UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Ligand and modulator binding in the skeletal muscle ryanodine receptor Lobo, Paolo Antonio 2012-12-31

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
[if-you-see-this-DO-NOT-CLICK]
[if-you-see-this-DO-NOT-CLICK]
ubc_2013_spring_lobo_paolo.pdf [ 15.12MB ]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0073428.json
JSON-LD: 1.0073428+ld.json
RDF/XML (Pretty): 1.0073428.xml
RDF/JSON: 1.0073428+rdf.json
Turtle: 1.0073428+rdf-turtle.txt
N-Triples: 1.0073428+rdf-ntriples.txt
Original Record: 1.0073428 +original-record.json
Full Text
1.0073428.txt
Citation
1.0073428.ris

Full Text

LIGAND	
  AND	
  MODULATOR	
  BINDING	
  IN	
  THE	
   SKELETAL	
  MUSCLE	
  RYANODINE	
  RECEPTOR	
   	
   by	
   	
   PAOLO	
  ANTONIO	
  LOBO	
   	
   B.Sc.,	
  The	
  University	
  of	
  British	
  Columbia,	
  2010	
   	
   	
   A	
  THESIS	
  SUBMITTED	
  IN	
  PARTIAL	
  FULFILLMENT	
  OF	
  THE	
  REQUIREMENTS	
  FOR	
  THE	
   DEGREE	
  OF	
   	
   MASTER	
  OF	
  SCIENCE	
   	
   in	
   	
   THE	
  FACULTY	
  OF	
  GRADUATE	
  STUDIES	
   	
   (Biochemistry)	
   	
   	
   THE	
  UNIVERSITY	
  OF	
  BRITISH	
  COLUMBIA	
   (Vancouver)	
   	
   December	
  2012	
   	
   ©	
  Paolo	
  Antonio	
  Lobo,	
  2012	
    Abstract	
   	
   Calcium	
  (Ca2+)	
  is	
  required	
  in	
  the	
  cytoplasm	
  as	
  a	
  potent	
  second	
  messenger	
  for	
  a	
   variety	
  of	
  vital	
  physiological	
  events	
  in	
  the	
  cell.	
  Its	
  supply	
  to	
  the	
  cytoplasm	
  can	
  either	
   be	
   from	
   extracellular	
   sources,	
   or	
   from	
   intracellular	
   sarcoplasmic	
   reticulum	
   stores,	
   predominantly	
   through	
   the	
   ryanodine	
   receptors	
   (RyRs).	
   The	
   latter	
   are	
   large,	
   ~2	
   MDa	
   homotetrameric	
   channels	
   that	
   sense	
   initial	
   Ca2+	
   blips	
   due	
   to	
   voltage-­‐gated	
   calcium	
   channel	
   influx	
   and	
   respond	
   by	
   opening	
   to	
   result	
   in	
   physiologically	
   significant	
   concentration	
  spikes.	
  This	
  action	
  is	
  known	
  as	
  calcium	
  induced	
  calcium	
  release	
  and	
  is	
   the	
  major	
  process	
  by	
  which	
  an	
  excitation	
  signal	
  is	
  translated	
  to	
  a	
  physical,	
  muscular	
   contraction.	
  Cytoplasmic	
  Ca2+	
  concentrations	
  have	
  to	
  be	
  very	
  well	
  regulated	
  and	
  must	
   return	
  to	
  resting	
  levels	
  for	
  subsequent	
  contractions	
  to	
  occur.	
  The	
  importance	
  of	
  this	
   with	
  regard	
  to	
  the	
  RyR	
  can	
  be	
  seen	
  in	
  two	
  different	
  ways:	
  Firstly,	
  its	
  large	
  cytoplasmic	
   bulk,	
  a	
  huge	
  docking	
  site	
  for	
  modulators,	
  emphasizes	
  the	
  necessity	
  for	
  regulation,	
  and	
   secondly,	
  mutations	
  in	
  the	
  RyR	
  can	
  cause	
  severe	
  genetic	
  diseases	
  as	
  a	
  result	
  of	
  Ca2+	
   mishandling.	
   Presented	
   here,	
   are	
   preliminary	
   structural	
   and	
   binding	
   studies	
   for	
   several	
  different	
  RyR	
  regulators,	
  both	
  physiological	
  and	
  pharmacological.	
   Skeletal	
   muscle	
   RyR	
   (RyR1)	
   1-­‐617	
   that	
   contains	
   the	
   drug	
   dantrolene’s	
   supposed	
   binding	
   site	
   was	
   crystallized	
   and	
   its	
   structure	
   determined.	
   Isothermal	
   Titration	
   Calorimetry	
   (ITC)	
   and	
   co-­‐crystallization	
   attempts	
   have	
   not	
   confirmed	
   the	
   binding	
   of	
   dantrolene.	
   The	
   structure	
   does	
   however	
   shed	
   light	
   on	
   the	
   physical	
   involvement	
  of	
  phosphatases	
  as	
  modulators	
  of	
  the	
  channel.	
   Caffeine	
   binding	
   was	
   detected	
   successfully	
   by	
   ITC	
   and	
   attributed	
   to	
   RyR1	
   217-­‐ 536.	
   A	
   co-­‐crystal	
   structure	
   yielded	
   a	
   binding	
   site	
   in	
   the	
   construct	
   that	
   could	
   not	
   be	
   knocked	
  out	
  by	
  site-­‐directed	
  mutagenesis	
  according	
  to	
  ITC.	
  	
   Three	
   types	
   of	
   modulators	
   were	
   shown	
   to	
   bind	
   RyR1	
   4071-­‐4128	
   by	
   ITC.	
   1)	
   Ca2+,	
   which	
   affects	
   the	
   channel	
   both	
   positively	
   and	
   negatively,	
   2)	
   magnesium	
   ions,	
   which	
  inhibit	
  the	
  channel,	
  and	
  3)	
  an	
  intrinsic	
  ligand	
  in	
  RyR1:	
  residues	
  4295-­‐4325;	
  a	
   peptide	
  that	
  shows	
  affinity	
  for	
  Calmodulin,	
  yet	
  another	
  modulator	
  of	
  RyR.	
    	
    ii	
    The	
  results	
  provide	
  insight	
  into	
  allosteric	
  reactions	
  as	
  a	
  result	
  of	
  RyR	
  ligand	
  or	
   modulator	
  binding.	
   	
    	
    iii	
    Preface	
   	
    Ching-­‐Chieh	
   Tung	
   optimized	
   the	
   purification	
   for	
   skeletal	
   ryanodine	
   receptor	
   (RyR1)	
   residues	
   1-­‐536	
   and	
   RyR1	
   217-­‐536.	
   In	
   addition,	
   for	
   the	
   latter	
   construct,	
   he	
   performed	
  the	
  initial	
  Isothermal	
  Titration	
  Calorimetry	
  (ITC)	
  experiments	
  in	
  Chapter	
   3.2	
   and	
   discovered	
   the	
   caffeine	
   co-­‐crystallization	
   condition	
   discussed	
   in	
   Chapter	
   3.2.1.	
   Kelvin	
   Lau	
   carried	
   out	
   the	
   ITC	
   experiments	
   on	
   RyR1	
   1-­‐617	
   with	
   dantrolene	
   and	
   azumolene	
   (Chapter	
   3.1.3).	
   He	
   is	
   also	
   responsible	
   for	
   all	
   Mass	
   Spectrometry	
   discussed	
  in	
  this	
  thesis.	
   Lynn	
   Kimlicka	
   maintains	
   a	
   human	
   mutation	
   database	
   for	
   both	
   RyR1	
   and	
   the	
   cardiac	
   isoform.	
   The	
   database	
   was	
   referred	
   to	
   several	
   times	
   in	
   this	
   discussion	
   (Chapters	
  1.3,	
  3.1.1,	
  4.1.3,	
  and	
  4.3.2).	
   Ulrika	
   Brath	
   completed	
   the	
   NMR	
   experiments	
   on	
   RyR1	
   4071-­‐4138	
   to	
   detect	
   calcium	
  ion	
  binding	
  to	
  RyR1	
  4071-­‐4138	
  (Chapter	
  3.3.1).	
    	
    iv	
    Table	
  of	
  Contents	
   	
   Abstract .................................................................................................................. ii	
   Preface ....................................................................................................................iv	
   Table	
  of	
  Contents..................................................................................................v	
   List	
  of	
  Tables ......................................................................................................viii	
   List	
  of	
  Figures .......................................................................................................ix	
   List	
  of	
  Abbreviations ..........................................................................................xi	
   Acknowledgements ......................................................................................... xiv	
   Dedication.............................................................................................................xv	
   	
   1	
    Introduction ................................................................................................1	
    	
    1.1	
  Calcium	
  Release	
  and	
  Regulation........................................................................................1	
    	
    	
    1.1.1	
  Calcium	
  Induced	
  Calcium	
  Release	
  and	
  E-­‐C	
  coupling...............................2	
    	
    	
    1.1.2	
  Calmodulin	
  and	
  Calmodulin	
  Binding	
  Domains..........................................4	
    	
    1.2	
  Structural	
  Information...........................................................................................................9	
    	
    	
    1.2.1	
  Cryo-­‐Electron	
  Microscopy ..................................................................................9	
    	
    	
    1.2.2	
  Crystallographic	
  Insight ................................................................................... 12	
    	
    	
    1.2.3	
  The	
  IP3	
  Receptor .................................................................................................. 15	
    	
    1.3	
  Disease	
  Mutations................................................................................................................. 17	
    	
    	
    1.3.1	
  Phosphorylation	
  and	
  Disease......................................................................... 19	
    	
    	
    1.3.2	
  The	
  Molecular	
  Cause	
  of	
  Disease.................................................................... 20	
    	
    1.4	
  Small	
  Molecule	
  Drugs	
  and	
  Ligands................................................................................ 21	
    	
    	
    1.4.1	
  Dantrolene.............................................................................................................. 22	
    	
    	
    1.4.2	
  Purine	
  Derivatives............................................................................................... 25	
    	
    1.5	
  Hypotheses	
  and	
  Goals ......................................................................................................... 29	
    	
    	
    v	
    2	
    Materials	
  and	
  Methods ......................................................................... 31	
    	
    2.1	
  Cloning	
  and	
  Expression...................................................................................................... 31	
    	
    2.2	
  Protein	
  Purification.............................................................................................................. 34	
    	
    2.3	
  X-­‐ray	
  Crystallography ......................................................................................................... 36	
    	
    2.4	
  ITC................................................................................................................................................ 37	
    	
    2.5	
  in	
  silico	
  Docking	
  -­‐	
  RyR1	
  1-­‐617	
  into	
  Cryo-­‐EM	
  Maps ................................................ 38	
    	
    2.6	
  in	
  silico	
  Docking	
  -­‐	
  Caffeine	
  into	
  RyR1	
  1-­‐536.............................................................. 39	
    	
    3	
    Results ........................................................................................................ 40	
    	
    3.1	
  A	
  New	
  Structural	
  Domain.................................................................................................. 40	
    	
    	
    3.1.1	
  New	
  loops	
  and	
  Disease	
  Mutations ............................................................... 45	
    	
    	
    3.1.2	
  Docking	
  into	
  Cryo-­‐EM	
  Maps ........................................................................... 46	
    	
    	
    3.1.3	
  Dantrolene	
  Binding............................................................................................. 49	
    	
    	
    3.1.4	
  N-­‐terminal	
  Construct	
  Summary.................................................................... 50	
    	
    3.2	
  Caffeine	
  Binding..................................................................................................................... 52	
    	
    	
    3.2.1	
  Structural	
  Insight?............................................................................................... 53	
    	
    	
    3.2.2	
  W269A	
  Mutant	
  Binding .................................................................................... 58	
    	
    	
    3.2.3	
  An	
  in	
  silico	
  Search	
  for	
  a	
  Caffeine	
  Binding	
  Site .......................................... 60	
    	
    	
    3.2.4	
  ATP	
  Binding	
  Experiments................................................................................ 64	
    	
    3.3	
  RyR1	
  4071-­‐4138 ................................................................................................................... 66	
    	
    	
    3.3.1	
  Ca2+	
  and	
  Mg2+	
  Binding	
  to	
  RyR1	
  EF-­‐hands ................................................. 72	
    	
    	
    3.3.2	
  Interaction	
  with	
  RyR1	
  CaMBDs..................................................................... 73	
    	
   4	
    Discussion ................................................................................................. 75	
    	
    4.1	
  RyR1	
  ABCd ............................................................................................................................... 75	
    	
    	
    4.1.1	
  A	
  Larger	
  Surface	
  Area	
  for	
  Binding ............................................................... 75	
    	
    	
    4.1.2	
  PP1	
  Recruitment	
  by	
  a	
  Leucine	
  Zipper ........................................................ 77	
    	
    	
    4.1.3	
  Molecular	
  Insight	
  from	
  Docking.................................................................... 81	
    	
    	
    4.1.4	
  Interface	
  1 ............................................................................................................... 84	
    	
    	
    4.1.5	
  Continuing	
  the	
  Debate	
  on	
  Dantrolene........................................................ 86	
    	
    vi	
    	
    4.2	
  Purines	
  and	
  RyR1.................................................................................................................. 86	
    	
    4.3	
  Regulation	
  by	
  EF-­‐hands ..................................................................................................... 88	
    	
    	
    4.3.1	
  The	
  Involvement	
  of	
  CaMBDs .......................................................................... 89	
    	
    	
    4.3.2	
  Disease	
  and	
  RyR	
  EF-­‐hands .............................................................................. 89	
    	
    4.4	
  Allostery	
  in	
  RyR1................................................................................................................... 91	
    	
   References ........................................................................................................... 93	
    	
    vii	
    List	
  of	
  Tables	
   	
   NOTE:	
   In	
   tables	
   in	
   which	
   residues	
   have	
   been	
   numbered,	
   the	
   numbering	
   used	
   is	
   denoted	
   at	
   the	
   end	
   of	
   each	
   legend.	
   In	
   general,	
   tables	
   with	
   disease	
   mutations	
   are	
   numbered	
   according	
   to	
   RyR1	
   or	
   RyR2	
   Homo	
   sapiens,	
   and	
   all	
   others	
   adopt	
   the	
   RyR1	
   Oryctolagus	
  cuniculus	
  numbering.	
   	
   TABLE	
  1.	
   List	
  of	
  primers	
  to	
  make	
  constructs.............................................................................. 33	
   TABLE	
  2.	
   Data	
  collection	
  and	
  refinement	
  statistics	
  for	
  RyR1	
  1-­‐617................................. 43	
   TABLE	
  3.	
   Additional	
  mutations	
  or	
  their	
  interactions	
  in	
  the	
  RyR1	
  1-­‐617	
  structure... 45	
   TABLE	
  4.	
   N-­‐terminal	
  constructs ....................................................................................................... 51	
   TABLE	
  5.	
   Data	
   collection	
   and	
   refinement	
   statistics	
   for	
   RyR1BC	
   and	
   RyR1ABC	
   with	
   caffeine..................................................................................................................................... 55	
   TABLE	
  6.	
   EF-­‐hand	
  constructs............................................................................................................. 67	
   TABLE	
  7.	
   Disease	
  mutations	
  in	
  RyR1/2	
  EF-­‐hands.................................................................... 90	
   	
    	
    viii	
    List	
  of	
  Figures	
   	
    NOTE:	
   In	
   figures	
   in	
   which	
   residues	
   have	
   been	
   numbered,	
   the	
   numbering	
   used	
   is	
   denoted	
   at	
   the	
   end	
   of	
   each	
   legend.	
   In	
   general,	
   figures	
   with	
   labelled	
   disease	
   mutations	
   are	
   numbered	
   according	
   to	
   RyR1	
   or	
   RyR2	
   Homo	
   sapiens,	
   and	
   all	
   others	
   adopt	
   the	
   RyR1	
  Oryctolagus	
  cuniculus	
  numbering.	
   	
   FIGURE	
  1.	
    RyR,	
  Ca2+	
  and	
  muscle	
  contraction .............................................................................4	
    FIGURE	
  2.	
    Calmodulin	
  and	
  EF-­‐hands.............................................................................................7	
    FIGURE	
  3.	
  	
    A	
  9.6	
  Å	
  cryo-­‐EM	
  reconstruction	
  of	
  RyR1............................................................. 10	
    FIGURE	
  4	
    Closed	
  vs	
  open	
  states	
  of	
  RyR .................................................................................... 11	
    FIGURE	
  5.	
    Docking	
  of	
  RyR1ABC .................................................................................................... 14	
    FIGURE	
  6.	
    IP3R1ABC	
  vs	
  RYR1ABC............................................................................................... 17	
    FIGURE	
  7.	
    The	
  molecular	
  structures	
  of	
  dantrolene	
  and	
  azumolene ............................. 23	
    FIGURE	
  8.	
    Adenine	
  nucleotides	
  and	
  xanthines....................................................................... 26	
    FIGURE	
  9.	
    Purine	
  binding	
  by	
  aromatic	
  stacking .................................................................... 28	
    FIGURE	
  10.	
   Method	
  flowchart .......................................................................................................... 30	
   FIGURE	
  11.	
   Expression,	
  solubility	
  and	
  purification	
  of	
  N-­‐terminal	
  constructs ............ 41	
   FIGURE	
  12.	
   RyR1	
  1-­‐617	
  crystals ..................................................................................................... 42	
   FIGURE	
  13.	
   RyR1	
  1-­‐617	
  structure .................................................................................................. 44	
   FIGURE	
  14.	
   Docking	
  contrast ............................................................................................................ 46	
   FIGURE	
  15.	
   Docking	
  of	
  the	
  RyR1	
  1-­‐617	
  crystal	
  structure	
  into	
  EMDB	
  1606................. 48	
   FIGURE	
  16.	
   N-­‐terminal	
  construct	
  schematic.............................................................................. 50	
   FIGURE	
  17.	
  	
   Caffeine	
  binds	
  RyR1BC................................................................................................ 53	
   FIGURE	
  18.	
   Crystallographic	
  caffeine	
  binding	
  interactions................................................. 56	
    	
    ix	
    FIGURE	
  19.	
   Structural	
  insight	
  into	
  caffeine	
  binding	
  to	
  RyR?	
   ............................................. 57	
   FIGURE	
  20.	
   W269A	
  mutation............................................................................................................ 59	
   FIGURE	
  21.	
   DOCK	
  and	
  AutoDock	
  scoring	
  bubble	
  charts....................................................... 61	
   FIGURE	
  22.	
   Docking	
  caffeine	
  in	
  silico ............................................................................................ 63	
   FIGURE	
  23.	
   ATP	
  binding	
  to	
  RyR1ABC ........................................................................................... 65	
   FIGURE	
  24.	
   Expression	
  of	
  PHYRE2	
  guided	
  domains ............................................................... 68	
   FIGURE	
  25.	
   Potential	
  EF-­‐hands	
  in	
  RyR ......................................................................................... 70	
   FIGURE	
  26.	
   Cysteine	
  dimerisation	
  in	
  RyR1	
  4071-­‐4138 ........................................................ 71	
   FIGURE	
  27.	
   RyR1	
  4071-­‐4138	
  binds	
  Ca2+	
  and	
  Mg2+ ................................................................. 72	
   FIGURE	
  28.	
   RyR1	
  4071-­‐4138	
  binds	
  a	
  CaMBD3	
  mutant ........................................................ 74	
   FIGURE	
  29.	
   A	
  comparison	
  between	
  IP3R	
  and	
  RyR1	
  1-­‐617................................................... 77	
   FIGURE	
  30.	
   A	
  LZ	
  motif	
  in	
  RyR1d...................................................................................................... 79	
   FIGURE	
  31.	
   An	
  antiparallel	
  LZ	
  in	
  the	
  RyR1ABCd	
  crystal	
  structure.................................. 80	
   FIGURE	
  32.	
  	
   RyR1	
  ABCd	
  docking	
  analysis .................................................................................... 83	
   FIGURE	
  33.	
  	
   RyR1	
  ABCd	
  in	
  channel	
  opening ............................................................................... 85	
   FIGURE	
  34.	
  	
   Allosteric	
  communication	
  in	
  RyR ........................................................................... 92	
   	
   	
   	
    	
    x	
    List	
  of	
  Abbreviations	
   	
   (2/3)D	
    (Two/three)	
  dimensional	
    (Apo-­‐/Ca2+-­‐)CaM	
    (Calcium-­‐free/Calcium-­‐bound)	
  Calmodulin	
    (d)NTP	
    (Deoxyribo)	
  Nucleotide	
  triphospate	
    (n/µ/m)M	
  or	
  l	
    (nano/micro/milli-­‐)	
  molar	
  or	
  litre	
    ∆G	
    Gibb’s	
  free	
  energy	
    ∆H	
    Change	
  in	
  enthalpy	
    ∆S	
    Change	
  in	
  entropy	
    ACh	
    Acetylcholine	
    AD#	
    Rank	
  number	
  #	
  AutoDock	
  solution	
    ANSA	
    8-­‐anilino-­‐1-­‐naphthalenesulfonic	
  acid	
  ammonium	
    APS	
    Advanced	
  Photon	
  Source	
    ARVD2	
    Arrhythmogenic	
  Right	
  Ventricular	
  Dysplasia	
  Type	
  2	
    ATP	
    Adenosine	
  triphosphate	
    Ca2+	
  	
    Calcium	
  ions	
    CaMBD	
    Calmodulin	
  Binding	
  Domain	
    CaMKII	
    Ca2+/CaM-­‐dependent	
  kinase	
  II	
    CaV	
    Voltage-­‐Gated	
  Calcium	
  Channel	
    CCD	
    Central	
  Core	
  Disease	
    CICR	
    Calcium	
  Induced	
  Calcium	
  Release	
    CLS	
    Canadian	
  Lightsource	
    CPVT	
    Catecholaminergic	
  Polymorphic	
  Ventricular	
  Tachycardia	
    CSQ	
    Calsequestrin	
    CV	
    Column	
  volume	
    D#	
    Rank	
  number	
  #	
  DOCK	
  solution	
    DAG	
    1,2-­‐diacylglycerol	
    DNA	
    Deoxyribonucleic	
  acid	
    E-­‐C	
    Excitation-­‐Contraction	
    E.	
  coli	
    Escherichia	
  coli	
    	
    xi	
    EC50	
    Half	
  maximal	
  effective	
  concentration	
    EDTA	
    Ethylenediaminetetraacetic	
  acid	
    EM	
    Electron	
  Microscopy	
    EMDB	
    Electron	
  Microscopy	
  Data	
  Bank	
    ER	
    Endoplasmic	
  Reticulum	
    FRET	
    Förster	
  Resonance	
  Energy	
  Transfer	
    GFP	
    Green	
  Fluorescent	
  Protein	
    GST	
    Glutathione	
  S-­‐Transferase	
    GUI	
    Graphical	
  User	
  Interface	
    HEPES	
    4-­‐(2-­‐hydroxyethyl)-­‐1-­‐piperazineethanesulfonic	
  acid	
    IP3	
    Inositol	
  1,4,5-­‐triphosphate	
    IP3R[#](A,B,C)	
    IP3	
  receptor	
  [isoform	
  1,	
  2	
  or	
  3](domains	
  A,	
  B	
  and	
  C)	
    IPTG	
    Isopropyl-­‐β-­‐D-­‐thiogalactoside	
    ITC	
    Isothermal	
  Titration	
  Calorimetry	
    K+	
    Potassium	
  ions	
    kDa	
    Kilodalton	
    LIC	
    Ligation	
  Independent	
  Cloning	
    LZ	
  (1,2,	
  or	
  3)	
    Leucine/Isoleucine/Valine	
  zipper	
  (1,	
  2	
  or	
  3)	
    MBP	
    Maltose	
  Binding	
  Protein	
    MDa	
    Megadalton	
    MH	
    Malignant	
  Hyperthermia	
    Na+	
    Sodium	
  ions	
    NaV	
    Voltage-­‐Gated	
  Sodium	
  Channel	
    NCX	
    Sodium-­‐Calcium	
  Exchanger	
    OD600	
    Optical	
  density	
  at	
  600	
  nm	
    P	
  or	
  S/N	
    Pellet	
  or	
  Supernatant	
    PAC	
    Phosphatase	
  Access	
  Channel	
    PCR	
    Polymerase	
  Chain	
  Reaction	
    PIP2	
    Phosphatidylinositol	
  4,5-­‐triphosphate	
    RyR[#](A,B,C)	
    Ryanodine	
  Receptor	
  [isoform	
  1,	
  2	
  or	
  3]	
  (domains	
  A,	
  B	
  and	
  C)	
    RyR1ABC	
    RyR1	
  1-­‐532	
  (structure,	
  the	
  construct	
  is	
  longer)	
    	
    xii	
    RyR1ABCd	
    RyR1	
  1-­‐577	
  (structure,	
  the	
  construct	
  is	
  longer)	
    SDS	
  PAGE	
    Sodium	
  dodecyl	
  sulphate	
  polyacrylamide	
  gel	
  electrophoresis	
    SERCA	
    Sarcoplasmic/Endoplasmic	
  Reticulum	
  Calcium	
  ATPase	
    SOICR	
    Store	
  overload-­‐induced	
  Ca2+	
  release	
    SR	
    Sarcoplasmic	
  Reticulum	
    SSRL	
    Stanford	
  Synchrotron	
  Radiation	
  Lightsource	
    TEV	
    Tobacco	
  Etch	
  Virus	
    Tris	
    tris(hydroxymethyl)aminomethane	
    βME	
    β-­‐mercaptoethanol	
    	
    	
    xiii	
    	
   Acknowledgements	
   	
   I	
  would	
  like	
  to	
  thank	
  Dr.	
  Filip	
  Van	
  Petegem	
  for	
  several	
  years	
  of	
  patient	
  teaching	
   and	
   leading	
   in	
   his	
   laboratory.	
   All	
   of	
   the	
   Van	
   Petegem	
   lab	
   members	
   have	
   been	
   a	
   pleasure	
  to	
  work	
  with.	
  Special	
  appreciation	
  goes	
  to	
  Dr.	
  Michael	
  Yuchi,	
  Kelvin	
  Lau,	
  Jett	
   (C-­‐C.)	
   Tung	
   and	
   Lynn	
   Kimlicka	
   for	
   their	
   valuable	
   input	
   and	
   discussions.	
   From	
   neighbouring	
  labs,	
  Sarah	
  Chow	
  and	
  Maen	
  Sarhan	
  attended	
  lab	
  meetings	
  have	
  always	
   been	
  keen	
  to	
  provide	
  useful	
  insight.	
   	
    I	
  am	
  very	
  grateful	
  for	
  the	
  PyMOL	
  Molecular	
  Graphics	
  System	
  (Version	
  1.5.0.4	
    Schrödinger,	
   LLC),	
   which	
   was	
   used	
   to	
   create	
   all	
   structural	
   figures	
   and	
   in	
   addition	
   carry	
  out	
  some	
  manual	
  docking	
  (Figure	
  34).	
   	
    With	
  regard	
  to	
  crystallographic	
  data	
  collection	
  I	
  would	
  like	
  to	
  thank	
  all	
  of	
  the	
    staff	
  at	
  APS	
  23-­‐ID-­‐D-­‐GM/CA	
  and	
  CLS	
  08-­‐ID1,	
  as	
  well	
  as	
  Gunnar	
  Olovsson	
  and	
  Anson	
   Chan	
  at	
  UBC	
  homesources.	
   	
    Finally,	
   thank	
   you	
   to	
   my	
   committee	
   members	
   Dr.	
   Joerg	
   Gsponer	
   and	
   Dr.	
   Calvin	
    Yip	
  for	
  their	
  input	
  and	
  general	
  approachability.	
   	
   	
    	
    xiv	
    Dedication	
   	
   	
    To	
   my	
   parents	
   in	
   Kenya,	
   who	
   from	
   thousands	
   of	
   miles	
   away,	
   have	
   lovingly	
    provided	
   the	
   support	
   I	
   have	
   needed	
   to	
   succeed.	
  To	
  my	
  sister	
  Ariana,	
  who	
  inspires	
   me	
   with	
   her	
   confidence	
   and	
   her	
   independence.	
   And	
   in	
   Canada,	
   to	
   Julio	
   and	
   Debbie,	
   the	
   strongest	
  pillars	
  in	
  my	
  life.	
  I	
  strive	
  to	
  make	
  you	
  all	
  proud!	
   	
    	
    xv	
    1	
  Introduction	
   	
   	
   	
    1.1	
   Calcium	
  Release	
  and	
  Regulation	
   	
   Extracellular	
   signals	
   such	
   as	
   membrane	
   depolarization	
   or	
   adrenaline	
   binding	
   have	
   to	
   be	
   relayed	
   from	
   receptors	
   at	
   the	
   cell	
   surface	
   to	
   intracellular	
   target	
   molecules.	
   To	
   do	
   this	
   second	
   messengers	
   are	
   required,	
   which	
   act	
   via	
   signal	
   transduction	
   and	
   amplify	
  the	
  signal.	
  One	
  of	
  the	
  most	
  potent	
  second	
  messengers	
  is	
  calcium	
  (Ca2+),	
  which	
   is	
  required	
  for	
  a	
  large	
  variety	
  of	
  physiological	
  functions.	
  As	
  a	
  result,	
  the	
  demand	
  for	
   Ca2+	
   in	
  the	
  cell	
  can	
  be	
  substantial	
  and	
  although	
  extracellular	
  concentrations	
  are	
  high,	
   the	
  main	
  supply	
  of	
  the	
  ions	
  for	
  these	
  physiological	
  functions	
  are	
  intracellular	
  stores:	
   namely	
   the	
   Sarcoplasmic	
   and	
   Endoplasmic	
   Reticula	
   (SR	
   and	
   ER).	
   Here	
   the	
   main	
   sentinels	
  of	
  these	
  stores	
  are	
  Ryanodine	
  Receptors	
  (RyR),	
  which	
  are	
  activated	
  by,	
  and	
   conduct	
  Ca2+	
  (Endo	
  et	
  al.	
  1970,	
  Fabiato,	
  1983).	
   RyRs	
  are	
  large	
  homotetrameric	
  ion	
  channels	
  located	
  in	
  the	
  SR/ER	
  membranes	
   of	
   the	
   cell.	
   With	
   each	
   monomer	
   measuring	
   ~565	
   kDa,	
   a	
   complete	
   structure	
   exceeds	
   2.2	
  MDa.	
  Despite	
  its	
  main	
  function	
  in	
  controlling	
  calcium	
  release	
  into	
  the	
  cell,	
  it	
  gets	
   its	
  name	
  from	
  the	
  observed	
  high	
  affinity	
  binding	
  of	
  the	
  plant	
  toxin	
  Ryanodine,	
  which	
   locks	
  the	
  channel	
  in	
  a	
  subconductance	
  state	
  at	
  lower	
  (nanomolar)	
  concentrations	
  and	
   completely	
   blocks	
   it	
   at	
   higher	
   (micromolar)	
   concentrations	
   (Meissner	
   1986).	
   There	
   are	
  three	
  identified	
  isoforms	
  in	
  mammalian	
  organisms:	
  RyR1,	
  which	
  is	
  predominantly	
   expressed	
  in	
  skeletal	
  muscle	
  (Takeshima	
  1989),	
  RyR2,	
  mostly	
  in	
  cardiac	
  muscle	
  (Otsu	
   et	
  al.	
  1990),	
  and	
  the	
  more	
  ubiquitously	
  expressed	
  RyR3,	
  which	
  was	
  first	
  discovered	
  in	
   the	
  brain	
   (Hakamata	
  et	
  al.	
  1992).	
  Although	
  they	
  do	
  generally	
  follow	
  the	
  this	
  trend	
  in	
   expression,	
   all	
   three	
   types	
   are	
   found	
   in	
   other	
   tissues	
   as	
   well,	
   and	
   are	
   close	
   to	
   70%	
   identical	
   with	
   only	
   a	
   few	
   divergent	
   regions	
   (Sorrentino	
   and	
   Volpe	
   1993).	
   Here,	
   the	
   focus	
  will	
  be	
  predominantly	
  on	
  RyR1.	
    	
    1	
    In	
  proof	
  of	
  the	
  reliance	
  on	
  intracellular	
  stores,	
  it	
  has	
  been	
  demonstrated	
  that	
   an	
   extracellular	
   influx	
   of	
   Ca2+	
   through	
   voltage-­‐gated	
   Ca2+	
   channels	
   (CaVs)	
   is	
   not	
   necessary	
   for	
   contraction	
   in	
   skeletal	
   muscle.	
   Both	
   twitches	
   (Armstrong	
   et	
   al.	
   1972)	
   and	
   sarcoplasmic	
   Ca2+	
   release	
   (Miledi	
   et	
   al.	
   1984)	
   can	
   still	
   occur	
   in	
   muscle	
   cells	
   exposed	
   to	
   Ca2+-­‐free	
   solutions.	
   Instead	
   there	
   is	
   thought	
   to	
   be	
   a	
   direct	
   physical	
   link	
   between	
   RyR1	
   and	
   L-­‐type	
   CaV	
   (CaV1.1)	
   (Rios	
   and	
   Brum	
   1987,	
   Block	
   et	
   al.	
   1988,	
   Tanabe	
   et	
   al.	
   1988,	
   Tanabe	
   et	
   al.	
   1990).	
   Here	
   CaVs	
   are	
   recruited	
   as	
   voltage	
   sensors	
   and	
   in	
   turn	
   cause	
   Ca2+	
   release	
   through	
   RyRs.	
   This	
   initial	
   spike	
   in	
   concentration	
   would	
   then	
   trigger	
   further	
   Ca2+	
   release	
   via	
   neighbouring	
   RyRs	
   that	
   are	
   physiologically	
   clustered,	
   corner-­‐to-­‐corner,	
   in	
   ‘chessboard’	
   arrays	
   (Saito	
   et	
   al.	
   1984,	
   Ferguson	
   et	
   al.	
   1988,	
  Saito	
  et	
  al.	
  1988).	
  Evidently	
  then,	
  Ca2+-­‐binding	
  is	
  sufficient,	
  but	
  not	
  necessary	
  to	
   activate	
  RyR1.	
   	
   	
    1.1.1	
  Calcium	
  Induced	
  Calcium	
  Release	
  and	
  E-­C	
  Coupling	
   	
   As	
   stated,	
   RyR	
   is	
   both	
   triggered	
   by	
   and	
   conducts	
   Ca2+	
   in	
   a	
   process	
   known	
   as	
   Calcium	
   Induced	
   Calcium	
   Release	
   (CICR)	
   (Endo	
   et	
   al.	
   1970,	
   Fabiato,	
   1983).	
   This	
   seems	
  to	
  set	
  up	
  a	
  positive	
  feedback	
  loop	
  that	
  would	
  completely	
  deplete	
  the	
  SR	
  of	
  Ca2+,	
   but	
  Meissner	
  et	
  al.	
  (1986)	
  have	
  shown	
  that	
  higher	
  concentrations	
  of	
  Ca2+	
  decrease	
  the	
   open	
  probability	
  of	
  RyR.	
  This	
  means	
  that	
  there	
  must	
  be	
  at	
  least	
  two	
  binding	
  sites	
  for	
   Ca2+	
   in	
   the	
   RyR,	
   one	
   for	
   activation	
   and	
   one	
   for	
   inhibition.	
   In	
   addition,	
   they	
   showed	
   that	
   another	
   divalent	
   cation,	
   Mg2+,	
   binds	
   and	
   inhibits	
   the	
   channel	
   by	
   two	
   proposed	
   mechanisms:	
  direct	
  competition	
  with	
  Ca2+	
  at	
  the	
  activation	
  site,	
  or	
  by	
  binding	
  to	
  a	
  low	
   affinity	
   inhibitory	
   site	
   that	
   can	
   accept	
   either	
   Ca2+	
   or	
   Mg2+.	
   Consequently	
   RyR	
   is	
   activated	
   by	
   micromolar	
   (µM)	
   concentrations	
   of	
   Ca2+	
   and	
   inhibited	
   by	
   millimolar	
   (mM)	
   concentrations	
   of	
   Ca2+	
   or	
   Mg2+	
   (Laver	
   et	
   al.	
   1997).	
   The	
   subsequent	
   amplification	
   of	
   the	
   Ca2+	
   signal	
   upon	
   activation	
   is	
   a	
   necessary	
   requirement	
   for	
   all	
   muscle	
  contraction.	
    	
    2	
    60	
   years	
   ago,	
   Sandow	
   (1952)	
   described	
   the	
   relationship	
   between	
   an	
   excitation	
   signal	
   and	
   the	
   consequent	
   activation	
   of	
   muscular	
   contraction.	
   It	
   occurs	
   in	
   a	
   process	
   now	
   known	
   as	
   excitation-­‐contraction	
   (E-­‐C)	
   coupling,	
   which	
   heavily	
   relies	
   on	
   Ca2+.	
   Already	
   mentioned,	
   it	
   was	
   later	
   understood	
   that	
   this	
   process	
   involved	
   a	
   physical	
   CaV1.1-­‐RyR1	
   link	
   near	
   transverse	
   tubules	
   (T-­‐tubules),	
   invaginations	
   in	
   the	
   sarcolemma	
   or	
   plasma	
   membrane	
   of	
   muscle	
   cells,	
   and	
   the	
   terminal	
   cisternae	
   of	
   SR	
   (Rios	
  and	
  Brum	
  1987,	
  Block	
  et	
  al.	
  1988,	
  Tanabe	
  et	
  al.	
  1988,	
  Tanabe	
  et	
  al.	
  1990).	
  	
   According	
   to	
   the	
   sliding	
   filament	
   model	
   of	
   muscle	
   contraction	
   (Huxley	
   and	
   Niedergerke,	
   1954,	
   Huxley	
   and	
   Hanson,	
   1954),	
   shortening	
   of	
   sarcomeres	
   occurs	
   by	
   interactions	
   between	
   thin	
   actin	
   and	
   thick	
   myosin	
   filaments.	
   Melzer	
   et	
   al.	
   (1995)	
   describe	
   the	
   role	
   of	
   Ca2+	
   in	
   binding	
   to	
   troponin	
   C,	
   located	
   on	
   actin	
   filaments.	
   This	
   causes	
  the	
  troponin	
  complex	
  to	
  pull	
  tropomyosin	
  away	
  from	
  the	
  myosin	
  binding	
  sites	
   on	
   these	
   actin	
   filaments.	
   Once	
   free	
   to	
   bind,	
   myosin	
   heads	
   can	
   attach	
   and	
   cause	
   sarcomere	
  shortening	
  by	
  pulling	
  actin	
  filaments	
  and	
  induce	
  contraction	
  of	
  the	
  muscle	
   cell	
   as	
   a	
   whole.	
   Following	
   this,	
   binding	
   of	
   adenosine	
   triphosphate	
   (ATP)	
   to	
   myosin	
   returns	
  the	
  heads	
  to	
  their	
  resting	
  state,	
  primed	
  for	
  re-­‐attachment	
  to	
  actin	
  and	
  thereby	
   causing	
   further	
   sarcomere	
   contraction	
   as	
   long	
   as	
   Ca2+	
   is	
   still	
   present.	
   Sarcoplasmic/Endoplasmic	
   Reticulum	
   Calcium	
   ATPases	
   (SERCAs)	
   must	
   then	
   pump	
   Ca2+	
   back	
   into	
   the	
   SR	
   to	
   re-­‐establish	
   normal	
   concentrations	
   (Carafoli	
   1987,	
   MacLennan	
   et	
   al.	
   2002,	
   Dulhunty,	
   1992,	
   Stephenson	
   et	
   al.	
   1998).	
   This	
   costs	
   a	
   lot	
   of	
   energy	
   provided	
   by	
   ATP	
   hydrolysis.	
   Other	
   Ca2+	
   concentration	
   restorers	
   include	
   Plasma	
  Membrane	
  Ca2+	
   ATPases,	
  which	
  work	
  in	
  a	
  similar	
  way	
  but	
  pump	
  ions	
  out	
  of	
   the	
  cell,	
  and	
  Na+/Ca2+	
  exchangers	
  (NCX),	
  which	
  swap	
  three	
  sodium	
  ions	
  (Na+)	
  into	
  the	
   cell	
  for	
  every	
  Ca2+	
  that	
  they	
  take	
  out.	
   	
    Ca2+	
   is	
  also	
  involved	
  prior	
  to	
  T-­‐Tubule	
  excitation	
  by	
  an	
  action	
  potential,	
  at	
  the	
    neuromuscular	
   junction.	
   At	
   the	
   presynaptic	
   neuron	
   terminal,	
   Ca2+	
   causes	
   vesicles	
   filled	
   with	
   acetylcholine	
   (ACh)	
   to	
   fuse	
   with	
   the	
   cell	
   membrane	
   and	
   release	
   the	
   neurotransmitter	
   into	
   the	
   synaptic	
   cleft.	
   Receptors	
   on	
   the	
   postsynaptic	
   membrane	
   recognize	
   ACh	
   and	
   open	
   to	
   allow	
   the	
   influx	
   of	
   sodium	
   ions	
   (Na+));	
   this	
   initiates	
   the	
   action	
  potential.	
  Being	
  a	
  vital	
  part	
  of	
  both	
  the	
  excitation	
  and	
  contraction,	
  Ca2+	
   can	
  be	
   considered	
  to	
  be	
  the	
  major	
  force	
  behind	
  E-­‐C	
  coupling.	
  Its	
  cytoplasmic	
  concentration	
   	
    3	
    in	
   the	
   cell	
   is	
   controlled	
   by	
   a	
   variety	
   of	
   Ca2+-­‐conductors,	
   most	
   importantly	
   RyR	
   with	
   SERCA,	
  NCX	
  and	
  PMCA	
  (Figure	
  1).	
  	
   	
    	
    	
   2+  FIGURE 1. RyR, Ca and muscle contraction. 1) An action potential reaches the T2+  Tubules. 2) RyR recruits Ca as a voltage sensor and releases Ca from SR. Nearby RyRs V  2+  2+  respond by CICR to amplify the Ca signal (not shown). 3) Myoplasmic Ca causes sliding of 2+  actin filaments in the direction of the red arrows. 4) SERCA pumps use ATP to return Ca to the SR. 	
   	
    1.1.2	
  Calmodulin	
  and	
  Calmodulin	
  Binding	
  Domains	
    	
    	
   The	
   large	
   size	
   of	
   the	
   RyR,	
   with	
   ~4/5ths	
   of	
   its	
   volume	
   in	
   the	
   cytoplasm,	
   provides	
  a	
  huge	
  area	
  for	
  different,	
  smaller	
  proteins	
  to	
  bind.	
  The	
  channels	
  even	
  exhibit	
   corner-­‐to-­‐corner	
   interactions	
   with	
   themselves,	
   resulting	
   in	
   a	
   two-­‐dimensional	
   (2D)	
   RyR	
   ‘checkerboard’	
   (Saito	
   et	
   al.	
   1984,	
   Ferguson	
   et	
   al.	
   1988,	
   Saito	
   et	
   al.	
   1988).	
   This	
    	
    4	
    emphasizes	
   the	
   essential	
   requirement	
   for	
   Ca2+	
   concentrations	
   to	
   be	
   well	
   regulated.	
   Understandably	
   then,	
   a	
   plethora	
   of	
   protein	
   chess	
   pieces	
   have	
   been	
   shown	
   to	
   bind	
   these	
   RyR	
   chessboard	
   arrangements.	
   The	
   immunosuppressant	
   FK-­‐506	
   binding	
   proteins	
   (FKBP12	
   and	
   12.6)	
   attach	
   and	
   stabilize	
   the	
   channel	
   in	
   the	
   closed	
   state	
   (Brilliantes	
   et	
   al.	
   1994,	
   Ma	
   et	
   al.	
   1995,	
   Timerman	
   et	
   al.	
   1996).	
   Oligomers	
   of	
   calsequestrin	
   (CSQ),	
   which	
   can	
   hold	
   over	
   60	
   ions	
   per	
   monomer	
   (Park	
   et	
   al.	
   2004)	
   are	
   responsible	
   for	
   the	
   high	
   concentration	
   of	
   Ca2+	
   in	
   SR.	
   CSQ-­‐bound	
   Ca2+	
   is	
   released	
   in	
   communication	
  with	
  RyR1,	
  but	
  whether	
  it	
  activates	
  or	
  inhibits	
  the	
  channel	
  depends	
   on	
  which	
  CSQ	
  isoform	
  is	
  involved	
  (Beard	
  et	
  al.	
  2004).	
  Other	
  examples	
  include	
  the	
  SR	
   anchoring	
  proteins	
  triadin	
  and	
  junctin	
  (Beard	
  et	
  al.	
  2009)	
  and	
  more	
  relevant	
  for	
  this	
   discussion,	
  calmodulin	
  (CaM)	
  (Chen	
  and	
  MacLennan,	
  1994,	
  Yang	
  et	
  al.	
  1994,	
  Tripathy	
   et	
   al.	
   1995).	
   Together,	
   and	
   among	
   many	
   others,	
   they	
   form	
   the	
   core	
   of	
   a	
   RyR	
   macrocomplex	
  that	
  defines	
  the	
  release	
  of	
  Ca2+.	
   Using	
  lipid	
  bilayer	
  studies,	
  CaM	
  in	
  fact	
  was	
  the	
  first	
  protein	
  to	
  be	
  identified	
  as	
   an	
   interacting	
   partner	
   with	
   single	
   RyR	
   channels	
   (Smith	
   et	
   al.	
   1989).	
   It	
   binds	
   stoichiometrically,	
   with	
   four	
   CaMs	
   binding	
   one	
   tetramer	
   (Tripathy	
   et	
   al.	
   1995,	
   Moore	
   et	
   al.	
   1999,	
   Balshaw	
   et	
   al.	
   2001)	
   even	
   though	
   there	
   are	
   thought	
   to	
   be	
   multiple	
   Calmodulin	
  Binding	
  Domains	
  (CaMBDs)	
  (Zorzatto	
  et	
  al.	
  1990,	
  Menegazzi	
  et	
  al.	
  1994,	
   Chen	
   and	
   MacLennan	
   1995).	
   Cryo-­‐electron	
   Microscopy	
   (cryo-­‐EM)	
   and	
   Förster	
   Resonance	
   Energy	
   Transfer	
   (FRET)	
   experiments	
   further	
   confirmed	
   the	
   4:1	
   stoichiometry	
   (Wagenknecht	
   et	
   al.	
   1994;	
   Wagenknecht	
   et	
   al.	
   1997;	
   Samsó	
   and	
   Wagenknecht	
  2002).	
  It	
  is	
  clear	
  that	
  the	
  binding	
  of	
  CaM	
  to	
  RyR	
  is	
  far	
  from	
  simple	
  and	
   is	
  yet	
  to	
  be	
  fully	
  understood.	
  Keeping	
  with	
  the	
  theme,	
  CaM	
  was	
  also	
  part	
  of	
  the	
  first	
   RyR	
   crystal	
   structure	
   as	
   part	
   of	
   a	
   complex	
   with	
   residues	
   3614-­‐3643	
   in	
   RyR1	
   (Maximciuc	
  et	
  al.	
  2006).	
  The	
  structure	
  unearthed	
  an	
  unusual,	
  novel	
  mode	
  of	
  binding,	
   even	
  for	
  the	
  promiscuous	
  CaM.	
  For	
  the	
  first	
  time	
  CaM	
  was	
  seen	
  to	
  bind	
  two	
  anchoring	
   hydrophobic	
   residues	
   located	
   17	
   residues	
   apart	
   (inclusive)	
   along	
   the	
   primary	
   structure	
  of	
  RyR1	
  in	
  an	
  antiparallel	
  organization.	
   To	
  further	
  complicate	
  matters,	
  CaM	
  contains	
  two	
  structurally	
  similar	
  lobes,	
  but	
   functionally	
   quite	
   different,	
   each	
   of	
   which	
   can	
   bind	
   Ca2+	
   and	
   change	
   conformation.	
   Their	
   Ca2+-­‐dependent	
   structural	
   changes	
   can	
   have	
   dramatic	
   effects	
   on	
   CaM’s	
   	
    5	
    regulatory	
   effect	
   on	
   RyR.	
   Ca2+-­‐free	
   CaM	
   (Apo-­‐CaM)	
   for	
   example	
   has	
   been	
   shown	
   to	
   partially	
  activate	
  RyR1,	
  whereas	
  Ca2+-­‐bound	
  CaM	
  (Ca2+-­‐CaM)	
  has	
  an	
  inhibitory	
  effect	
   on	
   Ca2+	
   release	
   from	
   the	
   SR	
   (Ikemoto	
   et	
   al.	
   1995,	
   Buratti	
   et	
   al.	
   1995,	
   Tripathy	
   et	
   al.	
   1995,	
   Rodney	
   et	
   al.	
   2000).	
   In	
   addition,	
   Samsó	
   and	
   Wagenknecht	
   (2002)	
   showed	
   by	
   cryo-­‐EM	
   that	
   the	
   centres	
   of	
   bound	
   Ca2+-­‐CaM	
   and	
   Apo-­‐CaM	
   lie	
   ~33	
   Å	
   apart,	
   and	
   visualized	
  sites	
  that	
  were	
  definitely	
  distinct,	
  but	
  sterically	
  overlapping.	
  This	
  implied	
   that	
   both	
   binding	
   events	
   cannot	
   not	
   occur	
   at	
   the	
   same	
   time	
   and	
   instead	
   that	
   some	
   kind	
   of	
   lobe	
   shifting	
   must	
   occur.	
   This	
   evidence	
   then	
   suggests	
   that	
   CaM	
   must	
   play	
   a	
   role	
  in	
  sensing	
  high	
  Ca2+	
   concentrations	
  achieved	
  after	
  signal	
  amplification	
  by	
  RyR1,	
   and	
  in	
  turn	
  shift	
  to	
  cause	
  the	
  channel	
  to	
  close.	
   From	
  available	
  crystal	
  structures,	
  we	
  see	
  that	
  its	
  ability	
  to	
  bind	
  Ca2+	
   is	
  due	
  to	
   four	
  EF-­‐hands,	
  two	
  on	
  both	
  lobes,	
  each	
  of	
  which	
  complex	
  a	
  single	
  Ca2+	
  ion	
  resulting	
  in	
   four	
   Ca2+	
   per	
   CaM	
   (Figure	
   2)	
   (Babu	
   et	
   al.	
   1988,	
   Chattopadhyaya	
   et	
   al.	
   1992,	
   Wilson	
   and	
   Brunger	
   2000).	
   EF-­‐hand	
   domains,	
   consisting	
   of	
   a	
   helix-­‐loop-­‐helix	
   motif,	
   are	
   commonly	
  associated	
  with	
  divalent	
  ions	
  such	
  as	
  Ca2+	
  or	
  Mg2+.	
  S100A,	
  another	
  EF-­‐hand	
   protein,	
  also	
  binds	
  Ca2+,	
  and	
  in	
  addition,	
  can	
  also	
  bind	
  the	
  RyR1	
  3614-­‐3643	
  peptide	
   (Wright	
   et	
   al.	
   2008),	
   suggesting	
   that	
   different	
   EF-­‐hand	
   containing	
   proteins	
   can	
   compete	
  for	
  the	
  same	
  binding	
  sites.	
  Given	
  that	
  RyR	
  itself	
  binds	
  Ca2+,	
  it	
  is	
  reasonable	
  to	
   assume	
  it	
  may	
  contain	
  its	
  own	
  EF-­‐hand	
  domains	
  that	
  could	
  presumably	
  bind	
  CaMBDs	
   as	
  intrinsic	
  ligands.	
  In	
  three-­‐dimensional	
  (3D)	
  space,	
  RyR1	
  EF-­‐hands	
  could	
  be	
  located	
   close	
   to	
   CaMBDs	
   in	
   the	
   full-­‐length	
   channel.	
   This	
   would	
   put	
   their	
   local,	
   physiological	
   concentrations	
   beyond	
   competitive	
   capabilities.	
   It	
   is	
   therefore	
   possible	
   that	
   in	
   vivo,	
   neither	
  CaM,	
  nor	
  S100A	
  bind	
  the	
  RyR1	
  3614-­‐3643	
  peptide.	
   	
    	
    6	
    	
    FIGURE 2. Calmodulin and EF-hands. (A) CaM (PDB 1CLL, Chattopadhyaya et al. 1992) is coloured in a spectrum from its N- (blue) to C- (red) terminus. Ca ions are shown 2+  as grey spheres. (B) A zoomed in view on the C-terminal helix-loop-helix EF-hand. Highlighted in black are the side-chains of some negatively charged residues involved in 2+  binding Ca . 	
   Xiong	
  et	
  al.	
  (2006)	
  added	
  serious	
  strength	
  to	
  this	
  argument	
  with	
  their	
  studies	
   on	
  RyR1	
  4064-­‐4210.	
  Firstly,	
  they	
  used	
  circular	
  dichroism	
  to	
  show	
  that	
  it	
  has	
  α-­‐helical	
   propensity,	
   hinting	
   to	
   the	
   possibility	
   of	
   an	
   EF-­‐hand.	
   Using	
   equilibrium	
   dialysis	
   with	
   45Ca2+,	
  they	
  then	
  determined	
  that	
  two	
  Ca2+	
  ions	
  bind	
  the	
  peptide	
  co-­‐operatively	
  with	
    an	
   affinity	
   of	
   ~60	
   µM	
   and	
   a	
   Hill	
   coefficient	
   of	
   ~1.6.	
   Furthermore,	
   the	
   fluorescent	
   probe	
   8-­‐anilino-­‐1-­‐naphthalenesulfonic	
   acid	
   ammonium	
   (ANSA),	
   which	
   changes	
   its	
    	
    7	
    fluorescent	
   properties	
   in	
   the	
   company	
   of	
   hydrophobic	
   residues,	
   was	
   used	
   to	
   show	
   binding	
   to	
   a	
   CaMBD	
   binding	
   site	
   in	
   RyR1	
   3614-­‐3643.	
   Here	
   the	
   affinity	
   was	
   Ca2+-­‐ dependent,	
  ranging	
  from	
  ~800	
  nM	
  in	
  apo	
  conditions	
  to	
  ~180	
  nM	
  in	
  the	
  presence	
  of	
  5	
   mM	
   Ca2+.	
   Lastly,	
   they	
   demonstrated	
   that	
   Histidine-­‐tagged	
   RyR1	
   4064-­‐4210	
   pulled	
   down	
   CaV1.1.	
   Their	
   results	
   hint	
   towards	
   the	
   identification	
   of	
   single	
   RyR1	
   peptide	
   as	
   a	
   Ca2+-­‐binding	
  site,	
  a	
  CaMBD	
  binding	
  site	
  and	
  the	
  physical	
  link	
  between	
  RyR1	
  and	
  CaV.	
   Care	
  must	
  be	
  taken	
  though,	
  not	
  to	
  interpret	
  too	
  much	
  from	
  experiments	
  on	
  isolated	
   peptides	
   that	
   physiologically	
   could	
   be	
   inaccessible	
   to	
   any	
   or	
   all	
   of	
   the	
   above.	
   Removing	
   such	
   peptides	
   from	
   their	
   folded	
   domains	
   exposes	
   hydrophobic	
   residues	
   that	
  could	
  result	
  in	
  a	
  general	
  ‘stickiness’.	
  Interestingly,	
  following	
  lysis,	
  their	
  construct	
   was	
   found	
   almost	
   exclusively	
   in	
   inclusion	
   bodies	
   and	
   had	
   to	
   be	
   refolded	
   in	
   purification.	
  	
   	
    In	
   addition,	
   as	
   discussed	
   in	
   Chapter	
   1.3,	
   hundreds	
   of	
   mutations	
   in	
   RyR1	
   can	
    cause	
   Malignant	
   Hyperthermia	
   (MH).	
   Volatile	
   anaesthetics,	
   for	
   example	
   halothane,	
   trigger	
   the	
   disorder	
   and	
   presumably	
   bind	
   RyR1.	
   Given	
   that	
   CaM	
   has	
   been	
   shown	
   to	
   bind	
  halothane	
  (Streiff	
  et	
  al.	
  2004),	
  and	
  NMR	
  structures	
  have	
  visualized	
  both	
  the	
  N-­‐	
   and	
   C-­‐lobe	
  bound	
  in	
  solution	
  (PDB	
  2KUG	
  and	
  2KUH,	
  Juranic	
  et	
  al.	
  unpublished),	
  it	
  yet	
   again	
   hints	
   that	
   there	
   may	
   not	
   be	
   any	
   discrimination	
   between	
   EF-­‐hands	
   in	
   binding.	
   It	
   would	
   therefore	
   not	
   be	
   too	
   far-­‐fetched	
   to	
   conclude	
   that	
   RyR1	
   4064-­‐4210	
   may	
   also	
   interact	
  with	
  the	
  anaesthetic.	
  At	
  present,	
  a	
  binding	
  site	
  for	
  anaesthetics	
  in	
  RyR1	
  is	
  yet	
   to	
  be	
  established.	
   	
   	
   	
    	
   	
   	
   	
    	
    8	
    1.2 Structural	
  Information	
   	
   	
    1.2.1	
  Cryo-­Electron	
  Microscopy	
    	
    	
   Historically,	
   structural	
   studies	
   on	
   RyRs	
   have	
   been	
   dominated	
   by	
   electron	
   microscopy	
   (EM)	
   with	
   their	
   first	
   documented	
   observation	
   in	
   thin-­‐layer	
   or	
   negative	
   stain	
  EM	
  images,	
  as	
  simple	
  protrusions	
  from	
  SR,	
  occurring	
  half	
  a	
  century	
  ago	
  (Revel	
   1962).	
  With	
  the	
  progress	
  in	
  cryo-­‐EM	
  technology,	
  structures	
  of	
  the	
  full	
  RyR1	
  channel	
   both	
   in	
   a	
   closed	
   and	
   open	
   state	
   have	
   been	
   made	
   available	
   in	
   the	
   Electron	
   Microscopy	
   Data	
  Bank	
  (EMDB)	
  at	
  resolutions	
  up	
  to	
  ~10	
  Å	
  determined	
  by	
  fourier	
  shell	
  correlation	
   at	
  cut-­‐offs	
  of	
  either	
  0.5	
  or	
  0.143	
  (Serysheva	
  et	
  al.	
  2005,	
  Ludtke	
  et	
  al.	
  2005,	
  Samsó	
  et	
   al.	
  2005,	
  Samsó	
  et	
  al.	
  2009).	
  These	
  reconstructions	
  have	
  extensively	
  been	
  studied	
  and	
   as	
   a	
   result,	
   several	
   globular	
   domains	
   within	
   the	
   channel	
   have	
   been	
   identified.	
   The	
   different	
  cytoplasmic	
  domains	
  or	
  ‘subregions’	
  have	
  been	
  assigned	
  arbitrary	
  numbers	
   (Figure	
   3)	
   (Serysheva	
   et	
   al.	
   2008).	
   Previously	
   lower	
   resolution	
   images	
   had	
   only	
   allowed	
   larger,	
   more	
   general	
   regions	
   to	
   be	
   named:	
   the	
   ‘central	
   rim’,	
   ‘handles’	
   and	
   ‘clamps’	
  (Serysheva	
  et	
  al.	
  1995).	
   	
    Comparison	
   between	
   the	
   open	
   and	
   closed	
   states	
   (Samsó	
   et	
   al.	
   2009)	
   of	
   the	
    channel	
   has	
   also	
   shed	
   some	
   light	
   on	
   the	
   allosteric	
   properties	
   of	
   RyRs	
   (Figure	
   4).	
   Besides	
   the	
   obvious	
   widening	
   of	
   the	
   central	
   rim	
   in	
   opening,	
   there	
   are	
   clear	
   movements	
  in	
  the	
  handles	
  and	
  clamps,	
  some	
  of	
  which	
  occur	
  more	
  than	
  100	
  Å	
  away	
   from	
  the	
  pore.	
  It	
  is	
  then	
  useful	
  to	
  portray	
  the	
  channel	
  as	
  a	
  set	
  of	
  globular	
  gears	
  that	
   can	
   communicate	
   in	
   both	
   directions;	
   opening	
   of	
   the	
   pore	
   can	
   cause	
   structural	
   rearrangements	
  near	
  the	
  periphery	
  of	
  receptor,	
  and	
  conversely,	
  binding	
  of	
  a	
  ligand	
  in	
   a	
  clamp	
  could	
  cause	
  opening	
  or	
  closing	
  at	
  the	
  pore.	
    	
    9	
    FIGURE 3. A 9.6 Å cryo-EM reconstruction of RyR1. Top (above) and side (below) views of the EMDB 1275 map (Ludtke et al. 2005). Subregions are coloured and numbered (Serysheva et al. 2008). The central rim, a handle and a clamp are also labelled for clarity. (Serysheva et al. 1995). 4/5ths of the channel resides in the cytoplasmic portion of the ‘mushroom’-like receptor. As labelled the diameter of the channel measures ~270 Å.  	
    10	
    FIGURE 4. Closed vs open states of RyR. Comparison of the closed and open cryoEM reconstructions of RyR1 from Samsó et al. 2009. The closed state (EMDB 1606) is shown as a blue mesh on top of the green surface of the open state (EMDB 1607). The cutoff selection for both maps are set to the same value and chosen so that a clear channel is observed through the open state.	
   	
   	
   High	
   quality	
   cryo-­‐EM	
   maps	
   have	
   been	
   very	
   useful	
   in	
   looking	
   at	
   the	
   complete	
   ~2.2	
   MDa	
   structure	
   but	
   pin-­‐pointing	
   the	
   location	
   of	
   individual	
   domains	
   from	
   the	
   primary	
   sequence,	
   has	
   proven	
   to	
   be	
   a	
   little	
   more	
   challenging.	
   To	
   overcome	
   this	
    	
    11	
    barrier,	
  several	
  insertion	
  studies	
  and	
  difference	
  maps	
  with	
  larger	
  protein	
  ligands	
  or	
   antibodies	
  were	
  carried	
  out.	
  Kimlicka	
  and	
  Van	
  Petegem	
  (2011)	
  have	
  reviewed	
  these	
   studies	
  and	
  mapped	
  out	
  each	
  position	
  onto	
  a	
  single	
  reconstruction	
  of	
  RyR1.	
  In	
  each	
   case	
   though,	
   they	
   stress	
   that	
   the	
   exact	
   location	
   of	
   the	
   ligand,	
   insertion	
   or	
   antibody	
   may	
   carry	
   a	
   significant	
   degree	
   of	
   uncertainty	
   due	
   to	
   a	
   combination	
   of	
   factors.	
   Besides	
   the	
   resolution	
   of	
   the	
   EM	
   map,	
   the	
   length	
   of	
   linkers	
   and	
   the	
   size	
   of	
   the	
   insert	
   or	
   ligand	
   must	
  also	
  be	
  taken	
  into	
  account.	
  Longer	
  linkers	
  could	
  result	
  in	
  global	
  differences	
  very	
   far	
  from	
  the	
  true	
  location,	
  and	
  smaller	
  ligands	
  would	
  be	
  difficult	
  to	
  distinguish	
  from	
   noise.	
   Still,	
   the	
   results	
   have	
   proven	
   to	
   be	
   very	
   helpful.	
   The	
   supposed	
   binding	
   location	
   of	
   the	
   drug	
   dantrolene	
   (see	
   Chapter	
   1.4.1)	
   for	
   example,	
   which	
   resides	
   in	
   RyR1	
   residues	
  590-­‐609	
  (Paul-­‐Pletzer	
  et	
  al.	
  2002,	
  2005),	
  could	
  be	
  attributed	
  to	
  subregion	
  9	
   due	
  to	
  cryo-­‐EM	
  images	
  of	
  RyR2	
  in	
  which	
  GFP	
  was	
  inserted	
  at	
  tyrosine	
  846	
  (Wang	
  et	
   al.	
   2011).	
   However,	
   the	
   insertion	
   point	
   is	
   over	
   200	
   residues	
   away	
   from	
   the	
   presumed	
   binding	
  site	
  and	
  the	
  observed	
  mass	
  difference	
  could	
  therefore	
  be	
  far	
  from	
  the	
  point	
  of	
   interest.	
   But	
   assuming	
   an	
   average	
   protein	
   domain	
   to	
   be	
   ~200	
   residues,	
   they	
   could	
   argue	
   this	
   would	
   put	
   the	
   GFP	
   no	
   more	
   than	
   single	
   subregion	
   away.	
   Additionally,	
   an	
   insertion	
  into	
  arginine	
  626,	
  which	
  could	
  not	
  be	
  expressed	
  well	
  enough	
  for	
  cryo-­‐EM,	
   but	
   was	
   usable	
   for	
   further	
   FRET	
   studies,	
   confirmed	
   that	
   the	
   distant	
   side	
   (from	
   the	
   pore)	
  of	
  subregion	
  9	
  might	
  indeed	
  house	
  the	
  dantrolene	
  binding	
  site	
  of	
  RyRs.	
   	
   	
    1.2.2	
  Crystallographic	
  Insight	
   	
   Cryo-­‐EM	
  is	
  without	
  doubt	
  a	
  powerful	
  tool	
  in	
  determining	
  the	
  general	
  structure	
   of	
  RyRs,	
  but	
  falls	
  short	
  when	
  it	
  comes	
  to	
  understanding	
  some	
  of	
  the	
  molecular	
  details	
   involved.	
   The	
   search	
   for	
   the	
   subregion	
   that	
   belonged	
   to	
   the	
   N-­‐terminus	
   in	
   the	
   full-­‐ length	
  channel	
  highlighted	
  the	
  limitations	
  of	
  using	
  difference	
  imaging	
  in	
  cryo-­‐EM	
  to	
   unambiguously	
   locate	
   domains.	
   Glutathione	
   S-­‐Transferase	
   (GST)	
   fused	
   at	
   the	
   N-­‐ terminus	
  of	
  RyR3	
  (Liu	
  et	
  al.	
  2001),	
  a	
  RyR1	
  416-­‐434	
  sequence	
  specific	
  antibody	
  (Baker	
   et	
   al.	
   2002)	
   and	
   a	
   GFP	
   insertion	
   at	
   serine	
   437	
   (Wang	
   et	
   al.	
   2007)	
   all	
   showed	
   major	
    	
    12	
    difference	
  peaks	
  in	
  the	
  clamp	
  regions	
  of	
  the	
  RyR.	
  This	
  corroborating	
  evidence	
  led	
  to	
  a	
   common	
   belief	
   in	
   the	
   field	
   that	
   the	
   N-­‐terminus	
   was	
   localized	
   between	
   subregions	
   5	
   and	
   9.	
   It	
   was	
   not	
   until	
   the	
   crystal	
   structure	
   of	
   the	
   first	
   3	
   domains	
   of	
   RyR1	
   in	
   residues	
   1-­‐532	
   (RyR1ABC)	
   was	
   solved	
   that	
   it	
   became	
   clear	
   where	
   the	
   N-­‐terminus	
   really	
   is	
   (Tung	
  et	
  al.	
  2010).	
  Here	
  we	
  performed	
  extensive	
  docking	
  of	
  RyR1ABC	
  using	
  Laplacian	
   filters,	
  an	
  application	
  that	
  has	
  been	
  shown	
  to	
  be	
  vital	
  in	
  the	
  docking	
  of	
  relatively	
  small	
   crystal	
   structures	
   into	
   large	
   EM	
   maps	
   (Wriggers	
   and	
   Chacón,	
   2001,	
   Chacón	
   and	
   Wriggers,	
   2002).	
   In	
   the	
   study	
   we	
   demonstrated	
   that	
   RyR1ABC	
   actually	
   forms	
   the	
   central	
   rim	
   on	
   the	
   cytoplasmic	
   surface	
   of	
   RyR	
   (Figure	
   5).	
   Following	
   the	
   results,	
   we	
   stressed	
   the	
   need	
   for	
   publishing	
   the	
   docking	
   contrast	
   between	
   the	
   top	
   solution	
   and	
   the	
   remaining	
   ones.	
   A	
   high	
   docking	
   contrast	
   would	
   eliminate	
   false	
   positives	
   and	
   in	
   turn	
  mean	
  a	
  high	
  probability	
  of	
  correctness.	
  	
   In	
   the	
   study	
   of	
   large	
   membrane	
   proteins	
   like	
   RyR,	
   crystallography	
   though,	
   is	
   definitely	
  not	
  a	
  replacement	
  for,	
  or	
  an	
  upgrade	
  from	
  cryo-­‐EM.	
  With	
  the	
  full	
  channel	
   having	
   so	
   far	
   evaded	
   crystallographic	
   success,	
   viewing	
   atomic	
   detail	
   in	
   ‘the	
   bigger	
   picture’	
   can	
   only	
   be	
   achieved	
   in	
   combining	
   the	
   two	
   techniques,	
   done	
   by	
   docking	
   crystal	
   structures	
   into	
   EM	
   maps.	
   Yuchi	
   et	
   al	
   (2012)	
   highlight	
   an	
   example	
   of	
   this	
   in	
   their	
  mapping	
  of	
  phosphorylation	
  domains	
  into	
  RyRs.	
   Full	
   length	
   RyRs	
   initially	
   intrigued	
   the	
   crystallographic	
   world	
   with	
   their	
   observed	
  formation	
  of	
  2D	
  ‘chessboard’	
  lattices	
  (Saito	
  et	
  al.	
  1984,	
  Ferguson	
  et	
  al.	
  1988,	
   Saito	
   et	
   al.	
   1988).	
   This	
   innate	
   ability	
   of	
   the	
   receptors	
   to	
   network	
   like	
   this	
   meant	
   crystals	
   were	
   easier	
   than	
   thought	
   to	
   obtain.	
   The	
   failure	
   to	
   form	
   3D	
   lattices	
   though,	
   crippled	
   the	
   progress	
   of	
   determining	
   a	
   full-­‐length	
   structure	
   for	
   several	
   years.	
   Thankfully,	
  high	
  resolution	
  EM	
  images	
  of	
  RyR,	
  showed	
  that	
  as	
  previously	
  mentioned,	
   the	
  complete	
  structure	
  can	
  be	
  broken	
  down	
  into	
  much	
  smaller,	
  globular	
  subregions,	
   each	
  of	
  which	
  provide	
  a	
  much	
  more	
  feasible	
  crystallographic	
  target.	
  	
   	
    	
    13	
    FIGURE 5. Docking of RyR1ABC. Docked (EMDB 1275, Ludtke et al. 2005) RyR1 A (blue) B (green) and C (red) are shown from the cytoplasmic top-side (A) and from a perpendicular cross-section at the dashed line (B). Arrows show the potential for allosteric interactions in the direction of the pore. Distant regions of the EM map (mesh) have been clipped for clarity. 	
    14	
    Three	
   years	
   ago,	
   the	
   structures	
   of	
   the	
   N-­‐terminal	
   ~200	
   residues	
   of	
   RyR1	
   (Amador	
   et	
   al.	
   2009,	
   Lobo	
   and	
   Van	
   Petegem	
   2009)	
   as	
   well	
   as	
   RyR2	
   (Lobo	
   and	
   Van	
   Petegem	
   2009)	
   were	
   solved	
   (RyR1A	
   and	
   RyR2A	
   respectively).	
   Since	
   then	
   several	
   wild-­‐type	
   large	
   domains	
   have	
   been	
   solved	
   and	
   docked	
   including	
   the	
   confusingly	
   named	
  Ryanodine	
  Receptor	
  (RYR)	
  domain	
  (Sharma	
  et	
  al.	
  2012),	
  the	
  aforementioned	
   RyR1ABC	
   domains	
   (Tung	
   et	
   al.	
   2010)	
   and	
   the	
   phosphorylation	
   domains	
   in	
   all	
   three	
   isoforms	
  (Yuchi	
  et	
  al.	
  2012,	
  Sharma	
  et	
  al.	
  2012).	
  Mutations	
  in	
  some	
  of	
  these	
  domains	
   have	
  been	
  shown	
  to	
  cause	
  severe	
  disorders,	
  and	
  crystal	
  structures	
  of	
  some	
  of	
  these	
   mutants	
   have	
   enlightened	
   some	
   of	
   the	
   mechanistic	
   effects	
   (discussed	
   in	
   Chapter	
   1.3.2).	
  Part	
  of	
  the	
  work	
  described	
  in	
  this	
  thesis	
  aims	
  to	
  increase	
  the	
  number	
  of	
  high-­‐ resolution	
  structures	
  of	
  RyR	
  domains.	
   	
   	
    1.2.3	
  The	
  IP3	
  Receptor	
   	
   RyRs	
   are	
   not	
   the	
   only	
   Ca2+	
   release	
   channels	
   located	
   in	
   the	
   ER	
   or	
   SR.	
   Another	
   powerful	
  second	
  messenger,	
  inositol	
  1,4,5-­‐triphosphate	
  (IP3),	
  and	
  its	
  receptor	
  (IP3R)	
   have	
   been	
   studied	
   extensively	
   as	
   well.	
   The	
   IP3	
   molecule	
   is	
   the	
   result	
   of	
   cleavage	
   of	
   phosphatidylinositol	
   4,5-­‐bisphosphate	
   (PIP2)	
   by	
   phospholipase	
   C.	
   An	
   additional	
   by-­‐ product	
   of	
   the	
   separation	
   is	
   1,2-­‐diacylglycerol	
   (DAG),	
   which	
   is	
   another	
   second	
   messenger	
   that	
   activates	
   Protein	
   Kinase	
   C.	
   DAG,	
   like	
   its	
   precursor	
   PIP2,	
   is	
   membrane-­‐ bound	
  whereas	
  IP3	
   is	
  water-­‐soluble	
  and	
  diffuses	
  in	
  the	
  cytoplasm	
  to	
  trigger	
  opening	
   of	
  IP3Rs.	
   IP3Rs	
   also	
   have	
   three	
   isoforms	
   in	
   vertebrates.	
   Most	
   cell-­‐types	
   express	
   more	
   than	
   one	
   of	
   these	
   isoforms	
   but	
   IP3R1	
   is	
   enriched	
   in	
   neurons	
   and	
   smooth	
   muscle,	
   IP3R2	
   in	
   the	
   brain,	
   heart	
   and	
   secretory	
   organs,	
   and	
   IP3R3	
   is	
   also	
   found	
   in	
   secretory	
   organs,	
  but	
  is	
  more	
  ubiquitously	
  expressed	
  (Newton	
  et	
  al.	
  1994,	
  De	
  Smedt	
  et	
  al.	
  1994,	
   Wojcikiewicz,	
  1995).	
  At	
  ~	
  1	
  MDa,	
  it	
  is	
  less	
  than	
  half	
  the	
  size	
  of	
  RyR,	
  due	
  mostly	
  to	
  its	
   much	
  smaller	
  cytoplasmic	
  portion.	
  Similar	
  to	
  RyR	
  though,	
  it	
  was	
  thought	
  to	
  have	
  the	
   same	
  ‘thumb-­‐tack’	
  structure	
  with	
  four-­‐fold,	
  tetrameric	
  symmetry.	
  Ludtke	
  et	
  al	
  (2011)	
    	
    15	
    provided	
  a	
  9.5	
  Å	
  cryo-­‐EM	
  structure	
  that	
  confirmed	
  this.	
  The	
  crystal	
  structure	
  of	
  the	
   N-­‐terminal	
  three	
  domains	
  of	
  IP3R	
  was	
  initially	
  solved	
  in	
  two	
  segments:	
  the	
  latter	
  two	
   forming	
  an	
  IP3	
  binding	
  domain	
  (Bosanac	
  et	
  al.	
  2002)	
  and	
  the	
  former	
  domain,	
  which	
   suppressed	
   IP3	
   binding	
   (Serysheva	
   et	
   al.	
   2003,	
   Bosanac	
   et	
   al.	
   2005).	
   Later,	
   structures	
   of	
   the	
   first	
   three	
   domains	
   together	
   (IP3R1ABC)	
   were	
   solved	
   (Lin	
   et	
   al.	
   2011,	
   Seo	
   et	
   al.	
   2012),	
  but	
  to	
  lower	
  resolutions.	
  	
   A	
   detailed	
   comparison	
   of	
   RyR1ABC	
   to	
   the	
   corresponding	
   domains	
   in	
   IP3R1	
   shows	
  that	
  their	
  overall	
  structures	
  are	
  conserved,	
  suggesting	
  that	
  the	
  channels	
  must	
   rely	
  on	
  similar	
  allosteric	
  mechanisms	
  in	
  regulation	
  (Yuchi	
  and	
  Van	
  Petegem,	
  2011).	
  In	
   both	
   cases,	
   a	
   pair	
   of	
   beta-­‐trefoil	
   domains	
   followed	
   by	
   a	
   third	
   helical	
   domain	
   can	
   describe	
  the	
  structure	
  (Figure	
  6).	
  However,	
  two	
  notable	
  differences	
  can	
  be	
  observed.	
   The	
   first	
   domain	
   of	
   IP3R1	
   (IP3R1A)	
   contains	
   two	
   large	
   helices	
   in	
   comparison	
   with	
   RyR1A	
  that	
  has	
  only	
  one,	
  and	
  the	
  construct	
  length	
  of	
  IP3R1ABC,	
  which	
  results	
  in	
  a	
  35-­‐ residue	
   extension	
   in	
   the	
   third,	
   helical	
   domain.	
   Despite	
   these	
   differences,	
   Seo	
   et	
   al.	
   (2012)	
   used	
   some	
   intricate	
   chimera	
   studies	
   to	
   show	
   that	
   RyR1A	
   can	
   functionally	
   replace	
  the	
  suppressor	
  domain	
  in	
  IP3R1.	
    	
    16	
    FIGURE 6. IP3R1ABC vs RYR1ABC. IP3R1 domain A (yellow), B (orange) and C (purple) are individually superposed onto RyR1ABC (black). Dashed circles highlight the major differences between the two: a large extra helix in IP3RA, and an extended domain C. 	
   	
   	
    1.3 Disease	
  Mutations	
   	
   Ca2+	
   regulation	
  in	
  the	
  cell	
  is	
  vital.	
  With	
  E-­‐C	
  coupling	
  relying	
  so	
  heavily	
  on	
  Ca2+	
   as	
   a	
   second	
   messenger,	
   any	
   disruption	
   in	
   this	
   regulation	
   causes	
   severe	
   muscular	
   pathologies.	
   Mutations	
   in	
   both	
   RyR1	
   and	
   RyR2	
   can	
   cause	
   acute	
   genetic	
   diseases.	
   Kimlicka	
   et	
   al.	
   (unpublished)	
   have	
   maintained	
   a	
   RyR	
   mutation	
   database	
   that	
   	
    17	
    currently	
   shows	
   over	
   290	
   mutations	
   in	
   RyR1	
   cause	
   skeletal	
   muscle	
   disorders	
   such	
   as	
   MH	
   or	
   Central	
   Core	
   Disease	
   (CCD),	
   and	
   over	
   150	
   mutations	
   cause	
   diseases	
   such	
   as	
   Catecholaminergic	
   Polymorphic	
   Ventricular	
   Tachycardia	
   (CPVT)	
   or	
   Arrhythmogenic	
   Right	
   Ventricular	
   Dysplasia	
   Type	
   2	
   (ARVD2)	
   in	
   RyR2.	
   In	
   both	
   isoforms,	
   disease	
   mutations	
   seem	
   to	
   be	
   clustered	
   into	
   ‘hotspots’	
   in	
   the	
   channel	
   that	
   seem	
   distant	
   in	
   primary	
  structure,	
  but	
  may	
  interact	
  in	
  3D	
  space	
  according	
  to	
  the	
  ‘zipper	
  hypothesis’	
   (Ikemoto	
   and	
   Yamamoto,	
   2002,	
   Oda	
   et	
   al.	
   2005,	
   Liu	
   et	
   al.	
   2010).	
   For	
   this	
   discussion,	
   I	
   will	
  concentrate	
  on	
  MH	
  and	
  CCD.	
   MH	
   is	
   a	
   life-­‐threatening	
   genetic	
   disorder	
   triggered	
   by	
   volatile	
   anaesthetics.	
   Some	
   of	
   the	
   symptoms	
   include	
   metabolic	
   acidosis,	
   tachycardia,	
   skeletal	
   muscle	
   rigidity,	
  rhabdomyolysis	
  and	
  most	
  notably,	
  a	
  dramatic	
  rise	
  in	
  body	
  core	
  temperature	
   (Pamukcoglu	
  1988,	
  Denborough	
  1998).	
  Its	
  record	
  as	
  a	
  human	
  disorder	
  began	
  in	
  1960	
   with	
  a	
  twenty	
  one	
  year	
  old	
  man	
  who	
  required	
  surgery	
  after	
  being	
  run	
  over	
  by	
  a	
  car,	
   and	
   his	
   accident-­‐prone	
   family.	
   With	
   10	
   of	
   the	
   man’s	
   24	
   relatives	
   thought	
   to	
   have	
   died	
   in	
  response	
  to	
  ether-­‐based	
  anaesthetics,	
  there	
  was	
  more	
  concern	
  on	
  how	
  to	
  proceed	
   anaesthetically	
   than	
   there	
   was	
   about	
   how	
   to	
   deal	
   with	
   the	
   compound	
   fractures	
   in	
   his	
   leg.	
   They	
   decided	
   to	
   abandon	
   ethers	
   and	
   proceed	
   cautiously	
   with	
   the	
   recently	
   available	
   halothane,	
   but	
   within	
   twenty	
   minutes	
   this	
   had	
   to	
   be	
   stopped	
   when	
   both	
   the	
   man’s	
   body	
   temperature	
   and	
   heart	
   rate	
   rose	
   rapidly.	
   In	
   somewhat	
   of	
   a	
   miracle,	
   not	
   only	
  did	
  they	
  manage	
  to	
  save	
  the	
  man’s	
  life	
  using	
  ice	
  packs	
  to	
  cool	
  him	
  down,	
  but	
  also	
   successfully	
   completed	
   the	
   required	
   surgery.	
   Still,	
   Denborough	
   and	
   Lovell	
   (1960)	
   actually	
  published	
  the	
  article	
  ‘Anaesthetic	
  Deaths	
  in	
  a	
  family’	
  with	
  the	
  hope	
  of	
  finding	
   others	
  that	
  had	
  come	
  across	
  the	
  clinical	
  problem.	
  Eventually,	
  by	
  1966	
  enough	
  cases	
   and	
  interest	
  evolved	
  that	
  a	
  symposium	
  in	
  Toronto	
  was	
  held	
  that	
  was	
  dedicated	
  solely	
   to	
  MH	
  (Gordon,	
  1996).	
   Mutation	
   in	
   RyR1	
   became	
   a	
   candidate	
   for	
   blame	
   when	
   pigs	
   with	
   a	
   single	
   cysteine	
   to	
   arginine	
   mutation	
   at	
   position	
   615	
   (C615R)	
   in	
   RyR1	
   exhibited	
   the	
   symptoms	
   of	
   MH	
   (Fujii	
   et	
   al.	
   1991).	
   Since	
   then	
   a	
   large	
   number	
   of	
   MH-­‐causing	
   mutations	
  in	
  human	
  RyR1	
  have	
  been	
  found	
  (McCarthy	
  et	
  al.	
  2000,	
  Girard	
  et	
  al.	
  2006,	
   Robinson	
   et	
   al.	
   2006,	
   Groom	
   et	
   al.	
   2011).	
   Thankfully,	
   in	
   the	
   late	
   1970s,	
   the	
   muscle	
   relaxant	
   dantrolene	
   was	
   clinically	
   approved	
   to	
   treat	
   MH	
   and	
   successfully	
   reduced	
   the	
   	
    18	
    mortality	
   rate	
   of	
   MH	
   from	
   ~80%	
   to	
   less	
   than	
   5%	
   (Rosenberg	
   et	
   al.	
   2007).	
   The	
   mechanistic	
  details	
  of	
  the	
  treatment	
  have	
  not	
  fully	
  been	
  understood	
  although	
  RyR1	
  is	
   considered	
  to	
  be	
  the	
  major	
  molecular	
  target	
  (Parness	
  and	
  Palnitkar	
  1995,	
  Palnitkar	
  et	
   al.	
   1999,	
   Krause	
   et	
   al.	
   2004).	
   Jiang	
   et	
   al.	
   (2008)	
   popularized	
   a	
   theory	
   involving	
   the	
   misregulated	
   sensing	
   of	
   luminal	
   Ca2+	
   concentration.	
   RyRs	
   were	
   shown	
   to	
   recognize	
   when	
  the	
  SR	
  Ca2+	
  store	
  is	
  beyond	
  its	
  capacity	
  and	
  to	
  trigger	
  Store	
  Overload-­‐Induced	
   Ca2+	
  release	
  (SOICR).	
  Physiological	
  Ca2+	
  concentrations	
  inside	
  the	
  SR	
  rarely	
  get	
  high	
   enough	
  to	
  cause	
  SOICR,	
  but	
  mutations	
  in	
  RyR	
  result	
  in	
  leaky	
  channels	
  that	
  lower	
  this	
   threshold.	
  	
   CCD	
   is	
   a	
   rare	
   congenital	
   myopathy	
   associated	
   with	
   reduced	
   oxidative	
   activity	
   in	
   muscle	
   fibre.	
   Muscle	
   weakness	
   causes	
   an	
   inability	
   to	
   independently	
   walk,	
   and	
   with	
   no	
   curative	
   treatment	
   current	
   therapeutic	
   management	
   is	
   limited	
   to	
   physiotherapy	
   (Jungbluth	
  et	
  al.	
  2002).	
  With	
  regards	
  to	
  RyR1,	
  CCD	
  can	
  be	
  explained	
  in	
  two	
  separate	
   ways.	
  The	
  first	
  runs	
  parallel	
  with	
  MH	
  and	
  involves	
  a	
  gain-­‐of-­‐function	
  in	
  RyR1.	
  Here	
  SR	
   Ca2+	
   stores	
   get	
   depleted	
   by	
   constitutively	
   active	
   channels.	
   As	
   a	
   result,	
   there	
   is	
   an	
   inability	
   to	
   sufficiently	
   amplify	
   Ca2+	
   signals	
   required	
   for	
   muscle	
   contraction	
   (Zhang	
   et	
   al.	
   1993,	
   Dirksen	
   and	
   Avila,	
   2002).	
   The	
   second	
   involves	
   a	
   loss	
   of	
   function	
   in	
   the	
   coupling	
   between	
   RyR1	
   and	
   CaV.	
   ‘Uncoupled’	
   receptors	
   would	
   then	
   cause	
   muscle	
   weakness	
  by	
  decreasing	
  RyR1-­‐mediated	
  SR	
  Ca2+	
   release	
  (Quane	
  et	
  al.	
  1993,	
  Dirksen	
   and	
  Avila,	
  2002).	
  Which	
  of	
  the	
  two	
  mechanisms	
  best	
  describes	
  any	
  particular	
  case	
  of	
   CCD	
  depends	
  on	
  the	
  location	
  of	
  the	
  mutation.	
   	
   	
    1.3.1	
  Phosphorylation	
  and	
  disease	
   	
   The	
  underlying	
  cause	
  of	
  MH	
  has	
  been	
  attributed	
  to	
  a	
  gain-­‐of-­‐function	
  in	
  RyR1	
   channels,	
   which	
   become	
   leaky	
   and	
   as	
   a	
   result	
   cause	
   irregular	
   Ca2+	
   homeostasis	
   (Tong	
   et	
   al.	
   1999).	
   The	
   symptoms	
   of	
   disease	
   have	
   been	
   attributed	
   to	
   a	
   variety	
   of	
   cellular	
   processes	
  that	
  occur	
  to	
  restore	
  proper	
  Ca2+	
  levels.	
  For	
  example,	
  in	
  an	
  attempt	
  to	
  re-­‐ establish	
   normal	
   Ca2+	
   concentrations,	
   the	
   ATP-­‐consuming	
   SERCA	
   pump	
   overworks,	
    	
    19	
    thereby	
   generating	
   an	
   increase	
   in	
   oxidative/nitrosative	
   stress.	
   As	
   part	
   of	
   a	
   brutal	
   positive	
  feedback	
  loop,	
  this	
  stress	
  causes	
  S-­‐Nitrosylation	
  of	
  RyR1	
  that	
  makes	
  it	
  even	
   more	
   leaky	
   (Durham	
   et	
   al.	
   2008).	
   Others	
   have	
   shown	
   a	
   link	
   between	
   hyperphosphorylation	
  of	
  RyR2	
  in	
  the	
  heart	
  and	
  SR	
  Ca2+	
   leak	
  (Wehrens	
  et	
  al.	
  2003,	
  Ai	
   et	
  al.	
  2005).	
  It	
  is	
  thought	
  that	
  this	
  is	
  due	
  to	
  the	
  phosphorylation	
  causing	
  detachment	
   of	
  the	
  stabilizing	
  FKBP	
  proteins	
  (Marx	
  et	
  al.	
  2000,	
  Reiken	
  2003),	
  but	
  this	
  mechanism	
   is	
  yet	
  to	
  be	
  fully	
  accepted.	
  A	
  similar	
  link	
  has	
  been	
  used	
  to	
  explain	
  Ca2+	
  leak	
  in	
  RyR1	
   (Bellinger	
   et	
   al.	
   2009).	
   The	
   perpetrators	
   of	
   RyR	
   phosphorylation	
   are	
   Protein	
   kinase	
   A	
   (PKA),	
  (Mayrleitner	
  et	
  al.	
  1995)	
  and	
  Ca2+/CaM-­‐dependent	
  kinase	
  II	
  (CaMKII)	
  (Wang	
   and	
   Best,	
   1992).	
   There	
   is	
   some	
   debate	
   in	
   the	
   field	
   though	
   as	
   to	
   which	
   sites	
   in	
   the	
   channel	
  are	
  phosphorylated.	
  Most	
  of	
  the	
  conflict	
  involves	
  RyR2,	
  and	
  ultimately	
  there	
   has	
   been	
   a	
   shift	
   in	
   focus	
   away	
   from	
   the	
   notion	
   of	
   one	
   particular	
   site	
   being	
   hyperphosphorylated	
  in	
  an	
  RyR	
  cluster.	
  Instead	
  the	
  idea	
  of	
  multiple	
  phosphorylation	
   sites	
   in	
   a	
   single	
   receptor	
   has	
   been	
   included	
   (Huttlin	
   et	
   al.	
   2010,	
   Yuchi	
   et	
   al.	
   2012).	
   With	
   regard	
   to	
   RyR1,	
   so	
   far	
   serine-­‐2843	
   (S2843)	
   has	
   been	
   identified	
   as	
   a	
   PKA	
   and	
   CaMKII	
  phosphorylation	
  site	
  (Suko	
  et	
  al.	
  1993).	
  	
   From	
   a	
   reverse	
   perspective,	
   phosphatases	
   play	
   their	
   role	
   in	
   the	
   RyR	
   macromolecular	
  complex	
  as	
  phosphate	
  removers,	
  thereby	
  reducing	
  the	
  activity	
  of	
  the	
   channels.	
   In	
   RyR2,	
   heart	
   failure	
   associated	
   with	
   PKA	
   hyperphosphorylation	
   was	
   shown	
  to	
  coincide	
  with	
  a	
  drop	
  in	
  abundance	
  of	
  protein	
  phosphatases	
  1	
  and	
  2A	
  (PP1	
   and	
  PP2A)(Marx	
  et	
  al.	
  2000).	
  	
  Interestingly	
  PKA,	
  as	
  well	
  as	
  PP1	
  and	
  PP2A	
  are	
  thought	
   to	
   use	
   the	
   same	
   mode	
   of	
   attachment	
   via	
   leucine	
   zippers	
   in	
   regulatory	
   subunits	
   to	
   tether	
  to	
  RyR	
  (Marx	
  et	
  al.	
  2001).	
  	
   	
   	
    1.3.2	
  The	
  Molecular	
  Cause	
  of	
  Disease	
   	
   The	
  molecular	
  details	
  that	
  actually	
  cause	
  RyR	
  leakiness	
  can	
  be	
  understood	
  in	
   analysis	
   of	
   the	
   available	
   crystal	
   structures	
   and	
   their	
   docking	
   to	
   cryo-­‐EM	
   maps.	
   So	
   far,	
   the	
  Van	
  Petegem	
  lab	
  has	
  been	
  responsible	
  for	
  all	
  published	
  mutant	
  crystal	
  structures.	
    	
    20	
    With	
   regard	
   to	
   point	
   mutations,	
   the	
   molecular	
   effects	
   of	
   two	
   disease	
   mutations	
   in	
   RyR2A	
   (Lobo	
   and	
   Van	
   Petegem,	
   2009)	
   and	
   three	
   disease	
   mutations	
   in	
   the	
   RyR1	
   phosphorylation	
   domain	
   (Yuchi	
   et	
   al.	
   2012)	
   have	
   been	
   revealed.	
   Looking	
   at	
   the	
   primary	
  structures	
  alone,	
  in	
  each	
  case,	
  all	
  that	
  was	
  involved	
  was	
  the	
  replacement	
  of	
  a	
   single	
   amino	
   acid.	
   The	
   little	
   structural	
   changes	
   observed	
   as	
   a	
   result	
   though,	
   are	
   drastic	
   enough	
   to	
   cause	
   the	
   aforementioned,	
   severe	
   genetic	
   disorders.	
   In	
   a	
   particularly	
   interesting	
   example	
   though,	
   the	
   deletion	
   of	
   an	
   entire	
   exon	
   in	
   RyR2A	
   is	
   involved.	
   The	
   removal	
   takes	
   away	
   some	
   major	
   wild-­‐type	
   structural	
   elements,	
   including	
   a	
   vital	
   beta	
   strand	
   that	
   is	
   then	
   replaced	
   by	
   a	
   neighbouring	
   stretch	
   of	
   residues.	
  Remarkably,	
  these	
  residues	
  are	
  completely	
  unstructured	
  in	
  a	
  normal	
  wild-­‐ type	
   channel	
   and	
   seem	
   to	
   be	
   able	
   to	
   mould	
   into	
   place.	
   People	
   with	
   this	
   disease	
   mutation	
  have	
  a	
  much	
  worse	
  version	
  of	
  CPVT	
  (Lobo	
  et	
  al.	
  2011).	
  	
   Most	
   of	
   the	
   molecular	
   insight	
   into	
   disease	
   in	
   RyR	
   came	
   from	
   analysing	
   the	
   mutation	
  sites	
  in	
  RyR1ABC,	
  which	
  are	
  all	
  clustered	
  at	
  interfaces	
  between	
  domains	
  in	
   the	
  full-­‐length	
  channel	
  (Tung	
  et	
  al.	
  2010).	
  Evidently,	
  RyRs	
  must	
  function	
  as	
  complex	
   allosteric	
   gear	
   systems	
   and	
   mutations	
   that	
   blunt	
   the	
   teeth	
   of	
   these	
   gears	
   result	
   in	
   leaky	
   channels.	
   This	
   breakdown	
   of	
   communication	
   between	
   subregions	
   in	
   RyR1	
   added	
   some	
   weight	
   to	
   the	
   zipper	
   hypothesis,	
   which	
   as	
   mentioned,	
   theorized	
   that	
   disease	
  hotspots	
  interact	
  physically	
  with	
  each	
  other	
  (Ikemoto	
  and	
  Yamamoto,	
  2002,	
   Oda	
  et	
  al.	
  2005,	
  Liu	
  et	
  al.	
  2010).	
  However,	
  most	
  disease	
  mutations	
  in	
  the	
  N-­‐terminal	
   disease	
   hot	
   spots	
   are	
   already	
   involved	
   in	
   contacts	
   with	
   other	
   domains	
   within	
   the	
   same	
   hot	
   spot,	
   so	
   only	
   a	
   small	
   subset	
   of	
   mutations	
   remain	
   available	
   for	
   interacting	
   with	
   the	
   central	
   disease	
   hot	
   spot.	
   As	
   such,	
   the	
   original	
   zipper	
   hypothesis	
   can	
   only	
   apply	
  to	
  a	
  small	
  subset	
  of	
  mutations.	
   	
    1.4	
   Small	
  Molecule	
  Drugs	
  and	
  Ligands	
   	
   	
    In	
   addition	
   to	
   the	
   numerous	
   proteins	
   that	
   interact	
   with	
   RyR,	
   various	
   small	
    molecules	
  modulate	
  the	
  channel	
  as	
  well.	
  Ca2+	
   and	
  Mg2+	
   have	
  already	
  been	
  discussed,	
   but	
   other	
   physiological	
   ligands	
   include	
   natural	
   toxins	
   that	
   can	
   either	
   activate	
   RyR,	
    	
    21	
    such	
  as	
  imperatoxin	
  A	
  in	
  scorpion	
  venom	
  (Tripathy	
  et	
  al.	
  1998,	
  Gurrola	
  et	
  al.	
  1999),	
   or	
   cause	
   inhibition,	
   for	
   example	
   natrin	
   in	
   snake	
   venom	
   (Thomas	
   et	
   al.	
   2010).	
   [3H]-­‐ Ryanodine	
   should	
   be	
   mentioned	
   in	
   addition	
   to	
   ryanodine	
   as	
   a	
   ligand,	
   as	
   its	
   production	
  provided	
  a	
  new	
  approach	
  to	
  studying	
  RyR	
  structure	
  and	
  function.	
  Binding	
   to	
   ryanodine	
   occurs	
   only	
   for	
   open	
   channels,	
   and	
   thus	
   the	
   amount	
   of	
   radiolabelled	
   ligand	
   bound	
   represents	
   a	
   measure	
   of	
   the	
   open	
   probability	
   of	
   the	
   RyRs.	
   (Pessah	
   et	
   al.	
   1985).	
   In	
   an	
   example,	
   the	
   immunosuppressants	
   rapamycin	
   (Kaftan	
   et	
   al.	
   1996)	
   and	
   FK-­‐506	
   (Kraus-­‐Friedmann	
   and	
   Feng,	
   1994)	
   have	
   been	
   shown	
   to	
   reduce	
   ryanodine	
   binding.	
   Single-­‐channel	
   experiments	
   on	
   RyR	
   showed	
   that	
   the	
   anticoagulant	
   heparin	
   increased	
   the	
   PO	
   of	
   the	
   channel	
   (Bezprozvanny	
   et	
   al.	
   1993).	
   4-­‐chloro-­‐m-­‐cresol	
   (Hermann-­‐Frank	
  et	
  al.	
  1996)	
  and	
  volatile	
  anesthetics	
  such	
  as	
  halothane	
  (Connelly	
  and	
   Coronado,	
  1994,	
  Bull	
  and	
  Marengo,	
  1994)	
  are	
  examples	
  of	
  non-­‐physiological	
  ligands	
   that	
  activated	
  RyR	
  in	
  bilayer	
  studies.	
  In	
  contrast,	
  the	
  polycationic	
  dye	
  ruthenium	
  red	
   inhibits	
  the	
  binding	
  of	
  ryanodine	
  (Pessah	
  et	
  al.	
  1985).	
  Relevant	
  to	
  this	
  discussion	
  are	
   the	
  drug	
  dantrolene,	
  and	
  the	
  purine	
  derivatives:	
  adenine	
  nucleotides	
  and	
  caffeine.	
   	
   	
    1.4.1	
  Dantrolene	
   	
   Dantrolene	
   is	
   a	
   hydantoin	
   derivative	
   (Figure	
   7A)	
   originally	
   synthesized	
   as	
   a	
   muscle	
  relaxant	
  (Snyder	
  et	
  al.	
  1967),	
  used	
  to	
  treat	
  skeletal	
  muscle	
  spasticity	
  (Dykes,	
   1975).	
   Its	
   muscle	
   relaxant	
   properties	
   were	
   pinned	
   on	
   the	
   depression	
   of	
   E-­‐C	
   coupling.	
   With	
  the	
  number	
  of	
  fatal	
  MH	
  cases	
  rising,	
  this	
  led	
  to	
  animal	
  model	
  testing	
  using	
  MH-­‐ susceptible	
  pigs	
  (Harrison,	
  1975).	
  The	
  remarkable	
  results	
  jumpstarted	
  the	
  drug	
  into	
   clinical	
   trials	
   just	
   two	
   years	
   later	
   that	
   led	
   to	
   an	
   approved	
   treatment	
   for	
   MH	
   shortly	
   after	
  (Kolb	
  et	
  al.	
  1982).	
  Initially	
  there	
  were	
  problems	
  in	
  administering	
  the	
  drug	
  due	
  to	
   its	
  inability	
  to	
  dissolve	
  well	
  in	
  water,	
  but	
  doses	
  today	
  are	
  added	
  to	
  mannitol,	
  which	
   significantly	
   improves	
   its	
   solubility	
   (Krause	
   et	
   al.	
   2004).	
   A	
   more	
   water-­‐soluble	
   alternative	
  is	
  azumolene	
  (Figure	
  7B)	
  (Sudo	
  et	
  al.	
  2008).	
   	
    	
    22	
    FIGURE 7. The molecular structures of dantrolene and azumolene. (A) Dantrolene is mostly planar and very poorly soluble in water. (B) Azumolene is a more water-soluble analog and differs by the presence of a bromide group in place of the nitro group in dantrolene. 	
   The	
  drug	
  was	
  shown	
  to	
  inhibit	
  SR	
  Ca2+	
  release	
  at	
  a	
  concentration	
  of	
  ~10-­‐50	
  µM	
   (Van	
   Winkle,	
   1976,	
   Herbette	
   et	
   al.	
   1982,	
   Loke	
   and	
   MacLennan,	
   1998),	
   and	
   as	
   the	
   therapeutic	
   concentration	
   is	
   in	
   the	
   same	
   range	
   (Flewellen	
   et	
   al.	
   1983),	
   the	
   clinical	
   effects	
   are	
   most	
   likely	
   due	
   to	
   this	
   inhibition.	
   But	
   with	
   regard	
   to	
   RyR,	
   a	
   clear	
   mechanistic	
  story	
  is	
  yet	
  to	
  be	
  achieved.	
  The	
  most	
  recent	
  theories	
  deal	
  with	
  SOICR	
  in	
   both	
  RyR1	
  and	
  RyR2	
  (Jiang	
  et	
  al.	
  2004,	
  Jiang	
  et	
  al.	
  2005,	
  Jiang	
  et	
  al.	
  2008,	
  Maxwell	
  et	
   al.	
   2011).	
   Dantrolene	
   is	
   suspected	
   to	
   remedy	
   the	
   inability	
   of	
   mutant	
   RyRs	
   to	
   withstand	
   a	
   proper	
   SR	
   luminal	
   Ca2+	
   concentration.	
   Additionally,	
   Kobayashi	
   et	
   al.	
   (2005)	
  used	
  a	
  fluorescent	
  probe	
  to	
  track	
  RyR1	
  domain	
  movements	
  upon	
  mimicking	
    	
    23	
    MH	
   and	
   noticed	
   that	
   dantrolene	
   prevented	
   them	
   –	
   apparently	
   stabilizing	
   the	
   interdomain	
   interactions.	
   Whether	
   or	
   not	
   these	
   effects	
   are	
   due	
   to	
   a	
   direct	
   interaction	
   with	
  RyR	
  though	
  is	
  debatable,	
  with	
  strong	
  arguments	
  both	
  for	
  and	
  against.	
   As	
  mentioned,	
  [3H]-­‐Ryanodine	
  binding	
  has	
  been	
  a	
  diagnostic	
  tool	
  to	
  recognize	
   an	
  open	
  RyR.	
  Palnitkar	
  et	
  al.	
  (1997)	
  showed	
  no	
  effect	
  of	
  dantrolene	
  on	
  [3H]-­‐ryanodine	
   binding,	
   and	
   that	
   in	
   linear	
   centrifugation	
   of	
   SR	
   membranes,	
   distinct	
   peaks	
   for	
   both	
   [3H]-­‐ryanodine	
   and	
   [3H]-­‐dantrolene	
   were	
   attained.	
   This	
   initially	
   led	
   to	
   the	
   conclusion	
   that	
   dantrolene	
   and	
   ryanodine	
   bind	
   two	
   different	
   molecules.	
   Later	
   though,	
   Zhao	
   et	
   al.	
   (2001)	
   described	
   that	
   [3H]-­‐ryanodine	
   binding	
   was	
   indeed	
   affected	
   by	
   10	
   µM	
   dantrolene.	
  Together	
  these	
  conflicting	
  results	
  hint	
  towards	
  configuration-­‐dependent	
   ryanodine	
   and	
   dantrolene	
   binding,	
   and	
   that	
   perhaps	
   open	
   and	
   closed	
   RyRs	
   fractionate	
  differently	
  in	
  centrifugation	
  due	
  to	
  changes	
  in	
  the	
  RyR	
  macrocomplex.	
  One	
   interesting	
   difference	
   between	
   the	
   two	
   studies	
   was	
   the	
   presence	
   of	
   an	
   adenine	
   nucleotide,	
   required	
   for	
   dantrolene-­‐inhibition	
   of	
   RyR	
   (Zhao	
   et	
   al.	
   2001),	
   suggesting	
   that	
  if	
  there	
  is	
  a	
  dantrolene	
  binding	
  site	
  in	
  RyR,	
  that	
  it	
  is	
  intimately	
  linked	
  to	
  a	
  purine	
   binding	
  site	
  by	
  some	
  allosteric	
  mechanism.	
   Single-­‐channel	
   experiments	
   only	
   provided	
   more	
   topics	
   of	
   discussion.	
   In	
   early	
   lipid	
  bilayer	
  studies,	
  Bull	
  and	
  Marengo	
  (1994)	
  stated	
  that	
  100	
  µM	
  dantrolene	
  or	
  less	
   failed	
   to	
   affect	
   halothane-­‐induced	
   activation	
   of	
   frog	
   skeletal	
   RyR,	
   evidence	
   that	
   was	
   corroborated	
  more	
  recently	
  in	
  recordings	
  using	
  20	
  µM	
  dantrolene	
  on	
  both	
  RyR1	
  and	
   RyR2	
  (Diaz-­‐Sylvester	
  et	
  al.	
  2008).	
  This	
  went	
  against	
  reports	
  that	
  showed	
  a	
  biphasic	
   pattern	
   of	
   regulation	
   with	
   activation	
   at	
   nanomolar	
   and	
   inhibition	
   at	
   micromolar	
   dantrolene	
   concentrations	
   (Nelson	
   et	
   al.	
   1996).	
   Both	
   sets	
   of	
   experiments	
   incorporated	
   SR	
   vesicles	
   into	
   lipid	
   bilayers	
   so	
   again,	
   differences	
   in	
   experimental	
   conditions	
   and	
   preparations	
   resulting	
   in	
   the	
   purification	
   of	
   distinct	
   RyR	
   macrocomplexes	
   could	
   explain	
   the	
   discrepancies.	
   Szentesi	
   et	
   al.	
   (2001)	
   purified	
   RyRs	
   alone	
   by	
   centrifugation	
   through	
   a	
   sucrose	
   gradient,	
   and	
   showed	
   that	
   in	
   lipid	
   bilayers,	
   they	
  were	
  not	
  affected	
  by	
  dantrolene.	
  This	
  convincing	
  evidence	
  though,	
  still	
  does	
  not	
   clarify	
  if	
  RyR	
  is	
  the	
  direct	
  molecular	
  target	
  or	
  not.	
  More	
  accurately,	
  it	
  proves	
  that	
  RyR	
   and	
  dantrolene	
  alone	
  do	
  not	
  result	
  in	
  channel	
  inhibition;	
  the	
  two	
  could	
  still	
  bind	
  and	
    	
    24	
    exhibit	
   a	
   latent	
   effect	
   on	
   Ca2+	
   release.	
   This	
   direct	
   mechanism	
   would	
   then	
   require	
   subsequent	
  change	
  in	
  RyR	
  structure,	
  due	
  to	
  the	
  binding	
  of	
  modulators.	
  Similarly,	
  an	
   indirect	
  mechanism	
  could	
  just	
  as	
  easily	
  be	
  described	
  in	
  which	
  one	
  of	
  the	
  modulators	
   is	
   in	
   fact	
   the	
   molecular	
   target	
   of	
   dantrolene.	
   Strengthening	
   the	
   latter	
   theory	
   are	
   observations	
   that	
   dantrolene	
   inhibits	
   the	
   binding	
   of	
   dihydropyridine	
   analogs	
   to	
   CaV1.1	
   more	
   effectively	
   than	
   the	
   binding	
   of	
   [3H]-­‐ryanodine	
   to	
   RyR1	
   (el-­‐Hayek	
   et	
   al.	
   1992).	
   In	
   general,	
   it	
   can	
   be	
   concluded	
   that	
   other	
   proteins	
   are	
   definitely	
   involved	
   in	
   the	
   mechanism	
   of	
   inhibition	
   by	
   dantrolene;	
   in	
   an	
   example,	
   it	
   was	
   shown	
   to	
   be	
   dependent	
  on	
  CaM	
  (Zhao	
  et	
  al.	
  2001,	
  Kobayashi	
  et	
  al.	
  2005).	
   The	
   most	
   compelling	
   argument	
   for	
   a	
   direct	
   mode	
   of	
   action	
   comes	
   from	
   photoaffinity	
   label	
   experiments	
   (Paul-­‐Pletzer	
   et	
   al.	
   2001,	
   Paul-­‐Pletzer	
   et	
   al.	
   2002)	
   using	
   [3H]-­‐azidodantrolene.	
   This	
   reactive	
   analog	
   of	
   the	
   drug	
   was	
   specifically	
   synthesized	
   to	
   locate	
   a	
   dantrolene	
   binding	
   site	
   in	
   skeletal	
   muscle	
   SR	
   in	
   the	
   hope	
   of	
   clearing	
  some	
  of	
  the	
  controversy	
  in	
  the	
  field	
  (Palnitkar	
  et	
  al.	
  1999).	
  Upon	
  activation,	
   [3H]-­‐azidodantrolene	
   covalently	
   bound	
   to	
   porcine	
   sarcoplasmic	
   reticulum,	
   and	
   after	
   purification	
   and	
   cleavage	
   by	
   n-­‐calpain,	
   the	
   N-­‐terminal	
   1400	
   RyR1	
   residues	
   were	
   credited	
   as	
   containing	
   the	
   drug	
   binding	
   site.	
   This	
   prompted	
   later	
   experiments	
   showing	
  that	
  it	
  selectively	
  attached	
  to	
  a	
  synthetic	
  peptide	
  of	
  RyR1	
  residues	
  590-­‐609.	
   Similar	
   to	
   the	
   experiments	
   with	
   RyR1	
   4064-­‐4120	
   mentioned	
   earlier	
   though,	
   results	
   from	
  experiments	
  using	
  isolated	
  peptides	
  have	
  to	
  be	
  interpreted	
  carefully.	
   	
    	
    	
    	
    1.4.2	
  Purine	
  derivatives	
   	
   These	
  are	
  heterocyclic,	
  aromatic	
  compounds	
  consisting	
  of	
  two	
  fused	
  rings:	
  one	
   imidazole	
   and	
   one	
   pyrimidine.	
   Even	
   in	
   early	
   RyR	
   research,	
   the	
   adenine	
   nucleotide	
   purines	
  in	
  particular	
  were	
  shown	
  to	
  be	
  important	
  activators	
  of	
  the	
  channel	
  (Morii	
  and	
   Tonomura,	
   1983;	
   Nagasaki	
   and	
   Kasai,	
   1983;	
   Meissner,	
   1984)	
   with	
   a	
   half	
   maximal	
   effective	
  concentration	
  (EC50)	
  in	
  the	
  millimolar	
  range	
  (Meissner	
  et	
  al.	
  1986).	
  Earlier	
   still,	
   it	
   was	
   known	
   that	
   the	
   methylxanthine	
   caffeine	
   also	
   favours	
   the	
   release	
   of	
   Ca2+	
    	
    25	
    from	
  the	
  SR	
  (Endo,	
  1977)	
  and	
  shows	
  a	
  similar	
  effect	
  on	
  RyR	
  as	
  adenine	
  nucleotides.	
   Given	
  their	
  closeness	
  in	
  structure	
  (Figure	
  8),	
  this	
  is	
  initially	
  not	
  surprising,	
  but	
  other	
   evidence	
   suggests	
   that	
   their	
   interaction	
   sites	
   may	
   not	
   be	
   the	
   same:	
   Ogawa	
   and	
   Harafuji	
   (1990)	
   for	
   example	
   showed	
   that	
   the	
   in	
   vitro	
   effect	
   of	
   combining	
   caffeine	
   and	
   adenine	
   nucleotides	
   on	
   Ca2+-­‐dependent	
   RyR1	
   stimulation,	
   was	
   additive	
   in	
   reaction	
   media	
   of	
   high	
   osmolarity,	
   whereas	
   it	
   was	
   potentiating	
   in	
   lower,	
   physiological	
   salt	
   concentrations.	
   	
    	
   FIGURE 8. Adenine nucleotides and xanthines. The molecular structures of purine, adenine, xanthine, and the methylxanthine caffeine are shown. The purine ring exists in all three structures, highlighting their stuctural similarity. 	
   	
    	
    26	
    Still,	
   two	
   cases	
   in	
   the	
   PDB	
   database	
   exist,	
   where	
   both	
   an	
   adenine	
   nucleotide	
   and	
  caffeine	
  have	
  been	
  separately	
  co-­‐crystallized	
  in	
  the	
  same	
  receptor:	
  1)	
  Adenosine	
   (Lebon	
   et	
   al.	
   2011)	
   and	
   caffeine	
   (Doré	
   et	
   al.	
   2011)	
   bound	
   to	
   the	
   adenosine	
   A2A	
   receptor,	
   and	
   2)	
   Adenine	
   (Lukacs	
   et	
   al.	
   2006)	
   and	
   caffeine	
   (Oikonomakos	
   et	
   al.	
   2000)	
   bound	
   to	
   a	
   glycogen	
   phosphorylase.	
   In	
   both	
   cases,	
   the	
   two	
   species	
   are	
   found	
   in	
   the	
   same	
  pocket	
  (Figure	
  9A	
  and	
  9B).	
  Close	
  analysis	
  of	
  the	
  interactions	
  show	
  that	
  all	
  that	
   may	
  be	
  required	
  is	
  a	
  hydrophobic	
  pocket,	
  ideally	
  containing	
  an	
  aromatic	
  ring	
  to	
  stack	
   against.	
   A	
   few	
   charged	
   and	
   polar	
   residues	
   may	
   supply	
   some	
   specificity.	
   In	
   fact,	
   all	
   published	
  structures	
  with	
  caffeine	
  bound	
  seem	
  to	
  take	
  advantage	
  of	
  an	
  aromatic	
  ring.	
   Figures	
  9C,	
  D	
  and	
  E	
  highlight	
  three	
  examples	
  (Ekstrom	
  et	
  al.	
  2002,	
  Yang	
  et	
  al.	
  2010,	
   Rao	
   et	
   al.	
   2005).	
   Interestingly,	
   the	
   structure	
   in	
   figure	
   9C	
   involves	
   a	
   glycogen	
   phosphorylase	
   as	
   well	
   and	
   although	
   it	
   superposes	
   well	
   with	
   the	
   structure	
   from	
   9B,	
   contains	
  an	
  extra	
  binding	
  site	
  for	
  caffeine.	
  Similarly,	
  Figures	
  9D	
  and	
  E	
  are	
  both	
  from	
   easily	
   superposable	
   chitinase	
   structures,	
   but	
   the	
   latter	
   contains	
   an	
   extra	
   binding	
   site.	
   This	
   kind	
   of	
   interaction	
   seems	
   very	
   general	
   and	
   hints	
   to	
   the	
   possibility	
   of	
   multiple	
   binding	
  sites	
  in	
  RyR.	
  	
   The	
   lack	
   of	
   specificity	
   in	
   binding	
   is	
   emphasized	
   further	
   when	
   weighing	
   the	
   effects	
  of	
  other	
  methylxanthines	
  against	
  each	
  other.	
  Rousseau	
  et	
  al.	
  (1988)	
  compared	
   seven	
   different	
   methylxanthines,	
   including	
   caffeine,	
   and	
   although	
   they	
   established	
   a	
   hierarchy	
   in	
   effectiveness,	
   the	
   differences	
   were	
   not	
   great.	
   They	
   did	
   however	
   show	
   that	
  the	
  five-­‐member	
  imidazole	
  ring,	
  which	
  is	
  also	
  present	
  in	
  adenine,	
  was	
  necessary,	
   as	
  1,3-­‐dimethyluracil	
  was	
  ineffective.	
  It	
  is	
  likely	
  then,	
  that	
  there	
  are	
  multiple	
  binding	
   sites	
  for	
  both	
  adenine	
  nucleotides	
  and	
  methylxanthines	
  in	
  RyR	
  and	
  that	
  some	
  of	
  these	
   pockets,	
  maybe	
  most,	
  can	
  accommodate	
  either.	
   	
    	
    27	
    	
   FIGURE 9. Purine binding by aromatic stacking. A) PDBs 2YDO (Lebon et al. 2011) and 3RFM (Doré et al. 2011) respectively show adenosine (white) and caffeine (grey) bound to the adenosine A receptor B) PDBs 1Z8D (Lukacs er al. 2000) and 1GFZ 2A  (Oikonomakos et al. 2006)respectively show adenine (white) and caffeine (grey) bound to a glycogen phosphorylase. C) PDB 1L5Q (Ekstrom et al. 2002) shows another glycogen phosphorylase with an extra caffeine molecule (*) bound in the crystal structure. D) PDB 3G6M (Yang et al. 2010) shows caffeine bound to a chitinase. E) PDB 2A3B (Rao et al. 2005) shows another chitinase with an extra caffeine molecule (*) bound in the crystal structure. Together, the structures hint towards aromatic promiscuity in caffeine. 	
   	
   	
    	
    28	
    1.5	
   Hypotheses	
  and	
  Goals	
   	
   	
    This	
   thesis	
   aims	
   to	
   uncover	
   some	
   of	
   the	
   structural	
   and	
   functional	
    characteristics	
  of	
  ligand	
  binding	
  in	
  RyR1,	
  findings	
  that	
  are	
  transferable	
  to	
  other	
  RyR	
   isoforms.	
   Given	
   the	
   size	
   of	
   the	
   channel	
   we	
   hypothesize	
   that	
   distant	
   allosteric	
   mechanisms	
   must	
   be	
   involved	
   in	
   most	
   ligand-­‐related	
   processes	
   including	
   general	
   regulation	
   and	
   drug	
   action.	
   To	
   test	
   this	
   theory	
   a	
   range	
   of	
   known	
   modulators,	
   both	
   physiological	
  and	
  pharmacological,	
  have	
  been	
  chosen:	
   	
   1. The	
   drug	
   dantrolene	
   and	
   the	
   protein	
   PP1,	
   both	
   of	
   which	
   negatively	
   regulate	
   RyR1,	
  have	
  been	
  linked	
  to	
  the	
  N-­‐terminal	
  ~600	
  residues	
  of	
  the	
  channel.	
  	
   2. Activators	
   like	
   caffeine	
   (pharmacological)	
   and	
   adenine	
   nucleotides	
   (physiological),	
  which	
  may	
  interact	
  with	
  RyR1	
  via	
  similar	
  mechanisms.	
   3. Ca2+	
   and	
   Mg2+,	
   which	
   are	
   thought	
   to	
   interact	
   with	
   RyR	
   EF-­‐hands,	
   domains	
   that	
   could	
  regulate	
  the	
  RyR	
  by	
  competing	
  with	
  CaM	
  for	
  CaMBDs.	
   	
   Specifically,	
   to	
   further	
   our	
   understanding	
   of	
   the	
   effects	
   of	
   modulators	
   on	
   RyR1,	
   the	
  techniques	
  of	
  X-­‐ray	
  crystallography,	
  ITC	
  and	
  in	
  silico	
  docking	
  and	
  modelling	
  are	
   implemented	
  (Figure	
  10).	
  	
    	
    29	
    	
   FIGURE 10. Method flowchart. A flow-chart describing the general procedures and aims of the projects in this thesis. The structural and thermodynamic properties of ligand binding in RyR1 are obtained from ITC and crystallographic studies. Information on large subregion movements in RyR1 is obtained from the docking studies. Together, the results are used to attempt deciphering the details of allosteric interactions between domains in RyR1 that occur as a result of ligand binding. All experiments were carried out on the RyR1  Oryctolagus cuniculus clone, a sequence that is >96% identical to RyR1 Homo sapiens and in general is shifted in register by +1 residues. 	
    	
    	
    30	
    2	
   Materials	
  and	
  Methods	
   	
   	
   	
    2.1	
  Cloning	
  and	
  Expression	
   	
    	
    The	
  program	
  PHYRE2	
  (Kelly	
  and	
  Stenberg,	
  2009)	
  was	
  used	
  to	
  aid	
  in	
  construct	
    selection.	
  A	
  modified	
  pET28	
  vector	
  was	
  used	
  for	
  all	
  cloning	
  (Van	
  Petegem	
  et	
  al.	
  2004).	
   A	
   full-­‐length	
   RyR1	
   clone	
   from	
   Oryctolagus	
   cuniculus	
   (rabbit)	
   provided	
   the	
   template	
   for	
   all	
   construct	
   selection	
   using	
   the	
   Polymerase	
   Chain	
   Reaction	
   (PCR).	
   Constructs	
   were	
   inserted	
   at	
   the	
   C-­‐terminus	
   of	
   hexahistidine	
   tagged	
   Maltose	
   Binding	
   Protein	
   (MBP).	
  In	
  resultant	
  fusion	
  proteins,	
  a	
  Tobacco	
  Etch	
  Virus	
  (TEV)	
  protease	
  cleavage	
  site	
   (ENLYFQSNA)	
  separates	
  MBP	
  and	
  the	
  construct	
  of	
  interest.	
  As	
  a	
  result	
  of	
  cleavage,	
  a	
   serine-­‐asparagine-­‐alanine	
   cloning	
   artefact	
   remained	
   at	
   the	
   N-­‐terminus	
   of	
   all	
   constructs.	
   Part	
   of	
   the	
   vector	
   design	
   allows	
   for	
   Ligation	
   Independent	
   (LIC)	
   cloning.	
   Here,	
   primers	
   for	
   PCR	
   of	
   constructs	
   of	
   interest	
   were	
   designed	
   to	
   include	
   a	
   complementary	
   LIC	
   sequence	
   to	
   the	
   one	
   in	
   the	
   vector	
   (see	
   Table	
   1).	
   Constructs	
   and	
   the	
   vector	
   were	
   treated	
   with	
   T4	
   deoxyribonucleic	
   acid	
   (DNA)	
   Polymerase	
   in	
   the	
   absence	
   of	
   all	
   but	
   one	
   deoxyribonucleotide	
   triphosphate	
   (dNTP).	
   Under	
   these	
   conditions	
  the	
  polymerase	
  exhibits	
  3’5’	
  exonuclease	
  activity	
  until	
  it	
  reaches	
  a	
  point	
   where	
  the	
  provided	
  dNTP	
  can	
  be	
  added.	
  In	
  this	
  way,	
  long	
  complementary	
  overhangs	
   (between	
  vector	
  and	
  constructs)	
  can	
  be	
  created	
  that	
  hydrogen	
  bond	
  well	
  enough	
  that	
   in	
   vitro	
   ligation	
   is	
   not	
   required.	
   Following	
   annealing	
   of	
   constructs	
   to	
   the	
   vector,	
   transformation	
  into	
  a	
  calcium	
  chloride	
  (CaCl2)	
  competent	
  DH5α	
  strain	
  of	
  Escherichia	
   coli	
  (E.	
  coli)	
  (Invitrogen)	
  was	
  performed	
  and	
  resultant	
  colonies	
  were	
  grown	
  in	
  2xYT	
   media	
   (16g/L	
   Tryptone,	
   10	
   g/L	
   yeast	
   extract	
   and	
   5	
   g/L	
   sodium	
   chloride)	
   for	
   DNA	
   amplification.	
   	
    The	
   RyR1	
   217-­‐536	
   W269A	
   and	
   4071-­‐4138	
   C4114A	
   mutations	
   were	
   made	
    using	
  the	
  QuikChange	
  protocol	
  (Stratagene).	
  The	
  process	
  involves	
  designing	
  primers	
    	
    31	
    that	
   flank	
   the	
   desired	
   mutation	
   site	
   and	
   code	
   for	
   an	
   alanine	
   instead	
   of	
   tryptophan.	
   Extension	
   around	
   the	
   plasmid	
   is	
   done	
   by	
   Pfu	
   Turbo	
   Polymerase	
   in	
   a	
   PCR	
   reaction.	
   The	
   restriction	
   enzyme	
   DpnI	
   was	
   used	
   to	
   cut	
   the	
   methylated	
   template	
   DNA	
   and	
   leave	
   a	
   mutated	
   plasmid	
   (albeit	
   with	
   nicks	
   in	
   the	
   DNA)	
   for	
   transformation	
   yet	
   again	
   into	
   DH5α	
  for	
  amplification.	
   	
    For	
   protein	
   expression	
   the	
   E.	
   coli	
   Rosetta	
   (DE3)	
   pLacI	
   cells	
   (Novagen)	
   were	
    used.	
  Cultures	
  were	
  grown	
  at	
  37°C	
  until	
  the	
  optical	
  density	
  at	
  600nm	
  (OD600)	
  reached	
   ~0.2-­‐0.3	
  and	
  then	
  the	
  temperature	
  was	
  lowered	
  to	
  18°C.	
  At	
  an	
  OD600	
  of	
  ~0.6,	
  cultures	
   were	
  induced	
  with	
  isopropyl-­‐β-­‐D-­‐thiogalactoside	
  (IPTG)	
  and	
  grown	
  for	
  a	
  further	
  ~24	
   hours.	
   Cultures	
   were	
   then	
   pelleted	
   and	
   stored	
   at	
   -­‐20°C.	
   In	
   the	
   case	
   of	
   EF-­‐hand	
   constructs,	
  the	
  temperature	
  was	
  kept	
  at	
  37°C	
  throughout	
  expression	
  and	
  cells	
  were	
   pelleted	
  after	
  ~4	
  hours	
  of	
  induction.	
   	
    RyR1	
   4295-­‐4325,	
   a	
   CaMBD	
   (Takeshima	
   et	
   al.	
   1989,	
   Chen	
   and	
   MacLennan	
    1994),	
   has	
   been	
   named	
   CaMBD3	
   by	
   the	
   Van	
   Petegem	
   lab.	
   A	
   synthesized	
   mutant	
   peptide	
  with	
  a	
  4320-­‐4322	
  LRR	
  duplication	
  (RyR1	
  CaMBD3dLRR)	
  was	
  available	
  in	
  the	
   lab	
  and	
  purchased	
  from	
  Lifetein	
   	
   	
   	
   	
   	
   	
   	
   	
   	
   	
   	
   	
   	
   	
    	
    32	
    Forward	
  primers	
   N1	
    TACTTCCAATCCAATGCAATGGGTGACGGAGGAGAGGG	
    N217	
    TACTTCCAATCCAATGCAGGCCACGTCCTCCGCCTC	
    N394	
    TACTTCCAATCCAATGCACAGCAGGAGGAGTCCCAGGC	
    N533	
    TACTTCCAATCCAATGCAAATCGTGCCAACTGTGCCCTC	
    N570	
    TACTTCCAATCCAATGCAGAGAGTCCCGAGGTGCTGAAC	
    N3487	
    TACTTCCAATCCAATGCATCGGGCGGCTCGGACCAGGAAC	
    N3497	
    TACTTCCAATCCAATGCAAAGAAGCGCCGGGGGGACAGG	
    N3507	
    TACTTCCAATCCAATGCACAGACGTCACTGATCGTGGCC	
    N3818	
    TACTTCCAATCCAATGCAGATTATCTGAAGGACAAGAAG	
    N3828	
    TACTTCCAATCCAATGCATTCTTCCAGAGCATCCAGGCG	
    N3838	
    TACTTCCAATCCAATGCAACATGCAGCGTCCTGGATCTC	
    N3997	
    TACTTCCAATCCAATGCAGCCCACATGATGATGAAGCTC	
    N4024	
    TACTTCCAATCCAATGCAGTGGTGATGCTGCTGTCCCTAC	
    N4031	
    TACTTCCAATCCAATGCACTGGAAGGGAACGTGGTGAAC	
    N4041	
    TACTTCCAATCCAATGCAGCCCGGCAGATGGTGGACATG	
    N4071	
    TACTTCCAATCCAATGCAATCGTGGGCTCCGAGGCCTTC	
    Reverse	
  primers	
   C536	
    TTATCCACTTCCAATGTTATTAGTTGGCACGATTGCCCCGGATC	
    C574	
    TTATCCACTTCCAATGTTATTACACCTCGGGACTCTCAATCAGG	
    C589	
    TTATCCACTTCCAATGTTATTAGAGGGAGATGATGGACTTG	
    C608	
    TTATCCACTTCCAATGTTATTACACACACAGGGAGCACAGC	
    C617	
    TTATCCACTTCCAATGTTATTAGTTGGAGCGCACAGCCACG	
    C627	
    TTATCCACTTCCAATGTTATTAAGGGAGCAGATTCTCGGTG	
    C635	
    TTATCCACTTCCAATGTTATTAGATGCTGGTGACGTAGTTG	
    C644	
    TTATCCACTTCCAATGTTATTAGATGCTGGTGACGTAGTTG	
    C654	
    TTATCCACTTCCAATGTTATTACTCGGCTCGGCCCACGAAG	
    C796	
    TTATCCACTTCCAATGTTATTAGCGGCCGCCAAGGAGAAACCG	
    C4070	
    TTATCCACTTCCAATGTTATTAGTCCTTGAGTTTCAGGAACATG	
    C4128	
    TTATCCACTTCCAATGTTATTAGAACTCCTCGAAGTTGATCATC	
    C4138	
    TTATCCACTTCCAATGTTATTAGTCCCGGGCTGGCTCCTGGAAG	
    C4148	
    TTATCCACTTCCAATGTTATTAGGTCAGCAGCACGGCCACGTTG	
    Quikchange	
  primers	
   W269A_F	
    GAACCCCTGAGAATCAGCGCGAGTGGAAGCCACCTGCGC	
    W269A_R	
    GCGCAGGTGGCTTCCACTCGCGCTGATTCTCAGGGGTTC	
    C4114A_F	
    GAAATCCAGTTTCTGCTCTCGGCCTCCGAAGCCGACGAGAATG	
    C4114A_R	
    CATTCTCGTCGGCTTCGGAGGCCGAGAGCAGAAACTGGATTTC	
    	
    Table 1. List of primers used to make constructs. All sequences are shown 5’3’. The LIC sequence for forward primers is: 5’-TACTTCCAATCCAATGCA-3’, whereas for reverse primers it is: 5’-TTATCCACTTCCAATGTTATTA-3’ which encodes two stop codons after the construct. Note that aside from Quikchange primers, in all cases an LIC sequence precedes a coding sequence. For Quikchange primers, highlighted in red are the codons that were changed and code for alanine. 	
    33	
    2.2	
  Protein	
  Purification	
   	
   	
    The	
  following	
  buffers	
  are	
  referred	
  to	
  in	
  the	
  purification	
  protocol:	
    	
    	
    Buffer	
  A:	
  250	
  mM	
  KCl,	
  10	
  mM	
  HEPES*	
  at	
  pH	
  7.4	
    	
    	
    Buffer	
  B:	
  250	
  mM	
  KCl,	
  500	
  mM	
  Imidazole	
  at	
  pH	
  7.4	
    	
    	
    Buffer	
  C:	
  250	
  mM	
  KCl,	
  10	
  mM	
  HEPES	
  at	
  pH	
  7.4,	
  10	
  mM	
  Maltose	
    	
    	
    Buffer	
  D(pH):	
  50	
  mM	
  KCl,	
  (10	
  mM	
  HEPES	
  pH	
  7.4	
  or	
  Tris**	
  at	
  pH	
  8.0)	
  	
    	
    	
    Buffer	
  E(pH):	
  1	
  M	
  KCl,	
  (10	
  mM	
  HEPES	
  pH	
  7.4	
  or	
  Tris	
  at	
  pH	
  8.0)	
    	
    	
    Buffer	
  F:	
  150	
  mM	
  KCl,	
  10	
  mM	
  HEPES	
  at	
  pH	
  7.4	
    *HEPES:	
  4-­‐(2-­‐hydroxyethyl)-­‐1-­‐piperazineethanesulfonic	
  acid	
    	
    **Tris:	
  tris(hydroxymethyl)aminomethane	
   	
   Frozen	
   pellets	
   were	
   thawed	
   in	
   a	
   room	
   temperature	
   water	
   bath	
   but	
   in	
   all	
   purification	
   steps	
   following,	
   care	
   was	
   taken	
   to	
   keep	
   samples	
   at	
   a	
   maximum	
   of	
   4°C.	
   Cells	
  were	
  then	
  lysed	
  by	
  sonication	
  in	
  ~60	
  mL	
  of	
  buffer	
  A,	
  containing	
  in	
  addition	
  25	
   μg/ml	
   of	
   deoxyribonuclease	
   I,	
   25	
   μg/ml	
   of	
   lysozyme,	
   1	
   mM	
   phenylmethylsulphonyl	
   fluoride,	
  10%	
  (vol/vol)	
  glycerol	
  and	
  14	
  mM	
  β-­‐mercaptoethanol	
  (βME).	
  In	
  the	
  case	
  of	
   RyR1	
   4071-­‐4138	
   C4114A,	
   the	
   addition	
   of	
   βME	
   was	
   omitted,	
   with	
   no	
   need	
   for	
   the	
   reducing	
  agent.	
  The	
  lysate	
  was	
  centrifuged	
  at	
  30,000x	
  g	
  for	
  30	
  minutes	
  to	
  pellet	
  any	
   insoluble	
  material.	
   Using	
   protein	
   liquid	
   chromatography	
   (AKTA	
   purifier,	
   GE	
   Healthcare),	
   the	
   supernatant	
   was	
   injected	
   onto	
   a	
   25	
   ml	
   Poros	
   MC	
   column	
   (Tosoh	
   Bioscience);	
   a	
   nickel	
   column	
   that	
   binds	
   hexahistidine	
   tags.	
   Samples	
   were	
   then	
   washed	
   with	
   five	
   column	
   volumes	
   (CV)	
   of	
   buffer	
   A	
   to	
   remove	
   unbound	
   E.	
   coli	
   proteins,	
   followed	
   by	
   5	
   CV	
   of	
   1%	
   buffer	
   B	
   in	
   buffer	
   A	
   (vol/vol)	
   to	
   remove	
   non-­‐specifically	
   bound	
   bacterial	
   proteins.	
   Fusion	
  protein	
  was	
  eluted	
  with	
  30%	
  buffer	
  B	
  in	
  buffer	
  A	
  (vol/vol).	
  Samples	
  were	
  then	
   dialyzed	
  in	
  the	
  presence	
  of	
  recombinant	
  TEV	
  protease	
  for	
  at	
  least	
  4	
  hours	
  twice	
  at	
  4°C	
   in	
   buffer	
   A	
   and	
   14	
   mM	
   βME.	
   The	
   addition	
   of	
   the	
   reducing	
   agent	
   here	
   was	
   done	
   regardless	
   of	
   whether	
   or	
   not	
   the	
   sample	
   contained	
   cysteines;	
   being	
   a	
   cysteine-­‐ protease,	
  the	
  TEV	
  protease	
  worked	
  better	
  in	
  the	
  presence	
  of	
  βME.	
  For	
  RyR1	
  217-­‐536,	
    	
    34	
    wild-­‐type	
   and	
   W269A,	
   cleavage	
   was	
   more	
   difficult:	
   more	
   TEV	
   protease	
   and	
   longer	
   dialysis	
   were	
   required	
   to	
   completely	
   cleave	
   the	
   constructs.	
   Samples	
   were	
   then	
   applied	
  onto	
  a	
  25  ml	
  amylose	
  column	
  (New	
  England	
  Biolabs)	
  followed	
  by	
  a	
  second	
  25	
   ml	
  Poros	
  MC	
  column.	
  In	
  both	
  cases	
  the	
  flowthrough	
  was	
  collected	
  and	
  following	
  the	
   nickel	
   column,	
   samples	
   were	
   dialyzed	
   in	
   buffer	
   D(8.0)	
   for	
   RyR1	
   1-­‐536	
   and	
   217-­‐536	
   wild-­‐type	
   or	
   mutant	
   constructs,	
   or	
   buffer	
   D(7.4)	
   for	
   RyR1	
   4071-­‐4138	
   and	
   1-­‐617	
   constructs.	
   Here	
   and	
   in	
   further	
   purification,	
   14mM	
   βME	
   was	
   added	
   to	
   all	
   buffers	
   in	
   which	
   cysteine-­‐containing	
   constructs	
   were	
   purified.	
   In	
   some	
   cases,	
   the	
   amylose	
   column	
  was	
  instead	
  run	
  before	
  TEV	
  cleaving	
  to	
  increase	
  purity.	
   Following	
  dialysis,	
  samples	
  were	
  transferred	
  to	
  a	
  HiLoad	
  Q	
  Sepharose	
  column	
   (GE	
   Healthcare)	
   and	
   subject	
   to	
   a	
   0%	
   to	
   30%	
   (vol/vol)	
   buffer	
   E	
   in	
   buffer	
   D	
   gradient	
   (pHs	
   for	
   each	
   were	
   as	
   mentioned	
   for	
   buffer	
   D	
   above).	
   After	
   the	
   gradient,	
   two	
   CV	
   of	
   100%	
   buffer	
   E	
   were	
   used	
   to	
   wash	
   the	
   column.	
   Constructs	
   would	
   elute	
   during	
   the	
   gradient	
   and	
   observed	
   as	
   peaks	
   of	
   UV	
   absorbance	
   at	
   280	
   nm	
   on	
   resulting	
   chromatograms.	
  In	
  a	
  final	
  purification	
  step,	
  a	
  HiLoad	
  16/60	
  Superdex	
  200	
  prep	
  grade	
   (GE	
   Healthcare)	
   gel	
   filtration	
   column	
   was	
   used	
   to	
   confirm	
   a	
   well-­‐behaved,	
   single	
   species	
  was	
  purified.	
  Ideally,	
  a	
  single	
  symmetric	
  peak,	
  not	
  in	
  the	
  void	
  volume	
  would	
   be	
  observed	
  in	
  the	
  chromatogram.	
   Amicon	
  concentrators	
  (3K	
  cut-­‐off	
  for	
  RyR1	
  4071-­‐4138	
  constructs	
  and	
  10K	
  cut-­‐ off	
   for	
   RyR1	
   1,217-­‐536	
   and	
   1-­‐617	
   constructs)	
   were	
   used	
   to	
   exchange	
   the	
   samples	
   into	
  10	
  mM	
  KCl,	
  10	
  mM	
  HEPES	
  7.4	
  and	
  5	
  mM	
  dithiothreitol,	
  and	
  then	
  concentrate	
  the	
   protein	
   to	
   ~10-­‐20	
   mg/ml	
   (for	
   crystallography).	
   Mass	
   spectrometry	
   and	
   sodium	
   dodecyl	
   sulphate	
   polyacrylamide	
   gel	
   electrophoresis	
   (SDS	
   PAGE)	
   were	
   used	
   to	
   confirm	
  the	
  purity	
  of	
  final	
  preparations.	
   	
   	
   	
   	
   	
   	
    	
    35	
    2.3	
  X-­ray	
  Crystallography	
   	
   	
   	
    864	
   random	
   crystallization	
   conditions	
   (Qiagen)	
   were	
   set	
   up	
   using	
   a	
   Phoenix	
    robot	
   in	
   96-­‐well	
   Intelliplates	
   (Hampton	
   Research)	
   and	
   implementing	
   sitting-­‐drop	
   vapour	
   diffusion.	
   Fine	
   screens	
   in	
   24-­‐well	
   VDX	
   plates	
   (Hampton	
   Research)	
   were	
   performed	
  around	
  the	
  conditions	
  of	
  any	
  hits	
  from	
  the	
  random	
  screen.	
  Crystals	
  were	
   obtained	
   in	
   these	
   fine	
   screens	
   by	
   hanging-­‐drop	
   vapour	
   diffusion.	
   RyR1	
   1-­‐536	
   for	
   soaking	
  experiments	
  was	
  crystallized	
  as	
  before	
  (Tung	
  et	
  al.	
  2010)	
  at	
  4°C	
  in	
  solution	
  of	
   20-­‐40%	
  (vol/vol)	
  saturated	
  ammonium	
  sulphate	
  ((NH4)2SO4),	
  0.1	
  M	
  Bicine	
  pH	
  9	
  and	
   12%	
   glycerol	
   (vol/vol).	
   Resulting	
   crystals	
   were	
   soaked	
   in	
   20-­‐40%	
   (NH4)2SO4,	
   0.1	
   M	
   Bicine	
  pH	
  9,	
  25%	
  glycerol	
  and	
  90%	
  (vol/vol)	
  caffeine	
  for	
  at	
  least	
  20	
  minutes	
  before	
   being	
  flash-­‐frozen	
  in	
  liquid	
  nitrogen.	
  RyR1	
  217-­‐536	
  was	
  co-­‐crystallized	
  by	
  C-­‐C.	
  Tung	
   in	
   0.1M	
   Tris	
   8.0,	
   15-­‐25%	
   PEG	
   3350	
   and	
   30	
   mM	
   caffeine	
   and	
   placed	
   in	
   a	
   cryoprotectant	
   of	
   the	
   same	
   condition	
   but	
   with	
   25%	
   glycerol	
   before	
   flash-­‐freezing.	
   RyR1	
   1-­‐617	
   crystals	
   were	
   obtained	
   at	
   room	
   temperature	
   in	
   a	
   condition	
   of	
   5-­‐15%	
   (w/vol)	
  PEG	
  2000	
  MME,	
  0.1	
  M	
  NaCl,	
  0.1	
  M	
  Tris	
  6.5-­‐8.0,	
  with	
  or	
  without	
  90%	
  (vol/vol)	
   saturated	
   dantrolene	
   (in	
   drop)	
   and	
   frozen	
   in	
   cryoprotectant	
   of	
   the	
   same	
   condition	
   with	
  25%	
  glycerol.	
  	
   	
    Datasets	
   were	
   collected	
   at	
   two	
   Rigaku	
   homesources	
   and	
   two	
   synchrotron	
    beamlines:	
   the	
   Canadian	
   Light	
   Source	
   (CLS)	
   beamline	
   08ID-­‐1	
   and	
   the	
   Advanced	
   Photon	
  Source	
  (APS)	
  beamline	
  23-­‐ID-­‐D-­‐GM/CA.	
  RyR1	
  1-­‐536	
  caffeine-­‐soaked	
  crystals	
   produced	
  datasets	
  for	
  final	
  refinement	
  at	
  CLS	
  08ID-­‐1.	
  RyR1	
  217-­‐536	
  and	
  caffeine	
  co-­‐ crystals	
  produced	
  datasets	
  for	
  final	
  refinement	
  at	
  APS	
  23-­‐ID-­‐D-­‐GM/CA.	
  Two	
  datasets	
   were	
   chosen	
   for	
   refinement	
   on	
   RyR1	
   1-­‐617:	
   A	
   native	
   crystal	
   dataset	
   from	
   a	
   homesource	
  to	
  2.7	
  Å	
  and	
  a	
  dataset	
  from	
  a	
  crystal	
  grown	
  and	
  frozen	
  in	
  the	
  presence	
  of	
   dantrolene,	
  which	
  was	
  taken	
  at	
  CLS	
  08ID-­‐1	
  and	
  exhibited	
  diffraction	
  to	
  2.2	
  Å.	
   	
    Data	
   processing	
   from	
   CLS	
   was	
   done	
   using	
   their	
   autoprocess	
   program,	
   whereas	
    data	
  processing	
  from	
  APS	
  was	
  done	
  using	
  their	
  online	
  graphical	
  user	
  interface	
  (GUI).	
   In	
  both	
  cases,	
  their	
  automated	
  programs	
  used	
  XDS	
  (Kabsch,	
  2010).	
  Processing	
  from	
   the	
  homesource	
  was	
  done	
  using	
  HKL2000	
  (Minor	
  et	
  al.	
  2006).	
    	
    36	
    	
    In	
  all	
  cases,	
  individual	
  domains	
  from	
  PDB	
  2XOA	
  (Tung	
  et	
  al.	
  2010)	
  were	
  used	
    as	
  models	
  for	
  molecular	
  replacement	
  using	
  the	
  program	
  Phaser	
  (Storoni	
  et	
  al.	
  2004).	
   Successive	
  rounds	
  of	
  manual	
  building	
  in	
  COOT	
  (Emsley	
  and	
  Cowtan,	
  2004)	
  followed	
   by	
   refinement	
   with	
   REFMAC	
   (Murshudov	
   et	
   al.	
   1997)	
   led	
   to	
   the	
   structures	
   presented	
   here.	
   	
   	
   	
    2.4	
  ITC	
   	
    	
    	
    Proteins	
  were	
  dialyzed	
  against	
  buffer	
  F	
  at	
  4°C	
  with	
  14	
  mM	
  βME	
  if	
  the	
  construct	
    contained	
   cysteines.	
   In	
   the	
   case	
   of	
   RyR1	
   4071-­‐4138,	
   samples	
   were	
   dialyzed	
   once	
   in	
   the	
   buffer	
   F	
   supplemented	
   with	
   10	
   mM	
   ethylenediaminetetraacetic	
   acid	
   (EDTA)	
   pH	
   8.0,	
   then	
   dialyzed	
   three	
   to	
   four	
   times	
   in	
   buffer	
   F	
   to	
   ensure	
   neither	
   Ca2+	
   nor	
   EDTA	
   were	
   present.	
   Small	
   molecules	
   and	
   the	
   RyR1	
   CaMBD	
   4295-­‐4325	
   mutant	
   were	
   dissolved	
   in	
   buffer	
   from	
   the	
   dialysis	
   chamber	
   on	
   the	
   day	
   of	
   the	
   experiment.	
   A	
   6M	
   guanidine	
   hydrochloride,	
   20	
   mM	
   pH	
   6.5	
   phosphate	
   solution	
   was	
   used	
   to	
   dilute	
   protein	
   samples	
   for	
   concentration	
   measurement	
   by	
   absorbance	
   at	
   280	
   nm	
   (Edelhoch	
   1967).	
  	
   In	
   each	
   experiment,	
   40	
   µl	
   of	
   titrant	
   (small	
   molecule	
   or	
   peptide)	
   was	
   added	
   either	
   as	
   20x2µl	
   or	
   40x1µl	
   injections	
   to	
   protein	
   in	
   the	
   cell	
   at	
   a	
   ~10	
   or	
   ~20	
   fold	
   lower	
   concentration	
  and	
  stirred	
  at	
  1000	
  rpm.	
  Experiments	
  on	
  RyR1	
  4071-­‐4138	
  and	
  1-­‐617	
   were	
  performed	
  at	
  25°C	
  while	
  those	
  on	
  RyR1	
  1-­‐536	
  and	
  217-­‐536	
  were	
  performed	
  at	
   4°C,	
  on	
  an	
  ITC200	
  instrument	
  (GE	
  Healthcare).	
  Origin	
  7.0	
  was	
  used	
  to	
  process	
  all	
  data	
   and	
   determine	
   values	
   for	
   binding	
   constants	
   (Ka),	
   Gibb’s	
   free	
   energy	
   (∆G),	
   enthalpy	
   (∆H),	
   entropy	
   (∆S)	
   and	
   stoichiometry	
   (n).	
   The	
   measured	
   change	
   in	
   heat	
   between	
   injections	
   (∆Q(i))	
   due	
   to	
   the	
   injection	
   of	
   dVi	
   µl	
   of	
   titrant	
   into	
   the	
   active	
   cell	
   volume	
   (Vo)	
  is	
  calculated	
  by	
  the	
  equation:	
   	
    	
    37	
    	
   	
   	
    From	
  this	
  Ka,	
  ∆H,	
  and	
  n	
  are	
  determined	
  by	
  fitting	
  to	
  the	
  equation:	
    	
   	
   	
   	
    The	
   macromolecular	
   (Mt)	
   and	
   ligand	
   (Xt)	
   concentrations	
   are	
   additionally	
    corrected	
  to	
  account	
  for	
  displaced	
  volume	
  effects	
  which	
  occur	
  with	
  each	
  injection.	
   ∆G	
  and	
  ∆S	
  are	
  obtained	
  from	
  their	
  relationship	
  to	
  the	
  calculated	
  values	
  of	
  ∆H	
   and	
  Ka	
  (1/Kd,	
  where	
  Kd	
  is	
  the	
  dissociation	
  constant):	
   	
   	
  and	
    	
    	
   	
   	
    2.5	
   in	
  silico	
  Docking	
  -­	
  RyR1	
  1-­617	
  into	
  Cryo-­EM	
  Maps	
   	
   	
    The	
  program	
  ADP_EM	
  (Garzón	
  et	
  al.	
  2007)	
  was	
  used	
  to	
  dock	
  the	
  RyR1	
  1-­‐617	
    crystal	
  structure	
  into	
  three	
  different	
  cryo-­‐EM	
  maps:	
  EMDB	
  1275	
  (Ludtke	
  et	
  al.	
  2005),	
   EMDB	
  1606	
  (Samsó	
  et	
  al.,	
  2009)	
  and	
  EMBD	
  1607	
  (Samsó	
  et	
  al.,	
  2009).	
  The	
  following	
   command	
  was	
  used	
  for	
  all	
  docking:	
   	
   adp_em	
  <em_map.ccp4>	
  <crystal_structure.pdb>	
  <bw>	
  <em_map_resolution>	
  <cutoff>	
  	
    	
   In	
   any	
   run,	
   the	
   bandwidth	
   (<bw>)	
   and	
   density	
   threshold	
   value	
   of	
   the	
   map	
   (<cutoff>)	
   were	
   set	
   to	
   16	
   and	
   0.06	
   respectively.	
   The	
   program	
   does	
   not	
   consider	
   density	
  levels	
  below	
  the	
  density	
  threshold	
  value.	
  Higher	
  bandwidth	
  values,	
  used	
  in	
  a	
   harmonic	
  transformation	
  step	
  of	
  the	
  docking	
  would	
  increase	
  the	
  rotational	
  accuracy.	
   	
    38	
    Laplacian	
  filters	
  were	
  used	
  by	
  default	
  in	
  all	
  runs	
  given	
  their	
  necessity	
  in	
  the	
  docking	
   of	
  small	
  crystal	
  structures	
  into	
  large	
  EM	
  maps	
  (Wriggers	
  and	
  Chacón,	
  2001,	
  Chacón	
   and	
  Wriggers,	
  2002).	
   	
    ADP_EM	
  works	
  by	
  rotating	
  and	
  translating	
  a	
  simulated	
  probe	
  map,	
  obtained	
  by	
    blurring	
   the	
   resolution	
   of	
   the	
   atomic	
   structure	
   to	
   be	
   docked,	
   to	
   maximize	
   a	
   correlation	
   function	
   between	
   it	
   and	
   a	
   target	
   experimental	
   EM	
   map	
   (Garzón	
   et	
   al.	
   2007).	
  Bar	
  charts	
  of	
  docking	
  contrast	
  were	
  made	
  by	
  taking	
  these	
  correlation	
  values	
   from	
  the	
  docking	
  log	
  file	
  and	
  plotting	
  them	
  in	
  Microsoft	
  Excel	
  (2008).	
  	
   	
   	
   	
    2.6	
   in	
  silico	
  Docking	
  -­	
  Caffeine	
  into	
  RyR1	
  1-­536	
   	
   	
    Two	
  separate	
  programs	
  were	
  used	
  to	
  dock	
  caffeine	
  to	
  the	
  RyR1	
  1-­‐536	
  (RyR1	
    ABC)	
  crystal	
  structure:	
  AutoDock	
  (Morris	
  et	
  al.	
  2009)	
  and	
  the	
  DOCK	
  6	
  program	
  suite	
   from	
   the	
   University	
   of	
   California,	
   San	
   Francisco	
   (Shoichet	
   et	
   al.	
   1992,	
   Meng	
   et	
   al.	
   1992,	
   Kuntz	
   et	
   al.	
   1982).	
   In	
   both	
   programs,	
   10	
   standard	
   docking	
   runs	
   were	
   carried	
   out	
   and	
   results	
   were	
   grouped	
   into	
   clusters	
   that	
   contained	
   similar	
   docking	
   sites.	
   Microsoft	
  Excel	
  (2008)	
  was	
  implemented	
  to	
  rank	
  the	
  hits	
  according	
  to	
  binding	
  energy	
   (AutoDock)	
   and	
   grid	
   score	
   (DOCK).	
   Initial	
   files	
   for	
   DOCK	
   were	
   created	
   using	
   UCSF	
   Chimera	
  (Pettersen	
  et	
  al.	
  2004).	
   	
    	
    	
    39	
    3	
   Results	
   	
   	
   	
    3.1	
   	
  A	
  New	
  Structural	
  Domain	
   	
    	
    In	
   the	
   crystal	
   structure	
   of	
   the	
   first	
   three	
   domains	
   of	
   RyR1,	
   there	
   was	
    intrepretable	
   electron	
   density	
   up	
   to	
   residue	
   532	
   (Tung	
   et	
   al.	
   2010).	
   The	
   Protein	
   Homology/analogY	
   Recognition	
   Engine	
   version	
   2.0	
   (PHYRE2)	
   (Kelley	
   and	
   Sternberg	
   2009)	
   helped	
   in	
   defining	
   the	
   rough	
   boundaries	
   for	
   a	
   fourth	
   domain.	
   To	
   account	
   for	
   any	
  errors	
  in	
  theoretical	
  modelling,	
  six	
  different	
  N-­‐terminal	
  constructs	
  with	
  different	
   ending	
   points	
   were	
   cloned:	
   RyR1	
   1-­‐608,	
   617,	
   627,	
   635,	
   644	
   and	
   654	
   (N1	
   to	
   6	
   respectively).	
   	
    Expression	
   was	
   detectable	
   for	
   all	
   six	
   fusion	
   proteins	
   by	
   SDS-­‐PAGE	
   but	
   only	
   the	
    1-­‐617	
   construct	
   had	
   a	
   significant	
   amount	
   of	
   soluble	
   material	
   (Figure	
   11A	
   and	
   11B).	
   The	
   calculated	
   molecular	
   weights	
   of	
   the	
   smallest	
   and	
   largest	
   fusion	
   proteins	
   were	
   ~112	
   and	
   ~117	
   kDa	
   respectively.	
   RyR1	
   1-­‐617	
   was	
   subjected	
   to	
   purification	
   by	
   column	
   chromatography	
   and	
   yielded	
   soluble,	
   monomeric	
   protein	
   on	
   a	
   gel	
   filtration	
   superdex	
   column	
   (Figure	
   11C).	
   Figure	
   11D	
   shows	
   an	
   example	
   SDS-­‐PAGE	
   gel	
   from	
   a	
   purification.	
  The	
  gel	
  filtration	
  chromatogram	
  shows	
  that	
  multiple	
  oligomeric	
  states	
  of	
   the	
  protein	
  exist	
  in	
  solution,	
  but	
  the	
  monomeric	
  fraction	
  re-­‐runs	
  as	
  a	
  monomer	
  on	
  a	
   subsequent	
  superdex	
  column.	
  	
   	
    RyR1	
   1-­‐617	
   crystals	
   diffracted	
   to	
   2.7	
   Å	
   at	
   the	
   in-­‐house	
   X-­‐ray	
   source	
   (Figure	
    12).	
   After	
   solving	
   the	
   phase	
   problem	
   by	
   using	
   PDB	
   2XOA	
   (Tung	
   et	
   al.	
   2010)	
   as	
   a	
   molecular	
   replacement	
   model,	
   its	
   structure	
   was	
   deciphered.	
   Crystals	
   grown	
   in	
   the	
   presence	
  of	
  dantrolene	
  diffracted	
  to	
  2.2	
  Å	
  at	
  the	
  CLS	
  synchrotron.	
  Table	
  2	
  shows	
  the	
   statistics	
  from	
  both	
  datasets.	
    	
    40	
    	
   FIGURE 11. Expression, solubility and purification of N-terminal constructs. A) Expression of N1 through N6. The negative control is not shown but comparison of the six constructs shows a clear trend in expression of an increasing MW band from N1 to N6. B) Solubility tests on N1, N2 and N3 clearly show that only N2 contains soluble material (red box) in the supernatant following lysis and centrifugation. Note: the N6 expression is from a different SDS-PAGE gel and is shown here for illustrative purposes only. C) A gel filtration column superdex column elution profile shows that a monomer (*) does exist and can be separated. D) A purification gel of RyR1 1-617 (triangle) showing the lysis supernatant (L), first nickel column elution (MC), amylose elution (Am), TEV-cleaved sample (TEV), post-TEV nickel flowthrough (MC2FT) and elution (MC2E), anion exchange column (HQ) and gel filtration elution (SD) from a single prep. 	
    41	
    	
   FIGURE 12. RyR1 1-617 crystals. Crystals of RyR1 1-617 (A) diffracted to 2.7 Å (B) at a homesource. 	
   With	
   the	
   previously	
   identified	
   dantrolene	
   binding	
   site	
   in	
   RyR1	
   presumably	
   located	
   in	
   residues	
   590-­‐609	
   (Paul-­‐Pletzer	
   et	
   al.	
   2002),	
   there	
   was	
   a	
   lot	
   of	
   interest	
   in	
   the	
   RyR1	
   1-­‐617	
   structure.	
   There	
   were	
   two	
   molecules	
   in	
   the	
   asymmetric	
   unit,	
   but	
   unfortunately,	
  electron	
  density	
  was	
  only	
  visible	
  to	
  residue	
  ~577	
  in	
  both	
  (Figure	
  13A).	
   This	
   meant	
   that	
   the	
   C-­‐terminal	
   residues	
   were	
   too	
   flexible	
   to	
   be	
   trapped	
   in	
   a	
   crystal	
   lattice.	
   Often,	
   constructs	
   are	
   degraded	
   to	
   a	
   stable	
   core	
   that	
   can	
   be	
   crystallized,	
   but	
   mass	
   spectrometry	
   suggested	
   that	
   with	
   a	
   molecular	
   weight	
   of	
   68.6	
   kDa,	
   the	
   construct	
   was	
   still	
   intact.	
   The	
   co-­‐crystals	
   with	
   dantrolene,	
   which	
   diffracted	
   to	
   ~2.2	
   Å	
   in	
   the	
   same	
   space	
   group	
   again	
   showed	
   interpretable	
   electron	
   density	
   to	
   residue	
   ~577,	
   suggesting	
   that	
   dantrolene	
   did	
   not	
   induce	
   structure	
   in	
   the	
   flexible	
   part.	
   The	
   concentration	
   of	
   dantrolene	
   in	
   the	
   soaking	
   solution	
   though	
   was	
   very	
   limited	
   due	
   to	
   the	
  drug’s	
  inherent	
  lack	
  of	
  solubility.	
   As	
   expected,	
   the	
   first	
   three	
   domains	
   look	
   very	
   similar	
   to	
   the	
   ones	
   already	
   described	
   by	
   RyR1ABC:	
   two	
   beta	
   trefoil	
   domains	
   followed	
   by	
   an	
   α-­‐helical	
   bundle	
   (Figure	
  13B).	
  	
   	
   	
    	
    42	
    	
   Data	
  collection	
   Space	
  group	
   Cell	
  dimensions	
   	
  	
  	
  	
  a,	
  b,	
  c	
  (Å)	
   	
  	
  	
  	
  α,β,γ	
  (°)	
  	
   Resolution	
  (Å)	
   Rsym	
  or	
  Rmeas.	
   I	
  /	
  σ(I)	
   Completeness	
  (%)	
   	
   Refinement	
   Resolution	
  (Å)	
   No.	
  reflections	
   Rwork	
  /	
  Rfree	
   No.	
  atoms	
   	
  	
  	
  	
  Protein	
   	
  	
  	
  	
  Water	
  	
   B-­‐factors	
   	
  	
  	
  	
  Protein	
   	
  	
  	
  	
  Water	
   R.m.s.	
  deviations	
   	
  	
  	
  	
  Bond	
  lengths	
  (Å)	
   	
  	
  	
  	
  Bond	
  angles	
  (°)	
   Ramachandran	
   (core/allowed	
  %)	
    RyR1	
  1-­‐617	
   Homesource	
   P21	
   	
   69.3,	
  103.2,	
  128.7	
   90.0,	
  96.7,	
  90.0	
   50.00-­‐2.70	
  	
  	
  (2.80-­‐2.70)	
   11.7	
  	
  	
  (35.3)	
  (Rsym)	
   4.47	
  	
  	
  (1.83)	
   82.6	
  (45.3)	
   	
   	
   127.86-­‐2.70	
   38924	
   27.98/31.23	
   8377	
   	
  	
  	
  	
  8309	
   	
  	
  	
  	
  68	
   39.57	
   	
  	
  	
  	
  39.33	
   	
  	
  	
  	
  68.54	
   	
   0.005	
   0.861	
    RyR1	
  1-­‐617	
  &	
  Dantrolene	
   CLS	
   P21	
   	
   69.2,	
  103.3,	
  127.9	
   90.0,	
  95.9,	
  90.0	
    96.41/3.59	
    92.25/4.75	
    50.00-­‐2.16	
  	
  	
  (2.22-­‐2.16)	
  	
   16.8	
  	
  	
  (152.2)	
  (Rmeas)*	
   7.93	
  	
  	
  (1.16)	
   99.5	
  	
  	
  (98.9)	
   	
   	
   47.86-­‐2.35	
   70630	
   29.12/31.49	
   8339	
   	
  	
  	
  	
  8272	
   	
  	
  	
  	
  67	
   35.34	
   	
  	
  	
  	
  35.13	
   	
  	
  	
  61.46	
   	
   0.004	
   0.776	
    Table 2. Data collection and refinement statistics for RyR1 1-617. Values in parentheses are for the highest-resolution shell. The R-factor R (Arndt al. 1968), obtained sym  by HKL2000 (Minor et al. 2006) measures the qualtity of data collection. R , obtained meas  from XDS (Kabsch, 2010), is version of R , modified to be redundancy independent sym  (Diederichs and Karplus, 1997).  	
    43	
    FIGURE 13. RyR1 1-617 structure. The protein crystallized with two molecules in the asymmetric unit (A), both of which show a similar structure to domains A (blue), B (green) and C (red) of PDB 2XOA (Tung et al. 2010). A 34-residue α-helix (yellow) describes the main structural addition (B). (C) RyR1 1-617 is shown from the direction of the arrow in (B), with everything gray except for the new helix (yellow) and some buildable loops (teal). Black side-chains highlight the sites of newly built disease mutation; due to inherent flexibility in the crystal structure, some side-chains have been truncated. Construct numbering: RyR1 Oryctolagus cuniculus, mutation numbering: RyR1 or RyR2 (*) Homo  sapiens. 	
    44	
    3.1.1	
  New	
  Loops	
  and	
  Disease	
  Mutations	
    	
    	
   The	
   new	
   structure	
   extends	
   ~45	
   residues	
   at	
   the	
   C-­‐terminus,	
   compared	
   to	
   the	
   previously	
  published	
  RyR1ABC	
  structure	
  (Tung	
  et	
  al.	
  2010).	
  In	
  addition,	
  several	
  loop	
   regions,	
  previously	
  displaying	
  poor	
  density,	
  could	
  now	
  be	
  modelled.	
  A	
  detailed	
  look	
  at	
   the	
   sites	
   of	
   disease-­‐causing	
   mutations	
   has	
   already	
   been	
   performed	
   on	
   RyR1ABC	
   (Tung	
  et	
  al.	
  2010),	
  but	
  due	
  to	
  flexibility	
  in	
  the	
  crystal	
  structure,	
  some	
  of	
  the	
  molecular	
   insight	
  was	
  limited	
  for	
  a	
  number	
  of	
  mutation	
  sites.	
  In	
  this	
  extended	
  structure,	
  we	
  now	
   observe	
   seven	
   additional	
   mutation	
   sites	
   in	
   RyR1,	
   and	
   two	
   that	
   correspond	
   to	
   mutations	
  in	
  RyR2	
  (Figure	
  13C).	
  In	
  addition,	
  the	
  structure	
  provided	
  some	
  molecular	
   insight	
   into	
   some	
   mutation	
   sites	
   whose	
   structures	
   were	
   already	
   known,	
   but	
   whose	
   interaction	
  networks	
  were	
  not	
  (Table	
  3	
  and	
  Chapter	
  4.1.3).	
   New	
  Information	
   RyR1	
  Human	
  Mutation	
   Structure	
   M226K	
   Structure	
   D227V	
   Structure	
   R367Q	
   Structure	
   R367L	
   Structure	
   S427L	
   Interaction	
   Q474H	
   Interaction	
   Y522C	
   Interaction	
   Y522S	
   Structure	
  	
   R533H	
   Interface	
   Structure	
   R533C	
   Interface	
   Structure	
   D544Y	
   Interface	
   Structure	
   R552W	
   Interaction	
   	
   RyR2	
  Human	
  Mutation	
   Structure	
   H240R	
   Structure	
   V507I	
   Interaction	
   Structure	
   A549V	
   Interface	
   TABLE 3. Additional mutations or their  Disease	
   MH	
   MH	
   MH	
   MH	
   CCD	
   MH,	
  CCD	
   MH	
   MH,	
  CCD	
   MH	
    Reference	
   Levano	
  et	
  al.	
  2009	
   Monnier	
  et	
  al.	
  2005	
   Galli	
  et	
  al.	
  2006	
   Levano	
  et	
  al.	
  2009	
   Wu	
  et	
  al.	
  2006	
   Ibarra	
  et	
  al.	
  2006	
   Yeh	
  et	
  al.	
  2006	
   Quane	
  et	
  al.	
  1994	
   Brandt	
  et	
  al.	
  1999	
    MH	
    Tammaro	
  et	
  al.	
  2003	
    MH	
    Levano	
  et	
  al.	
  2009	
    MH	
    Keating	
  et	
  al.	
  1997	
    Disease	
   CPVT	
   CPVT	
    Reference	
   Tester	
  et	
  al.	
  2006	
   Medeiros-­‐Domingo	
   et	
   al.	
  2009	
   CPVT	
   Medeiros-­‐Domingo	
   et	
   al.	
  2009	
   interactions in the RyR1 1-617  structure. The first column indicates whether the new information on a mutation site was solely in its structure, in its interaction network or its interface predictions. 	
    45	
    3.1.2	
  Docking	
  into	
  Cryo-­EM	
  Maps	
   	
   The	
  RyR1	
  1-­‐617	
  structures	
  (2.2	
  Å	
  and	
  2.7	
  Å)	
  were	
  docked	
  into	
  three	
  available	
   cryo-­‐EM	
  maps:	
  EMDB	
  1275	
  (Ludtke	
  et	
  al.	
  2005),	
  EMDB	
  1606	
  and	
  EMDB	
  1607	
  (Samsó	
   et	
   al.	
   2009)	
   using	
   Laplacian	
   filters.	
   In	
   all	
   cases,	
   the	
   top	
   four	
   solutions	
   corroborated	
   the	
   tetrameric	
   vestibule	
   found	
   in	
   the	
   docking	
   of	
   RyR1ABC	
   (Tung	
   et	
   al.	
   2010).	
   To	
   validate	
   the	
   docking,	
   correlation	
   coefficients	
   were	
   grouped	
   into	
   sets	
   of	
   four,	
   to	
   accommodate	
   the	
   inherent	
   symmetry	
   of	
   the	
   channel,	
   and	
   averaged	
   for	
   comparison.	
   The	
   top	
   sets	
   exhibited	
   correlation	
   coefficients	
   that	
   were	
   27,	
   92,	
   and	
   98	
   times	
   the	
   standard	
  deviation	
  from	
  the	
  mean	
  of	
  the	
  next	
  ten	
  sets	
  for	
  EMDB	
  1275,	
  1606	
  and	
  1607	
   respectively	
   (Figure	
   14).	
   These	
   values	
   for	
   docking	
   contrast	
   are	
   very	
   high	
   and	
   add	
   a	
   great	
  deal	
  of	
  confidence	
  to	
  the	
  placement	
  of	
  the	
  structure.	
   	
   	
   	
    FIGURE 14. Docking contrast. The correlation coefficients of solutions from docking are shown for different maps. Each bar represents an average of four solutions that group due to the symmetry of the tetrameric channel.  	
    46	
    Visualizing	
  the	
  interactions	
  between	
  subunits	
  in	
  RyR1	
  1-­‐617	
  is	
  best	
  observed	
    	
    in	
  structures	
  docked	
  into	
  an	
  EM	
  map	
  of	
  a	
  channel	
  in	
  the	
  closed	
  conformation	
  (EMDB	
   1606).	
  When	
  examined	
  we	
  see	
  that	
  four	
  B	
  domains	
  form	
  the	
  bulk	
  of	
  the	
  central	
  rim	
   and	
  combined	
  with	
  domain	
  A,	
  form	
  the	
  aforementioned	
  vestibule.	
  With	
  the	
  additional	
   structural	
   information	
   from	
   the	
   extended	
   helix	
   from	
   domain	
   C,	
   it	
   seems	
   that	
   this	
   third,	
   helical	
   moiety	
   acts	
   as	
   a	
   bridge	
   between	
   the	
   central	
   rim	
   and	
   the	
   clamp	
   region	
   (Figure	
   15A,	
   B	
   and	
   C).	
   Looking	
   at	
   the	
   new,	
   visible	
   disease	
   mutation	
   sites,	
   there	
   is	
   a	
   general	
   agreement	
   with	
   previous	
   findings.	
   Disease-­‐causing	
   mutations	
   occur	
   at	
   sites	
   that	
   would	
   affect	
   interactions	
   between	
   domains	
   in	
   RyR1.	
   One	
   interesting	
   exception	
   is	
   the	
  S427L	
  (human	
  RyR1)	
  CCD-­‐causing	
  mutation	
  that	
  seems	
  to	
  be	
  completely	
  solvent	
   accessible	
  and	
  not	
  involved	
  in	
  interface	
  interaction.	
  Upon	
  analysis	
  though,	
  Wu	
  et	
  al.	
   (2006)	
  showed	
  that	
  the	
  mutation	
  was	
  sequenced	
  as	
  heterozygous	
  and	
  accompanied	
   by	
   another	
   mutation	
   that	
   had	
   previously	
   been	
   linked	
   to	
   CCD	
   (Monnier	
   et	
   al.	
   2001,	
   Romero	
   et	
   al.	
   2003).	
   This	
   suggests	
   that	
   the	
   S427L	
   mutation	
   may	
   not	
   be	
   a	
   causative	
   mutation	
  for	
  CCD.	
    	
   	
   	
   	
    	
    47	
    	
   FIGURE 15. Docking of the RyR1 1-617 crystal structure into EMDB 1606. Docking into the closed conformation of RyR1 (Samsó et al. 2009) visually enhances the interactions between subunits. RyR1A (blue) and B (green) form a ring around the cytoplasmic pore as seen from the top view (A). Domain C (red) and its extended helix (yellow) attach between the two and seem to link the vestibule to domains in the clamp region. Zoomed in views from the top (B) and side (C) show that the new structural information on mutation sites from the structure generally agrees with previous results. Human mutations (black) except S427 (labelled) occur at interfaces between domains in the full-length channel and probably affect allosteric communication. Interface 1 (Tung et al. 2010) is labelled as a dashed white line. Note: blackened here are only newly structured mutation sites. A detailed analysis including mutation sites that interact with freshly built structural motifs is covered in Chapter 4.1.3. Constuct numbering: RyR1 Oryctolagus  cuniculus, mutation numbering: RyR1 Homo sapiens. 	
    48	
    3.1.3 Dantrolene	
  Binding	
   	
   Despite	
  the	
  lack	
  of	
  success	
  in	
  determining	
  the	
  structure	
  of	
  dantrolene	
  bound	
  to	
   RyR1	
  1-­‐617,	
  as	
  mentioned	
  the	
  measured	
  molecular	
  weight	
  of	
  the	
  protein	
  resembled	
   that	
   of	
   full-­‐length	
   construct.	
   Although	
   crystals	
   were	
   grown	
   in	
   the	
   presence	
   of	
   dantrolene,	
   they	
   were	
   grown	
   in	
   a	
   condition	
   optimized	
   for	
   unbound	
   RyR1	
   1-­‐617.	
   In	
   the	
   dynamic	
   equilibrium	
   of	
   binding,	
   large	
   structural	
   rearrangements	
   may	
   be	
   required.	
   Experiments	
   such	
   as	
   this	
   can	
   bias	
   structures	
   to	
   the	
   assembly	
   most	
   beneficial	
  for	
  crystal	
  lattice	
  formation,	
  and	
  not	
  one	
  suited	
  for	
  binding.	
  This	
  effectively	
   pulls	
   unbound	
   RyR1	
   1-­‐617	
   out	
   of	
   solution	
   and	
   into	
   crystal	
   formation.	
   It	
   is	
   then	
   possible	
  that	
  in	
  solution,	
  binding	
  to	
  dantrolene	
  could	
  still	
  be	
  occurring	
  and	
  we	
  could	
   not	
  eliminate	
  the	
  possibility	
  based	
  purely	
  on	
  the	
  lack	
  of	
  crystallographic	
  evidence.	
   ITC	
   experiments	
   performed	
   by	
   K.	
   Lau	
   were	
   carried	
   out	
   in	
   an	
   attempt	
   to	
   clarify	
   whether	
   RyR1	
   1-­‐617	
   does	
   in	
   fact	
   contain	
   the	
   dantrolene	
   binding	
   site.	
   Both	
   dantrolene,	
   and	
   its	
   more	
   water-­‐soluble	
   homolog	
   azumolene	
   showed	
   no	
   measurable	
   affinity	
   between	
   the	
   two	
   (data	
   not	
   shown).	
   This	
   is	
   in	
   direct	
   conflict	
   to	
   earlier	
   work	
   in	
   the	
   field	
   (Paul-­‐Pletzer	
   et	
   al.	
   2002).	
   However,	
   ITC	
   measurements	
   rely	
   on	
   changes	
   in	
   enthalpy.	
   If	
   a	
   reaction	
   is	
   driven	
   by	
   entropy	
   and	
   is	
   enthalpically	
   silent,	
   it	
   cannot	
   be	
   detected.	
   In	
   addition,	
   we	
   aim	
   to	
   retry	
   the	
   experiment	
   with	
   a	
   new	
   batch	
   of	
   purified	
   RyR1ABCD	
  to	
  make	
  sure	
  our	
  observations	
  are	
  reproducible.	
  	
   	
   	
    	
   	
   	
   	
   	
   	
   	
    	
    49	
    3.1.4	
  N-­terminal	
  Construct	
  Summary	
   	
   Several	
   different	
   constructs	
   were	
   tried	
   before	
   and	
   after	
   the	
   crystallographic	
   success	
  of	
  RyR1	
  1-­‐617.	
  Figure	
  16	
  shows	
  a	
  schematic	
  of	
  the	
  constructs	
  tried	
  and	
  Table	
   4	
  summarizes	
  the	
  problems	
  and	
  potentials	
  in	
  the	
  attempted	
  N-­‐terminal	
  clones.	
  	
   	
   	
    	
   FIGURE 16. N-terminal construct schematic. An illustration describing the constructs attempted in the search for the next N-terminal RyR1 domain. Domains A, B and C from PDB 2XOA are coloured blue, green and red respectively, and the new structure from RyR1 1-617 is labelled ‘d’ and coloured yellow. Forward (bright red) and reverse (bright  green) primers for construct PCR are numbered according to the first (forward) or last (reverse) amino acid they code for. Of the reverse primers ending at residues 607, 617, 627, 635, 644 and 654, only 617 is numbered. Numbering: RyR1 Oryctolagus cuniculus.  	
    50	
    Construct	
   1-­‐536	
   1-­‐574	
   1-­‐589	
    Expression	
   18°C,	
  ~24	
  hours	
   18°C,	
  ~24	
  hours	
   18°C,	
  ~24	
  hours	
    Solubility	
   very	
  soluble	
   slightly	
   	
    1-­‐608	
   1-­‐617	
   1-­‐627	
    18°C,	
  ~24	
  hours	
   18°C,	
  ~24	
  hours	
   18°C,	
  ~24	
  hours	
    slightly	
   slightly	
   slightly	
    1-­‐635	
   1-­‐644	
   1-­‐654	
    18°C,	
  ~24	
  hours	
   18°C,	
  ~24	
  hours	
   18°C,	
  ~24	
  hours	
    slightly	
   slightly	
   slightly	
    1-­‐796	
    18°C,	
  ~24	
  hours	
    217-­‐536	
    18°C,	
  ~24	
  hours	
    slightly	
   	
   very	
  soluble	
    217-­‐574	
    18°C,	
  ~24	
  hours	
    slightly	
    217-­‐589	
    18°C,	
  ~24	
  hours	
    slightly	
    217-­‐654	
   394-­‐574	
   394-­‐654	
   533-­‐796	
    18°C,	
  ~24	
  hours	
   18°C,	
  ~24	
  hours	
   18°C,	
  ~24	
  hours	
   18°C,	
  ~24	
  hours	
    slightly	
   slightly	
   slightly	
   slightly	
    570-­‐796	
    18°C,	
  ~24	
  hours	
    slightly	
    Purification	
   Large	
  yield,	
  crystallized.	
   Aggregate	
  on	
  SD	
   Low	
  yield,	
  ~50%	
  aggregate	
  on	
   SD	
   N/A	
   Purified	
  successfully,	
  crystallized	
   Very	
  low	
  yield,	
  stopped	
  after	
  first	
   MC	
   N/A	
   N/A	
   Low	
  yield,	
  ~50%	
  aggregate	
  on	
   SD	
   Digested	
  into	
  smaller	
  domains	
  by	
   proteases,	
  aggregate	
  on	
  SD	
   Large	
  yield,	
  sometimes	
  crashes	
   out	
  upon	
  addition	
  of	
  TEV.	
  Lots	
  of	
   TEV	
  and	
  time	
  required	
  to	
  cleave.	
   Crystallized.	
   50%	
  aggregate	
  on	
  SD	
  as	
  fusion	
   protein,	
  low	
  yield.	
   Crashes	
  out	
  with	
  TEV,	
  associates	
   with	
  MBP	
   Associates	
  with	
  MBP	
   Lost	
  on	
  HQ	
   Associates	
  with	
  MBP	
   Some	
  monomer	
  attainable	
  as	
   fusion	
  protein	
  (~60%	
  aggregate	
   on	
  SD)	
   Mostly	
  monomer	
  on	
  SD	
  as	
  fusion	
   protein,	
  low	
  yield	
    TABLE 4. N-terminal constructs. A list of the different constructs attempted in the search for a crystallisable fourth domain from the N-terminus. The RyR1ABC and BC (1536 and 217-536) constructs were made based on the PDB 2XOA – purification was optimized for these clones by C-C. Tung. 	
    	
   	
    	
    51	
    3.2	
   Caffeine	
  Binding	
   	
    	
    RyR1	
   1-­‐617	
   is	
   responsible	
   for	
   a	
   large	
   portion	
   of	
   the	
   full-­‐length	
   channels	
    solvent	
  exposed	
  surface	
  area	
  with	
  many	
  large	
  valleys	
  and	
  nooks	
  available.	
  With	
  the	
   large	
   number	
   of	
   small	
   molecules	
   and	
   ligands	
   that	
   have	
   been	
   shown	
   to	
   modulate	
   RyR,	
   it	
   made	
   sense	
   to	
   systematically	
   test	
   some	
   of	
   them,	
   quantitatively	
   for	
   binding.	
   Being	
   relatively	
  inexpensive	
  and	
  readily	
  accessible,	
  caffeine	
  provided	
  a	
  great	
  starting	
  point.	
   The	
   EC50	
   of	
   caffeine	
   on	
   full-­‐length	
   RyR1	
   is	
   ~2.8	
   mM	
   (Groom	
   et	
   al.	
   2011).	
   In	
   order	
  to	
  detect	
  binding	
  with	
  low	
  affinities,	
  high	
  concentrations	
  of	
  protein	
  and	
  ligand	
   are	
   required	
   for	
   a	
   better	
   signal.	
   Previously,	
   C-­‐C.	
   Tung	
   optimized	
   the	
   purification	
   procedure	
   for	
   two,	
   well-­‐expressed	
   constructs	
   in	
   RyR1	
   1-­‐536	
   (RyR1ABC)	
   and	
   RyR1	
   217-­‐536	
  (RyR1BC).	
  Their	
  levels	
  of	
  expression	
  resulted	
  in	
  large	
  yields	
  that	
  made	
  high-­‐ concentration	
  ITC	
  feasible.	
   ITC	
   experiments	
   showed	
   that	
   caffeine	
   bound	
   RyR1BC	
   and	
   ABC	
   with	
   affinities	
   of	
  ~150	
  µM	
  and	
  ~520	
  µM	
  respectively,	
  (Figures	
  17A	
  and	
  17B).	
  With	
  similar	
  ΔH	
  (~	
  -­‐ 6000	
   cal/mol)	
   and	
   ΔS	
   (~	
   -­‐6	
   cal/mol/K)	
   values	
   this	
   suggested	
   that	
   the	
   caffeine-­‐ binding	
   site	
   was	
   located	
   in	
   RyR1BC.	
   The	
   stoichiometry	
   of	
   the	
   interaction	
   remains	
   somewhat	
   of	
   a	
   mystery.	
   An	
   N	
   value	
   of	
   ~0.5	
   could	
   be	
   explained	
   either	
   by	
   errors	
   in	
   concentration	
   measurement	
   and	
   preparation	
   purity,	
   or	
   claim	
   that	
   a	
   dimer	
   of	
   RyR1BC	
   or	
   RyR1ABC	
   binds	
   a	
   single	
   caffeine	
   molecule.	
   Interestingly,	
   both	
   dynamic	
   light	
   scattering	
   and	
   a	
   calibrated	
   gel	
   filtration	
   column	
   output	
   a	
   molecular	
   weight	
   of	
   ~110	
   kDa	
   for	
   RyR1ABC	
   (MW:	
   59	
   kDa).	
   However,	
   these	
   techniques	
   provide	
   a	
   more	
   accurate	
   estimation	
  of	
  hydrodynamic	
  radius	
  than	
  they	
  do	
  molecular	
  weight.	
   	
    	
    52	
    	
   FIGURE 17. Caffeine binds RyR1BC. ITC experiments in which (A) 15 mM caffeine was titrated into 1.5 mM RyR1BC and (B) 8.97 mM caffeine was added to 0.88 mM RyR1ABC. The integrations (lower panels) of the raw heats (upper panels) were fit to a one site binding model and outputted similar thermodynamic details of interaction suggesting RyR1BC accommodates the caffeine binding site. 	
   	
    3.2.1	
  Structural	
  Insight?	
   	
   	
    In	
  order	
  to	
  visualize	
  the	
  molecular	
  details	
  involved	
  in	
  caffeine	
  binding,	
  RyR1BC	
    crystals	
  were	
  grown	
  in	
  the	
  presence	
  of	
  60	
  mM	
  caffeine	
  and	
  RyR1ABC	
  crystals	
  were	
   soaked	
   in	
   a	
   90%	
   saturated	
   caffeine	
   cryoprotectant	
   solution	
   before	
   flash-­‐freezing.	
   Data	
   collection	
   and	
   refinement	
   statistics	
   for	
   both	
   RyR1BC	
   and	
   caffeine	
   co-­‐crystals,	
   and	
  RyR1ABC	
  crystals	
  soaked	
  in	
  caffeine,	
  are	
  shown	
  in	
  Table	
  5.	
    	
    53	
    Analysis	
   of	
   the	
   resultant	
   structures	
   showed	
   that	
   in	
   both	
   cases	
   caffeine	
   is	
   indeed	
   bound.	
   To	
   prove	
   this,	
   the	
   FO-­‐FC	
   maps	
   were	
   calculated	
   using	
   models	
   without	
   the	
   ligand.	
   Positive	
   difference	
   density	
   was	
   visible	
   to	
   9.56σ	
   and	
   18.92σ	
   in	
   the	
   RyR1BC	
   and	
   ABC	
   structures	
   respectively.	
   In	
   combination	
   with	
   the	
   compelling	
   density	
   exhibited	
   in	
   the	
   2FO-­‐FC	
   maps	
   created	
   from	
   refinement	
   on	
   a	
   model	
   with	
   caffeine,	
   we	
   could	
  unquestionably	
  confirm	
  its	
  presence	
  (Figure	
  18).	
   	
    However,	
  in	
  both	
  cases,	
  caffeine	
  takes	
  advantage	
  of	
  hydrophobic	
  pockets	
  and	
    aromatic	
   faces	
   to	
   stack	
   somewhat	
   promiscuously.	
   Furthermore,	
   the	
   binding	
   sites	
   differ	
   considerably	
   between	
   the	
   two	
   structures.	
   Looking	
   closely	
   at	
   the	
   interactions	
   involved	
   in	
   binding,	
   we	
   see	
   that	
   in	
   the	
   RyR1BC	
   structure,	
   caffeine	
   recruits	
   two	
   tryptophan-­‐269	
   (W269)	
   residues	
   from	
   different	
   RyR1B	
   domains	
   and	
   two	
   tyrosine-­‐ 523	
   (Y523)	
   residues	
   from	
   different	
   RyR1C	
   domains.	
   It	
   stacks	
   between	
   the	
   tryptophans	
  and	
  sits	
  on	
  a	
  hydrophobic	
  bed	
  provided	
  by	
  the	
  two	
  tyrosines	
  ~	
  4	
  Å	
  away	
   (Figures	
  18A	
  and	
  18B).	
  With	
  RyR1ABC	
  the	
  interaction	
  is	
  a	
  little	
  more	
  complex.	
  Five	
   RyR1A	
   residues	
   in	
   total	
   are	
   involved,	
   three	
   from	
   one	
   protein	
   chain,	
   and	
   two	
   from	
   a	
   symmetry-­‐related	
   molecule.	
   The	
   main	
   hold	
   on	
   the	
   ligand	
   can	
   be	
   attributed	
   to	
   an	
   aromatic	
   stacking	
   sandwich	
   between	
   histidine-­‐151	
   (H151)	
   in	
   the	
   ‘master’	
   structure	
   and	
  phenylalanine-­‐195	
  (F195)	
  in	
  a	
  neighbouring	
  molecule.	
  Some	
  specificity	
  is	
  created	
   by	
   proline-­‐151’s	
   main-­‐chain	
   oxygen	
   (P151)	
   and	
   a	
   polar	
   tyrosine-­‐179	
   (Y179)	
   in	
   the	
   master	
  molecule,	
  and	
  a	
  symmetry-­‐related	
  methionine-­‐196	
  (M196),	
  which	
  completes	
   a	
  hydrophobic	
  pocket	
  (Figures	
  18C	
  and	
  18D).	
   	
  	
   	
   	
   	
   	
   	
   	
   	
   	
    	
    54	
    	
   Data	
  collection	
   Space	
  group	
   Cell	
  dimensions	
   	
  	
  	
  	
  a,	
  b,	
  c	
  (Å)	
   	
  	
  	
  	
  α,β,γ	
  (°)	
  	
   Resolution	
  (Å)	
   Rmeas.	
   I	
  /	
  σ(I)	
   Completeness	
  (%)	
   	
   Refinement	
   Resolution	
  (Å)	
   No.	
  reflections	
   Rwork	
  /	
  Rfree	
   No.	
  atoms	
   	
  	
  	
  	
  Protein	
   	
  	
  	
  	
  Water	
  	
   	
  	
  	
  	
  Caffeine	
   	
  	
  	
  	
  Other	
   B-­‐factors	
   	
  	
  	
  	
  Protein	
   	
  	
  	
  	
  Water	
   	
  	
  	
  	
  Caffeine	
   	
  	
  	
  	
  Other	
   R.m.s.	
  deviations	
   	
  	
  	
  	
  Bond	
  lengths	
  (Å)	
   	
  	
  	
  	
  Bond	
  angles	
  (°)	
   Ramachandran	
   (core/allowed	
  %)	
    RyR1BC	
  &	
  Caffeine	
   APS	
   P3121	
   	
   68.6,	
  68.6,	
  133.0	
   90.0,	
  120.0,	
  90.0	
   50.00-­‐2.13	
  	
  	
  (2.19-­‐2.13)	
   8.0	
  	
  	
  (163.1)	
   19.75	
  	
  	
  (1.56)	
   99.8	
  	
  	
  (98.0)	
   	
   	
   59.38-­‐2.30	
   15842	
   23.92/30.63	
   2308	
   	
  	
  	
  	
  2257	
   	
  	
  	
  	
  37	
   	
  	
  	
  	
  14	
   	
   45.89	
   	
  	
  	
  	
  45.71	
   	
  	
  	
  	
  50.44	
   	
  	
  	
  	
  62.41	
   	
  	
  	
  	
  N/A	
   	
   0.018	
   1.756	
    RyR1ABC	
  &	
  Caffeine	
   CLS	
   R32	
   	
   170.6,	
  170.6,	
  300.9	
   90.0,	
  90.0,	
  120.0	
    93.80/6.20	
    91.10/8.90	
    50.00-­‐2.05	
  	
  	
  (2.10-­‐2.05)	
  	
   22.0	
  	
  	
  (763.7)	
   7.93	
  	
  	
  (1.16)	
   99.5	
  	
  	
  (98.9)	
   	
   	
   132.62-­‐2.40	
   62534	
   20.23/23.14	
   3931	
   	
  	
  	
  	
  3802	
   	
  	
  	
  	
  95	
   	
  	
  	
  	
  14	
   	
  	
  	
  	
  20	
   49.53	
   	
  	
  	
  	
  49.13	
   	
  	
  	
  	
  57.81	
   	
  	
  	
  	
  52.02	
   	
  	
  	
  	
  90.06	
   	
   0.035	
   2.638	
    TABLE 5. Data collection and refinement statistics for RyR1BC and RyR1ABC with caffeine. Values in parentheses are for the highest-resolution shell. R , meas  obtained from XDS (Kabsch, 2010) is a redundancy independent R-factor (Diederichs and Karplus, 1997). 	
    	
    55	
    	
   	
   FIGURE 18. Crystallographic caffeine binding interactions. A detailed look at the binding of caffeine to RyR1BC (A, B) and RyR1ABC (C, D). The 2F -F maps calculated O  C  from a model with caffeine built in are shown as blue meshes (A, C) and the F -F maps O  C  calculated from a model without caffeine built in are shown as green meshes (B, D). The sigma cut-offs for each map are labelled. Caffeine is shown as black sticks. Side-chains of interacting residues are shown as sticks and coloured according to the domain they belong (blue – RyR1A, green – RyR1B, red – RyR1C). Some non-carbon atoms are highlighted red (oxygen), blue (nitrogen) or yellow (sulphur). Residues are marked with a superscript number denoting which symmetry-related molecule they belong to (1 - master, 2,3,4 symmetry-related). Numbering: RyR1 Oryctolagus cuniculus. 	
    	
    56	
    In	
   RyR1BC,	
   four	
   different	
   symmetry-­‐related	
   structures	
   including	
   the	
   master	
   molecule	
   are	
   involved	
   in	
   holding	
   caffeine	
   and	
   both	
   domains	
   B	
   and	
   C	
   are	
   engaged	
   (Figure	
   19A).	
   Conversely,	
   only	
   RyR1A	
   plays	
   part	
   in	
   binding	
   caffeine	
   to	
   RyR1ABC	
   between	
   two	
   symmetry-­‐related	
   molecules	
   (Figure	
   19B).	
   The	
   requirement	
   of	
   symmetry-­‐related	
  molecules	
  strongly	
  suggested	
  that	
  the	
  binding	
  sites	
  observed	
  might	
   not	
  occur	
  in	
  solution	
  and	
  prompted	
  ITC	
  experiments	
  to	
  confirm	
  them.	
    	
   FIGURE 19. Structural insight into caffeine binding to RyR? Crystal structures of both RyR1BC (A) and RyR1ABC (B) are shown bound to caffeine. In both cases, interactions involve aromatic rings from symmetry-related molecules. Caffeine is shown in the dashed ring interacting with four RyR1BC molecules (A) or two RyR1ABC molecules (B). The side-chains of interacting residues inside the dashed ring are shown, and highlighted in figure 18. For RyR1ABC, only the residues (from both molecules) interacting with the caffeine from the ‘master’ molecule are shown. Domains A, B and C are coloured blue, green and red respectively.	
   	
    57	
    3.2.2	
  W269A	
  Mutant	
  Binding	
   	
   	
    With	
  conflicting	
  results	
  between	
  crystal	
  structures,	
  the	
  search	
  instead	
  for	
  the	
    binding	
  site	
  in	
  solution	
  began.	
  With	
  the	
  knowledge	
  that	
  both	
  RyR1	
  BC	
  and	
  ABC	
  bind	
   caffeine	
   with	
   the	
   same	
   general	
   thermodynamic	
   parameters,	
   we	
   focused	
   first	
   on	
   the	
   information	
   from	
   the	
   RyR1BC	
   co-­‐crystal	
   structure,	
   as	
   the	
   soaking	
   experiments	
   with	
   RyR1ABC	
   yielded	
   a	
   binding	
   pocket	
   in	
   domain	
   A	
   alone.	
   More	
   specifically,	
   we	
   began	
   investigation	
  into	
  W269,	
  and	
  its	
  potential	
  as	
  a	
  caffeine	
  binder.	
   	
    To	
   test	
   this,	
   W269	
   was	
   mutated	
   to	
   an	
   alanine	
   in	
   RyR1BC	
   (Figure	
   20A).	
   The	
    mutant	
  construct	
  expressed	
  and	
  purified	
  as	
  normal,	
  indicating	
  the	
  mutation	
  was	
  not	
   destructive	
  to	
  the	
  overall	
  folding	
  of	
  the	
  protein.	
  An	
  ITC	
  experiment	
  showed	
  binding	
   with	
  an	
  affinity	
  ~300	
  µM,	
  and	
  similar	
  values	
  of	
  ΔH,	
  ΔS	
  and	
  N	
  to	
  those	
  of	
  the	
  wild-­‐type	
   protein	
  (Figure	
  20B).	
  Evidently,	
  W269	
  is	
  not	
  the	
  binding	
  site	
  of	
  caffeine	
  in	
  solution.	
   The	
   RyR1BC-­‐caffeine	
   structure	
   represents	
   a	
   crystallographically	
   trapped	
   conformation.	
  The	
  high	
  concentrations	
  of	
  caffeine	
  used	
  in	
  crystallography	
  seemed	
  to	
   promote	
  caffeine’s	
  non-­‐specific	
  attachment	
  to	
  aromatic	
  residues.	
  This	
  conclusion	
  can	
   be	
  extended	
  to	
  the	
  RyR1ABC-­‐caffeine	
  structure,	
  which	
  disagrees	
  with	
  our	
  ITC	
  results.	
    	
    58	
    FIGURE 20. W269A mutation. (A) A section of the chromatogram from a sequencing reaction (Eurofins) showing the successful mutation of W269 to an alanine. Two bases were changed simultaneously by Quikchange in the mutation from a TGG triplet codon to GCG. (B) 6mM caffeine was titrated into 0.575 mM RyR1BC W269A. The calculated thermodynamic values were similar to those obtained from the same experiment with wildtype RyR1BC (Figure 17). Numbering: RyR1 Oryctolagus cuniculus. 	
    59	
    3.2.3	
  An	
  in	
  silico	
  Search	
  for	
  a	
  Caffeine	
  Binding	
  Site	
   	
   	
    The	
   programs	
   DOCK	
   (Shoichet	
   et	
   al.	
   1992,	
   Meng	
   et	
   al.	
   1992,	
   Kuntz	
   et	
   al.	
   1982)	
    and	
   AutoDock	
   (Morris	
   et	
   al.	
   2009)	
   were	
   implemented	
   to	
   search	
   for	
   potential	
   sites	
   for	
   caffeine	
   binding	
   in	
   RyR1ABC.	
   Each	
   program	
   outputs	
   different	
   scoring	
   numbers	
   that	
   were	
   used	
   in	
   choosing	
   hits	
   for	
   analysis.	
   AutoDock	
   scores	
   based	
   on	
   intermolecular	
   energy,	
   internal	
   energy	
   and	
   torsional	
   energy.	
   A	
   ‘docking	
   energy’	
   is	
   computed	
   by	
   combining	
  the	
  first	
  two,	
  and	
  a	
  ‘binding	
  energy’	
  from	
  the	
  first	
  and	
  third.	
  DOCK	
  uses	
  a	
   grid	
   score	
   that	
   is	
   based	
   on	
   the	
   electrostatic	
   energy	
   and	
   Van	
   der	
   Waals	
   forces.	
   	
   In	
   addition,	
   given	
   their	
   importance	
   in	
   small-­‐molecule	
   docking	
   (Mackey	
   and	
   Melville,	
   2009),	
  the	
  cluster	
  numbers	
  of	
  each	
  hit	
  were	
  taken	
  into	
  account.	
  It	
  is	
  worth	
  noting	
  that	
   in	
   both	
   cases,	
   the	
   number	
   of	
   docking	
   runs	
   is	
   very	
   low	
   (10	
   runs)	
   and	
   a	
   more	
   exhaustive	
  dock	
  will	
  be	
  done	
  in	
  the	
  future	
  to	
  add	
  validity	
  to	
  the	
  results.	
   	
    From	
  DOCK,	
  seven	
  solutions	
  from	
  ten	
  runs	
  were	
  ranked	
  according	
  to	
  their	
  grid	
    score	
  (the	
  second	
  ranked	
  run	
  had	
  three	
  species	
  in	
  its	
  cluster,	
  and	
  the	
  third	
  had	
  two).	
   Analysis	
   of	
   the	
   hits	
   though	
   showed	
   that	
   the	
   top	
   five	
   clusters	
   contained	
   caffeine	
   in	
   the	
   same	
  site	
  but	
  rotated	
  or	
  flipped.	
  This	
  is	
  unsurprising	
  given	
  the	
  pseudo-­‐symmetry	
  of	
   the	
  molecule.	
  As	
  a	
  result,	
  these	
  runs	
  were	
  combined	
  manually,	
  increasing	
  the	
  number	
   of	
   species	
   in	
   the	
   top	
   cluster	
   (D1)	
   to	
   eight.	
   AutoDock	
   offered	
   five	
   solutions	
   from	
   ten	
   runs;	
   the	
   first	
   ranked	
   solution	
   (AD1)	
   had	
   the	
   best	
   binding	
   energy	
   and	
   a	
   cluster	
   number	
   of	
   6.	
   With	
   both	
   programs,	
   there	
   seemed	
   to	
   be	
   a	
   clear	
   docking	
   contrast	
   between	
  the	
  top	
  solutions	
  (D1	
  and	
  AD1)	
  and	
  the	
  rest	
  (Figures	
  21A	
  and	
  21B),	
  but	
  the	
   physical	
  locations	
  of	
  these	
  solutions	
  did	
  not	
  match	
  each	
  other.	
  Besides	
  the	
  top	
  ranked	
   solutions,	
  one	
  more	
  from	
  DOCK	
  (D2)	
  and	
  three	
  from	
  AutoDock	
  (AD2,	
  AD3	
  and	
  AD4)	
   were	
  analysed	
  based	
  on	
  their	
  docking	
  positions.	
   	
    	
    60	
    FIGURE 21. DOCK and AutoDock scoring bubble charts . (A) DOCK (Shoichet et al. 1992, Meng et al. 1992, Kuntz et al. 1982) grid scores for the docking of caffeine into RyR1ABC. The cluster number is reflected in the size of each bubble. The top five hits based on grid score were all located in the same pocket but with caffeine rotated, consequently for examination, they were averaged and their cluster numbers added to produce the last bubble (*) labelled D1 (cluster number = 8). In addition, D2 (cluster number = 1) was chosen for analysis. (B) The binding energies from AutoDock (Morris et al. 2009) are shown with cluster numbers represented by bubble size. Solutions AD1-4 were chosen for analysis. 	
    	
    	
    61	
    In	
  summary,	
  four	
  different	
  sites	
  were	
  identified	
  as	
  having	
  potential	
  for	
  caffeine	
   binding	
   (Figure	
   22A).	
   Site	
   1	
   contains	
   D1	
   and	
   AD2,	
   further	
   increasing	
   the	
   cluster	
   number	
   (Figure	
   22B).	
   Binding	
   involves	
   four	
   residues,	
   all	
   in	
   domain	
   B	
   and	
   therefore	
   agreeing	
  with	
  ITC	
  results	
  that	
  suggested	
  caffeine	
  bound	
  in	
  RyR1BC.	
  Site	
  2	
  also	
  houses	
   a	
  solution	
  from	
  each	
  computational	
  program	
  in	
  D2	
  and	
  AD3.	
  It	
  locates	
  at	
  the	
  interface	
   between	
  domains	
  A	
  and	
  B,	
  engaging	
  residues	
  from	
  both,	
  and	
  is	
  in	
  close	
  proximity	
  to	
   W269,	
  the	
  culprit	
  for	
  binding	
  in	
  the	
  RyR1BC	
  structure	
  (Figure	
  22C).	
  Site	
  3	
  represents	
   the	
  top	
  hit	
  from	
  AutoDock	
  (AD1)	
  and	
  suggests	
  only	
  residues	
  in	
  domain	
  A	
  are	
  required	
   for	
   binding	
   (Figure	
   22D).	
   Lastly,	
   site	
   4	
   places	
   caffeine	
   close	
   to	
   the	
   location	
   determined	
  by	
  the	
  RyR1ABC	
  caffeine-­‐soaked	
  crystal	
  structure	
  (Figure	
  22E).	
   With	
  our	
  preliminary,	
  rigid	
  docking,	
  one	
  of	
  the	
  sites	
  determined	
  does	
  stand	
  out	
   in	
  site	
  1.	
  As	
  mentioned,	
  it	
  is	
  the	
  only	
  site	
  that	
  does	
  not	
  involve	
  RyR1A,	
  and	
  therefore	
   agrees	
   with	
   our	
   ITC	
   results.	
   Adding	
   to	
   its	
   validity	
   is	
   the	
   presence	
   of	
   an	
   MH	
   causing	
   disease	
  mutation	
  R328W	
  (R329	
  in	
  Oryctolagus	
  cuniculus)(Loke	
  et	
  al.	
  2003),	
  which	
  is	
   located	
   in	
   very	
   close	
   proximity	
   to	
   this	
   caffeine	
   binding	
   site	
   (Figure	
   22B).	
   Caffeine	
   sensitivity	
  could	
  be	
  explained	
  in	
  this	
  disease	
  mutant	
  by	
  an	
  increase	
  in	
  affinity	
  for	
  the	
   drug	
   due	
   to	
   stacking	
   interactions	
   with	
   the	
   tryptophan.	
   Future	
   ITC	
   experiments	
   on	
   RyR1	
   ABC	
   mutants	
   in	
   site	
   1,	
   including	
   R329W,	
   will	
   hopefully	
   shed	
   light	
   on	
   the	
   physiological	
  and	
  functional	
  binding	
  site	
  of	
  caffeine	
  in	
  RyR1ABC.	
    	
    62	
    	
   FIGURE 22. Docking caffeine in silico. (A) A look at some docking solutions from both AutoDock (Morris et al. 2009) and DOCK (Shoichet et al. 1992, Meng et al. 1992, Kuntz et al. 1982) programs. Caffeine is coloured according to which method positioned it, AutoDock (dark blue), DOCK (yellow) or crystal structure (black). The results could be summarized into four locations: (B) Site 1, which contained D1 and AD2. The position of an R329 disease mutation site is marked by a black main-chain. (C) Site 2, which contained D2 with AD3 nearby in the same binding pocket. Not labelled are residues I267 and S268 directly behind the blue caffeine molecule. W269, the residue involved in binding caffeine in the RyR1BC structure is in close proximity to site 2. (D) AD1, the top solution from AutoDock. (E) AD4, near one of the sites seen in the RyR1ABC crystal structure. Interacting residues are labelled and domains are coloured blue (domain A) and green (domain B). Note: although the docking studies were performed using RyR1ABC, domain C is not shown for clarity. Boxes on (A) are drawn to indicate the rough positions of the sites highlighted in the following panels. Numbering: RyR1 Oryctolagus cuniculus. 	
    	
    	
   	
    	
   	
    63	
    3.2.4	
  ATP	
  Binding	
  Experiments	
   	
    	
    	
    Given	
   the	
   similarities	
   in	
   structure	
   between	
   different	
   purine	
   derivatives,	
   it	
    made	
  sense	
  to	
  test	
  binding	
  to	
  ATP	
  as	
  well.	
  Some	
  problems	
  though	
  halted	
  progress	
  in	
   determining	
   whether	
   or	
   not	
   the	
   coenzyme	
   bound	
   RyR1ABC.	
   Presumably	
   due	
   to	
   its	
   reactive	
  nature,	
  titrating	
  high	
  concentrations	
  of	
  ATP	
  into	
  buffer	
  alone	
  resulted	
  in	
  very	
   large	
  heat	
  signals	
  (data	
  not	
  shown).	
  As	
  a	
  result	
  more	
  dilute	
  solutions	
  had	
  to	
  be	
  used	
   in	
   ITC	
   measurements;	
   a	
   serious	
   problem	
   given	
   the	
   compound’s	
   millimolar	
   EC50	
   on	
   RyR	
  (Meissner	
  et	
  al.	
  1986).	
  	
   ITC	
   did	
   indicate	
   some	
   heat	
   that	
   is	
   significantly	
   different	
   from	
   background	
   dilution	
  experiments,	
  but	
  the	
  curve	
  was	
  too	
  shallow	
  to	
  fit	
  reliably	
  (Figure	
  23).	
  It	
  will	
   be	
   interesting	
   to	
   try	
   the	
   experiment	
   with	
   higher	
   concentrations	
   of	
   adenine	
   instead	
   of	
   ATP,	
  which	
  may	
  lead	
  to	
  smaller	
  dilution	
  heats.	
    	
    64	
    FIGURE 23. ATP binding to RyR1ABC . 5 mM ATP was titrated into 500 μM RyR1ABC and an affinity of ~1.3 mM was attained by ITC measurements. With the errors associated though, the thermodynamic values shown cannot be fully trusted. 	
    	
    	
    65	
    3.3	
   RyR1	
  4071-­4138	
   	
   Xiong	
   et	
   al.	
   (2006)	
   have	
   previously	
   described	
   a	
   construct	
   in	
   RyR1,	
   encompassing	
   residues	
   4064-­‐4210,	
   which	
   after	
   refolding,	
   binds	
   Ca2+,	
   CaV1.1	
   and	
   RyR1	
  CaMBDs.	
  A	
  few	
  main	
  areas	
  of	
  improvement	
  remain	
  from	
  their	
  work.	
  Firstly,	
  a	
   well-­‐behaved,	
   soluble	
   construct	
   that	
   does	
   not	
   require	
   refolding	
   would	
   strengthen	
   their	
   claims	
   of	
   specific	
   binding.	
   Secondly,	
   quantifying	
   the	
   thermodynamic	
   details	
   of	
   the	
   interactions,	
   using	
   ITC	
   for	
   example,	
   would	
   allow	
   us	
   to	
   test	
   whether	
   the	
   binding	
   affinity	
   is	
   in	
   a	
   physiological	
   range.	
   Ideally,	
   co-­‐crystallization	
   with	
   proposed	
   binding	
   partners	
   would	
   provide	
   the	
   most	
   insight.	
   Both	
   goals	
   rely	
   heavily	
   on	
   the	
   ability	
   to	
   obtain	
  a	
  soluble	
  domain	
  that	
  does	
  not	
  aggregate.	
  PHYRE2	
  (Kelley	
  and	
  Sternberg	
  2009)	
   was	
  implemented	
  again	
  to	
  aid	
  in	
  the	
  location	
  of	
  a	
  soluble	
  EF-­‐hand-­‐containing	
  domain.	
   Several	
   possible	
   constructs	
   were	
   tried	
   from	
   these	
   results,	
   based	
   on	
   alignment	
   with	
   actinin	
  (RyR1	
  ~3500-­‐4140),	
  Ca2+-­‐binding	
  protein	
  40	
  –	
  CaBP40	
  (RyR1	
  ~3830-­‐4140),	
   parvalbumin	
   (RyR1	
   ~4030-­‐4140)	
   and	
   CaM	
   (RyR1	
   ~3980-­‐4140).	
   A	
   summary	
   of	
   the	
   successes	
   and	
   failures	
   in	
   purification	
   are	
   described	
   in	
   Table	
   6	
   and	
   a	
   schematic	
   that	
   sums	
   up	
   the	
   PHYRE2	
   results	
   is	
   shown	
   in	
   Figure	
   24A.	
   Constructs	
   were	
   tested	
   for	
   expression	
  (Figure	
  24B)	
  and	
  solubility	
  by	
  small-­‐scale	
  purification	
  of	
  ~30	
  ml	
  cultures	
   on	
  cobalt	
  gravity	
  columns.	
  Successful	
  constructs	
  were	
  scaled	
  up	
  and	
  purified	
  by	
  fast	
   protein	
  liquid	
  chromatography.	
  	
   	
   	
   	
   	
   	
   	
   	
   	
   	
   	
  	
    	
    66	
    Construct	
    Expression	
    Solubility	
    Purification	
    3487/3497/3507	
   Slightly	
    All	
  nine	
    Not	
  soluble	
    -­‐  constructs	
  not	
    30°C,	
  ~5-­‐6	
  hrs	
    4128/4138/4148	
   induction.	
    soluble	
    3818/3828/3838	
   Slightly	
    All	
  nine	
    -­‐	
    constructs	
  not	
    30°C,	
  ~5-­‐6	
  hrs	
    4128/4138/4148	
   induction	
    soluble	
    3997-­‐4070	
    Soluble	
    3997-­‐4138	
    Success	
    Not	
  Soluble	
    Difficult	
  to	
  purify-­‐	
  does	
  not	
  absorb	
  at	
    30°C,	
  ~5-­‐6	
  hrs	
    280	
  nm	
  	
    induction	
    Crashes	
  out	
  in	
  concentrator	
    Success	
    Soluble	
    30°C,	
  ~5-­‐6	
  hrs	
    Multiple	
  species	
  in	
  gel	
  filtration	
   Variable	
  success	
  on	
  HQ	
    induction	
   4024/4031/4041	
   Success	
    All	
  nine	
    -­‐	
    constructs	
  are	
   soluble	
  (see	
  below)	
    30°C,	
  ~5-­‐6	
  hrs	
    RyR1	
  4041-­‐4138	
  was	
  the	
  most	
    4128/4138/4148	
   induction	
    soluble	
    4041-­‐4138	
    Soluble	
    Successfully	
  purified,	
  but	
  low	
  yield	
    Soluble	
    A	
  lot	
  of	
  pure,	
  soluble	
  protein	
    Success	
   30°C,	
  ~5-­‐6	
  hrs	
   induction	
    4071-­‐4138	
    Success	
   30°C,	
  ~5-­‐6	
  hrs	
    attainable	
    induction	
    TABLE 6. EF-hand constructs. A list of the different constructs attempted in the search for an EF-hand-containing domain.  	
    67	
    FIGURE 24. Expression of PHYRE guided domains . (A) A schematic describing 2  2  the constructs attempted using homology information from PHYRE (Kelley and Sternberg 2009). Forward (bright red) and reverse (bright green) primers for construct PCR are numbered according to the first (forward) or last (reverse) amino acid they code for. For systematic primers (± ~10 residues), only the middle one is labelled by number (see Table 1 for a list of primers). (B) An example gel showing the expression of some of the constructs tried in the search for RyR1 EF-hands. The molecular weights (MW) of the fusion proteins range from ~54 kDa (RyR1 4041-4128 + MBP) to ~58 kDa (RyR1 4031-4148). Numbering: RyR1 Oryctolagus cuniculus. 	
   	
    	
    	
   	
    	
    68	
    The	
   most	
   promising	
   constructs	
   showed	
   homology	
   to	
   CaM:	
   Individual	
   CaM	
   lobes	
  (PDB	
  1CLL,	
  Chattopadhyaya	
  et	
  al.	
  1992)	
  could	
  be	
  superposed	
  onto	
  a	
  RyR1	
  EF-­‐ hand	
  model	
  generated	
  from	
  PHYRE2	
  (Kelley	
  and	
  Sternberg	
  2009)	
  (Figure	
  25A).	
  From	
   the	
   structural	
   alignment	
   we	
   could	
   analyze	
   the	
   potential	
   for	
   Ca2+-­‐binding	
   in	
   the	
   predicted	
  EF-­‐hands.	
  Of	
  the	
  four	
  modelled,	
  only	
  the	
  most	
  C-­‐terminal	
  one	
  (EF-­‐hand	
  4)	
   portrays	
   a	
   probable	
   binding	
   site.	
   In	
   the	
   others,	
   hydrophobic	
   or	
   basic	
   residues	
   position	
   themselves	
   in	
   locations	
   that	
   would	
   need	
   to	
   co-­‐ordinate	
   Ca2+	
   (Figure	
   25B).	
   Looking	
  closer	
  at	
  the	
  different	
  lobes,	
  we	
  were	
  able	
  to	
  predict	
  the	
  hurdles	
  that	
  would	
   have	
   to	
   be	
   overcome	
   to	
   bind	
   Ca2+	
   in	
   the	
   first	
   three	
   EF-­‐hands.	
   In	
   addition	
   to	
   the	
   mentioned	
  hydrophobic	
  and	
  basic	
  clashes,	
  this	
  included	
  for	
  example,	
  the	
  presence	
  of	
   positively	
   charged	
   residues	
   in	
   the	
   vicinity	
   of	
   EF-­‐hand	
   3,	
   which	
   could	
   obstruct	
   Ca2+	
   ions	
  from	
  access	
  (Figures	
  25C	
  and	
  25D).	
  Analysing	
  the	
  superposition,	
  we	
  saw	
  that	
  the	
   C-­‐lobe	
  of	
  CaM	
  lined	
  up	
  with	
  RyR1	
  4071-­‐4134,	
  a	
  finding	
  that	
  eventually	
  resulted	
  in	
  the	
   RyR1	
  4071-­‐4138	
  construct.	
   RyR1	
   4017-­‐4138	
   expressed	
   and	
   purified	
   very	
   well	
   yielding	
   pure,	
   soluble	
   protein	
   and	
   confirmed	
   by	
   mass	
   spectrometry	
   to	
   be	
   the	
   correct	
   molecular	
   weight	
   (8111	
   kDa).	
   The	
   construct	
   did	
   however	
   readily	
   dimerize	
   in	
   the	
   absence	
   of	
   a	
   reducing	
   agent	
  as	
  shown	
  on	
  a	
  gel	
  filtration	
  column	
  as	
  well	
  as	
  by	
  SDS-­‐PAGE	
  (Figures	
  26A	
  and	
   26B).	
   A	
   mutant	
   C4114A	
   form	
   removed	
   the	
   only	
   cysteine	
   residue	
   and	
   resulted	
   in	
   a	
   purifiable	
   monomer	
   even	
   in	
   the	
   absence	
   of	
   a	
   reducing	
   agent.	
   RyR1	
   4071-­‐4138	
   was	
   used	
   for	
   ITC	
   experiments	
   and	
   kept	
   permanently	
   in	
   a	
   reducing	
   condition.	
   Random	
   crystallographic	
  screens	
  were	
  applied	
  to	
  both	
  the	
  wild-­‐type	
  and	
  C4114A	
  constructs.	
   So	
  far,	
  there	
  are	
  no	
  promising	
  crystallographic	
  hits	
  to	
  report.	
   	
    	
    69	
    	
   FIGURE 25. Potential EF-hands in RyR . (A) A model of RyR ~4000-4140 from PHYRE (Kelley and Sternberg 2009) is coloured in a spectrum from its N-teminus (blue) to 2  C-terminus (red) and superposed to CaM (PDB 1CLL, Chattopadhyaya et al. 1992), 2+  coloured grey. (B) Structural alignments of the EF-hands. Ca binding residues in the CaM sequences are highlighted green, as are compatible residues in RyR1. Positively charged (bold) and hydrophobic residues in Ca binding positions are coloured red, and in addition, 2+  2+  two lysines (K4090 and K4091), which may hinder access to Ca are bolded. The highlighted residues are shown as sticks in zoomed in views of the N-lobe (C) and C-lobe (D) superpositions. Ca ions are shown as green spheres. Numbering: RyR1 Oryctolagus 2+  cuniculus. 	
    70	
    	
    FIGURE 26. Cysteine dimerisation in RyR1 4071-4138 . (A) Wild-type (WT) RyR1 4071-4138 was run on a gel filtration column without βME and showed a showed a shoulder peak. The cysteine-4114 alanine mutant ran as a monomer in the absence of a reducing agent. (B) In the presence of a strong reducing agent (DTT), fractions (Frn) from the superdex column in (A) ran as a ~8 kDa monomer on an SDS PAGE gel. Numbering: RyR1 Oryctolagus cuniculus.  	
    71	
    3.3.1	
  Ca2+	
  and	
  Mg2+	
  Binding	
  to	
  RyR1	
  EF-­hands	
   	
   	
    ITC	
  detected	
  very	
  weak	
  binding	
  between	
  RyR1	
  4071-­‐4138	
  and	
  the	
  Ca2+	
  (Figure	
    27A)	
  as	
  well	
  as	
  Mg2+	
   (Figure	
  27B).	
  The	
  ~10	
  mM	
  Kd	
  values	
  indicate	
  an	
  affinity	
  much	
   lower	
  than	
  something	
  physiologically	
  relevant	
  but	
  an	
  explanation	
  may	
  be	
  suggested	
   in	
   construct	
   selection.	
   Neighbouring	
   domains	
   could	
   stablilize	
   the	
   EF-­‐hand	
   and	
   enhance	
  its	
  binding	
  to	
  both	
  Ca2+	
   and	
  Mg2+.	
  From	
  our	
  experiments,	
  although	
  absolute	
   thermodynamic	
   values	
   cannot	
   be	
   quoted	
   with	
   certainty,	
   we	
   can	
   say	
   that	
   the	
   ions	
   definitely	
   bind	
   RyR1	
   4071-­‐4138.	
   The	
   raw	
   endothermic	
   heats	
   of	
   binding	
   observed	
   were	
   in	
   sharp	
   contrast	
   to	
   the	
   exothermic	
   background	
   titrations	
   of	
   Ca2+/Mg2+	
   into	
   buffer.	
    	
   FIGURE 27. RyR1 4071-4138 binds Ca and Mg . Weak binding was detectable by 2+  2+  2+  2+  2+  2+  ITC for RyR1 4071 with both Ca (A) and Mg (B). In both cases, 30 mM of Ca /Mg was titrated into 1.5 mM RyR1 4071-4138. Given the weak affinity, the absolute thermodynamic values, although shown here, cannot be taken to be accurate. 	
    72	
    	
   Binding	
   to	
   Ca2+	
   was	
   further	
   tested	
   using	
   nuclear	
   magnetic	
   resonance	
   experiments.	
   U.	
   Brath	
   compared	
   Heteronuclear	
   Single	
   Quantum	
   Coherence	
   spectra	
   from	
   15N-­‐labelled	
   RyR1	
   4071-­‐4138	
   with	
   and	
   without	
   Ca2+.	
   The	
   dramatic	
   change	
   observed	
  confirmed	
  that	
  Ca2+	
  does	
  indeed	
  bind	
  the	
  protein	
  (data	
  not	
  shown).	
   	
   	
    3.3.2	
  Interaction	
  with	
  RyR1	
  CaMBDs	
   	
   RyR1	
   4071-­‐4138	
   was	
   identified	
   as	
   homologous	
   to	
   CaM.	
   In	
   addition,	
   RyR1	
   4064-­‐4210	
   has	
   been	
   shown	
   to	
   bind	
   to	
   the	
   second	
   CaMBD	
   (in	
   sequence)	
   in	
   RyR1	
   3614-­‐3643	
   (Xiong	
   et	
   al.	
   2006).	
   We	
   therefore	
   wished	
   to	
   perform	
   binding	
   experiments	
   to	
  additional	
  CaMBDs	
  that	
  we	
  have	
  previously	
  characterized	
  in	
  the	
  lab.	
  	
   A	
   synthesized	
   RyR1	
   CaMBD3dLRR	
   peptide	
   was	
   available	
   in	
   the	
   lab	
   for	
   ITC	
   experiments	
   with	
   RyR1	
   4071-­‐4138.	
   Preliminary	
   ITC	
   experiments	
   have	
   showed	
   an	
   affinity	
   of	
   ~7µM	
   in	
   conditions	
   without	
   Ca2+	
   (Figure	
   28).	
   The	
   success	
   in	
   binding	
   has	
   opened	
  a	
  multitude	
  of	
  opportunities	
  for	
  the	
  construct.	
  	
   	
    	
    73	
    	
    FIGURE 28. RyR1 4071-4138 binds a CaMBD3 mutant . 1.2 mM RyR1 4071-4138 was titrated into 0.12 mM of a RyR1 CaMBD3 mutant peptide. The concentration of the peptide was not accurately measured, perhaps explaining the N value of ~1.5. The thermodynamic values obtained from ITC are displayed on the graph.  	
    74	
    4	
   Discussion 	
   	
    4.1 RyR1	
  ABCd	
   	
   The	
   addition	
   of	
   a	
   small	
   helix,	
   and	
   the	
   structuring	
   of	
   previously	
   flexible	
   loops	
   are	
   more	
   than	
   just	
   requirements	
   for	
   crystal	
   lattice	
   formation.	
   The	
   relative	
   orientations	
   of	
   these	
   new	
   loops	
   may	
   just	
   be	
   crystallographic	
   artefacts,	
   but	
   it	
   is	
   worth	
   noting	
  that	
  they	
  can	
  adopt	
  structure	
  when	
  in	
  proximity	
  to	
  hydrophilic	
  domains.	
  Given	
   the	
   observed	
   polar	
   and	
   electrostatic	
   interactions	
   between	
   domains	
   in	
   RyR1ABC,	
   it	
   can	
  be	
  concluded	
  that	
  this	
  sort	
  of	
  dynamic	
  structuring	
  must	
  take	
  place	
  in	
  full-­‐length	
   RyR1	
  allosteric	
  communication.	
   Still,	
  looking	
  at	
  the	
  big	
  picture,	
  does	
  the	
  structure	
  presented	
  here	
  expand	
  the	
   understanding	
  of	
  RyR1?	
  Majority	
  of	
  the	
  structure	
  is	
  already	
  known	
  in	
  RyR1ABC,	
  the	
   theory	
   of	
   disease	
   mutations	
   affecting	
   interface	
   interactions	
   is	
   repeated	
   here,	
   and	
   in	
   comparing	
   the	
   structure	
   docked	
   in	
   the	
   open	
   and	
   closed	
   EM	
   maps,	
   the	
   observed	
   domain	
   movements	
   are	
   very	
   similar	
   to	
   those	
   seen	
   with	
   the	
   first	
   three	
   domains	
   alone.	
   Described	
   here	
   however,	
   are	
   the	
   exciting,	
   new	
   conclusions	
   that	
   can	
   be	
   drawn	
   from	
   this	
  structure.	
   	
   	
    4.1.1	
  A	
  Larger	
  Surface	
  Area	
  for	
  Binding	
   	
   For	
  a	
  channel	
  to	
  be	
  well	
  regulated,	
  there	
  needs	
  to	
  be	
  a	
  relative	
  ease	
  of	
  access	
   for	
  small-­‐molecule	
  modulators	
  to	
  bind.	
  Although	
  the	
  extended	
  helix	
  only	
  represents	
   ~9%	
   of	
   the	
   structure,	
   docking	
   studies	
   suggest	
   that	
   a	
   large	
   majority	
   of	
   it	
   is	
   exposed	
   to	
   the	
  cytoplasm.	
  This	
  enhances	
  the	
  potential	
  for	
  finding	
  regulatory	
  binding	
  sites	
  using	
   ITC	
   as	
   a	
   high-­‐throughput	
   screen.	
   In	
   the	
   search	
   for	
   RyR	
   ligands,	
   it	
   makes	
   sense	
   to	
   consult	
  a	
  related	
  channel	
  in	
  IP3R.	
    	
    75	
    IP3R	
   and	
   RyR	
   have	
   travelled	
   down	
   similar	
   evolutionary	
   roads	
   and	
   still	
   today	
   share	
   similar	
   functions	
   as	
   tightly-­‐regulated	
   Ca2+	
   channels.	
   Proving	
   this,	
   Seo	
   et	
   al.	
   (2012)	
   completely	
   replaced	
   the	
   N-­‐terminal	
   domain	
   of	
   an	
   IP3R	
   with	
   RyR1A	
   and	
   reproduced	
   a	
   functional	
   channel.	
   Similarly	
   Yuchi	
   and	
   Van	
   Petegem	
   (2010)	
   have	
   done	
   some	
   preliminary	
   structural	
   analysis	
   on	
   the	
   two	
   N-­‐termini	
   and	
   concluded	
   that	
   that	
   they	
   operate	
   with	
   parallel	
   allosteric	
   mechanisms.	
   Consequently,	
   anything	
   we	
   can	
   structurally	
  determine	
  from	
  one	
  channel	
  can	
  be	
  translated	
  to	
  extend	
  the	
  knowledge	
   in	
   the	
   field	
   of	
   the	
   other.	
   Compared	
   to	
   the	
   structure	
   of	
   IP31ABC,	
   we	
   learn	
   that	
   the	
   structured	
   residues	
   presented	
   here	
   (~10-­‐577)	
   define	
   a	
   better	
   ending	
   point	
   for	
   domain	
  C	
  than	
  we	
  previously	
  described	
  (Tung	
  et	
  al.	
  2010)	
  (Figure	
  29A).	
  The	
  solved	
   structure	
  will	
  therefore	
  be	
  referred	
  to	
  as	
  RyR1ABCd.	
  The	
  binding	
  sites	
  of	
  regulators	
   to	
   IP3R1	
   can	
   help	
   in	
   locating	
   similar	
   sites	
   in	
   RyR.	
   PDB	
   1N4K	
   (Bosanac	
   et	
   al.	
   2002)	
   represents	
   a	
   high-­‐resolution	
   structure	
   of	
   IP3RBC	
   in	
   which	
   its	
   most	
   important	
   triggering	
   ligand,	
   IP3	
   itself,	
   is	
   bound.	
   Given	
   the	
   potency	
   of	
   IP3	
   on	
   IP3R,	
   a	
   ligand	
   that	
   could	
  bind	
  in	
  the	
  same	
  pocket	
  in	
  RyR,	
  could	
  be	
  of	
  serious	
  importance.	
   However,	
   RyR1	
   is	
   not	
   modulated	
   by	
   IP3,	
   and	
   analysis	
   of	
   the	
   superposed	
   structures	
   helps	
   to	
   explain	
   why.	
   In	
   the	
   binding	
   pocket,	
   only	
   two	
   residues	
   are	
   conserved	
   between	
   the	
   receptors	
   (Figure	
   29B):	
   in	
   IP3R,	
   arginine-­‐504	
   and	
   arginine-­‐ 568,	
   both	
   of	
   which	
   are	
   involved	
   in	
   binding	
   the	
   1-­‐phosphate	
   of	
   IP3,	
   with	
   the	
   latter	
   being	
  essential	
  for	
  it	
  (Yoshikawa	
  et	
  al.	
  2006).	
  In	
  human	
  RyR1	
  these	
  represent	
  R471	
   and	
   R533.	
   Fascinatingly,	
   R533	
   has	
   twice	
   been	
   documented	
   to	
   cause	
   malignant	
   hyperthermia	
  as	
  a	
  result	
  of	
  a	
  mutation	
  to	
  a	
  histidine	
  (Brandt	
  et	
  al.	
  1999)	
  or	
  a	
  cysteine	
   (Tammaro	
  et	
  al.	
  2003).	
  Here	
  the	
  molecular	
  mechanism	
  of	
  disease	
  can	
  be	
  explained	
  by	
   directly	
   affecting	
   the	
   binding	
   of	
   a	
   crucial	
   modulator.	
   Finding	
   the	
   modulator	
   in	
   RyR	
   though,	
  is	
  yet	
  to	
  be	
  done;	
  something	
  far	
  from	
  trivial.	
  In	
  RyR1,	
  the	
  positively	
  charged	
   IP3	
  pocket	
  has	
  been	
  altered	
  dramatically,	
  both	
  in	
  sterics	
  and	
  electrostatics.	
   	
    	
    76	
    FIGURE 29. A comparison between IP R and RyR1 1-617 . (A) Superposition of 3  the 2.2 Å IP R1BC (PDB 1N4K, Bosanac et al. 2002) onto RyR1 1-617 shows that in the 3  search for a fourth domain, we have actually redefined domain C. The extra structural information from ‘RyR1ABCd’ provides more of a surface for potential binders. RyR1 A, B, C and d domains are coloured blue, green, red and yellow respectively and the IP R is 3  shown as grey. (B) A detailed view of the residues involved in binding IP . In the structural 3  alignments, residues that directly interact with IP in IP R are bolded and underlined. Shown 3  3  in bold, underlined and marked by an asterisk (*), are RyR1 R471 and R533: the only conserved residues involved in binding. R533 is also a disease mutation (black sticks). Numbering: RyR1 Homo sapiens. 	
   	
    4.1.2	
  PP1	
  Recruitment	
  by	
  a	
  Leucine	
  Zipper	
   	
   	
    Leucine	
   zippers	
   (LZ)	
   are	
   coiled	
   coils	
   that	
   force	
   the	
   adhesion	
   of	
   two	
   or	
   more	
    parallel	
   (common)	
   or	
   antiparallel	
   (rare)	
   α-­‐helices	
   (Lupas,	
   1996,	
   Kohn	
   et	
   al.	
   1997).	
   Originally	
   found	
   in	
   DNA-­‐binding	
   proteins,	
   they	
   described	
   a	
   mechanism	
   for	
   dimerisation	
   between	
   two	
   LZ-­‐containing	
   proteins.	
   Orthodox	
   descriptions	
   define	
   a	
   heptad	
  repeat	
  (abcdefg)	
  as	
  having	
  hydrophobic	
  residues	
  at	
  positions	
  ‘a’	
  and	
  ‘d’	
  with	
    	
    77	
    leucine	
   residues	
   occupying	
   the	
   latter	
   (Crick,	
   1953,	
   Landschulz	
   et	
   al.	
   1988).	
   Isoleucine	
   and	
  valine	
  have	
  also	
  been	
  accepted	
  as	
  residues	
  that	
  can	
  contribute	
  to	
  this	
  ‘zipping’.	
   Marx	
  et	
  al.	
  (2001)	
  manually	
  identified	
  three	
  LZ	
  motifs	
  in	
  RyR2.	
  LZ1,	
  in	
  residues	
   555-­‐604	
   was	
   shown	
   to	
   bind	
   Protein	
   Phosphatase	
   1	
   (PP1)	
   via	
   its	
   regulatory	
   subunit	
   spinophilin.	
   LZ2,	
   defined	
   by	
   RyR2	
   1603-­‐1631	
   pulled	
   down	
   Protein	
   Phosphatase	
   2A	
   (PP2A),	
  and	
  LZ3	
  (RyR2	
  3003-­‐3039)	
  interacted	
  with	
  PKA.	
  They	
  went	
  on	
  to	
  show	
  that	
   LZ1	
   and	
   LZ3	
   are	
   conserved	
   between	
   the	
   cardiac	
   and	
   skeletal	
   RyR	
   and	
   as	
   expected,	
   RyR1	
  fusion	
  peptides	
  containing	
  these	
  zippers	
  pull	
  down	
  PP1	
  and	
  PKA	
  as	
  well.	
  A	
  LZ-­‐ motif	
   in	
   spinophilin	
   (residues	
   485-­‐510)	
   was	
   classified	
   as	
   the	
   binding	
   site	
   for	
   RyR2,	
   and	
  mutating	
  out	
  ‘d’	
  leucines	
  in	
  either	
  RyR2	
  or	
  spinophilin	
  abolished	
  the	
  interaction.	
   LZ1	
   in	
   RyR1	
   corresponds	
   to	
   residues	
   534-­‐592,	
   almost	
   exactly	
   describing	
   the	
   gain	
  in	
  structure	
  due	
  to	
  RyR1ABCd.	
  The	
  structure	
  agrees	
  well	
  with	
  their	
  prediction,	
   lining	
   hydrophobic	
   residues	
   for	
   dimer	
   interactions	
   down	
   one	
   face	
   of	
   the	
   new	
   helix	
   and	
  scattering	
  charged	
  and	
  polar	
  amino	
  acids	
  on	
  the	
  other	
  side	
  (Figure	
  30).	
  	
   The	
   hydrophilic	
   face	
   contains	
   several	
   RyR1	
   and	
   RyR2	
   disease	
   mutations	
   that	
   cause	
  MH	
  or	
  CPVT	
  respectively.	
  This	
  side	
  of	
  the	
  coil	
  could	
  therefore	
  play	
  an	
  important	
   role	
  in	
  anchoring	
  the	
  helix	
  in	
  position,	
  primed	
  for	
  PP1	
  attachment.	
  Phosphatases	
  are	
   essential	
   players	
   in	
   the	
   inhibition	
   of	
   RyR	
   channels	
   (Marx	
   et	
   al.	
   2000).	
   Hyperphosphorylation	
   of	
   RyR	
   is	
   thought	
   to	
   cause	
   FKBP	
   dissociation	
   and	
   result	
   in	
   leaky	
  channels	
  (Bellinger	
  et	
  al.	
  2009).	
  From	
  our	
  results,	
  we	
  can	
  propose	
  yet	
  another	
   mechanism	
  for	
  disease	
  in	
  RyR1.	
  Mutations	
  in	
  RyR1d	
  could	
  directly	
  affect	
  the	
  binding	
   of	
  the	
  phosphatase	
  PP1,	
  required	
  in	
  recovery	
  from	
  hyperphosphorylation	
  (Marx	
  et	
  al.	
   2000).	
   In	
   wild-­‐type	
   channels	
   then,	
   RyR1d	
   may	
   serve	
   as	
   the	
   recruitment	
   domain	
   for	
   PP1,	
   transiently	
   increasing	
   the	
   phosphatase’s	
   local	
   concentration	
   near	
   sites	
   of	
   requirement.	
  	
    	
    78	
    	
   FIGURE 30. A LZ motif in RyR1d . Below the sequence and marked by up-arrows (↑) are human disease mutation sites. Above the sequence and marked by down-arrows (↓) are the valine or leucine residues that are arranged in the correct canonical spacing for the formation of a leucine zipper. In addition, other hydrophobic residues that could contribute to a zipper are bolded and underlined in the sequence. Hydrophobic, basic and acidic residues have their transparent surfaces coloured green, blue and red respectively. Disease mutations locate on the opposite helical side to potential zipper ‘action’, requiring a 180° rotation of the structure in (A) to view them (B). Numbering: RyR1 (black) or RyR2 (grey)  Homo sapiens. Interestingly,	
   upon	
   analysis	
   of	
   the	
   crystal	
   lattice,	
   we	
   observed	
   that	
   the	
   additional	
   helix	
   in	
   our	
   structure	
   contacts	
   a	
   symmetry-­‐related	
   molecule	
   through	
   a	
   coiled-­‐coil	
   interaction	
   (Figure	
   31).	
   The	
   observed	
   LZ	
   differs	
   slightly	
   from	
   the	
   strict	
   definition	
   of	
   interacting	
   heptads,	
   hinting	
   that	
   some	
   degree	
   of	
   plasticity	
   may	
   be	
   involved	
  in	
  RyR1	
  modulator	
  recruitment.	
   	
    79	
    FIGURE 31. An antiparallel LZ in the RyR1ABCd crystal structure. (A) and (B) show two perpendiuclar views of a coiled coil formed between RyR1d domains from different asymmetric units in the crystal lattice. Leucine, Isoleucine and valine residues involved in ‘zipping’ are green or lime green and are aligned in the sequence between (A) and (B) to the black helices. An unusual bend occurs at a polar S556-S557 gap (red) between canonical LZ heptads (HR1 and HR2). (C) The bend is allowed due to a possible register shift employing the HR2’ LZ residues in lime green. (D) A schematic explaining the atypical interactions in the zipper. L553 initiates a zipper with L560 in the other helix (asterisk). Seven positions from there on both helices, V546 in HR1 would be expected to interact with L567 in HR2, but it instead is closer to V566 in HR2’. Numbering: RyR1  Homo sapiens. 	
    80	
    4.1.3	
  Molecular	
  Insight	
  from	
  Docking	
    	
   	
   	
    There	
  are	
  two	
  related	
  hurdles	
  to	
  the	
  theory	
  of	
  LZ-­‐mediated	
  dephosphorylation	
    of	
   RyR1	
   by	
   PP1/spinophilin.	
   Firstly,	
   PP1	
   (37	
   kDa)	
   and	
   spinophilin	
   (90	
   kDa),	
   together	
   form	
   a	
   large,	
   bulky	
   complex	
   and	
   secondly,	
   considerable	
   structural	
   rearrangements	
   would	
   be	
   required	
   to	
   allow	
   for	
   dimerisation	
   by	
   coiled-­‐coils.	
   As	
   such,	
   a	
   special	
   right	
   of	
   entry	
  may	
  be	
  necessary	
  for	
  recruitment.	
  In	
  analysis	
  of	
  RyR1ABCd	
  docked	
  into	
  an	
  EM	
   map	
  (EMDB	
  1606,	
  Samso	
  et	
  al,	
  2009)	
  from	
  the	
  cytoplasmic	
  surface	
  (Figure	
  32A)	
  we	
   see	
  such	
  a	
  portal,	
  or	
  Phosphatase	
  Access	
  Channel	
  (PAC)	
  (Figure	
  32B).	
   	
    From	
  further	
  docking	
  investigation,	
  new	
  interaction	
  surfaces	
  within	
  RyR1	
  have	
    been	
  characterized.	
  Previously,	
  we	
  have	
  described	
  six	
  interfaces	
  in	
  the	
  full-­‐length	
  RyR	
   structure	
  that	
  gear	
  against	
  facial	
  planes	
  in	
  RyR1ABC:	
  Interface	
  1	
  marks	
  the	
  junction	
   between	
  RyR1A	
  of	
  one	
  monomer	
  and	
  RyR1B	
  of	
  the	
  next	
  in	
  the	
  tetrameric	
  assembly.	
   Interfaces	
  2	
  and	
  3	
  describe	
  respectively	
  as	
  the	
  lateral	
  junctions	
  of	
  RyR1A	
  and	
  RyR1C	
   to	
  adjacent	
  domains	
  in	
  the	
  full	
  channel.	
  And	
  interfaces	
  4,	
  5	
  and	
  6	
  link	
  RyR1A,	
  RyR1B	
   and	
   RyR1C	
   ‘downwards’	
   in	
   the	
   direction	
   of	
   the	
   channel	
   pore	
   (Tung	
   et	
   al.	
   2010).	
   Docking	
   RyR1ABCd	
   has	
   redefined	
   interface	
   3	
   and	
   brought	
   to	
   light	
   some	
   details	
   about	
   a	
  novel	
  interaction	
  surface	
  in	
  interface	
  7.	
  	
   The	
  RyR1	
  MH/CCD	
  causing	
  mutations	
  Q474H,	
  Y522C/S	
  and	
  R530H,	
  as	
  well	
  as	
   the	
  CPVT	
  mutation	
  in	
  RyR2	
  V507I,	
  were	
  all	
  previously	
  predicted	
  to	
  affect	
  interface	
  3	
   (Tung	
  et	
  al.	
  2010).	
  In	
  addition,	
  three	
  further	
  MH	
  mutations	
  in	
  RyR1	
  (R533C/H,	
  D544Y	
   and	
   R552W)	
   and	
   one	
   in	
   RyR2	
   that	
   causes	
   CPVT	
   (A549V)	
   can	
   be	
   mapped	
   onto	
   the	
   RyR1ABCd	
   structure	
   for	
   analysis	
   (Figure	
   32C).	
   An	
   intricate	
   web	
   of	
   polar	
   and	
   electrostatic	
   interactions	
   decorates	
   the	
   junction	
   between	
   RyR1C	
   and	
   RyR1d.	
   Two	
   props	
  from	
  below,	
  in	
  RyR1C	
  stabilize	
  the	
  LZ	
  raised	
  arm:	
  Q474	
  and	
  Y522.	
  The	
  former	
   interacts	
   with	
   N532,	
   the	
   very	
   first	
   residue	
   in	
   RyR1d,	
   and	
   the	
   latter	
   pokes	
   up	
   from	
   domain	
   C	
   below,	
   into	
   the	
   RyR1C-­‐d	
   hydrophilic	
   interaction	
   network.	
   On	
   the	
   inner	
   side	
   of	
  this	
  arm,	
  RyR1d	
  contributes	
  its	
  own	
  variety	
  of	
  hydrophilic	
  binders.	
  These	
  include	
   R552	
  and	
  surprisingly	
  a	
  rogue	
  tryptophan	
  that	
  has	
  abandoned	
  its	
  hydrophobic	
  nature	
   to	
  hydrogen	
  bond	
  via	
  its	
  N1	
  atom	
  (Figure	
  32C).	
    	
    81	
    RyR1	
  D544	
  and	
  RyR2	
  A549	
  lie	
  at	
  interface	
  3,	
  at	
  the	
  shoulder	
  of	
  the	
  RyR1d	
  arm.	
   Mutations	
   at	
   these	
   sites	
   that	
   cause	
   disease	
   introduce	
   more	
   bulky	
   and	
   hydrophobic	
   side-­‐chains	
   in	
   tyrosine	
   and	
   valine	
   respectively.	
   This	
   corroborates	
   well	
   with	
   our	
   earlier	
   studies	
   that	
   blamed	
   disease	
   on	
   a	
   breakdown	
   of	
   hydrophilic	
   communication	
   between	
  domains	
  in	
  RyR	
  (Tung	
  et	
  al.	
  2010).	
   The	
   mapped	
   RyR2	
   V507	
   is	
   shown	
   as	
   a	
   black	
   surface	
   for	
   clarity	
   in	
   figure	
   32B	
   and	
   grey	
   sticks	
   in	
   figure	
   32C.	
   Mutation	
   to	
   an	
   isoleucine	
   causes	
   CPVT,	
   emphasizing	
   that	
  even	
  very	
  small	
  perturbations	
  in	
  an	
  interface	
  are	
  large	
  enough	
  to	
  cause	
  serious	
   problems.	
  This	
  demonstrates	
  the	
  fragility	
  of	
  the	
  interactions	
  in	
  the	
  area.	
  Evidently	
  a	
   stable	
  LZ-­‐motif	
  in	
  RyR2d	
  as	
  well,	
  is	
  an	
  absolute	
  requirement.	
  For	
  RyR	
  in	
  general	
  then,	
   our	
   mutational	
   analysis	
   suggests	
   that	
   a	
   well-­‐positioned	
   LZ	
   arm	
   is	
   required	
   for	
   the	
   proper	
  function	
  and	
  regulation	
  of	
  RyR.	
  	
   Finally,	
   our	
   docking	
   studies	
   have	
   allowed	
   the	
   definition	
   of	
   a	
   new	
   interaction	
   plane	
  in	
  interface	
  7	
  (Figures	
  32B	
  and	
  32E).	
  This	
  can	
  be	
  seen	
  as	
  the	
  elbow	
  of	
  the	
  LZ	
   arm	
   and	
   pokes	
   out	
   in	
   a	
   radial	
   direction	
   towards	
   the	
   clamp	
   region.	
   It	
   must	
   be	
   noted	
   though,	
   that	
   the	
   absolute	
   location	
   seen	
   in	
   the	
   crystal	
   structure	
   of	
   this	
   elbow	
   may	
   be	
   a	
   result	
  of	
  LZ	
  dimerisation	
  and	
  not	
  truly	
  reflect	
  a	
  physiological	
  positioning.	
  In	
  support	
   of	
   this,	
   the	
   forearm,	
   past	
   the	
   elbow,	
   extrudes	
   out	
   of	
   the	
   EM	
   map	
   into	
   the	
   PAC	
   (figures	
   32B	
  and	
  15B).	
  The	
  general	
  location	
  of	
  interface	
  7	
  though,	
  is	
  undeniable	
  and	
  supplies	
   the	
   first	
   piece	
   of	
   structural	
   evidence	
   for	
   communication	
   between	
   the	
   clamp	
   and	
   central	
   rim.	
   With	
   this	
   we	
   can	
   begin	
   to	
   understand	
   how	
   modulators	
   that	
   bind	
   in	
   the	
   periphery	
  or	
  the	
  RyR	
  cytoplasmic	
  face	
  can	
  affect	
  the	
  channel	
  close	
  to	
  the	
  pore.	
  RyR1d	
   can	
  therefore	
  be	
  designated	
  two	
  hypothetical	
  roles.	
  The	
  first	
  involves	
  a	
  LZ	
  arm-­‐grip	
   with	
   spinophilin,	
   an	
   interaction	
   that	
   would	
   mediate	
   the	
   dephosphorylation	
   of	
   the	
   channel,	
  and	
  the	
  second	
  is	
  as	
  a	
  vestibule-­‐clamp	
  translator	
  that	
  sets	
  up	
  the	
  potential	
   for	
  allosteric	
  communication.	
    	
    82	
    FIGURE 32. RyR1ABCd docking analysis. Legend on next page.  	
    83	
    FIGURE 32. RyR1ABCd docking analysis . (A), (B) and (C) represent progressively zoomed in ‘top’ views of RyR1ABCd docked into a closed EMDB 1606 (Samsó et al. 2009). Similarly (D), (E) and (F) do the same but from a ‘side’ view. In surface representation, RyR1/2 mutations are coloured black but in stick representation RyR2 mutations are grey. A new interaction plane in interface 7 is defined and marks the join between the central rim of RyR1 and its clamps. Interface 3, or the lateral face to domain C (Tung et al. 2010) can now be modified to include areas adjacent to domain d. A Phosphatase Access Channel (PAC) is arrowed in (B) and has access to LZ-motif residues coloured green in (C). For illustration, in (F), the side-chains of RyR1 M226 and D227, as well as RyR2 H240 have been modelled in despite their flexibility in the crystal, and are seen to protrude into interface 1. A transparent cartoon shows interface 1 from RyR1ABCd docked into the open EMDB 1607. Numbering: RyR1 (black) or RyR2 (grey) Homo sapiens. 	
   	
    4.1.4	
  Interface	
  1	
   	
   Interactions	
   at	
   interface	
   1	
   (labelled	
   in	
   Figure	
   32D	
   and	
   32E)	
   are	
   of	
   special	
   interest	
   as	
   they	
   sew	
   the	
   vestibule	
   together	
   forming	
   seams	
   between	
   monomers.	
   Previously	
  we	
  hypothesized	
  that	
  gain	
  of	
  function	
   mutations	
   in	
   the	
   region	
   could	
   easily	
   be	
  explained	
  by	
  causing	
  these	
  subunit	
  seams	
  to	
  loosen	
  (Tung	
  et	
  al.	
  2010).	
  The	
  RyR1	
   MH	
   mutations	
   M226K,	
   D227V	
   and	
   R367L/Q	
   as	
   well	
   as	
   the	
   RyR2	
   CPVT	
   mutation	
   H240R	
   can	
   now	
   be	
   positioned	
   and	
   agree	
   with	
   this	
   previous	
   model.	
   Although	
   these	
   residues	
  are	
  flexible	
  in	
  the	
  crystal	
  structure,	
  they	
  can	
  presumably	
  interact	
  transiently	
   in	
   the	
   full-­‐length	
   channel.	
   As	
   an	
   example,	
   in	
   Figure	
   32E,	
   the	
   residues	
   labelled	
   H240	
   and	
  Q156	
  could	
  form	
  a	
  potential	
  hydrogen	
  bond.	
  If	
  Ca2+-­‐release	
  involves	
  movement	
   around	
   the	
   cytoplasmic	
   opening,	
   interactions	
   such	
   as	
   these	
   necessarily	
   have	
   to	
   be	
   weak	
  and	
  transient	
  to	
  allow	
  for	
  channel	
  opening.	
   During	
  opening	
  and	
  closing	
  of	
  the	
  channel	
  to	
  accommodate	
  Ca2+,	
  an	
  increase	
  in	
   the	
  transmembrane	
  pore	
  size	
  is	
  expected.	
  But	
  in	
  addition,	
  analysis	
  of	
  open	
  and	
  closed	
   EM	
   maps	
   has	
   indeed	
   shown	
   that	
   large	
   allosteric	
   movements	
   in	
   the	
   central	
   rim	
   accompany	
  pore	
  opening	
  (Kimlicka	
  and	
  Van	
  Petegem,	
  2011).	
  Molecular	
  details	
  can	
  be	
   	
    84	
    attained	
  by	
  docking	
  RyR1ABCd	
  into	
  cryo-­‐EM	
  reconstructions	
  of	
  the	
  same	
  resolution	
   in	
  both	
  the	
  open	
  (EMDB	
  1607)	
  and	
  closed	
  (EMDB	
  1606)	
  RyR1	
  conformations	
  (Samsó	
   et	
   al.	
   2009).	
   Comparing	
   these	
   docked	
   structures,	
   it	
   is	
   clear	
   that	
   interface	
   1	
   must	
   ‘break’	
   during	
   channel	
   opening,	
   allowing	
   the	
   movement	
   of	
   subunits	
   ~4	
   Å	
   further	
   apart	
  (Figure	
  32E).	
  A	
  look	
  at	
  the	
  overall	
  structure	
  motion	
  portrays	
  four	
  RyR1ABCds	
   as	
   a	
   cap,	
   tightening	
   to	
   close	
   the	
   channel	
   (Figures	
   33A	
   and	
   33B).	
   This	
   screw-­‐like	
   constriction	
   may	
   continue	
   towards	
   the	
   pore,	
   providing	
   an	
   allosteric	
   explanation	
   for	
   channel	
  closing	
  upon	
  ligand	
  binding.	
  	
   	
   	
    FIGURE 33. RyR1ABCd in channel opening . (A) Top and (B) side views of RyR1ABCd docked into the open (EMDB 1607) and closed (EMDB 1606) reconstructions of full-length channel (Samsó et al. 2009). The surface representations of the structure in the open and closed state are shown in blue and red respectively. In opening, RyR1ABCd motion can be described as a ~3 Å twist, lifting the domains ~5 Å towards the cytoplasm and widening interface 1 by ~4 Å. 	
    	
   	
   	
    85	
    4.1.5	
  Continuing	
  the	
  Debate	
  on	
  Dantrolene	
   	
   The	
  possibility	
  of	
  RyR1	
  as	
  the	
  molecular	
  target	
  of	
  dantrolene	
  has	
  been	
  a	
  topic	
   of	
  serious	
  discussion.	
  Evidence	
  in	
  the	
  field,	
  on	
  both	
  sides	
  of	
  the	
  argument	
  have	
  been	
   described,	
   but	
   work	
   by	
   the	
   Parness	
   lab	
   (Paul-­‐Pletzer	
   et	
   al.	
   2001,	
   Paul-­‐Pletzer	
   et	
   al.	
   2002)	
   has	
   strongly	
   suggested	
   that	
   the	
   ligand	
   does	
   directly	
   interact	
   with	
   RyR1,	
   and	
   specifically	
  residues	
  590-­‐609.	
   Characterizing	
   an	
   accurate,	
   thermodynamically	
   well-­‐defined	
   interaction	
   between	
  RyR1	
  590-­‐609	
  would	
  finally	
  put	
  to	
  rest	
  further	
  discussion	
  on	
  the	
  topic	
  and	
   confirm	
   RyR	
   as	
   a	
   molecular	
   target	
   of	
   dantrolene.	
   Our	
   results	
   though,	
   could	
   not	
   provide	
  this	
  valuable	
  information.	
  	
  ITC	
  experiments	
  failed	
  to	
  detect	
  binding	
  between	
   RyR1	
   1-­‐617	
   and	
   dantrolene	
   or	
   azumolene.	
   Initial	
   attempts	
   at	
   explaining	
   the	
   results	
   involved	
   low	
   protein	
   concentrations.	
   	
   Weak	
   binders	
   would	
   require	
   higher	
   concentrations	
   to	
   detect	
   their	
   interactions.	
   But	
   mentioned	
   already,	
   therapeutically	
   both	
   drugs	
   work	
   at	
   ~10	
   µM	
   (Flewellen	
   et	
   al.	
   1983),	
   well	
   below	
   the	
   concentration	
   tested	
   here	
   (final	
   cell	
   concentrations	
   of	
   drug	
   and	
   protein	
   were	
   ~125	
   µM	
   and	
   ~30	
   µM	
   respectively).	
  If	
  binding	
  does	
  take	
  place	
  then,	
  there	
  is	
  little	
  enthalpic	
  contribution	
  to	
   it.	
  If	
  this	
  were	
  the	
  case	
  ligands	
  that	
  bind	
  at	
  the	
  same	
  site	
  could	
  be	
  used	
  in	
  competition	
   experiments	
  to	
  obtain	
  an	
  apparent	
  affinity.	
  As	
  a	
  separate	
  explanation	
  for	
  the	
  lack	
  of	
   detectable	
  binding,	
  it	
  is	
  possible	
  that	
  a	
  prerequisite	
  binder	
  such	
  as	
  ATP	
  is	
  necessary	
   to	
  put	
  RyR	
  in	
  a	
  dantrolene-­‐accessible	
  state.	
   	
   	
   	
    4.2 Purines	
  and	
  RyR1	
   	
   Increased	
  sensitivity	
  to	
  caffeine	
  is	
  a	
  diagnostic	
  tool	
  for	
  MH	
  mutations	
  in	
  RyR1.	
   Zhao	
  et	
  al.	
  (2001)	
  have	
  shown	
  that	
  dantrolene	
  tames	
  this	
  sensitivity	
  and	
  lowers	
  the	
   affinity	
  for	
  caffeine.	
  In	
  the	
  same	
  study	
  they	
  illustrate	
  a	
  necessity	
  for	
  the	
  presence	
  of	
   an	
  ATP	
  analog	
  in	
  dantrolene	
  inhibition.	
  This	
  could	
  be	
  explained	
  by	
  the	
  existence	
  of	
  a	
   	
    86	
    purine	
   binding	
   site	
   in	
   intimate,	
   allosteric	
   communication	
   with	
   a	
   dantrolene	
   binding	
   site	
   in	
   RyR.	
   Our	
   results	
   indicate	
   that	
   indeed,	
   the	
   two	
   could	
   be	
   closely	
   related.	
   Already	
   discussed,	
  Paul-­‐Pletzer	
  et	
  al.	
  (2002)	
  have	
  identified	
  a	
  dantrolene	
  binding	
  site	
  in	
  RyR1	
   590-­‐609.	
   Here	
   we	
   have	
   presented	
   data	
   confirming	
   that	
   RyR1ABC	
   and	
   RyR1BC	
   both	
   interacted	
  with	
  caffeine,	
  narrowing	
  the	
  caffeine	
  binding	
  site	
  to	
  RyR1	
  217-­‐536	
  and	
  in	
   addition	
   we	
   show	
   potential	
   for	
   an	
   ATP	
   binding	
   site	
   in	
   the	
   same	
   domains.	
   Whether	
   or	
   not	
  these	
  interaction	
  sites	
  are	
  the	
  same,	
  remains	
  to	
  be	
  seen.	
   In	
   the	
   aim	
   of	
   structurally	
   characterizing	
   these	
   positions,	
   both	
   RyR1BC	
   and	
   RyR1ABC	
   were	
   crystallized	
   with	
   caffeine,	
   but	
   provided	
   different	
   binding	
   solutions.	
   These	
  results	
  although	
  outwardly	
  contradictory	
  may	
  not	
  be	
  completely	
  irrelevant	
  in	
   vivo.	
   Given	
   the	
   relatively	
   non-­‐specific	
   modes	
   of	
   attachment	
   to	
   aromatic	
   residues	
   discussed	
   in	
   Chapter	
   1.4.3	
   (Figure	
   9),	
   there	
   could	
   be	
   multiple	
   sites	
   of	
   attachment,	
   whether	
  these	
  are	
  all	
  functional	
  sites	
  is	
  a	
  different	
  story.	
  Further	
  experiments	
  would	
   need	
  to	
  be	
  carried	
  out	
  to	
  distinguish	
  between	
  sites	
  of	
  casual	
  interaction,	
  and	
  sites	
  of	
   biological	
   relevance.	
   Still,	
   the	
   crystallographic	
   binding	
   we	
   observed,	
   in	
   both	
   cases	
   heavily	
   involved	
   symmetry	
   related	
   molecules	
   arranged	
   in	
   a	
   physiologically	
   incompatible	
   manner.	
   A	
   couple	
   of	
   options	
   are	
   available	
   to	
   bypass	
   this	
   hurdle.	
   1)	
   Crystallization	
  in	
  lower	
  concentrations	
  of	
  caffeine,	
  although	
  given	
  the	
  low	
  measured	
   affinity,	
  this	
  may	
  not	
  be	
  an	
  effective	
  solution.	
  2)	
  Random	
  screens	
  in	
  the	
  presence	
  of	
   caffeine,	
  here	
  the	
  hope	
  would	
  be	
  to	
  attain	
  a	
  different	
  crystal	
  form.	
  In	
  a	
  different	
  space	
   group,	
   the	
   non-­‐physiologically	
   binding	
   aromatic	
   residues	
   would	
   be	
   in	
   different	
   orientations	
  and	
  be	
  unable	
  to	
  ‘sandwich’	
  caffeine.	
  	
   In	
   addition,	
   we	
   plan	
   to	
   use	
   ITC	
   to	
   probe	
   the	
   interaction	
   in	
   solution.	
   Through	
   in	
   silico	
   docking	
   experiments,	
   we	
   have	
   identified	
   a	
   few	
   possible	
   locations	
   for	
   caffeine	
   binding.	
  We	
  aim	
  to	
  prepare	
  mutants	
  in	
  these	
  locations,	
  and	
  will	
  test	
  via	
  ITC	
  whether	
   they	
  alter	
  the	
  binding	
  affinity.	
  Sequential	
  ITC	
  experiments	
  with	
  purines	
  followed	
  by	
   dantrolene	
  might	
  provide	
  an	
  added	
  benefit	
  in	
  finding	
  the	
  drug’s	
  illusive	
  binding	
  site.	
   	
   	
   	
    	
    87	
    4.3 Regulation	
  by	
  EF-­hands	
   	
   RyR1	
   4064-­‐4210	
   was	
   suggested	
   to	
   interact	
   with	
   Ca2+,	
   CaMBD	
   and	
   CaV1.1	
   (Xiong	
   et	
   al.	
   2006).	
   The	
   major	
   drawback	
   from	
   these	
   experiments	
   was	
   inherent	
   behavioural	
   problems	
   of	
   the	
   isolated	
   peptide	
   –	
   refolding	
   following	
   denaturation	
   by	
   urea	
   was	
   required	
   to	
   solubilize	
   this	
   fragment.	
   To	
   convince	
   the	
   field	
   though,	
   CD	
   was	
   used	
   to	
   confirm	
   that	
   the	
   protein	
   was	
   rich	
   in	
   α-­‐helices,	
   as	
   would	
   be	
   expected	
   from	
   their	
  initial	
  sequence	
  alignment	
  with	
  CaM.	
  But	
  a	
  helical,	
  CaM-­‐like	
  structure	
  should	
  not	
   be	
  exclusively	
  expressed	
  in	
  inclusion	
  bodies.	
  It	
  is	
  likely	
  that	
  even	
  after	
  refolding,	
  there	
   could	
  be	
  flexible	
  N-­‐	
  and/or	
  C-­‐terminal	
  ends	
  that	
  are	
  misfolded	
  and	
  bias	
  results.	
  In	
  our	
   construct	
  selection,	
  there	
  was	
  a	
  stringent	
  requirement	
  for	
  proteins	
  to	
  behave	
  as	
  well-­‐ folded,	
   soluble	
   monomers	
   as	
   portrayed	
   by	
   a	
   single	
   Gaussian	
   peak	
   on	
   a	
   gel	
   filtration	
   column.	
  This	
  was	
  achieved	
  in	
  RyR1	
  4071-­‐4138	
  (Figure	
  27A).	
   ITC	
   investigation	
   into	
   Ca2+	
   and	
   Mg2+	
   binding	
   provided	
   affinities	
   in	
   the	
   millimolar	
   range,	
   much	
   lower	
   than	
   the	
   ~60	
   µM	
   Kd	
   reported	
   for	
   RyR1	
   4064-­‐4120	
   (Xiong	
   et	
   al.	
   2006).	
   The	
   construct	
   length	
   though,	
   matches	
   a	
   full	
   length	
   CaM	
   and	
   equilibrium	
  dialysis	
  showed	
  two	
  Ca2+	
   bound	
  per	
  molecule	
  of	
  RyR1	
  4064-­‐4210	
  with	
  a	
   Hill	
   co-­‐efficient	
   of	
   ~1.6.	
   The	
   added	
   affinity	
   observed	
   could	
   be	
   explained	
   by	
   this	
   positive	
   co-­‐operativity.	
   Additionally,	
   with	
   respect	
   to	
   Ca2+	
   binding,	
   the	
   protein	
   must	
   have	
  contained	
  only	
  two	
  out	
  of	
  four	
  functional	
  EF-­‐hands,	
  agreeing	
  with	
  our	
  modelling	
   of	
  ‘dead’	
  EF-­‐hands	
  (Figure	
  26).	
  Interestingly,	
  the	
  CaM	
  alignment	
  we	
  present	
  in	
  RyR1	
   ~4000-­‐4138	
  is	
  partially	
  upstream	
  of	
  4064-­‐4210,	
  and	
  both	
  show	
  a	
  sequence	
  identity	
   of	
  ~20%.	
  This	
  hints	
  at	
  the	
  possibility	
  of	
  three	
  repeats	
  of	
  EF-­‐hand	
  domains.	
  It	
  will	
  be	
   interesting	
  to	
  test	
  constructs	
  that	
  are	
  inclusive	
  of	
  all	
  putative	
  EF-­‐hands.	
  	
   	
   	
   	
   	
   	
   	
    	
    88	
    4.3.1	
  The	
  Involvement	
  of	
  CaMBDs	
   	
   	
    Introducing	
  CaMBDs	
  adds	
  to	
  the	
  complexity	
  of	
  Ca2+	
   regulation.	
  Here	
  we	
  show	
    binding	
   to	
   a	
   RyR1	
   CaMBD3	
   mutant,	
   and	
   Xiong	
   et	
   al.	
   (2006)	
   in	
   addition	
   have	
   shown	
   Ca2+-­‐dependent	
   binding	
   to	
   RyR1	
   CaMBD2.	
   There	
   is	
   evidence	
   for	
   multiple	
   CaMBDs	
   (Zorzatto	
   et	
   al.	
   1990,	
   Menegazzi	
   et	
   al.	
   1994,	
   Chen	
   and	
   MacLennan,	
   1995),	
   and	
   now	
   as	
   well,	
  multiple	
  EF-­‐hands.	
  Slowly,	
  a	
  complex	
  picture	
  is	
  beginning	
  to	
  form	
  that	
  involves	
   all	
  of	
  the	
  above	
  and	
  includes	
  in	
  addition,	
  CaM,	
  Ca2+	
  and	
  Mg2+;	
  CaM	
  and	
  RyR1	
  EF-­‐hands	
   exhibit	
   Ca2+/Mg2+-­‐dependent	
   regulation	
   of	
   RyR1	
   via	
   CaMBDs	
   each	
   of	
   which	
   have	
   different	
   affinities.	
   In	
   summary,	
   some	
   kind	
   of	
   Ca2+-­‐regulated	
   competition	
   between	
   CaM	
  and	
  RyR1	
  EF-­‐hands	
  must	
  take	
  place.	
   	
    The	
   obvious	
   next	
   experiments	
   would	
   involve	
   the	
   other	
   CaMBDs	
   and	
    determining	
   their	
   affinities;	
   an	
   outstanding	
   binder	
   might	
   give	
   us	
   clues	
   as	
   to	
   where	
   RyR1	
   4071-­‐4138	
   is	
   located	
   in	
   3D	
   space	
   relative	
   to	
   the	
   CaMBDs.	
   Competition	
   experiments	
  can	
  now	
  be	
  used	
  to	
  further	
  analyze	
  the	
  roles	
  of	
  Ca2+	
   or	
  Mg2+.	
  The	
  change	
   in	
  structure	
  of	
  RyR1	
  4071-­‐4138	
  as	
  a	
  result	
  of	
  Ca2+	
  binding	
  or	
  unbinding,	
  as	
  suggested	
   by	
   the	
   NMR	
   experiments,	
   may	
   affect	
   the	
   interactions	
   with	
   CaMBDs.	
   Conversely,	
   CaMBD	
   binding	
   may	
   stabilize	
   the	
   structure	
   and	
   allow	
   for	
   a	
   quantifiable	
   Ca2+/Mg2+	
   affinity	
  by	
  ITC.	
   	
   	
    4.3.2	
  Disease	
  and	
  RyR1	
  EF-­hands	
   	
   	
    The	
   importance	
   of	
   the	
   RyR	
   EF-­‐hands	
   can	
   be	
   seen	
   in	
   the	
   number	
   of	
   disease	
    mutations	
   found	
   in	
   the	
   region.	
   Within	
   the	
   three	
   predicted	
   lobes	
   (RyR1	
   4000-­‐4210,	
   RyR2	
   3954-­‐4164),	
   11	
   and	
   23	
   mutations	
   have	
   been	
   shown	
   to	
   cause	
   disease	
   in	
   RyR1	
   and	
  RyR2	
  (summarized	
  in	
  Table	
  7).	
  In	
  effect,	
  this	
  defines	
  a	
  ‘hotspot’	
  of	
  mutations	
  and	
   emphasizes	
  the	
  regions	
  importance	
  in	
  Ca2+	
  regulation.	
  	
   	
    	
    	
    	
    89	
    RyR1	
  Human	
  Mutation	
   M4022V	
   R4041W	
   S4050Y	
   T4081M	
   G4104R	
   S4112L	
   N4119Y	
   R4136S	
   R4179H	
   E4181K	
   A4185T	
  (&	
  V4842M)	
   	
   RyR1	
  Human	
  Mutation	
   S3959L	
   M3972I	
   D3973H	
   L3974Q	
   K3997E	
   F4020L	
   E4076K	
   N4097S	
   N4104I	
   N4104K	
   L4105F	
   H4108N	
   H4108Q	
   M4109R	
   S4124G	
   S4124T	
   R4144G	
   E4146K	
   Y4149S	
   S4153R	
   R4157Q	
   T4158P	
   Q4159P	
    Disease	
   MH	
   MH	
   MH	
   MH	
   MH	
   CNM	
   MH	
   MH	
   UCM	
   UCM	
   MH	
   	
   Disease	
   CPVT	
   CPVT	
   CPVT	
   CPVT	
   CPVT	
   CPVT	
   CPVT	
   E-­‐IVF	
   CPVT	
   CPVT	
   CPVT	
   CPVT	
   CPVT	
   CPVT/SCD	
   CPVT	
   UCP	
   CPVT	
   E-­‐IVF	
   CPVT	
   CPVT	
   CPVT	
   SUD	
   CPVT	
    Reference	
   Lee	
  et	
  al.	
  2010	
   Galli	
  et	
  al.	
  2006	
   Robinson	
  et	
  al.	
  2006	
   Ibarra	
  et	
  al.	
  2006	
   Kraeva	
  et	
  al.	
  2011	
   Jungbluth	
  et	
  al.	
  2005	
   Sambuughin	
  et	
  al.	
  2005	
   Galli	
  et	
  al.	
  2002	
   Berilacqua	
  et	
  al.	
  2011	
   Berilacqua	
  et	
  al.	
  2011	
   Kraeva	
  et	
  al.	
  2011	
   	
   Reference	
   Tester	
  et	
  al.	
  2006	
   Medeiros-­‐Domingo	
  et	
  al.	
  2009	
   Medeiros-­‐Domingo	
  et	
  al.	
  2009	
   Medeiros-­‐Domingo	
  et	
  al.	
  2009	
   Medeiros-­‐Domingo	
  et	
  al.	
  2009	
   Postma	
  et	
  al.	
  2005	
   Postma	
  et	
  al.	
  2005	
   Tester	
  et	
  al.	
  2004	
   Postma	
  et	
  al.	
  2005	
   Priori	
  et	
  al.	
  2001	
   Hasdemir	
  et	
  al.	
  2008	
   Postma	
  et	
  al.	
  2005	
   Postma	
  et	
  al.	
  2005	
   Nof	
  et	
  al.	
  2011	
   Medeiros-­‐Domingo	
  et	
  al.	
  2009	
   Tester	
  et	
  al.	
  2005	
   Berge	
  et	
  al.	
  2008	
   Tester	
  et	
  al.	
  2004	
   Medeiros-­‐Domingo	
  et	
  al.	
  2009	
   Kazemian	
  et	
  al.	
  2011	
   Medeiros-­‐Domingo	
  et	
  al.	
  2009	
   Tester	
  et	
  al.	
  2004	
   Medeiros-­‐Domingo	
  et	
  al.	
  2009	
    Table 7. Disease mutations in RyR1/2 EF-hands. A list of all currently documented disease mutations within RyR1 4000-4210 and the corresponding residues in RyR2 (39544164). CNM: Centronuclear myopathy, Idiopathic Ventricular Fibrillation induced by emotion or exercise, SCD: Sudden Cardiac Death, SUO: Sudden Unexplained Death, UCP: Undefined Clinical Phenotype. 	
    90	
    Concerning	
   MH,	
   RyR1	
   EF-­‐hands	
   may	
   provide	
   a	
   means	
   of	
   furthering	
   our	
   understanding	
   of	
   its	
   mechanistic	
   details.	
   Volatile	
   anaesthetics	
   trigger	
   MH	
   episodes,	
   and	
   RyR1	
   could	
   potentially	
   be	
   their	
   target	
   as	
   shown	
   by	
   activation	
   in	
   lipid	
   bilayers	
   (Connelly	
   and	
   Coronado,	
   1994,	
   Bull	
   and	
   Marengo,	
   1994).	
   Testing	
   the	
   binding	
   of	
   anaesthetics	
  to	
  RyR1	
  EF-­‐hands	
  may	
  give	
  clues	
  as	
  to	
  whether	
  or	
  not	
  this	
  is	
  the	
  case.	
   	
   	
   	
    4.4	
   Allostery	
  in	
  RyR1	
   	
   Samsó	
  and	
  Wagenknecht	
  (2002)	
  have	
  provided	
  a	
  cryo-­‐EM	
  structure	
  depicting	
   the	
  locations	
  of	
  Ca2+-­‐	
  and	
  apocalmodulin	
  in	
  RyR1	
  (Figure	
  34A).	
  Ca2+-­‐CaM	
  appears	
  to	
   localize	
   to	
   the	
   PAC,	
   within	
   reach	
   of	
   an	
   extended	
   LZ	
   arm.	
   Taken	
   with	
   the	
   data	
   presented	
   here,	
   this	
   portrays	
   the	
   PAC	
   and	
   regions	
   around	
   it	
   as	
   a	
   complex	
   hub	
   for	
   allosteric	
  communication	
  in	
  RyR	
  regulation	
  (Figure	
  34B)	
  in	
  which:	
   	
   1) Purine	
  activators	
  bind	
  to	
  RyR1ABC	
   2) PP1	
  targets	
  RyR1d,	
  due	
  to	
  a	
  LZ	
  coiled	
  coil	
  with	
  spinophilin	
   3) Dantrolene	
  may	
  interact	
  with	
  the	
  flexible	
  residues	
  past	
  RyR1d	
   4) The	
  inhibitory	
  Ca2+-­‐CaM	
  attaches,	
  presumably	
  to	
  a	
  CaMBD	
   5) RyR1	
  EF-­‐hands	
  may	
  be	
  involved	
  in	
  Ca2+-­‐dependent	
  competition	
  with	
  CaM	
   	
   In	
  all,	
  these	
  findings	
  just	
  scratch	
  the	
  surface	
  of	
  the	
  molecular	
  and	
  mechanistic	
   details	
   involved	
   in	
   RyR	
   regulation	
   by	
   ligands	
   and	
   modulators.	
   Many	
   more	
   details	
   about	
  the	
  structure	
  and	
  function	
  of	
  the	
  channel	
  remain	
  to	
  be	
  discovered.	
    	
    91	
    	
   FIGURE 34. Allosteric communication in RyR. Side (A) and top (B) views of apo(orange) and Ca - (lime green) CaM in their approximate locations in RyR1. Manual 2+  placement of CaM was done in PyMOL based on cryo-EM maps identifying its positions (Samsó and Wagenknecht, 2002). CaM comes in contact with the PAC upon binding of 2+  Ca . RyR1 A, B, C and d are coloured blue, green, red and yellow and shown docked into EMDB 1606 (Samsó et al. 2009). Labelled are some of the modulators mentioned in this thesis. Their interactions are arrowed to emphasize the complexity of the allosteric reactions that must be involved in RyR ligand regulation. 	
    92	
    References Ai,	
   X.,	
   Curran,	
   J.W.,	
   Shannon,	
   T.R.,	
   Bers,	
   D.M.	
   and	
   Pogwizd,	
   S.M.	
   Ca2+/calmodulin-­‐ dependent	
   protein	
   kinase	
   modulates	
   cardiac	
   ryanodine	
   receptor	
   phosphorylation	
   and	
  sarcoplasmic	
  reticulum	
  Ca2+	
  leak	
  in	
  heart	
  failure.	
  Circ	
  Res.	
  2005.	
  97	
  :1314–22	
  	
   Amador,	
  F.J.,	
  Liu,	
  S.,	
  Ishiyama,	
  N.,	
  Plevin,	
  M.J.,	
  Wilson,	
  A.,	
  MacLennan,	
  D.H.	
  and	
  Ikura,	
   M.	
   Crystal	
   structure	
   of	
   type	
   I	
   ryanodine	
   receptor	
   amino-­‐terminal	
   β-­‐trefoil	
   domain	
   reveals	
  a	
  disease-­‐associated	
  mutation	
  “hot	
  spot”	
  loop.	
  Proc.	
  Natl.	
  Acad.	
  Sci.	
  U.S.A.	
  2009	
   106:	
  11040–11044	
   Armstrong,	
  C.M.,	
  Bezanilla,	
  F.M.	
  and	
  Horowicz,	
  P.	
  Twitches	
  in	
  the	
  presence	
  of	
  ethylene	
   glycol	
   bis(beta-­‐aminoethyl	
   ether)-­‐N,N'-­‐tetracetic	
   acid.	
   Biochimica	
   et	
   biophyica	
   acta.	
  	
   1972.	
  267:	
  605-­‐608	
   Arndt,	
   U.W.,	
   Crowther,	
   R.A.	
   and	
   Mallett,	
   J.F.	
   A	
   computer-­‐linked	
   cathode-­‐ray	
   tube	
   microdensitometer	
  for	
  x-­‐ray	
  crystallography.	
  J	
  Sci	
  Instrum.	
  1968.	
  1:	
  510-­‐6.	
   Babu,	
  YS,	
  Bugg,	
  CE	
  and	
  Cook	
  W.J.	
  Structure	
  of	
  calmodulin	
  refined	
  at	
  2.2	
  A	
  resolution.	
  J	
   Mol	
  Biol.	
  1988.	
  204:	
  191-­‐204	
   Baker,	
  M.L.,	
  Serysheva,	
  I.I.,	
  Sencer,	
  S.,	
  Wu,	
  Y.,	
  Ludtke,	
  S.J.,	
  Jiang,	
  W.,	
  Hamilton,	
  S.L.	
  and	
   Chiu,	
  W.	
  The	
  skeletal	
  muscle	
  Ca2+	
  release	
  channel	
  has	
  an	
  oxidoreductase-­‐like	
  domain.	
   Proc.	
  Natl	
  Acad.	
  Sci.	
  2002	
  USA	
  99:	
  12155–12160	
   Balshaw,	
  D.M.,	
  Xu,	
  L.,	
  Yamaguchi,	
  N.,	
  Pasek,	
  D.A.	
  and	
  Meissner,	
  G.	
  Calmodulin	
  binding	
   and	
  inhibition	
  of	
  cardiac	
  muscle	
  calcium	
  release	
  channel	
  (ryanodine	
  receptor).	
  J	
  Biol	
   Chem.	
  2001.	
  276:	
  20144–20153	
   Beard,	
   N.A.,	
   Laver,	
   D.R.	
   and	
   Dulhunty,	
   A.F.	
   Calsequestrin	
   and	
   the	
   calcium	
   release	
   channel	
  of	
  skeletal	
  and	
  cardiac	
  muscle.	
  Prog	
  Biophys	
  Mol	
  Biol.	
  2004.	
  85:	
  33-­‐69	
    	
    93	
    Beard,	
   N.A.,	
   Wei,	
   L.	
   and	
   Dulhunty	
   A.F.	
   Ca2+	
   signaling	
   in	
   striated	
   muscle:	
   the	
   elusive	
   roles	
  of	
  triadin,	
  junctin,	
  and	
  calsequestrin.	
  Eur.	
  Biophys.	
  J.	
  2009.	
  39:	
  27–36	
   Bellinger,	
   A.M.,	
   Reiken,	
   S.,	
   Carlson,	
   C.,	
   Mongillo,	
   M.,	
   Liu,	
   X.,	
   Rothman,	
   L.,	
   Matecki,	
   S.,	
   Lacampagne,	
   A.	
   and	
   Marks,	
   A.R.	
   Hypernitrosylated	
   ryanodine	
   receptor	
   calcium	
   release	
  channels	
  are	
  leaky	
  in	
  dystrophic	
  muscle.	
  Nat	
  Med.	
  2009.	
  15	
  :325–30	
   Berge,	
   K.E.,	
   Haugaa,	
   K.H.,	
   Früh,	
   A.,	
   Anfinsen,	
   O.G.,	
   Gjesdal,	
   K.,	
   Siem,	
   G.,	
   Oyen,	
   N.,	
   Greve,	
   G.,	
   Carlsson,	
   A.,	
   Rognum,	
   T.O.,	
   Hallerud,	
   M.,	
   Kongsgård,	
   E.,	
   Amlie	
   J.P.	
   and	
   Leren,	
   T.P.	
  	
  	
   Molecular	
   genetic	
   analysis	
   of	
   long	
   QT	
   syndrome	
   in	
   Norway	
   indicating	
   a	
   high	
   prevalence	
  of	
  heterozygous	
  mutation	
  carriers.	
  Scand	
  J	
  Clin	
  Lab	
  Invest.	
  2008.	
  68:	
  362-­‐ 8.	
   Betzenhauser,	
   M.J.	
   and	
   Marks,	
   A.R.	
   Ryanodine	
   receptor	
   channelopathies.	
   Pflugers	
   Arch.	
  2010.	
  460:	
  467–80	
   Bevilacqua,	
   J.A.,	
   Monnier,	
   N.,	
   Bitoun,	
   M.,	
   Eymard,	
   B.,	
   Ferreiro,	
   A.,	
   Monges,	
   S.,	
   Lubieniecki,	
   F.,	
   Taratuto,	
   A.L.,	
   Laquerrière,	
   A.,	
   Claeys,	
   K.G.	
   Marty,	
   I.,	
   Fardeau,	
   M.,	
   Guicheney,	
  P.,	
  Lunardi,	
  J.	
  and	
  Romero,	
  N.B.	
  Recessive	
  RYR1	
  mutations	
  cause	
  unusual	
   congenital	
   myopathy	
   with	
   prominent	
   nuclear	
   internalization	
   and	
   large	
   areas	
   of	
   myofibrillar	
  disorganization.	
  Neuropathol	
  Appl	
  Neurobiol.	
  2011.	
  37:	
  271-­‐84	
   Bezprozvanny,	
   I.B.,	
   Ondrias,	
   K.,	
   Kaftan,	
   E.,	
   Stoyanovsky,	
   D.A.	
   and	
   Ehrlich,	
   B.E.	
   Activation	
  of	
  the	
  calcium	
  release	
  channel	
  (ryanodine	
  receptor)	
  by	
  heparin	
  and	
  other	
   polyanions	
  is	
  calcium	
  dependent.	
  Mol.	
  Biol.	
  Cell.	
  1993.	
  4:	
  347–52	
   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.	
   1988.	
   107:	
   2587–2600	
   Bosanac,	
  I.,	
  Alattia,	
  J.R.,	
  Mal,	
  T.K,.	
  Chan,	
  J.,	
  Talarico,	
  S.,	
  Tong,	
  F.K.,	
  Tong,	
  K.I.,	
  Yoshikawa,	
   F.,	
   Furuichi,	
   T.,	
   Iwai,	
   M.,	
   Michikawa,	
   T.,	
   Mikoshiba,	
   K.	
   and	
   Ikura	
   M.	
   Structure	
   of	
   the	
    	
    94	
    inositol	
  1,4,5-­‐trisphosphate	
  receptor	
  binding	
  core	
  in	
  complex	
  with	
  its	
  ligand.	
  Nature	
   2002;	
  420:	
  696-­‐700	
   Bosanac,	
   I.,	
   Yamazaki,	
   H.,	
   Matsu-­‐Ura,	
   T.,	
   Michikawa,	
   T.,	
   Mikoshiba,	
   K.	
   and	
   Ikura,	
   M.	
   Crystal	
   structure	
   of	
   the	
   ligand	
   binding	
   suppressor	
   domain	
   of	
   type	
   1	
   inositol	
   1,4,5-­‐ trisphosphate	
  receptor.	
  Mol	
  Cell.	
  2005;	
  17:	
  193-­‐203	
   Brandt,	
  A.,	
  Schleithoff,	
  L.,	
  Jurkat-­‐Rott,	
  K.,	
  Klingler,	
  W.,	
  Baur,	
  C.	
  and	
  Lehmann-­‐Horn,	
  F.	
   Screening	
   of	
   the	
   ryanodine	
   receptor	
   gene	
   in	
   105	
   malignant	
   hyperthermia	
   families:	
   novel	
  mutations	
  and	
  concordance	
  with	
  the	
  in	
  vitro	
  contracture	
  test.	
  Hum	
  Mol	
  Genet.	
   1999.	
  8:	
  	
  2055-­‐62	
   Brillantes,	
   A.B.,	
   Ondrias,	
   K.,	
   Scott,	
   A.,	
   Kobrinsky,	
   E.,	
   Ondriasova,	
   E.,	
   Moschella,	
   M.C.,	
   Jayaraman,	
   T.,	
   Landers,	
   M.,	
   Ehrlich,	
   B.E.	
   and	
   Marks,	
   A.R.	
   Stabilization	
   of	
   calcium	
   release	
  channel	
  (ryanodine	
  receptor)	
  function	
  by	
  FK506-­‐binding	
  protein.	
  Cell.	
  1994.	
   77:	
  513–523	
   Bull,	
   R.	
   and	
   Marengo,	
   J.J.	
   Calcium-­‐dependent	
   halothane	
   activation	
   of	
   sarcoplasmic	
   reticulum	
   calcium	
   channels	
   from	
   frog	
   skeletal	
   muscle.	
   Am.	
   J.	
   Physiol.	
   1994.	
   266:	
   C391–6	
   Buratti,	
  R.,	
  Prestipino,	
  G.,	
  Menegazzi,	
  P.,	
  Treves,	
  S.	
  and	
  Zorzato,	
  F.	
  Calcium	
  dependent	
   activation	
   of	
   skeletal	
   muscle	
   Ca2+	
   release	
   channel	
   (ryanodine	
   receptor)	
   by	
   calmodulin.	
  Biochem	
  Biophys	
  Res	
  Commun,	
  1995,	
  213:	
  1082–1090	
   Carafoli,	
  E.	
  Intracellular	
  calcium	
  homeostasis.	
  Annu	
  Rev	
  Biochem.	
  1987.	
  56:	
  395–433	
   Chacón,	
  P.	
  and	
  Wriggers,	
  W.	
  Multi-­‐resolution	
  contour-­‐based	
  fitting	
  of	
  macromolecular	
   structures.	
  J.	
  Mol.	
  Biol.	
  2002,	
  317:	
  375–384	
   Chattopadhyaya,	
  R.,	
  Meador,	
  W.E.,	
  Means,	
  A.R.	
  and	
  Quiocho,	
  F.A.	
  Calmodulin	
  structure	
   refined	
  at	
  1.7	
  A	
  resolution.	
  J	
  Mol	
  Biol.	
  1992.	
  228:	
  1177-­‐92	
    	
    95	
    Chen,	
   S.R.	
   and	
   MacLennan,	
   D.H.	
   Identification	
   of	
   calmodulin-­‐,	
   Ca2+-­‐,	
   and	
   ruthenium	
   red-­‐binding	
   domains	
   in	
   the	
   Ca2+	
   release	
   channel	
   (ryanodine	
   receptor)	
   of	
   rabbit	
   skeletal	
  muscle	
  sarcoplasmic	
  reticulum.	
  J	
  Biol	
  Chem.	
  1994.	
  269:	
  22698–22704	
   Connelly,	
   T.J.	
   and	
   Coronado,	
   R.	
   Activation	
   of	
   the	
   Ca2+	
   release	
   channel	
   of	
   cardiac	
   sarcoplasmic	
  reticulum	
  by	
  volatile	
  anesthetics.	
  Anesthesiology.	
  1994.	
  81:	
  459–69	
   Crick,	
   F.H.C.	
   The	
   packing	
   of	
   α-­‐helices:	
   simple	
   coiled-­‐coils.	
   Acta	
   Crystallgr.	
   1953.	
   1:	
   689–97	
   De	
   Smedt,	
   H.,	
   Missiaen,	
   L.,	
   Parys,	
   J.B.,	
   Bootman,	
   M.D.,	
   Mertens,	
   L.,	
   Van	
   Den	
   Bosch,	
   L.	
   and	
  Casteels,	
  R.	
  Determination	
  of	
  relative	
  amounts	
  of	
  inositol	
  trisphosphate	
  receptor	
   mRNA	
  isoforms	
  by	
  ratio	
  polymerase	
  chain	
  reaction.	
  J	
  Biol	
  Chem.	
  1994.	
  269:	
  21691–8	
   Denborough,	
  M.	
  and	
  Lovell,	
  R.	
  Anaesthetic	
  deaths	
  in	
  a	
  family.	
  Lancet.	
  1960	
  2:	
  45	
   Denborough,	
  M.	
  Malignant	
  hyperthermia.	
  Lancet.	
  1998.	
  352:	
  1131–1136	
   Diaz-­‐Sylvester,	
   P.L.,	
   Porta,	
   M.	
   and	
   Copello,	
   J.A.	
   Halothane	
   modulation	
   of	
   skeletal	
   muscle	
   ryanodine	
   receptors:	
   dependence	
   on	
   Ca2+,	
   Mg2+,	
   and	
   ATP.	
   Am	
   J	
   Physiol	
   Cell	
   Physiol.	
  2008.	
  294:	
  C1103–12	
   Diederichs,	
   K.	
   and	
   Karplus,	
   P.A.	
   Improved	
   R-­‐factors	
   for	
   diffraction	
   data	
   analysis	
   in	
   macromolecular	
  crystallography.	
  Nat	
  Struct	
  Biol.	
  1997.	
  4:	
  269-­‐75	
   Dirksen	
   R.T.	
   and	
   Avila	
   G.	
   Altered	
   ryanodine	
   receptor	
   function	
   in	
   central	
   core	
   disease:	
   leaky	
  or	
  uncoupled	
  Ca2+	
  release	
  channels?	
  Trends	
  Cardiovasc	
  Med.	
  2002.	
  12:	
  189–97	
   Doré,	
   A.S.,	
   Robertson,	
   N.,	
   Errey,	
   J.C.,	
   Ng,	
   I.,	
   Hollenstein,	
   K.,	
   Tehan,	
   B.,	
   Hurrell,	
   E.,	
   Bennett,	
  K.,	
  Congreve,	
  M.,	
  Magnani,	
  F.,	
  Tate,	
  C.G.,	
  Weir,	
  M.	
  and	
  Marshall,	
  F.H.	
  Structure	
   of	
  the	
  adenosine	
  A(2A)	
  receptor	
  in	
  complex	
  with	
  ZM241385	
  and	
  the	
  xanthines	
  XAC	
   and	
  caffeine.	
  Structure.	
  2011.	
  19:	
  1283-­‐93	
   Dulhunty,	
  A.F.	
  The	
  voltage-­‐activation	
  of	
  contraction	
  in	
  skeletal	
  muscle.	
  Prog	
  Biophys	
   Mol	
  Biol.	
  1992.	
  57:	
  181–223	
    	
    96	
    Durham,	
  W.J.,	
  Aracena-­‐Parks,	
  P.,	
  Long,	
  C.,	
  Rossi,	
  A.E.,	
  Goonasekera,	
  S.A.,	
  Boncompagni,	
   S.,	
   Galvan,	
   D.L.,	
   Gilman,	
   C.P.,	
   Baker,	
   M.R.,	
   Shirokova,	
   N.,	
   Protasi,	
   F.,	
   Dirksen,	
   R.	
   and	
   Hamilton,	
  S.L.	
  RyR1	
  S-­‐nitrosylation	
  underlies	
  environmental	
  heat	
  stroke	
  and	
  sudden	
   death	
  in	
  Y522S	
  RyR1	
  knock-­‐in	
  mice.	
  Cell.	
  2008.	
  133:	
  53–65	
   Dykes,	
  M.H.	
  Evaluation	
  of	
  a	
  muscle	
  relaxant:	
  dantrolene	
  sodium	
  (Dantrium).	
  J	
  Am	
  Med	
   Ass.	
  1975.	
  231:	
  862–4	
   Edelhoch,	
   H.	
   Spectroscopic	
   determination	
   of	
   tryptophan	
   and	
   tyrosine	
   in	
   proteins.	
   Biochemistry.	
  1967.	
  6:	
  1948–54	
   Ekstrom,	
  J.L.,	
  Pauly,	
  T.A.,	
  Carty,	
  M.D.,	
  Soeller,	
  W.C.,	
  Culp,	
  J.,	
  Danley,	
  D.E.,	
  Hoover,	
  D.J.,	
   Treadway,	
   J.L.,	
   Gibbs,	
   E.M.,	
   Fletterick,	
   R.J.,	
   Day,	
   Y.S.,	
   Myszka,	
   D.G.	
   and	
   Rath,	
   V.L.	
   Structure-­‐activity	
   analysis	
   of	
   the	
   purine	
   binding	
   site	
   of	
   human	
   liver	
   glycogen	
   phosphorylase.	
  Chem	
  Biol.	
  2002.	
  9:	
  915-­‐24	
   el-­‐Hayek,	
   R.,	
   Parness,	
   J.,	
   Valdivia,	
   H.H.,	
   Coronado,	
   R.	
   and	
   Hogan,	
   K.	
   Dantrolene	
   and	
   azumolene	
   inhibit	
   [3H]-­‐PN200-­‐110	
   binding	
   to	
   porcine	
   skeletal	
   muscle	
   dihydropyridine	
  receptor.	
  Biochem	
  Biophys	
  Res	
  Commun.	
  1992.	
  187:	
  894–900	
   Emsley,	
   P.,	
   Cowtan,	
   K.	
   Coot:	
   model-­‐building	
   tools	
   for	
   molecular	
   graphics.	
   Acta	
   Crystallogr.	
  D	
  Biol.	
  Crystallogr.	
  2004.	
  60:	
  2126–2132	
   Endo,	
   M.	
   Calcium	
   release	
   from	
   the	
   sarcoplasmic	
   reticulum.	
   Physiol.	
   Rev.	
   1977	
   57:	
   71– 106	
   Endo,	
   M.,	
   Tanaka,	
   M.	
   and	
   Ogawa,	
   Y.	
   Calcium-­‐induced	
   release	
   of	
   calcium	
   from	
   the	
   sarcoplasmic	
  reticulum	
  of	
  skinned	
  skeletal	
  muscle	
  fibers.	
  Nature.	
  1970.	
  228:	
  34–36	
   Fabiato,	
   A.	
   Calcium-­‐induced	
   release	
   of	
   calcium	
   from	
   the	
   cardiac	
   sarcoplasmic	
   reticulum.	
  Am.	
  J.	
  Physiol.	
  1983.	
  245:	
  C1–C14	
    	
    97	
    Ferguson,	
   D.G.,	
   Schwartz,	
   H.W.	
   and	
   Franzini-­‐Armstrong,	
   C.	
   Subunit	
   structure	
   of	
   junctional	
  feet	
  in	
  triads	
  of	
  skeletal	
  muscle:	
  A	
  freezedrying,	
  rotary-­‐shadowing	
  study.	
  J	
   Cell	
  Biol,	
  1984,	
  99:	
  1735–1742	
   Flewellen,	
  E.H.,	
  Nelson,	
  T.E.,	
  Jones,	
  W.P.,	
  Arens,	
  J.F.	
  and	
  Wagner,	
  D.L.	
  Dantrolene	
  dose	
   response	
   in	
   awake	
   man:	
   implications	
   for	
   management	
   of	
   malignant	
   hyperthermia.	
   Anesthesiology.	
  1983.	
  59:	
  275–280	
   Fujii,	
   J.,	
   Otsu,	
   K.,	
   Zorzato,	
   F.,	
   de	
   Leon,	
   S.,	
   Khanna,	
   V.K.,	
   Weiler,	
   J.E.,	
   O'Brien,	
   P.J.	
   and	
   Maclennan,	
  D.H.	
  Identification	
  of	
  a	
  mutation	
  in	
  porcine	
  ryanodine	
  receptor	
  associated	
   with	
  malignant	
  hyperthermia.	
  Science.	
  1991	
  253:	
  488–551	
   Galli,	
  L.,	
  Orrico,	
  A.,	
  Cozzolino,	
  S.,	
  Pietrini,	
  V.,	
  Tegazzin,	
  V.	
  and	
  Sorrentino,	
  V.	
  Mutations	
   in	
  the	
  RYR1	
  gene	
  in	
  Italian	
  patients	
  at	
  risk	
  for	
  malignant	
  hyperthermia:	
  evidence	
  for	
  a	
   cluster	
  of	
  novel	
  mutations	
  in	
  the	
  C-­‐terminal	
  region.	
  Cell	
  Calcium.	
  2002.	
  32:	
  143-­‐51	
   Galli,	
  L.,	
  Orrico,	
  A.,	
  Lorenzini,	
  S.,	
  Censini,	
  S.,	
  Falciani,	
  M.,	
  Covacci,	
  A.,	
  Tegazzin,	
  V.	
  and	
   Sorrentino,	
  V.	
  Frequency	
  and	
  localization	
  of	
  mutations	
  in	
  the	
  106	
  exons	
  of	
  the	
  RYR1	
   gene	
  in	
  50	
  individuals	
  with	
  malignant	
  hyperthermia.	
  Hum	
  Mutat.	
  2006.	
  27:	
  830	
   Garzón,	
   J.I.,	
   Kovacs,	
   J.,	
   Abagyan,	
   R.	
   and	
   Chacón,	
   P.	
   ADP_EM:	
   fast	
   exhaustive	
   multi-­‐ resolution	
  docking	
  for	
  high-­‐throughput	
  coverage.	
  Bioinformatics.	
  2007.	
  23:	
  427–33	
   Girard,	
   T.,	
   Johr,	
   M.,	
   Schaefer,	
   C.	
   and	
   Urwyler,	
   A.	
   Perinatal	
   diagnosis	
   of	
   malignant	
   hyperthermia	
  susceptibility.	
  Anesthesiology.	
  2006.	
  104:	
  1353–4	
   Gordon,	
   R.A.	
   Malignant	
   hyperpyrexia	
   during	
   general	
   anaesthesia.	
   Can	
   Anaesth	
   Soc	
   J.	
   1996.	
  13:	
  415–416	
   Groom,	
   L.,	
   Muldoon,	
   S.M.,	
   Tang,	
   Z.Z.,	
   Brandom,	
   B.W.,	
   Bayarsaikhan,	
   M.,	
   Bina,	
   S.,	
   Lee,	
   H.S.,	
  Qiu,	
  X.,	
  Sambuughin,	
  N.	
  and	
  Dirksen,	
  R.T.	
  Identical	
  de	
  novo	
  mutation	
  in	
  the	
  type	
  1	
   ryanodine	
   receptor	
   gene	
   associated	
   with	
   fatal,	
   stress-­‐induced	
   malignant	
   hyperthermia	
  in	
  two	
  unrelated	
  families.	
  Anesthesiology.	
  2011.	
  	
  115:	
  938–45	
    	
    98	
    Gurrola,	
   G.B.,	
   Arevalo,	
   C.,	
   Sreekumar,	
   R.,	
   Lokuta,	
   A.J.,	
   Walker,	
   J.W.	
   and	
   Valdivia,	
   H.H.	
   Activation	
  of	
  ryanodine	
  receptors	
  by	
  imperatoxin	
  A	
  and	
  a	
  peptide	
  segment	
  of	
  the	
  II-­‐ III	
  loop	
  of	
  the	
  dihydropyridine	
  receptor.	
  J	
  Biol	
  Chem.	
  1999.	
  274:	
  7879–86	
   Hakamata,	
   Y.,	
   Nakai,	
   J.,	
   Takeshima,	
   H.	
   and	
   Imoto,	
   K.	
   Primary	
   structure	
   and	
   distribution	
  of	
  a	
  novel	
  ryanodine	
  receptor/calcium	
  release	
  channel	
  from	
  rabbit	
  brain.	
   FEBS	
  Lett,	
  1992,	
  312:	
  229–235	
   Hamilton,	
   S.	
   L.	
   and	
   Serysheva,	
   I.	
   I.	
   Ryanodine	
   receptor	
   structure:	
   progress	
   and	
   challenges.	
  J.	
  Biol.	
  Chem.	
  2009.	
  284:	
  4047–51	
   Harrison,	
   G.G.	
   Control	
   of	
   malignant	
   hyperpyrexia	
   syndrome	
   in	
   MHS	
   swine	
   by	
   dantrolene	
  sodium.	
  Br	
  J	
  Anaesth.	
  1975	
  47:	
  62–65	
   Hasdemir,	
   C.,	
   Aydin,	
   H.H.,	
   Sahin,	
   S.	
   and	
   Wollnik,	
   B.	
   Catecholaminergic	
   polymorphic	
   ventricular	
  tachycardia	
  caused	
  by	
  a	
  novel	
  mutation	
  in	
  the	
  cardiac	
  ryanodine	
  receptor.	
   Anadolu	
  Kardiyol	
  Derg.	
  2008.	
  8:	
  E35-­‐6.	
   Herbette,	
   L.,	
   Messineo,	
   F.C.	
   and	
   Katz,	
   A.M.	
   The	
   interaction	
   of	
   drugs	
   with	
   the	
   sarcoplasmic	
  reticulum.	
  Annu.	
  Rev.	
  Pharmacol.	
  Toxicol.	
  1982.	
  22:	
  413–434	
   Herrmann-­‐Frank,	
   A.,	
   Richter,	
   M.,	
   Sarközi,,	
   S.,	
   Mohr	
   U.	
   and	
   Lehmann-­‐Horn,	
   F.	
   4-­‐chloro-­‐ m-­‐cresol,	
   a	
   potent	
   and	
   specific	
   activator	
   of	
   the	
   skeletal	
   muscle	
   ryanodine	
   receptor.	
   Biochim.	
  Biophys.	
  Acta.	
  1996.	
  1289:	
  31–40	
   Huttlin,	
  E.L.,	
  Jedrychowski,	
  M.P.,	
  Elias,	
  J.E.,	
  Goswami,	
  T.,	
  Rad,	
  R.,	
  Beausoleil,	
  S.A.,	
  Villén,	
   J.,	
   Haas,	
   W.,	
   Sowa,	
   M.E.	
   and	
   Gygi,	
   S.P.	
   A	
   tissue-­‐specific	
   atlas	
   of	
   mouse	
   protein	
   phosphorylation	
  and	
  expression.	
  Cell.	
  2010.	
  143:	
  1174–89	
   Huxley,	
   A.F.	
   and	
   Niedergerke,	
   R.	
   Structural	
   Changes	
   in	
   Muscle	
   During	
   Contraction:	
   Interference	
  Microscopy	
  of	
  Living	
  Muscle	
  Fibres.	
  Nature.	
  1954.	
  173:	
  971–973	
   Huxley,	
  H.	
  and	
  Hanson,	
  J.	
  Changes	
  in	
  the	
  cross-­‐striations	
  of	
  muscle	
  during	
  contraction	
   and	
  stretch	
  and	
  their	
  structural	
  interpretation.	
  Nature.	
  1954.	
  173:	
  973–976	
    	
    99	
    Ikemoto,	
   N.,	
   and	
   Yamamoto,	
   T.	
   Regulation	
   of	
   calcium	
   release	
   by	
   interdomain	
   interaction	
  within	
  ryanodine	
  receptors.	
  Front	
  Biosci.	
  2002.	
  7:	
  d671-­‐683	
   Ikemoto,	
  T.,	
  Iino,	
  M.	
  and	
  Endo,	
  M.	
  Enhancing	
  effect	
  of	
  calmodulin	
  on	
  Ca2+-­‐induced	
  Ca2+	
   release	
  in	
  the	
  sarcoplasmic	
  reticulum	
  of	
  rabbit	
  skeletal	
  muscle	
  fibres.	
  J	
  Physiol,	
  1995,	
   487:	
  573–582	
   Jiang,	
  D.,	
  Chen,	
  W.,	
  Xiao,	
  J.,	
  Wang,	
  R.,	
  Kong,	
  H.,	
  Jones,	
  P.P.,	
  Zhang,	
  L.,	
  Fruen,	
  B.	
  and	
  Chen,	
   S.R.W.	
   Reduced	
   threshold	
   for	
   luminal	
   Ca2+	
   activation	
   of	
   RyR1	
   underlies	
   a	
   causal	
   mechanism	
  of	
  porcine	
  malignant	
  hyperthermia.	
  J	
  Biol	
  Chem.	
  2008.	
  283:	
  20813-­‐20	
   Jiang,	
  D.,	
  Wang,	
  R.,	
  Xiao,	
  B.,	
  Kong,	
  H.,	
  Hunt,	
  D.	
  J.,	
  Choi,	
  P.,	
  Zhang,	
  L.,	
  and	
  Chen,	
  S.	
  R.	
  W.	
   Circ.	
  Res.	
  2005.	
  97:	
  1173–81	
   Jiang,	
  D.,	
  Xiao,	
  B.,	
  Yang,	
  D.,	
  Wang,	
  R.,	
  Choi,	
  P.,	
  Zhang,	
  L.,	
  Cheng,	
  H.,	
  and	
  Chen,	
  S.	
  R.	
  W.	
   Proc	
  Natl	
  Acad	
  Sci	
  USA	
  .2004.	
  101:	
  13062–7	
   Jungbluth,	
   H.,	
   Müller,	
   C.R.,	
   Halliger-­‐Keller,	
   B.,	
   Brockington,	
   M.,	
   Brown,	
   S.C.,	
   Feng,	
   L.,	
   Chattopadhyay,	
   A.,	
   Mercuri,	
   E.,	
   Manzur,	
   A.Y.,	
   Ferreiro,	
   A.,	
   Laing,	
   N.G.,	
   Davis,	
   M.R.,	
   Roper,	
  H.P.,	
  Dubowitz,	
  V.,	
  Bydder,	
  G.,	
  Sewry,	
  C.A.	
  and	
  Muntoni,	
  F.	
  Autosomal	
  recessive	
   inheritance	
  of	
  RYR1	
  mutations	
  in	
  a	
  congenital	
  myopathy	
  with	
  cores.	
  Neurology.	
  2002.	
   59:284–7	
   Jungbluth,	
   H.,	
   Zhou,	
   H.,	
   Hartley,	
   L.,	
   Halliger-­‐Keller,	
   B.,	
   Messina,	
   S.,	
   Longman,	
   C.,	
   Brockington,	
  M.,	
  Robb,	
  SA,	
  Straub,	
  V.,	
  Voit,	
  T.	
  Swash,	
  M.,	
  Ferreiro,	
  A.,	
  Bydder,	
  G.,	
  Sewry,	
   C.A.,	
   Müller,	
   C.	
   and	
   Muntoni	
   F.	
   Minicore	
   myopathy	
   with	
   ophthalmoplegia	
   caused	
   by	
   mutations	
  in	
  the	
  ryanodine	
  receptor	
  type	
  1	
  gene.	
  	
  Neurology.	
  2005.	
  65:	
  1930-­‐5	
   Juranic,	
   N.,	
   Macura,	
   S.,	
   Simeonov,	
   M.V.,	
   Jones,	
   K.A.,	
   Penheiter,	
   A.R.,	
   Hock,	
   T.J.	
   and	
   Streiff,	
   J.H.	
   Halothane	
   binds	
   to	
   druggable	
   sites	
   in	
   calcium-­‐calmodulin:	
   Solution	
   structure	
  of	
  halothane-­‐CaM	
  C-­‐terminal	
  domain.	
  To	
  be	
  published.	
  PDB	
  ID:	
  2KUH	
    	
    100	
    Juranic,	
   N.,	
   Macura,	
   S.,	
   Simeonov,	
   M.V.,	
   Jones,	
   K.A.,	
   Penheiter,	
   A.R.,	
   Hock,	
   T.J.	
   and	
   Streiff,	
   J.H.	
   Halothane	
   binds	
   to	
   druggable	
   sites	
   in	
   calcium-­‐calmodulin:	
   Solution	
   structure	
  of	
  halothane-­‐CaM	
  N-­‐terminal	
  domain.	
  To	
  be	
  published.	
  PDB	
  ID:	
  2KUG	
   Kabsch,	
  W.	
  Xds.	
  Acta	
  Crystallogr.	
  D	
  Biol.	
  Crystallogr.	
  2010.	
  66:	
  125-­‐32	
   Kaftan,	
   E.,	
   Marks,	
   A.R.	
   and	
   Ehrlich,	
   B.E.	
   Effects	
   of	
   rapamycin	
   on	
   ryanodine	
   receptor/Ca2+	
  release	
  channels	
  from	
  cardiac	
  muscle.	
  Circ.	
  Res.	
  1996.	
  78:	
  990–7	
   Kazemian,	
  P.,	
  Gollob,	
  M.H.,	
  Pantano,	
  A.	
  and	
  Oudit,	
  G.Y.	
  A	
  novel	
  mutation	
  in	
  the	
  RYR2	
   gene	
   leading	
   to	
   catecholaminergic	
   polymorphic	
   ventricular	
   tachycardia	
   and	
   paroxysmal	
   atrial	
   fibrillation:	
   dose-­‐dependent	
   arrhythmia-­‐event	
   suppression	
   by	
   β-­‐ blocker	
  therapy.	
  Can	
  J	
  Cardiol.	
  2011.	
  27:	
  870.e7-­‐10	
   Keating,	
  K.E.,	
  Giblin,	
  L.,	
  Lynch,	
  P.J.,	
  Quane,	
  K.A.,	
  Lehane,	
  M.,	
  Heffron,	
  J.J.	
  and	
  McCarthy,	
   T.V.	
   Detection	
   of	
   a	
   novel	
   mutation	
   in	
   the	
   ryanodine	
   receptor	
   gene	
   in	
   an	
   Irish	
   malignant	
  hyperthermia	
  pedigree:	
  correlation	
  of	
  the	
  IVCT	
  response	
  with	
  the	
  affected	
   and	
  unaffected	
  haplotypes.	
  J	
  Med	
  Genet.	
  1997.	
  34:	
  291-­‐6	
   Kelley,	
  L.A.	
  and	
  Sternberg,	
  M.J.E.	
  Protein	
  structure	
  prediction	
  on	
  the	
  web:	
  a	
  case	
  study	
   using	
  the	
  Phyre	
  serve.	
  Nature	
  Protocols.	
  2009.	
  4:	
  363-­‐71	
   Kim,	
   D.H.,	
   Ohnishi,	
   S.T.	
   and	
   Ikemoto,	
   N.	
   Kinetic	
   studies	
   of	
   calcium	
   release	
   from	
   sarcoplasmic	
  reticulum	
  in	
  vitro.	
  J.	
  Biol.	
  Chem.	
  1983.	
  258:	
  9662–8	
   Kimlicka,	
  L.	
  and	
  Van	
  Petegem	
  F.	
  Ryanodine	
  Receptor	
  Mutation	
  Database.	
  Unpublished	
   Kimlicka,	
   L.	
   and	
   Van	
   Petegem,	
   F.	
   2011.	
   The	
   structural	
   biology	
   of	
   ryanodine	
   receptors.	
   Sci	
  China	
  Life	
  Sci.	
  54:	
  712–724	
   Kobayashi,	
  S.,	
  Bannister,	
  M.L.,	
  Gangopadhyay,	
  J.P.,	
  Hamada,	
  T.,	
  Parness,	
  J.	
  and	
  Ikemoto,	
   N.	
   Dantrolene	
   stabilizes	
   domain	
   interactions	
   within	
   the	
   ryanodine	
   receptor.	
   J	
   Biol	
   Chem.	
  2005.	
  280:	
  6580-­‐7	
    	
    101	
    Kohn,	
   W.D.,	
   Mant,	
   C.T.	
   and	
   Hodges,	
   RS.	
   Alpha-­‐helical	
   protein	
   assembly	
   motifs.	
   J	
   Biol	
   Chem.	
  1997.	
  272:	
  2583-­‐6	
   Kolb,	
   M.E.,	
   Horne,	
   M.L.	
   and	
   Martz,	
   R.	
   Dantrolene	
   in	
   human	
   malignant	
   hyperthermia.	
   Anesthesiology.	
  1982.	
  56:	
  254–62	
   Kraeva,	
   N.,	
   Riazi,	
   S.,	
   Loke,	
   J.,	
   Frodis,	
   W.,	
   Crossan,	
   M.L.,	
   Nolan,	
   K.,	
   Kraev,	
   A.	
   and	
   MacLennan,	
   D.H.	
   Ryanodine	
   receptor	
   type	
   1	
   gene	
   mutations	
   found	
   in	
   the	
   Canadian	
   malignant	
  hyperthermia	
  population.	
  Can	
  J	
  Anaesth.	
  2011.	
  58:	
  504-­‐13	
   Kraus-­‐Friedmann,	
  N.	
  and	
  Feng,	
  L.	
  Reduction	
  of	
  ryanodine	
  binding	
  and	
  cytosolic	
  Ca2+	
   levels	
   in	
   liver	
   by	
   the	
   immunosuppressant	
   FK506.	
   Biochem.	
   Pharmacol.	
   1994.	
   48:	
   2157–62	
   Krause,	
  T.,	
  Gerbershagen,	
  M.U.,	
  Fiege,	
  M.,	
  Weishorn,	
  R.	
  and	
  Wappler,	
  F.	
  Dantrolene	
  –	
  a	
   review	
   of	
   its	
   pharmacology,	
   therapeutic	
   use	
   and	
   new	
   developments.	
   Anaesthesia.	
   2004	
  59:	
  364–373	
   Kuntz,	
   I.D.,	
   Blaney,	
   J.M.,	
   Oatley,	
   S.J.,	
   Langridge,	
   R.	
   and	
   Ferrin,	
   T.E.	
   A	
   geometric	
   approach	
  to	
  macromolecule-­‐ligand	
  interactions.	
  J.	
  Mol.	
  Biol.	
  1982	
  161:	
  269-­‐88	
   Landschulz,	
  W.H.,	
  Johnson,	
  P.F.	
  and	
  McKnight,	
  S.L.	
  The	
  leucine	
  zipper:	
  a	
  hypothetical	
   structure	
  common	
  to	
  a	
  new	
  class	
  of	
  DNA	
  binding	
  proteins.	
  Science.	
  1988.	
  240:	
  1759-­‐ 64	
   Laver,	
  D.R.,	
  Baynes,	
  T.M.,	
  Dulhunty,	
  A.F.	
  Magnesium	
  inhibition	
  of	
  ryanodine-­‐receptor	
   calcium	
   channels:	
   evidence	
   for	
   two	
   independent	
   mechanisms.	
   J	
   Membr	
   Biol.	
   1997.	
   156:	
  213-­‐29	
   Lebon,	
  G.,	
  Warne,	
  T.,	
  Edwards,	
  P.C.,	
  Bennett,	
  K.,	
  Langmead,	
  C.J.,	
  Leslie,	
  A.G.	
  and	
  Tate,	
   C.G.	
   Agonist-­‐bound	
   adenosine	
   A2A	
   receptor	
   structures	
   reveal	
   common	
   features	
   of	
   GPCR	
  activation.	
  Nature.	
  2011.	
  474:	
  521-­‐5	
    	
    102	
    Lee,	
  H.,	
  Kim,	
  D.C.,	
  Lee,	
  J.H.,	
  Cho,	
  Y.G.,	
  Lee,	
  H.S.,	
  Choi,	
  S.I.	
  and	
  Kim,	
  D.S.	
  Molecular	
  genetic	
   analysis	
   of	
   the	
   ryanodine	
   receptor	
   gene	
   (RYR1)	
   in	
   Korean	
   malignant	
   hyperthermia	
   families.	
  Korean	
  J	
  Lab	
  Med.	
  2010.	
  30:	
  702-­‐10	
   Levano,	
   S.,	
   Vukcevic,	
   M.,	
   Singer,	
   M.,	
   Matter,	
   A.,	
   Treves,	
   S.,	
   Urwyler,	
   A.	
   and	
   Girard,	
   T.	
   Increasing	
   the	
   number	
   of	
   diagnostic	
   mutations	
   in	
   malignant	
   hyperthermia.	
   Hum	
   Mutat.	
  2009.	
  30:	
  590-­‐8	
   Lin,	
   C.	
   C.,	
   Baek,	
   K.	
   and	
   Lu,	
   Z.	
   Apo	
   and	
   InsP3-­‐bound	
   crystal	
   structures	
   of	
   the	
   ligand-­‐ binding	
  domain	
  of	
  an	
  InsP3	
  receptor.	
  Nat.	
  Struct.	
  Mol.	
  Biol.	
  2011.	
  18:	
  1172–74	
   Liu,	
   Z.,	
   Wang,	
   R.,	
   Tian,	
   X.,	
   Zhong,	
   X.,	
   Gangopadhyay,	
   J.,	
   Cole,	
   R.,	
   Ikemoto,	
   N.,	
   Chen,	
   S.	
   R.,	
   and	
  Wagenknecht,	
  T.	
  Dynamic,	
  inter-­‐subunit	
  interactions	
  between	
  the	
  N-­‐terminal	
  and	
   central	
   mutation	
   regions	
   of	
   cardiac	
   ryanodine	
   receptor.	
   J	
   Cell	
   Sci.	
   2010.	
   123:	
   1775-­‐ 1784	
   Liu,	
   Z.,	
   Zhang,	
   J.,	
   Sharma,	
   M.R.,	
   Li,	
   P.,	
   Chen,	
   S.R.	
   and	
   Wagenknecht,	
   T.	
   Three-­‐ dimensional	
   reconstruction	
   of	
   the	
   recombinant	
   type	
   3	
   ryanodine	
   receptor	
   and	
   localization	
  of	
  its	
  amino	
  terminus.	
  Proc.	
  Natl	
  Acad.	
  Sci.	
  2001,	
  USA	
  98:	
  6104–6109	
   Lobo,	
   P.A.	
   and	
   Van	
   Petegem,	
   F.	
   Crystal	
   structures	
   of	
   the	
   N-­‐terminal	
   domains	
   of	
   cardiac	
   and	
   skeletal	
   muscle	
   ryanodine	
   receptors:	
   insights	
   into	
   disease	
   mutations.	
   Structure.	
  2009,	
  17	
  1505–1514	
   Lobo,	
   P.A.,	
   Kimlicka,	
   L.,	
   Tung,	
   C.C.,	
   Van	
   Petegem,	
   F.	
   The	
   deletion	
   of	
   exon	
   3	
   in	
   the	
   cardiac	
   ryanodine	
   receptor	
   is	
   rescued	
   by	
   β-­‐strand	
   switching.	
   Structure.	
   2011.	
   19:	
   790–8	
   Loke,	
   J.	
   and	
   MacLennan.	
   D.H.	
   Malignant	
   hyperthermia	
   and	
   central	
   core	
   disease:	
   disorders	
  of	
  Ca2+	
  release	
  channels.	
  Am	
  J	
  Med.	
  1998.	
  104:	
  470–86	
   Loke,	
   J.C.,	
   Kraev,	
   N.,	
   Sharma,	
   P.,	
   Du,	
   G.,	
   Patel,	
   L.,	
   Kraev,	
   A.	
   and	
   MacLennan	
   DH.	
   Detection	
  of	
  a	
  novel	
  ryanodine	
  receptor	
  subtype	
  1	
  mutation	
  (R328W)	
  in	
  a	
  malignant	
    	
    103	
    hyperthermia	
   family	
   by	
   sequencing	
   of	
   a	
   leukocyte	
   transcript.	
   Anesthesiology.	
   2003.	
   99:	
  297-­‐302	
   Ludtke	
   S.J.,	
   Tran	
   T.P.,	
   Ngo	
   Q.T.,	
   Moiseenkova-­‐Bell	
   V.Y.,	
   Chiu	
   W.	
   and	
   Serysheva	
   I.I.	
   Flexible	
  architecture	
  of	
  IP3R1	
  by	
  cryo-­‐EM.	
  Structure.	
  2011,	
  19:1192–1199	
   Ludtke,	
  S.J.,	
  Serysheva,	
  I.I.,	
  Hamilton,	
  S.L.	
  and	
  Chiu,	
  W.	
  The	
  pore	
  structure	
  of	
  the	
  closed	
   RyR1	
  channel.	
  Structure,	
  2005,	
  13:	
  1203–1211	
   Lukacs,	
   C.M.,	
   Oikonomakos,	
   N.G.,	
   Crowther,	
   R.L.,	
   Hong,	
   L.N.,	
   Kammlott,	
   R.U.,	
   Levin,	
   W.,	
   Li,	
   S.,	
   Liu,	
   C.M.,	
   Lucas-­‐McGady,	
   D.,	
   Pietranico,	
   S.	
   and	
   Reik,	
   L.	
   The	
   crystal	
   structure	
   of	
   human	
   muscle	
   glycogen	
   phosphorylase	
   a	
   with	
   bound	
   glucose	
   and	
   AMP:	
   an	
   intermediate	
   conformation	
   with	
   T-­‐state	
   and	
   R-­‐state	
   features.	
   Proteins.	
   2006.	
   63:	
   1123-­‐6	
   Lupas,	
   A.	
   Coiled	
   coils:	
   new	
   structures	
   and	
   new	
   functions.	
   Trends	
   Biochem	
   Sci.	
   1996.	
   21:	
  375-­‐82	
   Ma,	
   J.,	
   Bhat,	
   M.B.,	
   and	
   Zhao,	
   J.	
   Rectification	
   of	
   skeletal	
   muscle	
   ryanodine	
   receptor	
   mediated	
  by	
  FK506	
  binding	
  protein.	
  Biophys	
  J.	
  1995.	
  69:	
  2398–2404	
   Mackey,	
   M.D.	
   and	
   Melville,	
   J.L.	
   Better	
   than	
   random?	
   The	
   chemotype	
   enrichment	
   problem.	
  J	
  Chem	
  Inf	
  Model.	
  2009.	
  49:	
  1154-­‐62	
   Maclennan,	
   D.H.,	
   Abu-­‐Abed	
   M.,	
   Kang	
   C.	
   Structure-­‐function	
   relationships	
   in	
   Ca	
   2+	
   cycling	
  proteins.	
  J	
  Mol	
  Cell	
  Cardiol.	
  2002.	
  34:	
  897–918	
   Marx,	
   S.O.,	
   Reiken,	
   S.,	
   Hisamatsu,	
   Y.,	
   Gaburjakova,	
   M.,	
   Gaburjakova,	
   J.,	
   Yang,	
   Y.M.,	
   Rosemblit,	
   N.	
   and	
   Marks,	
   A.R.	
   Phosphorylation-­‐dependent	
   regulation	
   of	
   ryanodine	
   receptors:	
  a	
  novel	
  role	
  for	
  leucine/isoleucine	
  zippers.	
  J	
  Cell	
  Biol.	
  2001.	
  153:	
  699-­‐708.	
   Marx,	
   S.O.,	
   Reiken,	
   S.,	
   Hisamatsu,	
   Y.,	
   Jayaraman,	
   T.,	
   Burkhoff,	
   D.,	
   Rosemblit,	
   N.	
   and	
   Marks,	
   A.R.	
   PKA	
   phosphorylation	
   dissociates	
   FKBP12.6	
   from	
   the	
   calcium	
   release	
    	
    104	
    channel	
   (Ryanodine	
   receptor):	
   defective	
   regulation	
   in	
   failing	
   hearts.	
   Cell.	
   2000	
   101	
   :365–76	
  	
  	
   Maximciuc,	
  A.A.,	
  Putkey,	
  J.A.,	
  Shamoo,	
  Y.	
  and	
  MacKenzie,	
  K.R.	
  Complex	
  of	
  Calmodulin	
   with	
  a	
  Ryanodine	
  Receptor	
  Target	
  Reveals	
  a	
  Novel,	
  Flexible	
  Binding	
  Mode.	
  Structure.	
   2006.	
  14:	
  1547–1556	
   Maxwell,	
   J.T.,	
   Domeier,	
   T.L.	
   and	
   Blatter,	
   L.A.	
   Dantrolene	
   prevents	
   arrhythmogenic	
   Ca2+	
  release	
  in	
  heart	
  failure.	
  Am	
  J	
  Physiol	
  Heart	
  Circ	
  Physiol.	
  2012.	
  302:	
  H953-­‐63	
   Mayrleitner,	
   M.,	
   Chandler,	
   R.,	
   Schindler,	
   H.	
   and	
   Fleischer	
   S.	
   Phosphorylation	
   with	
   protein	
   kinases	
   modulates	
   calcium	
   loading	
   of	
   terminal	
   cisternae	
   of	
   sarcoplasmic	
   reticulum	
  from	
  skeletal	
  muscle.	
  Cell	
  Calcium.	
  1995.	
  18:	
  197–206	
   McCarthy,	
  T.V.,	
  Quane,	
  K.A.	
  and	
  Lynch,	
  P.J.	
  Ryanodine	
  receptor	
  mutations	
  in	
  malignant	
   hyperthermia	
  and	
  central	
  core	
  disease.	
  Hum	
  Mutat.	
  2000.	
  15:	
  410–7	
   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	
  mutational	
   analysis.	
  J	
  Am	
  Coll	
  Cardiol.	
  2009.	
  54:	
  2065-­‐74	
   Meissner,	
   G.	
   Adenine	
   nucleotide	
   stimulation	
   of	
   Ca2+-­‐induced	
   Ca2+	
   release	
   in	
   sarcoplasmic	
  reticulum.	
  J.	
  Biol.	
  Chem.	
  1984.	
  259:	
  2365–2374	
   Meissner,	
   G.	
   Ryanodine	
   activation	
   and	
   inhibition	
   of	
   the	
   Ca2+	
   release	
   channel	
   of	
   sarcoplasmic	
  reticulum.	
  J	
  Biol	
  Chem.	
  1986.	
  261:	
  6300–6306	
   Meissner,	
  G.,	
  Darling,	
  E.	
  and	
  Eveleth,	
  J.	
  Kinetics	
  of	
  rapid	
  Ca2+	
  release	
  by	
  sarcoplasmic	
   reticulum.	
  Effects	
  of	
  Ca2+,	
  Mg2+,	
  and	
  adenine	
  nucleotides.	
  Biochemistry.	
  1986.	
  25:	
  236– 244	
    	
    105	
    Melzer,	
  W.,	
  Herrmann-­‐Frank,	
  A.	
  and	
  Lüttgau,	
  H.C.	
  The	
  role	
  of	
  Ca	
   2+	
  ions	
  in	
  excitation-­‐ contraction	
  coupling	
  of	
  skeletal	
  muscle	
  fibres.	
  Biochim	
  Biophys	
  Acta.	
  1995.	
  1241:	
  59– 116	
  	
   Menegazzi,	
   P.,	
   Larini,	
   F.,	
   Treves,	
   S.,	
   Guerrini,	
   R.,	
   Quadroni,	
   M.	
   and	
   Zorzato,	
   F.	
   Identification	
   and	
   characterization	
   of	
   three	
   calmodulin	
   binding	
   sites	
   of	
   the	
   skeletal	
   muscle	
  ryanodine	
  receptor.	
  Biochemistry.	
  1994.	
  33:	
  9078–9084	
   Miledi,	
   R.,	
   Parker,	
   I.	
   and	
   Zhu,	
   P.H.	
   Extracellular	
   ions	
   and	
   excitation-­‐contraction	
   coupling	
  in	
  frog	
  twitch	
  muscle	
  fibres.	
  Journal	
  of	
  Physiology,	
  1984.	
  351:	
  687-­‐710	
   Minor,	
   W.,	
   Cymborowski,	
   M.,	
   Otwinowski,	
   Z.	
   and	
   Chruszcz,	
   M.	
   HKL-­‐3000:	
   the	
   integration	
   of	
   data	
   reduction	
   and	
   structure	
   solution	
   -­‐	
   from	
   diffraction	
   images	
   to	
   an	
   initial	
  model	
  in	
  minutes.	
  Acta	
  Cryst.	
  2006.	
  D62:	
  859-­‐66	
   Monnier,	
  N.,	
  Kozak-­‐Ribbens,	
  G.,	
  Krivosic-­‐Horber,	
  R.,	
  Nivoche,	
  Y.,	
  Qi,	
  D.,	
  Kraev,	
  N.,	
  Loke,	
   J.,	
  Sharma,	
  P.,	
  Tegazzin,	
  V.,	
  Figarella-­‐Branger,	
  D.,	
  Roméro,	
  N.,	
  Mezin,	
  P.,	
  Bendahan,	
  D.,	
   Payen	
  J.F.,	
  Depret,	
  T.,	
  Maclennan,	
  D.H.	
  and	
  Lunardi,	
  J.	
  Correlations	
  between	
  genotype	
   and	
   pharmacological,	
   histological,	
   functional,	
   and	
   clinical	
   phenotypes	
   in	
   malignant	
   hyperthermia	
  susceptibility.	
  Hum	
  Mutat.	
  2005.	
  26:413-­‐25	
   Monnier,	
  N.,	
  Romero,	
  N.B.,	
  Lerale,	
  J.,	
  Landrieu,	
  P.,	
  Nivoche,	
  Y.,	
  Fardeau,	
  M.	
  and	
  Lunardi,	
   J.	
  Familial	
  and	
  sporadic	
  forms	
  of	
  central	
  core	
  disease	
  are	
  associated	
  with	
  mutations	
  in	
   the	
   C-­‐terminal	
   domain	
   of	
   the	
   skeletal	
   muscle	
   ryanodine	
   receptor.	
   Hum	
   Mol	
   Genet.	
   2001.	
  10:	
  2581-­‐92	
   Moore,	
  C.P.,	
  Rodney,	
  G.,	
  Zhang,	
  J.Z.,	
  Santacruz-­‐Tolozak,	
  L.,	
  Strasburg,	
  G.	
  and	
  Hamilton,	
   S.L.	
   Apocalmodulin	
   and	
   Ca2+	
   calmodulin	
   bind	
   to	
   the	
   same	
   region	
   on	
   the	
   skeletal	
   muscle	
  Ca2+	
  release	
  channel.	
  Biochemistry.	
  1999.	
  38:	
  8532–8537	
   Morii,	
  H.	
  and	
  Tonomura,	
  Y.	
  The	
  gating	
  behavior	
  of	
  a	
  channel	
  for	
  Ca2+-­‐induced	
  Ca2+	
   release	
  in	
  fragmented	
  sarcoplasmic	
  reticulum.	
  J.	
  Biochem.	
  1983.	
  93:	
  1271–1285	
    	
    106	
    Morris,	
  G.	
  M.,	
  Huey,	
  R.,	
  Lindstrom,	
  W.,	
  Sanner,	
  M.	
  F.,	
  Belew,	
  R.	
  K.,	
  Goodsell,	
  D.	
  S.	
  and	
   Olson,	
   A.	
   J.	
   AutoDock4	
   and	
   AutoDockTools4:	
   automated	
   docking	
   with	
   selective	
   receptor	
  flexiblity.	
  J.	
  Computational	
  Chemistry.	
  2009.	
  16:	
  2785-­‐91	
   Murshudov,	
   G.N.,	
   Vagin,	
   A.A.	
   and	
   Dodson,	
   E.J.,	
   Refinement	
   of	
   macromolecular	
   structures	
   by	
   the	
   maximum-­‐likelihood	
   method.	
   Acta	
   Crystallogr.	
   D	
   Biol.	
   Crystallogr.	
   1997.	
  53:	
  240–55	
   Nagasaki,	
   K.	
   and	
   Kasai,	
   M.	
   Fast	
   release	
   of	
   calcium	
   from	
   sarcoplasmic	
   reticulum	
   vesicles	
  monitored	
  by	
  chlortetracycline	
  fluorescence.	
  J.	
  Biochem.	
  94:	
  1101–1109	
   Nelson,	
  T.	
  E.,	
  Lin,	
  M.,	
  Zapata-­‐Sudo,	
  G.	
  and	
  Sudo,	
  R.	
  T.	
  Dantrolene	
  sodium	
  can	
  increase	
   or	
   attenuate	
   activity	
   of	
   skeletal	
   muscle	
   ryanodine	
   receptor	
   calcium	
   release	
   channel.	
   Anesthesiology.	
  1996.	
  84:	
  1368–79	
   Newton	
   C.L.,	
   Mignery,	
   G.A.	
   and	
   Südhof,	
   T.C.	
   Co-­‐expression	
   in	
   vertebrate	
   tissues	
   and	
   cell	
   lines	
   of	
   multiple	
   inositol	
   1,4,5-­‐trisphosphate	
   (InsP3)	
   receptors	
   with	
   distinct	
   affinities	
  for	
  InsP3.	
  J	
  Biol	
  Chem.	
  1994.	
  269:	
  28613-­‐9	
   Nof,	
  E.,	
  Belhassen,	
  B.,	
  Arad,	
  M.,	
  Bhuiyan,	
  Z.A.,	
  Antzelevitch,	
  C.,	
  Rosso,	
  R.,	
  Fogelman,	
  R.,	
   Luria,	
   D.,	
   El-­‐Ani,	
   D.,	
   Mannens,	
   M.M.,	
   Viskin,	
   S.,	
   Eldar,	
   M.,	
   Wilde,	
   A.A.	
   and	
   Glikson,	
   M.	
   Postpacing	
   abnormal	
   repolarization	
   in	
   catecholaminergic	
   polymorphic	
   ventricular	
   tachycardia	
   associated	
   with	
   a	
   mutation	
   in	
   the	
   cardiac	
   ryanodine	
   receptor	
   gene.	
   Heart	
   Rhythm.	
  2011.	
  8:	
  1546-­‐52	
   Oda,	
  T.,	
  Yano,	
  M.,	
  Yamamoto,	
  T.,	
  Tokuhisa,	
  T.,	
  Okuda,	
  S.,	
  Doi,	
  M.,	
  Ohkusa,	
  T.,	
  Ikeda,	
  Y.,	
   Kobayashi,	
   S.,	
   Ikemoto,	
   N.,	
   and	
   Matsuzaki,	
   M.	
   Defective	
   regulation	
   of	
   interdomain	
   interactions	
   within	
   the	
   ryanodine	
   receptor	
   plays	
   a	
   key	
   role	
   in	
   the	
   pathogenesis	
   of	
   heart	
  failure.	
  Circulation.	
  2005.	
  111:	
  3400-­‐3410	
   Ogawa	
   Y.,	
   Harafuji	
   H.	
   Osmolarity-­‐dependent	
   characteristics	
   of	
   [3H]ryanodine	
   binding	
   to	
  sarcoplasmic	
  reticulum.	
  J.	
  Biochem.	
  1990.	
  107:	
  894–898	
    	
    107	
    Oikonomakos,	
   N.G.,	
   Schnier,	
   J.B.,	
   Zographos,	
   S.E.,	
   Skamnaki,	
   V.T.,	
   Tsitsanou,	
   K.E.	
   and	
   Johnson,	
  L.N.	
  Flavopiridol	
  inhibits	
  glycogen	
  phosphorylase	
  by	
  binding	
  at	
  the	
  inhibitor	
   site.	
  J	
  Biol	
  Chem.	
  2000.	
  275:	
  34566-­‐73	
   Otsu,	
   K.,	
   Willard,	
   H.F.,	
   Khanna,	
   V.K.,	
   Zorzato,	
   F.,	
   Green,	
   N.M.	
   and	
   MacLennan,	
   D.H.	
   Molecular	
   cloning	
   of	
   cDNA	
   encoding	
   the	
   Ca2+	
   release	
   channel	
   (ryanodine	
   receptor)	
   of	
   rabbit	
  cardiac	
  muscle	
  sarcoplasmic	
  reticulum.	
  J	
  Biol	
  Chem,	
  1990,	
  265:	
  13472–13483	
   Palnitkar,	
  S.S.,	
  Bin,	
  B.,	
  Jimenez,	
  L.S.,	
  Morimoto,	
  H.,	
  Williams,	
  P.G.,	
  Paul-­‐Pletzer,	
  K.	
  and	
   Parness,	
   J.	
   [3	
   H]Azidodantrolene:	
   synthesis	
   and	
   use	
   in	
   identifi	
   cation	
   of	
   a	
   putative	
   skeletal	
  muscle	
  dantrolene	
  binding	
  site	
  in	
  sarcoplasmic	
  reticulum.	
  J	
  Med	
  Chem.	
  1999.	
   42:1872–1880	
  	
  	
   Palnitkar,	
   S.S.,	
   Mickelson,	
   J.R.,	
   Louis,	
   C.F.	
   and	
   Parness,	
   J.	
   Pharmacological	
   distinction	
   between	
   dantrolene	
   and	
   ryanodine	
   binding	
   sites:	
   evidence	
   from	
   normal	
   and	
   malignant	
   hyperthermia-­‐susceptible	
   porcine	
   skeletal	
   muscle.	
   Biochem	
   J.	
   1997.	
   326:	
   847-­‐52	
   Pamukcoglu,	
   T.	
   Sudden	
   death	
   due	
   to	
   malignant	
   hyperthermia.	
   Am	
   J	
   Forensic	
   Med	
   Pathol.	
  1988.	
  9:	
  161–162	
   Park,	
   H.,	
   Park,	
   I.Y.,	
   Kim,	
   E.,	
   Youn,	
   B.,	
   Fields,	
   K.,	
   Dunker,	
   A.K.	
   and	
   Kang,	
   C.	
   Comparing	
   skeletal	
   and	
   cardiac	
   calsequestrin	
   structures	
   and	
   their	
   calcium	
   binding:	
   a	
   proposed	
   mechanism	
   for	
   coupled	
   calcium	
   binding	
   and	
   protein	
   polymerization.	
   J	
   Biol	
   Chem.	
   2004.	
  279:	
  18026-­‐33	
   Parness,	
   J.	
   and	
   Palnitkar,	
   S.S.	
   Identification	
   of	
   dantrolene	
   binding	
   sites	
   in	
   porcine	
   skeletal	
  muscle	
  sarcoplasmic	
  reticulum.	
  J	
  Biol	
  Chem.	
  1995.	
  270:	
  18465–18472	
   Paul-­‐Pletzer,	
  K.,	
  Palnitkar,	
  S.S.,	
  Jimenez	
  L.S.,	
  Morimoto	
  H.	
  and	
  Parness,	
  J.	
  The	
  skeletal	
   muscle	
   ryanodine	
   receptor	
   identified	
   as	
   a	
   molecular	
   target	
   of	
   [3H]azidodantrolene	
   by	
   photoaffinity	
  labeling	
  Biochemistry.	
  2001.	
  40:	
  531–542	
    	
    108	
    Paul-­‐Pletzer,	
  K.,	
  Yamamoto,	
  T.,	
  Bhat,	
  M.B.,	
  Ma,	
  J.,	
  Ikemoto,	
  N.,	
  Jimenez,	
  L.S.,	
  Morimoto,	
   H.,	
   Williams,	
   P.G.	
   and	
   Parness,	
   J.	
   Identification	
   of	
   a	
   dantrolene-­‐binding	
   sequence	
   on	
   the	
  skeletal	
  muscle	
  ryanodine	
  receptor.	
  J	
  Biol	
  Chem,	
  2002,	
  277:	
  34918–34923	
   Paul-­‐Pletzer,	
   K.,	
   Yamamoto,	
   T.,	
   Ikemoto,	
   N.,	
   Jimenez,	
   L.S.,	
   Morimoto,	
   H.,	
   Williams,	
   P.G.,	
   Ma,	
   J.	
   and	
   Parness,	
   J.	
   Probing	
   a	
   putative	
   dantrolene-­‐binding	
   site	
   on	
   the	
   cardiac	
   ryanodine	
  receptor.	
  Biochem	
  J,	
  2005,	
  387:	
  905–909	
   Pessah,	
   I.N.,	
   Waterhouse,	
   A.L.	
   and	
   Casida,	
   J.E.	
   The	
   calcium-­‐ryanodine	
   receptor	
   complex	
   of	
   skeletal	
   and	
   cardiac	
   muscle.	
   Biochem.	
   Biophys.	
   Res.	
   Commun.	
   1985.	
   128:	
   449–56	
   Pettersen,	
  E.F.,	
  Goddard,	
  T.D.,	
  Huang,	
  C.C.,	
  Couch,	
  G.S.,	
  Greenblatt,	
  D.M.,	
  Meng,	
  E.C.	
  and	
   Ferrin,	
   T.E.	
   UCSF	
   Chimera	
   -­‐	
   a	
   visualization	
   system	
   for	
   exploratory	
   research	
   and	
   analysis.	
  J	
  Comput	
  Chem.	
  2004.	
  25	
  :1605-­‐12.	
   Postma,	
   A.V.,	
   Denjoy,	
   I.,	
   Kamblock,	
   J.,	
   Alders,	
   M.,	
   Lupoglazoff,	
   JM,	
   Vaksmann,	
   G.,	
   Dubosq-­‐Bidot,	
   L.,	
   Sebillon,	
   P.,	
   Mannens,	
   M.,	
   Guicheney,	
   P.	
   and	
   Wilde,	
   A.A.	
   Catecholaminergic	
    polymorphic	
    ventricular	
    tachycardia:	
    RYR2	
    mutations,	
    bradycardia,	
  and	
  follow	
  up	
  of	
  the	
  patients.	
  J	
  Med	
  Genet.	
  2005.	
  42:	
  863-­‐70	
   Priori,	
  S.G.,	
  Napolitano,	
  C.,	
  Tiso,	
  N.,	
  Memmi,	
  M.,	
  Vignati,	
  G.,	
  Bloise,	
  R.,	
  Sorrentino,	
  V.	
  and	
   Danieli,	
   G.A.	
   Mutations	
   in	
   the	
   cardiac	
   ryanodine	
   receptor	
   gene	
   (hRyR2)	
   underlie	
   catecholaminergic	
  polymorphic	
  ventricular	
  tachycardia.	
  Circulation.	
  2001.	
  103:	
  196-­‐ 200	
   Quane,	
   K.A.,	
   Healy,	
   J.M.,	
   Keating,	
   K.E.,	
   Manning,	
   B.M.,	
   Couch,	
   F.J.,	
   Palmucci,	
   L.M.,	
   Doriguzzi,	
   C.,	
   Fagerlund,	
   T.H.,	
   Berg,	
   K.,	
   Ording	
   H.,	
   Bendixen,	
   D.,	
   Mortier,	
   W.,	
   Linz,	
   U.,	
   Muller,	
   C.R.	
   and	
   McCarthy,	
   T.V.	
   Mutations	
   in	
   the	
   ryanodine	
   receptor	
   gene	
   in	
   central	
   core	
  disease	
  and	
  malignant	
  hyperthermia.	
  Nat	
  Genet.	
  1993.	
  5:	
  51–5	
   Quane,	
   K.A.,	
   Keating,	
   K.E.,	
   Healy,	
   JM,	
   Manning,	
   B.M.,	
   Krivosic-­‐Horber,	
   R.,	
   Krivosic,	
   I.,	
   Monnier,	
   N.,	
   Lunardi,	
   J.	
   and	
   McCarthy,	
   T.V.	
   Mutation	
   screening	
   of	
   the	
   RYR1	
   gene	
   in	
    	
    109	
    malignant	
  hyperthermia:	
  detection	
  of	
  a	
  novel	
  Tyr	
  to	
  Ser	
  mutation	
  in	
  a	
  pedigree	
  with	
   associated	
  central	
  cores.	
  Genomics.	
  1994.	
  23:	
  236-­‐9	
   Rao,	
   F.V.,	
   Andersen,	
   O.A.,	
   Vora,	
   K.A.,	
   Demartino,	
   J.A.	
   and	
   Van	
   Aalten,	
   D.M.F.	
   Methylxanthine	
   drugs	
   are	
   chitinase	
   inhibitors:	
   investigation	
   of	
   inhibition	
   and	
   binding	
   modes.	
  Chem	
  Biol.	
  2005.	
  12:	
  973-­‐80	
   Reiken,	
   S.,	
   Lacampagne,	
   A.,	
   Zhou,	
   H.,	
   Kherani,	
   A.,	
   Lehnart,	
   S.E.,	
   Ward,	
   C.,	
   Huang,	
   F.,	
   Gaburjakova,	
  M.,	
  Gaburjakova,	
  J.,	
  Rosemblit,	
  N.,	
  Warren,	
  M.S.,	
  He,	
  K-­‐l.,	
  Yi,	
  G-­‐h.,	
  Wang,	
  J.,	
   Burkhoff,	
   D.,	
   Vassort,	
   G.	
   and	
   Marks,	
   A.R.	
   PKA	
   phosphorylation	
   activates	
   the	
   calcium	
   release	
   channel	
   (ryanodine	
   receptor)	
   in	
   skeletal	
   muscle:	
   defective	
   regulation	
   in	
   heart	
   failure.	
  J	
  Cell	
  Biol.	
  2003.	
  160	
  :919–28	
  	
   Revel,	
  J.P.	
  The	
  sarcoplasmic	
  reticulum	
  of	
  the	
  bat	
  cricothroid	
  muscle.	
  J	
  Cell	
  Biol,	
  1962,	
   12:	
  571–588	
   Rios,	
   E.	
   and	
   Brum,	
   G.	
   Involvement	
   of	
   dihydropyridine	
   receptors	
   in	
   excitation-­‐ contraction	
  coupling	
  in	
  skeletal	
  muscle.	
  Nature.1987.	
  325:	
  717–720	
   Robinson,	
  R.,	
  Carpenter,	
  D.,	
  Shaw,	
  M.A.,	
  Halsall,	
  J.	
  and	
  Hopkins,	
  P.	
  Mutations	
  in	
  RYR1	
   in	
  malignant	
  hyperthermia	
  and	
  central	
  core	
  disease.	
  Hum	
  Mutat.	
  2006.	
  27:	
  977-­‐89	
   Rodney,	
   G.G,.	
   Williams,	
   B.Y.,	
   Strasburg,	
   G.M.,	
   Beckingham,	
   K.	
   and	
   Hamilton,	
   S.L.	
   Regulation	
  of	
  RYR1	
  activity	
  by	
  Ca2+	
  and	
  calmodulin.	
  Biochem.	
  2000.	
  39:	
  7807–7812	
   Romero,	
   N.B.,	
   Monnier,	
   N.,	
   Viollet,	
   L.,	
   Cortey,	
   A.,	
   Chevallay,	
   M.,	
   Leroy,	
   J.P.,	
   Lunardi,	
   J.	
   and	
   Fardeau,	
   M.	
   Dominant	
   and	
   recessive	
   central	
   core	
   disease	
   associated	
   with	
   RYR1	
   mutations	
  and	
  fetal	
  akinesia.	
  Brain.	
  2003.	
  126:	
  2341-­‐9	
   Rosenberg,	
  H.,	
  Davis,	
  M.,	
  James,	
  D.,	
  Pollock,	
  N.	
  and	
  Stowell,	
  K.	
  Malignant	
  hyperthermia.	
   Orphanet	
  J	
  Rare	
  Dis.	
  2007.	
  2:	
  21	
    	
    110	
    Rousseau,	
   E.,	
   Ladine,	
   J.,	
   Liu,	
   Q.Y.	
   and	
   Meissner,	
   G.	
   Activation	
   of	
   the	
   Ca2+	
   release	
   channel	
   of	
   skeletal	
   muscle	
   sarcoplasmic	
   reticulum	
   by	
   caffeine	
   and	
   related	
   compounds.	
  Arch.	
  Biochem.	
  Biophys.	
  1988.	
  267:	
  75–86	
   Saito,	
   A.,	
   Inui,	
   M.,	
   Radermacher,	
   M.,	
   Frank,	
   J.	
   and	
   Fleischer,	
   S.	
   Ultrastructure	
   of	
   the	
   calcium	
  release	
  channel	
  of	
  sarcoplasmic	
  reticulum.	
  J	
  Cell	
  Biol,	
  1988,	
  107:	
  211–219	
   Saito,	
   A.,	
   Seiler,	
   S.,	
   Chu,	
   A.	
   and	
   Fleischer,	
   S.	
   Preparation	
   and	
   morphology	
   of	
   sarcoplasmic	
   reticulum	
   terminal	
   cisternae	
   from	
   rabbit	
   skeletal	
   muscle.	
   J	
   Cell	
   Biol.	
   1984,	
  99:	
  875–885	
   Sambuughin,	
   N.,	
   Holley,	
   H.,	
   Muldoon,	
   S.,	
   Brandom,	
   B.W.,	
   de	
   Bantel,	
   A.M.,	
   Tobin,	
   J.R.,	
   Nelson,	
   T.E.	
   and	
   Goldfarb,	
   L.G.	
   Screening	
   of	
   the	
   entire	
   ryanodine	
   receptor	
   type	
   1	
   coding	
   region	
   for	
   sequence	
   variants	
   associated	
   with	
   malignant	
   hyperthermia	
   susceptibility	
  in	
  the	
  north	
  american	
  population.	
  Anesthesiology.	
  2005.	
  102:	
  515-­‐21	
   Samsó,	
   M.	
   and	
   Wagenknecht,	
   T.	
   Apocalmodulin	
   and	
   Ca2+-­‐Calmodulin	
   Bind	
   to	
   Neighboring	
   Locations	
   on	
   the	
   Ryanodine	
   Receptor.	
   J	
   Biol	
   Chem.	
   2002	
   	
   277:	
   1349– 1353.	
   Samsó,	
  M.,	
  Feng,	
  W.,	
  Pessah,	
  I.N.	
  and	
  Allen,	
  P.D.	
  Coordinated	
  movement	
  of	
  cytoplasmic	
   and	
  transmembrane	
  domains	
  of	
  RyR1	
  upon	
  gating.	
  PLoS	
  Biol,	
  2009,	
  7:	
  e85	
   Samsó,	
   M.,	
   Wagenknecht,	
   T.	
   and	
   Allen	
   P.D.	
   Internal	
   structure	
   and	
   visualization	
   of	
   transmembrane	
   domains	
   of	
   the	
   RyR1	
   calcium	
   release	
   channel	
   by	
   cryo-­‐EM.	
   Nat	
   Struct	
   Mol	
  Biol,	
  2005,	
  12:	
  539–544	
   Sandow,	
   A.	
   Excitation-­‐contraction	
   coupling	
   in	
   muscular	
   response.	
   Yale	
   J	
   Biol	
   Med.	
   1952.	
  25:	
  176-­‐201	
   Seo,	
  M.D.,	
  Velamakanni,	
  S.,	
  Ishiyama,	
  N.,	
  Stathopulos,	
  P.B.,	
  Rossi,	
  A.M.,	
  Khan,	
  S.A.,	
  Dale,	
   P.,	
  Li,	
  C.,	
  Ames,	
  J.B.,	
  Ikura,	
  M.,	
  and	
  Taylor,	
  C.W.	
  Structural	
  and	
  functional	
  conservation	
   of	
  key	
  domains	
  in	
  InsP3	
  and	
  ryanodine	
  receptors.	
  Nature.	
  2012.	
  483:	
  108–12	
    	
    111	
    Serysheva,	
  I.I.,	
  Bare,	
  D.J.,	
  Ludtke,	
  S.J.,	
  Kettlun,	
  C.S.,	
  Chiu,	
  W.	
  and	
  Mignery,	
  G.A.	
  Structure	
   of	
   the	
   type	
   1	
   inositol	
   1,	
   4,	
   5-­‐trisphosphate	
   receptor	
   revealed	
   by	
   electron	
   cryomicroscopy.	
  J	
  Biol	
  Chem,	
  2003.	
  24:	
  21319-­‐22	
   Serysheva,	
   I.I.,	
   Hamilton,	
   S.L.,	
   Chiu,	
   W.	
   and	
   Ludtke,	
   S.J.,	
   Structure	
   of	
   Ca2+	
   release	
   channel	
  at	
  14	
  Å	
  resolution.	
  J	
  Mol	
  Biol.	
  	
  2005,	
  345:	
  427-­‐431	
   Serysheva,	
   I.I.,	
   Ludtke,	
   S.J.,	
   Baker,	
   M.L.,	
   Cong,	
   Y.,	
   Topf,	
   M.,	
   Eramian,	
   D.,	
   Sali,	
   A.,	
   Hamilton,	
  S.L.	
  and	
  Chiu,	
  W.	
  Proc	
  Natl	
  Acad	
  Sci	
  USA,	
  2008,	
  105:	
  9610–9615	
   Serysheva,	
  I.I.,	
  Orlova,	
  E.V.,	
  Chiu,	
  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.	
  Nat	
  Struct	
  Biol,	
  1995,	
  2:	
  18–24	
   Sharma,	
   P.,	
   Ishiyama,	
   N.,	
   Nair,	
   U.,	
   Li,	
   W.,	
   Dong,	
   A.,	
   Miyake,	
   T.,	
   Wilson,	
   A.,	
   Ryan,	
   T.,	
   MacLennan,	
   D.H.,	
   Kislinger,	
   T.,	
   Ikura,	
   M.,	
   Dhe-­‐Paganon,	
   S.	
   and	
   Gramolini,	
   A.O.	
   Structural	
   determination	
   of	
   the	
   phosphorylation	
   domain	
   of	
   the	
   ryanodine	
   receptor.	
   FEBS	
  J.	
  2012,	
  doi:	
  10.1111/j.1742-­‐4658.2012.08755.x	
   Smith,	
   J.S.,	
   Rousseau,	
   E.	
   and	
   Meissner,	
   G.	
   Calmodulin	
   modulation	
   of	
   single	
   sarcoplasmic	
   reticulum	
   Ca-­‐release	
   channels	
   from	
   cardiac	
   and	
   skeletal	
   muscle.	
   Circ	
   Res.	
  1989.	
  64:	
  352–359	
   Snyder,	
   H.R.	
   Jr,	
   Davis,	
   C.S.,	
   Bickerton,	
   R.K.	
   and	
   Halliday,	
   R.P.	
   1-­‐[(5-­‐arylfurfurylidene)	
   amino]-­‐hydantoins.	
  A	
  new	
  class	
  of	
  muscle	
  relaxants.	
  J	
  Med	
  Chem.	
  1967.	
  10:	
  807–10	
   Sorrentino,	
  V.	
  and	
  Volpe,	
  P.	
  Ryanodine	
  receptors:	
  How	
  many,	
  where	
  and	
  why?	
  Trends	
   Pharmacol	
  Sci,	
  1993,	
  14:	
  98–103	
   Stephenson,	
   D.G.,	
   Lamb,	
   G.D.	
   and	
   Stephenson,	
   G.M.	
   Events	
   of	
   the	
   excitation-­‐ contractionrelaxation	
   (E-­‐C-­‐R)	
   cycle	
   in	
   fast-­‐	
   and	
   slow-­‐twitch	
   mammalian	
   muscle	
   fi	
   bres	
  relevant	
  to	
  muscle	
  fatigue.	
  Acta	
  Physiol	
  Scand.	
  1998.	
  162	
  :229–245	
    	
    112	
    Storoni,	
  	
  L.C.,	
  McCoy,	
  A.J.	
  	
  and	
  Read,	
  R.J.	
  Likelihood-­‐enhanced	
  fast	
  rotation	
  functions.	
   Acta	
  Cryst.	
  2004.	
  D60:	
  432-­‐8	
   Streiff,	
   J.H.,	
   Juranic,	
   N.O.,	
   Macura,	
   S.I.,	
   Warner,	
   D.O.,	
   Jones,	
   K.A.	
   and	
   Perkins,	
   WJ.	
   Saturation	
  transfer	
  difference	
  nuclear	
  magnetic	
  resonance	
  spectroscopy	
  as	
  a	
  method	
   for	
  screening	
  proteins	
  for	
  anesthetic	
  binding.	
  Mol	
  Pharmacol.	
  2004.	
  66:	
  929-­‐35	
   Sudo,	
   R.T.,	
   Carmo,	
   P.L.,	
   Trachez,	
   M.M.	
   and	
   Zapata-­‐Sudo,	
   G.	
   Effects	
   of	
   azumolene	
   on	
   normal	
   and	
   malignant	
   hyperthermia-­‐susceptible	
   skeletal	
   muscle.	
   Basic	
   Clin	
   Pharmacol	
  Toxicol.	
  2008.	
  102:	
  308–16	
   Suko,	
   J.,	
   Maurer-­‐Fogy,	
   I.,	
   Plank,	
   B.,	
   Bertel,	
   O.,	
   Wyskovsky,	
   W.,	
   Hohenegger,	
   M.	
   and	
   Hellmann,	
   G.	
   Phosphorylation	
   of	
   serine	
   2843	
   in	
   ryanodine	
   receptor-­‐calcium	
   release	
   channel	
   of	
   skeletal	
   muscle	
   by	
   cAMP-­‐,	
   cGMP-­‐	
   and	
   CaM-­‐dependent	
   protein	
   kinase.	
   Biochim.	
  Biophys.	
  Acta.	
  1993.	
  1175:	
  193–206	
   Szentesi,	
   P.,	
   Collet,	
   C.,	
   Sárközi,	
   S.,	
   Szegedi,	
   C.,	
   Jona,	
   I.,	
   Jacquemond,	
   V.,	
   Kovács,	
   L.	
   and	
   Csernoch,	
   L.	
   Effects	
   of	
   dantrolene	
   on	
   steps	
   of	
   excitation-­‐contraction	
   coupling	
   in	
   mammalian	
  skeletal	
  muscle	
  fibers.	
  J	
  Gen	
  Physiol.	
  2001.	
  118:	
  355-­‐75	
   Takeshima,	
   H.,	
   Nishimura,	
   S.,	
   Matsumoto,	
   T.,	
   Ishida,	
   H.,	
   Kangawa,	
   K.,	
   Minamino,	
   N.,	
   Matsuo,	
   H.,	
   Ueda,	
   M.,	
   Hanaoka,	
   M.	
   and	
   Hirose,	
   T.	
   Primary	
   structure	
   and	
   expression	
   from	
  complementary	
  DNA	
  of	
  skeletal	
  muscle	
  ryanodine	
  receptor.	
  Nature,	
  1989,	
  339:	
   439–445	
   Tammaro,	
  A.,	
  Bracco,	
  A.,	
  Cozzolino,	
  S.,	
  Esposito,	
  M.,	
  Di	
  Martino,	
  A.,	
  Savoia,	
  G.,	
  Zeuli,	
  L.,	
   Piluso,	
  G.,	
  Aurino,	
  S.	
  and	
  Nigro,	
  V.	
  Scanning	
  for	
  mutations	
  of	
  the	
  ryanodine	
  receptor	
   (RYR1)	
   gene	
   by	
   denaturing	
   HPLC:	
   detection	
   of	
   three	
   novel	
   malignant	
   hyperthermia	
   alleles.	
  Clin	
  Chem.	
  2003.	
  49:	
  761-­‐8	
   Tanabe,	
  T.,	
  Beam,	
  K.G.,	
  Adams,	
  B.A.,	
  Niidome,	
  T.	
  and	
  Numa,	
  S.	
  Regions	
  of	
  the	
  skeletal	
   muscle	
  dihydropyridine	
  receptor	
  critical	
  for	
  excitation-­‐contraction	
  coupling.	
  Nature.	
   1990.	
  346:	
  567–569	
    	
    113	
    Tanabe,	
   T.,	
   Beam,	
   K.G.,	
   Powell,	
   J.A.	
   and	
   Numa,	
   S.	
   Restoration	
   of	
   excitation-­‐contraction	
   coupling	
   and	
   slow	
   calcium	
   current	
   in	
   dysgenic	
   muscle	
   by	
   dihydropyridine	
   receptor	
   complementary	
  DNA.	
  Nature.	
  1988.	
  336:	
  134–139	
   Tester,	
   D.J.,	
   Arya,	
   P.,	
   Will,	
   M.,	
   Haglund,	
   C.M.,	
   Farley,	
   A.L.,	
   Makielski,	
   J.C.	
   and	
   Ackerman,	
   M.J.	
   Genotypic	
   heterogeneity	
   and	
   phenotypic	
   mimicry	
   among	
   unrelated	
   patients	
   referred	
   for	
   catecholaminergic	
   polymorphic	
   ventricular	
   tachycardia	
   genetic	
   testing.	
   Heart	
  Rhythm.	
  2006.	
  3:	
  800-­‐5	
   Tester,	
   D.J.,	
   Kopplin,	
   L.J.,	
   Will,	
   M.L.	
   and	
   Ackerman,	
   M.J.	
   Spectrum	
   and	
   prevalence	
   of	
   cardiac	
   ryanodine	
   receptor	
   (RyR2)	
   mutations	
   in	
   a	
   cohort	
   of	
   unrelated	
   patients	
   referred	
   explicitly	
   for	
   long	
   QT	
   syndrome	
   genetic	
   testing.	
   Heart	
   Rhythm.	
   2005.	
   2:	
   1099-­‐105	
   Tester,	
   D.J.,	
   Spoon,	
   D.B.,	
   Valdivia,	
   H.H.,	
   Makielski,	
   J.C.	
   and	
   Ackerman,	
   M.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.	
  2004.	
  79:	
  1380-­‐4	
   Timerman,	
  A.P.,	
  Onoue,	
  H.,	
  Xin,	
  H-­‐B.,	
  Barg,	
  S.,	
  Copello,	
  J.,	
  Wiederrecht,	
  G.	
  and	
  Fleischer,	
   S.	
   Selective	
   Binding	
   of	
   FKBP12.6	
   by	
   the	
   Cardiac	
   Ryanodine	
   Receptor.	
   J	
   Biol	
   Chem.	
   1996.	
  271:	
  20385–20391	
   Tong,	
   J.,	
   McCarthy,	
   T.V.	
   and	
   MacLennan,	
   D.H.	
   Measurement	
   of	
   resting	
   cytosolic	
   Ca2+	
   concentrations	
   and	
   Ca2+	
   store	
   size	
   in	
   HEK-­‐293	
   cells	
   transfected	
   with	
   malignant	
   hyperthermia	
  or	
  central	
  core	
  disease	
  mutant	
  Ca2+	
  release	
  channels.	
  J	
  Biol	
  Chem.	
  1999.	
   274:	
  693–702	
   Tripathy,	
  A.,	
  Resch,	
  W.,	
  Xu,	
  L.E.,	
  Valdivia,	
  H.H.	
  and	
  Meissner,	
  G.	
  	
  Imperatoxin	
  A	
  induces	
   subconductance	
  states	
  in	
  Ca2+	
  release	
  channels	
  (ryanodine	
  receptors)	
  of	
  cardiac	
  and	
   skeletal	
  muscle.	
  J	
  Gen	
  Physiol.	
  1998.	
  111:	
  679–90	
  	
    	
    114	
    Tripathy,	
  A.,	
  Xu,	
  L.,	
  Mann,	
  G.	
  and	
  Meissner,	
  G.	
  Calmodulin	
  activation	
  and	
  inhibition	
  of	
   skeletal	
   muscle	
   Ca2+	
   release	
   channel	
   (ryanodine	
   receptor).	
   Biophy	
   J.	
   1995.	
   69:	
   106– 119	
   Tung,	
   C-­‐C.,	
   Lobo,	
   P.A.,	
   Kimlicka,	
   L.,	
   Van	
   Petegem,	
   F.	
   The	
   amino-­‐terminal	
   disease	
   hotspot	
   of	
   ryanodine	
   receptors	
   forms	
   a	
   cytoplasmic	
   vestibule.	
   Nature,	
   2010,	
   468:	
   585–588	
   Van	
   Petegem,	
   F.	
   Ryanodine	
   receptors:	
   structure	
   and	
   function.	
   J	
   Biol	
   Chem.	
   2012.	
   287:	
   31624-­‐32.	
   Van	
  Petegem,	
  F.,	
  Clark,	
  K.A.,	
  Chatelain,	
  F.C.	
  and	
  Minor,	
  D.L.	
  Jr.	
  Structure	
  of	
  a	
  complex	
   between	
  a	
  voltage-­‐gated	
  calcium	
  channel	
  beta-­‐subunit	
  and	
  an	
  alpha-­‐subunit	
  domain.	
   Nature.	
  2004.	
  429:	
  671–5	
   Van	
  Winkle,	
  W.B.	
  Calcium	
  release	
  from	
  skeletal	
  muscle	
  sarcoplasmic	
  reticulum:	
  site	
  of	
   action	
  of	
  dantrolene	
  sodium?	
  Science.	
  1976.	
  193:	
  1130–1131	
   Wagenknecht,	
   T.,	
   Berkowitz,	
   J.,	
   Grassucci,	
   R.,	
   Timerman,	
   A.P.	
   and	
   Fleischer,	
   S.	
   Localization	
   of	
   calmodulin	
   binding	
   sites	
   on	
   the	
   ryanodine	
   receptor	
   from	
   skeletal	
   muscle	
  by	
  electron	
  microscopy.	
  Biophys	
  J.	
  1994.	
  67:	
  2286–2295	
   Wagenknecht,	
  T.,	
  Radermacher,	
  M.,	
  Grassucci,	
  R.,	
  Berkowitz,	
  J.,	
  Xin,	
  H-­‐B.	
  and	
  Fleischer,	
   S.	
   Locations	
   of	
   Calmodulin	
   and	
   FK506-­‐binding	
   Protein	
   on	
   the	
   Three-­‐dimensional	
   Architecture	
   of	
   the	
   Skeletal	
   Muscle	
   Ryanodine	
   Receptor.	
   J	
   Biol	
   Chem.	
   1997.	
   272:	
   32463–32471	
   Wang,	
  J.	
  and	
  Best,	
  PM.	
  Inactivation	
  of	
  the	
  sarcoplasmic	
  reticulum	
  calcium	
  channel	
  by	
   protein	
  kinase.	
  Nature.	
  1992.	
  359:	
  739–741	
   Wang,	
   R.,	
   Chen,	
   W.,	
   Cai,	
   S.,	
   Zhang,	
   J.,	
   Bolstad,	
   J.,	
   Wagenknecht,	
   T.,	
   Liu,	
   Z.	
   and	
   Chen,	
   S.R.W.	
   Localization	
   of	
   an	
   NH2-­‐terminal	
   disease-­‐causing	
   mutation	
   hot	
   spot	
   to	
   the	
   ‘clamp’	
  region	
  in	
  the	
  three-­‐dimensional	
  structure	
  of	
  the	
  cardiac	
  ryanodine	
  receptor.	
  J.	
   Biol.	
  Chem.	
  2007	
  282:	
  17785–17793	
  	
    	
    115	
    Wang,	
  R.,	
  Zhong,	
  X.,	
  Meng,	
  X.,	
  Koop,	
  A.,	
  Tian,	
  X.,	
  Jones,	
  P.P.,	
  Fruen,	
  B.R.,	
  Wagenknecht,	
   T.,	
  Liu,	
  Z.	
  and	
  Chen,	
  S.R.W.	
  Localization	
  of	
  the	
  dantrolene	
  binding	
  sequence	
  near	
  the	
   FK506-­‐binding	
   protein-­‐binding	
   site	
   in	
   the	
   three-­‐dimensional	
   structure	
   of	
   the	
   ryanodine	
  receptor.	
  J	
  Biol	
  Chem,	
  2011,	
  286:	
  12202–12212	
   Wehrens,	
   X.H.,	
   Lehnart,	
   S.E.,	
   Huang,	
   F.,	
   Vest,	
   J.A.,	
   Reiken,	
   S.R.,	
   Mohler,	
   P.J.,	
   Sun,	
   J.,	
   Guatimosim,	
  S.,	
  Song,	
  L.S.,	
  Rosemblit,	
  N.,	
  D’Armiento,	
  J.M.,	
  Napolitano,	
  C.,	
  Memmi,	
  M.,	
   Priori,	
  S.G.,	
  Lederer,	
  W.J.	
  and	
  Marks,	
  A.R.	
  FKBP12.6	
  deficiency	
  and	
  defective	
  calcium	
   release	
   channel	
   (ryanodine	
   receptor)	
   function	
   linked	
   to	
   exercise-­‐induced	
   sudden	
   cardiac	
  death.	
  Cell.	
  2003.	
  113:	
  829–40	
  	
  	
   Wilson,	
   M.A.	
   and	
   Brunger,	
   A.T.	
   The	
   1.0	
   A	
   crystal	
   structure	
   of	
   Ca(2+)-­‐bound	
   calmodulin:	
   an	
   analysis	
   of	
   disorder	
   and	
   implications	
   for	
   functionally	
   relevant	
   plasticity.	
  J	
  Mol	
  Biol.	
  2000	
  301:	
  1237-­‐56	
   Wojcikiewicz,	
   R.J.	
   Type	
   I,	
   II,	
   and	
   III	
   inositol	
   1,4,5-­‐trisphosphate	
   receptors	
   are	
   unequally	
   susceptible	
   to	
   down-­‐regulation	
   and	
   are	
   expressed	
   in	
   markedly	
   different	
   proportions	
  in	
  different	
  cell	
  types.	
  J	
  Biol	
  Chem.	
  1995	
  270:	
  11678–83	
  	
   Wriggers,	
  W.	
  and	
  Chacón,	
  P.	
  Modeling	
  tricks	
  and	
  fitting	
  techniques	
  for	
  multiresolution	
   structures.	
  Structure.	
  2001,	
  9:	
  779–788	
   Wright,	
  N.T.,	
  Prosser,	
  B.L.,	
  Varney,	
  K.M.,	
  Zimmer,	
  D.B.,	
  Schneider,	
  M.F.	
  and	
  Weber,	
  D.J.	
   S100A1	
   and	
   calmodulin	
   compete	
   for	
   the	
   same	
   binding	
   site	
   on	
   ryanodine	
   receptor.	
   J	
   Biol	
  Chem,	
  2008,	
  283:	
  26676–26683	
   Wu,	
  S.,	
  Ibarra,	
  M.C.,	
  Malicdan,	
  M.C.,	
  Murayama,	
  K.,	
  Ichihara,	
  Y.,	
  Kikuchi,	
  H.,	
  Nonaka,	
  I.,	
   Noguchi,	
   S.,	
   Hayashi,	
   Y.K.	
   and	
   Nishino,	
   I.	
   Central	
   core	
   disease	
   is	
   due	
   to	
   RYR1	
   mutations	
  in	
  more	
  than	
  90%	
  of	
  patients.	
  Brain.	
  2006.	
  129:	
  1470-­‐80	
   Xiong,	
   L.,	
   Zhang,	
   J-­‐Z.,	
   He,	
   R.,	
   and	
   Hamilton,	
   S.L.	
   A	
   Ca2+-­‐Binding	
   Domain	
   in	
   RyR1	
   that	
   Interacts	
   with	
   the	
   Calmodulin	
   Binding	
   Site	
   and	
   Modulates	
   Channel	
   Activity.	
   Biophys	
   J.	
   90:	
  173-­‐182	
    	
    116	
    Yang,	
   H.C.,	
   Reedy,	
   M.M.,	
   Burke,	
   C.L.	
   and	
   Strasburg,	
   G.M.	
   Calmodulin	
   interaction	
   with	
   the	
   skeletal	
   muscle	
   sarcoplasmic	
   reticulum	
   calcium	
   channel	
   protein.	
   Biochemistry.	
   1994.	
  33:	
  518–25	
   Yang,	
  J.,	
  Gan,	
  Z.,	
  Lou,	
  Z.,	
  Tao,	
  N.,	
  Mi,	
  Q.,	
  Liang,	
  L.,	
  Sun,	
  Y.,	
  Guo,	
  Y.,	
  Huang,	
  X.,	
  Zou,	
  C.,	
  Rao,	
   Z.,	
   Meng,	
   Z.	
   and	
   Zhang,	
   K.Q.	
   Crystal	
   structure	
   and	
   mutagenesis	
   analysis	
   of	
   chitinase	
   CrChi1	
   from	
   the	
   nematophagous	
   fungus	
   Clonostachys	
   rosea	
   in	
   complex	
   with	
   the	
   inhibitor	
  caffeine.	
  Microbiology.	
  2010.	
  156:	
  3566-­‐74	
   Yeh,	
   H.M.,	
   Tsai,	
   M.C.,	
   Su,	
   Y.N.,	
   Shen,	
   R.C.,	
   Hwang,	
   J.J.,	
   Sun,	
   W.Z.	
   and	
   Lai,	
   L.P.	
   Denaturing	
   high	
   performance	
   liquid	
   chromatography	
   screening	
   of	
   ryanodine	
   receptor	
   type	
   1	
   gene	
   in	
   patients	
   with	
   malignant	
   hyperthermia	
   in	
   Taiwan	
   and	
   identification	
   of	
   a	
   novel	
   mutation	
  (Y522C).	
  Anesth	
  Analg.	
  2005.	
  101:	
  1401-­‐6	
   Yoshikawa,	
  F.,	
  Morita,	
  M.,	
  Monkawa,	
  T.,	
  Michikawa,	
  T.,	
  Furuichi,	
  T.	
  and	
  Mikoshiba,	
  K.	
   Mutational	
   analysis	
   of	
   the	
   ligand	
   binding	
   site	
   of	
   the	
   inositol	
   1,4,5-­‐trisphosphate	
   receptor.	
  J.	
  Biol.	
  Chem.	
  1996.	
  271:	
  18277–84	
   Yuchi,	
   Z.,	
   Lau,	
   K.	
   and	
   Van	
   Petegem,	
   F.	
   Disease	
   mutations	
   in	
   the	
  ryanodine	
  receptor	
   central	
   region:	
   crystal	
   structures	
   of	
   a	
   phosphorylation	
   hot	
   spot	
   domain.	
   Structure.	
   2012,	
  20:	
  1201-­‐11	
   Zhang,	
  Y.,	
  Chen,	
  H.S.,	
  Khanna,	
  V.K.,	
  De	
  Leon,	
  S.,	
  Phillips,	
  M.S.,	
  Schappert,	
  K.,	
  Britt,	
  B.A.,	
   Browell,	
   A.K.	
   and	
   MacLennan,	
   D.H.	
   A	
   mutation	
   in	
   the	
   human	
   ryanodine	
   receptor	
   gene	
   associated	
  with	
  central	
  core	
  disease.	
  Nat	
  Genet.	
  1993.	
  5:	
  46–50	
  	
  	
   Zhao,	
   F.,	
   Li,	
   P.,	
   Chen,	
   S.	
   R.,	
   Louis,	
   C.	
   F.,	
   and	
   Fruen,	
   B.	
   R.	
   Dantrolene	
   Inhibition	
   of	
   Ryanodine	
   Receptor	
   Ca2+Release	
   Channels	
   –	
   Molecular	
   Mechanism	
   and	
   Isoform	
   Selectivity.	
  J.	
  Biol.	
  Chem.	
  2001.	
  276:	
  13810–16	
   Zhou,	
   Q.,	
   Wang,	
   Q.L.,	
   Meng,	
   X.,	
   Shu,	
   Y.,	
   Jiang,	
   T.,	
   Wagenknecht,	
   T.,	
   Yin,	
   C.C.,	
   Sui,	
   S.F.	
   and	
   Liu,	
   Z.	
   Structural	
   and	
   functional	
   characterization	
   of	
   ryanodine	
   receptor-­‐natrin	
   toxin	
   interaction.	
  Biophys	
  J.	
  2008.	
  95:	
  4289–99	
    	
    117	
    Zorzato,	
   F.,	
   Fujii,	
   J.,	
   Otsu,	
   K.,	
   Phillips,	
   M.,	
   Green,	
   N.M.,	
   Lai,	
   F.A.,	
   Meissner,	
   G.	
   and	
   MacLennan,	
  D.H.	
  Molecular	
  cloning	
  of	
  cDNA	
  encoding	
  human	
  and	
  rabbit	
  forms	
  of	
  the	
   Ca2+	
  release	
  channel	
  (ryanodine	
  receptor)	
  of	
  skeletal	
  muscle	
  sarcoplasmic	
  reticulum.	
   J	
  Biol	
  Chem.	
  1990.	
  265:	
  2244–56	
    	
    118	
    

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Country Views Downloads
Canada 14 7
Poland 10 0
United States 9 1
China 6 25
United Kingdom 3 1
Ukraine 1 0
Romania 1 0
Japan 1 0
City Views Downloads
Vancouver 10 4
Warsaw 10 0
Ashburn 6 0
Shenzhen 5 25
Unknown 3 1
Richmond 3 3
London 1 0
Boardman 1 0
New York 1 0
Houston 1 1
Iași 1 0
Leeds 1 0
Oxford 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0073428/manifest

Comment

Related Items