., The University of British Columbia, 2013  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Biochemistry and Molecular Biology)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  June 2018  © Siobhan Meagan Wong King Yuen, 2018  ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled:  Structural and biochemical insights into the cardiac and skeletal muscle excitation-contraction coupling  machinery  submitted by Siobhan Meagan Wong King Yuen  in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biochemistry and Molecular Biology  Examining Committee: Filip Van Petegem, Biochemistry and Molecular Biology Supervisor  Natalie Strynadka, Biochemistry and Molecular Biology Supervisory Committee Member  Edwin Moore, Cellular and Physiological Sciences Supervisory Committee Member Calvin Yip, Biochemistry and Molecular Biology University Examiner Hakima Moukhles, Cellular and Physiological Sciences University Examiner   Additional Supervisory Committee Members:  Supervisory Committee Member  Supervisory Committee Member  iii  Abstract               Excitation-contraction (EC) coupling describes the process whereby the depolarizing action potential is transduced into a rapid increase of cytosolic calcium (Ca2+) that initiates muscle contraction. Proper execution of EC coupling relies on the coordinated communication between two calcium channels: plasma membrane-bound, L-type voltage-gated calcium channels (CaVs) and the intracellular Ryanodine Receptors (RyRs). CaVs respond to membrane depolarization by conveying an intracellular signal to the RyR. In skeletal muscle, CaV1.1 mechanically couples to the RyR; in cardiac tissue, extracellular Ca2+ entry via CaVs trigger RyR opening. The net effect of RyR activation is elevation of intracellular Ca2+ levels, activating the contractile machinery. In skeletal muscle, the nature of the physical CaV-RyR coupling has been an area of intense interest: do the channels directly interact or are auxiliary proteins required? Recently, a novel adaptor protein, STAC3, has been identified as playing a role in trafficking and maintaining components of the EC coupling machinery in a functional state. Indeed, STAC3-null mice and fish exhibit failure of skeletal muscle EC coupling. Chapter 2 presents x-ray crystallographic and isothermal titration calorimetry (ITC) data showing a direct interaction between STAC3 and CaV1.1. EC coupling assays reveal the importance of this interaction in EC coupling. The CaV1.1-STAC3 interaction is perturbed by the Native American Myopathy STAC3 mutation.                L-type voltage-gated calcium channels fulfill dual roles as voltage-sensors for EC coupling and calcium ion conduits. In non-muscle cells, STAC3 facilitates CaV1.1’s functional membrane expression and alters the current properties of CaV1.2, suggesting a role of STAC proteins as a CaV regulator. Chapter 3 presents electrophysiology data illustrating the significant effect of STAC3 on modulating CaV1.2 currents. Detection of an interaction to Calmodulin iv  (CaM), a well-known CaV regulator, suggests that STAC proteins may exert its effect on ion conduction via CaM. Genetic defects in the EC coupling machinery underlie numerous congenital myopathies and life-threatening cardiac arrhythmias. Chapter 4 explores the implications of disease-associated mutations within the cardiac Ryanodine Receptor (RyR2) using structural, spectroscopic, and thermal stability assays. An anion binding site within the N-terminal RyR2 region and maintenance of proper domain interfaces are key to RyR2 stability and normal functioning.     v  Lay Summary Skeletal and cardiac muscle contraction is the product of an intricate process termed Excitation-Contraction (EC) Coupling, converting an electrical stimulus (“excitation”) into muscle contraction. Two calcium channels, pores within cellular membranes which conduct calcium ions, are fundamental to EC coupling. However, the identity of novel proteins involved in this process are emerging, enriching our understanding of EC coupling. Once an electrical stimulus reaches the muscle fiber, it alters the voltage across the cell membrane, which is sensed by voltage-sensitive calcium channels (CaVs). Consequently, CaVs activate an intracellular calcium channel, the Ryanodine Receptor (RyR). Finally, the RyR releases Ca2+ into the fiber, triggering muscle contraction. This thesis uses an interdisciplinary approach – x-ray crystallography, calorimetry, and electrophysiology – to explore the communication between CaVs and RyRs. We identify a crucial role of a muscle-specific adaptor protein, STAC3, in EC coupling. The structural implications of disease-associated mutations within STAC3 and RyR are examined.  vi  Preface This thesis is composed of sections that has been adapted from original publications or that are currently in preparation for publication in peer-review journals.  Chapter 2. A version of this material has been published as Wong King Yuen, S.M., Campiglio, M., Tung, C.C., Flucher, B.E., and Van Petegem, F. Structural insights into binding of STAC proteins to voltage-gated calcium channels. Proc. Natl. Acad. Sci. U.S.A (2017) 114(45): E9520-E9528. In this work, I was principally involved in cloning, expressing, and purifying the constructs used for x-ray crystallography and isothermal titration calorimetry. Jett Tung generously dedicated his time to aid in the cloning and purification of the constructs listed in Figure 2.5, in addition to preparing crystallographic screens for the STAC2 Q347I construct. All the remaining crystallographic screens discussed were prepared by me. I obtained the diffraction data of the structures presented in this chapter. Apart from the structure of the STAC2-CaV1.1 complex, which was solved by Dr. Filip Van Petegem, I solved all other structures. All ITC data presented in this chapter was obtained and analyzed by me. The immunostaining and EC coupling assays were performed and data analyzed by Dr. Marta Campiglio and Dr. Bernhard Flucher (Medical University of Innsbruck). The figures presented in the manuscript were prepared by me. Drs. Van Petegem, Campiglio and Flucher wrote the manuscript. Dr. Van Petegem designed and supervised the project.  Chapter 3. I was involved in the cloning, acquisition, and analysis of the electrophysiological data. Figure 3.2 is obtained from the manuscript Wong King Yuen, S.M., Campiglio, M., Tung, C.C., Flucher, B.E., and Van Petegem, F. Structural insights into binding of STAC proteins to voltage-gated calcium channels. Proc. Natl. Acad. Sci. U.S.A (2017) 114(45): E9520-E9528. The cloning, expression, purification, and ITC experiments described vii  were completed in tandem with Jett Tung. Ethics approval for experiments involving Xenopus laevis was obtained from the University of British Columbia’s Animal Care Committee, Animal Care Certificate no.: A13-0212. Dr. Van Petegem designed and supervised the project.  Chapter 4. I performed the cloning, expression, purification, and all facets of structure determination of the RyR2 Y125H R420H and RyR2 R420W constructs. I was involved in the cloning and purification of the RyR2 “Cys-Lite” and RyR2 R383/K441C constructs. The time-resolved FRET experiments were performed and analyzed by Megan McCarthy, Dr. Robyn Rebbeck and Dr. Razvan Cornea (University of Minnesota).  The discussion involving the RyR2 H29D mutant has been adapted from the original publication Xiao, Z., Guo, W., Wong King Yuen, S.M., Wang, R., Zhang, L., Van Petegem, F., and Chen, S.R.W. The H29D mutation does not enhance cytosolic Ca2+ activation of the cardiac ryanodine receptor. PLOS One (2015) 10(9):e0139058. doi: 10.1371/journal.pone.0139058. Cloning, expression, purification, and thermal stability assays of the RyR2ABC H29D construct was performed by me. The remainder of the data discussed in the manuscript was collected and analyzed by Dr. Wayne Chen’s lab (University of Calgary).  Work discussing the RyR2 G357S mutant has been adapted from the publication Liu, Y., Wei, J., Wong King Yuen, S.M., Sun, B., Tang, Y., Wang, R., Van Petegem, F., and Chen, S.R.W. CPVT-associated cardiac ryanodine receptor mutation G357S with reduced penetrance impairs Ca2+ release termination and diminishes protein expression. PLOS One (2017) 12(9):e0184177. doi: 10.1371/journal.pone.0184177. Cloning, expression, purification, and thermal stability assays of the RyR2ABC G357S construct was completed jointly by Jett Tung and I.  The remainder of the data discussed in the manuscript was collected and analyzed by Dr. Wayne Chen’s lab. viii  Table of Contents Abstract ......................................................................................................................................... iii Lay Summary .................................................................................................................................v Preface ........................................................................................................................................... vi Table of Contents ....................................................................................................................... viii List of Tables .............................................................................................................................. xiii List of Figures ............................................................................................................................. xiv List of Symbols ........................................................................................................................... xvi List of Abbreviations ................................................................................................................ xvii Acknowledgements .................................................................................................................... xxi Dedication ................................................................................................................................. xxiii Chapter 1: Introduction ................................................................................................................1 1.1 Calcium ions – the universal biological signal ............................................................... 1 1.2 The role of Ca2+ in muscle contraction ........................................................................... 2 1.3 L-type voltage-gated calcium channels: voltage-sensors for EC coupling ..................... 5 1.3.1 Structure of the α1 ion conducting subunit ................................................................. 5 1.3.2 Auxiliary subunits: Function and Structure ................................................................ 9 1.3.3 Involvement of CaV1.1 II-III loop in skeletal muscle EC coupling .......................... 11 1.3.4 Additional properties of the L-type voltage-gated calcium channel: Channel inactivation ............................................................................................................................ 12 1.4 Ryanodine Receptors: calcium signal amplifiers .......................................................... 15 1.4.1 Ryanodine Receptors and disease ............................................................................. 17 1.4.1.1 RyR1 Channelopathies...................................................................................... 18 ix  1.4.1.2 RyR2 Channelopathies...................................................................................... 20 1.4.2 Structure of Ryanodine Receptors ............................................................................ 22 1.4.2.1 RyR structural insights obtained by cryo-EM .................................................. 23 1.4.2.2 RyR N-terminal domains .................................................................................. 27 1.4.3 Regulation of Ryanodine Receptors ......................................................................... 29 1.4.3.1 RyR regulation by Ca2+ ..................................................................................... 31 1.4.3.2 RyR regulation by protein-binding partners: FKBPs........................................ 33 1.4.3.3 RyR regulation by post-translational modification: phosphorylation ............... 34 1.5 Advances in understanding EC coupling: new players in the EC coupling machinery 34 1.5.1 STAC3: an essential component in skeletal muscle EC coupling ............................ 36 1.5.1.1 Native American Myopathy .............................................................................. 36 1.5.1.2 Defining the role of STAC3 as a member of the EC coupling machinery ....... 39 1.5.2 Defining the core components of the EC coupling machinery ................................. 40 1.6 Research Question ........................................................................................................ 41 Chapter 2: Structural insights into binding of STAC proteins to voltage-gated calcium channels .........................................................................................................................................43 2.1 Introduction ................................................................................................................... 43 2.2 Experimental Procedures .............................................................................................. 44 2.2.1 Expression Constructs ............................................................................................... 44 2.2.2 Protein expression and purification .......................................................................... 45 2.2.3 Crystallization and structure determination .............................................................. 46 2.2.4 Isothermal titration calorimetry ................................................................................ 49 2.2.5 Immunolabeling ........................................................................................................ 49 x  2.2.6 EC coupling analysis................................................................................................. 50 2.3 Results ........................................................................................................................... 51 2.3.1 The tandem SH3 domains form a rigid interaction ................................................... 51 2.3.2 The tandem SH3 domains interact with the II-III loop of CaV1.1 and CaV1.2 ......... 53 2.3.3 Crystal structure of STAC2 in complex with a peptide from the II-III loop ............ 58 2.3.4 Effect of the NAM mutation W284S ........................................................................ 63 2.3.5 Functional role of the interaction .............................................................................. 63 2.4 Discussion ..................................................................................................................... 67 Chapter 3: STAC3 modulates L-type voltage-gated calcium channel inactivation ...............70 3.1 Introduction ................................................................................................................... 70 3.2 Experimental Procedures .............................................................................................. 75 3.2.1 Electrophysiology ..................................................................................................... 75 3.2.2 Expression constructs................................................................................................ 76 3.2.3 Protein expression and purification .......................................................................... 76 3.2.4 Isothermal titration calorimetry ................................................................................ 77 3.3 Results ........................................................................................................................... 78 3.3.1 STAC3 modulates CaV1.2 channel inactivation ....................................................... 78 3.3.2 The STAC1 C1 domain does not interfere with the CaM-CaV1.2 IQ domain interaction ............................................................................................................................. 80 3.3.3 An intrinsically disordered STAC3 peptide binds to Calmodulin ............................ 86 3.4 Discussion ..................................................................................................................... 89 Chapter 4: Structural insights into the unique anion binding site within the cardiac ryanodine receptor N-terminal region .......................................................................................93 xi  4.1 Introduction ................................................................................................................... 93 4.2 Experimental Procedures .............................................................................................. 96 4.2.1 Expression constructs................................................................................................ 96 4.2.2 Protein expression and purification .......................................................................... 97 4.2.3 Thermal melt analysis ............................................................................................... 98 4.2.4 Crystallization and structure determination .............................................................. 99 4.2.5 Time-resolved FRET .............................................................................................. 100 4.3 Results ......................................................................................................................... 102 4.3.1 A RyR1-RyR2ABC hybrid alleviates RyR2ABC’s chloride-dependence but does not restore the hydrogen bond and salt bridge network of RyR1ABC. .................................... 102 4.3.2 The RyR2 disease mutant, R420W, abolishes chloride binding ............................. 105 4.3.3 Spectroscopic measurements of conformational dynamics incurred upon chloride binding ................................................................................................................................ 108 4.3.4 The H29D mutation does not affect the thermal stability of the N-terminal region of RyR2 nor does it enhance cytosolic Ca2+ activation of the cardiac ryanodine receptor ..... 111 4.3.5 The CPVT-associated G357S RyR2 mutation, with reduced penetrance, diminishes protein expression and impairs Ca2+ release termination ................................................... 113 4.4 Discussion ................................................................................................................... 117 4.4.1 Chloride binding to RyR2 induced global conformation changes in the N-terminal domains ............................................................................................................................... 117 4.4.2 The H29D mutation does not alter the intrinsic properties of RyR2 or the thermal stability of the N-terminal domains .................................................................................... 118 xii  4.4.3 The CPVT-associated G357S RyR2 mutation impairs Ca2+ release termination and reduces the thermal stability of the N-terminal domains .................................................... 119 Chapter 5: Concluding remarks and future directions ..........................................................122 5.1 STAC3 is an essential component of the skeletal muscle EC coupling machinery .... 124 5.2 STAC proteins: A Swiss-Army knife for the EC coupling complex? ........................ 126 5.3 Ligand-dependent modulation of RyRs ...................................................................... 129 5.4 Understanding the mechanisms of RyR disease-associated mutations ....................... 133 5.5 Muscling in on EC coupling? ..................................................................................... 135 Bibliography ...............................................................................................................................137  xiii  List of Tables Table 2.1 Human STAC and CaV1.1/1.2 constructs described in the crystallographic and ITC experiments ................................................................................................................................... 45 Table 2.2 Crystallographic data collection and refinement statistics for STAC crystals ............. 48 Table 2.3 Thermodynamic parameters for the interaction between individual or tandem SH3 STAC domains and the II-III loop constructs of CaV1.1 and 1.2 ................................................. 56 Table 3.1 Thermodynamic parameters for the interaction between full-length, N- and C-lobe CaM and STAC1 C1 ..................................................................................................................... 86 Table 3.2 Thermodynamic parameters for the interaction between full-length, N- and C-lobe CaM and a STAC3 peptide. .......................................................................................................... 87 Table 4.1 Crystallographic data collection and refinement statistics for RyR2ABC mutant constructs .................................................................................................................................... 101 Table 4.2 Mid-point unfolding temperature, TM, for RyR2ABC Y125H R420H in buffers of varying KCl concentration .......................................................................................................... 104  xiv  List of Figures Figure 1.1 Schematic diagram illustrating the T-tubule arrangement and components involved in EC coupling.. .................................................................................................................................. 4 Figure 1.2 Structure of the CaV1.1 channel complex... ................................................................... 8 Figure 1.3 CaV1.2 channel inactivation. ....................................................................................... 13 Figure 1.4 Disease hotspots within RyR1 and RyR2. ................................................................... 18 Figure 1.5 Overall structure of the RyR........................................................................................ 24 Figure 1.6 The N-terminal ABC domains and their location within RyR1 .................................. 28 Figure 1.7 RyR regulation by small molecules, protein binding-partners and post-translational modification. ................................................................................................................................. 30 Figure 2.1 Crystal structures of the STAC SH3 domains ............................................................. 52 Figure 2.2 Superposition of the SH3 domains of the STAC isoforms.......................................... 53 Figure 2.3 ITC experiments reveal binding between peptides of the CaV1.1 II-III loop and the SH3 domains of the STAC proteins.............................................................................................. 55 Figure 2.4 Determining the binding determinants for the STAC tandem SH3 domains and II-III loop interaction. ............................................................................................................................ 57 Figure 2.5 Titrations between the tandem SH3 domains of STAC2 (296-411) and predicted SH3 binding sites within CaV1.2 reveal no additional binding sites. ................................................... 59 Figure 2.6 Residues of the CaV1.1 “core II-III loop” have an extensive interaction network with the first SH3 domain of STAC2.................................................................................................... 62 Figure 2.7 Examining the binding interface of the CaV1.1 II-III loop peptide and the tandem SH3 domains of STAC2. ...................................................................................................................... 64 Figure 2.8 Mutation of the IPR motif perturbs EC coupling in skeletal muscle myotubes.. ........ 66 xv  Figure 3.1  NMR structure of the STAC3 C1 domain.. ................................................................ 74 Figure 3.2 STAC3 influences the inactivation kinetics of CaV1.2 α1c.. ....................................... 81 Figure 3.3 The STAC1 C1 domain does not interfere with CaM’s ability to bind to a CaV1.2 IQ peptide ........................................................................................................................................... 84 Figure 3.4 The STAC1 C1 domain interacts directly with CaM.. ................................................ 85 Figure 3.5 ITC experiments reveal binding between an intrinsically disordered STAC3 peptide and Ca2+-CaM.. ............................................................................................................................. 88 Figure 4.1 Mapping disease-associated mutations in the amino-terminal domains of RyR1 and RyR2.. ........................................................................................................................................... 95 Figure 4.2 Comparison of the amino-terminal domains of RyR1 and RyR2. .............................. 97 Figure 4.3 Thermal stability curves of RyR2ABC Y125H R420H in buffers containing various concentrations of KCl. ................................................................................................................ 103 Figure 4.4 RyR2ABC Y125H R420H abolishes halide ion binding and imparts local conformational rearrangement .................................................................................................... 105 Figure 4.5 Removal of chloride in RyR2ABC causes local residue re-arrangement and global domain reorientation ................................................................................................................... 107 Figure 4.6 TR-FRET detection of chloride’s structural effect on RyR2ABC.. .......................... 110 Figure 4.7 Effect of the H29D mutation on the thermal stability of the N-terminal domains of RyR2. .......................................................................................................................................... 113 Figure 4.8 The G357S mutation reduces the stability of the ABC domains of RyR2. ............... 116 Figure 4.9 Gating model proposed for RyR channels and the closely related IP3R. ................. 118  xvi  List of Symbols Å   Angstrom oC   Degree Celsius ∆H   Change in enthalpy Kd   Dissociation constant Ω   Ohms ∆S   Change in entropy V   Volts  xvii  List of Abbreviations Amino Acid One Letter Code: A  Ala  Alanine C  Cys  Cysteine D  Asp  Aspartate/aspartic acid E  Glu  Glutamate/glutamic acid F  Phe  Phenylalanine G  Gly  Glycine H  His  Histidine I  Ile  Isoleucine K  Lys  Lysine L  Leu  Leucine M  Met  Methionine N  Asn  Asparagine P  Pro  Proline Q  Gln  Glutamine R  Arg  Arginine S  Ser  Serine T  Thr  Threonine V  Val  Valine W  Trp  Tryptophan  Y  Tyr  Tyrosine    Additional Abbreviations: α-subunit  alpha-subunit α2δ-subunit  alpha-2-delta subunit ADP   Adenosine diphosphate AF350   Alexa Fluor 350 C5 maleimide AF488   Alexa Fluor 488 C5 maleimide ARDV2  Arrhythmogenic right ventricular dysplasia type 2 ATP   Adenosine triphosphate Ba2+   Barium ion BAPTA  1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid  Ba(OH)2  Barium hydroxide β-me   β-mercapotoethanol BSA   Bovine serum albumin β-subunit  Beta-subunit C-lobe   Carboxy-terminus lobe Ca2+   Calcium ion CaCl2    Calcium chloride Ca(NO3)2  Calcium nitrate CaM   Calmodulin CaV   Voltage-gated calcium channel xviii  CaV1.x   L-type voltage-gated calcium channel CaVAb   Bacterial (Arcobacter butzleri) voltage-gated calcium channel CCD   Central core disease cDNA   Complementary deoxyribonucleic acid CDF   Calcium-dependent facilitation CDI   Calcium-dependent inactivation CHAPS  3-((3-Cholamidopropyl) dimethylammonio)-1-propanesulfonate CHO   Chinese hamster ovary (cell) CICR   Calcium-induced calcium release Cl-   Chloride ion COOT   Crystallographic Object-Oriented Toolkit CPVT   Catecholaminergic ventricular tachycardia CTD   C-terminal domain C-terminus  Carboxy-terminus CV   Column volume Da   Dalton DADs   Delayed afterdepolarizations DAG   sn-1,2-Diacylglycerol DSSP   Define Secondary Structure of Proteins DHPR   Dihydropyridine receptor EC   Excitation-contraction ECG   Electrocardiogram EDTA   Ethylenediaminetetraacetic acid EM   Electron microscopy EMD   Electron Microscopy Databank EPR   Electron Paramagnetic Resonance ER   Endoplasmic reticulum FF   Fast flow FKBP   FK506-binding protein FRET   Förster resonance energy transfer FWHM  Full width at half-maximum γ-subunit  Gamma-subunit GFP   Green fluorescent protein GLT    muscular dysGenic Line transfected with the large T antigen HD1   Helical Domain 1 HEK293  Human embryonic kidney 293 (cell) HEPES  4-(2-Hydroxyethyl)-1-piperazineethanesulfonic acid HNO3   Nitric acid IBa   Barium current ICa   Calcium current IP3R   Inositol 1,4,5-triphosphate receptor IQ motif  Isoleucine-glutamine motif IRF   Instrument response function ITC   Isothermal titration calorimetry JP2   Junctophilin2 xix  K+    Potassium ion KCl   Potassium chloride KPO4   Monopotassium phosphate KO   Knock-out KOH   Potassium hydroxide MH   Malignant hyperthermia MBP   Maltose binding protein MWCO  Molecular weight cut-off N-lobe   Amino-terminal lobe NaI   Sodium iodide NaOH   Sodium hydroxide NaV   Voltage-gated sodium channel NaVRh   Bacterial (Rickettsiales sp. HIMB114, denoted as Rh) voltage-gated  sodium channel NaVAb   Bacterial (Arcobacter butzleri) voltage-gated sodium channel NCBI   National Center for Biotechnology Information (NH4)2SO4  Ammonium sulphate N-terminus  Amino-terminus NAM   Native American Myopathy NCX   Na+/Ca2+ exchanger NMR   Nuclear magnetic resonance PCR   Polymerase chain reaction PDB   Protein Data Bank PEG   Polyethylene glycol PHENIX  Python-based Hierarchical Environment for Integrated Xtallography Pi   Inorganic phosphate PKA    Protein kinase A PKC   Protein kinase C PMSF   Phenylmethane sulfonyl fluoride Po    Open probability psi   Pounds-per-square inch R   Distance REFMAC  Refinement of macromolecular structures RMSD   Root-mean square deviation RNA   Ribonucleic acid RPM   Rotations per minute RyR   Ryanodine Receptor SAD   Single-wavelength anomalous diffraction SDS-PAGE  Sodium dodecyl sulfate-Polyacrylamide gel electrophoresis SERCA  Sarcoplasmic/endoplasmic reticulum Ca2+ ATPase SH3   Src homology 3  SOE   Splicing by overlap extension SOICR  Store overload-induced calcium release SPA   Single particle analysis SR    Sarcoplasmic reticulum xx  T-tubular  Transverse-tubular TEV   Tobacco-etch virus TEVC   Two-electrode voltage clamp  TCEP   Tris(2-carboxyethyl)phosphine TCSPC  Time-correlated single photon counting TM   Transmembrane TM   Mid-point unfolding temperature TR-FRET  Time-resolved FRET Tris-HCl  Tris-hydrochloride VDI   Voltage-dependent inactivation VSD   Voltage-sensing domain v/v   Volume-per-volume VWA   Von Willebrand factor domain A WT   Wildtype w/v   Weight-per-volume Zn2+    Zinc ion ZnSO4   Zinc sulphate 2xYT   2x Yeast extract tryptone (media) 3D    3-dimensional   xxi  Acknowledgements My journey through graduate school would not have been possible without my supervisor, Dr. Filip Van Petegem. I am immensely grateful to Dr. Van Petegem for his guidance and contributions of time and ideas to make my PhD experience productive and stimulating. The passion and enthusiasm he demonstrates for his research motivated me to elevate my research and has ultimately served to make me a more well-rounded scientist and researcher. As a token of my gratitude, I will lay it all out on the line as I say a heartfelt, “Baie Dankie!”. Members of my thesis committee, Drs. Natalie Strynadka, Ed Moore, and Harley Kurata, have been wonderful sources of insight into my research projects. I am grateful for the time they have invested to contribute to my thesis.  Producing this body of research would not have been possible without the funding received from the University of British Columbia’s Four-Year Fellowship. My projects could ambitiously advance to another level of understanding with the help of our collaborators. Much gratitude is extended to Dr. Wayne Chen, Dr. Bernhard Flucher and Dr. Razvan Cornea’s labs.  The members – both past and present – of the Van Petegem lab provided a collaborative, intellectually stimulating and fun environment to learn in as I navigated my way through the Ph.D. experience. The camaraderie and friendships have truly enriched my journey. I have endured many a nail-biting and nerve-wrangling adventure in my Ph.D. tenure, which would have been insurmountable if it were not for the support of our lab technician-extraordinaire, Jett Tung, multi-talented and ever-knowledgeable post-doc, Dr. Bernd Gardill, and Honorary Van Petegem Lab Members, Drs. Sylvia Cheung and Leo Ng. Thus, to Jett, Bernd, Sylvia and Leo: your friendship has been invaluable, and I send my heartfelt thanks.  xxii  My mom’s humble story of a Christmas tree star sparked my inspiration to embark on this remarkable journey. To my cheerleaders – mom and dad – it is with your zeal and unyielding love and support that I am able to reach for the stars. Words simply fail to convey my gratitude and appreciation for the many sacrifices you have endured to ensure that I could achieve my full potential. A humongous hug is owed to both of you.     xxiii  Dedication  For Mom, Dad and Koong-Koong … ... the stars of my Christmas tree                             *             *                                   _/^\_                                  <     >                 *                 /.-.\         *                          *        `/&\`                   *                                  ,@.*;@,                                 /_o.I %_\    *                    *           (`'--:o(_@;                               /`;--.,__ `')             *                              ;@`o % O,*`'`&\                         *    (`'--)_@ ;o %'()\      *                             /`;--._`''--._O'@;                            /&*,()~o`;-.,_ `""`)                 *          /`,@ ;+& () o*`;-';\                           (`""--.,_0 +% @' &()\                           /-.,_    ``''--....-'`)  *                      *    /@%;o`:;'--,.__   __.'\                          ;*,&(); @ % &^;~`"`o;@();         *                          /(); o^~; & ().o@*&`;&%O\                          `"="==""==,,,.,="=="==="`                     __.------.(\-''#####---...___...-----._    1  Chapter 1: Introduction 1.1 Calcium ions – the universal biological signal Calcium ions are ubiquitous intracellular signaling molecules that are versatile enough to operate over a broad temporal range to regulate numerous cellular processes, beginning with the creation of life at fertilization and culminating in cellular apoptosis1,2. The effectiveness and speed of Ca2+ signaling is mediated by the 20 000-fold Ca2+ concentration gradient that is maintained between cells and their external environment3. Whilst cells are usually bathed in a solution containing millimolar concentrations of Ca2+, resting intracellular Ca2+ concentrations are vastly different, typically being in the ~100nM range3. A homeostatic balance avoids excessive energy expenditure to significantly raise intracellular Ca2+ levels4 and circumvents organellar damage, protein and nucleic acid aggregation, and phosphate precipitation associated with elevated Ca2+ concentrations5. The cell’s ability to precisely regulate its concentration of free and sequestered Ca2+ both temporally and spatially is a feature which makes Ca2+ such a versatile signaling ion5. Buffering of cytosolic Ca2+ levels is achieved by sequestering the ion via calcium-binding proteins and transport of the ion, from both the extracellular environment and intracellular stores, via calcium-specific ion channels and pumps4,5.  The biological significance of Ca2+ was first demonstrated in 1883 by Sydney Ringer. Serendipitously, Ringer observed that the presence of Ca2+ in a fluid mimicking the constituents of blood was required for sustained ventricular contraction6. This seminal finding cultivated a flourishing interest to establish the cellular roles of the calcium ion. It has since been found that Ca2+ is involved in a plethora of processes from fertilization, cellular proliferation and motility, metabolism, muscle contraction, conduction of nerve impulses, to the imminent cellular apoptosis1,3,5.  2  1.2 The role of Ca2+ in muscle contraction Contraction and relaxation of striated muscle is orchestrated by rapid changes in the myoplasmic free- Ca2+ concentration7. Prior to even knowing the molecular components of the “Ca2+ release unit”, it had been established that Ca2+ was the activator of the contractile machinery7,8. Tension generation in the muscle fiber is a consequence of an upswing in the myoplasmic Ca2+ concentration, elicited by electrical stimulation7.  The apparent dependence of contraction on membrane excitation was indicative that the excited membrane initiated a process which moved inward to the core of the fiber to activate the contractile elements9. Alexander Sandow aptly coined the term “excitation-contraction (EC) coupling” to describe the communication between electrical events occurring at the sarcolemma and intracellular Ca2+ release, which precipitates muscle contraction9. However, the rapid conduction (1-2 milliseconds) of the action potential to initiate contraction in the center of fibers of 50-100µm in diameter remained intriguing8,10. Insights into this could be gained by examination of the morphology of the sarcolemma. Skeletal and cardiac muscle sarcolemma is characterized by invaginations called transverse- or T-tubules that run perpendicular to the surface of the cell deep into its body (Fig1.1)11. The T-tubule network ensures that no part of the fiber interior is more than 1µm away from a membrane that is continuous with the sarcolemma8,12–15. Electron microscopy revealed a close association between the T-tubule network and the expanded terminal cisternae sacs of the sarcoplasmic reticulum (SR)8,14. In skeletal muscle, the junction between the external and internal membrane is called a triad due to the T-tubule being flanked on either side by a SR terminal cisternum8,14. An equivalent feature in cardiac tissue is the dyad: a single T-tubule paired with a terminal cisternum16. Triads and dyads form the anatomical basis for skeletal and cardiac muscle EC coupling, respectively8.  3  The final stage of EC coupling involves the release of Ca2+ from intracellular SR stores, which then activates the contractile proteins17,18. A link, however, was missing regarding how a change in potential across the T-tubule membrane could result in Ca2+ release from the neighbouring SR. During the 1970s, the skeletal muscle T-tubule membrane was identified to be a rich source of the L-type voltage-gated calcium channel (CaV), also known as the dihydropyridine receptor (DHPR)8. Work by Armstrong and Bezanilla on squid giant axon membranes led to the discovery of novel and tiny electrical signals, which have now been termed gating currents19,20. Gating current was thought to be the movement of charged residues within the T-tubule membrane and reflected a voltage-sensor response to depolarization that initiated EC coupling8. Sensitivity of EC coupling to dihydropyridine-based compounds21,22, that were corresponding agonists or antagonists of CaVs, further fueled the hypothesis that CaVs were the voltage-sensors for EC coupling8. The 1980s provided insight into the molecular identity of the elusive SR Ca2+-release channel, which was identified as a very high molecular weight protein with a high affinity for the plant alkaloid ryanodine23–26, hence acquiring the moniker Ryanodine Receptor (RyR). Electron microscopy provided pivotal evidence for mechanical EC coupling in skeletal muscle. Examining the architecture of the junctional SR and T-tubule membranes in skeletal muscle triads and the morphology of the proteins isolated from these membranes, the RyR and CaV, provided the first hints that a direct interaction existed between the RyR and a protein component of the junctional T-tubule membrane27. Thus, the electrical stimulus via the voltage-gated calcium channel signals the opening of the RyR that then triggers Ca2+ release from intracellular stores. In cardiac tissue, EC coupling occurs via an alternate process termed calcium-induced calcium release (CICR)28, whereby entry of extracellular Ca2+ through the CaV activates Ca2+ release from the SR29.    4   Figure 1.1 Schematic diagram illustrating the T-tubule arrangement and components involved in EC coupling. Plasma membrane depolarization activates L-type voltage-gated calcium (CaV) channels. In skeletal muscle, this results in a physical coupling between the CaV and the neighbouring RyR in the SR membrane. In cardiac tissue, Ca2+ influx upon CaV activation triggers opening of the RyR. In both cases, activation of RyR leads to a massive efflux of luminal Ca2+ into the cytosol, which activates the contractile machinery leading to muscle contraction. Muscle relaxation is a consequence of Ca2+ removal from the cytosol, a process achieved by both the SERCA pump in the SR membrane and Na+/Ca2+-exchanger (NCX) in the plasma membrane. RyR activity can be upregulated by PKA phosphorylation upon activation of the β-adrenergic receptor in the plasma membrane.  Contracture of sarcomere length is described by the sliding filament theory30,31. When Ca2+ is released from the SR into the cytoplasm, Ca2+ allosterically binds to troponin C, resulting in a conformational change in the troponin-tropomyosin complex, exposing the myosin-binding sites on actin. Multiple cycles of ADP+Pi release and ATP hydrolysis ensues allowing for 5  myosin head movement along the actin filament producing force and motion, and ultimately contraction. Muscle relaxation occurs when removal of Ca2+ from the cytoplasm exceeds the amount of Ca2+ released into it. The majority of Ca2+ is returned back into the SR via the action of the sarcoplasmic/endoplasmic reticulum Ca2+-ATPase (SERCA) pump; additionally, Ca2+ is pumped into the extracellular environment by the sodium-calcium exchanger (NCX)11.   1.3 L-type voltage-gated calcium channels: voltage-sensors for EC coupling L-type, high-voltage activated calcium channels are multi-subunit membrane proteins that regulate Ca2+ influx into excitable cells. During the infancy of the EC coupling field, it was thought that the gating current of CaV, which reflected the movement of a dipole in the T-tubule membrane, was like a lever that pulled a plug from the terminal cisternae to dump Ca2+ in the myoplasm8. The skeletal muscle CaV channel was first identified by photoaffinity labeling, purification and reconstitution32; and its amino acid sequence determined by cDNA cloning and sequencing33. In mammals, ten CaV subtypes have been identified and classified into three families: CaV1, CaV2 and CaV3, based on their ion conducting α1-subunit34. Within the CaV1 subfamily, four isoforms of the channel have been identified, CaV1.1-1.4. Due to the enrichment of CaVs in the skeletal muscle T-tubule membrane, studies have focused on CaV1.1, the skeletal muscle isoform35.  1.3.1 Structure of the α1 ion conducting subunit CaVs are composed of five subunits, the membrane-spanning α1-, γ-, and δ-subunits, a cytosolic β-subunit and an extracellular α2-subunit, which is disulfide-linked to the δ-subunit32,36. The α1-subunit is the fundamental structural building block of the channel, 6  consisting of four repeats of a transmembrane domain containing six membrane-spanning helices. The first four transmembrane helices, S1-S4, form the voltage-sensing module that responds to changes in membrane potential (Fig. 1.2). Transmembrane helices S5 and S6, and the connecting pore-forming (P) loop, form the ion conducting pore.  Structural insights on the L-type voltage-gated calcium channel has relied heavily on electron microscopy (EM) techniques and single particle analysis (SPA)35. Recent advancements in the cryo-EM field heralded a new era, the “resolution revolution”, in structural biology and allowed the nanometer resolution barrier (3-2nm) of structures obtained of the skeletal muscle CaV1.1 channel to be exceeded35. The molecular architecture of CaV1.1, with its complete set of auxiliary subunits, has been resolved to a nominal resolution of 3.6Å34,37. The overall structure is ~170Å in height and 100Å in the longest dimension of width. Although the transmembrane region could be unambiguously identified in the EM density map, the lack of density for the inter-repeat cytosolic loops yielded challenges to distinguish the four homologous repeats in the α1-subunit. Analysis of distinctive extracellular loops of the pore domain, unique sequence patterns, and mass spectroscopy analysis of cross-linking characterizations allowed for the eventual identification of the α1-subunit repeats, and revealed a clockwise arrangement of the four homologous repeats in the extracellular view – a feature conserved in all eukaryotic CaV and NaV channels34,37. The pore domain of CaV1.1 exhibits a pseudo-four-fold symmetry due to marked differences in primary sequence and conformations of the extracellular loops linking transmembrane segments S5 and S6 among the four repeats. A kink in the S6 segment of repeat IV (close to the intracellular end) destructs the symmetry at the activation gate.  Despite variation in primary sequence, the backbone configuration of CaV1.1’s selectivity filter is comparable to those of bacterial homotetrametric CaV and NaV channels, CaVAb and 7  NaVRh, where two pore helices, P1 and P2, support the selectivity filter37. The high-quality density of the P1 and P2 helices provided an accurate template for backbone assignment of the selectivity filter residues, including the critical EEEE residues (Glu292/614/1014/1323) that provide the sidechains critical for ion selectivity and the two preceding residues in each repeat that contribute carbonyl oxygens (C=O)34. Two Ca2+ ions were assigned to density within the selectivity filter vestibule, since the channel was purified in the presence of Ca2+. The height of the two Ca2+ ions in the 3.6Å map are similar to those in CaVAb34. Polar and negatively charged residues line the entrance to the selectivity filter vestibule. These residues may constitute extracellular calcium ion binding sites, or alternatively, act as an electronegative sink attracting cations37.   The presented rabbit CaV1.1 structure revealed that the four voltage-sensing domains share similar, but non-identical structural features34. Sequence alignment of rabbit CaV1.1 with the ten human CaV channels reveals up to six gating charge residues, a mixture of arginines and lysines, on each S4 segment. One side of the 310 helix of S4 was decorated with the gating charges in all four voltage-sensing domains – a feature comparable to NaVAb and its chimaera with NaV1.7, but different from NaVRh34. The voltage-dependent movement of the S4 segment results in channel opening and closing38.   Unexpectedly, the C-terminal domain (CTD) and the cytoplasmic loop linking repeats III and IV (III-IV loop) form a globular domain37 (Fig 1.2C). While there are marked structural similarities between CaVs and NaVs, the CaV1.1 α6 helix is substantially longer than its NaV counterpart. This is an interesting difference to note between the two channels since this helix contains the so-called isoleucine-glutamine (IQ-) motif. Within CaV1.1, the α6 helix adopts an extended conformation which is then capable of binding to Ca2+-CaM, a crucial channel  8  Figure 1.2 Structure of the CaV1.1 channel complex. A. Topology of the α1-subunit. N- and C- termini have been denoted. Roman numerals (I-IV) indicate the four repeats of the α1-9  subunit. The blue transmembrane helices represent S1-S4, the voltage-sensing module of the channel. The grey helices represent S5 and S6, which form the ion conducting pore of the channel. B. Reconstruction of the entire channel complex, as determined from a 3.6Å cryo-EM map (EMD: 9513; PDB: 5GJV). In addition to the main α1-subunit (blue = voltage sensor; grey = pore-forming region), the beta (sand colour), gamma (magenta), alpha-2 (green) and delta (mauve) auxiliary subunits have been resolved. The modelled calcium ions are represented as yellow spheres. C. A side view of the α1-subunit. The AID motif, the key motif for binding the β-subunit has been coloured brown. The globular CTD has been coloured purple. Key selectivity filter glutamate residues have been shown in stick form and are coloured black D. A top view of the channel. The clockwise arrangement of repeats I-IV can be observed.  regulator39,40. The III-IV loop plays an essential role in NaV channel regulation. Several studies have reported that the III-IV loop is involved in regulating the Na+ current following channel activation41, a feature of the channel known as inactivation (discussed further in section 1.3.4). Thus, the structural similarity between the III-IV loop of the intact CaV1.1 channel and an isolated fragment within NaV will facilitate the mechanistic understanding of CaV channel regulation37.  1.3.2 Auxiliary subunits: Function and Structure The α1-subunit of the L-type voltage-gated calcium channel co-assembles with the extracellular α2-δ, the intracellular β-, and the transmembrane γ- auxiliary subunits. The auxiliary subunits of the calcium channel complex are all products of distinct genes42–44, except for the α2-δ subunit, which is encoded by a single gene45. Purified CaV1 and CaV2 channels contain a tightly bound cytosolic CaVβ protein38. CaVα1 and CaVβ associate through a high affinity (Kd ~ nM) interaction between a short peptide segment within the cytosolic loop linking domains I and II, known as the α-interacting domain (AID), and a groove within the CaVβ, the α-binding pocket38. There are four subfamilies of CaVβ (β1-β4), all capable of dramatically enhancing calcium channel currents when they are co-expressed in heterologous expression 10  systems along with the α1-subunit of CaV1 or CaV238,46. The β-subunit also influences the channel’s voltage dependence and kinetics of activation and inactivation, however CaVβ does not affect ion permeation38. Additionally, the β-subunit is involved in the regulation or modulation of CaV1 and CaV2 channels by protein kinases, G proteins and small RGK (Rem, Rem2, Rad, Gem/Kir) proteins38. Similarly, the α2δ-subunit is involved in modifying the channel’s biophysical properties, but its main role is to increase calcium channel current by promoting trafficking of the α1-subunit to the plasma membrane and/or increasing its retention there38,47. Four isoforms of the α2δ-subunit, α2δ-1 to α2δ-4, have been identified46. Based on phylogenetic analysis, sequence homology and tissue distribution, the γ-subunits have been divided into two groups: skeletal γ- (γ1 and γ6) and neuronal γ- (γ2-5 and γ7-8) subunits46. Studies with γ1-null mice revealed an inhibitory effect of γ1 on the functional activity of calcium channels in skeletal muscle46. A more recent study demonstrated the role of γ1 in CaV1.1’s function and membrane trafficking in tsa201 cells48. The association of numerous disorders – epilepsy, ataxia, cardiac anomalies – with mutations in the channel subunits has spurred the study of calcium channel regulation by its auxiliary subunits46. The recent cryo-EM structures were able to capture CaV1.1 in complex with its auxiliary subunits at a nominal resolution of 4.2-3.6Å34,37. The cytoplasmic β1a-subunit was placed near the voltage-sensing domain of repeat II (VSDII). Although a limited resolution in this region only allowed assignment of 25% of the sidechains, secondary structural elements of the β-subunit supported reliable docking of a crystal structure of the β-subunit in complex with the AID peptide49. The α2δ-subunit comprises four tandem cache domains and one von Willebrand factor domain A (VWA). The VWA and two cache domains interact with the extended extracellular loops of α1. Although the δ-subunit was predicted to possess a single transmembrane helix, 11  recent mass spectrometry and structural characterizations suggest that the δ-subunit may be anchored to the membrane through a glycophosphatidylinositol modification. The γ-subunit contains four transmembrane helices and shares a structural fold with claudins. A detailed interaction between the transmembrane helices 2 and 3 of γ and the S3 and S4 of VDSIV of the α1-subunit has been observed. However, due to the highly hydrophobic nature of the interface, there is unlikely to be much specificity between γ and VDSIV34.  1.3.3 Involvement of CaV1.1 II-III loop in skeletal muscle EC coupling Failure of skeletal muscle EC coupling, and subsequent lethality, in naturally occurring CaV1.1 α1s-null dysgenic mice50 fueled the hypothesis that CaV1.1 is the voltage sensor for skeletal muscle EC coupling8. Injection of cDNA encoding the CaV1.1 α1-subunit restored EC coupling and the L-type Ca2+ current that was missing in skeletal muscle myotubes in dysgenic mice51,52. The restored coupling resembled normal skeletal muscle EC coupling, whereby entry of extracellular Ca2+ is not required. Conversely, injection of CaV1.2 α1-subunit cDNA into dysgenic myotubes yielded L-type Ca2+ current and cardiac-type EC coupling, which requires entry of extracellular Ca2+ 52. Chimeras of the α1-subunit of CaV1.1 and CaV1.2 showed that a skeletal muscle-specific stretch of residues in the cytoplasmic loop linking domains II and III (II-III loop) was essential for skeletal muscle EC coupling52,53. In addition to transmitting an orthograde EC coupling signal to the Ryanodine Receptor type 1 (RyR1, expressed predominantly in skeletal muscle), the II-III loop also receives a retrograde, current-enhancing signal from the RyR1 to CaV1.154–56. Thus, the II-III loop plays an essential role in this bidirectional coupling mechanism between CaV1.1 and the RyR.  12  1.3.4 Additional properties of the L-type voltage-gated calcium channel: Channel inactivation Calcium ion influx into myocytes via voltage-gated calcium channels is fundamental to instigating the contractile response. Equally important is a negative feedback mechanism, inactivation of ion channels: a process which prevents the collapse of ionic gradients, determines the duration of action potentials and refractory period in excitable tissues57. Channel inactivation can be defined as a transition into a non-conducting state following channel opening (activation)57; however, it has been observed that several channels – like voltage-gated sodium channels – can also inactivate from a closed state58,59. On a whole-cell level, inactivation is observed as a decay in current amplitude over the course of membrane depolarization, and a diminished ability for the channel to open at more depolarized membrane potentials57,60. Inactivation of calcium channels is an essential mechanism by which cells can tightly regulate their intracellular Ca2+ levels, a requisite for optimal cellular function and survival. In excitable cells, calcium channel inactivation modulates cellular excitability. Additionally, conformational changes incurred upon inactivation can dramatically alter the affinity of the channel for numerous pharmacological agents, many of which are clinically used as calcium channel therapeutics57. All ten of the mammalian CaV subtypes can be inactivated by intrinsic mechanisms that are triggered by the same stimulating depolarization61. This process is known as voltage-dependent inactivation (VDI). The extent of VDI varies among the CaV isoforms, with the high-voltage activated calcium channel subfamilies (encoded by the CaV1.x or CaV2.x pore forming subunits) being potently modulated by the auxiliary β-subunit61. The high-voltage activated channels have an additional layer of complexity in their pore-opening modulation. The Ca2+ 13  influx generated by the channel’s own opening is capable of modulating channel activation through a process known as calcium-dependent inactivation (CDI). Dual regulation of the high-voltage activated calcium channel by both voltage and Ca2+ ensures that the channels maintain their sensitivity to the electrical and chemical signals generated by channel opening in both myocytes and neurons60. Most studies on calcium-dependent inactivation have focused on the L- (CaV1.2) and the P/Q- (CaV2.1) type channels, which represent the major channel type in cardiac ventricular and cerebellar Purkinje cells, respectively60. Since voltage and Ca2+ drive inactivation in CaV1.2 channels, these two processes can be distinguished by replacing extracellular Ca2+ with Ba2+, a divalent cation unable to induce CDI.              Figure 1.3 CaV1.2 channel inactivation. Representative traces of Xenopus laevis oocytes co-expressing the α1-subunit of CaV1.2 and the β1- and α2δ-auxiliary subunits. Channels were conducting either calcium or barium ions, as indicated. Local Ca2+ entry causes a decay in current amplitude, representing CDI. On the contrary, Ba2+ entry does not result in CDI.    14  The phenomenon of L-type CDI is characterized by diminished channel open probability during prolonged depolarization, and is only seen when calcium is the carrier ion60. During subsequent depolarizing stimuli, expected to induce maximal channel opening in resting conditions, CDI inhibits further channel opening. By recording such inhibition at different membrane potentials, a typical U-shaped inactivation curve is produced, illustrating that CDI is maximal at voltages where Ca2+ influx is at its peak. The prototypical U-shaped inactivation curve is absent when Ba2+ is used as a carrier ion.   Calcium ion binding to a calcium-binding EF-hand motif within the carboxy-terminus of the main pore-forming α1c channel subunit was a popular model proposed to explain the mechanism underlying CDI62. However, studies where an over-expressed calcium-insensitive mutant calmodulin (CaM) ablated CDI in a dominant negative manner demonstrated that CaM is the calcium sensor for inactivation39. As its name suggests, CaM is a Ca2+-modulated protein. A prototypical EF-hand containing protein, CaM contains four EF-hands, organized within two lobes, the N- and C-terminal lobes, which are joined by a flexible linker. Both lobes are capable of binding to Ca2+, albeit with different affinity: CaM’s N-lobe binds with a lower affinity of ~1µM, while C-lobe binds with an affinity of 0.1µM63. Combined with biochemical binding studies, it was demonstrated that CaM is constitutively tethered to the channel complex, in a Ca2+-independent manner39,40. Inactivation occurs via CDI when Ca2+ binds to this tethered CaM, and the resulting Ca2+-CaM complex binds to the IQ- motif in the carboxy tail of α1c39,40. Mutations in the IQ-motif severely impair CDI, but have minor effects on VDI40,60.  The cytosolic loop connecting domains I and II (I-II loop) and the adjacent S6 segment of repeat I were the first molecular determinants identified for VDI64. Characterizing the effects of mutations in this loop on VDI and CDI confirmed its role in channel inactivation. Interactions 15  with the membrane-anchored auxiliary β2-subunit blocks VDI, suggesting that the I-II loop could act as a blocking particle that occludes the pore during inactivation60. An additional CaM binding site lies within the cytosolic N-terminus of CaV1.2. However, since binding to this site only occurs with Ca2+-CaM, it was rejected as the constitutive tethering site for CaM. Deletion of the N-terminal CaM binding site severely perturbs CDI, with moderate changes to VDI65. Multiple mutations within the S6 segment of all four repeats of CaV1.2 have also been demonstrated to perturb VDI, suggesting a role of the inner channel vestibule as a determinant for VDI60. The role of the inner vestibule appears to work independently of VDI regulation by the I-II loop-β-subunit complex60.  As an additional layer of complexity to the inactivation of CaV1.2, the auxiliary CaVβ-subunit and various signaling pathways, including phosphorylation/dephosphorylation and cytoskeletal elements (both microtubules and microfilaments) modulate CDI and VDI60. Although the mechanisms by which these factors affect CDI and VDI remain enigmatic, it has been suggested that these additional factors target the set of fundamental interactions responsible for CDI and VDI described above60. Thus, these interactions represent a crucial regulatory means for the specific and fine spatiotemporal tuning of Ca2+ entry into excitable cells.   1.4 Ryanodine Receptors: calcium signal amplifiers Even prior to their isolation, the elusive SR Ca2+ release channel had been visualized via thin section or negative stain microscopy studies of muscle ultrastructure. These images revealed a tight connection between the T-tubule system flanked on either side by the expanded terminal cisternae sacs of the SR. A feature of the aptly named triad junction was periodic electron-dense protrusions, named “feet”, which spanned the ~10nm junctional gap between the cytoplasmic 16  leaflets of the T-tubule and the SR membrane14. The early 1980s saw an exciting breakthrough for the EC coupling field when these SR calcium ion gateways were isolated and purified26,66. Observations that ryanodine, the poisonous alkaloid compound derived from the South American plant Ryania speciosa, can potently stimulate or inhibit Ca2+ efflux from SR vesicles loaded with the channel conferred it the name Ryanodine Receptor (RyR)24. Since each monomer contains a high-affinity binding site for ryanodine in its pore-forming region67, this alkaloid has been instrumental in investigating the identity, structure, function, and levels of expression of RyR24,25. Ryanodine preferentially binds the channel in its open state, however, changing the concentration of the alkaloid in the surrounding environment yields marked differences in channel behavior. At low (nM)-activating concentrations, where even a single ryanodine molecule can occlude part of the pore, ryanodine “locks” the channel in a sub-conductance state; while at concentrations >100µM, ryanodine fully inhibits Ca2+ release24. Takeshima et al. cloned the receptor isolated from rabbit skeletal muscle in 198968. RyRs form homotetrameric assemblies, with each monomer consisting of ~5000 amino acids26,66. With an overall mass of ~2.2MDa, RyRs are the largest ion channels known to date.  There are three known mammalian isoforms of RyR. RyR1 was first detected, and is widely expressed, in skeletal muscle68,69. RyR2 was first isolated and is abundantly expressed in cardiac muscle70,71. The third distinct isoform, RyR3, was originally identified in the brain72. Although the three isoforms exhibit subtype-specific tissue expression patterns, the three isoforms are found in a variety of cell types - including neurons, exocrine cells, and lymphocytes - beyond their original tissue of identification. The three isoforms share ~65% sequence identity72, with the largest differences arising from three “divergent regions”: D1 (residues 4254-4631, in RyR1), D2 (residues 1342-1403) and D3 (1872-1932). Non-mammalian vertebrates 17  express two isoforms, RyRα and RyRβ, which are highly homologous to the mammalian isoforms73. A single isoform was identified in lower organisms, including nematodes, fruit flies and lobster73.   RyRs are mostly recognized for their involvement in EC coupling, releasing the SR Ca2+ to mobilize the contractile machinery. Their importance has been underscored in several RyR-knockout (KO) studies. Mice with a homozygous RyR1 KO died perinatally with gross skeletal muscle abnormalities74. Mice lacking RyR2 died during embryogenesis75. While a KO of RyR3 is not lethal, RyR3 KO mice demonstrate impairments in spatial learning and memory76.   1.4.1 Ryanodine Receptors and disease Given the potency of Ca2+ as a cellular messenger, unregulated or deficient Ca2+ signalling can have adverse cellular outcomes, especially in excitable cells such as skeletal and cardiac myocytes77. As RyRs play a central role in orchestrating Ca2+ release, it is not surprising that mutations in the ryr genes, which manifest in cytoplasmic Ca2+ mishandling, may develop and progress into severe muscle pathologies73,77. Indeed, over 300 disease mutations have been identified within the skeletal and cardiac RyR isoforms73. The majority of the mutations cluster in three so-called disease “hotspots” located in the N-terminal region (~ first 600 residues), a central region (residues ~2100-2500) and towards the C-terminal end of the protein (residues ~3900 to the end) (Fig. 1.4)73. The apparent organization of these disease hotspots may reflect a certain sequencing bias since mutations are increasingly being identified in regions outside of these hotspots73.   18   Figure 1.4 Disease hotspots within RyR1 and RyR2. A linear view of the RyR sequences with each vertical line representing a disease mutation. Shaded areas designate domains A, B, C, SPRY1, SPRY2, SPRY3, Repeats 1 and 2, the Handle, Helical and Central domains (containing the EF-hand domain and U-motif), and the pore-forming region. Red dotted boxes indicate the approximate location of the three divergent regions (D1-3). Non-shaded regions represent alpha-solenoid regions.    1.4.1.1 RyR1 Channelopathies Malignant hyperthermia (MH), the first identified RyR channelopathy, is a pharmacogenetic disorder triggered by the combination of a RyR1 mutation, inherited in an autosomal dominant fashion, and an external trigger, such as the administration of a volatile anesthetic, an example of which is halothane, or succinylcholine, a muscle relaxant73,77. Although the exact prevalence of MH susceptibility is difficult to gauge, as many as 1 in 2000-3000 individuals may be susceptible to anesthesia-induced hyperthermic episodes, and as such, this disease continues to be of major concern for anesthetic-induced death in otherwise healthy individuals77. Symptoms of MH present as hyperthermia to marked degree, tachycardia, tachypnea, increased carbon dioxide production and oxygen consumption, acidosis, severe 19  muscle rigidity, and rhabdomyolysis (breakdown of damaged muscle tissue) – all of which are related to a hypermetabolic state78. The pathophysiological changes of MH are due to an uncontrolled rise of myoplasmic Ca2+ following anesthesia, which activates the contractile machinery77,78. Continuous muscle contractions quickly deplete ATP, cause acidosis and compromises the muscle membrane integrity leading to hyperkalemia and rhabdomyolysis78. Unless treated, this condition can quickly be fatal. Fortunately, most operating rooms are supplied with dantrolene, a clinically approved drug to treat MH79. The protective effect of dantrolene occurs by the drug decreasing the intracellular Ca2+ concentration73 through inhibition of SR Ca2+ release80. Beyond this knowledge, dantrolene’s precise mechanism of action has been fraught with controversy. Several studies suggest a direct interaction between dantrolene and RyR181,82, and experiments with heterologously expressed RyR1 in HEK293 cells alluded to a mechanism whereby the drug inhibits store overload-induced Ca2+ release (SOICR)83. RyRs are sensitive to the SR luminal Ca2+ levels, thus, SOICR describes a process where RyRs are able to open spontaneously, usually due to Ca2+ overload in the SR73. A more recent collaboration between Knollmann and Laver (2015) has concluded that CaM binding to the RyR is essential for dantrolene’s inhibition of RyR1 and RyR280, albeit, the mechanism by which CaM facilitates dantrolene inhibition remains unclear.   RyR1 mutations are also associated with a number of rare congenital myopathies, such as Central Core Disease (CCD) and Multi-mini core disease77. These similar diseases exhibit autosomal dominant and recessive modes of inheritance and are defined by the characteristic presence of metabolically inactive cores in the center of muscle fibers73,84. The cores are completely devoid of mitochondria and evidence of oxidative metabolism. While the pathological significance of these cores is unclear, the most severe cases involve pronounced 20  muscle weakness, decreased muscle tone and skeletal anomalies77. Unlike MH, the symptoms of CCD and Multi-mini core disease appear in the absence of external triggers73.    1.4.1.2 RyR2 Channelopathies Irregularities in Ca2+ handling in cardiac myocytes are also responsible for human disease. Inheritable arrhythmogenic disorders, triggered by emotional or physical stress have been linked to mutations in RyR277. In lieu of any known structural heart defects, the hallmark of these diseases is the polymorphic ventricular tachycardia observed in patients’ electrocardiograms during cardiac stress tests77. Catecholaminergic polymorphic ventricular tachycardia (CPVT) leads to bidirectional ventricular tachycardia during exercise or stress, increasing patient’s susceptibility to syncope and sudden cardiac death73. The catecholamines released during exercise or stress trigger a β-adrenergic response, which correlates to an increase in RyR2 activity via direct phosphorylation of the channel or through phosphorylation of proteins that modulate RyR2 activity (Fig. 1.1)85. CPVT mutations increase susceptibility of RyR2 Ca2+ leak, which activates the Na+/Ca2+ exchanger (NCX) on the plasma membrane77. The activated NCX propagates the electrogenic influx of three sodium ions in exchange for the efflux of one calcium ion. Influx of positive charge generates inwardly depolarizing currents causing delayed after-depolarizations, leading to the arrhythmia. Since CPVT appears to be the cause of sudden death at a young age, early diagnosis is of paramount importance85. Current treatment regimes include β-blocker therapy to blunt the effects of catecholamines77. Interestingly, mutations within proteins that interact with RyR2 - calsequestrin86, triadin87 and CaM88,89 - have also been shown to give rise to the CPVT phenotype. These findings further demonstrate that altered SR calcium ion handling is the root cause of the arrhythmia77. 21   Mutations within RyR2 have also been implicated with arrhythmogenic right ventricular dysplasia, which results in the progressive replacement of the right ventricle muscle with fatty and fibrous tissue90.  The effects of RyR2 mutations in the brain, where the channel is also widely expressed, have also been reported: Lehnart et al. have studied mice with a heterozygous mutation in RyR2 that exhibited generalized tonic-clonic seizures91.   Comparing the nature of the disease mutations associated with RyR1 and RyR2, it appears that most RyR disease mutations cause a gain-of-function phenotype, although exceptions to this observation have been reported92. RyR mutations cause premature or prolonged, “leaky”, release of Ca2+ in the cytoplasm73. CPVT-associated mutations enhance RyR2’s sensitivity toward and decrease the threshold for activation by luminal Ca2+, thereby increasing with propensity for SOICR to occur92. Furthermore, studies have demonstrated that the disease mutations increase channel sensitivity toward activating agents73. Specific domain interactions within the RyR regulates channel gating, therefore defective regulation of interdomain interactions within the RyR has been hypothesized to play a key role in the pathogenesis of diseases93,94. In one model, Liu et al. suggest that a direct interaction exists between the channel’s N-terminal regions and the central disease hotspot. During channel opening, this interaction is “unzipped” through allosteric coupling. Thus, any disease mutations clustered at this interface weakens the interaction and facilitates channel opening. Acquired muscle pathologies, including skeletal muscle fatigue and heart failure, have been associated with alterations in RyR post-translational modifications and remodelling of the RyR channel macromolecular complexes 77.   Although RyR channelopathies cause significant human diseases, studying the underlying mechanisms of mutations not only grants us a greater understanding of disease propagation, and 22  novel therapeutic targets, but channelopathies provide model systems that can be studied to elucidate important structure-function relationships of these ion channels77.  1.4.2 Structure of Ryanodine Receptors While the massive size of RyRs has obstructed crystallographic studies of the channel as a whole, its massive size and stability as a detergent-purified protein in aqueous solution has made RyRs attractive targets for early negative-stain and, later, cryo-EM studies95. Due to the high abundance of RyR1 in the SR membrane and its relative ease of purification, it is the isoform that has received the most intensive study. Early structural work relied on thin-section or negative-stain electron microscopy to visualize the sub-cellular localization of RyRs along with other proteins that compose the EC coupling machinery. These micrographs captured square-shaped, electron-dense protrusions from the SR membrane14. Comparison with micrographs of purified RyRs revealed that the so-called “junctional-feet” were composed of four subunits (or protomers) exhibiting four-fold symmetry26,27,96,97.  Sections of the terminal cisternae, cut tangential to and intersecting the plane of the junctional face, revealed a two-dimensional array of the RyRs96–98. These micrographs illustrated an interesting feature of how the channels associate with each other in a physiological context. Rather than being connected at each pointed corner of the square, the channels overlapped with each other in an off-kilter checkerboard-like lattice of alternating square feet-shaped structures. In skeletal muscle preparations, the RyR tetramers, in their two-dimensional arrays, were strictly aligned with CaV tetrads in the opposing T-tubule membrane14. This finding provided further evidence of a physical coupling between the channels of the plasma and SR membranes.  23  1.4.2.1 RyR structural insights obtained by cryo-EM The first cryo-EM three-dimensional (3D) reconstitutions of the closed channel, at nominal resolutions of ~37-30Å99–102, provided enough details to identify the basic architecture of the channel. From a side-view, the channel is analogous to a mushroom: it has a large cytoplasmic region, encompassing ~80% of its volume, which forms the cap, and a smaller transmembrane region forming the stalk73,102. The cytoplasmic channel face measures ~270x270x100Å, whereas the transmembrane region measures 120x120x60Å73. Further description was given to the globular-shaped domains in the cytoplasmic region of the channel: masses at the corners were termed “clamps”, whereas areas which interconnected the four clamps were designated “handles”102 (Fig1.5). Assessing the 3D structure of the channel from different orientations showed a remarkably “empty” channel102. Indeed, the large cytoplasmic cap does not form a rigid block, instead the numerous globular masses, which may correspond to individual or groups of folded domains, are interspersed among many solvent-filled cavities73,102.  Advances to understanding the structural assembly of RyR1 was made in 2005, when two structures of the closed channel were solved at a nominal resolution of ~9.6-10Å103,104. These structures provided insights into the membrane-spanning α-helices and pore architecture. The structure by Samso et al. showed that the cytoplasmic and transmembrane regions are connected  via four, interconnected tubular columns104. To aid with description of the expansive cytoplasmic assembly, globular portions were designated identifying numbers, often referred to as “subregions”, and a “central rim” was assigned to the area which surrounds a central cavity 105. Comparison to the functionally homologous inositol 1,4,5-triphosphate receptor (IP3R) suggested structural homology between the two channels, particularly between the channels’ N-terminal region105. In 2009 Samso et al. published structures of the channel in both the open and 24  closed conformation, which provided a platform to assess conformational changes incurred upon channel gating. Channel opening requires coordinated movement between the cytoplasmic and transmembrane domains106. The arrangement of the inner transmembrane helices is similar to various tetrameric ion channels (e.g. K+ channels)106.                   Figure 1.5 Overall structure of the RyR. The ~3.8Å cryo-EM (EMD: 2807) reconstruction of RyR1 in the closed-state. A. Top view from the cytoplasm, looking toward the SR/ER. B. Side view from the plane of the SR membrane. Labels show the identifying structural elements. Subregions, assigned according to Serysheva et al.105, are numbered. Channel dimensions are as labelled.  25  Although the early RyR1 cryo-EM structures provided a wealth of information regarding the channel’s structural assembly and gave initial insights into channel gating, structural ambiguities remained, particularly in determining the number of transmembrane helices73. In 2015 RyR1 joined the cryo-EM resolution revolution: three structures were obtained for the mammalian RyR1, in the closed state, at resolutions ranging from 6.1 to 3.8Å107–109. Significant differences in local resolution between the various regions of the channel can be observed in all three cryo-EM maps. The highest resolution (~3Å), was registered at the channel’s pore, where most of the sidechains could be assigned de novo. Larger dynamic movements could account for the decreased resolution towards to the periphery of the cytosolic cap, where sequence registration and connectivity remain vague.   The RyR1 structure can be organized into three main regions: the N-terminal or cytosolic “shell”, the core solenoid/central domain, and the transmembrane domain95,107–109.  The cytosolic shell is by far the largest region of the channel, and is comprised of the N-terminal A, B, and C domains, three SPRY domains, two tandem repeat (phospho-) domains and the so-called handle domain, now referred to as the junctional and bridging solenoids. Docking previously determined high-resolution crystal structures110–113 into lower resolution regions of the EM structures aided domain localization and minimized discrepancies among the models. Built upon a scaffold of 37 α-solenoid repeats, the expansive cytosolic shell provides a platform for several RyR modulatory proteins and is involved in allosteric control of channel gating73,95. The cytosolic shell and channel pore is connected via the rigid core solenoid, which is involved in the allosteric transmission of information between the cytoplasmic assembly and pore region. These near-atomic resolution structures have clearly delineated the membrane-spanning architecture of the channel. The pore region resembles the six-transmembrane (6TM) cation 26  channel superfamily95,107–109. TM segments S5 and S6 form the pore and are connected to the S1-S4 segments via a juxta-membrane S4-S5 linker. S6 forms the ion-conduction pathway. Akin to the selectivity filter of Na+ and K+ channels, a short pore helix and an extended segment form a narrow path at the luminal side of the pore. Negatively charged residues on each protomer at the selectivity filter and luminal loops of the TM domain create a negatively charged “sink” that probably serves to concentrate calcium ions close to the pore mouth95. Acidic residues form negatively charged rings on the cytosolic side of the channel, further contributing to the negative charge of the region. Resembling other Ca2+-conducting channels, carboxylic sidechains (Asp and Glu) and carbonyl groups from backbone residues line the pore. The size and architecture of TM helices S1-S4 is similar to the voltage-sensors of members of the 6TM family. Since all but one of the positive charges that define voltage-sensors are missing, this region has been termed a “pseudo-voltage sensor domain”. The RyR1 C-terminal domain (CTD) is found at the cytosolic end of the S6 segment and is engulfed by the core solenoid. Since many RyR1 activators have been found surrounding the CTD, it could play a role in transmitting conformational changes from the cytosolic assembly to the pore upon activator binding95.  More recently, cryo-EM structures of RyR167,114 and RyR2115 have been obtained capturing the channels in multiple functional states and revealing long-range allosteric conformational changes incurred during channel gating and ligand-dependent activation. Activating ligands, such as ATP and Ca2+, induces a “primed” conformation – mediating changes in the cytoplasmic cap without causing pore dilation. Channel gating involves both global changes in the cytoplasmic assembly and local changes in the transmembrane domain. Bending of the S6 transmembrane helix, displacement, and deformation of the S4-S5 linker and conformational changes in the pseudo-voltage-sensor domain all contribute to pore dilation.  27   Comparison of cryo-EM reconstructions of the three RyR isoforms (albeit at low resolution for RyR3), reveal that the overall shape of the channel is conserved107–109,115,116.  1.4.2.2 RyR N-terminal domains Prior to the advent of cryo-EM technological advances, high-resolution studies on the RyR was restricted to crystallographic structures of the individual channel domains. The first crystal structures described an individually folded domain located in the N-terminal region of RyR1117,118 and RyR2118. Later studies described a larger N-terminal region of RyR1110 and RyR2119 encompassing the first three domains. These three domains have been designated domains A, B, and C. In both RyR1 and RyR2, the ABC domains are approximately 550 amino acids in length. Domains A (residues 1-205, rabbit RyR1) and B (residues 205-394) form β-trefoils, each containing twelve β-strands; domain C (residues 395-532) consists of a five-helix bundle (Fig. 1.6A). The RyR1 and RyR2 ABC domains form a compact triangular arrangement, with high structural homology between the two isoforms, and represent the bulk of the N-terminal disease  hotspot73. Docking the ABC domains of both RyR isoforms into several RyR1 cryo-EM maps revealed that this disease hotspot is located on the cytosolic face of the channel, forming a vestibule around the four-fold symmetry axis (Fig. 1.6 B and C). In addition to the domain-domain interfaces between the ABC domains, they also form extensive interactions with the remainder of the channel110. Although far from the transmembrane region, the pseudo-atomic model reveals that the N-terminal domains are connected to the pore via the four tubular columns.    28  Figure 1.6 The N-terminal ABC domains and their location within RyR1. A. High resolution crystal structure of the RyR1ABC domains (PDB: 2XOA). Domains A (green), B (blue), and C (red) have been indicated. B. Side view of the docked location of RyR1ABC in the 9.6Å cryo-EM map (EMD: 1275). C. View from the cytoplasmic side towards to endoplasmic reticulum shows that the RyR1ABC domains are located within the central rim region of the channel. The domains form a four-fold symmetric vestibule on the channel’s cytosolic face. Docking used the 3.8Å cryo-EM map (EMD: 2807).  The RyR1 and RyR2 ABC domains harbour over 80 disease mutations, mostly point substitutions, which can be located on the crystal structures and pseudo-atomic model. Interestingly, the majority of mutations are clustered at domain-domain boundaries: either in between the three domains or at the interfaces between neighboring protomer domains110,119. Since the N-terminal region is allosterically coupled to channel opening, it has been suggested that domain interactions may be disrupted during channel opening73. Thus, mutations located at these functional interfaces weaken domain-domain contacts, facilitating channel opening. Several mutations are buried within the individual domains. In these cases, the disease phenotype may be a result of the mutation affecting the local domain fold. Although there is high structural similarity between the ABC domains of RyR1 and RyR2, a unique anion binding site has been identified within RyR2ABC119. Nestled at the interface of the three domains, biochemical assays 29  have hinted at the necessity of chloride binding for the stability of the RyR2ABC domains. RyR2 mutations which target this chloride-binding site, not surprisingly, have repercussions on the overall channel function119. The location of the N-terminal disease hotspot on the cytosolic face of the channel reinforces the notion that allosteric modulation plays a critical role in pore gating95.  1.4.3 Regulation of Ryanodine Receptors In theory, a few transmembrane α-helices would suffice to make a calcium-selective pore. An explanation for the evolution of such an expansive cytoplasmic channel assembly, undoubtedly, lies in channel modulation. Due to the potency of Ca2+ as a cellular messenger, its entry into the cytoplasm via RyRs needs to be strictly controlled. The cytoplasmic mass and many solvent-filled cavities of the RyR provides ~500,000Å2 of surface area (including the transmembrane area) onto which a multitude of proteins and small molecules can dock and post-translational modification events can occur73. In addition to the voltage-gated calcium channels discussed before, RyR regulators include binding partners from both the cytoplasmic and SR luminal portions that can function by both stimulating or inhibiting channel function73 (Fig. 1.7). As such, RyRs have evolved to become sensitive signal integrators. As the list of RyR effectors is far too extensive, the effect of a subset regulators will be discussed below.       30   Figure 1.7 RyR regulation by small molecules, protein binding-partners and post-translational modification. A schematic overview of the RyR and L-type voltage-gated calcium channel (CaV), which are located on two different membranes. In skeletal muscle the CaV, via its II-III loop, physically interacts with RyR1, stimulating SR Ca2+ release. In cardiac muscle, RyR2 is activated via calcium-induced calcium release. RyR regulators can either stimulate (+) or inhibit (-) channel activity. A subset of regulators (Ca2+, CaM, Homer, and Ryanodine) have both activating and inhibitory effects on the channel. Regulatory molecules can exert their effects from the cytoplasmic and luminal side of the channel.   31  1.4.3.1 RyR regulation by Ca2+ RyRs not only conduct calcium ions, but they are also highly sensitive to Ca2+ regulation. In cardiac myocytes, RyRs are activated by calcium-induced calcium released (CICR)28,120. Depolarization of the plasma membrane triggers an influx of extracellular Ca2+ through the L-type voltage-gated calcium channels. The rapid change in cytoplasmic Ca2+ is sensed by calcium ion sensors within RyR2, which bind Ca2+ and mediate channel opening28. The resulting release of SR Ca2+ is a regenerative process120, further stimulating SR Ca2+ release. In this capacity the RyR acts as a signal amplifier. However, as the cytoplasmic Ca2+ levels begin to elevate, a negative feedback response is activated, whereby Ca2+ triggers closing of RyRs. This behaviour of RyRs alludes to multiple calcium-binding sites with different affinities and binding kinetics73. Calcium ion efflux assays and planar lipid bilayer studies revealed that the open probability (Po) of RyRs exhibit a biphasic-dependence on cytosolic Ca2+ concentration. Highest Po was recorded in the 10-100µM range with decreased channel activity at both higher and lower levels of Ca2+ 121,122. Additionally, RyRs can sense the Ca2+ concentrations in the SR/ER. Under conditions when the SR store Ca2+ content reaches a critical level, RyRs can open spontaneously and undergo SOICR123,124.  Calmodulin, the ubiquitous cellular Ca2+-sensor, is capable of associating with the RyR in both its apo- and Ca2+-loaded states, thereby fine-tuning the effect of Ca2+ on the RyR73. Early preparations of RyRs pointed to a functional role of CaM on RyR modulation as the two proteins were found to co-purify together125. Several groups have demonstrated that CaM binds directly to RyR and modulates the channel in both in vivo and in vitro assays126–129. RyR modulation by CaM depends on the concentration of Ca2+ and shows isoform-specific differences73. At high levels of Ca2+ (micromolar concentrations), CaM inhibits RyR1 and RyR2130,131. Conversely, at 32  low (sub-micromolar) Ca2+ levels, CaM activates RyR1 but inhibits RyR2126,131–133. These effects have been nicely captured in single channel recordings reported by Yamaguchi et al. (2004). In the presence of 50nM CaM and low Ca2+ levels (0.4µM), the Po for RyR1 is doubled, while the Po for RyR2 is decreased almost two-fold130. A higher Ca2+ concentration (2µM) decreased the Po for both channel isoforms. Cryo-EM reconstructions have mapped the binding locations of apo- and Ca2+-CaM on RyR1134. Interestingly, the location of CaM binding on the channel changes upon binding to Ca2+, providing a structural basis for the differential functional effects of CaM at various Ca2+ levels. Systematic attempts to explore CaM modulation on RyR2 demonstrated that CaM mutations which abolish Ca2+ binding to the N-lobe increases the channel’s termination threshold (i.e. facilitates termination)135. Mutations which diminish Ca2+ binding to both N- and C-lobe, or C-lobe alone, delayed termination of the SR Ca2+ release. Collectively, these results are reflective of individual, and possibly distinct, binding sites for CaM’s lobes on the channel. Currently, there is one crystal structure available (PDB: 2BCX) depicting the complex between Ca2+-CaM and a RyR1 peptide (3614-3643, mouse RyR1). Although the structure shows both CaM lobes wrapped around the amphipathic α-helix, nuclear magnetic resonance (NMR) data suggests that the Ca2+-N-lobe is only weakly associated136, leaving open the possibility that the N-lobe can bind to a different segment within the full-length channel since several other peptides have been found to bind CaM73. The relevance of proper RyR modulation by CaM is underscored by the fact that several CaM mutations have been associated with CPVT89.    33  1.4.3.2 RyR regulation by protein-binding partners: FKBPs The FK506-binding proteins (FKPBs) are well known RyR protein-binding partners, particularly since they were found tightly associated with RyRs isolated from skeletal muscle SR preparations137. FKBPs are cytosolic receptors for the immunosuppressive drug FK506. Two isoforms have been of particular interest to the RyR-field: FKBP12 (named based on its molecular masses of 12kDa) and FKBP12.6 (12.6kDa). Although both FKBP12 and FKBP12.6 can associate with all three RyR isoforms138, FKBP12 typically binds RyR1 and FKBP12.6 binds RyR2.  FKBP12 stabilizes and coordinates the gating activity of RyR1: removal of FKBP12 caused the channel to enter sub-conductance states and to have higher open probabilities, resulting in increased Ca2+ leak139. The RyR2-FKBP12.6 association has been of intense interest since dissociation of FKBP12.6 from RyR2 has been implicated in heart failure and arrhythmias140,141. Low resolution cryo-EM reconstitutions106,142,143, combined with fluorescence resonance energy transfer (FRET) measurements144 demonstrated that FKBP12 and FKBP12.6 bind in the same position and orientation to the periphery of the cytosolic cap, at the junction between the “clamp” and “handle domains”. More recent crystallographic, cryo-EM and FRET insights have revealed that FKBP interacts with the RyR SPRY1 domain, located in the clamp region of RyR113. Molecular dynamics flexible fitting and mutagenesis experiments suggest that a hydrophobic cluster within SPRY1 is crucial for FKBP binding. Not surprisingly then, a mutation which directly impacts FKBP binding has been implicated in central core disease.  34  1.4.3.3 RyR regulation by post-translational modification: phosphorylation  RyR activity is regulated by various kinases (PKA, PKG, and Ca2+/CaM-dependent protein kinase II, CaMKII) and phosphatases (PP1, PP2A and PDE4D3)73. The channel’s cytoplasmic assembly serves as a scaffold for several of these enzymes, which allows for specific and compartmentalized regulation95. Extensive interest has surrounded RyR phosphorylation by protein kinase A (PKA) as this provides a connection between RyR2 activation and physiological stress. Catecholamines released during exercise or stress trigger a β-adrenergic response, increases RyR2 activity via direct channel phosphorylation or phosphorylation of RyR2 modulatory proteins85. The role of phosphorylation and heart failure has been controversial. Within RyR2, Ser-2030 and Ser-2808 can be phosphorylated by PKA. One group has reported that PKA phosphorylation of RyR2 causes dissociation of FKBP12.6, increasing the channels’ open probability140. Further, in heart failure, Ser-2808 was found to be hyper-phosphorylated, resulting in increased sensitivity to Ca2+-induced activation and defective channel function. However, several groups have failed to observe PKA hyper-phosphorylation in heart failure and did not see FKBP dissociation by RyR phosphorylation145,146.  RyR regulation by protein-binding partners, small molecules and post-translational modification events reinforces the importance of allosteric regulation in channel activity. By proving a large binding scaffold, these modulators are able to impact the channel’s open probability by regulating conformational changes in the pore and surrounding TM segments95.   1.5 Advances in understanding EC coupling: new players in the EC coupling machinery Experiments on dysgenic (CaV1.1-null) myocytes and myocytes which are genetically null for RyR1 (dyspedic) revealed that CaV1.1 responds to T-tubular voltage changes, while 35  RyR1 gates SR Ca2+ release51,74. For skeletal muscle, multiple structural and biochemical studies have established that RyR1 and CaV1.1 are mechanically coupled22,27,51,53,74 and are involved in bidirectional signalling54–56. Due to the stringent conditions under which EC coupling operates, triad junctions contain a large complex of proteins, only some of which are directly involved in the transduction process147. This complexity, combined with the fact that SR Ca2+ release depends on the interaction between two separate membrane systems, has marred the use of reductionist approaches (such as heterologous expression systems) to define the molecular mechanism of EC coupling Ca2+ release147. Although the CaV-α1s and β1a-subunits and RyR1 are the essential and most well-studied proteins of the complex, the roles of other members of the complex remains to be defined, particularly to determine how the complex assembles at triadic junctions and how EC coupling is regulated.   Other core proteins include FKBP12, triadin, junctin, calsequestrin and SepN1148,149. FKBP12 has been discussed in section 1.4.3.2. Triadin, junctin, calsequestrin and SepN1 are SR proteins. Although knock-outs of these five proteins are non-lethal, genetic changes in these proteins are associated with myopathies, indicating their role in normal muscle function149,150. Triadin, junctin and calsequestrin maintain the ultrastructure of the membranes containing the RyR1 core complex for optimal function during EC coupling. The selenoprotein, SepN1, is required for the formation of slow twitch in zebrafish muscle151. CaM is a cytosolic protein which interacts with the EC coupling complex by binding both RyR and the CaV α1-subunit152, thereby modulating EC coupling148. The identification of these components hints at the sheer complexity of the EC coupling complex and poignantly demonstrates a dearth in understanding EC coupling regulation.   36  1.5.1 STAC3: an essential component in skeletal muscle EC coupling Genetic defects in the EC coupling machinery underlie numerous congenital myopathies, increased susceptibility to the pharmacogenetic disorder known as Malignant Hyperthermia and have been linked to life-threatening cardiac arrhythmias. Despite efforts to understand the nature of these debilitating myopathies, their pathologies remain largely elusive. The high heterogeneity of congenital myopathies means that the genetic basis of many disorders is unknown148.  1.5.1.1 Native American Myopathy In 1987, Bailey and Bloch first described a congenital myopathy, with susceptibility to malignant hyperthermia, in a three-month old American Indian infant of Lumbee descent153. The infant had multiple congenital anomalies including cleft palate, micrognathia (undersized jaw), talipes equinus (a foot deformity, where the sole is permanently flexed) and arthogryposis (stiffness and limited range joint motion). Since this initial diagnosis, many more individuals of Lumbee descent have been identified with similar congenital anomalies, giving rise to the name of this myopathy: Native American Myopathy (NAM). The Lumbee people are a multiracial group indigenous to Robeson County in south-central North Carolina. Additional features have since been associated with NAM: congenital muscle weakness, short stature, ptosis (drooping of the upper eye-lid), and kyphoscoliosis; however, it was MH susceptibility which warranted NAM of medical concern154,155. Pedigree analysis revealed an autosomal recessive mode of transmission154,155. Attempts to understand the genetic basis of this disease revealed that a missense mutation (W284S) in STAC3, a putative muscle-specific adaptor protein, was responsible148,154,155.  37  STAC3 is one of three isoforms of the STAC family of adaptor proteins. The family’s name has been derived from the src homology 3 (SH3) and cysteine-rich domains comprising proteins. Quantitative PCR analysis of mice has determined that STAC3 expression is highly restricted to skeletal muscle throughout development and into adulthood156–158. Low levels of STAC3 was detected in the cerebellum, forebrain, and eye. While STAC1 and STAC2 transcripts were excluded from skeletal muscle, high levels of STAC2 were detected in the cerebellum, fore- and mid-brain, and eye. STAC1 was also detected in the three neuronally-rich areas, albeit at lower levels than STAC2, as well as in the bladder and adrenal gland. Interestingly, the transcripts of the STAC family were absent in cardiac tissue156.  Quests to identify novel factors required for skeletal muscle development and function led to the discovery of STAC3, which was determined to be required for myogenic differentiation157,158. Published almost simultaneously, animal studies with STAC3 knock-outs and mutants established that this previously uncharacterized adaptor protein plays a vital role in skeletal muscle EC coupling148,156. Nelson et al. (2013) revealed that homozygous deletion of STAC3 in mice resulted in complete paralysis and perinatal lethality. Compared to wildtype (WT), KO mice exhibited a range of musculoskeletal defects. The gross morphology of KO neonates resembled mice with severe defects in muscle differentiation, neuromuscular junction formation or function, or EC coupling. Anatomical and histological comparison between WT and KO neonates revealed several interesting findings: KO lungs had a marked reduction in alveoli size, their diaphragm muscles were smaller and more translucent, and their lungs were not buoyant in saline. Although paralyzed, hearts from the KO neonates did beat for several minutes following birth – suggesting the cause of death in these animals was due to asphyxiation. Since muscle contractility and SR Ca2+ release of cultured myotubes from STAC3 mutant mice could 38  be restored with the application of a RyR agonist (4-chloro-m-cresol), it indicated that the defect did not reside in the contractile apparatus, SR Ca2+ storage, and that the ryanodine receptors were functional. The authors finally concluded that loss of STAC3 results in a blockade in skeletal muscle EC coupling.   An autosomal recessive mutation, mi34, was identified in zebrafish which results in mutants displaying defective motor behaviours (decreased amplitude of body coiling and ineffective touch-induced escape and swimming) and death in their early developmental stages148. Since tactile stimulation elicited synaptic responses in both slow and fast twitch muscles equally in WT and mutant embryos, Horstick et al. (2013) determined that the aberrant behaviour in mutants was due to skeletal muscle, and not nervous system, defects. Depolarization-evoked muscle contraction revealed that mutant muscles contracted significantly less than WT muscle. Since exposure to RyR agonists were able to restore muscle contraction in mutants, it suggested that the defect did not affect the function of the contractile machinery nor was the SR Ca2+ store grossly perturbed. The efficiency of EC coupling release was examined in vivo by imaging Ca2+ transients in skeletal myofibers expressing a Ca2+ indicator during swimming. Calcium ion transients were reduced in mutant slow and fast twitch fibers – suggesting that EC coupling was defective in the mi34 mutants. Of interest, no changes in the triadic junction anatomy was observed. Genetic analysis revealed that the STAC3 gene is responsible for the mutant. Labeling with an anti-STAC3 antibody confirmed skeletal muscle-specific expression and showed co-localization of STAC3 with the α1-subunit of CaV in WT, but not mutant, muscle triads. These findings synonymously illustrated that the STAC3 mutation was responsible for the defective EC coupling and established that STAC3 is an essential component of the skeletal muscle triadic molecular complex. 39  1.5.1.2 Defining the role of STAC3 as a member of the EC coupling machinery The structural and functional connectedness between CaV1.1 and RyR1 suggests that the trafficking of CaV1.1 to the plasma membrane must be precisely regulated to form the EC coupling triad junctions. CaV1.1 expresses poorly, or not at all, in mammalian cells that are not of muscle origin159. In comparison, heterologous expression of the closely related cardiac/neuronal CaV1.2 is not perturbed in such cells160,161. This finding led Polster et al. (2015) to hypothesize that there must be a skeletal muscle-specific factor which facilitates CaV1.1 expression. STAC3, with its restricted expression in skeletal muscle, was of particular interest to investigate. Using fluorescently-labeled constructs, it was observed that CaV1.1 and STAC3 trafficked together to the plasma membrane of tsa201 cells. In contrast, when expressed in the absence of STAC3, CaV1.1 was retained in the endoplasmic reticulum. Photobleaching experiments demonstrated that STAC3 does not intrinsically localize to the plasma membrane, but instead, it also depends on CaV1.1 for plasma membrane expression. Moreover, STAC3 promoted the functional channel expression at the plasma membrane. As discussed in section 1.3.2, the CaV-β- and α2δ-subunits play an enormous role in channel trafficking. Remarkably though, even in the absence of CaV-β and α2δ, STAC3 was able to support insertion of CaV1.1 into the plasma membrane. These results strongly hinted at the direct interaction between CaV1.1 and STAC3 that is required for the efficient trafficking of the channel to the plasma membrane162.  The next logical step was to establish a functional role of STAC3 in the crucial CaV1.1-RyR1 interaction. Polster et al. (2016) probed into this using STAC3 KO myotubes. As an addendum to their previous findings162, it was noted that STAC3 was not essential for CaV1.1 trafficking to the plasma membrane. Indeed, in STAC3 KO myotubes, the channel’s plasma 40  membrane expression was sustained by the co-expression of the CaV-β1a-, α2δ- and γ-auxiliary subunits48. However, CaV1.1 expressed in STAC3 KO myotubes were unable to mediate EC coupling and produced only small, rapidly inactivating Ca2+ currents. Introduction of WT STAC3 into the KO myotubes restored both channel functions (coupling and calcium ion conduction). In contrast, expression of the STAC3 NAM mutation in the KO myotubes resulted in partial recovery of normal Ca2+ current, but only a very weak restoration of EC coupling. These results reveal that STAC3 is involved in both functions of CaV1.1, and that these two effects of STAC3 rely on different regions of the protein, which function partially independent of each other48.   1.5.2 Defining the core components of the EC coupling machinery Given the demanding conditions – high firing rates, hypoxia, and metabolic acidosis – in which EC coupling operates, it is not surprising that the triad junction contains a large army of proteins147. Teasing out the essential components of the EC coupling machinery has been challenging. Suda et al. (1997) attempted to reconstitute EC coupling Ca2+ release in Chinese Hamster Ovary (CHO) cells by co-expressing a CaV1 construct, CaV auxiliary subunits, and RyR1. This system was hindered by two main problems: 1) low-level CaV1.1 expression, forcing the authors to make use of a chimeric construct that was largely (>90%) composed of CaV1.2; 2) the essential ER-plasma membrane junction did not form in the transfected cells163. The essence of skeletal muscle EC coupling was lost when these cells produced only very slow cytoplasmic Ca2+ release which depended on extracellular Ca2+ entry.  A more recent attempt by Perni et al. (2017) to reconstitute skeletal muscle EC coupling in a heterologous system was successful. Reconstitution of conformational coupling in tsa201 41  cells was achieved by expression of CaV1.1, CaV-β1a, STAC3, RyR1 and junctophilin2 (JP2)147. As previously demonstrated, STAC3 aided in the robust plasma membrane expression of CaV1.1. Co-expression of JP2 was key to the formation of ER-plasma membrane junctions, which contain RyR1, and that are morphologically similar to SR-plasma membrane junctions in native muscle tissue. Comparable to muscle, depolarization of this de novo reconstitution elicited Ca2+ transients independent of extracellular Ca2+ entry and having amplitude with a saturating dependence on voltage. Freeze-fracture microscopy revealed that these five proteins were sufficient to replicate the ER-plasma membrane junctions observed in native muscle tissue, where CaV1.1 and RyR1 can interact. Remarkably, this assembly allowed CaV1.1 to be arranged in tetrads indicative of physical links to RyR1. These results succinctly identify the core components required for skeletal muscle EC coupling.   1.6 Research Question Excitation-contraction coupling has been an area of immense interest. Despite the fantastic advances achieved by the field – understanding the ultrastructure of the triadic junction and identifying the core components of the EC coupling machinery – there is still much work to be done to fully elucidate the mechanism and regulation of EC coupling. It is still conceivable that we have yet to discover novel proteins involved in EC coupling, a notion made plausible given that, until recently, STAC3 was not known to be a triadic protein. Of the known elements of the EC coupling complex, there are still many questions that remain to be answered. Whether CaV1.1 and RyR1 interact with each other directly, through auxiliary proteins, or a combination of both remains to be confirmed. Given the infancy of the STAC3 field, there are many facets of the protein and its role in EC coupling which remain to be explored. The Ryanodine Receptor 42  represents a pinnacle of complexity in the ion channel field: its modulation by its many regulators remains of interest, as does deciphering the mechanism of action of disease mutations.  The second and third chapters of this thesis focuses on the STAC family of proteins. Structural information on the functional domains of the STAC proteins have largely been lacking. Part of chapter 2’s work aims to fill this void. The co-localization experiments by Polster et al. led us to explore the possibility of a direct interaction between STAC3 and CaV1.1. Both chapters employ an integrated approach of x-ray crystallography, isothermal titration calorimetry (ITC), electrophysiology, and EC coupling assays to explore the structure and function of the STAC proteins. The EC coupling assays were performed by Dr. Bernhard Flucher’s lab (Innsbruck Medical University). The fourth chapter focuses on the N-terminal region of the cardiac RyR isoform, RyR2. The unique anion binding site identified by Kimlicka et al. in the crystal structure of the RyR2ABC domains fueled many questions about its functional importance. Since an equivalent binding site is absent in the RyR1ABC domains, this chapter aims to understand the nature of the RyR2 chloride binding site. Attempts to understand the dynamic changes incurred upon chloride binding was done in collaboration with Dr. Razvan Cornea’s lab (University of Minnesota). Additionally, we explore a disease mutant which targets this binding site. In collaboration with Dr. Wayne Chen’s lab (University of Calgary), two additional RyR2 N-terminal disease mutants are studied. X-ray crystallography and thermal stability assays were employed in our explorations of the RyR2 N-terminal region.    43  Chapter 2: Structural insights into binding of STAC proteins to voltage-gated calcium channels1 2.1 Introduction A key tenet of EC coupling is depolarization of the T-tubular membrane which triggers release of Ca2+ from the SR, leading to muscle contraction17,18. For skeletal muscle, multiple studies have shown that RyR1 and CaV1.1 are coupled mechanically, with CaV1.1 serving as the voltage sensor for RyR122,27,51,53,164. Conversely, changes in RyR1 can also affect the function of CaV1.154,165. Freeze-fracture studies show particles, thought to correspond to CaV1.1 channels, grouped into tetrads, opposite foot structures corresponding to RyR127,166. However, whether CaV1.1 interacts with RyR1 directly, through auxiliary proteins, or a combination of both remains to be confirmed. The protein STAC3 has been identified as a factor required for myogenic differentiation157,158. Recently, two groups independently identified STAC3 as a novel component essential for skeletal muscle EC coupling148,156.  It belongs to a small family of three proteins in the SH3-and cysteine-rich adaptor proteins (STAC1, STAC2, STAC3). STAC3 is mostly expressed in skeletal muscle, whereas STAC1 and STAC2 are expressed in a variety of tissues, including the brain48.  Multiple roles have been identified for the STAC proteins. STAC3 allows for expression of CaV1.1 in the plasma membrane of heterologous cells162 and increases membrane expression in myotubes. Although CaV1.1 can also be expressed at the plasma membrane of heterologous cells in the presence of the CaV-γ1-subunit, the addition of STAC3 results in much higher current                                                  1 This chapter has been adapted from the original publication: Wong King Yuen, S.M., Campiglio, M., Tung, C.C., Flucher, B.E., and Van Petegem, F. Structural insights into binding of STAC proteins to voltage-gated calcium channels. Proc. Natl. Acad. Sci. U.S.A (2017) 114(45): E9520-E9528. 44  amplitudes48. Another role is in EC coupling, because myotubes from STAC3 knock-out mice, which still have some degree of CaV1.1 expression, have drastically reduced EC coupling. In addition, zebrafish embryos that are null for STAC3 still have normal levels of CaV1.1 but display highly reduced EC coupling148.  STAC3 is the target for a disease mutation (W284S) linked to Native American Myopathy (NAM). The mutation results in a large reduction in EC coupling. It was also found to reduce the recruitment of STAC3 into the skeletal muscle CaV1.1 complex167.  The STAC proteins contain three predicted structured domains, including a C1 domain near the N-terminus, flanked by regions of predicted intrinsic disorder, and two SH3 domains in the C-terminal half (Fig. 2.1A). Here we report crystal structures of individual and tandem SH3 domains of different STAC isoforms. We identify a binding site in the loop connecting repeats II and III of CaV1.1 and CaV1.2 and show the importance of this interaction for EC coupling.  2.2 Experimental Procedures 2.2.1  Expression Constructs All constructs for ITC and X-ray crystallography (see Table 2.3) were cloned into a modified pET28 vector (Novagen), containing a His6-tag, maltose-binding protein (MBP) and a cleavage site for the tobacco etch virus (TEV) protease49. Table 2.1 outlines the exact constructs used. Cloning procedures for GFP-CaV1.1 (NM_001101720) and GFP-CaV1.1 –II-IIIM were previously described168,169. Mutation of I752, P753, and R757 to alanines were introduced by splicing by overlap extension- (SOE-) PCR. Briefly, the II-III loop cDNA sequence of CaV1.1 was amplified by PCR with overlapping mutagenesis primers in separate PCR reactions using GFP-CaV1.1 as a template. The two separate PCR products were then used as templates for a 45  final PCR reaction with flanking primers to connect the nucleotide sequences. This fragment was digested with SmaI/XhoI and cloned in the respective sites of GFP-CaV1.1, yielding GFP-CaV1.1-IPRAAA. Sequence integrity was confirmed by sequencing (MWG Biotech).  Table 2.1 Human STAC and CaV1.1/1.2 constructs described in the crystallographic and ITC experiments    2.2.2 Protein expression and purification Proteins were expressed for 20-24 hours in Escherichia coli Rosetta (DE3) pLacI (Novagen) grown in 2xYT media at 18oC. Cells were lysed via sonication in 250mM KCl, 10mM HEPES pH7.4 (buffer A) supplemented with 1mM PMSF, 25µg/mL DNaseI and 25µg/mL lysozyme, 10% glycerol, 6.2mM β-mercapotoethanol (β-me) and 20mM imidazole. Lysates were applied to HisTrap FF Crude columns (GE Healthcare), washed with 10 column volumes (CVs) of buffer A plus 20mM imidazole and eluted with buffer B (containing 250mM KCl and 500mM imidazole pH 7.4). All CaV1.1 and 1.2 constructs were further purified by an Amylose column (New England Biolabs), washed with 2CVs of buffer A supplemented with 50mM CaCl2 and eluted with buffer A plus 10mM maltose. Following cleavage of both the STAC and CaV proteins with His-tagged TEV protease for 12-14 hours at 4oC, dialyzing against buffer A supplemented with 3mM β-me, the proteins were applied to a Talon column (Clontech)  STAC1 STAC2 STAC3 1st SH3 288-342 296-349 245-304 2nd SH3 not cloned 354-411 309-364 Tandem SH3’s 288-402 296-411 245-364  CaV1.1 α1s CaV1.2 α1c  Core II-III loop 728-775 829-876  Minimal II-III loop  747-760 (synthetic)   46  in buffer A and eluted with buffer B. The collected flow-through fractions were concentrated using 3kDa MWCO Amicon concentrators prior to loading onto a Superdex75 16/600 column (GE Healthcare) in buffer A supplemented with 2mM TCEP.   2.2.3 Crystallization and structure determination The STAC3 SH3-2 domain was crystallized using the sitting drop or hanging drop vapor diffusion method at room temperature by mixing equal volumes of protein (16mg/mL) and well solution, containing 0.1M Bis-Tris pH 6.5, 22.5% (w/v) PEG3350 and 0.2M ammonium acetate. Crystals were soaked in a mixture of mother liquor and 30% ethylene glycol and flash-frozen. The tandem SH3 domain construct of STAC1 was crystallized by sitting drop vapor diffusion at room temperature and mixing equal volumes of protein (10mg/mL) and well solution, which contained 0.1M sodium-HEPES pH 7.5 and 25% (w/v) PEG1000. Crystals were transferred to a cryo-solution containing the mother liquor and 35% (v/w) glycerol and flash-frozen. The WT STAC2 tandem SH3 domain construct was crystallized in a 350µL dialysis button against 150mM KCl, 10mM HEPES pH 7.4 and 2mM TCEP at 4oC. Its complex with the synthetic minimal CaV1.1 peptide (synthesized by LifeTein) was crystallized in a 1:2 ratio of STAC2:peptide by sitting drop or hanging drop vapor diffusion at 4oC by mixing an equal amount of protein and well solution which contained 2.24M (NH4)2SO4, 0.1M sodium acetate, pH 5.5. Prior to being flash frozen in liquid nitrogen, the apo- and complex STAC2 crystals were transferred to a drop containing mother liquor and 35% and 30% (v/v) glycerol, respectively. The STAC2 Q347I mutant was crystallized by hanging drop vapor diffusion at 4oC upon mixing equal volumes of protein (20mg/mL) and well solution (0.1M Bis-Tris pH 6.4, 0.2M lithium sulfate and 35% (w/v) PEG3350). The Q347I mutant was cryoprotected with 25% PEG200 and 47  mother liquor. Diffraction data was collected at the Stanford Synchrotron Radiation Lightsource beamline BL9-2, the Advanced Photon Source beamline 23-ID-D, the Canadian Light Source  beamline 08ID-1, and our homesource (Micromax 007 HF, Mar345 detector).  Data sets were processed using XDS170. Initial phases were collected by iodide-SAD for the 2nd SH3 domain of STAC3. Crystals of the 2nd SH3 domain of STAC3 were soaked in 1M NaI for 1 to 2 minutes before being frozen. Phases were determined via PHENIX171 and ARP/wARP was used to build an initial model of the structure of this individual domain. Subsequent models were refined with PHENIX using high-resolution native datasets at 1.3Å. The structure of apo-STAC2 (tandem SH3 domains) was solved by molecular replacement via Phaser172, using an unpublished structure of the 1st SH3 domain of STAC1 and the SH3-2 domain of STAC3 as search models. Structures of STAC2 Q347I, apo-STAC1 and the complexed STAC2 tandem SH3 domains were solved by molecular replacement using the structure of apo-STAC2 as a search model. All models were completed with iterative cycles of manual model building in Coot173 and refinement with Refmac5174. Table 2.2 highlights the statistics for crystallographic data collection and refinement statistics.          48  Table 2.2 Crystallographic data collection and refinement statistics for STAC crystals  STAC1 Tandem SH3 STAC2 Tandem SH3 apo STAC2 Tandem SH3 Q347I STAC2 Tandem SH3 complex STAC3 SH3-2 PDB ACCESSION CODE 6B25 6B26 6B28 6B27 6B29  DATA COLLECTION      I-SAD    Space group P212121 P21 21 21 P21 21 21 P212121 C2221 C2221    Cell dimensions             a, b, c (in Å) 28.96, 36.73, 113.00 36.86, 48.00, 73.59 35.38, 49.71, 89.67 47.92, 114.65, 144.67 49.54, 57.96, 156.82 49.54, 57.96, 156.82       α, β, γ (in o) 90.0, 90.0, 90.0 90.0, 90.0, 90.0 90.0, 90.0, 90.0 90.0, 90.0, 90.0 90.0, 90.0, 90.0 90.0, 90.0, 90.0    Wavelength 0.97946 0.97946 1.54179 0.97946 0.97949 1.7712    Resolution (in Å) 56.50 - 2.39 (2.46 - 2.39) 32.95 - 1.20 (1.24 - 1.20) 44.83-2.55 (2.63-2.55) 35.00 - 1.73 (1.78 - 1.73) 39.21 - 1.30  (1.35 - 1.30) 37.7 - 2.06 (2.13 - 2.06)    Rmerge 0.124 (0.520) 0.02397 (0.2007) 0.135 (0.769) 0.058 (0.493) 0.05199 (0.371) 0.1661 (1.104)    I/σI 7.40 (2.4) 29.89 (5.84) 6.5 (1.5) 7.80 (2.38) 20.96 (3.41) 21.08 (1.99) Completeness 98.00 (93.80) 99.33 (98.95) 99.3 (98.6) 99.10 (99.60) 98.97 (91.63) 93.66 (53.79)     Redundancy 4.0 (3.2) 4.4 (4.1) 3.4 (3.1) 3.6 (3.6) 6.7 (4.3) 24.2 (8.4) REFINEMENT          Resolution (in Å) 56.50-2.39 32.95-1.20 44.83-2.55 35.00-1.73 39.21-1.30     No. of reflections 4819 39292 5222 79187 55372     Rwork/Rfree 20.26/25.88 15.24/16.91 22.17/25.17 17.86/21.63 15.68/18.87     No. atoms             Protein 960 1009 817 5940 1830        Ligand 15   42         Water 32 240 33 744 305     B-factors             Protein 35.79 12.89 40.20 24.50 19.40        Ligand 50.33   55.94         Water 35.74 26.36 36.02 32.25 35.32     Ramachandran      favored (%) 98.23 100 98.04 98.60 97.75     Ramachandran      allowed (%) 1.77 0.00 1.96 1.40 2.25     Ramachandran    Outliers (%) 0.00 0.00 0.0 0.00 0.00    rmsd             Bond lengths (Å) 0.011 0.007 0.012 0.021 0.013        Bond angles (o) 1.55 1.31 1.54 1.94 1.22  49  2.2.4 Isothermal titration calorimetry All proteins were concentrated and dialyzed against 150mM KCl, 10mM HEPES pH 7.4, 2mM TCEP at 4oC.  The synthetic CaV1.1 peptide was dissolved in the dialysis buffer. Protein absorbances were measured at 280nm using a Nanodrop2000 spectrophotometer (ThermoFisher) and concentrations were calculated using the extinction coefficient obtained from sequence input into the Peptide Properties Calculator (Innovagen)175. Titrations consisted of 20 injections of 2µL with concentrations noted in the figure legends. Experiments were performed at 25oC and using a stirring speed of 750RPM on an ITC200 instrument (GE Healthcare). All data were processed and modeled using a single-site fitting model on Origin 7.0 and isotherms were generated following a point-by-point subtraction of a reference titration of ligand into buffer.  2.2.5 Immunolabeling Myotubes of the homozygous dysgenic (CaV1.1mdg/mdg) cell line GLT were cultured as described in Powell, Petherbridge and Flucher176. At the onset of myoblast fusion (two days after addition of differentiation medium), GLT cultures were transfected using FuGene-HD according to manufacturer’s instructions (Promega). Cultures were fixed with paraformaldehyde five days after the transfection and double immunolabeled with rabbit anti-GFP (serum, 1:10,000; Molecular Probes) and mouse monoclonal anti-RyR (34-C, 1:1000; ThermoScientific) and fluorescently labeled with Alexa-488 and Alexa-594 conjugated secondary antibodies (ThermoScientific), respectively. Samples were observed using a 63x, 1.4 NA objective Axioimager microscope (Carl Zeiss Inc.) and 14-bit images were acquired with a cooled CCD camera (SPOT, Diganostic Instruments) and Metaview image processing software. Images were 50  arranged in Adobe Photoshop and where necessary, linear adjustments were performed to correct black level and contrast.  2.2.6 EC coupling analysis Depolarization-induced Ca2+ transients were recorded in 4-5 day post-transfection cultures loaded with 5µM Fluo4-AM (ThermoFisher) plus 0.1% Pluronic F-127 in Tyrode solution (130mM NaCl, 2.5mM KCl, 2mM CaCl2, 2mM MgCl2, 10mM HEPES, 30mM glucose) for 40 min at room temperature. Calcium ion transients were elicited by passing 2-ms pulses of 40V across the 19-mm incubation chamber, a condition approximately two-fold above the stimulation threshold. In both conditions, myotubes showed a gradual response to increasing stimulation voltages (10-60V), indicating that depolarization directly activated EC coupling without eliciting action potentials. Tetanic stimulation was performed with 500ms trains of 2ms 40V stimuli at increasing frequencies between 10 and 240Hz, in 4-second intervals. To analyze the efficacy of the CaV1.1 constructs to reconstitute EC coupling, the cover glasses were systematically scanned and myotubes responding to continuous 0.5Hz stimulation at 40V were counted. Fluorescence signals from single myotubes were recorded with a PTI RatioMaster microphotometry system (Horiba Scientific). Traces were normalized by calculating the ∆F/F ratio in Microsoft Excel and analyzed with the Student’s t-test in GraphPad. Results are expressed as mean ± standard error and graphs were assembled in GraphPad.   51  2.3 Results 2.3.1 The tandem SH3 domains form a rigid interaction All three STAC isoforms have a predicted set of SH3 domains (SH3-1 and SH3-2) in their C-terminal half. We set out to solve high-resolution structures of the tandem SH3 domains of both STAC1 and STAC2, with resolutions of 2.5Å and 1.2Å, respectively (Fig. 2.1A). Each domain consists of a five-stranded antiparallel β-sheet with additional short 310 helices. SH3 domains are found in a plethora of proteins177, but the STAC SH3 domains are unique because they are connected by a very short five-residue linker and form a rigid interaction through an extensive interdomain interface. In both isoforms, the relative domain orientation is very similar, shown by a superposition with a RMSD value of 1.28Å for 104 Cα atoms (Fig. 2.2).  Figure 2.1B shows a zoomed-in view of the interdomain interface in STAC2. A large portion of the interface is made up by hydrophobic residues, with the exception of Gln347, located on the SH3-1 domain, which forms hydrogen bonds with main chain atoms of the SH3-2 domain and neighboring water molecules. Gln347 is conserved in human STAC1 and STAC2 but is replaced by an isoleucine in STAC3 (Fig. 2.1C).  To verify its importance, we also solved a crystal structure of the Q347I mutant in STAC2 (Fig. 2.2). Superposition of the STAC tandem SH3 structures with one harboring the Q347I mutant shows that the glutamine is not essential for maintaining the domain orientation. Since a similar interaction is seen in two different isoforms and for various molecules in the asymmetric unit, we conclude that this is a stable interface. Although we have been unable to produce a structure of the STAC3 tandem SH3 domains, we did succeed in crystallizing its second SH3 domain (Fig. 2.1D). The resulting structure at 1.3Å superposes very well with the SH3-2 domains of STAC1 and STAC2, with a 52  RMSD value of 0.82Å for 55 Cα atoms (based on superposition with STAC1). Previously, it had been suggested that STAC1 and STAC2 would lack a second SH3 domain167.  Figure 2.1 Crystal structures of the STAC SH3 domains. A. Schematic diagram of the domain arrangement within the STAC proteins (top right corner). Crystal structures of the two tandem SH3 domains of STAC1 (top left) and STAC2 (beneath). Beta-strands and 310 helices are labeled. The conserved Trp residue implicated in NAM for STAC3 is shown in black sticks. B. Close-up of the SH3 domain interface for STAC2, showing hydrophobic residues in black and Gln347 in white sticks. C. Sequence alignment of the SH3 domains of the three STAC isoforms with the secondary structure of STAC2 shown above. Trp284 in STAC3 is highlighted in red. D. Crystal structure of the second SH3 domain of STAC3.  53   Figure 2.2 Superposition of the SH3 domains of the STAC isoforms. The tandem SH3 domains of STAC1 (blue), STAC2 (red) and STAC2 harboring the Q347I mutation (orange) are superimposed on each other. In green is a superposition of STAC3 SH3-2 on the STAC2 SH3-2 domain. The ellipses indicate loops within SH3-2 which demonstrate the largest degree of difference among the STAC isoforms.   2.3.2 The tandem SH3 domains interact with the II-III loop of CaV1.1 and CaV1.2 In CaV1.1, the linker connecting transmembrane repeats II and III (‘II-III loop’) has been shown to be crucial for EC coupling52,56,178. We therefore hypothesized that it may form a binding site for the STAC3 tandem SH3 domains. SH3 domains typically associate with proline-rich segments and an interaction prediction server179 suggested three putative sites within the 54  loop, all contained within the fragment 728-775 (Figure 2.6A). Figure 2.3A shows the results of ITC experiments, where we titrated this fragment (‘core II-III loop’) into the STAC2 tandem SH3 domains. These show binding with a Kd of 1.85µM. Both STAC1 and STAC3 also bind to the same core II-III loop, with affinities ~2.4 and 2.2-fold higher than STAC2 (Figures 2.3 E and F; Table 2.3). We next tested binding of each predicted site individually to STAC3 but failed to detect any significant interaction with either peptide (Table 2.3). However, we noticed an additional set of proline residues next to the second predicted site and utilized a larger peptide, encoded by residues 747-760. This peptide showed binding to the tandem SH3 domains of STAC3 (Kd of 10.6µM), as well as to STAC2 (Kd = 9.3µM) and STAC1 (Kd = 3.9µM) (Figure 2.3 and Table 2.3). Since the peptide 747-760 is the shortest sequence for which we could detect reliable binding to STAC proteins, we call this the minimal peptide. To determine which individual SH3 domain is required for binding, we utilized the individual domains of STAC2 for ITC experiments (Fig. 2.3 C and D). Only SH3-1 showed appreciable binding to the minimal peptide (Kd = 84µM), but this affinity is lower compared to the construct containing both SH3 domains (Kd = 9.3µM). The SH3-2 domain thus adds a small contribution, either by altering the conformation of the SH3-1 domain or by providing an additional interaction surface for the peptide. In order to verify that the minimal peptide forms the main binding site within the core II-III loop, we generated a double mutant whereby Pro756 and Pro758 were both mutated to alanine in the core II-III loop (residues 728-775). This mutant failed to show binding to STAC3, indicating that these prolines, contained within the minimal peptide, are a major binding determinant (Fig. 2.4A).  55  Figure 2.3 ITC experiments reveal binding between peptides of the CaV1.1 II-III loop and the SH3 domains of the STAC proteins. A. Shown is 1mM CaV1.1 728-775 (“core II-III loop”) titrated into 0.1mM STAC2 tandem SH3 domains (residues 296-411). B. Shown is 1mM CaV1.1 747-760 (minimal peptide) titrated into 0.1mM STAC2 tandem SH3 domains. C. Shown is 1mM CaV1.1 747-760 titrated into 0.1mM STAC2 SH3-1 (residues 296-349). D. Shown is 1mM CaV1.1 728-775 titrated into 0.1mM STAC2 SH3-2 (residues 354-411).  E. Shown is 1mM CaV1.1 728-775 titrated into 0.1mM STAC1 tandem SH3 domains (residues 288-402). F. Shown is 390µmM CaV1.1 728-775 titrated into 39µM STAC3 tandem SH3 domains (residues 245-364).   56  Table 2.3 Thermodynamic parameters for the interaction between individual or tandem SH3 STAC domains and the II-III loop constructs of CaV1.1 and 1.2   Values are presented as the average ± the standard error of the mean (where applicable).  * These short CaV1.1 constructs contained an N-terminal MBP tag for stability and a C-terminal His6-tag  ** N-value value fixed.    Titrated constructs Kd (µM) N-value ΔH  (cal.mol-1) ΔS  (cal.mol-1.K-1) Number of Replicates CaV1.1 728-775/STAC1 288-402 0.78 ± 0.02 1.16 ± 0.02 -9887 ± 72 -5.22 ± 0.28 2 CaV1.1 728-775/STAC2 296-411 1.85 ± 0.08 1.14 ± 0.04 -9494 ± 1.5 -5.61 ± 0.09 2 CaV1.1 728-775/STAC3 245-364 0.83 ± 0.45 0.46 ± 0.08 -11111 ± 1421 -10.09 ± 4.70 3 CaV1.1 747-760/STAC1 288-402 3.92 ± 0.08 1.07 ± 0.02 -10056 ± 244 -8.98 ± 0.85  CaV1.1 747-760/STAC2 296-411 9.31 ± 1.11 0.90 ± 0.02 -9880 ± 370 -10.12 ± 1.49 2 CaV1.1 747-760/STAC2 296-411 Q306L  22.4 0.67 -10740 -14.7 1 CaV1.1 747-760/STAC3 245-364 10.63 ± 7.32 1.14 ± 0.22 -4709 ± 1092 7.58 ± 5.32 2 CaV1.1 728-775/STAC2 354-411 No binding detected 1 CaV1.1 747-760/STAC2 296-349 84 0.708 -15360 -32.9 1 CaV1.1 729-735/STAC3 245-364 * (predicted binding site 1) No binding detected 1 CaV1.1 747-756/STAC3 245-364 * (predicted binding site 2) No binding detected 1 CaV1.1 768-774/STAC3 245-364 * (predicted binding site 3) No binding detected 1 CaV1.1 728-775/STAC2 296-411 W329S No binding detected 3 CaV1.1 728-775 P756A P758A/STAC3 245-364  15.9 0.911 -3406 10.5 1 CaV1.1 728-775 I752A P753A R757A/STAC2 296-411 No binding detected 2 CaV1.1 728-775 I752A P753A R757A/STAC3 245-364 No binding detected 1 CaV1.2 829-876/STAC2 296-411 19.3 ± 0.94 1.00 ± 0.02 -8665 ± 242 -7.48 ± 0.91 3 57  It was recently reported that STAC proteins can alter the function of CaV1.2162. We therefore tested whether the tandem SH3 domains could bind to a similar site in this channel isoform. Figure 2.4C shows an ITC between the STAC2 tandem SH3 domains and the core II-III loop of CaV1.2 (equivalent to the core II-III loop of CaV1.1). This interaction has a Kd of ~19µM, representing a 10-fold weaker binding compared to CaV1.1. Figure 2.6A shows a sequence alignment of the core II-III loop regions of all four CaV1 isoforms. Although we did not formally test binding to CaV1.3 and CaV1.4, based on sequence conservation we predict that binding of the tandem SH3 domains may occur with the II-III loop of CaV1.3, but is unlikely to be significant for CaV1.4.  Figure 2.4 Determining the binding determinants for the STAC tandem SH3 domains and II-III loop interaction. A. ITC experiments using proline mutants within the CaV1.1 core II-III loop aided in determining the binding interface between the core II-III loop and the SH3 domains of the STAC proteins. Shown is 300µM CaV1.1 728-775 P756A P758A titrated into 30µM STAC3 tandem SH3 domains. B. Substituting a key residue within the STAC2-CaV1.1 binding interface with its STAC3 counterpart retains binding to the minimal CaV1.1 peptide. Gln306 was mutated to leucine, the residue found at the equivalent position within STAC3. Shown is 1mM 58  CaV1.1 747-760 titrated into 0.1mM STAC2 296-411 Q306L. C. The CaV1.2 equivalent of the core II-III loop also reveals binding to the tandem SH3 domains of STAC2. Shown is 1mM CaV1.2 829-876 titrated into the STAC2 tandem SH3 domains (residues 296-411).  Many more proline-rich segments are present in the CaV sequences. We therefore also tested interactions between various cytosolic loops and termini of CaV1.2 with the STAC2 tandem SH3 domains. Together, the cytosolic segments tested cover all proline-rich regions predicted to bind SH3 domains179, but no significant interactions could be detected outside of the II-III loop (Fig. 2.5). This suggests that the II-III loop is the primary binding site for the STAC SH3 domains.  2.3.3 Crystal structure of STAC2 in complex with a peptide from the II-III loop We have solved a crystal structure of the STAC2 isoform in complex with the CaV1.1 minimal peptide (residues 747-760) at 1.76Å resolution. The structure contains six complexes in the asymmetric unit, all with highly similar interfaces. Here we describe the complex formed by STAC2 chain A, which showed the best density for the peptide. Although STAC2 is not co-expressed with CaV1.1 in skeletal muscle, the high conservation of the interface among STAC isoforms suggests a very similar interface for STAC3.  The interaction buries 547Å2 of surface area and mostly involves the SH3-1 domain (Fig. 2.6B and C; Fig. 2.7A). Trp329, conserved in all three STAC isoforms, is central in the interaction and is the target for a disease mutation linked to Native American Myopathy (NAM) in STAC3148.  Its sidechain nitrogen forms a hydrogen bond with the main chain of the peptide.  On the CaV1.1 side, Arg757 is involved in multiple interactions with SH3-1. It sits in a pocket lined by Trp329 and is involved in an extensive hydrogen bond network with Gln306, Asp310  59   Figure 2.5 Titrations between the tandem SH3 domains of STAC2 (296-411) and predicted SH3 binding sites within CaV1.2 reveal no additional binding sites. A. Titrating 1mM CaV1.1 728-775 into 0.1mM STAC2 296-411 revealed a binding affinity on the order of 1.85µM. In contrast, no significant binding was detected by calorimetry for titrations with peptides B-H. B. 1mM STAC2 was titrated into 0.1mM CaV1.2 446-516. C. 743µM STAC2 was titrated into 59µM CaV1.2 1579-1646. D. 819µM CaV1.2 1170-1218 was titrated into 81.9µM STAC2. E. 228µM CaV1.2 1632-1860 was titrated into 22.8µM STAC2. Although this titration yields large heats, the background titration of CaV1.2 1632-1860 into buffer produced equally large exothermic heats.  F. 1mM CaV1.2 1949-1971 was titrated into 0.1mM STAC2. G. 1mM STAC2 was titrated into 0.1mM CaV1.2 1-112 as a thioredoxin fusion protein to aid with solubility of the construct. H. 1mM STAC2 was titrated into 0.1mM CaV1.2 2080-2092 as a thioredoxin fusion protein. All STAC2 constructs used in the titrations A-H contain the tandem SH3 domains, encompassing residues 296-411.  60  and several additional water-mediated hydrogen bonds. It also forms an additional salt bridge with Glu307 and a cation-pi interaction with Trp329. Other interactions of interest are made by CaV1.1 residues Pro753, which sits in another small pocket lined by Trp329, and by Ile752, which forms hydrophobic interactions. To show the importance of these residues, we mutated all three residues (Ile752, Pro753, Arg757) to alanines within the core II-III loop and were no longer able to detect an interaction with the STAC2 tandem SH3 domains (Fig. 2.6D). Although we do not observe a direct interaction between the CaV1.1 peptide and the SH3-2 domain, the sidechain of the CaV1.1 residue Glu749 forms multiple water-mediated hydrogen bonds with STAC2, including Lys374, located in the SH3-2 domain. This observation likely explains our observation that the tandem SH3 domains bind more strongly to this peptide than SH3-1 alone (Table 2.3). At the N-terminus of the CaV1.1 peptide, two negatively charged residues display weak electron density that precluded building their structure. They would be pointing towards a highly positively charged region of the SH3-2 domain (Fig. 2.6C) and may thus also contribute to the affinity. We hypothesize that this additional interaction displays different conformational states, thus not providing clear electron density.  These interactions, described for STAC2, are most likely very similar for STAC1 and STAC3. First, all three isoforms were able to bind the same peptide (Fig. 2.3, Table 2.3) and the surface residues of the binding region in STAC2 are highly conserved in STAC1 and STAC3 (Fig. 2.1C). The only exception is Gln306, which is replaced by a Leu in STAC3. To test its importance in binding, we mutated Gln306 to leucine in the STAC2 tandem SH3 domains and found the affinity for the minimal peptide to be roughly two-fold lower than wild type (Fig. 2.4B, Table 2.3). So, although Gln306 contributes to binding, it is not essential. 61  Because part of the SH3-2 surface seems to contribute to binding the II-III loop peptide, the question arises whether it may still bind a different peptide, residing either within or outside of CaV1.1. A closer inspection shows that the region of SH3-2 that interacts with the II-III loop peptide is not at the canonical binding surface for SH3 domains. Figure 2.7B shows a superposition of the STAC2 SH3-2 domain with the structure of the SH3 domain of Abl, in complex with a target peptide (PDB: 3EG1). This shows that the canonical contact interface resides distally away from the binding site for the short II-III loop peptide. Another peptide could thus bind to the canonical face of SH3-2, without clashing with the minimal II-III loop peptide, implying that the STAC tandem SH3 domains may bridge two discontinuous segments within CaV1.1, or possibly between CaV1.1 and another protein like RyR1. The tight association between the two SH3 domains would ensure a rigid link. However, the canonical tryptophan residue is replaced by a phenylalanine in the SH3-2 domain (Phe391 in STAC2; Figure 2.7B). It thus remains to be determined whether the SH3-2 domain can form significant interactions on its own.  62  Figure 2.6 Residues of the CaV1.1 “core II-III loop” have an extensive interaction network with the first SH3 domain of STAC2. A. Sequence alignment of the cytosolic loop linking domains II and III in various CaV1 isoforms (CaV1.1-CaV1.4). Conserved residues are highlighted in grey. In bold are the three predicted SH3 binding sites. Outlined in the red box is the minimal peptide used in the complex of CaV1.1 and STAC2. B. Crystal structure of the CaV1.1 747-760 and STAC2 tandem SH3 domains complex. The CaV1.1 peptide is colored in gold with residue identities denoted in regular font. The key Trp329 STAC2 residue is colored in black. Critical STAC2 residues involved in the interaction with CaV1.1 are indicated in the bold italic font. Ionic networks and water-mediated hydrogen bond networks are indicated by dashed 63  lines. C. Electrostatic surface representation of the STAC2 tandem SH3 domains interacting with the CaV1.1 747-760 peptide. The ellipse represents the region where we predict the N-terminus of the CaV1.1 peptide, not visible in the crystal structure, would be located. D. The core CaV1.1 II-III loop (residues 728-775, 1mM) harboring the triple mutation, Ile752A P753A R757A, dramatically diminished binding to the STAC2 tandem SH3 domains (residues 296-411, 0.1mM). E. ITC showing 1mM CaV1.1 728-775 (core II-III loop) titrated into 0.1mM STAC2 tandem SH3 domains (residues 296-411) containing the W329S mutation (equivalent of the NAM mutation, W284S in STAC3). No significant binding is detected.   2.3.4 Effect of the NAM mutation W284S In STAC3, Trp284 is the target for a disease mutation (W284S) linked to NAM148. The equivalent residue in STAC2 is Trp329, which forms the multiple interactions with the CaV1.1 peptide. We introduced the W284S mutation in STAC3, as well as the equivalent W329S mutation in STAC2. In both cases, we were still able to produce purified protein, indicating that the mutation did not cause complete misfolding of the domain. However, using ITC experiments, we were no longer able to detect an interaction with the II-III loop (Figure 2.6E). One implication of this may be a loss of mechanical coupling to the skeletal muscle Ryanodine  Receptor (RyR1). This agrees with previous experiments showing that the W284S mutation in STAC3 causes a drastic reduction in depolarization-induced Ca2+ release in reconstituted stac-/- myotubes48,180.  2.3.5 Functional role of the interaction If the SH3:II-III loop interaction is truly occurring in myocytes, then point mutations in the II-III loop that disrupt the interaction should have a functional effect similar to the NAM mutation. We therefore made use of the triple I752A/P753A/R757A mutation that knocks out binding to the tandem SH3 domains (CaV1.1-IPRAAA). We reconstituted dysgenic (CaV1.1-/-)  64  Figure 2.7 Examining the binding interface of the CaV1.1 II-III loop peptide and the tandem SH3 domains of STAC2. A. Fo-Fc density of the CaV1.1 II-III loop peptide prior to building in the peptide, contoured at 3σ. The final peptide model is shown. B. Superposition of STAC3 SH3-2 with the coordinates of the Abl-SH3 domain (PDB: 3EG1) in complex with a proline-rich peptide. This superposition reveals that it may be possible for SH3-2 of the STAC proteins to bind an additional peptide. The CaV1.1 core II-III loop peptide (shown in gold) is bound to SH3-1 of STAC2. The key Trp329 residue of STAC2 is colored in black. In red is the proline-rich peptide in complex with the Abl-SH3 domain. A potentially key residue, Phe391, is shown in stick format. 65  myotubes with wildtype and GFP-CaV1.1-IPRAAA and analyzed depolarization-induced Ca2+ transients. Even though expression and triad targeting of the mutant construct was normal (Fig.2.8A), the potency of GFP-CaV1.1-IPRAAA to reconstitute EC coupling was dramatically reduced compared to the wildtype control. The fraction of myotubes responding to electrical field stimulation with detectable Ca2+ transients was only 4.1 ± 2.0% of cultures transfected with GFP-CaV1.1 (Fig. 2.8B). Moreover, in the few responding myotubes, Ca2+ transients in response to single stimuli or to tetanic stimulation with increasing frequencies were significantly weaker with GFP-CaV1.1-IPRAAA than with GFP-CaV1.1 (Fig. 2.8 C and D). The peak amplitude of the transients was significantly reduced (0.13 ± 0.01 ΔF/F) compared to control values (0.31 ± 0.04 ΔF/F, p<0.0001) and the time-to-peak was approximately doubled, from 57.6 ± 4.4 ms to 125.5 ± 14.0 ms (p<0.0001) (Fig. 2.8E). In contrast, the decay of the Ca2+ transient was not affected by the II-III loop mutations (377.5 ± 41.4 ms and 452.1 ± 44.1 ms, p=0.22), indicating that overall Ca2+ handling in the reconstituted myotubes was normal. In variance to the weak response of cultures reconstituted with GFP-CaV1.1-IPRAAA, transfection with GFP-CaV1.1-II-IIIM, a II-III loop chimera known to lack skeletal muscle EC coupling168, did not at all reconstitute depolarization-induced Ca2+ transients in dysgenic myotubes (Fig. 2.8D). In summary, this analysis demonstrates that disruption of the interaction between the II-III loop of CaV1.1 and the tandem SH3 domains of STAC3 strongly perturbs skeletal muscle EC coupling.      66   Figure 2.8 Mutation of the IPR motif perturbs EC coupling in skeletal muscle myotubes. A. Dysgenic (CaV1.1-/-) myotubes reconstituted with GFP-CaV1.1 (left) or the mutated GFP-CaV1.1-IPRAAA (right) were double-immunolabeled with anti-GFP (upper) and anti-RyR (middle). Clusters of GFP-CaV1.1 and GFP-CaV1.1-IPRAAA colocalized with the RyR1 (lower, yellow clusters in color overlay) indicate the correct targeting of both channel variants into T-tubule/SR 67  or plasma membrane/SR junctions. Scale bar, 10μm. B. A severely reduced fraction of the cells transfected with GFP-CaV1.1-IPRAAA (4.1 ± 2.0 %, N=3 transfections, 6 dishes, p<0.0001) responded with Ca2+ transients to 2-ms pulses of 40V, compared to cells transfected with GFP-CaV1.1. C. Representative Ca2+ transients evoked by single 2-ms 40V stimuli in myotubes expressing GFP-CaV1.1-IPRAAA (grey) or GFP-CaV1.1 (black). D. GFP-CaV1.1-IPRAAA myotubes responded to tetanic stimulation up to 240Hz, although with much reduced amplitudes, whereas myotubes reconstituted with the EC coupling-deficient mutant GFP-CaV1.1-II-IIIM showed no response at all. E. Compared to GFP-CaV1.1, individual Ca2+ transients in GFP-CaV1.1-IPRAAA expressing cells showed a significantly reduced peak amplitude (p<0.0001) and time to peak (p<0.0001), but comparable decay (N=4 transfections, nIPRAAA = 21, nWT = 24). Values are expressed as mean ± SEM; Statistics: unpaired t-test, **** p<0.0001.  2.4 Discussion Skeletal muscle EC-coupling relies on a mechanical link between the L-type calcium channel, CaV1.1, located in the plasma membrane, and RyR1, located in the SR membrane. Although several reports have suggested direct interactions between both proteins, an unambiguous interaction through a quantitative method or through structures of complexes have thus far been lacking112.  It is thus possible that the interaction is mediated, in full or in part, by auxiliary proteins. Recently, several reports have revealed a key function for the adaptor protein STAC348,148,156–158,162,167,180. A mutation in STAC3, W284S, is responsible for Native American Myopathy (NAM), a rare disorder found in Lumbee Native Americans153–155. In addition to its role in EC coupling, STAC3 was also found to aid in expression of CaV1.1 at the plasma membrane of tsa201 cells162.   Here we show the first structural insights into the STAC proteins, by solving high-resolution crystal structures of the tandem SH3 domains of STAC1 and STAC2, and the 2nd SH3 domain of STAC3.  These domains form a compact arrangement by virtue of a short linker and form a micromolar affinity binding site for the II-III loops of CaV1.1 and CaV1.2. The binding 68  site is located within a short proline-rich peptide encoded by residues 747-760 in human CaV1.1. The corresponding region also forms a binding site in CaV1.2.  We postulate that the interaction of the CaV1.1 II-III loop with the STAC3 tandem SH3 domains is required for EC coupling. Indeed, using field stimulation experiments in dysgenic myotubes we show that a mutant CaV1.1, unable to bind the tandem SH3 domains via its II-III loop, severely perturbs skeletal muscle EC coupling (Fig. 2.8). This is also in agreement with previous reports that have shown the importance of the II-III loop in bidirectional coupling with RyR152,55. The region 720-765, which encompasses the described binding site, was shown to be essential for normal EC coupling168 and insertion of YFP immediately adjacent to the site, between residues 760 and 761, completely knocked out bidirectional coupling181. Given these findings, this stretch of CaV1.1 residues has been termed the “critical domain” for EC coupling. Chimeric CaV1.1-1.2 channels, which were entirely cardiac except for a small stretch of residues which were of skeletal origin in the II-III loop, revealed that residues 725-742 in CaV1.1 are critical for mediating skeletal muscle EC coupling178.  Encouragingly, our results have been corroborated by the recent work of Polster et al. (2018). Using colocalization as an indicator of molecular interactions, Polster and colleagues demonstrated that STAC3, as well as STAC1 and STAC2, interact with the critical domain of the II-III loop of CaV1.1182. They conclude that CaV1.1 residues 745-765, which encompasses our identified minimal CaV1.1-STAC binding interface, represents the likeliest region for high-affinity binding to STAC3. Although all three STAC isoforms were able to support the functional expression of CaV1.1 and EC coupling in STAC3-null myotubes, co-expression of STAC3 produced the most robust effect. Further consolidating our findings, Polster et al. concluded that the interaction of STAC3 with the II-III loop critical domain did not require the 69  C1 domain or the intrinsically disordered stretch of residues flanking the C1 domain, but instead relied on the first of the two SH3 domains.  Our structure also directly explains the effect of the W284S mutation in STAC3, responsible for NAM. The equivalent residue in STAC2, W329, forms key interactions with the II-III loop peptide, and either the W329S mutation in STAC2, or the W284S mutation in STAC3 decreases the affinity to levels below detection. Previous functional experiments have shown that the W284S mutation results in diminished EC coupling48 and in reduced recruitment of STAC3 into the skeletal muscle CaV1.1 complex167.  Could STAC3 be a protein that links both CaV1.1 and RyR1? Previous experiments have shown coimmunoprecipitation of STAC3 with RyR1148 in muscle tissue. However, recombinant STAC proteins expressed in dysgenic myotubes, which contain RyR1 but not CaV1.1, were not targeted into triad junctions162,167, and also photobleaching experiments in tsa201 cells could not confirm this interaction162. It is still possible that there is a weak interaction between STAC3 and RyR1 in native cells, only occurring by virtue of a high local concentration. In our structure of the STAC2:CaV1.1 II–III loop peptide complex, the canonical binding surface of the SH3-2 domain is still available for binding another peptide (Fig. 2.7B). But whether RyR1 and STAC3 interact remains to be determined. In conclusion, STAC3 has emerged as a novel player in EC coupling and our results provide the first glimpse into the specific interactions with CaV1.1. Further experiments will be required to identify additional binding sites within both CaV1.1 and other EC coupling proteins.    70  Chapter 3: STAC3 modulates L-type voltage-gated calcium channel inactivation 3.1 Introduction The L-type voltage-gated calcium channel fulfills two roles in the T-tubular membrane of skeletal muscle: voltage sensor for EC coupling and the slowly activating calcium ion conduit51. A fascinating component of CaV behaviour is their ability to inactivate in a voltage- and/or Ca2+ dependent manner to circumvent excess extracellular Ca2+ entry61. Voltage-dependent inactivation (VDI) is a property observed in all voltage-gated calcium channel subtypes, although the extent of VDI varies among the channel isoforms61. Elucidating the mechanisms of VDI has identified the I-II loop and the S6 segment of the transmembrane domains as major components of this process183–189. Chimeric and mutagenesis studies by Zamponi et al. led to the model whereby structural rearrangement in the S6 segments in response to prolonged membrane depolarization exposes a binding site for the I-II loop to act as a hinged-lid-like gating particle, thereby inactivating the channel187,189. Multiple studies have demonstrated the potency of β-subunit modulation of VDI60,190. Multiple mutations in the S6 segments of all four CaV1.2 repeats has established the critical role of the internal vestibule as a determinant of VDI60. Indeed, Findeisen and Minor support the idea that inactivation involves some type of constriction of the pore involving the S6 segments from each domain190.  A distinct mechanism of inactivation, calcium-dependent inactivation (CDI), becomes apparent when comparing Ba2+ and Ca2+ current kinetics61. Ca2+ accelerates current decay kinetics. The finding that CDI is abolished with over-expression of a dominant-negative mutant of CaM, which lacks the ability to bind Ca2+, was fundamental to understanding the mechanism 71  underlying CDI39,40,191. One model suggests that apo-CaM is constitutively anchored to the C-terminus of the channel such that its C-lobe interacts with the channel’s IQ domain and the N-lobe with an upstream EF-hand region61,192. In response to elevated intracellular Ca2+, the anchored CaM binds to Ca2+ which promotes a conformational change in the C-terminus-CaM complex, giving rise to CDI. In this manner CaM is the Ca2+-sensor for CDI. CaM is a highly versatile and ubiquitously expressed Ca2+ sensor, which regulates the function of ion channels in addition to many enzymes152. The N- and C-terminal lobes each contain two Ca2+-binding helix-loop-helix motifs, “EF-hands”. Ca2+ binding to any or all of these sites exposes hydrophobic pockets that bind and modulate the activity of target proteins193. In the presence of Ca2+, CaM can adopt numerous conformations depending on the target sequence. There is increasing evidence suggesting that these multiple CaM configurations can regulate the activity of voltage-gated calcium channels in response to different Ca2+ signals152. Ca2+-CaM is also involved in another regulatory feature of voltage-gated CaV1.2 channels: calcium-dependent facilitation (CDF)40,194–196. CDF describes a positive-feedback mechanism whereby increased basal Ca2+ or repeated transient depolarization leads to increased channel opening. Thus, Ca2+ binding to CaM  drives both CDI and CDF40,159. DeMaria et al. proposed that the ability of CaM to serve a bifunctional role in voltage-gated calcium channel regulation reflects the bifurcation of the Ca2+ signal arising from the differential binding of Ca2+ to the N- and C-terminal lobes of CaM197. Apo-CaM binding to the channel may either influence: 1) the rate at which the channel responds to the Ca2+ signal due to an increased local concentration of CaM; 2) the rate of Ca2+ association with either or both CaM lobes due to binding to the targets; and/or 3) stabilization of a channel state that is able to respond more rapidly to elevated Ca2+ levels.  72  Combined, VDI and CDI, are important regulatory mechanisms for the channel to limit excess extracellular Ca2+ influx. VDI represents an intrinsic channel property which only depends on voltage, in comparison, CDI is a tunable process, providing feedback inhibition in response to rising Ca2+ levels61. Since many clinically used drugs – e.g. anti-epileptics and dihydropyridines – block CaVs via a drug-induced inactivated channel conformation, understanding the molecular mechanisms of inactivation is crucial61. Ongoing research interest in this area reflects the many intricacies of channel inactivation that remain elusive, particularly if there are novel proteins involved in the regulation of channel inactivation.   STAC3 has been found to be intertwined into both functional roles of the L-type voltage-gated calcium channels. In addition to being a vital player in skeletal muscle EC coupling48,148,156, the STAC proteins – particularly STAC2 and STAC3 – have been shown to interact with CaV1.2162, the channel isoform predominantly expressed in the heart and neurons. Unlike CaV1.1, STAC co-expression was not required for trafficking CaV1.2 to the plasma membrane, however, the interaction is of functional importance since the STAC proteins significantly decreased the channel’s speed of inactivation162. Polster et al.’s study, where STAC3 KO myotubes reconstituted with STAC3-NAM caused partial recovery of normal Ca2+ currents but only very weak EC coupling restoration, hinted that these two effects rely on STAC3 domains that function partially independent of each other48. The STAC family of proteins share a conserved protein kinase C (PKC) C1 domain, located towards the N-terminus, and two SH3 domains, located at the C-terminus (Chapter 2, Fig. 2.1A). A large intrinsically disordered stretch of amino acids (~100 residues) connects the C1 and SH3 domains. Currently, there are structures for two domains of STAC3: the C1 domain structure has been solved by NMR198 (Fig 3.1), and the second SH3 domain structure has been solved via x-ray 73  crystallography199. The STAC2 tandem SH3 domains and CaV1.1/1.2 II-III loop interaction199 is the first report of a direct interaction between the STAC proteins and the CaVs. Here, it was discovered that the first SH3 domain was the main contact point for the CaV II-III loop. Although both ITC and structural analysis alluded to a contribution of the second SH3 domain to the interaction, the canonical binding surface of SH3-2 resides distally away from the II-III loop binding site. Thus, another peptide could bind to the canonical face of SH3-2, without clashing with the minimal II-III loop peptide. This implies that the STAC tandem SH3 domains may bridge two discontinuous segments within CaV1.1/1.2199.  A recent study has demonstrated that the C1 domain of STAC3 is crucial for the stability of the complex between STAC3 and L-type voltage-gated calcium channels167. While the NAM mutation did not affect the stability of the STAC3-CaV1.1 interaction, mutation of only two residues in the C1 binding pocket increased the turnover of STAC3 in skeletal muscle triads. More recently, Campiglio et al. (2018) used dysgenic myotubes to identify that amino acid residues 1641-1668 in the C-terminus of CaV1.2 are necessary for the domain association of STAC proteins200. This stretch of residues contains the IQ domain. The critical residues involved in STAC association overlap with those known to interact with the C-lobe of Ca2+-CaM and mediating CDI. Collectively, their results reflect that the entire family of STAC proteins are capable of modulating Ca2+ entry through the CaV1.2 channel: STAC1 and STAC2 also interfered with CDI, albeit to a lower extent to compared to STAC3. The authors’ previous findings with the STAC C1 domain167 led them to postulate that a direct C1-IQ interaction is crucial for the stable association of the STAC-CaV1 complex in skeletal muscle triads and for the modulation of inactivation properties of CaV1.2 currents200. However, the high affinity interaction between Ca2+-CaM and the IQ domain201 may impede a direct interaction between 74  STAC and the IQ domain, thus, an alternate possibility is that STAC proteins interact directly with the Ca2+-CaM-IQ complex, preventing the conformational changes that trigger CDI200. This speculation reflects the vast amount of work still to be achieved in understanding the complete effect of STAC proteins on the L-type calcium channels and identifying all the regions within the channel to which the STAC proteins can bind.                         Figure 3.1  NMR structure of the STAC3 C1 domain. The compact fold of the C1 domain coordinates two zinc ions (represented as blue spheres) via a pair of three cysteine residues. The model depicted here is from PDB: 2DB6.    75  3.2 Experimental Procedures 3.2.1 Electrophysiology Constructs for electrophysiology consisted of human CaV1.2 (splice variant a1c77) in pcDNA3.1(+)/hygro (Invitrogen), rabbit CaV-β1 in pSP65, rabbit CaV-α2δ in pcDNA3 (Invitrogen) and human STAC3 isoform a (NCBI Reference Sequence: NP_659501.1) in pcDNA3.1(-) (Invitrogen). STAC3 constructs encoded either the full-length transcript (1-364), a truncated STAC3 missing both SH3 domain (1-244), or a truncated STAC3 containing only the SH3 domains (residues 247-364). All RNA transcripts were prepared from cDNA using a T7 mMessage mMachine kit except for CaV-β1 which was synthesized using a SP6 mMessage mMachine kit (Ambion).  Collagenased Xenopus leavis stage V-VI oocytes were injected with 46.6nL of a mixture containing 8.2-10.9ng CaV1.2 α1c and 11.7-15.5ng β1. Certain RNA mixtures, as described in the figure legends, were supplemented with 11.7ng CaV-α2δ and 9.8-13.0ng of either of the STAC3 constructs described. Oocytes were kept at 18oC in a modified Ringer’s OR3 media (50% (v/v) L-15 medium, 0.5% (v/v) L-glutamine, 0.5% (v/v) gentamycin and 15mM HEPES, adjusted to pH 7.6 using NaOH). Recordings were performed 3-4 days following injection. Prior to recording, oocytes were injected with 50nL of 500mM of the tetrapotassium salt of BAPTA (ThermoFisher) to minimize contaminating Ca2+-activated Cl- current. Oocytes were then perfused with a Ba2+-containing solution (40mM Ba(OH)2, 50mM NaOH, 1mM KOH, 10mM sodium-HEPES, adjusted to pH 7.4 using HNO3) or a Ca2+-containing solution (where the Ba(OH)2 was replaced with Ca(NO3)2) as indicated in the figure legends. Two-electrode voltage-clamp (TEVC) experiments were performed using an Axoclamp 900A amplifier (Molecular Devices) and digitized with a Digidata1440A digitizer (Molecular Devices). Electrodes were 76  filled with 3M KCl and had resistances of 0.1–1.2MΩ. Ionic currents were collected and analyzed using pClamp10, leak currents were subtracted using a P/4 protocol. Data was normalized in Excel. Statistical significance was determined using a two-tailed unpaired t-test, with a confidence level of 95%. Statistical analysis was performed and graphed using GraphPad Prism 5.   3.2.2 Expression constructs A human STAC1 (residues 98-159) and STAC3 (residues 80-140) C1 domain construct, full-length human CaM, CaM N-lobe (1-78), and CaM C-lobe (79-149) were cloned into a modified pET28 vector (Novagen), containing a His6-tag, MBP and a cleavage site for the TEV protease49. Sequence integrity was confirmed by sequencing (Eurofins Operon).  3.2.3 Protein expression and purification The STAC C1 domains and CaM constructs were expressed for 3-5 hours in Escherichia coli Rosetta (DE3) pLacI (Novagen) grown in 2xYT media at 37oC. Since C1 domains typically are zinc-binding proteins202, the induction media of the STAC C1 constructs was boosted with zinc sulphate (20µM). Cells were lysed via sonication as described in Chapter 2. However, lysis buffers for the STAC C1 domains additionally contained 25µg/mL RNAase and >0.06unit/µL Benzonase (SigmaAldrich), and 20µM ZnSO4. Full-length CaM lysates were applied to HisTrap FF Crude columns (GE Healthcare), washed with 10 column volumes (CVs) of buffer A plus 20mM imidazole and eluted with buffer B. Following a 12-14 hour digestion with His-tagged TEV protease at 4oC, CaM was further purified using a Phenyl-Sepharose HP column (GE Healthcare). CaM was applied to the column using 150mM KCl, 10mM HEPES pH 7.4 and 77  10mM CaCl2, and eluted in an equivalent buffer containing 10mM EDTA instead of CaCl2. Similarly, the CaM N- and C- lobe lysates were applied to HisTrap FF Crude columns and digested with TEV protease. A Talon column (Clontech) was used to remove the His-Tagged TEV protease. The lobes were applied to the Talon column with buffer A. Finally, the lobes were purified to homogeneity using a HiLoad Q-sepharose HP column, which had been equilibrated with 10mM KCl and 10mM Tris pH 8.0, and eluted with a gradient of 30% of buffer containing an additional 1M KCl over 15CVs. The STAC C1 lysates were applied to gravity-flow Amylose columns (New England Biolabs), using buffer A supplemented with 10mM CaCl2 and 20µM ZnSO4. Following elution in buffer A supplemented with 10mM maltose, the proteins were digested with TEV protease overnight and purified to homogeneity by size. Final protein quality for all the above constructs was assessed via size-exclusion chromatography (Superdex75 16/600 column, GE Healthcare) using a buffer containing 150mM KCl, 10mM HEPES 7.4. The Superdex buffer for the STAC constructs contained 20µM ZnSO4.  3.2.4 Isothermal titration calorimetry Since Ca2+ binding causes CaM to undergo large conformational changes, which vastly alters CaM’s behaviour, interactions with CaM can be subdivided into Ca2+-dependent and Ca2+-independent modes of binding63. This chapter investigates two different STAC-CaM interaction sites:  1) CaM and CaV1.2 IQ domain binding to the STAC C1 domain. To test Ca2+-dependent modes of binding, all proteins were dialyzed against 25mM KCl, 10mM HEPES 7.4, 20µM ZnSO4 and 10mM CaCl2. Due to the Zn2+-binding nature of the STAC C1 domains, these proteins could not be dialyzed in an EDTA-containing buffer. Thus, to 78  ensure the CaM constructs were Ca2+-free, they were first dialyzed in a buffer containing 150mM KCl, 10mM HEPES 7.4 and 10mM EDTA before being transferred to an EDTA-free buffer to dialyze out the EDTA. Finally, the apo-CaM constructs and the STAC1 C1 constructs were dialyzed together in 25mM KCl, 10mM HEPES 7.4, 20µM ZnSO4. A synthetic peptide encompassing the CaV1.2 IQ motif (Lifetein) was dissolved in each corresponding buffer.   2) CaM binding to a STAC3 peptide (residues 175-194, human). In preparation for ITCs testing Ca2+-dependent modes of binding, all proteins were concentrated and dialyzed against 150mM KCl, 10mM HEPES pH 7.4 and 10mM CaCl2 at 4oC. Proteins were dialyzed against a similar buffer except with 10mM EDTA instead of CaCl2 for ITCs testing Ca2+-independent binding. A synthetic STAC3 peptide (Lifetein) was dissolved in each dialysis buffer.  Protein concentrations were determined using a Nanodrop2000 spectrophotometer (ThermoFisher). Titrations consisted of 20 injections of 2µL with concentrations noted in the figure legends. Experiments were performed at 25oC and using a stirring speed of 750RPM on an ITC200 instrument (GE Healthcare). All data were processed on Origin 7.0 and isotherms were generated following a point-by-point subtraction of a reference titration of ligand into buffer.   3.3 Results 3.3.1 STAC3 modulates CaV1.2 channel inactivation STAC2 and STAC3 have been shown to associate with CaV1.2 in tsa201 cells causing a significant decrease in channel inactivation162. We have managed to recapitulate the result of STAC3 on CaV1.2’s rate of inactivation using Xenopus laevis oocytes199. To discern STAC3’s 79  effect on VDI and CDI, both Ba2+ and Ca2+ were used as the conducting ion. Co-expression of STAC3 dramatically reduced the speed of inactivation for both ions, although the effect on whole-cell Ca2+ currents was more pronounced (Fig3.2A). In all the electrophysiology experiments discussed, the β1 subtype was co-expressed. Not only is β1 required in skeletal muscle EC coupling203,204, but it is also the subtype which elicits the maximal amount of CDI, thus using β1 provided a measure of the full extent of STAC3 on CDI205. To assess whether the identified STAC3 – CaV II-III loop interaction is involved in mediating the effects on channel inactivation, the NAM mutation (W284S) was introduced into full-length STAC3 and a triple mutation, M853A P854A R858A was introduced into CaV1.2 α1c. The CaV1.2 triple mutant is the equivalent of the CaV1.1 triple mutant, I752A P753A R757A, which abolished binding between the core II-III loop and the tandem SH3 domains of STAC2199. Both sets of mutants still allowed slowing of inactivation by STAC3, similar to the effect observed upon WT STAC3 co-expression. This finding suggested that the slowing of inactivation does not involve the STAC3-CaV II-III loop interaction (Fig 3.2 A, C and D). An outright deletion of both SH3 domains (STAC3 residues 1-244) still resulted in slower inactivation kinetics, which implies that the SH3 domains are not involved in channel inactivation (Fig. 3.2 A-D). A corresponding STAC3 construct which only contained the two SH3 domains (residues 247-364) restored the inactivation seen in the absence of full-length STAC3. Combined, these results suggest that the slowing of inactivation is mediated either by the C1 domain or the intrinsically disordered region which immediately precedes the SH3 domains, underscoring the possibility of multiple binding interfaces between STAC proteins and CaV1.1 or CaV1.2.   Since inactivation was more pronounced in oocytes conducting Ca2+ current, we assessed whether STAC3’s effect on CDI proceeded independently from VDI. We used an approach 80  similar to that described by Barrett and Tsien (2008), where we analyzed ratio plots of normalized Ica/IBa currents205. As the Ica/IBa current ratio remains at unity in the presence of STAC3 (Fig. 3.2E), it implies that STAC3 influences both CDI and VDI.  3.3.2 The STAC1 C1 domain does not interfere with the CaM-CaV1.2 IQ domain interaction Campiglio and Flucher’s 2017 paper demonstrated that introducing the NAM mutation into STAC3 had normal stability, although reduced incorporation, in the CaV1.1 channel complex167. Their findings led them to identify a binding pocket within the C1 domain which is key to the protein’s stable incorporation into the CaV1.1 complex. Indeed, replacing the first 22 residues or the last 36 amino acids of the STAC3 C1 domain with those of STAC2 resulted in a complete loss of association with CaV1.2. This demonstrated that residues from the N- and C- termini of the C1 domain directly or indirectly participate in the interaction with the calcium channel167. These findings, combined with their more recent results have suggested that STAC proteins are also involved in regulating Ca2+ entry through CaV1.2200. Their chimeric CaV channel and mutagenesis approach identified that STAC proteins associate with a stretch of residues in the C-terminus of the channel encompassing the IQ domain. Intriguingly, the critical residues for STAC association overlap with those known to interact with the Ca2+-C-lobe of CaM, which is involved in mediating CDI39,40. These findings led us to explore the possibility of a direct interaction between STAC proteins’ C1 domain and the IQ motif of CaV1.2.    81  Figure 3.2 STAC3 influences the inactivation kinetics of CaV1.2 α1c. A. Ca2+ currents of oocytes expressing CaV1.2 α1c in the presence and absence of various STAC3 constructs. Oocytes are co-expressing CaVβ1 and CaVα2δ. B. Ba2+ currents of oocytes expressing CaV1.2 α1c in the presence and absence of various STAC3 constructs. Oocytes are co-expressing CaVβ1 and CaVα2δ. C. Fraction of peak Ca2+ current after 600ms with oocytes either null for STAC3 or co-expressing various constructs of STAC3. Oocytes are co-expressing CaVβ1 and CaVα2δ. Significance was determined using a two-tailed unpaired t-test. *** P<0.0001, * 82  P=0.0399, ns = not significant. D. Fraction of peak Ba2+ current after 600ms with oocytes either null for STAC3 or co-expressing various constructs of STAC3. Clear bars represent oocytes co-expressing CaVβ1 and CaVα2δ. Lined bars represent oocytes co-expressing only CaVβ1. The W284S mutation was introduced into full-length STAC3. The triple mutation M853A P854A R858A was introduced into full-length CaV1.2 α1c and is the equivalent of the CaV1.1 triple mutant (Ile752A P753A R757A) which abolished binding between the core II-III loop and the tandem SH3 domains of STAC2. E. Net-CDI current for oocytes null for full-length STAC3 (blue) or co-expressing full-length STAC3 (green). The curve is obtained by dividing the values of the normalized Ca2+ current by the values of the normalized Ba2+ current according to Barrett and Tsien205.  A direct interaction between the C1 domains and the IQ motif would be a likely explanation for STAC’s involvement in channel inactivation.  Although there is a NMR structure of the STAC3 C1 domain (PDB: 2DB6, Fig. 3.1), our initial attempts to express and purify the C1 domains of STAC3 (residues 80-140) and STAC1 (residues 98-159) via a heterologous express system was trying. The structure of the STAC3 C1 domain revealed two pairs of three cysteines each coordinating a zinc ion (Fig 3.1), thus we included zinc sulfate in our expression media and purification buffers and excluded reducing agents in all buffers. Despite these improvements in expression and purification, preparations of STAC3 C1 resulted in yields too low to be used in subsequent ITC experiments. As we were able to purify the C1 domain of STAC1 to appreciable yield, all our ITC experiments described hereon were completed with STAC1 C1. Although future investigations will warrant the use of STAC2 and STAC3 to determine if there are isoform-dependent differences in C1 domain function, all three STAC isoforms were found to oppose channel inactivation200. We used Campiglio et al.’s recent paper (2018) as a template to test for a direct interaction between the STAC1 C1 domain and a CaV1.2 IQ domain peptide. We used isothermal titration calorimetry to test our hypothesis but failed to detect a direct interaction between the C1 domain and IQ peptide (Fig. 3.3A). Crystal structures and ITC experiments both corroborate a 83  tight association between Ca2+-CaM and a CaV1.2 IQ peptide (Kd ~ nM), which has provided a framework for understanding CaM’s regulation of CaVs206,207. Thus, it is likely that this high affinity of association prevents the STAC C1 domain from directly interacting with the IQ domain200. Instead, STAC proteins could interact with the Ca2+-CaM-IQ complex, thereby exerting STAC’s effect on inactivation. The STAC1 C1 domain was titrated into a cell containing a complex of the Ca2+-CaM-CaV1.2 IQ peptide. Our results revealed that the STAC1 C1 domain was not able to bind directly to the Ca2+-CaM-IQ domain complex nor was the C1 domain able to compete with Ca2+-CaM for binding to the IQ peptide (Fig. 3.3B). Even for the weaker (micromolar) affinity apo-CaM-IQ complex208, STAC1 C1 did not compete with CaM for binding to the CaV1.2 IQ  peptide nor was there any evidence of a direct interaction with the IQ peptide (Fig. 3.3C). Intriguingly, our reference titrations, where the STAC C1 domain was titrated into either apo- or Ca2+-CaM, hinted that further investigation into a possible STAC-CaM interaction was warranted.  STAC1 C1 binding to CaM is not dependent on Ca2+, since binding was detected to both apo- and Ca2+-CaM (Fig. 3.4 A and D). A dissociation constant of 4.45µM was measured between Ca2+-CaM and the C1 domain (Table 3.1). While the endothermic heats obtained for the background titration of STAC1 C1 into buffer starkly contrasts to the exothermic heats obtained for the titration of STAC1 C1 into apo-CaM, suggesting an interaction, no reliable fit and Kd value could be discerned for the titration involving apo-CaM. These results reflect a complex mode of interaction. Indeed, from the onset, this is a highly unusual interaction since the N- and C-terminal hydrophobic pockets of CaM typically bind to amphipathic α-helical domains within their target domains49,152,207. The target α-helical regions usually span about twenty amino acid residues209,210. DSSP algorithm analysis of the STAC3 C1 NMR structure reveals a short three-84  residue 310 helix and four-residue α-helix in an otherwise beta-strand and disordered protein structure198,211. Thus, a systematic approach was employed to understand the binding interaction, testing each apo- or Ca2+-CaM lobe independently. Under calcium conditions, C-lobe had a measured affinity of ~42µM, whereas N-lobe revealed much weaker binding to the C1 domain (Fig. 3.4 A-C). Since full-length CaM had a stronger affinity to the C1 domain than each lobe individually, it suggests that both lobes contribute to the interaction. A similar trend was observed under apo-conditions (Fig. 3.4 D-F).    Figure 3.3 The STAC1 C1 domain does not interfere with CaM’s ability to bind to a CaV1.2 IQ peptide. A. No evidence of a direct interaction between STAC1 C1 and CaV1.2 IQ peptide. Shown is a titration of 500µM STAC1 C1 into 50µM CaV1.2 IQ peptide. B. Shown is a titration of 1mM STAC1 C1 titrated into a cell containing a complex of 0.1mM apo-CaM-IQ. Inset, background titration of 1mM STAC1 C1 into buffer. Comparison of the heats of reaction of STAC1 C1 titrated into the apo-CaM-IQ complex or buffer revealed similar heats. C. Titration of 1mM STAC1 C1 into a cell containing a complex of 0.1mM Ca2+-CaM-IQ.    85   Figure 3.4 The STAC1 C1 domain interacts directly with CaM. A. Titration experiment where 1mM STAC1 C1 is injected into a cell containing 0.1mM Ca2+-CaM. B. Shown is a titration of 1mM Ca2+-N-lobe titrated into 0.1mM STAC1 C1. C. Titration of 1mM Ca2+-C-lobe into 0.1mM STAC1C1. To assess for the interaction with all these CaM constructs under Ca2+-conditions, all components were dialyzed against a buffer containing 10mM CaCl2, D. Shown is the titration of 1mM STAC1 C1 into 500µM apo-CaM. E. Titration of 1mM STAC1 C1 into a cell containing 0.1mM apo-N-lobe. F. Titration of 1mM STAC1 C1 into 0.1mM apo-C-lobe. The CaM constructs in these apo-experiments were initially dialyzed against a buffer containing 10mM EDTA. Subsequent dialysis against buffers without EDTA was used to remove the EDTA. The final buffer composition used in titrations D-F is: 25mM KCl, 10mM HEPES pH 7.4, 20µM ZnSO4.    86  Table 3.1 Thermodynamic parameters for the interaction between full-length, N- and C-lobe CaM and STAC1 C1    3.3.3 An intrinsically disordered STAC3 peptide binds to Calmodulin The finding that STAC proteins can modulate CaV1.2 inactivation kinetics may in part be due to STAC proteins being able to interact with CaM. Many known CaM-binding proteins possess a region of approximately twenty residues that is often characterized by moderate hydrophobicity, a net positive charge and a propensity for forming basic amphipathic helices209,210. A peptide located in the intrinsically disordered region of STAC3 (residues 175-194) (Fig.3.5A), just preceding the two SH3 domains, was predicted to be a CaM binding site209. This interaction is Ca2+-dependent as binding to the STAC3 peptide was only detected with Ca2+-CaM (Kd ~ 19.2µM) but not apo-CaM (Fig. 3.5 B and E). To determine which CaM lobe is primarily responsible for the interaction, each Ca2+-loaded lobe was titrated into the STAC3 peptide. Binding was detected for each lobe, albeit dramatically weaker than full-length CaM (Table 3.2, Fig 3.5 C and D), which is an indication that both lobes contribute toward binding this STAC3 peptide. An interesting feature of the thermodynamic binding parameters was that endothermic heats of interaction was detected for STAC3 binding to full-length CaM; however, when tested individually, each lobe contributed weak exothermic heats of interaction. While a Titrated constructs Kd (µM) N-value ΔH  (cal.mol-1) ΔS  (cal.mol-1.K-1) Number of Replicates STAC1 C1/Ca2+-CaM 4.45 ± 2.04 0.78 ± 0.05 -1262 ± 49 20.5 ± 1.2  2 Ca2+-C-lobe/STAC1 C1 42.3 ± 14.4 0.27 ± 0.13 -9063 ± 4783 -10.5 ± 16.9 3 Ca2+-N-lobe/STAC1 C1 Unable to obtain reliable fit 1 STAC1 C1/Apo-CaM Binding detected - unable to obtain reliable fit 3 STAC1 C1/Apo-C-lobe Binding detected - unable to obtain reliable fit 3 STAC1 C1/Apo-N-lobe Binding detected - unable to obtain reliable fit 2 87  thorough explanation for this characteristic is lacking, it clearly demonstrates that this interaction needs to be fully investigated with functional experiments.  Again, making use of the putative CaM-binding site prediction server209, a modified STAC3 synthetic peptide was designed which converted three hydrophobic residues, Val185, Ile186 and M187, into a string of three glutamate residues. We hypothesized that this triple-STAC3 mutant would abolish the interaction with Ca2+-CaM. Glutamate was chosen preferentially over alanine, since we assumed the methyl sidechain of alanine residues could still create a hydrophobic patch to engage with Ca2+-CaM. ITC experiments confirmed this prediction (Fig 3.5E): the mutant peptide was ~300-fold weaker compared to the WT peptide. While this set of ITC experiments is the first reports of a direct interaction between CaM and STAC3, the functional relevance of these interaction – if it, as hypothesized, plays a role in CaV1.2 inactivation – is currently under investigation. Undoubtedly, evidence of these interactions adds to the rich complexity of how the L-type voltage-gated calcium channels may be regulated.  Table 3.2 Thermodynamic parameters for the interaction between full-length, N- and C-lobe CaM and a STAC3 peptide.     Titrated constructs Kd (µM) N-value ΔH  (cal.mol-1) ΔS  (cal.mol-1.K-1) Number of Replicates Ca2+-CaM/STAC3 175-194 19.2 ± 0.37 1.18 ± 0.19 -2584 ± 315 30.3 ± 1.10 3 Ca2+-CaM N-lobe/ STAC3 175-194 581 ± 128 1.0 (fixed) -3644 ± 228 2.64 ± 1.21 2 Ca2+-CaM C-lobe/ STAC3 175-194 325 ± 59 1.0 (fixed) -809 ± 28 13.25 ± 0.45 2 Ca2+-CaM/STAC3 175-194 V185E I186E M187E 6.07 ± 1.44mM 1.0 (fixed) -18130 ± 3680 -50.7 ± 12.9 2 Apo-CaM/ STAC3 175-194 No binding detected 2 88  Figure 3.5 ITC experiments reveal binding between an intrinsically disordered STAC3 peptide and Ca2+-CaM. A. Schematic arrangement of the STAC domains, with the sequence and approximate location of the intrinsically disordered peptide highlighted. B. Shown is 2mM Ca2+-CaM titrated into 0.2mM STAC3 175-194, an intrinsically disordered peptide. C. Titration between 2mM Ca2+-N-lobe and 0.2mM STAC3 175-194. D. Titration between 2mM Ca2+-C-lobe and 0.2mM STAC3 175-194. E. A triple mutation within the STAC3 peptide, V185E I186E M187E, dramatically reduces the binding affinity between STAC3 and Ca2+-CaM. Shown is 2mM Ca2+-CaM titrated into 0.2mM STAC3 175-194 V185E I186E M187E. A Ca2+-containing environment was created by dialyzing CaM constructs against a buffer containing 10mM CaCl2. The STAC peptide was dissolved in the CaCl2 dialysis buffer. F. No binding is detected, via ITC, between apo-CaM and STAC3 175-194. Shown is the titration of 2mM apo-CaM titrated into 0.2mM STAC3 175-194. Inset, background titration showing 2mM apo-CaM into buffer. Comparison of the heats of reaction between the titration of apo-CaM into STAC3 or buffer 89  reveal similar heats. To ensure a Ca2+-free experiment, CaM was dialyzed in a buffer containing 10mM EDTA.   3.4 Discussion Voltage-gated calcium channels possess remarkable features that allow them to autoregulate their activity, thereby preventing excessive Ca2+ influx during prolonged membrane depolarization61. Voltage-dependent inactivation is an intrinsic property of all subtypes of the voltage-gated calcium channels, while calcium-dependent inactivation is a tunable feedback mechanism of the high-voltage activated calcium channels (encoded by CaV1.x or CaV2.x pore forming subunits)60,61. Through interaction with a constitutively tethered Ca2+ sensor, CaM, the channel is able to adjust its Ca2+ flux in response to local rises in Ca2+ 39,40. Channel inactivation in high-voltage activated calcium channels is also potently modulated by the CaVβ-subunit61,161. With the advent of identifying new components within the calcium channel complex, channel inactivation remains under intense investigation as we try to understand the elements involved in this process.  Of late, the spotlight has focused on a previously uncharacterized component of the EC coupling machinery: STAC3. In addition to establishing that STAC3 plays an essential role in mediating skeletal muscle contraction48,148,156,199, STAC3 was found to aid expression of CaV1.1 in the plasma membrane, and to slow down the inactivation of CaV1.2, the cardiac and neuronal channel isoform, in tsa201 cells162. We have managed to recapitulate STAC3’s effect on CaV1.2 inactivation using Xenopus oocytes. Here, we aimed to explore the region(s) of STAC3 involved in mediating its effect on channel inactivation.  90  Introducing the NAM mutation into STAC3 or the CaV1.2 M853A P854A R858A triple mutant, which effectively abolished binding to the CaV1.1/1.2 II-III loop199, had no effect on STAC3’s ability to delay channel inactivation. These findings fueled the hypothesis that STAC3’s ability to modulate CaV1.2 inactivation lies within another region of the protein. Indeed, a truncated STAC3 construct, lacking the two C-terminal SH3 domains, was still able to significantly affect the channel’s inactivation kinetics. In contrast, a construct which only contained the two tandem SH3 domains had no effect on inactivation, and the channel displayed kinetics similar to when STAC3 is not co-expressed. Combined, these results suggested that the N-terminal portion of the STAC3 protein – either the C1 domain together or separately from the disordered region flanking this domain – may be involved in propagating the adaptor protein’s effect on channel inactivation.  Comparing CaV1.2’s inactivation kinetics when conducting either Ca2+ or Ba2+ current, revealed that conduction of both ions was affected in the presence of STAC3, although the effect on Ca2+ current was more pronounced. Using a similar approach to that described by Barrett and Tsien (2008), the ratio of ICa/IBa205 was plotted for oocytes either null or co-expressing STAC3 (Fig. 3.2E). The marked difference in the presence of STAC3 revealed that STAC3 influences both CDI and VDI.  Our finding that the STAC3 SH3 domains did not participate in channel inactivation shifted our attention to the N-terminal regions of the adaptor protein. Indeed, Campiglio et al. recently suggested that the STAC C1 domain plays an essential role in mediating complex formation in skeletal muscle triads and modulating the inactivation properties of CaV1.2 currents200. Their results highlighted the importance of the STAC C1 domain interacting with the channel’s IQ domain, however, whether a direct interaction existed between the two proteins or 91  if the C1 domain interacted with the Ca2+-CaM-IQ domain complex remained to be determined. Here, our ITC experiments failed to detect evidence of direct binding between the STAC1 C1 domain and a synthetic peptide encoding the IQ domain. Concurrently, no discernable binding isotherm could be generated from the titration of the C1 domain into a cell containing either apo- or Ca2+-CaM in complex with the CaV1.2 IQ peptide. However, our reference titration, where the C1 domain was titrated into a cell containing CaM, hinted of an interaction between these two proteins. STAC1 C1 binding to CaM is not Ca2+-dependent, as the C1 domain interacted with both apo- and Ca2+-CaM. In both conditions, both CaM lobes appeared to contribute to the binding of the STAC1 C1 domain.   To further complicate CaM binding to STAC proteins, we identified a putative CaM binding site, located in the intrinsically disordered region of STAC3 (residues 175-194), using the Calmodulin Target Database209. CaM binding to this STAC3 peptide is Ca2+-dependent, since binding was detected under Ca2+ conditions, but not in the presence of 10mM EDTA (apo-CaM) (Fig. 3.5). Systematic titrations with Ca2+-N- and C-lobes revealed that each lobe weakly contributes to binding. Collectively, these interactions provide the first evidence of direct binding between CaM and STAC proteins, reinforcing the notion that STAC proteins may exert their effect on channel inactivation through CaM.  These results add to the complex milieu of the channel’s regulation. We have described two independent binding sites for CaM on the STAC proteins, which need further functional investigation. However, these interactions suggest that STAC proteins can tether multiple components of the channel complex. The channel’s II-III loop can be brought into proximity to CaM via STAC3: the first SH3 domain binds to the II-III loop and CaM is capable of binding to the STAC C1 domain and intrinsically disordered region. STAC3 binding could serve as a 92  mechanism of coupling the channel’s two functions (EC coupling and conducting calcium ions) together. Furthermore, the structure of the complex between STAC2’s tandem SH3 domains and the CaV1.1 II-III loop peptide revealed that the canonical binding interface of the second SH3 domain is still available for binding199. Thus, the possibility remains that STAC3 could interact with RyR1, thereby acting as a bridge between CaV1.1 and RyR1. However, whether RyR1 and STAC3 interact remains to be determined.  In conclusion, STAC3 has roles in both EC coupling and CaV1.2 channel inactivation. Our results reveal that the N-terminal regions of STAC3 are involved in mediating its role in channel inactivation, and our ITC experiments reveal the first reported direct interactions between STAC proteins and CaM. By binding different segments of the channel complex via independent binding sites, STAC3 may act as a linker between components of the EC coupling machinery.          93  Chapter 4: Structural insights into the unique anion binding site within the cardiac ryanodine receptor N-terminal region2 4.1 Introduction Ryanodine Receptors are the pinnacle of ion channel complexity. The large size of the channel, exceeding a molecular weight of 2.2MDa, facilitated early cryo-EM studies which determined that the channel has a vast cytoplasmic cap connected to a narrower “stalk”, encompassing the channel’s transmembrane region102–104,106. Resembling a mushroom, the expansive cytoplasmic cap encompasses ~80% of the channel’s volume, and serves as the docking site for a myriad of small molecule ligands and auxiliary proteins73 which allosterically modulate channel gating. Indeed, Samso et al. (2009) published structures of the channel in both the open and closed conformation, which provided a platform to assess the coordinated movement in the cytoplasmic and transmembrane domains incurred upon channel opening106.  Dysregulation of Ca2+ homeostasis can have devastating consequences, particularly within skeletal and cardiac myocytes77. Disease severity imparted by sometimes subtle point mutations in the ryr genes reflect the importance of the channel in the tight regulation of Ca2+ release during muscle contraction73,118,212. The cardiac isoform, RyR2, is the target for over 150 mutations, many of which are associated with catecholaminergic polymorphic ventricular tachycardia, a condition that can degenerate into cardiac arrest and cause sudden death213,214.                                                  2 Sections of this chapter has been adapted from two original publications: Liu, Y., Wei, J., Wong King Yuen, S.M., Sun, B., Tang, Y., Wang, R., Van Petegem, F., and Chen, S.R.W. CPVT-associated cardiac ryanodine receptor mutation G357S with reduced penetrance impairs Ca2+ release termination and diminishes protein expression. PLOS One (2017) 12(9):e0184177. doi: 10.1371/journal.pone.0184177. Xiao, Z., Guo, W., Wong King Yuen, S.M., Wang, R., Zhang, L., Van Petegem, F., and Chen, S.R.W. The H29D mutation does not enhance cytosolic Ca2+ activation of the cardiac ryanodine receptor. PLOS One (2015) 10(9):e0139058. doi: 10.1371/journal.pone.0139058.  94  RyR2 mutations have also been linked with arrhythmogenic right ventricular dysplasia type 2 (ARDV2) and idiopathic ventricular fibrillation215. A general observation is that the majority of the disease mutations cause a gain-of-function phenotype, rendering the channel more sensitive toward stimuli, such as cytosolic or luminal Ca2+ 73. Crystal structures of amino-terminal domains of RyR1 were the first high-resolution glimpses into the channel110,117,118 and have provided insights into how disease-associated mutations affect channel function. Pseudo-atomic models have located the first three amino-terminal domains of RyR1 on the cytosolic face of the channel, forming a vestibule around the four-fold symmetry axis110. The crystal structure covers the first 559 residues of RyR1, a region that encompasses most of the N-terminal disease hotspot. This hotspot folds up as three independent domains – domains A, B and C – that interact with each other in a compact form. The domain-domain interfaces are mainly hydrophilic in nature and are inherently weak212. Mapping the location of RyR1 and RyR2 disease mutations onto the pseudo-atomic model has led to the observation that many are located at the interface between domains A, B, and C, and thereby seem to destabilize the domain-domain interactions (Fig. 4.1A). Several of the mutations have abolished hydrogen bond pairs or charge-coupling interactions bridging the domains110,216 (Fig. 4.1B). Another group of mutations are buried within individual domains, and thus propagate the disease phenotype by causing local misfolding or protein destabilization. The final cluster of disease-associated mutations are exposed to the surface of the domains, altering the local environment. These local surfaces could be involved in binding other RyR domains or auxiliary proteins118. In fact, one hypothesis has suggested that there is a direct interaction between the N-terminal and central disease hotspots which is “unzipped” during channel opening through allosteric coupling93,94,217. Disease mutations located at this interface weaken the 95  domain-domain interaction, increasing the propensity for “unzipping”, which causes channel activation and leakiness93. Kimlicka et al. (2013) have provided an additional model by illustrating that the intersubunit contacts of the N-terminal region, mediated by domains A and B from neighbouring subunits, can be disrupted upon channel opening (Fig. 4.1B). In this capacity, the N-terminal region acts as a “brake” against channel opening216. Figure 4.1 Mapping disease-associated mutations in the amino-terminal domains of RyR1 and RyR2. A. Location of disease mutations in RyR1 (magenta) and the equivalent positions of RyR2 mutations (blue). Regions not associated with mutations are in grey. For clarity, mutations in flexible loops have been omitted. B. A cluster of disease mutants located at the interface of domains A and B of neighbouring protomers. Disease-associated mutants are shown in stick format (grey). Hydrogen bonds between residues are indicated by black dashed lines. Inset, the location of domains A, B, and C (PDB: 2XOA) docked in the 3.8Å RyR1 cryo-EM map (EMD: 2807).  The fold of the three individual amino-terminal domains of RyR1 and RyR2 is conserved, as is their compact domain arrangement119. Despite the similarities, there are key differences which close examination of the RyR1ABC and RyR2ABC structures reveal. Residues from all 96  three of the RyR1 N-terminal domains contribute to an elaborate hydrogen bond and salt bridge interaction network which is key to the protein’s stability (Fig. 4.2B)119. Despite all the ionic interaction partners being conserved within RyR2, the replacement of two histidine residues with Tyr125 and Arg420 within RyR2 is enough to abolish the ionic network and introduce a unique central anion binding site within RyR2 (Fig. 4.2C). The central chloride anion is coordinated by electrostatic interactions with arginine residues from domains B and C119. Since this anion binding site is the target for disease mutations (R420Q, R420W)218–220 that cause CPVT, it suggests anion binding affects relative domain orientations and may regulate RyR2119. Attempts to remove chloride resulted in protein precipitation, further reflecting an essential role of anion binding in protein stability.   Here, we will use an integrated approach to understand the nature of the anion binding site within RyR2ABC. Following up from the work of Kimlicka et al. (2013), a structural analysis of the R420W disease mutation will be completed to assess how the mutation affects chloride binding. The N-terminal region of RyR2 harbours over 30 distinct disease mutations and more are continuously being identified. In addition to exploring anion binding in RyR2ABC, we will also analyze the effects of two disease-associated mutations, H29D221 and G357S222.  4.2 Experimental Procedures 4.2.1 Expression constructs All RyR2ABC constructs, residues 1-547 (mouse numbering), were cloned into a modified pET28 vector (Novagen), containing a His6-tag, MBP and a cleavage site for TEV protease49. Mutations, H29D, G357S, and R420W, was achieved by the QuikChange protocol (Stratagene). All constructs were confirmed by sequencing (Eurofins). 97  Figure 4.2 Comparison of the amino-terminal domains of RyR1 and RyR2. A. Superposition of RyR1ABC (grey) and RyR2ABC (domain A in blue, B in green, and C in red). The structures have been superposed based on domain A. The chloride ion present in the RyR2ABC structure is represented as a pink sphere. B. Extensive interaction network between the three domains of RyR1ABC. The salt bridge network involves residues in all three domains. Black dashed lines indicate hydrogen bonding pairs. Red dashed lines indicate a salt bridge between Arg283 and Asp61 that is not formed via hydrogen bonds. C. An equivalent interdomain view in RyR2ABC. Domains are coloured as in A. Arg298 and Asp61 form a salt bridge that is not mediated by hydrogen bonds, this is represented by red dashed lines. Two histidine residues in RyR1 are replaced by Arg420 and Tyr125. This results in a disruption of the ionic pair network and the creation of a chloride binding site. A hydrogen bond is formed by an interaction between Arg420 and Tyr125. Key arginine residues which coordinate the chloride anion (pink sphere) are highlighted.    4.2.2 Protein expression and purification RyR2ABC constructs were expressed for 20-24 hours in Escherichia coli Rosetta (DE3) pLacI (Novagen) grown in 2xYT media at 18oC. Cells were lysed by sonication in buffer A (250mM KCl and 10mM HEPES pH7.4) supplemented with 25μg/mL DNase I, 25μg/mL lysozyme, 14mM β-me, and 1mM PMSF. Lysates were applied to a 25mL Poros MC column (Tosoh Bioscience), washed with 5CVs of buffer A and 5CVs of buffer A plus 2% (vol/vol) buffer B (250mM KCl and 500mM imidazole, pH 7.4) and eluted with 60% (vol/vol) buffer B. The protein was dialyzed against buffer A plus 14mM β-me and was cleaved simultaneously 98  with recombinant TEV protease. The protein was then applied to a 25mL amylose column (New England Biolabs), washed with 10CVs of buffer A plus 14mM β-me, and eluted with buffer A supplemented with 10mM maltose and 14mM β-me. The protein was applied again to the 25mL Poros MC column in buffer A and then run on a HiLoad 16/10 Phenyl Sepharose HP column (GE Healthcare) in buffer D (1M KCl, 10mM Tris, pH8.8, and 14mM β-me) and eluted with a gradient of 0% to 50% buffer E (50mM KCl, 10mM Tris, pH8.8, and 14mM β-me). The protein was subsequently applied to a HiLoad Q-Sepharose HP column (GE Healthcare) with buffer E and eluted with a gradient of 0% to 50% with buffer D. Lastly, the quality of the proteins was assessed by a HiLoad 16/60 Superdex200 preparatory-grade gel filtration column (GE Healthcare) equilibrated with buffer A. The RyR2ABC H29D construct followed a similar purification protocol, except buffer A contained 10mM phosphate, pH8.5 and buffer B consisted of 500mM imidazole, pH 8.5. The H29D mutant also required an extra amylose column prior to the Superdex200 column. The identity and purity of each RyR2ABC construct was confirmed via SDS-PAGE.   4.2.3 Thermal melt analysis Mid-point unfolding temperature curves of the RyR2ABC constructs were measured by ThermoFluor experiments223, as described before for other RyR constructs110,118,119. Briefly, samples for melting curves contained 50µL of purified protein at 0.1mg/mL and 1x SYPRO Orange solution (Invitrogen) using manufacturer’s instructions in a buffer consisting of 250mM KCl, 10mM HEPES pH7.4, and 14mM β-me. Data was obtained in a DNA Engine Opticon®2 real-time PCR machine (Bio-Rad) using the SYBR green filter option. The temperature incrementally increased from 20°C to 95°C in 0.5°C steps. At every step, the temperature was 99  kept constant for 15 seconds. Curves were normalized and mid-point unfolding temperatures were obtained by taking the maxima of the first derivative of the curves. The values described are the averages from three replicates, unless otherwise described.  4.2.4 Crystallization and structure determination In preparation for crystallography, the RyR2 Y125H R420H double mutant was exchanged to a buffer containing 250mM KBr, 10mM HEPES pH 7.4, and 14mM β-me using a 10kDa MWCO concentrator (Millipore). RyR2 Y125H R420H was crystallized using the sitting drop vapor diffusion method at room temperature by mixing equal volumes of protein (8-10mg/mL) and well solution. Large protein crystals were obtained in a buffer spectrum covering 0.1M Tris pH 6.93 to 8.0 and a PEG8000 range of 4% to 8% (w/v). Similarly, the RyR2 R420W mutant was exchanged to a buffer containing 250mM KBr, 10mM HEPES pH 7.4, and 14mM β-me. RyR2 R420W was crystallized using the sitting drop vapor diffusion method at room temperature by mixing equal volumes of protein (~10mg/mL) and well solution. Crystals were obtained in precipitants containing 0.1M HEPES pH’ed to either 7.3 or 7.5 and 6% (w/v) PEG6000. Prior to being flash frozen in liquid nitrogen, both RyR2 Y125H R420H and RyR2 R420W were transferred to a drop containing mother liquor and 30 and 35% (v/v) glycerol, respectively. Diffraction data was collected at the Stanford Synchrotron Radiation Lightsource beamline BL9-2.  Data sets were processed using XDS170. The structure of both mutant RyR2ABC constructs was solved by molecular replacement via Phaser172, using the WT-RyR2ABC structure (PDB: 4L4H) as the search model. All models were completed with iterative cycles of 100  manual model building in Coot173 and refinement with Refmac5174. Table 4.1 highlights the statistics for crystallographic data collection and refinement statistics.   4.2.5 Time-resolved FRET A RyR2ABC construct (residues 1-547) with all its surface cysteines removed (C24/36/47/131/132/244/361A) was generated to prevent non-specific labelling. Mutations were generated sequentially by QuikChange site-directed mutagenesis (Stratagene). For simplicity, this construct has been termed “Cys-lite” and was used as a labelling control. Two cysteines were engineered onto the Cys-lite construct at strategic locations that would facilitate fluorophore-labelling and could allow different conformational changes to be detected. R383 (located in Domain B) and K441 (domain C) were mutated to cysteines. Success of mutagenesis was assessed by sequence analysis (Eurofins). Constructs were expressed and purified as described in sections 4.2.2. The RyR2ABC R383C/K441C construct was labelled with Alexa Fluor 350 C5 maleimide (AF350, donor) and Alexa Fluor 488 C5 (AF488, acceptor). Labelling was completed by incubating 50µM RyR2ABC R383C/K441C with 167µM AF350 and 333µM AF488, 1mM TCEP, 50mM KPO4, 150mM KCl, 100mM Tris-HCl, pH 7.5 for 3 hours at 21oC with rotation. The sample was dialyzed in 50mM KPO4 buffer, pH 7.4 at 4oC and excess dye was removed via an Amicon stirred cell (10 kDa MWCO, ~20 psi).  Time-resolved fluorescence decay of dual labeled RyR2ABC R383C/K441C samples was measured by time-correlated single-photon counting (TCSPC, Becker-Hickl, Berlin, Germany). Excitation was followed at 385 nm using a sub-nanosecond pulsed diode laser (PicoQuant, Berlin, Germany), filtering the emitted light using a 440/40 filter (Semrock, New 101  York) and detection with a PMH-100 photomultiplier (Becker-Hickl). The instrument response function (IRF) was recorded from water. Global multi-exponential analysis of the TR-FRET data was used to test a series of structural models. A three component (donor only fraction + two distance distribution) was found to provide the best fit for the TR-FRET data. A two-Gaussian distribution was the best fit for the data based on chi-square (χ2). TR-FRET data was collected by Dr. Razvan Cornea’s lab (University of Minnesota).  Table 4.1 Crystallographic data collection and refinement statistics for RyR2ABC mutant constructs   RyR2 Y125H R420H RyR2 R420W DATA COLLECTION      Space group P42 21 2                P42 21 2                    Cell dimensions         a, b, c (in Å) 78.69, 78.69, 250.24 79.20, 79.20, 249.06       α, β, γ (in o) 90.0, 90.0, 90.0 90.0, 90.0, 90.0    Wavelength 0.826570 0.826570    Resolution (in Å) 38.87-2.57 (2.66-2.57) 46.43-2.85 (2.95-2.85)    Rmerge 7.71 (1.20) 17.02 (1.72)    I/σI 28.74 (1.86) 16.92 (2.24) Completeness  99.74 (98.24) 99.84 (100.0)     Redundancy 13.7 (10.4) 12.0 (12.3) REFINEMENT      Resolution (in Å) 38.87-2.57 46.42-2.85    No. of reflections 26011 (2505) 19399 (1901)    Rwork/Rfree 23.24/26.80 26.78/31.77    No. atoms         Protein 3678 3565       Water  1    B-factors         Protein 58.04 80.37       Water  56.06    Ramachandran favored (%) 93.46 92.81    Ramachandran allowed (%) 5.45 6.97    Ramachandran outliers (%) 1.09 0.22   rmsd         Bond lengths (in Å) 0.011 0.007       Bond angles (o) 1.23 1.065 102  4.3 Results 4.3.1 A RyR1-RyR2ABC hybrid alleviates RyR2ABC’s chloride-dependence but does not restore the hydrogen bond and salt bridge network of RyR1ABC. Although the domain arrangement is conserved between the ABC domains of RyR1 and RyR2, noticeable differences between the isoforms exist in the interdomain region of their N-terminal domains. As described by Kimlicka et al. (2013), residues from the three N-terminal domains of RyR1 participate in a network of hydrogen bonds and salt bridge interactions119. Arg283 (Domain B) forms a salt bridge with Asp61 (domain A), which in turn interacts with the positively charged Arg402 (domain C). Arg402 makes an additional salt bridge with Glu40 (domain A). Furthermore, Glu40, Asp61, and Arg402 form multiple hydrogen bonds with one another. This network of interactions confers greater stability to the protein compared to individual pairs, and thus is a predominant factor in RyR1ABC’s stability. In RyR2ABC, this network is disrupted. Despite conservation of the four residues involved in the network, His405 (RyR1) on domain C is replaced with Arg420 in RyR2. Instead of the four-residue network described for RyR1, RyR2ABC only has a single salt bridge formed by Arg298 and Asp61. All hydrogen bonds within the network are lost. Re-arrangement of the residues within this interdomain region introduces a binding pocket for a chloride ion. The anion is coordinated by domains B and C’s arginine sidechains’ electrostatic interactions (Arg420, Arg298, and Arg276). RyR2’s Tyr125 takes the place of RyR1’s His113 and forms a hydrogen bond with its sidechain hydroxyl group and the Arg420 sidechain.  To assess the nature of the anion binding site in RyR2ABC, we opted to create a RyR1-RyR2ABC hybrid, where we re-introduced the two histidine (His405, His113) residues present in RyR1 into RyR2: Y125H and R420H. As predicted, introducing the two histidine residues into 103  RyR2ABC alleviated the protein’s dependence on chloride (or any halide anion) for stability. ThermoFluor assays where completed with RyR2ABC Y125H R420H in buffers containing 0mM, 16mM, 26.5mM, 44mM, 135mM, 225mM KCl, and a condition in which the protein was initially at 0mM KCl, and then dialyzed against a buffer containing 250mM KCl (Table 4.2, Fig. 4.3). Kimlicka et al. (2013) reported a mid-point unfolding temperature, TM, for wildtype RyR2ABC, solubilized in a buffer containing 150mM KCl, of 43.1 ± 0.4oC119.  Comparing the range of TM values for RyR2ABC Y125H R420H (42.17oC -45.00oC) against the measured value of WT RyR2ABC, we can see that the double mutant removes the dependence of the protein on chloride.   Figure 4.3 Thermal stability curves of RyR2ABC Y125H R420H in buffers containing various concentrations of KCl.  Introducing the two histidine residues, which are native to RyR1ABC, into RyR2ABC alleviates RyR2’s dependence on chloride for protein stability. WT-RyR2ABC has a mid-point unfolding temperature of 43.1 ± 0.4 oC as determined by Kimlicka et al. (2013).  00.250.50.75132.00 37.00 42.00 47.00 52.00 57.00Normalized FluorescenceTemperature (oC)  0mM KCl  16mM KCl  26.5mM KCl  44mM KCl  135mM KCl  225mM KCl  Re-introduced KCl104  Table 4.2 Mid-point unfolding temperature, TM, for RyR2ABC Y125H R420H in buffers of varying KCl concentration  For all conditions, the data presented reflects the TM ± SEM, where the TM was obtained from taking the average maxima of the first derivative of three replicates.   * Assay for the 0mM KCl and re-introduced KCl conditions were performed on a separate day  In order to see whether the RyR2ABC Y125H R420H double mutant indeed disrupted the chloride binding site, we solved its crystal structure. As wildtype RyR2ABC can bind both chloride and bromide119, we crystallized the mutant in the presence of 250mM KBr. Analysis of the interdomain region of RyR2ABC Y125H R420H yielded no density for a halide ion, confirming our hypothesis that removal of Arg420 abolishes halide ion binding. Surprisingly, introducing the histidine residues did not restore the salt bridge and hydrogen bond network that is observed in the RyR1ABC structure. The salt bridge between Arg298 and Asp61 is the only interaction that remains conserved in this RyR2 double mutant (Fig. 4.4B). The double mutation and removal of chloride imparted small perturbations of the residues around the chloride binding site and consequently, relative domain reorientations are incurred - a result of tilting around an axis near the domain-domain interfaces. Superposition with WT-RyR2ABC revealed that domains A and B maintain the same relative orientation, but both domains are displaced relative to domain C with a maximal shift of ~0.45Å furthest away from the tilting axis (Fig 4.4A)    [KCl] 0mM* 16mM 26.5mM 44mM 135mM 225mM Re-introduced KCl* TM (oC)  45.00 ± 0.00 42.50 ± 0.00 42.17 ± 0.17 43.00 ± 0.50  43.00 ± 0.29  42.50 ± 0.00 44.17 ± 0.17 105   Figure 4.4 RyR2ABC Y125H R420H abolishes halide ion binding and imparts local conformational rearrangement. A. Superposition of WT-RyR2ABC (grey) and RyR2ABC Y125H R420H (domain A in blue, B in green, and C in red). The chloride ion seen in the WT-RyR2ABC structure is represented as a grey sphere for reference. The two structures are superimposed relative to domain C. Removal of chloride from the RyR2ABC double mutant causes domains A and B to move upward relative WT-RyR2ABC. The range of motion spans approximately 0.452Å. B. Replacing Tyr125 and R420 with histidine removes the halide ion binding site but does not restore the hydrogen bond and salt bridge network observed in RyR1ABC. The salt bridge between Arg298 and Asp61 is the only interaction that remains conserved.    4.3.2 The RyR2 disease mutant, R420W, abolishes chloride binding Two RyR2 disease mutants directly target the chloride binding site: R420Q and R420W119. The structure of RyR2ABC R420Q has been described by Kimlicka et al. (2013). Introducing the R420Q mutation abolished chloride binding and resulted in relative domain orientations due to tilting around an axis near the domain-domain interfaces119. Orientations on the order of ~1.4Å from the tilting axis was measured, again suggesting that chloride plays a role in mediating relative domain orientations. We followed up on this analysis by assessing 106  structural consequences of introducing R420W to RyR2ABC. R420W has been associated with effort-induced polymorphic ventricular arrhythmia, arrhythmogenic right ventricular dysplasia type 2, and reducing the threshold for Ca2+-release termination219,220. Similar to R420Q, the R420W mutation did not appear to perturb the protein’s stability as thermal stability assays yielded a mid-point unfolding temperature of 43.70 ± 0.12oC (average TM ± SEM, n=5 replicates) which is of comparable stability to WT-RyR2ABC (43.1 ± 0.4oC, Kimlicka et al.).  Structural analysis revealed that chloride binding was abolished. Fo-Fc density, contoured at 1σ, revealed no evidence of a halide ion in the interdomain region. The mutation coupled with removal of chloride imparted small perturbations of the residues around the chloride binding site and consequently, global domain reorientations were observed. Superposition with WT-RyR2ABC revealed that Domain A has a maximal displacement of 1.06Å relative to Domain C; and Domain B has a maximal shift of 0.8Å relative to Domain C (Fig. 4.5A). It is prudent to note that these domain reorientations are observed in isolated channel fragments, and thus may not represent the full extent of conformational change in full-length RyR2. The salt bridge between Arg298 and Asp61 remains conserved in this disease mutant. Interestingly, the local rearrangement of Arg417 and Try125 results in the formation of two additional salt bridges: Asp61 interacts with Arg417, which also interacts with the sidechain of Glu40 (Fig4.5B). Therefore, the R420W mutation seems to cause a salt bridge network similar to the one observed in RyR1. This network is likely key to the stability of R420W and accounts for the similar thermal stability compared to WT-RyR2ABC.     107  Figure 4.5 Removal of chloride in RyR2ABC causes local residue re-arrangement and global domain reorientation.  A. Superposition of WT-RyR2ABC (grey) with the R420W mutant (Domain A in blue, B in green, and C in red). The chloride ion of WT-RyR2ABC is represented as a grey sphere for context. Domain conformational changes can be observed in Domains A and B since the R420W structure was superimposed relative to WT-Domain C. B. Introducing Trp420 results in local residue re-arrangement in the interdomain region, which restores several salt bridge interactions (red dashed lines). C. Comparison of large-scale conformational changes of human RyR2ABC (domains colour-coded as in A, PDB: 4JKQ) vs. mouse RyR2ABC (grey, PDB: 4L4H). No halide ion was observed in the structure of human RyR2ABC, while chloride was present in the crystal of mouse RyR2ABC. The two structures are superimposed relative to Domain C, where shifts up to 5.9Å were measured.   Borko et al. (2014) published the structure of the first 606 N-terminal residues of human RyR2. Interestingly, despite >90% sequence conservation between the N-terminal region of human and mouse RyR2, particularly Arg420, Arg298, and Arg276 (mouse numbering), the anion binding site is notably absent in the human RyR2 N-terminal region224. Instead, much like RyR1ABC, the N-terminus is held together by a network of interactions involving residues from Domains A, B, and C, which play an essential role in protein stability. Superimposing human (-chloride) and mouse (+chloride) RyR2ABC relative to domain C reveals global conformational changes as a result of a tilt around an axis near the domain-domain interfaces. Although Domains A and B maintain the same relative orientation, both domains are displaced relative to Domain C 108  with maximal shifts of ~5.9Å furthest away from the tilting axis (Fig4.5C). This finding again alludes to a potential role chloride may have in modulating channel conformational changes.  4.3.3 Spectroscopic measurements of conformational dynamics incurred upon chloride binding Analyzing structures containing or lacking chloride, as described in sections 4.3.1. and 4.3.2., has suggested that chloride binding to RyR2ABC causes the protein to undergo global conformational changes. In collaboration with Dr. Razvan Cornea’s lab, time-resolved FRET (TR-FRET) measurements were used to quantitatively resolve the extent of conformational changes incurred upon chloride binding225. A RyR2ABC construct with all its surface cysteines removed (C24/36/47/131/132/244/361A) was generated to prevent non-specific labelling. For simplicity, this construct has been termed “Cys-lite”. Although three buried cysteines remain in the Cys-lite construct, these are crucial for protein stability, as removal caused protein precipitation during purification attempts. To measure the extent to which the domains re-arrange upon introduction of chloride, two cysteines were engineered onto the Cys-lite construct at strategic locations that would facilitate labelling and could allow different conformational changes to be detected. R383 (located in Domain B) and K441 (Domain C) were mutated to cysteines. These particular residues were chosen due to their position in flexible loops (instead, for example, disturbing residues located in β-strands and α-helices) and because their sidechains are not involved in crucial interactions with the rest of the construct. The engineered R383C and K441C was labelled with Alexa Fluor 350 C5 maleimide (donor) and Alexa Fluor 488 C5 maleimide (acceptor). This pair (R0 = 50 Å) was ideal due to a predicted distance of 43Å between residues R388 and K441 (based on the structure PDB: 4L4H). The efficiency and 109  specificity of labelling was determined against a control of Cys-lite RyR2ABC construct which did not contain any surface-exposed cysteine sidechains.   Using fluorescence lifetime measurements of FRET and model-based multi-exponential analysis of the acquired data identified two major distance components: R1 and R2. R1 has a donor-acceptor distance centered around 40Å, which matches the predicted distance measured from the crystal structure and had a relatively narrow Gaussian distribution (FWHM = 9Å) (Fig. 4.6B). R1 was not significantly responsive to KCl. In the absence of KCl, R2 is centered at 55Å and shifts to 60Å in response to KCl (Fig. 4.6C). However, the R2 component revealed a broader distribution (FWHM = 28Å) (Fig. 4.6B). These results suggest the coexistence of two structural states in RyR2ABC: a relatively ordered R1 and a more disordered R2. However, the shifts measured for R2 correlate well with the measurements made upon superimposing mouse (+Cl-) and human (-Cl-) RyR2ABC structures, where the largest shift observed was ~5.9Å (Fig.4.5C). Importantly, this demonstrated that the structural effects of chloride binding and unbinding, suggested by comparing crystal structures, are also observed in solution.           110   Figure 4.6 TR-FRET detection of chloride’s structural effect on RyR2ABC. Two engineered cysteines, R383C and K441C, were labeled with Alexa Fluor 350 C5 maleimide (donor) and Alexa Fluor 488 C5 maleimide (acceptor) onto an otherwise surface-cysteine-free RyR2ABC construct. A. Fluorescence decays of 800nM Donor (D)-RyR2ABC and Donor and Acceptor (DA)-RyR2ABC samples were acquired using TCSPC. D-only decays were fitted to 2-exponentials. Both D and D+A datasets were fitted to a two-Gaussian distance distribution. The concentration of KCl was varied. D-only decays overlap almost perfectly. DA decays show small differences due to [KCl], indicating subtle changes in the donor-acceptor distance relationship. B. Multi-exponential analysis of the TR-FRET data yielded a two-distance (R1 and R2) Gaussian distribution model of the separation between Domains B and C within RyR2ABC under various concentrations of KCl. C. R2 is centered around 55Å (FWHM=28Å) and its donor-acceptor distance increases to 60Å with increasing [KCl]. D. The R2 mole-fraction slightly decreases, while R1 increases, with increasing [KCl]. Figures were kindly provided by Dr. Razvan Cornea.   111  4.3.4 The H29D mutation does not affect the thermal stability of the N-terminal region of RyR2 nor does it enhance cytosolic Ca2+ activation of the cardiac ryanodine receptor To date, over 150 naturally-occurring, disease-associated mutations in RyR2 have been identified. Although a small fraction of these RyR2 mutations have been functionally characterized, the functional consequences for the vast majority remain to be fully explored. Improving the diagnosis and treatment of RyR2-linked diseases relies on understanding the functional impact of disease-associated RyR2 mutations.  The N-terminal region of RyR2 harbours over 30 distinct disease mutations, many of which are clustered at domain interfaces or buried within domains. Mutations located at these regions alter domain-domain interactions or the stability/folding of domains. Recently, a novel RyR2 mutation, H29D, has been identified and is associated with short-coupled polymorphic ventricular arrhythmia at rest221. Functional characterization using single channels revealed that the mutation, in comparison to single WT-RyR2 channels, significantly enhanced cytosolic Ca2+ activation at diastolic cytosolic Ca2+ concentrations. Cheung et al. (2015) proposed that the H29D mutation causes a “leaky” channel at low cytosolic Ca2+ concentrations, which may be a mechanism for the polymorphic ventricular tachycardia at rest.  Unlike other N-terminal disease mutations, the H29D mutation is located on the surface of the N-terminal domain, not at a domain interface (Fig. 4.7 A and B). Residue H29 is in a loop between two β-sheets within domain A and mapped to the surface of the RyR2 structure which faces the T-tubular membrane. Thus, it is unclear how this surface-exposed H29D mutation, that does not appear to interact with other parts of the RyR2 structure, could alter the intrinsic properties of the channel. Moreover, multiple crystal structures of the RyR2 amino-terminal region reveal that the H29 sidechain is largely flexible, further indicating that it forms no 112  significant interactions with any other residues119. To understand the mechanism by which the H29D mutation affects the intrinsic properties of the RyR2 channel, the H29D mutant was characterized at the molecular and cellular level using several functional and biochemical assays.   In partnership with Dr. Wayne Chen’s lab, they found that the H29D mutation did not alter the basal level or the Ca2+-dependence of [3H]ryanodine binding to RyR2226. Single channel analysis revealed that the H29D mutation did not affect cytosolic Ca2+ activation of single RyR2 channels nor was caffeine-induced calcium release in HEK293 cells altered. Furthermore, the H29D mutation did not alter the propensity for spontaneous Ca2+ release or the threshold for Ca2+ release activation or termination (SOICR).   The location of the H29 residue on the surface of the RyR2, away from any interface with other RyR2 domains or known auxiliary proteins (Fig. 4.7 A and B), supports the lack of functional effects described above. Disease-causing mutations may also exert their physiological phenotype by destabilizing the general protein fold, which could then indirectly lead to altered domain interactions. Indeed, many disease-causing mutations in RyRs have been found to significantly destabilize the fold, as indicated by a decreased thermal stability112,118,119,216. The surface location of H29D makes this an unlikely explanation to explain this mutation’s mechanism of action, however, we decided to confirm this by purifying the H29D variant of RyR2 (1-547) and subjecting it to thermal melt analysis. As shown in Fig 4.7C, the H29D mutant N-terminal region displayed a temperature-dependence of protein unfolding similar to that of the WT N-terminal region. Thus, this illustrates that the H29D mutation has no significant effect on the stability of the N-terminal domains of the RyR2 channel.    113   Figure 4.7 Effect of the H29D mutation on the thermal stability of the N-terminal domains of RyR2. A and B. Location of residue H29 in the three-dimensional structure of the N-terminal domains of RyR2 in the top view (A) and side view (B). Domains are coloured as follows: Domain A in blue, B in green and C in red. C. Representative thermal stability curves of the WT (filled circles) and the H29D mutant (open circles) N-terminal domains of RyR2 (RyR2ABC) (n=5 for each).    4.3.5 The CPVT-associated G357S RyR2 mutation, with reduced penetrance, diminishes protein expression and impairs Ca2+ release termination The mean age of CPVT symptom onset in individuals with RyR2 mutations is ~17 years old. Due to the highly lethal nature of the disease, most families with CPVT-associated RyR2 mutations are relatively small and demonstrate a high penetrance (50-90%)219,227–229. Intriguingly, a novel RyR2 mutation, G357S, has been identified in a large family with ten 114  generations and 1404 members222. Within this family, 179 carry the G357S mutant, 36 of whom suffered from sudden cardiac death.  The G357S mutation is located within Domain B in the N-terminal region of the RyR2 channel, and Liu et al. (2014) have suggested that Domain B is important for RyR2 protein expression and the activation and termination of SOICR230. In 2015, Wanguemert et al. demonstrated that the G357S mutation requires PKA-dependent phosphorylation to display enhanced SOICR activity in HEK293 cells, which is unusual compared to other N-terminal CPVT-associated mutations222. This dependence on sympathetic activation may contribute to the reduced penetrance (28%) of CPVT in the G357S mutant carriers who may experience varied levels of sympathetic activation. However, in addition to altering the response to PKA activation, it remains to be determined whether the G357S mutation could affect other properties of RyR2.  Considering the roles of Domain B in RyR2 function, Dr. Wayne Chen’s lab assessed the impact of the G357S mutation on the properties of SOICR and the protein expression of RyR2231. Relative to WT-RyR2 expressing HEK293 cells, the G357S mutation reduced the maximum fraction of cells that displayed SOICR. Of the cells that did display SOICR, the G357S mutation enhanced the susceptibility to spontaneous Ca2+ release by reducing the SOICR activation threshold and augmented the amplitude of spontaneous Ca2+ release by delaying SOICR termination. A potential explanation for the reduction in SOICR is finding that the G357S mutation dramatically reduced the expression of the full-length RyR2 protein in HEK239 cells. However, the G357S mutation did not alter the cytosolic Ca2+-dependent activation of [3H]ryanodine binding or the cytosolic calcium-regulated Ca2+ release in HEK293 cells.  Since the G357S mutation reduces RyR2 protein expression, we wondered whether this effect is intrinsic to the N-terminal region where the mutation is located. We expressed the N-115  terminal region of RyR2 (RyR2ABC) containing the G357S mutation as a fusion protein with a His-tag and MBP. Upon removal of the purification tags, the RyR2ABC protein was purified to homogeneity. Fig. 4.8C shows thermal stability curves of the G357S mutant compared to WT-RyR2ABC. There are two transitions in the unfolding curve for the G357S mutant, with the first transition exhibiting a mid-point unfolding temperature significantly lower than that of the WT-RyR2ABC (G357S-RyR2ABC: 39.8 ± 0.17oC vs. WT-RyR2ABC: 42.8 ± 0.17oC, P = 0.0002). These results indicate that the G357S mutation has an intrinsically destabilizing effect on RyR2ABC. In the context of the full-length RyR2, the G357S mutation is close to an interface with the Central domain that encompasses disease hotspot 3109,115 (Fig. 4.5E). Within the individual RyR2ABC domains, G357 is exposed to the surface and the introduction of a serine sidechain would not cause a steric clash with another sidechain. While this may contradict the dramatically reduced expression of the RyR2 protein, close inspection of the backbone dihedral angles explains the repercussions of this mutation. Because glycine residues lack a C-beta atom, they can adopt conformations that are not allowed for other amino acids. Fig. 4.5D shows a Ramachandran plot for the WT-RyR2ABC protein (PDB: 4L4H). The G357 residue is located in the middle of a disallowed region for non-glycine residues. Substitution by serine would thus cause an intrinsic destabilization of the protein, and we postulate that this underlies the effect on the stability of the G357S mutant protein.      116   Figure 4.8 The G357S mutation reduces the stability of the ABC domains of RyR2. A. Crystal structure of the mouse RyR2ABC (residues 1-547), showing three domains and a central chloride anion that links the three domains together. B. Details showing G357S. The backbone conformation is allowed for glycine, but not serine. C. Thermal stability curves of WT RyR2ABC vs. G357S RyR2ABC. The mid-point unfolding temperature, obtained by taking the maxima of the first derivatives, is 42.8 ± 0.17oC for WT RyR2ABC and 39.8 ± 0.17oC for G357S RyR2ABC (errors indicating SEM). Statistical significance was determined using an unpaired two-tailed t-test with a 95% confidence interval, P=0.0002. D. Ramachandran plot for the RyR2ABC crystal structure, with backbone phi angles on the x-axis and psi angles on the y-axis. The shaded areas indicate preferred regions for non-glycine residues. Glycine residues are indicated with squares, with G357 indicated by an arrow. E. Cryo-EM structure of RyR108,109 showing the location of the corresponding glycine. One subunit, at the “front” of the view has been omitted for clarity. The N-terminal region, the Central domains and the S6 inner helix bundle with the C-terminal domain (CTD) are indicated. The equivalent of G357S in RyR1 is near the interface between the N-terminal region and the Central domain.     117  4.4 Discussion 4.4.1 Chloride binding to RyR2 induced global conformation changes in the N-terminal domains The 4.2- and 4.4Å resolution structures of the open- and closed-states of RyR2 were solved by Peng et al. in 2016 and provided a glimpse into the conformational changes incurred upon gating for this cardiac channel isoform. When compared individually, the amino-terminal, handle and helical (HD1) domains revealed little intradomain rearrangements as the channel transitioned from a closed to open state115. Thus, the authors suggest that the overall motion of the cytoplasmic assembly may stem from domain-wise displacement and relative domain shifts. For example, the cytoplasmic vestibule formed by the N-terminal domains appear to rotate upward and counter-clockwise from the closed to open state. Major structural re-orientation occurs at the interface between the N-terminal domains, handle and HD1 domains, and the Central domain. Thus, conglomerating the analysis from the cryo-EM, RyR2ABC Y125H R420H and R420W structures, we can propose a mechanism of channel gating as it pertains to the N-terminal domains. Neighbouring amino-terminal domains need to move apart for the channel to open. As a result, this means that certain interactions need to be broken for the channel to transition from the closed to open state, requiring an input of energy (Fig. 4.9). Our structure of RyR2ABC Y125H R420H and R420W illustrate that chloride induces domain reorientation. One possibility is that chloride binding directly affects interactions between the neighbouring protomers, affecting the energetics of channel opening. This notion of ligand binding inducing domain-domain re-orientations is not a foreign concept: the N-terminal domains of the closely-related IP3-receptor undergoes alterations upon binding to IP3232. Thus, ligand-induced domain re-orientation can affect channel opening.  118   Figure 4.9 Gating model proposed for RyR channels and the closely related IP3R. Under normal conditions (upper panel), channel opening requires re-arrangement of the N-terminal domains, which forms a cytoplasmic vestibule. Domain re-arrangement is energetically costly as certain interactions need to be broken before re-arrangement can occur. Chloride has been shown to induce domain-domain conformational changes, thus, one possibility is that chloride binding directly affects interactions between the neighbouring protomers, affecting the energetics of channel opening (middle panel). This hypothesis is analogous to the IP3-receptor, where inositol triphosphate binding to the amino-terminal domains causes similar domain-domain re-orientations232 (lower panel).  4.4.2 The H29D mutation does not alter the intrinsic properties of RyR2 or the thermal stability of the N-terminal domains Unlike the majority of disease-linked RyR2 mutations in the N-terminal region, the H29 residue is not located at a domain-domain interface. Further examination of the recent cryo-EM structures of RyR37,107,108,115 demonstrated that the mutation would also not be in contact with neighboring RyR domains nor is it located at a known interface with RyR2 auxiliary proteins. Additionally, His29 is not involved in coordinating the central chloride ion, which is key to the 119  stability of RyR2ABC119, nor is it a buried residue that, when mutated, could interfere with protein folding. On the contrary, the H29 sidechain has been found to be highly flexible in multiple crystal structures, implying that it does not even interact with sidechains of neighbouring residues.  The finding that the H29D mutation does not alter the intrinsic function of the RyR2 channel does not explain how the mutation causes ventricular tachyarrhythmia at rest. One possibility may be that, while the H29D mutation does not affect the isolated RyR2 channel, it may alter the interactions between RyR2 and its regulatory proteins in cardiac cells, given its location on the surface of the cytoplasmic assembly of RyR2. Currently, there have been no reports of an auxiliary protein which interacts directly with the RyR N-terminal region, but the possibility remains that such a binding partner exists. The cytoplasmic region of this residue may interact with proteins associated with the T-tubular membrane. In this capacity, the RyR2-H29D mutation would alter EC coupling or SR Ca2+ release by affecting protein-protein interactions that regulate RyR2 function. Another possibility is that the H29D mutation is not disease-causing, but coincidentally this mutation was isolated in the presence of another dysfunctional RyR2-associated protein.    4.4.3 The CPVT-associated G357S RyR2 mutation impairs Ca2+ release termination and reduces the thermal stability of the N-terminal domains Although CPVT is thought to be one of the most lethal of the inherited channelopathies, not all CPVT-susceptible individuals experience the highly lethal phenotypes associated with the disease214,218,229,233–235. Indeed, in 2015 Wanguemert et al. described a large family of more than 1400 members, many of whom had CPVT and sudden cardiac death222. Contrary to what is 120  normally associated with the disease, many of the G357S mutant carriers lived beyond the age of 50, with some members even reaching their 70-80’s. A mechanistic explanation underlying this variable phenotypic expression is unclear. However, Wanguemert et al. demonstrated that the functional impact of the G357S mutation requires sympathetic activation and suggested that the dependence on β-adrenergic stimulation may underlie the incomplete disease penetrance within this family.  In our present study, we have found that the G357S mutation, located in Domain B of the N-terminal region of the channel, reduces the activation and termination thresholds for arrhythmogenic spontaneous Ca2+ release (SOICR) – a finding that is similar in other N-terminal CPVT RyR2 mutations that have been previously characterized220,236. The G357S mutation also caused a substantial decrease in the expression level of the RyR2 protein. Combined, the altered SOICR thresholds and markedly reduced protein expression of the G357S mutant likely contribute to the variable disease phenotypes observed in this extended family. The augmented SOICR activity caused by the G357S mutation may be an underlying explanation for the lethal CPVT phenotypes displayed in some of the mutant carriers in this large family.  RyR channels are homotetramers26,66. In G357S heterozygous patients, the RyR2 channel will be a mixture of both WT and G357S-mutants. Thus, the impact of the mutant would depend on the ratio of WT and G357S mutant, which depends on the expression level of the WT and mutant alleles. The G357S mutant was found to significantly decrease RyR2 protein expression. Thus, reduced expression of the mutant RyR2 protein may be an explanation for the variable phenotypic expression observed in the large family. Despite the G357S mutation affecting the properties of SOICR, there was no observed increase in the fraction of cells that displayed SOICR. This finding is likely due to the reduced level of expression of the mutant protein, which 121  in turn may mask any enhanced SOICR activity of the G357S mutant, especially when measuring SOICR in a cell population based-assay.   The G357S mutation lies within a loop which connects two β-strands in the β-trefoil fold of Domain B. Assessing the high-resolution crystal structure of the RyR2ABC region, the G357 backbone adopts a conformation that is only allowed for a glycine residue. A mutation to a serine residue would interfere with the folding of the G357S backbone, which could be reflected in the decreased thermal stability measured for G357S-RyR2ABC compared to WT-RyR2ABC. Interestingly, G357 appears to be located close to an interface between the N-terminal domain (disease hotspot 1) and the Central domains, encompassing disease hotspot 3109,115. This location is of significance given the recent suggestions that the Central domain acts as the transducer, which integrates conformational changes in the cytoplasmic assembly to the channel pore domain, controlling RyR2 gating109,114,115. Thus, the interface between the N-terminal region and the Central domains may be required for normal allosteric coupling between the channel’s N-terminal region and channel pore. Understandably, a mutation in this region, such as the G357S mutation, may interfere with the coupling, explaining its effects on SOICR.     122  Chapter 5: Concluding remarks and future directions Muscle contraction, the physiological process underlying our beating heart or body movements, displays an exquisite complexity reflected by ongoing efforts to uncover its detailed underlying mechanism. The muscle contraction field gained momentum when Alexander Sandow coined the term “excitation-contraction coupling”, aptly describing the link between sarcolemma depolarization and the generation of muscle tension9. Remarkably, even before the molecular components required for muscle contraction was discovered, it had been established that Ca2+ was an activator of the contractile machinery7,8. The rapid speed, on the order of 1-2 milliseconds, of the action potential to initiate contraction in the center of fibers of 50-100µm in diameter was an intriguing feature of this system8,10. Ultrastructure studies of the twitch muscle fibers of the frog provided insights into this mystery14.  The invaginations of the sarcolemma (T-tubule network) formed a tight association with the expanded terminal cisternae sacs of the sarcoplasmic reticulum. In skeletal muscle, this organization is called the triad, due to the T-tubule being flanked on either side by a SR terminal cisternum; in cardiac tissue, a dyad is formed8,14. Triads and dyads are the anatomical basis for skeletal and cardiac muscle EC coupling, respectively8.   Abundance of the dihydropyridine receptor (DHPR/L-type voltage-gated calcium channel) in skeletal muscle T-tubule membranes combined with the sensitivity of EC coupling to dihydropyridine-based compounds21,22 fueled hypotheses that the CaV was the molecular entity which initiated the SR Ca2+ release essential for contraction. Further examination of the architecture of the skeletal muscle triads showed regularly spaced electron-dense protrusions, termed “feet”, extending into the ~12nm gap from the SR membrane14. The molecular identity of these gateways which release SR Ca2+ later become known as the Ryanodine Receptor, after this 123  channel was cloned and purified23–26,68. Much work has been dedicated to understanding how these two Ca2+ channels, located within two different membrane systems, communicate with each other to release intracellular SR Ca2+. In cardiac tissue, Fabiato proposed the mechanism of calcium-induced calcium release (CICR)28, where depolarization-induced opening of CaVs cause an influx of extracellular Ca2+ which in turn activates the RyR. While CICR was initially proposed for skeletal muscle237, this mechanism has since been dismissed in favor of compounding evidence suggesting a mechanical coupling between the CaV and RyR22,27,51,53,74.   Failure of skeletal muscle EC coupling, and subsequent lethality, in CaV1.1 dysgenic mice50 established CaV1.1 as the voltage sensor for skeletal muscle EC coupling8. Indeed, studies utilizing CaV1.1 and CaV1.2 chimeras have identified a sequence specifically in the skeletal muscle channel’s long cytosolic loop linking transmembrane domains II and III (the II-III loop) that is involved in EC coupling. These experiments have consistently shown that residues 720-765 of CaV1.1 are required for EC coupling to occur54,178,238–241. In addition to transmitting an orthograde EC coupling signal to RyR1, the II-III loop also receives a retrograde, current-enhancing signal from the RyR1 to CaV1.154–56. Thus, the II-III loop plays an essential role in this bidirectional coupling mechanism between CaV1.1 and the RyR.  Given the intricacy and stringency under which EC coupling operates, it is not surprising that triad junctions contain large complexes of proteins147. This complexity, combined with the fact that SR Ca2+ release depends on the interaction of two separate membrane systems, has marred the use of reductionist approaches to define the molecular mechanism of EC coupling Ca2+ release147. By far, the CaV-α1s- and β1a-subunits and RyR1 are the most well-studied proteins of the EC coupling complex and have been established to be the essential components of the complex. Currently, there is a shift to identify additional core proteins in EC coupling 124  transduction, particularly to determine how the complex assembles at triadic junctions and how EC coupling is regulated.    5.1 STAC3 is an essential component of the skeletal muscle EC coupling machinery Recent movements to identify novel factors required for skeletal muscle development and function led to the discovery of STAC3, which was determined to be required for myogenic differentiation157,158. Genetic analysis of zebrafish larvae exhibiting defective motor behaviours and premature death revealed an underlying mutation in the gene encoding the putative muscle-specific adaptor protein, STAC3148. Given the myopathic features of the zebrafish mutation, the possibility was explored for stac3 mutations being involved in propagating congenital human myopathies. A tryptophan-to-serine mutation (W284S)148 was determined to be the genetic basis of a debilitating myopathy first described by Bailey and Bloch in a three-month old American Indian infant of Lumbee descent153. Since the initial diagnosis in 1987, many more individuals of Lumbee descent have been identified with similar congenital anomalies, including cleft palate, skeletal abnormalities and susceptibility to malignant hyperthermia, giving rise to the name of this myopathy: Native American Myopathy (NAM)154,155. Further work by Nelson et al. revealed that homozygous deletion of STAC3 in mice resulted in complete paralysis and perinatal lethality156, cementing the role of this adaptor protein in skeletal muscle EC coupling.  In skeletal muscle, mechanical CaV1.1-RyR1 coupling is thought to underlie both EC coupling SR Ca2+ release and retrograde coupling, whereby RyR1 increases the amplitude of CaV1.1 Ca2+ current22,27,48,51,53,164. Despite numerous probes into this interaction, it remained elusive whether CaV1.1 interacts with RyR1 directly, through auxiliary proteins, or a combination of both112. A recent study by Polster et al. (2016) provided insight into the role of 125  STAC3 in the crucial CaV1.1-RyR1 interaction. CaV1.1 expression in STAC3 KO myotubes was unable to mediate EC coupling and produced only small, rapidly inactivating Ca2+ currents48. Introduction of WT STAC3 into the KO myotubes restored both coupling and Ca2+ conduction. In contrast, expression of the STAC3 NAM mutation resulted in partial recovery of normal Ca2+ current, but only a very weak restoration of EC coupling. Collectively, these results revealed that STAC3 is involved in both functions of CaV1.1, ion conduction and EC coupling, and that these two effects of STAC3 rely on different regions of the protein, which function partially independent of each other48.  In our present study, we provided structural insights into the STAC proteins by solving high-resolution crystal structures of the two C-terminal SH3 domains of STAC1 and STAC2 and the second SH3 domain of STAC3199. These domains form a compact arrangement by virtue of a short linker between the two domains. ITC measurements allowed detection of a micromolar affinity binding site for the II-III loops of CaV1.1 and CaV1.2. The binding site is located within a short proline-rich peptide encoded by residues 747-760 in human CaV1.1. An equivalent region forms a binding site in CaV1.2. Field stimulation experiments in dysgenic myotubes demonstrated that a mutant CaV1.1, unable to bind the tandem SH3 domains via its II-III loop, severely perturbs skeletal muscle EC coupling. Our identified II-III loop binding site corresponds well to previous studies reporting the importance of the sequence encompassed by residues 720-756 for EC coupling to occur54,178,238–241, and was recently confirmed by Polster and colleagues (2018)182. Our structure of the complex between the tandem SH3 domains of STAC2 and a minimal CaV1.1 II-III loop peptide revealed that the peptide forms key interactions with Trp329. Interaction studies using either STAC2 W329S or STAC3 W284S (the mutation implicated in NAM), significantly reduced the affinity to the CaV1.1 peptide. This outcome would explain the 126  previous functional evidence where it was shown that the W284S mutation results in diminished EC coupling48 and reduced recruitment of STAC3 into the skeletal muscle CaV1.1 complex167. Collectively, our data reveals that SH3-1 is the primary determinant for the CaV1.1 II-III loop interaction. Our structure of the STAC2:CaV1.1 II-III loop peptide reveals that the canonical binding surface of the SH3-2 domain remains available for binding another peptide. Thus, it is entirely possible for the SH3-2 domain to bind another region of the CaV channel or one of the EC coupling proteins, thereby bridging multiple components of the EC coupling complex. Although previous experiments have shown coimmunoprecipitation of STAC3 and RyR1 in muscle tissue148, a direct interaction between the two proteins remains to be determined.     5.2 STAC proteins: A Swiss-Army knife for the EC coupling complex? In addition to its key role in mediating skeletal muscle EC coupling, several additional roles for the STAC3 protein has emerged. Polster et al. (2015) demonstrated that co-expression of STAC3 in a heterologous expression system (tsa201 cells) promoted the robust and functional plasma membrane expression of CaV1.1162, a construct which has notoriously been difficult to express in mammalian cells that are not of muscle origin159. Remarkably, even in the absence of CaV-β and α2δ, STAC3 was able to support insertion of CaV1.1 into the plasma membrane162. Additionally, the members of the STAC family functionally interact with CaV1.2, the channel isoform predominantly expressed in cardiac tissue and neurons, as co-expression significantly decreased the channel’s speed of inactivation162,200. As STAC proteins are composed of three functional domains – a N-terminal C1 domain and the two C-terminal SH3 domains, interspersed by a large flexible linker – there is much to discover of the complete functional roles of STAC proteins, especially in the diverse tissue-types where the proteins are expressed.  127  Utilizing various truncation constructs, our TEVC electrophysiology studies have suggested that the C1 domain or the intrinsically disordered region flanking this domain is involved in mediating STAC3’s effect on CaV1.2 inactivation199. Indeed, an outright deletion of both SH3 domains still resulted in slower inactivation. Recently, it has been demonstrated that the C1 domain of STAC3 is crucial for the stability of the complex between STAC3 and L-type voltage-gated calcium channels167. Furthermore, Campiglio et al. have suggested that a potential direct interaction exists between the STAC C1 domain and the IQ domain of CaV1.2200. If such an interaction exists, it would explain STAC’s effect on channel inactivation as Ca2+-CaM’s association with the IQ domain underlies CDI39,40. However, calorimetric assays failed to detect a direct interaction between these two proteins. Intriguingly, we did detect evidence of direct binding (~ micromolar affinity) between the C1 domain and CaM. CaM encompasses complex binding states by virtue of its two lobes, which can bind to a protein target collectively as a unit or as separate entities, combined with CaM’s ability to bind its targets in its apo- or -Ca2+-bound state. STAC C1-CaM binding was determined to be independent of Ca2+, as binding was detected using both apo- and Ca2+-CaM. CaM C-lobe appears to be the primary contact for the interaction with the C1 domain, however, C-lobe binding was weaker compared to full-length CaM. Very weak binding was detected between N-lobe and the C1 domain, confirming that the N-lobe does contribute to the interaction. These findings lay the foundation for future work to be done in this area. From the onset, CaM binding to a protein fold which is enriched with β-strands, such as the C1 domain, is highly unusual. The hydrophobic pockets of N- and C- lobe typically bind to amphipathic α-helical domains within their target domains49,152,207. Thus, structural insights, either via NMR or crystallography, would provide a wealth of information to understand this unusual interaction. Functional experiments are required to grasp the 128  physiological ramifications of this interaction and to enrich our current understanding of channel inactivation regulation.  In non-muscle cells, STAC3 was shown to facilitate functional membrane expression of CaV1.1162 and its C1 domain has been deemed responsible for the stable incorporation into the CaV1.1 complex167. The role of the C1 domain located in protein kinase C (PKC) is well studied. sn-1,2-Diacylglycerol (DAG) is one of the central intracellular second messengers that bind to typical C1 domains (reported Kd‘s in the nano- to micromolar range)202. Phorbol esters, tetracyclic diterpenoids, extracted from plants mimic the actions of DAG242,243 and have binding affinities to C1 domains at least three-orders of magnitude higher than DAGs244,245. Given this property, phorbol esters are the most widely used molecules to study PKCs and related proteins. Through their interaction with PKC’s C1 domain, both DAG and phorbol ester dramatically increase the affinity of PKC for membranes, serving as a “molecular glue” to recruit PKC to membranes246. Extrapolating the role of PKC’s C1 domain in membrane recruitment, perhaps the C1 domain of STAC proteins function in a similar manner. Potentially, the C1 domain could serve as a membrane anchor and the STAC SH3 domains could bind to and thereby traffic CaV1.1 to the plasma membrane. Given the partial conservation of residues involved in phorbol ester binding in the PKC C1 domain202 and STAC proteins, binding assays exploring the interaction between the STAC C1 domains and DAG, phorbol esters, and lipids enriched in the T-tubular membrane needs to be completed.  The STAC3 protein has an additional CaM binding site located in the intrinsically disordered region (residues 175-194) between the C1 and SH3 domains. Interaction with this STAC3 peptide is Ca2+-dependent as a ~20µM affinity was measured using Ca2+-CaM, but no evidence of binding was detected with apo-CaM. The combined binding of both CaM lobes 129  appears to be required, since the Ca2+-lobes have substantially weaker binding affinities when tested individually. A triple mutation targeting three consecutive hydrophobic residues (V185E I186E M187E) dramatically reduces the measured affinity to Ca2+-CaM. Currently, identification of these STAC-CaM interactions raises more possibilities and questions than they answer. Do multiple CaM molecules bind to the STAC protein? Or does the C1 domain and this intrinsically disordered region cooperate to bind a single CaM with high affinity? The discovery of independent CaM binding sites certainly raises these possibilities. The next task will be to assess the physiological role of both CaM binding sites and tease apart the contributions of both sites. Perhaps calmodulin acts as an intermediate for STAC proteins to carry out their effect on channel inactivation. These findings certainly suggest multiple roles of STAC proteins beyond their contributions to skeletal muscle EC coupling, and reflect the exciting new discoveries which lie ahead to fully understand the intricacies of each component of the EC coupling complex and how they synchronously work together to coordinate intracellular Ca2+ levels.    5.3 Ligand-dependent modulation of RyRs The crux of EC coupling involves the release of Ca2+ from intracellular SR stores which then activates the proteins involved in muscle contraction17,18. The molecular identity of the SR Ca2+-release gateway has been determined to be the Ryanodine Receptor26,66,68. The channel’s membrane protein nature and immense size and have been an impediment to study the numerous interactions between its domains and modulators. Indeed, the RyRs are regulated by a plethora of small molecules (e.g. Ca2+, caffeine, volatile anesthetics) and protein binding partners (e.g. CaM, FKBPs)73. Improvements in cryo-EM technology, and the dawn of the “resolution revolution”, has catapulted our advancements in knowledge of the structure of the full-length channel. 130  Indeed, in 2016 des George et al. were able to probe into the structural basis of channel gating and ligand-dependent activation. Channel activators Ca2+, ATP and caffeine bind to the RyR’s C-terminal domain, whereas ryanodine was located in the channel’s TM conduction pathway67. Despite having different binding sites, a common theme of ligand binding and gating is the global conformational changes incurred in the cytosolic assembly which accompanies local changes in the transmembrane domain67. While the recent cryo-EM RyR structures report a nominal resolution of ~6.1-3.8Å107–109, an important caveat to note is that the resolution of the cytosolic cap tapers off dramatically due to the inherently dynamic nature of this region113. As such, structural details of ligand binding sites and disease-associated mutations in the cytoplasmic assembly remains limited. Previous successes in obtaining and docking high-resolution crystal structures of globular domains located within the N-terminal channel region110–113,117–119,216 reflect an attractive means to fill this void.    The structure of the N-terminal ABC domains of the cardiac ryanodine receptor was published in 2013 by Kimlicka et al.119. Interestingly, the N-terminal region of RyR2 contains an anion-binding site located at the interdomain interface, where a chloride ion appears to shield the repulsive charges of several arginine residues contributed by each of the three domains119. This anion binding site replaces the extensive and stabilizing hydrogen bond and salt bridge network observed in RyR1. Notably, RyR2ABC anion binding is crucial to the protein’s stability, as removal of chloride from the purification buffers resulted in protein precipitation. Furthermore, mutations which target Arg420, a key anion-coordinating residue, are associated with increased susceptibility to the lethal cardiac arrhythmia, CPVT. Here, we analyzed the structural ramifications of removing the chloride anion from the N-terminal RyR2 domains. A RyR1-RyR2ABC hybrid introduced two RyR1 histidine residues, involved in mediating salt bridges, 131  into the RyR2ABC structure and abolished anion binding. Surprisingly, while the RyR1-RyR2ABC hybrid (hereon referred to as RyR2 Y125H R420H) removed the anion from the interdomain interface, it did not restore the salt bridges observed in the RyR1ABC structure216. Removal of the chloride anion resulted in local residue re-arrangements at the interdomain interface as well as global domain-domain conformational changes. As predicted, the RyR2 R420W disease mutation abolished a key arginine residue involved in anion coordination. Introduction of this mutation also caused sidechain rearrangement in the interdomain interface, particularly, the sidechain of Arg417 was able to mediate two salt bridges with Asp61 and Glu40. These salt bridges likely compensate for the stability lost as a result of anion removal, as the extent of the domain-domain conformational changes was smaller than that measured for RyR2 Y125H R420H. Additionally, the formation of this salt bridge network accounts for the similar thermal stability compared to WT-RyR2ABC. Compared to the mouse RyR2ABC structure216 described before by Kimlicka et al., the human RyR2ABC structure, solved by Borko et al. (2014)224, did not contain an anion binding site. Although the overall domain arrangement is conserved between the two structures, superposition of human RyR2ABC (-Cl-) onto mouse RyR2ABC (+Cl-) revealed large-scale conformational changes on the order of ~6Å. Although a potential explanation for these conformation changes could be due to the presence of detergent (1mM CHAPs, used in the Borko et al. study), we followed up with these findings by taking a more systematic approach via spectroscopic measurements. TR-FRET measurements with a labeled RyR2ABC construct revealed two major distance components: R1 and R2. R1 had a distance measurement centered around 40Å, which matches the predicted probe distance measured from the crystal structure. In the absence of KCl, R2 is centered at 55Å and shifts to 60Å in response to KCl. These results 132  suggest the coexistence of two structural states in RyR2ABC: a relatively ordered R1 and a more disordered R2 with differential responses to chloride.  RyR2 modulation by anions is not a foreign concept. Fruen et al. reported that anions are able to potentiate SR Ca2+ release via RyRs247,248. For RyR2, ryanodine affinity was higher in 0.25M KCl vs. 0.25M KMes, suggesting that chloride can increase the activity of RyR2249. However, another study reported that chloride had a minimal effect on cardiac muscle (RyR2) Ca2+ release248. These contradicting results may be a product of differences in channel extraction methods. The collective results from Kimlicka et al. and the results presented here, suggest that chloride’s ability to modulate RyR2 may be due to binding the N-terminal region. Interestingly, chloride was able to stimulate SR Ca2+ release in RyR1248, yet the skeletal muscle channel isoform does not have an N-terminal chloride binding site. Chloride dependence of activity may reflect a complex event involving multiple sites in both RyR1 and RyR2. Affinity measurements of chloride binding would reveal whether chloride binding in the physiological context is a dynamically regulated event or if it were a “structural” binding event (i.e. always bound)119. A high affinity site would be indicative of a constitutively bound chloride ion; a ~mM affinity would imply a dynamic event, with an equilibrium between the unbound and bound states. The dependence on chloride for protein stability has made affinity measurements inherently difficult, however, a preliminary measurement revealed that the affinity of chloride toward RyR2ABC appears to be at the physiological level of chloride (0-4mM)250. Given that the intracellular concentration of chloride in skeletal muscle and neurons is ~3-4mM251,252, we can expect an equilibrium between the unbound and bound states. Cell types which have a substantially higher intracellular chloride concentration, such as cardiac myocytes (10-25mM)253,254 and epithelial cells (~40mM)255, would shift the equilibrium further toward the bound state. It is possible that 133  RyR2 activity could be allosterically coupled to chloride binding to the channel’s N-terminus. Binding could function as a regulatory mechanism in a cell-type dependent manner, responding to the chloride flux in a dynamic manner or in cells with high chloride levels, chloride could be permanently bound to provide structural stability250. Other physiologically relevant anions, like phosphate, have been reported to modulate RyR activity247. Thus, future work will involve examining the functional effect of chloride binding, as well as exploring other anions capable of similarly modulating the channel.  Conglomerating the analysis from our current structural investigation, we can propose a mechanism of channel gating as it pertains to the N-terminal domains. Channel opening requires allosteric movement of the cytoplasmic domains, particularly involving the neighbouring amino-terminal domains which need to move apart for the channel to open. As a consequence, this requires certain interactions to be broken for the channel to transition from the closed to open state, requiring an input of energy. The conformational changes induced by chloride (or more broadly, by anions) binding directly affects interactions between the neighbouring protomers, affecting the energetics of channel opening. This mechanism of ligand binding inducing domain-domain re-orientations is analogous to the closely-related IP3-receptor, which undergoes alterations upon binding to IP3232. Thus, ligand-induced domain re-orientation can affect channel opening.  5.4 Understanding the mechanisms of RyR disease-associated mutations RyRs are allosteric membrane proteins, requiring coordinated movement in the massive cytoplasmic assembly and the pore domains to gate the channel95. Pseudo-atomic models of the N-terminal region in the open and closed states of RyR1 revealed that channel opening is 134  accompanied by a ~7Å widening of an inter-subunit contact area216. Kimlicka et al. proposed a mechanism whereby the N-terminal domains act as a “brake” against channel opening216. Direct contacts between domains A and B of neighbouring subunits in the closed state are disrupted upon channel opening. In this capacity, the interdomain interactions act as an energetic barrier that needs to be overcome for the channel to open. Binding channel activators, like Ca2+, delivers the energy by causing low-range conformational changes that couple allosterically to the N-terminal region. For this mechanism to be feasible, the interdomain contacts must be extremely weak, since a high-affinity interaction would be a severe energetic penalty against disruption and channel opening. Indeed, several approaches have failed to detect any interactions between individual RyR1ABC monomers in solution, implying that they do not drive tetramerization216.  RyR1 and RyR2 disease mutations interfere with channel gating. Although there are some exceptions, the majority of RyR disease mutations have a gain-of-function phenotype, facilitating channel activation by increasing sensitivity toward cytoplasmic or luminal activation mechanisms and decreasing the propensity for binding regulators which favor the closed channel state73. Within the N-terminal disease hotspot, most mutations cluster around the domain-domain interfaces, lowering the energetic barrier and facilitating channel opening. The R420W mutation is an example of a mutation at the interface between domains A, B, and C. By perturbing chloride binding, the imbued global conformational changes likely potentiates susceptibility to CPVT. The glycine involved in the G357S disease mutation appears close to an interface between the N-terminal region (disease hotspot1) and the Central domain (disease hotpot2)108,109,115. Since the Central domain has been proposed to transduce conformational changes in the cytoplasmic assembly to the channel’s pore, a mutation at this interface could alter the allosteric coupling between these two channel regions, explaining this mutation’s effect 135  on SOICR. Additionally, we noted that the G357S mutation decreases the thermal stability of RyR2ABC. Altered protein-fold temperature sensitivity may imply that there are temperature-dependent conformational changes, which could then also affect one or more individual domain interactions216.   As we advance toward complete models of RyRs, by combining the recent near-atomic-resolution cryo-EM structures of RyR1107–109 and RyR2256 with high-resolution crystal structures of the channel’s globular domains we can determine domain-domain interfaces, map the location of disease-causing RyR mutations, grasp the structural impact and long-range allosteric consequences of these disease mutations.   5.5 Muscling in on EC coupling? A remarkable feature of EC coupling involves the communication of two calcium channels across two different membrane systems. In cardiac tissue, communication between the two channels occurs through Ca2+, where extracellular Ca2+ activates the SR-bound RyR2, in a process termed calcium-induced calcium release28. Understanding the mechanism of signal transduction in skeletal muscle has been more arduous: for many years it was contested whether communication occurred via chemical, mechanical or electrical means8. Experiments on muscle cells which are genetically null for triadic proteins, established that CaV1.1 is the voltage sensor for RyR1 and that SR Ca2+ release occurs via mechanical coupling51,74. Over the last ~30 years, gene knockouts have revealed that CaV1.1/RyR1 coupling requires additional proteins, but leave open the possibility that currently unidentified and untested proteins may also be necessary147.   Approaches to reconstitute EC coupling in heterologous systems have been marred by the dual (plasma and ER) membrane dependence of the process163. However, recently Perni et al. 136  (2017) were able to reconstitute skeletal muscle EC coupling in tsa201 cells co-expressing CaV1.1, CaV-β1a, STAC3, RyR1 and junctophilin2 (JP2)147. The ER-plasma membrane junctions which formed were morphologically similar to SR-plasma membrane junctions in native muscle tissues. Depolarization of this de novo reconstitution elicited Ca2+ transients independent of extracellular Ca2+ entry, comparable to skeletal muscle. Remarkably, this assembly allowed CaV1.1 to be arranged in tetrads indicative of the physical links to RyR1. These results succinctly identify the core components required for skeletal muscle EC coupling.  While we may now have a scaffold of the EC coupling complex, future work in the EC coupling field will need to assess how all these components interact with each other, identifying how the “jig-saw puzzle” pieces align to form the bigger picture. Whether CaV1.1 interacts directly with RyR1 or via auxiliary proteins is an area of intense focus. The de novo reconstitution achieved by Perni et al. certainly has identified the essential EC coupling components, however, within the physiological setting of the triad junction, several other proteins such as CaM, FKBP12, triadin, junctin and calsequestrin148,149 are also involved in the regulation of the process. In this context, experiments to understand the regulation of EC coupling is required. As we have seen from recent studies, we cannot exclude the possibility of identifying currently unknown entities involved in the process. Excitingly, uncovering the intricacies of the EC coupling complex has become a tangible project given the advancements in cryo-EM technology combined with crystallographic, biochemical, and biophysical insights.  137  Bibliography 1. Berridge, M. J., Lipp, P. & Bootman, M. D. The versatility and universality of calcium signalling. Nat. Rev. Mol. Cell Biol. 1, 11–21 (2000). 2. Berridge, M. J., Bootman, M. D. & Roderick, H. L. Calcium signalling: dynamics, homeostasis and remodelling. Nat. Rev. Mol. Cell Biol. 4, 517–29 (2003). 3. Clapham, D. E. Calcium Signaling. Cell 131, 1047–1058 (2007). 4. Carafoli, E. Calcium – a universal carrier of biological signals. FEBS J. 272, 1073–1089 (2005). 5. Jaiswal, J. K. Calcium - How and why? J. Biosci. 26, 357–363 (2001). 6. Ringer, S. A further contribution regarding the influence of the different constituents of the blood on the contraction of the heart. J. Physiol. 4, 29–42 (1883). 7. Calderón, J. C., Bolaños, P. & Caputo, C. The excitation – contraction coupling mechanism in skeletal muscle. 133–160 (2014). doi:10.1007/s12551-013-0135-x 8. Dulhunty, A. F. Excitation-contraction coupling from the 1950s into the new millennium. Clin. Exp. Pharmacol. Physiol. 33, 763–772 (2006). 9. Sandow, A. Excitation-contraction coupling in muscular response. Yale joural Biol. Med. 25, 176–201 (1952). 10. Huxley, A. F. & Taylor, R. E. Local ativation of striated muscle fibres. J. Physiol. 144, 426–441 (1958). 11. Hopkins, P. M. Skeletal muscle physiology. Contin. Educ. Anaesthesia, Crit. Care Pain 6, 1–6 (2006). 12. Porter, K.R., and Palade, G. E. Studies on the endoplasmic reticulum III. Its form and distribution in striated muscle cells. J. Biophys. Biochem. Cytol. 3, 269–300 (1957). 138  13. Eisenberg, R. S. & Gage, P. W. Ionic conductances of the surface and transverse tubular membranes of frog sartorius fibers. J. Gen. Physiol. 53, 279–97 (1969). 14. Franzini-Armstrong, C. Structure of the junction in frog slow fibers. J. Cell Biol. 47, 488–499 (1970). 15. Adrian, R.H., Costantin, L.L., and Peachey, L. D. Radial spread of contraction in frog muscle fibres. J. Physiol. 204, 231–257 (1969). 16. Langer, G. A. & Peskoff, A. Calcium concentration and movement in the diadic cleft space of the cardiac ventricular cell. Biophys. J. 70, 1169–1182 (1996). 17. Ashley, C. C. & Ridgway, E. B. Simultaneous recording of membrane potential, calcium transient and tension in single muscle fibres. Nature 219, 1168–1169 (1968). 18. Schneider, M. F. & Chandler, W. K. Voltage dependent charge movement in skeletal muscle: A possible step in excitation-contraction coupling. Nature 242, 244–246 (1973). 19. Armstrong, C. M. & Bezanilla, F. Currents related to movement of the gating particles of the sodium channels. Nature 242, 459–461 (1973). 20. Armstrong, C. M. & Bezanilla, F. Charge movement associated with the opening and closing of the activation gates of the Na channels. J. Gen. Physiol. 63, 533–52 (1974). 21. Lamb, G. D. Asymmetric charge movement in polarized and depolarized muscle fibres of the rabbit. J. Physiol. 387, 349–367 (1987). 22. Rios, E. & Brum, G. Involvement of dihydropyridine receptors in excitation-contraction coupling in skeletal muscle. Nature 325, 717–20 (1987). 23. Smith, J. S., Coronado, R. & Meissner, G. Sarcoplasmic reticulum contains adenine nucleotide-activated calcium channels. Nature 316, 446–449 (1985). 24. Meissner, G. Ryanodine activation and inhibition of the Ca2+ release channel of 139  sarcoplasmic reticulum. J. Biol. Chem. 261, 6300–6306 (1986). 25. Fleischer, S., Ogunbunmi, E. M., Dixon, M. C. & Fleer, E. A. Localization of Ca2+ release channels with ryanodine in junctional terminal cisternae of sarcoplasmic reticulum of fast skeletal muscle. Proc. Natl. Acad. Sci. U. S. A. 82, 7256–7259 (1985). 26. Inui, M., Saito, A. & Fleischer, S. Purification of the ryanodine receptor and identity with feet structures of junctional terminal cisternae of sarcoplasmic reticulum from fast skeletal muscle. J. Biol. Chem. 262, 1740–1747 (1987). 27. Block, B.A., Imagawa, T., Campbell, K.P., and Franzini-Armstrong, C. Structural evidence for direct interaction between the molecular components of the transverse tubule/sarcoplasmic reticulum junction in skeletal muscle. J. Cell Biol. 107, 2587–2600 (1988). 28. Fabiato, A. Calcium-induced release of calcium from the cardiac sarcoplasmic reticulum. Am. J. Physiol. 245, C1-14 (1983). 29. Bers, D. M. Cardiac Excitation-Contraction coupling. Nature 415, 198–205 (2002). 30. Huxley, A. F. & Niedergerke, R. Structural changes in muscle during contraction: Interference microscopy of living muscle fibres. Nature 173, 971–973 (1954). 31. Huxley, H.E., H. J. Changes in the cross-striations of muscle during contraction and stretch and their structural interpretation. Group 173, 973–976 (1954). 32. Takahashi, M., Seagar, M. J., Jones, J. F., Reber, B. F. & Catterall, W. A. Subunit structure of dihydropyridine-sensitive calcium channels from skeletal muscle. Proc. Natl. Acad. Sci. U. S. A. 84, 5478–82 (1987). 33. Tanabe, T., Takeshima, H., Mikami, A., Flockerzi, V., Takahashi, H., Kangawa, K., Kojima, M., Matsuo, H., Hirose, T., and Numa, S. Primary structure of the receptor for 140  calcium channel blockers from skeletal muscle. Nature 328, 313–8 (1987). 34. Wu, J. et al. Structure of the voltage-gated calcium channel Cav1.1 at 3.6 Å resolution. Nature 537, 191–196 (2016). 35. Wang, M. C., Dolphin, A. & Kitmitto, A. L-type voltage-gated calcium channels: Understanding function through structure. FEBS Lett. 564, 245–250 (2004). 36. Ahlijanian, M. K., Westenbroek, R. E. & Catterall, W. A. Subunit structure and localization of dihydropyridine-sensitive calcium channels in mammalian brain, spinal cord, and retina. Neuron 4, 819–832 (1990). 37. Wu, J., Yan, Z., Li, Z., Yan, C., Lu, S., Dong, M., and Yan, N. Structure of the voltage-gated calcium channel Cav1.1 complex. Science (80-. ). 350, aad2395 (2015). 38. Buraei, Z. & Yang, J. The β Subunit of Voltage-Gated Ca2+ Channels. Physiol Rev 90, 1461–1506 (2010). 39. Peterson, B. Z., DeMaria, C. D. & Yue, D. T. Calmodulin Is the Ca2+ Sensor for Ca2+-Dependent Inactivation of L-Type Calcium Channels. Neuron 22, 549–558 (1999). 40. Zühlke, R. D., Pittt, G. S., Deisseroth, K., Tsien, R. W. & Reuter, H. Calmodulin supports both inactivation and facilitation of L-type calcium channels. Nature 399, 159–162 (1999). 41. Goldin, A. L. Mechanisms of sodium channel inactivation. Curr. Opin. Neurobiol. 13, 284–290 (2003). 42. Ellis, S. B., Williams, M.E., Ways, N.R., Brenner, R., Sharp, A.H., Leung, A.T., Campbell, K.P., McKenna, E., Koch, W.J., Hui, A., Schwartz, A., and Harpold, M.M. Sequence and expression of mRNAs encoding the alpha-1 and alpha-2 subunits of a DHP-sensitive calcium channel. Science, 241, 1661–1664 (1988). 141  43. Ruth, P., Rohrkasten, A., Beel, M., Bosse, E., Regulla, S., Meyer, H.E., Flockerzi, V., and Hofmann, F. Primary Structure of the Beta Subunit of the DHP-Sensitive Calcium Channel from Skeletal Muscle. Science, 245, 1115–1118 (1989). 44. Jay, S. D., Ellis, S.B., McCue, A.F., Williams, M.E., Vedvick, T.S., Harpold, M.M., and Campbell, K.P. Primary structure of the gamma subunit of the DHP-sensitive calcium channel from skeletal muscle. Science, 248, 490–492 (1990). 45. De Jongh, K. S., Warner, C. & Catterall, W. A. Subunits of purified calcium channels: alpha-2 and delta are encoded by the same gene. J. Biol. Chem. 265, 14738–14741 (1990). 46. Kang, M. G. & Campbell, K. P. γ Subunit of Voltage-Activated Calcium Channels. J. Biol. Chem. 278, 21315–21318 (2003). 47. Davies, A. et al. Functional biology of the α2δ subunits of voltage-gated calcium channels. Trends Pharmacol. Sci. 28, 220–228 (2007). 48. Polster, A., Nelson, B. R., Olson, E. N. & Beam, K. G. Stac3 has a direct role in skeletal muscle-type excitation – contraction coupling that is disrupted by a myopathy-causing mutation. Proc. Natl. Acad. Sci. 113, 10986–10991 (2016). 49. Van Petegem, F., Clark, K.A., Chatelain, F.C., Minor, D. L. J. Structure of a complex between a voltage-gated calcium channel beta-subunit and an alpha-subunit domain. Nature 429, 671–675 (2004). 50. Beam, K. G., Knudson, C.M., Powell, J. A. A lethal mutation in mice eliminates the slow calcium current in skeletal muscle cells. Nature 324, 148–152 (1986). 51. Tanabe, T., Beam, K. G., Powell, J. A. & Numa, S. Restoration of excitation-contraction coupling and slow calcium current in dysgenic muscle by dihydropyridine receptor complementary DNA. Nature 336, 134–9 (1988). 142  52. Tanabe, T., Beam, K. G., Adams, B. a, Niidome, T. & Numa, S. Regions of the skeletal muscle dihydropyridine receptor critical for excitation-contraction coupling. Nature 346, 567–569 (1990). 53. Adams, B.A., Tanabe, T., Mikami, A., Numa, S., and Beam, K. G. Intramembrane charge movement restored in dysgenic skeletal muscle by injection of dihydropyridine receptor cDNA. Nature 346, 569–572 (1990). 54. Nakai, J., Dirksen, R.T., Nguyen, H.T., Pessah, I.N., Beam, K.G., and Allen, P. . Enhanced dihydropyridine receptor channel activity in the presence of ryanodine receptor. Nature 380, 72–75 (1996). 55. Grabner, M., Dirksen, R. T., Suda, N. & Beam, K. G. The II-III loop of the skeletal muscle dihydropyridine receptor is responsible for the Bi-directional coupling with the ryanodine receptor. J. Biol. Chem. 274, 21913–21919 (1999). 56. Kugler, G., Weiss, R. G., Flucher, B. E. & Grabner, M. Structural Requirements of the Dihydropyridine Receptor a1S II-III Loop for Skeletal-type Excitation-Contraction Coupling. J. Biol. Chem. 279, 4721–4728 (2004). 57. An, M.T., Zamponi, G. W. in Voltage-gated Calcium Channels 194–204 (Springer, Boston, MA, 2005). doi:https://doi.org/10.1007/0-387-27526-6_12 58. Bean, B. P. Sodium Channel Inactivation in the Crayfish Giant Axon. Biophys. J. 35, 595–614 (1981). 59. Aldrich, R.W. and Stevens, C. F. Inactivation of open and closed sodium channels determined separately. Cold Spring Harb. Symp. Quant. Biol. 48, 147–153 (1983). 60. Cens, T., Rousset, M., Leyris, J. P., Fesquet, P. & Charnet, P. Voltage- and calcium-dependent inactivation in high voltage-gated Ca 2+ channels. Prog. Biophys. Mol. Biol. 143  90, 104–117 (2006). 61. Simms, B.A., and Zamponi, W. Neuronal voltage-gated calcium channels: Structure, function, and dysfunction. Neuron 82, 24–45 (2014). 62. De Leon, M., Wang, Y., Jones, L., Perez-Reyes, E., Wei, X., Soong, T.W., Snutch, T.P., and Yue, D. T. Essential Ca2+-binding motif for Ca 2+-sensitive inactivation of L-type Ca2+ channels. Science (80-. ). 270, 1502–1506 (1995). 63. Chin, D. & Means, A. R. Calmodulin: A prototypical calcium sensor. Trends Cell Biol. 10, 322–328 (2000). 64. Zhang, J. F., Ellinor, P. T., Aldrich, R. W. & Tsien, R. W. Molecular determinants of voltage-dependent inactivation in calcium channels. Nature 372, 97–100 (1994). 65. Ivanina, T., Blumenstein, Y., Shistik, E., Barzilai, R. & Dascal, N. Modulation of L-type Ca2+channels by Gβγ and calmodulin via interactions with N and C termini of α1c. J. Biol. Chem. 275, 39846–39854 (2000). 66. Lai, A.F., Erickson, H.P., Rousseau, E., Liu, Q.Y., and Meissner, G. Purification and reconstitution of the calcium release channel from skeletal muscle. Nature 331, 315–319 (1988). 67. des Georges, A. et al. Structural Basis for Gating and Activation of RyR1. Cell 167, 145–157.e17 (2016). 68. Takeshima, H. et al. Primary structure and expression from complementary DNA of skeletal muscle ryanodine receptor. Nature 339, 439–45 (1989). 69. Zorzato, F. et al. Molecular cloning of cDNA encoding human and rabbit forms of the Ca2+ release channel (ryanodine receptor) of skeletal muscle sarcoplasmic reticulum. J Biol Chem 265, 2244–2256 (1990). 144  70. Nakai, J. et al. Primary structure and functional expression from cDNA of the cardiac ryanodine receptor/calcium release channel. Febs Lett. 271, 169–177 (1990). 71. Otsus, K. et al. Molecular cloning of cDNA encoding the Ca2+ release channel (ryanodine receptor) of rabbit cardiac muscle sarcoplasmic reticulum. J. Biol. Chem. 265, 13472–83 (1990). 72. Hakamata, Y., Nakai, J., Takeshima, H. & Imoto, K. Primary structure and distribution of a novel ryanodine receptor/Ca release channel from rabbit brain. FEBS Lett 312, 229–235 (1992). 73. Van Petegem, F. Ryanodine receptors: Structure and function. J. Biol. Chem. 287, 31624–31632 (2012). 74. Takeshima, H. et al. Excitation-contraction uncoupling and muscular degeneration in mice lacking functional skeletal muscle ryanodine-receptor gene. Nature 369, 556–559 (1994). 75. Takeshima, H. et al. Embryonic lethality and abnormal cardiac myocytes in mice lacking ryanodine receptor type 2. EMBO J. 17, 3309–3316 (1998). 76. Balschun, D. et al. Deletion of the ryanodine receptor type 3 (RyR3) impairs forms of synaptic plasticity and spatial learning. EMBO J. 18, 5264–5273 (1999). 77. Betzenhauser, M.J., and Marks, A. R. Ryanodine receptor channelopathies. Pflugers Arch. 460, 467–480 (2010). 78. Rosenberg, H., Davis, M., James, D., Pollock, N. & Stowell, K. Malignant hyperthermia. Orphanet J. Rare Dis. 2, 1–14 (2007). 79. Harrison, G. G. Control of the malignant hyperpyrexic syndrome in MHS swine by dantrolene sodium. Br. J. Anaesth. 47, 62–65 (1975). 80. Oo, Y. W. et al. Essential Role of Calmodulin in RyR Inhibition by Dantrolene. Mol. 145  Pharmacol. 88, 57–63 (2015). 81. Zhao, F., Li, P., Chen, S. R. W., Louis, C. F. & Fruen, B. R. Dantrolene inhibition of ryanodine receptor Ca2+ release channels. Molecular mechanism and isoform selectivity. J. Biol. Chem. 276, 13810–13816 (2001). 82. Paul-Pletzer, K. et al. Identification of a dantrolene-binding sequence on the skeletal muscle ryanodine receptor. J. Biol. Chem. 277, 34918–34923 (2002). 83. Jiang, D. et al. Reduced threshold for luminal Ca2+ activation of RyR1 underlies a causal mechanism of porcine malignant hyperthermia. J. Biol. Chem. 283, 20813–20820 (2008). 84. Jungbluth, H. Multi-minicore disease. Orphanet J. Rare Dis. 2, 1–11 (2007). 85. Kontula, K. et al. Catecholaminergic polymorphic ventricular tachycardia: Recent mechanistic insights. Cardiovasc. Res. 67, 379–387 (2005). 86. Eldar, M., Pras, E. & Lahat, H. A missense mutation in a highly conserved region of CASQ2 is associated with autosomal recessive catecholamine-induced polymorphic ventricular tachycardia in bedouin families from Israel. Cold Spring Harb. Symp. Quant. Biol. 67, 333–337 (2002). 87. Roux-buisson, N. et al. Absence of triadin, a protein of the calcium release complex, is responsible for cardiac arrhythmia with sudden death in human. Hum. Mol. Genet. 21, 2759–2767 (2012). 88. Hwang, H. S. et al. Divergent regulation of ryanodine receptor 2 calcium release channels by arrhythmogenic human calmodulin missense mutants. Circ. Res. 114, 1114–1124 (2014). 89. Nyegaard, M. et al. Mutations in calmodulin cause ventricular tachycardia and sudden cardiac death. Am. J. Hum. Genet. 91, 703–712 (2012). 146  90. Muthappan, P. & Calkins, H. Arrhythmogenic Right Ventricular Dysplasia. Prog. Cardiovasc. Dis. 51, 31–43 (2008). 91. Lehnart, S. E. et al. Leaky Ca 2 + release channel / ryanodine receptor 2 causes seizures and sudden cardiac death in mice. J. Clin. Invest. 118, 2230–2245 (2008). 92. Jiang, D., Chen, W., Wang, R., Zhang, L. & Chen, S. R. W. Loss of luminal Ca2+ activation in the cardiac ryanodine receptor is associated with ventricular fibrillation and sudden death. Proc. Natl. Acad. Sci. U. S. A. 104, 18309–14 (2007). 93. Oda, T. et al. Defective regulation of interdomain interactions within the ryanodine receptor plays a key role in the pathogenesis of heart failure. Circulation 111, 3400–3410 (2005). 94. Liu, Z. et al. Dynamic, inter-subunit interactions between the N-terminal and central mutation regions of cardiac ryanodine receptor. J. Cell Sci. 123, 1775–1784 (2010). 95. Zalk, R. & Marks, A. R. Ca2+Release Channels Join the ‘Resolution Revolution’. Trends Biochem. Sci. 42, 543–555 (2017). 96. Saito, A., Inui, M., Radermacher, M., Frank, J. & Fleischer, S. Ultrastructure of the calcium release channel of sarcoplasmic reticulum. J. Cell Biol. 107, 211–9 (1988). 97. Ferguson, D. G., Schwartz, H. W. & Franzini-Armstrong, C. Subunit structure of junctional feet in triads of skeletal muscle: A freeze-drying, rotary-shadowing study. J. Cell Biol. 99, 1735–1742 (1984). 98. Saito, A., Seiler, S., Chu, A. & Fleischer, S. Preparation and morphology of sarcoplasmic reticulum terminal cisternae from rabbit skeletal muscle. J. Cell Biol. 99, 875–885 (1984). 99. Wagenknecht, T. et al. Three-dimensional architecture of the calcium channel/foot structure of sarcoplasmic reticulum. Nature 338, 167–70 (1989). 147  100. Radermacher, M. et al. Cryo-EM of the native structure of the calcium release channel/ryanodine receptor from sarcoplasmic reticulum. Biophys. J. 61, 936–940 (1992). 101. Radermacher, M. et al. Cryo-electron microscopy and three-dimensional reconstruction of the calcium release channel/ryanodine receptor from skeletal muscle. J. Cell Biol. 127, 411–423 (1994). 102. Serysheva, I.I., Orlova, E.V., Chio, W., Sherman, M.B., Hamilton, S.L., and van Heel, M. Electron cryomicroscopy and angular reconstitution used to visualize the skeletal muscle calcium release channel. Nature Structural Biology 2, 18–24 (1995). 103. Ludtke, S.J., Serysheva, I.i., Hamilton, S.L., and Chiu, W. The pore structure of the closed RyR1 channel. Structure 13, 1203–1200 (2005). 104. Samso, M., Wagenknecht, T., and Allen, P. D. Internal structure and visualization of transmemrbane domains of the RyR1 calcium release channel by cryo-EM. Nat. Struct. Mol. Biol. 12, 539–544 (2005). 105. Serysheva, I. I. et al. Subnanometer-resolution electron cryomicroscopy-based domain models for the cytoplasmic region of skeletal muscle RyR channel. Proc. Natl. Acad. Sci. U. S. A. 105, 9610–5 (2008). 106. Samsó, M., Feng, W., Pessah, I. N. & Allen, P. D. Coordinated movement of cytoplasmic and transmembrane domains of RyR1 Upon gating. PLoS Biol. 7, 0980–0995 (2009). 107. Efremov, R. G., Leitner, A., Aebersold, R. & Raunser, S. Architecture and conformational switch mechanism of the ryanodine receptor. Nature 517, 39–43 (2015). 108. Zalk, R. et al. Structure of a mammalian ryanodine receptor. Nature 517, 44–49 (2015). 109. Yan, Z. et al. Structure of the rabbit ryanodine receptor RyR1 at near-atomic resolution. Nature 517, 50–55 (2015). 148  110. Tung, C.-C., Lobo, P. a, Kimlicka, L. & Van Petegem, F. The amino-terminal disease hotspot of ryanodine receptors forms a cytoplasmic vestibule. Nature 468, 585–8 (2010). 111. Yuchi, Z., Lau, K. & Van Petegem, F. Disease mutations in the ryanodine receptor central region: Crystal structures of a phosphorylation hot spot domain. Structure 20, 1201–1211 (2012). 112. Lau, K. & Van Petegem, F. Crystal structures of wild type and disease mutant forms of the ryanodine receptor SPRY2 domain. Nat. Commun. 5, 1–11 (2014). 113. Yuchi, Z. et al. Crystal structures of ryanodine receptor SPRY1 and tandem-repeat domains reveal a critical FKBP12 binding determinant. Nat. Commun. 6, 1–13 (2015). 114. Bai, X. C., Yan, Z., Wu, J., Li, Z. & Yan, N. The Central domain of RyR1 is the transducer for long-range allosteric gating of channel opening. Cell Res. 26, 995–1006 (2016). 115. Peng, W., Shen, H., Wu, J., Guo, W., Pan, X., Wang, R., Chen, A.R.W., and Yan, N. Structural basis for the gating mechanism of the type 2 ryanodine receptor RyR2. Science (80-. ). 354, aah5324 (2016). 116. Sharma, M. R., Jeyakumar, L. H., Fleischer, S. & Wagenknecht, T. Three-dimensional structure of ryanodine receptor isoform three in two conformational states as visualized by cryo-electron microscopy. J. Biol. Chem. 275, 9485–9491 (2000). 117. Amador, F. J. et al. Crystal structure of type I ryanodine receptor amino-terminal beta-trefoil domain reveals a disease-associated mutation ‘hot spot’ loop. Proc. Natl. Acad. Sci. U. S. A. 106, 11040–11044 (2009). 118. Lobo, P. A. & Van Petegem, F. Crystal Structures of the N-Terminal Domains of Cardiac and Skeletal Muscle Ryanodine Receptors: Insights into Disease Mutations. Structure 17, 149  1505–1514 (2009). 119. Kimlicka, L. et al. The cardiac ryanodine receptor N-terminal region contains an anion binding site that is targeted by disease mutations. Structure 21, 1440–9 (2013). 120. Endo, M., Tanaka, M. & Ogawa, Y. Calcium Induced release of calcium from the Sarcoplasmic Reticulum of Skinned Skeletal Muscle Fibres. Nature 228, 34–36 (1970). 121. Meissner, G., Darling, E. & Eveleth, J. Kinetics of Rapid Ca2+ Release by Sarcoplasmic Reticulum. Effects of Ca2+, Mg2+, and Adenine Nucleotides. Biochemistry 25, 236–244 (1986). 122. Bezprozvanny, I., Watras, J. & Ehrlich, B. E. Bell-shaped calcium-response curves of Ins(1,4,5)P3- and calcium-gated channels from endoplasmic reticulum of cerebellum. Nature 351, 751–754 (1991). 123. Palade, P., Mitchell, R.D., and Fleischer, S. Spontaneous calcium release from sarcoplasmic reticulum. J. Biol. Chem. 258, 8098–8107 (1983). 124. Jiang, D. et al. RyR2 mutations linked to ventricular tachycardia and sudden death reduce the threshold for store-overload-induced Ca2+ release (SOICR). Proc. Natl. Acad. Sci. U. S. A. 101, 13062–7 (2004). 125. Seiler, S., Wegener, A. D., Whang, D. D., Hathaway, D. R. & Jones, L. R. High molecular weight proteins in cardiac and skeletal muscle junctional sarcoplasmic reticulum vesicles bind calmodulin, are phosphorylated, and are degraded by Ca2+-activated protease. J. Biol. Chem. 259, 8550–8557 (1984). 126. Fuentes, O., Valdivia, C., Vaughan, D., Coronado, R. & Valdivia, H. H. Calcium-dependent block of ryanodine receptor channel of swine skeletal muscle by direct binding of calmodulin. Cell Calcium 15, 305–316 (1994). 150  127. Yang, H. C., Reedy, M. M., Burke, C. L. & Strasburg, G. M. Calmodulin interaction with the skeletal muscle sarcoplasmic reticulum calcium channel protein. Biochemistry 33, 518–525 (1994). 128. Meissner, G. Evidence of a Role for Calmodulin in the Regulation of Calcium Release from Skeletal Muscle Sarcoplasmic Reticulum. Biochemistry 25, 244–251 (1986). 129. Ikemoto, T., Takeshima, H., Iino, M. & Endo, M. Effect of calmodulin on Ca2+-induced Ca2+ release of skeletal muscle from mutant mice expressing either ryanodine receptor type 1 or type 3. Pflugers Arch. 437, 43–8 (1998). 130. Yamaguchi, N., Xu, L., Evans, K. E., Pasek, D. a & Meissner, G. Different regions in skeletal and cardiac muscle ryanodine receptors are involved in transducing the functional effects of calmodulin. J. Biol. Chem. 279, 36433–36439 (2004). 131. Tripathy, A., Xu, L., Mann, G. & Meissner, G. Calmodulin activation and inhibition of skeletal muscle Ca2+ release channel (ryanodine receptor). Biophys. J. 69, 106–119 (1995). 132. Ikemoto, T., Iino, M. & Endo, M. Enhancing effect of calmodulin on Ca 2+ -induced Ca 2+  release in the sarcoplasmic reticulum of rabbit skeletal muscle fibres. J. Physiol. 487, 573–582 (1995). 133. Buratti, R., Prestipino, G., Menegazzi, P., Treves, S. & Zorzato, F. Calcium dependent activation of skeletal muscle Ca2+ release channel (ryanodine receptor) by calmodulin. Biochemical and biophysical research communications 213, 1082–90 (1995). 134. Samsó, M. & Wagenknecht, T. Apocalmodulin and Ca2+-calmodulin bind to neighboring locations on the ryanodine receptor. J. Biol. Chem. 277, 1349–1353 (2002). 135. Tian, X., Tang, Y., Liu, Y., Wang, R. & Chen, S. R. W. Calmodulin modulates the 151  termination threshold for cardiac ryanodine receptor-mediated Ca 2+ release. Biochem. J. 455, 367–375 (2013). 136. Maximciuc, A. A., Putkey, J. A., Shamoo, Y. & MacKenzie, K. R. Complex of Calmodulin with a Ryanodine Receptor Target Reveals a Novel, Flexible Binding Mode. Structure 14, 1547–1556 (2006). 137. Xin, H. B., Timerman, A. P., Onoue, H., Wiederrecht, G. J. & Fleischer, S. Affinity purification of the ryanodine receptor/calcium release channel from fast twitch skeletal muscle based on its tight association with FKBP12. Biochemical and Biophysical Research Communications 214, 263–270 (1995). 138. Chelu, M. G., Danila, C. I., Gilman, C. P. & Hamilton, S. L. Regulation of ryanodine receptors by FK506 binding proteins. Trends Cardiovasc. Med. 14, 227–234 (2004). 139. Ahern, G. P., Junankar, P. R. & Dulhunty, A. F. Subconductance states in single-channel activity of skeletal muscle ryanodine receptors after removal of FKBP12. Biophys. J. 72, 146–162 (1997). 140. Marx, S. O. et al. PKA Phosphorylation Dissociates FKBP12.6 from the Calcium Release Channel (Ryanodine Receptor). Cell 101, 365–376 (2000). 141. Marx, S. O. & Marks, A. R. Dysfunctional ryanodine receptors in the heart: New insights into complex cardiovascular diseases. J. Mol. Cell. Cardiol. 58, 225–231 (2013). 142. Samsó, M., Shen, X. & Allen, P. D. Structural characterization of the RyR1-FKBP12 interaction. J. Mol. Biol. 356, 917–927 (2006). 143. Sharma, M. R., Jeyakumar, L. H., Fleischer, S. & Wagenknecht, T. Three-dimensional visualization of FKBP12.6 binding to an open conformation of cardiac ryanodine receptor. Biophys. J. 90, 164–172 (2006). 152  144. Cornea, R. L., Nitu, F. R., Samsó, M., Thomas, D. D. & Fruen, B. R. Mapping the ryanodine receptor FK506-binding protein subunit using fluorescence resonance energy transfer. J. Biol. Chem. 285, 19219–19226 (2010). 145. Jiang, M. T. et al. Abnormal Ca2+ release, but normal ryanodine receptors, in canine and human heart failure. Circ. Res. 91, 1015–1022 (2002). 146. George, C. H., Higgs, G. V. & Lai, F. A. Ryanodine receptor mutations associated with stress-induced ventricular tachycardia mediate increased calcium release in stimulated cardiomyocytes. Circ. Res. 93, 531–540 (2003). 147. Perni, S., Lavorato, M. & Beam, K. G. De novo reconstitution reveals the proteins required for skeletal muscle voltage-induced Ca 2+ release. Proc. Natl. Acad. Sci. 201716461 (2017). doi:10.1073/pnas.1716461115 148. Horstick, E. J. et al. Stac3 is a component of the excitation-contraction coupling machinery and mutated in Native American myopathy. Nat. Commun. 4, 1–11 (2013). 149. Dulhunty, A.F., Wei-LaPierre, L., Casarotto, M., and Beard, N. A. Core skeletal muscle ryanodine receptor calcium release complex. Clin. Exp. Pharmacol. Physiol. 44, 3–12 (2017). 150. Roy, S. Q., Mer-, L. & Desguerre, I. Mutations in SEPN1 cause congenital muscular dystrophy with spinal rigidity. 29, 17–18 (2001). 151. Jurynec, M. J. et al. Selenoprotein N is required for ryanodine receptor calcium release channel activity in human and zebrafish muscle. (2008). 152. Halling, D. B., Aracena-parks, P. & Hamilton, S. L. Regulation of Voltage-Gated Ca 2 + Channels by Calmodulin. 1–12 (2006). 153. Bailey, A.G., and Bloch, E. C. Malignant Hyperthermia in a Three-Month-Old American 153  Indian Infant. Anesth. Analg. 66, 1043–1045 (1987). 154. Stamm, D. S. et al. Novel congenital myopathy locus identified in Native American Indians at. (2008). 155. Stamm, D. S. et al. Native American Myopathy : Congenital Myopathy With Cleft Palate , Skeletal Anomalies , and Susceptibility to Malignant Hyperthermia. 1841, 1832–1841 (2008). 156. Nelson, B.R., Wu, F., Liu, Y., Anderson, D.M., McAnally, J., Lin, W., Cannon, S.C., Bassel-Duby, R., and Olson, E. . Skeletal muscle-specific T-tubule protein STAC3 mediates voltage-induced Ca2+ release and contractility. Proc. Natl. Acad. Sci. 110, 11881–11886 (2013). 157. Bower, N. I. et al. Stac3 is required for myotube formation and myogenic differentiation in vertebrate skeletal muscle. J. Biol. Chem. 287, 43936–43949 (2012). 158. Reinholt, B. M., Ge, X., Cong, X., Gerrard, D. E. & Jiang, H. Stac3 Is a Novel Regulator of Skeletal Muscle Development in Mice. PLoS One 8, (2013). 159. Johnson, B. D. et al. Modulation of the cloned skeletal muscle L-type Ca2+ channel by anchored cAMP-dependent protein kinase. J. Neurosci. 17, 1243–1255 (1997). 160. Pérez-García, M. T., Kamp, T. J. & Marbán, E. Functional properties of cardiac L-type calcium channels transiently expressed in HEK293 cells. Roles of alpha 1 and beta subunits. J. Gen. Physiol. 105, 289–305 (1995). 161. Yasuda, T. et al. Auxiliary subunit regulation of high-voltage activated calcium channels expressed in mammalian cells. Eur. J. Neurosci. 20, 1–13 (2004). 162. Polster, A., Perni, S., Bichraoui, H. & Beam, K. G. Stac adaptor proteins regulate trafficking and function of muscle and neuronal L-type Ca 2+ channels. Proc. Natl. Acad. 154  Sci. 112, 602–606 (2015). 163. Suda, N., Franzius, D., Fleig, A., Nishimura, S., Bodding, M., Hoth, M., Takeshima, H., and Penner, R. Ca2+-induced Ca2+ release in Chinese hamster ovary (CHO) cells co-expressing dihydropyridine and ryanodine receptors. J. Gen. Physiol. 109, 619–631 (1997). 164. Takekura, H., Bennett, L., Tanabe, T., Beam, K. G. & Franzini-Armstrong, C. Restoration of junctional tetrads in dysgenic myotubes by dihydropyridine receptor cDNA. Biophys. J. 67, 793–803 (1994). 165. Avila, G., and Dirksen, R. T. Functional impact of the ryanodine receptor on the skeletal muscle L-type Ca(2+) channel. J.Gen.Physiol 115, 467–480 (2000). 166. Franzini-Armstrong, C., Protasi, F. & Ramesh, V. Comparative ultrastructure of Ca2+ release units in skeletal and cardiac muscle. Ann. N. Y. Acad. Sci. 853, 20–30 (1998). 167. Campiglio, M. & Flucher, B. E. STAC3 stably interacts through its C1 domain with Ca v 1.1 in skeletal muscle triads. Sci. Rep. 7, 1–12 (2017). 168. Wilkens, C. M., Kasielke, N., Flucher, B. E., Beam, K. G. & Grabner, M. Excitation-contraction coupling is unaffected by drastic alteration of the sequence surrounding residues L720-L764 of the alpha 1S II-III loop. Proc. Natl. Acad. Sci. U. S. A. 98, 5892–7 (2001). 169. Grabner, M., Dirksen, R. T. & Beam, K. G. Tagging with green fluorescent protein reveals a distinct subcellular distribution of L-type and non-L-type Ca2+ channels expressed in dysgenic myotubes. Proc. Natl. Acad. Sci. U. S. A. 95, 1903–1908 (1998). 170. Kabsch, W. Xds. Acta Crystallogr. Sect. D Biol. Crystallogr. 66, 125–132 (2010). 171. Adams, P. D. et al. PHENIX: A comprehensive Python-based system for macromolecular 155  structure solution. Acta Crystallogr. Sect. D Biol. Crystallogr. 66, 213–221 (2010). 172. McCoy, A. J. et al. Phaser crystallographic software. J. Appl. Crystallogr. 40, 658–674 (2007). 173. Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and development of Coot. Acta Crystallogr. Sect. D Biol. Crystallogr. 66, 486–501 (2010). 174. Murshudov, G. N. et al. REFMAC5 for the refinement of macromolecular crystal structures. Acta Crystallogr. Sect. D Biol. Crystallogr. 67, 355–367 (2011). 175. Innovagen. Peptide Property Calculator. (2015). at <https://pepcalc.com/> 176. Powell, J. A., Petherbridge, L. & Flucher, B. E. Formation of triads without the dihydropyridine receptor subunits in cell lines from dysgenic skeletal muscle. J. Cell Biol. 134, 375–387 (1996). 177. Kurochkina, N. & Guha, U. SH3 domains: Modules of protein-protein interactions. Biophys. Rev. 5, 29–39 (2013). 178. Nakai, J., Tanabe, T., Konno, T., Adams, B. & Beam, K. G. Localization in the II-III loop of the dihydropyridine receptor of a sequence critical for excitation-contraction coupling. J. Biol. Chem. 273, 24983–24986 (1998). 179. Dinkel, H. et al. ELM 2016 - Data update and new functionality of the eukaryotic linear motif resource. Nucleic Acids Res. 44, D294–D300 (2016). 180. Linsley, J. W. et al. Congenital myopathy results from misregulation of a muscle Ca 2+ channel by mutant Stac3. Proc. Natl. Acad. Sci. 114, E228–E236 (2017). 181. Bannister, R. A., Papadopoulos, S., Haarmann, C. S. & Beam, K. G. Effects of inserting fluorescent proteins into the alpha1S II-III loop: insights into excitation-contraction coupling. J.Gen.Physiol 134, 35–51 (2009). 156  182. Polster, A., Nelson, B. R., Papadopoulos, S., Olson, E. N. & Beam, K. G. Stac proteins associate with the critical domain for excitation – contraction coupling in the II – III loop of Ca V 1 . 1. J. Gen. Physiol. doi: 10.1085/jgp.201711917 (2018). 183. Berrou, L., Bernatchez, G. & Parent, L. Molecular determinants of inactivation within the I-II linker of alpha1E (CaV2.3) calcium channels. Biophys. J. 80, 215–228 (2001). 184. Herlitze, S., Hockerman, G., Scheuer, T. & Catterall, W. A. Molecular determinants of inactivation and G protein modulation in the intracellular loop connecting domains I and II of the calcium channel alpha1A subunit. Proc. Natl. Acad. Sci. U. S. A. 94, 1512–6 (1997). 185. Kraus, R. L., Sinnegger, M. J., Glossmann, H., Hering, S. & Striessnig, J. Familial hemiplegic migraine mutations change alpha1A Ca2+ channel kinetics. J. Biol. Chem. 273, 5586–5590 (1998). 186. Stotz, S. C. & Zamponi, G. W. Identification of Inactivation Determinants in the Domain IIS6 Region of High Voltage-activated Calcium Channels. J. Biol. Chem. 276, 33001–33010 (2001). 187. Stotz, S. C. & Zamponi, G. W. Structural determinants of fast inactivation of high voltage-activated Ca2+ channels. Trends Neurosci. 24, 176–182 (2001). 188. Spaetgens, R. L. & Zamponi, G. W. Multiple structural domains contribute to voltage-dependent inactivation of rat brain alpha(1E) calcium channels. J.Biol.Chem. 274, 22428–22436 (1999). 189. Stotz, S. C., Hamid, J., Spaetgens, R. L., Jarvis, S. E. & Zamponi, G. W. Fast inactivation of voltage-dependent calcium channels. A hinged-lid mechanism? J. Biol. Chem. 275, 24575–24582 (2000). 157  190. Findeisen, F. & Minor, D. L. Disruption of the IS6-AID Linker Affects Voltage-gated Calcium Channel Inactivation and Facilitation. J. Gen. Physiol. 133, 327–343 (2009). 191. Lee, A. et al. Ca 2 + / calmodulin binds to and modulates P / Q-type calcium channels. Nature 399, 155–159 (2000). 192. Johny, M. Ben, Yang, P. S., Bazzazi, H. & Yue, D. T. Dynamic switching of calmodulin interactions underlies Ca2+ regulation of CaV1.3 channels. Nat. Commun. 4, 1–30 (2013). 193. Babu, Y. S., Bugg, C. E. & Cook, W. J. Structure of calmodulin refined at 2.2 Å resolution. J. Mol. Biol. 204, 191–204 (1988). 194. Tiaho, F., Piot, C., Nargeot, J. & Richard, S. Regulation of the frequency-dependent facilitation of L-type Ca2+ currents in rat ventricular myocytes. J. Physiol. 477, 237–251 (1994). 195. Zygmunt, A. C. & Maylie, J. Stimulation-dependent facilitation of the high threshold calcium current in guinea-pig ventricular myocytes. J. Physiol. 428, 653–671 (1990). 196. Gurney, A. M., Charnet, P., Pye, J. M. & Nargeot, J. Augmentation of cardiac calcium current by flash photolysis of intracellular caged-Ca2+ molecules. Nature 341, 65–68 (1989). 197. DeMaria, C. D., Soong, T. W., Alseikhan, B. A., Alvania, R. S. & Yue, D. T. Calmodulin bifurcates the local Ca2+ signal that modulates P/Q-type Ca2+ channels. Nature 411, 484–489 (2001). 198. Abe, T., Kurosaki, C., Yoshida, M., Hayashi, F., Hirota, H., and Yokoyama, S. Solution structure of RSGI RUH-051, a C1 domain of STAC3 from human cDNA. Protein Data Bank (2006). doi:10.2210/pdb2db6/pdb 199. Wong King Yuen, S. M., Campiglio, M., Tung, C.-C., Flucher, B. E. & Van Petegem, F. 158  Structural insights into binding of STAC proteins to voltage-gated calcium channels. Proc. Natl. Acad. Sci. 114, E9520–E9528 (2017). 200. Campiglio, M. et al. STAC proteins associate to the IQ domain of Ca V 1.2 and inhibit calcium-dependent inactivation. Proc. Natl. Acad. Sci. 201715997 (2018). doi:10.1073/pnas.1715997115 201. Findeisen, F., Rumpf, C., and Minor, D. L. J. Apo-states of calmodulin and CaBP1 control CaV1 voltage-gated calcium channel function through direct competition for the IQ domain. J. Mol. Biol. 425, 3217–3234 (2013). 202. Das, J. & Rahman, G. M. C1 domains: Structure and ligand-binding properties. Chem. Rev. 114, 12108–12131 (2014). 203. Schredelseker, J. et al. The beta 1a subunit is essential for the assembly of dihydropyridine-receptor arrays in skeletal muscle. Proc. Natl. Acad. Sci. U. S. A. 102, 17219–24 (2005). 204. Gregg, R. G. et al. Absence of the beta subunit (cchb1) of the skeletal muscle dihydropyridine receptor alters expression of the alpha 1 subunit and eliminates excitation-contraction coupling. Proc. Natl. Acad. Sci. U. S. A. 93, 13961–13966 (1996). 205. Barrett, C. F. & Tsien, R. W. The Timothy syndrome mutation differentially affects voltage- and calcium-dependent inactivation of CaV1.2 L-type calcium channels. Proc. Natl. Acad. Sci. U. S. A. 105, 2157–2162 (2008). 206. Van Petegem, F., Chatelain, F. C. & Minor, D. L. Insights into voltage-gated calcium channel regulation from the structure of the CaV1.2 IQ domain-Ca2+/calmodulin complex. Nat. Struct. Mol. Biol. 12, 1108–1115 (2005). 207. Fallon, J. L., Halling, D. B., Hamilton, S. L. & Quiocho, F. A. Structure of calmodulin 159  bound to the hydrophobic IQ domain of the cardiac Cav1.2 calcium channel. Structure 13, 1881–1886 (2005). 208. Lian, L. Y., Myatt, D. & Kitmitto, A. Apo calmodulin binding to the L-type voltage-gated calcium channel Cav1.2 IQ peptide. Biochem. Biophys. Res. Commun. 353, 565–570 (2007). 209. Yap, K. L. et al. Calmodulin target database. J. Struct. Funct. Genomics 1, 8–14 (2000). 210. Rhoads, R. & Friedberg, F. Sequence motifs for calmodulin recognition. FASEB J. 11, 331–340 (2016). 211. Kabsch, W., and Sander, C. Dictionary of Protein Secondary Structure: Pattern Recognition of Hydrogen-Bonded and Geometrical Features. Biopolymers 22, 2577–2637 (1983). 212. Kimlicka, L. & Van Petegem, F. The structural biology of ryanodine receptors. Sci. China Life Sci. 54, 712–724 (2011). 213. Priori, S.G., Napolitano, C., Tiso, N., Memmi, M., Vignati, G., Bloise, R., Sorrentino, V., and Danieli, G. A. Mutation in the cardiac ryanodine receptor gene (hRyR2) underlie catecholaminergic polymorphic ventricular tachycardia. Circulation 103, 196–200 (2001). 214. Priori, S. G. et al. Clinical and molecular characterization of patients with catecholaminergic polymorphic ventricular tachycardia. Circulation 106, 69–74 (2002). 215. Tiso, N. et al. Identification of mutations in the cardiac ryanodine receptor gene in families affected with arrhythmogenic right ventricular cardiomyopathy type 2 (ARVD2). Hum. Mol. Genet. 10, 189–194 (2001). 216. Kimlicka, L., Lau, K., Tung, C.-C. & Van Petegem, F. Disease mutations in the ryanodine receptor N-terminal region couple to a mobile intersubunit interface. Nat. Commun. 4, 160  1506 (2013). 217. Ikemoto, N., and Yamamoto, T. Regulation of calcium release by interdomain interaction within ryanodine receptors. Front. Biosci. 7, d671–d683 (2002). 218. Medeiros-Domingo, A., Bhuiyan, Z.A., Tester, D.J., Hofman, N., Bikker, H., van Tintelen, J.P., Mannens, M.M., Wilde, A. A., and Ackerman, M. J. The RYR2-encoded ryanodine receptor/calcium release channel in patients diagnosed previously with either catecholaminergic polymorphic ventricular tachycardia or genotype negative, exercise-induced long QT syndrome: a comprehensive open reading frame muta. J. Am. Coll. Cardiol. 54, 2065–2074 (2009). 219. Bauce, B. et al. Screening for ryanodine receptor type 2 mutations in families with effort-induced polymorphic ventricular arrhythmias and sudden death: Early diagnosis of asymptomatic carriers. J. Am. Coll. Cardiol. 40, 341–349 (2002). 220. Tang, Y., Tian, X., Wang, R., Fill, M. & Chen, S. R. W. Abnormal termination of Ca 2+ release is a common defect of RyR2 mutations associated with cardiomyopathies. Circ. Res. 110, 968–977 (2012). 221. Cheung, J. W. et al. Short-coupled polymorphic ventricular tachycardia at rest linked to a novel ryanodine receptor (RyR2) mutation: Leaky RyR2 channels under non-stress conditions. Int. J. Cardiol. 180, 228–236 (2015). 222. Wangüemert, F. et al. Clinical and molecular characterization of a cardiac ryanodine receptor founder mutation causing catecholaminergic polymorphic ventricular tachycardia. Hear. Rhythm 12, 1636–1643 (2015). 223. Nettleship, J. E., Brown, J., Groves, M. R. & Geerlof, A. in Structural Proteomics. High-throughput methods. 299–318 (2008). doi:10.1007/978-1-60327-058-8_19 161  224. Borko, L., Bauerova-Hlinkova, V., Hostinova, E., Gasperik, J., Beck, K., Lai, F.A., Zahradnikova, A., and Sevcik, J. Structural insights into the human RyR2 N-terminal region involved in cardiac arrhythmias. Acta Crystallogr. Sect. D Biol. Crystallogr. 70, 2897–2912 (2014). 225. Rebbeck, R. T. et al. S100A1 protein does not compete with calmodulin for ryanodine receptor binding but structurally alters the ryanodine receptoro·calmodulin complex. J. Biol. Chem. 291, 15896–15907 (2016). 226. Xiao, Z. et al. The H29D Mutation Does Not Enhance Cytosolic Ca2+ Activation of the Cardiac Ryanodine Receptor. PLoS One 10, e0139058 (2015). 227. Haugaa, K. H. et al. High prevalence of exercise-induced arrhythmias in catecholaminergic polymorphic ventricular tachycardia mutation-positive family members diagnosed by cascade genetic screening. Europace 12, 417–423 (2010). 228. Hayashi, M. et al. Incidence and risk factors of arrhythmic events in catecholaminergic polymorphic ventricular tachycardia. Circulation 119, 2426–2434 (2009). 229. Van Der Werf, C. et al. Familial evaluation in catecholaminergic polymorphic ventricular tachycardia disease penetrance and expression in cardiac ryanodine receptor mutation-carrying relatives. Circ. Arrhythmia Electrophysiol. 5, 748–756 (2012). 230. Liu, Y. et al. Roles of the NH2-terminal domains of cardiac ryanodine receptor in Ca2+ release activation and termination. J. Biol. Chem. 290, 7736–7746 (2015). 231. Liu, Y. et al. CPVT-associated cardiac ryanodine receptor mutation G357S with reduced penetrance impairs Ca2+release termination and diminishes protein expression. PLoS One 12, 1–18 (2017). 232. Seo, M.-D. et al. Structural and functional conservation of key domains in InsP3 and 162  ryanodine receptors. Nature 483, 108–12 (2012). 233. Tester, D. J. Targeted mutational analysis of the RyR2-encoded cardiac ryanodine receptor in sudden unexplained death: a molecular autopsy of 49 medical examiner/coroner’s cases. Mayo Clin. Proc. 79, 1380–1384 (2004). 234. Napolitano, C. & Priori, S. G. Diagnosis and treatment of catecholaminergic polymorphic ventricular tachycardia. Hear. Rhythm 4, 675–678 (2007). 235. Roston, T.M., Vinocur, J.M., Maginot, K.R., Mohammed, S., Salerno, J.C., Etheridge, S.P.,  et al. Catecholaminergic polymorphic ventricular tachycardia in children: an analysis of therapeutic strategies and outcomes from an international multicenter registry. Circ. Arrhythmia Electrophysiol. 8, 633–642 (2015). 236. Jiang, D. et al. Enhanced store overload-induced Ca2+ release and channel sensitivity to luminal Ca2+ activation are common defects of RyR2 mutations linked to ventricular tachycardia and sudden death. Circ. Res. 97, 1173–1181 (2005). 237. Frank, G. B. The current view of the source of trigger calcium in excitation-contraction coupling in vertebrate skeletal muscle. Biochem. Pharamcology 29, 2399–2406 (1980). 238. Leong, P. & MacLennan, D. H. The cytoplasmic loops between domains II and III and domains III and IV in the skeletal muscle dihydropyridine receptor bind to a contiguous site in the skeletal muscle ryanodine receptor. J. Biol. Chem. 273, 29958–29964 (1998). 239. Xiangen lu, Le Xu,  and G. M. Activation of the Skeletal Muscle Calcium Release Channel by a. J. Biol. Chem. 269, 6511–6516 (1994). 240. Saiki, Y., el-Hayek, R. & Ikemoto, N. Involvement of the Glu724 -Pro760 Region of the Dihydropyridine Receptor 11-111 Loop in Skeletal Muscle-Type Excitation-Contraction Coupling. J.Biol.Chem. 274, 7825–7832 (1999). 163  241. Casarotto, M.G., Green, D., Pace, S., Young, J., and Dulhunty, A. F. Activating the ryanodine receptor with dihydropyridine receptor II-III loop segments: size and charge do matter. Front. Biosci. 1, 2860–2872 (2004). 242. Geiger, M., Wrulich, O.A., Jenny, M., Schwaiger, W., Grunicke, H.H., and Uberail, F. Defining the human targets of phorbol ester and diacylglycerol. Curr. Opin. Mol. Ther. 5, 631–641 (2003). 243. Goel, G., Makkar, H. P. S., Francis, G. & Becker, K. Phorbol Esters: Structure, Biological Activity, and Toxicity in Animals. Int. J. Toxicol. 26, 279–288 (2007). 244. Kazanietz, M. G., Krausz, K. W. & Blumberg, P. M. Differential irreversible insertion of protein kinase C into phospholipid vesicles by phorbol esters and related activators. J. Biol. Chem. 267, 20878–20886 (1992). 245. Sharkey, N. A. & Blumberg, P. M. Short communication: Comparison of the activity of phorbol 12-myristate 13-acetate and the diglyceride glycerol 1-myristate 2-acetate. Carcinogenesis 7, 677–679 (1986). 246. Mosior, M. & Newton, A. C. Mechanism of Interaction of Protein Kinase C with Phorbol Esters. 270, 25526–25533 (1995). 247. Fruen, B. R., Mickelson, J. R., Roghair, T. J., Cheng, H. L. & Louis, C. F. Anions that potentiate excitation-contraction coupling may mimic effect of phosphate on Ca2+ release channel. Am. J. Physiol. 266, C1729-35 (1994). 248. Fruen, B. R., Kane, P. K., Mickelson, J. R. & Louis, C. F. Chloride-dependent sarcoplasmic reticulum Ca2+ release correlates with increased Ca2+ activation of ryanodine receptors. Biophys. J. 71, 2522–2530 (1996). 249. Liu, W., Pasek, D.A., and Meissner, G. Modulation of Ca2+-gated cardiac muscle ca2+-164  release channel (ryanodine receptor) by mono- and divalent ions. Am. J. Physiol. 274, C120–C128 (1998). 250. Kimlicka, L. Structural and biochemical characterizations of the skeletal muscle and cardiac ryanodine receptor N-terminal disease-associated mutants. (2014). 251. Hironaka, T., and Morimoto, S. Intracellular chloride concentration and evidence for the existence of a chloride pump in frog skeletal muscle. Jpn. J. Physiol. 30, 357–363 (1980). 252. Payne, J. A., Rivera, C., Voipio, J. & Kaila, K. Cation-chloride co-transporters in neuronal communication, development and trauma. Trends Neurosci. 26, 199–206 (2003). 253. Duan, D. D. Phenomics of cardiac chloride channels. Compr. Physiol. 3, 667–692 (2013). 254. Desilets, M., Puceat, M., and Vassort, G. Chloride dependence of pH modulation by B-adrenergic agonist in rat cardiomyocytes. Circ Res 75, 862–869 (1994). 255. Treharne, K. J., Crawford, R. M. & Mehta, A. CFTR, chloride concentration and cell volume: Could mammalian protein histidine phosphorylation play a latent role? Exp. Physiol. 91, 131–139 (2006). 256. Peng, W. et al. Structural basis for the gating mechanism of the type 2 ryanodine receptor RyR2. Science. 354 (6310), aah5324, DOI: 10.1126/science.aah5324 (2016).  

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.24.1-0368619/manifest

Comment

Related Items