DELIVERY  OF  ZINC  TO  RED  BLOOD  CELLS  AND  THE  DOWNSTREAM     EFFECTS  IN  MULTIPLE  SCLEROSIS     By     Suzanne  Letourneau                         A  DISSERTATION     Submitted  to     Michigan  State  University   in  partial  fulfillment  of  the  requirements   for  the  degree  of     Chemistry  –  Doctor  of  Philosophy     2013   ABSTRACT     DELIVERY  OF  ZINC  TO  RED  BLOOD  CELLS  AND  THE  DOWNSTREAM  EFFECTS  IN  MULTIPLE   SCLEROSIS     By     Suzanne  Letourneau     The  research  presented  here  shows  that  zinc  is  delivered  to  the  red  blood  cell  (RBC)  and   investigates   the   effects   of   this   in   multiple   sclerosis.     A   review   of   MS,   as   well   as   prior   research   investigating   RBC-­‐derived   adenosine   triphosphate   (ATP)   and   the   stimulation   of   nitric  oxide  (NO)  are  presented  here.    This  dissertation  hypothesizes  that  the  increase  in   RBC-­‐derived  ATP  seen  in  MS  patients  may  be  the  result  of  increased  zinc  levels  and  leads   to  increased  levels  of  NO,  a  molecule  known  to  increase  the  permeability  of  the  blood   brain  barrier  (BBB),  which  is  a  precursor  to  lesion  formation.       C-­‐peptide,  a  biologically  active  byproduct  in  the  formation  of  insulin,  also  increases  the   2+ release   of   ATP   from   RBCs,   but   only   when   bound   to   Zn .     Evidence   is   presented   here   2+ showing  that  C-­‐peptide  can  deliver  Zn  to  RBCs.    When  bound  to  C-­‐peptide,  2.54  ±  0.23   2+ 2+ pmol  of  Zn  were  delivered  to  RBCs,  compared  to  0.09  ±  0.21  pmol  when  Zn  alone   was  introduced  RBCs.    The  significance  of  this  in  diabetes  mellitus  will  be  discussed.   2+ Because   of   these   findings,   C-­‐peptide   was   used   to   deliver   Zn   to   the   RBCs   of   MS   2+ patients.    Significantly  more  Zn  is  delivered  to  the  RBCs  of  MS  patients,  at  a  value  of   3.61   ±   0.22   pmol,   than   to   those   of   healthy   controls,   at   a   value   of   2.26   ±   0.24   pmol.     2+ Additionally,   the   basal   level   of   Zn   in   the   RBCs   of   MS   patients   and   those   of   healthy     controls  was  measured.    The  RBCs  of  MS  patients  were  found  to  have  41.8  ±  1.7  μg  of   2+ 2+ Zn /g  Hb  where  the  RBCs  of  healthy  controls  only  had  32.9  ±  2.2  μg  Zn /g  Hb.   To  further  the  research  into  the  increase  ATP  release  from  the  RBCs  of  MS  patients,  this   was   measured   in   a   flow   system   that   mimics   the   shear   stress   experienced   by   the   RBCs   in   vivo.     It   was   found   that   the   RBCs   of   MS   patients   release   significantly   more   ATP,   at   a   value   of   344.7   ±   46.8   nM,   than   those   of   healthy   controls,   at   a   value   of   132.1   ±   14.1   nM.     Glybenclamide,   an   inhibitor   of   ATP   release   from   the   RBC,   decreased   this   value   to   65.3   ±   11.6  nM  in  the  RBCs  of  MS  patients,  showing  that  this  is  indeed  the  result  of  increase   release   of   ATP,   as   opposed   to   cell   lysis.     The   glucose   uptake   into   these   cells   that   may   be   leading  to  the  increased  ATP  release  is  also  discussed.   Finally,   reports   have   shown   that   estrogens   have   a   protective   effect   in   MS.   Here,   the   effect   of   estradiol   and   estriol   on   the   RBC-­‐derived   ATP   was   measured.     Estradiol   and   estriol   reduced   the   ATP   release   from   healthy   RBCs   to   74   ±   4%   and   70   ±   11%   that   of   healthy  controls,  respectively.    Through  the  use  of  a  microfluidic  device,  the  effects  of   estradiol  on  RBC-­‐derived  ATP  and  the  subsequent  endothelial  cell  NO  production  were   investigated.   When   these   RBCs   were   incubated   estradiol,   the   NO   production   from   the   endothelial  cells  was  attenuated  to  a  value  that  was  only  59  ±  7%  of  RBCs  in  the  absence   of   estradiol.     The   ability   of   estrogens   to   decrease   the   ATP   release   from   RBCs   and   subsequently   the   NO   production   of   endothelial   cells   has   major   implications   in   the   treatment  of  MS.       ACKNOWLEDGMENTS     First  I  would  like  to  thank  Michigan  State  University  for  the  funding  that  made  this  work   possible.     The   chemistry   department   here   at   Michigan   State   has   provided   an   environment   that   fosters   cooperation   and   resource   sharing   that   creates   a   vast   knowledge  base  and  allows  for  increase  productivity.    In  this  area,  I  would  especially  like   to  thank  Kathy  Severin  for  all  of  her  assistance  in  the  training  on  and  maintenance  of  the   instruments  in  the  teaching  laboratories.    She  is  always  willing  to  spend  her  time  helping   others.     During   my   time   at   Michigan   State,   I   was   also   privileged   to   be   part   of   a   collaboration   with   Michigan   State   University’s   Neurology   and   Ophthalmology   Clinic.     I   would  like  to  thank  Dr.  Eric  Eggenberger,  Dr.  Jayne  Ward,  and  Doozie  Russell  for  their   help  in  obtaining  blood  samples  from  MS  patients,  to  who  I  am  also  grateful.       I’d  also  like  to  thank  my  advisor,  Dana  Spence,  for  his  role  in  helping  me  to  get  where  I   am  today.    His  unique  leadership  style  fosters  productive  and  publishable  results  while   leaving   time   for   students   to   have   a   life   outside   of   the   laboratory.     The   research   environment   he   has   created   promotes   critical   thinking   and   independent   investigation,   which  are  invaluable  skills,  no  matter  what  field  his  students  end  up  in.     A   lot   of   this   work   would   not   have   been   possible   without   the   teamwork   of   the   Spence   group  members,  past  and  present.    I’d  especially  like  to  thank  Wathsala  Medawala,  who   collected  the  C-­‐peptide  data  presented  in  this  thesis.    In  addition  to  working  together,  I   enjoyed  sitting  next  to  her  for  my  first  three  years  in  the  laboratory.    I  would  also  like  to     iv   thank   the   current   Spence   group   members:     Kari   Anderson,   Yimeng   Wang,   Yueli   Liu,   Jayda   Erkal,   Sarah   Lockwood,   Bethany   Gross,   Chengpeng   Chen,   Kristen   Entwistle,   and   Ruipeng  Mu.    Yueli  and  Jayda  were  extremely  instrumental  in  the  research  on  the  RBCs   of   MS   patients,   taking   time   away   from   their   own   research   to   help   with   mine,   and   for   that  I  am  truly  indebted  to  them.    Kari  and  I  entered  the  program  here  in  the  same  year,   and  she  was  not  only  an  excellent  labmate,  but  also  became  one  of  my  closest  friends.     I   would   like   to   than   my   parents,   AnneMarie   and   Peter,   and   my   sister,   Lisa,   for   their   support   all   of   these   years.     They   have   been   by   biggest   cheerleaders   throughout   this   process,   and   I   am   lucky   to   have   grown   up   in   such   a   wonderful   environment.     My   parents,  Lisa,  and  Jackie  and  Bella  Robbins  have  all  helped  to  keep  me  smiling.     Finally,   I   would   like   to   thank   my   soon-­‐to-­‐be   husband,   Jonathan,   for   all   the   love   and   support  he  has  shown  me  since  the  day  we  first  met.    His  patience  and  encouragement   throughout  the  dissertation  and  defense  process  kept  me  going.    He  is  my  sunshine.         v   TABLE  OF  CONTENTS     LIST  OF  TABLES   viii   LIST  OF  FIGURES   ix   Chapter  1  –  Introduction   1.1 Multiple  Sclerosis   1.2 Classification,  Diagnosis,  and  Treatment  of  MS   1.3 Animal  Model  of  MS   1.4 Adenosine  Triphosphate  and  Nitric  Oxide   1.5 Pregnancy  and  Estrogens  in  MS   1.6 Zinc  and  MS   REFERENCES     Chapter  2  –  The  Effect  on  Estrogen  on  Endothelial  Nitric  Oxide  Stimulation  via  Red       Blood  Cell  Derived  Adenosine  Triphosphate     2.1  Red  Blood  Cell  Derived  ATP  and  NO  Stimulation       2.1.1  Red  Blood  Cell  Glycolysis     2.1.2  ATP  Release  and  NO  Stimulation       2.1.3  Estrogens  and  MS     2.2  Microfluidic  Devices     2.3  Experimental       2.3.1  Collection  and  Purification  of  Rabbit  RBCs       2.3.2  Preparation  of  Reagents       2.3.3  Preparation  of  Microfluidic  Device       2.3.4  Chemiluminescence  Detection  of  ATP  Release  from  RBCs       2.3.5  Adhering  Cells  to  a  Microfluidic  Device  and  Fluorescence       Determination  of  NO     2.4  Results     2.5  Discussion   REFERENCES     1   1   3   11   12   15   20   22   Chapter  3  –  Delivery  of  Zn  to  the  Red  Blood  Cell  by  C-­‐Peptide       3.1  Diabetes  Mellitus       3.1.1  Classifications       3.1.2  Complications     3.2  Insulin  and  C-­‐Peptide       3.2.1  Discovery  of  Insulin       3.2.2  Insulin  and  C-­‐Peptide  Production  and  Release       3.2.3  Structure  of  C-­‐Peptide   72   72   72   75   77   77   78   81   2+   vi   30   30   30   33   37   38   41   41   42   43   44   48   52   63   66       3.2.4  Biological  Effects  of  C-­‐Peptide     3.3  Experimental       3.3.1  Preparation  of  Reagents       3.3.2  Collection  and  Preparation  of  Human  RBCs       3.3.3  Radiolabeled  Zn 2+ 2+   Assays  for  Determination  of  the  Amount       of  Zn  Interacting  with  the  RBC     3.4  Results     3.5  Discussion   REFERENCES     Chapter  4  –  Potential  Multiple  Sclerosis  Biomarkers     4.1  MS  Diagnosis  and  the  Need  for  Biomarkers     4.2  Experimental       4.2.1  Preparation  of  Reagents       4.2.2  Collection  and  Preparation  of  Patient  and  Control  Samples                           2+   4.2.3  Determination  of  Radiolabelled  Zn Interaction  with       MS  Patient  RBCs   4.2.4  Determination  of  Radiolabelled  Glucose  Uptake  of  MS     Patient  RBCs   4.2.5  Determination  of  ATP  Release  from  MS  Patient  RBCs  in     Response  to  Flow   4.2.6  Determination  of  C-­‐peptide  Interaction  with  MS  Patient     RBCs  and  the  Plasma  Concentration  of  C-­‐Peptide  by     Enzyme  Linked  Immunosorbent  Assay   2+     4.2.7  Determination  of  Basal  Zn  Levels  in  MS  Patient  RBCs       by  Atomic  Absorption  Spectroscopy     4.3  Results     4.4  Discussion   REFERENCES     Chapter  5  –  Overall  Conclusions  and  Future  Directions     5.1  Overall  Conclusions     5.2  Future  Directions   2+     5.2.1  Zn  and  C-­‐Peptide  Binding       5.2.2  Future  MS  Studies       5.2.3  Microfluidics  and  MS  Studies   REFERENCES         83   86   86   87   vii   87   88   97   105   113   113   117   117   118   119   119   121   123   123   126   131   139   143   143   153   153   155   156   158   LIST  OF  TABLES       Table  1.1  –  Medications  used  in  MS.  *IFNs  inhibit  the  proliferation  of  T  cells,  decreases   the  production  of  proinflammatory  cytokines   8           viii   LIST  OF  FIGURES       Figure  1.1  –  Common  Lesions  in  MS  –  The  diagnosis  of  MS  requires  lesions  to  be  found   in   two   of   the   four   following   areas   of   the   CNS:   juxtacortical,   periventricular,   infratentorial,  and  the  spinal  cord.    Above  are  MRI  images  of  lesions  in  these  regions  of   the  brain,  as  well  as  in  the  spinal  cord.    For  interpretation  of  the  references  to  color  in   this   and   all   other   figures,   the   reader   is   referred   to   the   electronic   version   of   this   dissertation.       5     Figure   1.2   –   Proposed   mechanism   of   ATP   release   from   RBCs   and   subsequent   NO   production   –   It   has   been   proposed   that   G-­‐protein   coupled   receptor   (GPCR),   cyclic   adenosine  monophosphate  (cAMP),  and  the  cystic  fibrosis  transmembrane  conductance     regulation   (CFTR)   protein are   all   required   for   the   release   of   ATP   by   mechanical   deformation,  though  how  the  ATP  leaves  the  cell  remains  in  question.    In  this  proposed   mechanism,  the  binding  of  ATP  to  the  P2Y  receptor  on  the  endothelial  cell  results  in  the   activation  NOS  and  the  production  of  NO.   14     Figure   1.3   –   Steroid   structures   –   Dehydroxyepiandosterone   (DHEA)   is   a   precursor   of   both   estradiol   and   estriol,   and   all   three   steroids   have   similar   structures   leading   to   the   hypothesis  that  the  estrogens  will  have  a  similar  effect  on  the  ATP  release  from  the  RBC   as  that  of  DHEA.   19     Figure   2.1   –   Glycolysis   –   This   process,   which   occurs   in   RBCs,   is   the   only   way   for   the   cells   to   metabolize   glucose.     Glycolysis   occurs   in   two   phases.     In   the   preparatory   phase,   on   the  left,  two  ATP  molecules  are  consumed.    During  the  payoff  phase,  on  the  right,  four   ATP  molecules  are  produced,  resulting  in  a  net  gain  of  two  ATP  molecules  per  glucose   molecule  consumed.         32     Figure   2.2   –   ATP   Release   From   MS   RBCs   (From   Letourneau   et   al.)   –   ATP   release   was   measured  from  RBCs  subjected  to  flow.    The  average  ATP  release  from  the  healthy  RBCs   was   138   ±   21   nM.     An   increase   to   375   ±   51   nM   was   seen   for   RBCs   obtained   from   MS   patients.    The  error  is  reported  as  standard  error  of  the  mean  for  11  controls  and  18  MS   samples.    The  values  are  statistically  different  at  a  value  of  p  <  0.001.   35     Figure   2.3   –   Proposed   mechanism   of   ATP   release   from   RBCs   and   subsequent   NO   production   –   It   has   been   proposed   that   G-­‐protein   coupled   receptor   (GPCR),   cyclic   adenosine  monophosphate  (cAMP),  and  the  cystic  fibrosis  transmembrane  conductance     regulation   (CFTR)   protein are   all   required   for   the   release   of   ATP   by   mechanical   deformation,  though  how  the  ATP  leaves  the  cell  remains  in  question.    In  this  proposed   mechanism,  the  binding  of  ATP  to  the  P2Y  receptor  on  the  endothelial  cell  results  in  the   activation  NOS  and  the  production  of  NO.         36     ix     Figure   2.4   –   Photolithography   –   In   order   to   perform   rapid   fabrication   using   PDMS,   a   reusable   master   is   needed.     This   can   be   prepared   through   photolithography.     In   this   process,   a   clean   silicon   wafer   is   coated   with   SU-­‐8   photoresist.     After   pre-­‐baking   the   wafer,   a   mask   is   placed   over   the   photoresist   with   the   desired   features.     The   wafer   is   then   exposed   to   UV   light.     After   a   post-­‐bake,   the   wafer   is   soaked   in   developer   to   removed   unpolymerized   photoresist.     The   result   is   a   wafer   with   raised   features   that   can   be  repeatedly  used  as  a  mold  for  PDMS.         40     Figure   2.5   –   Soft   Lithography   –   Through   this   process,   the   layers   of   the   microfluidic   device  were  created.    The  polymer  and  curing  agent  were  mixed  in  20:1  and  5:1  ratios  in   separate  plastic  cups  and  degassed  by  vacuum.    The  20:1  mixture  was  poured  over  the   channel  portion  of  the  master  and  baked  at  75°C  for  12  minutes.    The  5:1  mixture  was   then   pour   wafer   the   entire   master   and   bake   for   another   12   minutes   at   the   same   temperature.    The  PDMS  was  then  removed  from  the  master.    The  second  layer  of  the   microfluidic   device   was   prepared   using   the   same   steps   with   an   unpatterned   silicon   master.         45     Figure   2.6   –   Microfluidic   Device   –   A)   The   microfluidic   device   layers   are   diagramed.     A   polycarbonate  membrane  separated  the  bottom  layer,  containing  the  channels,  and  the   top  layer,  patterned  with  wells.    B)  A  photograph  of  a  completed  microfluidic  device.    C)   A  diagram  of  the  cross  section  of  a  well.  bPAECs  are  adhered  to  the  membrane  in  the   well.     The   porous   membrane   separates   the   bPAECs   from   the   channel   through   which   the   RBCs  are  pumped.         46     Figure  2.7  –  Measurement  of  ATP  Release  –  Estradiol  (E2)  or  estriol  (E3)  were  added  to   2+ PSS,   immediately   followed   by   RBCs   to   make   a   7%   RBC   solution.     When   Zn   bound   to   C-­‐ 2+ peptide   was   used,   the   Zn   and   C-­‐peptide   were   added   to   the   vial   first   and   allowed   to   incubate  for  2-­‐3  minutes  before  the  addition  of  PSS.    After  a  two  hour  incubation,  a  200   μL  sample  was  transferred  into  a  cuvette.    100  μL  of  the  luciferin/luciferase  solution  was   added  and  the  cuvette  was  lightly  shaken.    The  cuvette  was  placed  in  a  dark  box  over  a   PMT  and  at  15  seconds  the  luminescence  was  measured.         47     Figure   2.8   –   DAF-­‐FM   DA   –   DAF-­‐FM   DA   crosses   the   cell   membrane.   Esterases   then   transform   the   probe   to   DAF-­‐FM,   which   loses   its   cell   permeability.     The   DAF-­‐FM   then   reacts   with   NO,   producing   a   benzotiazol   derivative   that   is   fluorescent   at   the   above   excitation  and  emission.   50     Figure   2.9   –   Flowing   RBCs   Through   the   Device   –   Pictured   above,   the   RBC   solutions   were   pumped  at  the  rate  of  0.1  µL/min  through  the  underlying  microfluidic  channels  for  30   minutes.    Each  well  was  imaged  before  and  after  using  an  Olympus  MVX  fluorescence   macroscope,  fitted  with  a  mercury-­‐arc  lamp  and  a  fluoroscien  isothiocyanate  (FITC)  filter     x   cube   having   excitation   and   emission   wavelengths   of   470   and   525   nm,   respectively.         51     Figure  2.10  –  The  Effect  of  Estradiol  on  RBC  ATP  Release  -­‐  Compared  to  RBCs  with  the   absence   of   estradiol,   the   RBC-­‐derived   ATP   for   RBCs   treated   with   0.5   μM   estradiol   was   reduced   to   76   ±   7%   of   the   value   of   the   untreated   cells.     The   amount   of   ATP   released   decreased  with  increasing  concentration  of  estradiol;  for  the  1  µM  and  1.5  µM  estradiol   solutions,  the  RBC-­‐derived  ATP  was  reduced  to  62  ±  7%  and  56  ±  6%,  respectively.    Error   is   shown   as   standard   deviation   for   N   =   4   rabbits.   The   asterisks   denote   a   statistically   significant  difference  from  untreated  RBCs  at  p  <  0.05.   53     Figure   2.11   –   Estradiol   (E2)   Decreases   ATP   Release   at   Physiological   Concentrations   –   incubating  the  RBCs  with  30  nM  estradiol  reduces  the  chemiluminescence  in  response   to  ATP  release  to  74  ±  4%,  which  is  statistically  equivalent  to  the  decrease  in  the  amount   of  chemiluminescence  seen  with  0.5  µM  estradiol,  but  at  a  more  physiologically  relevant   concentration   of   the   hormone.     Error   is   standard   deviation   for   N   =   4   rabbits.     The   asterisk   denotes   the   decrease,   as   compared   to   untreated   RBCs,   is   statistically   significant   at  p  <  0.02.   54     Figure   2.12   –   The   Effect   of   Estriol   (E3)   on   RBC-­‐derived   ATP   –   Estriol   decreases   the   chemiluminescence   due   to   ATP   at   physiologically   relevant   concentrations.     30   nM   of   estriol  reduced  the  ATP  release  to  70  ±  11%  of  that  of  RBCs  alone.    The  RBCs  incubated   with   0.5   µM   and   1.0   µM   had   their   ATP   release   reduced   to   69   ±   13%   and   62   ±   11%,   respectively.     Error   is   shown   as   standard   deviation   for   N   =   4   rabbits,   and   the   asterisk   denotes  a  statistically  significant  decrease  at  p  <  0.05   56     2+ Figure   2.13   –   The   Effect   of   Estrogen   on   ATP   Release   Stimulated   By   Zn   bound   to   C-­‐ peptide  –  Note  that  the  concentrations  of  estradiol  and  estriol  are  not  same  looking  at   the   bars   left   to   right   in   an   attempt   to   compare   the   data   sets.     In   both   sets   of   results   the   2+ Zn   bound   to   C-­‐peptide   resulted   in   an   increase   in   chemiluminescence   to   more   than   double  that  seen  for  RBCs  alone.      For  estradiol  (E2),  a  significant  decrease  in  the  ATP   2+ release,  as  compared  to  RBCs  and  10  nM  Zn -­‐C-­‐peptide,  to  76  ±  15%  and  61  ±  24%  was   observed   for   the   0.75   µM   and   1.5   µM   samples,   respectively.     For   estriol   (E3),   the   statistically  significant  decreases  were  seen  for  the  0.5  µM  and  1.0  µM  samples.    They   were   decreased   to   73   ±   17%   and   61   ±   13%,   respectively.     Error   is   standard   deviation   for   N   =   5   rabbits   for   estradiol   and   N   =   4   rabbits   for   estriol.     The   asterisk   denotes   the   2+ significant  increase  of  ATP  release  with  10  nM  Zn  bound  to  C-­‐peptide  at  p  <  0.0005.   2+ The  #  denotes  a  decrease  as  compared  to  10  nM  Zn -­‐C-­‐peptide  at  p  <  0.05.   57     Figure   2.14   –   Fluorescence   Intensity   Images   –   DAF-­‐FM   reacts   with   NO   in   bPAECs   resulting   in   fluorescence   than   can   be   photographed   with   a   CCD   camera   and   the   pixel   intensity  can  be  measured.    As  seen  above,  when  RBCs  incubated  with  estradiol  (E2)  are     xi   flowed  beneath  wells  containing  bPAECs,  the  NO  production  of  those  cells  is  lower.    This   is  due  to  the  decrease  in  RBC-­‐derived  ATP,  stimulating  NO  production  in  bPAECs.         59     Figure   2.15   –   Normalize   Fluorescence   Intensity   –   The   data   in   this   figure   shows   the   normalized,   background   subtracted   emission.     As   shown,   the   emission   decreases   with   RBCs  incubated  in  buffer  containing  estradiol.    In  this  case,  the  estradiol  is  actually  able   to   reduce   the   NO   production   to   43   ±   0.1%   of   that   of   untreated   RBCs.     The   emission   intensity   begins   to   increase   with   again   with   increasing   concentration   of   estradiol,   although  a  significant  increase  in  emission,  in  comparison  to  RBCs  alone,  is  not  measure   until   the   estradiol   levels   reach   approximately   2   µM.     At   this   value,   the   fluorescence   intensity  is  13  ±  4%  higher  than  the  untreated  RBCs.    The  next  highest  concentration,  1.5   µM,  is  not  significantly  different  from  the  untreated  RBCs.    Error  is  shown  as  standard   deviation   for   N   =   4   rabbits   and   the   asterisk   denotes   a   significant   difference   from   untreated  RBCs  at  p  <  0.05.   60     Figure   2.16   –   Washed   and   Unwashed   RBCs   –   There   was   a   marked   decrease   in   NO   production  in  the  bPAECs,  denoted  by  the  decrease  in  fluorescent  emission,  when  the   RBCs   are   pre-­‐incubated   with   estradiol.     RBCs   that   have   been   washed   to   removed   excess   estradiol  do  not  result  in  a  statistically  significant  increase  in  NO  production.    The  black   bars  represent  the  data  from  the  previous  figure  and  are  included  for  comparison.    The   grey  bars  show  that  the  estradiol  is  able  to  decrease  the  NO  production  in  bPAECs,  as   detected  by  fluorescence  intensity.    The  NO  production  was  decreased  to  59  ±  7%  with   10   nM   estradiol,   and   this   decrease   remained   statistically   constant   across   the   range   of   concentrations,   down   to   51   ±   5%   that   of   the   NO   production   of   bPAECs   exposed   to   untreated   RBCs,   seen   at   2.0   µM.     Error   is   reported   as   standard   deviation   for   N   =   4   rabbits.   The   asterisks   denote   a   significant   difference   from   untreated   RBCs   (black   bar)   and  the  #  denote  a  significant  decrease  from  untreated  RBCs  (grey  bar).   62     Figure   3.1   –   Proinsulin   –   This   basic   diagram   shows   the   role   of   C-­‐peptide   in   the   structure   of   insulin,   as   well   as   the   amino   acid   sequence   of   C-­‐peptide.     The   acidic   residues   have   been  colored  purple.    When  the  C-­‐peptide  is  cleaved  from  proinsulin,  the  A  and  B  chains   remain,  forming  insulin.         80     Figure  3.2  –  Insulin  Packaging  and  Release  –    Insulin  is  first  synthesize  as  preproinsulin  in   the  cytosol  before  being  uptaken  into  the  rough  endoplasmic  reticulum  where  the  signal   peptide   is   cleaved   to   form   proinsulin.     It   is   then   transported   to   the   Golgi   apparatus   where   it   is   subsequently   packaged   into   vesicles.   C-­‐peptide   is   cleaved   from   proinsulin   2+ during  this  process.  At  the  same  time,  Zn  ions  are  entering  the  vesicle  through  ZnT8,  a   2+ 2+ Zn   transporter   in   the   vesicle   membrane.     This   Zn   is   involved   in   the   creation   of   2+ proinsulin   hexamers,   consisting   of   six   proinsulin   and   two   Zn .     After   C-­‐peptide   is   2+ removed,  the  Zn -­‐insulin  complex  becomes  less  soluble  and  is  stored  in  the  vesicles  in   a  crystalline  form.  Insulin  is  released  from  the  β-­‐cells  in  response  to  blood  glucose  levels,     xii   and   the   vesicles   undergo   exocytosis.     When   this   occurs,   insulin   and   C-­‐peptide   are   2+ released  in  equimolar  amounts  and  Zn  is  present  as  well.           2+ 65 82   2+ Figure   3.3   –   Radiolabeled   Zn   Assays   –   Zn   and   the   peptide   were   incubated   together   in   pure   water   for   three   minutes   to   ensure   peptide   activation.     PSS   was   then   added  to  the  mixture,  immediately  followed  by  the  addition  of  the  appropriate  volume   of   RBCs   to   prepare   1   mL   samples   with   a   7%   hematocrit.     Following   a   one   hour   incubation  samples  were  centrifuged  and  the  supernatant  was  collected.    In  some  of  the   experiments,  the  RBCs  were  then  washed  three  times  and  lysed  with  bleach.    In  a  96-­‐ well   plate,   200   μL   of   supernatant   or   lysate   solution   was   combined   with   100   μL   of   scintillation  cocktail.    The  amount  of  radioactivity  in  each  sample  was  then  determined   using  a  1450  Microbeta  Plus  liquid  scintillation  counter.         89     65 2+ Figure   3.4   –   Zn   Remaining   in   the   Supernatent   after   Incubation   with   RBCs   –   For   each   65 2+ set   of   bars,   when   C-­‐peptide   was   also   added,   it   was   added   in   a   1:1   ratio   with   Zn .     As   is   evident   from   the   data,   65 2+ Zn   alone,   black   bars,   does   not   interact   with   the   RBC.     However,   with   the   exception   of   the   1   nM   concentration,   when   65 65 2+ Zn   bound   to   C-­‐ 2+ peptide   is   incubated   with   the   RBCs,   grey   bars,   significantly   less   Zn   is   found   in   the   supernatant.    Error  is  represented  as  standard  deviation  for  N  =  7  humans.  The  asterisk   represents  p  <  0.05.   90     2+ Figure   3.5   –   C-­‐Peptide   and   Zn   on   the   RBC   –   The   data   on   the   left   shows   the   amount   of   C-­‐peptide   interaction   with   the   RBC.   Comparing   to   the   data   on   the   right,   showing   the   65 2+ 2+ amount   of   Zn   recovered   from   RBC   lysate,   it   is   evident   that   C-­‐peptide   and   Zn   interaction   with   the   RBC   in   a   1:1   ratio.     For   both   binding   curves,   saturation   seems   to   2+ begin  around  10  picomoles  of  added  Zn  bound  to  C-­‐peptide  and  results  in  a  maximum   65 2+ of  2.64  ±  1.03  picomoles  of  C-­‐peptide  and  2.96  ±  0.17  picomoles  of   Zn  interacting   2+ with  the  RBC  when  20  picomoles  of  Zn -­‐activated  C-­‐pepide  have  been  added.    Within   2+ error,  this  shows  that  Zn  and  C-­‐peptide  interact  with  the  RBC  in  a  1:1  manner.    Error  is   2+ shown  as  standard  deviation  for  N  =  4  humans  for  C-­‐peptide  and  N  =  5  for    Zn .   92     Figure   3.6   –   Saturation   Curve   –   After   total   and   specific   binding   (circles   and   triangle,   respectively)   were   found,   the   non-­‐specific   binding   (squares)   was   calculated.     2+ Comparable   to   the   C-­‐peptide   binding,   the   amount   of   specifically   bound   Zn   is   1.85   ±   65 2+ 0.53  picomoles  for  20  picomoles  of  added   Zn  bound  to  C-­‐peptide.    This  data  shows   2+ 2+ that  Zn ,  or  Zn  bound  to  C-­‐peptide,  is  interacting  with  a  receptor  on  the  RBC.    The   error  is  shown  as  standard  deviation  for  N  =  7  humans  for  2,  5,  10,  20  and  40  picomoles,   N  =  4  humans  for  80  picomoles,  and  N  =  5  humans  for  100  picomoles.   94     xiii     2+ Figure  3.7  –  C-­‐Peptide  and  Zn  in  the  Supernatent  –    The  data  on  the  left  shows  that   2+ whether   or   not   20   nM   C-­‐peptide   is   bound   to   20   nM   Zn ,   statistically   the   same   amount   41 2+ is   found   to   be   interacting   with   the   RBC.     However,   when   Zn   alone   is   added   to   an   2+ RBC   solution,   all   of   it   is   found   in   the   supernatant   after   incubation.     When   Zn   is   2+ 2+ introduced   to   the   sample   as   20   nM   Zn   bound   to   C-­‐peptide,   significantly   less   Zn   was   found  in  the  supernatant,  meaning  it  is  interacting  with  the  RBC,  but  only  when  added   to  the  sample  bound  to  C-­‐peptide.    Error  is  standard  deviation  for  N  =  4  humans  for  C-­‐ 2+ 2+ peptide   and   N   =   5   humans   for   Zn .     The   asterisk   denotes   p   <   0.01   as   compared   to   Zn   added  alone.   96     2+ Figure   3.8   –   Effect   of   Mutant,   Ratios,   and   BSA   on   Zn   Interaction   with   RBCs   –   The   mutation   in   E27A   prevents   65 2+ Zn   interaction   with   the   RBC   in   the   presences   of   BSA   (black   bars).     Also,   in   normal   PSS,   only   the   1:1   65 2+ 65 2+ Zn   to   C-­‐peptide   ratio   was   able   to   delivery   any   Zn   to   the   RBC.     When   the   experiments   were   repeated   using   a   BSA-­‐free   2+ PSS,   (grey   bars)   both   peptides   are   able   to   deliver   Zn   to   the   RBC,   regardless   of   ratio.     2+ However,  C-­‐peptide  is  able  to  deliver  roughly  double  the  amount  of  Zn  as  compared   to   the   mutant.     The   error   is   standard   deviation   for   N   =   4   humans.   The   asterisk   represents  p  <  0.05  and  the  pound  sign  represents  p  <  0.001  as  compared  to  1:1  ratio  of   2+ Zn  bound  to  C-­‐peptide.     98   2+ Figure  3.9  –  Hypothesis  of  Zn  Binding  to  C-­‐peptide  and  a  Mutant  –    Results  showing   2+ 2+ the   loss   of   Zn   interaction   with   the   RBC   with   higher   Zn   to   C-­‐peptide   ratios   led   to   the   2+ hypothesis   that   multiple   Zn   ions   may   be   binding   to   the   C-­‐peptide   in   those   cases.     2+ Then,  upon  addition  of  serum  albumin  containing  PSS,  the  Zn  may  be  removed  from   the   peptide   by   the   albumin   because   of   its   high   binding   affinity.   It   is   believed   that   the   2+ 2+ Zn  may  bind  to  the  E27A  peptide,  however  there  is  no  folding,  and  the  Zn  ions  are   easily  removed  by  the  BSA.    This  hypothesis  is  related  to  the  data  in  figure  3.8.   102     Figure   4.1  –   Determination   of  ATP  Release   from   MS  Patient   RBCs  in   Response   to   Flow   –   RBCs   were   diluted   to   a   7%   hematocrit   and   pumped   through   tubing,   one   syringe   containing   the   sample   and   the   other   containing   luciferin/luciferase.   Solutions   were   pumped,   at   6.7   µL/min,   through   50   µm   internal   diameter   microbore   tubing.     The   solutions  met  at  a  mixing  T-­‐junction  and  the  combined  stream  was  pumped  through  an   additional  segment  of  75  µm  internal  diameter  tubing,  with  a  portion  of  the  polyimide   coating   removed,   over   a   PMT   in   a   light   excluding   box.     This   allowed   for   the   chemiluminescence   resulting   from   the   reaction   of   ATP   with   luciferin/luciferase   to   be   detected.         122     xiv     Figure   4.2   –   Determining   C-­‐peptide   Interaction   with   the   RBCs   of   MS   Patients   using   2+ Enzyme-­‐Linked   Immunosorbent   Assay   –   The   interaction   between   Zn   bound   to   C-­‐ peptide  and  RBCs  was  studying  using  solutions  containing  7%  RBCs    and  20  nM  each  of   2+ 2+ Zn  and  C-­‐peptide.    Samples  were  prepared  by  mixing  the  appropriate  volumes  of  Zn   and  C-­‐peptide  in  water  for  several  minutes  before  adding  PSS,  immediately  followed   by   the   RBC   of   either   MS   patients   or   healthy   controls.     Following   a   three   hour   incubation   at   37°C,   the   samples   were   centrifuged   at   500g   for   four   minutes   and   the   C-­‐peptide   remaining  in  the  supernatant  was  measured  using  an  ELISA.         124     Figure  4.3  –  ATP  Release  from  RBCs  of  MS  Patients  –  The  results  from  the  ATP  release   studies   show   that   the   ATP   release   from   the   RBCs   of   MS   patients   was   found   to   be   an   average   of   344.7   ±   46.8   nM   where   the   average   release   from   healthy   controls   was   132.1   ±   14.1   nM.     When   the   RBCs   of   the   MS   patients   were   incubated   with   a   CFTR   inhibitor,   glybenclamide,   the   ATP   release   is   decreased   back   down   below   the   amount   of   the   healthy  controls,  to  a  level  of  65.3  ±  11.6  nM,  showing  that  the  increase  in  ATP  release   of  the  flowing  RBCs  of  MS  patients  is  not  the  result  of  RBC  lysis.    The  error  is  reported  as   standard   error   of   the   mean,   for   N   =   19   MS   patients,   10   healthy   controls   and   12   glybenclamide  inhibitions.    The  asterisk  represents  p  <  0.001.   127     2+ 2+ Figure   4.4   –   Basal   Levels   of   Zn   in   the   RBCs   of   MS   Patients   –   The   basal   amount   of   Zn   2+ in  the  RBCs  of  the  MS  patients  was  found  to  be  41.8  ±  1.7  µg  of  Zn /g  Hb,  which  is  a   2+ 27%  increase  of  the  basal  levels  of  Zn  of  the  RBCs  of  healthy  controls,  32.9  ±  2.2  µg  of   2+ Zn /g  Hb.    The  error  is  reported  as  standard  error  of  the  mean  for  N  =  21  MS  patients   and   11   healthy   controls.   The   asterisk   represents   p   <   0.01   as   compared   to   the   control   sample.   128     65 2+ 65 2+ Figure   4.5   –   Zn   interaction   with   the   RBCs   of   MS   Patients   –   The   amount   of   Zn   that  is  able  to  interact  with  the  RBCs  of  MS  patients  is  significantly  higher,  at  a  value  of   3.61  ±  0.22  picomoles,  than  that  of  healthy  controls,  at  a  value  of  2.26  ±  0.24  picomoles.     The   amount   of   C-­‐peptide   interaction   with   the   RBC   correlates   to   this   very   well,   as   shown   in  figure  4.6.    The  error  is  reported  as  standard  error  of  the  mean  for  N  =  22  MS  patients   and  11  healthy  controls.      The  asterisk  represents  p  <  0.001  as  compared  to  the  control   samples.   129     Figure  4.6  –  C-­‐peptide  Interaction  with  the  RBCs  of  MS  Patients  –  Correlating  to  the  data   in  figure  4.5,  it  can  be  seen  that  the  amount  of  C-­‐peptide  interacting  with  the  RBCs  is   65 2+ very   similar   to   the   amount   of   Zn .     3.61   ±   0.18   picomoles   of   C-­‐peptide   interacted   with  the  RBCs  of  MS  patients,  while  only  2.43  ±  0.20  picomoles  interacted  with  the  RBCs   of  healthy  controls.    The  error  is  reported  as  standard  error  of  the  mean  for  N  =  12  MS     xv   patients  and  6  healthy  controls.    The  asterisk  represents  p  <  0.001  as  compared  to  the   control  sample.   130     Figure  4.7  –  Glucose  Uptake  into  the  RBCs  of  MS  Patients  –  It  should  be  noted  that  the  x-­‐ axis   does   not   start   at   zero,   so   while   there   are   differences,   they   are   not   as   extreme   as   they  seem.    Comparing  the  untreated  RBCs  (black)  of  the  MS  patients  and  those  of  the   healthy   control,   no   significant   difference   was   seen   in   the   amount   of   glucose   uptake.     2+ However,   when   the   RBCs   were   stimulated   by   Zn   bound   to   C-­‐peptide   (grey),   those   from   the   MS   patients   took   in   6.0   ±   0.01%   more   glucose   than   those   of   the   healthy   controls.    While  this  seems  like  a  small  increase,  it  is  a  statistically  significant  one.    The   error  is  reported  as  standard  error  of  the  mean  for  N  =  22  MS  patient  RBC  samples  and   11  healthy  control  RBC  samples.  The  asterisk  represents  p  <  0.001  and  the  pound  sign   represents  p  <  0.01.   132     65 2+ Figure  4.8  –  The  Difference  in   Zn  interaction  with  the  RBCs  of  Male  and  Female  MS   65 2+ Patients  –  There  is  a  significant  difference  in  the  amount  of   Zn  interaction  with  the   RBC   between   male   (black   bars)   and   female   (white   bars)   patients   with   MS   that   is   not   65 2+ seen   in   healthy   controls.     The   highest   amount   of   Zn   interacting   with   the   RBC   s   is   seen   in   male   MS   patients   at   a   level   of   4.16   ±   0.18   picomoles.     The   error   is   shown   as   standard   error   of   the   mean   for   N   =   6   male   MS   patients,   3   male   healthy   controls,   16   female  MS  patients,  and  8  female  healthy  controls.    The  *  represents  p  <  0.001.   138     Figure   5.1   –   The   Effect   of   Estriol   (E3)   on   ATP   Release   from   the   RBCs   of   MS   Patients   –     Flow  based  studies  were  performed  on  one  sample  of  RBCs  from  an  MS  patient  with  the   addition   of   30   and   500   nM   estriol.     Without   estriol,   the   RBCs   from   the   MS   patient   released  208.1  nM  ATP,  while  the  healthy  control  released  103.9  nM.    When  incubated   for  a  half  hour  with  30  or  500  nM  estriol,  the  ATP  release  dropped  to  59.1  and  21.6  nM,   respectively.  N  =  1.   145       xvi   Chapter  1  –  Introduction     1.1  Multiple  Sclerosis       Multiple  Sclerosis  (MS)  is  a  disease  of  the  central  nervous  system  (CNS)  that  affects  over   400,000   people   in   the   United   States   and   2.1   million   people   worldwide.     Although   the   disease  was  first  described  in  1868,  by  Jean-­‐Martin  Charcot,  and  much  is  known  about   1 MS,   it   remains   difficult   to   diagnose   and   the   cause   remains   undetermined.     MS   can   affect   people   in   different   ways,   but   the   hallmark   feature   of   the   disease   is   the   break   down  of  the  myelin  sheath,  a  layer  of  lipids  and  proteins  that  covers  the  axon  of  nerve   cells.     When   demyelination   occurs,   characteristic   lesions   form   and   nerve   signals   are   slowed.     This   causes   a   variety   of   neurological   complications,   including   difficulties   with   vision,  motor  skills,  coordination  and  cognition.       MS   has   long   been   considered   to   be   an   autoimmune   disease.     Patients   are   typically   diagnosed   between   the   ages   of   20   and   30,   though   up   to   10%   of   patients   develop   2 symptoms   before   reaching   adulthood.     Women   are   diagnosed   with   more   than   twice   the  frequency  of  men,  and  though  Caucasians  are  twice  as  likely  to  be  affected  as  any   3 other   race,   MS   does   occur   in   most   ethnic   groups.     There   have   been   several   studies   suggesting   a   correlation   between   geographic   location   and   disease   prevalence,   with   4,5 incidences  of  the  disease  increasing  with  distance  from  the  equator.     1         While   MS   is   not   hereditary,   there   is   an   increase   familial   risk,   suggesting   some   genetic   component  to  the  disease.    There  is  a  20-­‐40  fold  increase  of  instance  MS  if  a  first-­‐degree   relative   has   the   disease   and   a   300   fold   increase   if   a   genetically   identical   twin   has   the   disease.    This  suggests  a  genetic  factor  in  the  disease  development,  but  with  only  one   identical   twin   often   being   affected,   it   can   be   assumed   that   genetics   are   not   the   only   6 factor   involved.     Because   of   the   varied   courses   in   the   disease,   it   can   be   difficult   to   predict   how   and   when   the   disease   will   worsen   for   an   individual   patient.     Neurological   problems  generally  develop  after  10-­‐15  years,  with  the  majority  of  patients  losing  their   ability   to   walk   by   15   years   post-­‐diagnosis.     Although   the   disease   is   debilitating,   most   people   with   MS   have   a   normal   life   expectancy.     There   have   been   several   treatments   developed   to   slow   down   the   disease   progression   or   manage   symptoms,   but   there   is   1 currently  no  cure  for  MS.     An   important   feature   of   MS   is   the   breakdown   of   the   blood   brain   barrier   (BBB).   The   BBB   is  the  inner  lining  of  the  capillaries  in  the  brain,  made  up  of  endothelial  cells  that  form   highly   organized   tight   junctions   to   create   a   selectively   permeable   barrier.     The   destruction   of   this   barrier   is   seen   early   in   the   course   of   MS,   occurring   before   the   7 8 appearance   of   lesions.     These   lesions   have   been   found   to   contain   nitric   oxide   (NO),   which  is  a  facet  of  MS  that  has  been  investigated  at  some  length.       In  addition  to  the  NO  in  the  lesions,  it  has  been  reported  that  elevated  levels  of  nitrite   and  nitrate  have  been  measured  in  the  cerebral  spinal  fluid  (CSF),  serum,  and  urine  of     2   8 MS  patients.    Interestingly,  Boje  et  al.  found  that  NO  may  have  a  role  in  the  change  in   9 BBB   permeability.     The   reaction   of   the   NO   radical   with   superoxide   forms   peroxynitrite,   which   has   been   shown   to   have   a   disruptive   effect   on   cerebral   capillaries.     Other   NO   redox   products   have   been   shown   to   have   similar   effects,   but   the   mechanism   remains   undefined.     1.2  Classification,  Diagnosis,  and  Treatment  of  MS       There  are  four  basic  types  of  MS,  each  with  a  different  disease  progression.    The  most   common   form   of   MS   is   relapse-­‐remitting,   characterized   by   periods   of   exacerbated   symptoms,  such  as  vision  problems  or  difficulty  with  motor  skills,  with  returns  to  normal   health   in   between.     As   the   disease   progresses,   many   patients   are   diagnosed   with   secondary-­‐progressive   MS.     This   form   is   also   characterized   by   relapses,   however   after   some   years   there   is   a   steady   worsening   of   overall   neurological   function   between   the   attacks.    More  rare  forms  of  MS  include  primary-­‐progressive  MS,  where  there  is  a  steady   worsening  over  time  with  no  acute  attacks,  and  progressive-­‐relapsing  MS,  which  starts   1 out  like  primary-­‐progressive  MS,  but  exacerbations  of  the  condition  are  also  seen.         It  is  the  nature  of  MS  that  the  disease  does  not  present  the  same  way  across  all  patients.     This   makes   the   diagnosis   of   MS   slow   and   difficult.     Diagnosis   requires   meeting   a   specific   1 set   of   criteria   and   can   take   months   or,   in   some   cases,   years.     Before   confirmed   diagnosis,   most   patients   present   with   a   clinically   isolated   syndrome   (CIS).     CIS   is   defined     3   as   an   isolated   event   that   affects   one   area   of   the   CNS,   in   patients   with   no   history   of   demyelinating  events.    The  event  is  characterized  as  a  neurological  episode  that  is  acute,   lasts   at   least   24   hours   and   is   assumed   to   be   demyelinating.     The   most   common   types   of   CIS  are  optic  neuritis  or  complications  with  the  CNS.    While  nearly  90%  of  MS  patients   were   originally   diagnosed   with   CIS,   not   all   patients   diagnosed   with   CIS   will   go   on   to   10 develop  the  disease.         The  diagnostic  criteria  for  MS  has  been  adapted  and  changed  several  times  in  keeping   with   current   disease   knowledge   and   imaging   technologies.     For   a   clinically   definite   diagnosis  of  MS,  most  criteria  require  the  characteristic  lesions  to  be  separated  in  both   space  and  time  and  any  alternative  diagnoses  to  be  eliminated.    The  current  diagnostic   criteria   were   first   described   by   McDonald   et   al.   in   2001   and   were   revised   in   2005   and   2010. 11-­‐13     To   show   dissemination   in   space,   magnetic   resonance   imaging   (MRI)   must   show  at  least  one  lesion  in  two  of  the  four  following  areas  of  the  CNS:  periventricular,   juxtacortical,  infratentorial  and  spinal  cord.    Examples  of  these  can  be  seen  in  figure  1.1.     The  viewing  of  these  lesions  does  not  require  any  MRI  contrast  media.    However,  for  the   13 demonstration   of   dissemination   in   time,   gadolinium   enhancing   may   be   used.     The   presence   of   gadolinium   enhancement   denotes   BBB   breakdown   and   has   been   14 consistently   found   in   new   lesions   in   MS   patients.     Therefore,   dissemination   in   time   requires   a   new   non-­‐enhanced   lesion   and/or   a   new   gadolinium-­‐enhanced   lesion   on   a         4   Juxtacortical   Periventricular       Infratentorial Spinal  Cord   Figure  1.1  –  Common  Lesions  in  MS  –   The   diagnosis   of   MS   requires   lesions   to   be   found   in   two   of   the   four   following   areas   of   the   CNS:   juxtacortical,   periventricular,   infratentorial,  and  the  spinal  cord.    Above  are  MRI  images  of  lesions  in  these  regions  of   the  brain,  as  well  as  in  the  spinal  cord.    For  interpretation  of  the  references  to  color  in   this   and   all   other   figures,   the   reader   is   referred   to   the   electronic   version   of   this   dissertation.       5     later   MRI   or   the   presence   of   both   non-­‐enhanced   and   enhanced   lesions   13 simultaneously.         In  2005,  criteria  were  added  for  diagnosis  with  progression  from  the  onset,  as  is  seem  in   12 primary-­‐progressive   and   progressive-­‐relapsing   forms   of   the   disease.     These   criteria   were  revised  in  2010  and  diagnosis  now  requires  one  year  of  disease  progression,  plus   two  of  three  of  the  following:  at  least  one  lesion  in  at  least  one  of  characteristic  part  of   the  brain  (periventricular,  juxtacortical  or  infretentorial),  at  least  2  lesions  in  the  spinal   cord,  or  CSF  that  tests  positive  for  immunoglobulin  G  (IgG).     There   is   a   great   need   for   the   discovery   of   biomarkers   in   MS.     Ideally,   diagnosis   and   treatment  would  be  possible  while  the  patients  are  in  the  early  stages  of  CIS.    This  would   potentially   allow   for   a   better   quality   of   life   both   in   the   short   term   and   over   longer   periods  of  time,  if  disease  slowing  treatments  are  effective.    While  the  search  for  an  MS   15 biomarker   has   been   ongoing   for   nearly   a   century,   there   is   still   no   biomarker   used   16 clinically.    Ideal  biomarkers  are  ones  that  can  be  objectively  measure  and  evaluated  as   17 indicators  of  biological  processes  or  drug  response.         There  are  four  proposed  biomarker  categories  for  MS:    diagnostic  biomarkers,  predictive   biomarkers,   process-­‐specific   biomarkers   and   treatment-­‐related   biomarkers.     These   are   18 discussed   at   length   in   chapter   4.     Currently,   the   only   potential   biomarker   used   clinically  is  IgG,  an  indicator  of  immune  response  that,  while  it  is  associated  with  MS,  it  is     6   not  specific  enough  to  be  considered  as  a  true  biomarker  for  the  disease.    There  have   been   other   recent   findings   in   the   field   of   MS   biomarkers,   however,   like   IgG,   most   use   invasive  CSF  testing.    The  CSF  is  traditionally  believed  to  be  the  most  likely  source  of  an   MS   biomarker   because   of   the   proximity   to   the   inflammation   and   lesions   in   the   CNS,   though  this  may  not  be  the  case,  as  CSF  is  collected  through  a  lumbar  puncture  in  the   lower   back   and   may   not   accurately   reflect   inflammatory   markers   in   the   brain   regions   19 were  most  MS  lesions  occur.     Although  there  is  still  no  cure  for  MS,  there  are  treatments  available  that  will  alter  the   course   of   the   disease,   slowing   the   rate   of   disability,   as   well   as   drugs   for   symptom   management   to   help   patients   cope.     The   first   drug   for   the   treatment   of   MS   was   approved  by  the  US  Food  and  Drug  Administration  (FDA)  in  1993.    Since  then,  nine  drugs   20 have  been  approved  by  the  FDA  for  treatment.  A  complete  list  of  FDA  approved  MS   drugs,  their  dosage  and  effects  can  be  seen  in  table  1.1.         Four  of  the  nine  drugs,  including  the  first,  Betaseron  (IFN-­‐β-­‐1b),  are  in  a  family  of  drugs   known   as   interferons   (IFNs).     In   1996,   the   second   IFN   drug,   Avonex   (IFN-­‐β-­‐1a)   was   approved,   followed   by   Rebif   (IFN-­‐β-­‐1a)   in   2002   and   Extavia   (IFN-­‐β-­‐1b)   in   2009.     These   medications  all  function  similarly,  differing  in  dosage,  effectiveness,  and  side  effects.  For   use   in   treatment,   IFNs   are   manufactured   using   recombinant   DNA,   but   IFNs   are   also   naturally  present  in  humans.  Their  role  is  complex  and  not  completely  understood,  but           7     Drug   Year   Type   Dosing   Effects   Side  Effects   2012   Pyrimidine  synthesis   One  tablet   Reduced   Severe  liver   Aubagio   inhibitor:    reduces   per  day   relapse  rates,   toxicity     proliferation  of   fewer  new  and   immune  cells   active  lesions   20 1996   INF-­‐β-­‐1a*   Intra-­‐ Reduced  risk  of   heart   Avonex   muscular   disability   problems,   injection   progression,   depression,   Once  per   fewer   seizures     week   exacerbations   Betaseron   1993   INF-­‐β-­‐1b*   Sub-­‐ Stops  the   Flu-­‐like   20 2009   cutaneous   increase  in  total   symptoms,   Extavia   injection   lesion  area   injection  site   Every  other   necrosis   day   24 1996   Polymer  of  amino   Sub-­‐ Decrease  in   Chest   Copaxone     acids,  shifts  the   cutaneous   annual  relapse   tightness,   immune  effect  to  be   injection   rates,  decrease   heart   less  inflammatory   Every  day   in  new  lesions   palpitations,     25 2010   Sphingosine-­‐1-­‐ One  capsule   Reduced   Decreased   Gilenya   phosphate  receptor   per  day   relapse  rate,   heart  rate,     modulator,  prevents   reduced  lesion   swelling  of   lymphocytes  from   activity   the  eye,  liver   leaving  nodes   toxicity   64 2000   Antineoplastic,   Intravenous Delays  time  to   Cardio-­‐ Novantrone   intercalates  in  the   Every  3   relapse  and   toxicity,     DNA,  suppresses  T   months   disability   increased   and  B  cells  and   progression,   leukemia  risk   macrophages     reduced  lesions   20 2002   INF-­‐β-­‐1a*   Sub-­‐ Lower  relapse   Depression,   Rebif   cutaneous   rate,  prolonged   liver  toxicity,   injection   time  to  first   seizures   Three  times   relapse,  delay   per  week   in  progression   65,66 2006   Monoclonal   Intravenous Reduced  risk  of   Increased  risk   Tysabri   antibody:    Lessens   Once  per   disability   of  leukemia     movement  of   month   progression,   and  serious   immune  cells  across   fewer  relapses   infection   the  BBB   Table  1.1:    Medications  used  in  MS.  *IFNs  inhibit  the  proliferation  of  T  cells,  decreases   63 23 the  production  of  proinflammatory  cytokines     8     it   is   known   that   they   are   involved   in   modulating   the   immune   system. 21,22     In   the   treatment  of  MS,  INF-­‐β  has  been  shown  to  inhibit  the  proliferation  of  T  cells,  which  have   a   variety   of   functions   in   vivo,   including   regulating   the   immune   system,   as   well   as   23 decrease  the  production  of  proinflammatory  cytokines.         In   1996,   the   first   non-­‐IFN   MS   drug   was   approved   by   the   FDA;   Copaxone   (glatiramer   acetate)   has   since   become   the   most   prescribed   single   drug   for   the   treatment   of   MS.     This   random   polymer   of   the   four   amino   acids   found   in   myelin   basic   protein   (MBP)   (glutamic   acid,   lysine,   alanine,   and   tyrosine)   was   originally   intended   for   use   to   mimic   MBP   in   the   investigation   of   its   ability   to   cause   irritation   and   swelling   in   the   brain,   but   was   found   to   have   positive,   rather   than   negative,   effects.   Now,   it   is   believed   to   result   in   a   shift   in   the   characteristics   of   the   immune   system   from   being   dominated   by   the   proinflammatory   T-­‐helper-­‐1   cells   to   the   less   harmful   T-­‐helper-­‐2   cells.     The   mechanism   for   this   is   not   well   understood,   and   there   may   be   more   immune   effects   than   first   24 realized.    All  of  the  previously  mentioned  drugs,  as  well  as  Novantrone  (mitoxantrone)   and  Tysabri  (natalizumab)  are  dosed  by  injection  or  intravenously.         In   2010,   the   first   oral   MS   medication   was   FDA   approved.     Gilenya   (fingolimod)   has   been   shown   to   reduce   relapse   rate   and   lesion   activity   with   a   capsule   medication   that   is   taken   once  per  day.    Gilenya  is  a  receptor  modulator  that  prevents  lymphocytes  from  leaving   nodes  and  crossing  the  BBB,  reducing  inflammatory  damage  to  nerve  cells;  however,  it   25 can   have   serious   cardiac   side   effects.     Two   years   later,   in   2012,   a   second   oral     9   medication,  Aubagio  (teriflunomide),  was  approved  by  the  FDA.    This  drug  also  works  by   interaction   with   the   immune   system,   reducing   the   proliferation   of   certain   types   of   immune   cells.     This   decreases   the   number   of   new   lesions   and   reduces   relapse   rates.     While  this  drug  is  new  to  the  market  at  the  time  of  writing,  the  treatment  has  been  well   26 tolerated  with  few  adverse  reactions.     While  disease-­‐modifying  treatments  have  been  used,  with  varying  success,  to  alter  the   course   of   the   disease,   patients   still   experience   exacerbations,   or   relapses,   of   MS.     The   20 most   common   agent   used   in   these   cases   is   steroid   therapy.     Not   all   exacerbations   require   treatment,   but   if   the   symptoms   are   impacting   a   patient’s   ability   to   function,   the   most  commonly  prescribed  steroid  treatment  for  MS  is  glucocorticosteroids  (GCS)  with   prednisone,   methylprednisone   and   dexamethasone   being   the   top   three   of   that   group.     Steroids  are  believed  to  impact  MS  through  modulation  of  the  immune  system,  reducing   27 both   lymphocyte   and   pro-­‐inflammatory   cytokine   levels.     GCSs   shorten   the   duration   of   exacerbations,  and  it  has  been  suggested  that  there  may  be  long-­‐term  benefits  as  well,   though   there   has   been   some   question   of   the   validity   of   the   studies   leading   to   those   28 conclusions.    High  dose  steroids,  like  those  used  to  treat  MS  are  often  associated  with   serious   side   effects,   such   as   depression,   weight   gain,   insomnia,   moodiness,   and   27 hyperglycemia.     However,   when   used   for   MS   treatment,   steroids   are   generally   administered   in   short   pulses   and   are   well   tolerated   with   only   mild   side   effects   28 reported.       10   1.3  Animal  Model  of  MS       In   1933,   in   order   to   more   effectively   study   MS,   an   animal   model,   experimental   autoimmune  encephalomyelitis  (EAE),  was  developed.    Although  there  are  limitations  to   29 using  EAE  as  an  MS  model,  including  the  inability  of  EAE  to  mimic  all  types  of  MS,  it   has   proven   to   be   a   valuable   tool   in   MS   research   since   its   development   began.     Discovered   initially   as   a   side   effect   of   the   rabies   vaccine,   it   was   subsequently   found   that   a   series   of   injections   containing   neural   tissue   would   induce   an   autoimmune   response   in   a  large  variety  of  animals,  including  primates  and  rodents.    It  has  been  found  that  EAE   can   be   induced   with   crude   brain   extract   and   whole   myelin,   as   well   as   with   isolated   antigens:   myelin   basic   protein   (MBP),   proteolipid   protein   (PLP)   or   myelin   oligodendrocyte  glycoprotein  (MOG),  or  by  using  synthetic  peptides.         The  antigens  are  generally  injected  along  with  complete  Freund’s  adjuvant  and  pertussis   30 toxin,  to  increase  the  immune  response.    Clinical  signs  of  EAE  are  generally  seen  8-­‐14   days   after   induction,   and   the   progression   of   the   disease   is   measured   using   a   clinical   disease   scale   from   0-­‐5,   with   0   representing   no   clinical   signs   and   5   representing   total   31 paralysis  or  death.    This  animal  model  is  often  used  for  preliminary  investigations  of   disease   modifying   drugs,   including   several   that   will   be   discussed   in   greater   detail   presently.           11   1.4  Adenosine  Triphosphate  and  Nitric  Oxide       Recently,   a   small   scale   study   by   the   Spence   group   revealed   that   the   red   blood   cells   (RBCs)   from   MS   patients   release   nearly   three   times   the   amount   of   adenosine   32 triphosphate   (ATP),   in   response   to   shear   stress,   than   RBCs   from   healthy   controls.     Importantly,   ATP   release   from   RBCs   is   known   to   increase   as   the   cell   deformability   increases  and  it  has  been  reported  that  RBCs  obtained  from  MS  patients  have  properties   that  would  suggest  an  increase  in  deformability.    The  release  of  ATP  from  RBCs  results  in   the  stimulation  of  NO  synthesis  in  endothelial  cells. 33,34    As  mentioned  previously,  MS   patients   have   elevated   levels   of   nitrite   and   nitrate   in   their   CSF,   serum,   and   urine,   and   the  lesions  that  occur  as  a  result  of  demyelination  have  been  reported  to  have  increased   8 levels  of  NO.       Sprague  et  al.  have  shown  that  the  mechanical  deformation  experienced  by  circulating   33 RBC   results   in   ATP   release.     While   the   exact   mechanism   for   this   release   is   not   fully   35 known,   a   G-­‐protein   coupled   receptor   (GPCR),   cyclic   adenosine   monophosphate   36 (cAMP),   and   the   cystic   fibrosis   transmembrane   conductance   regulator   (CFTR)   37 protein  have  all  been  implicated  in  this  process.    It  is  expected  that  the  increase  in  ATP   release   from   RBCs   would   lead   to   a   subsequent   increase   in   NO   production   in   endothelial   33,34 cells   because   ATP   is   a   stimulus   of   NO   production   in   endothelial   cells.   12       In   this   construct,  the  binding  of  ATP  to  the  P2Y  receptor  on  the  endothelial  cell  results  in  the   activation   of   nitric   oxide   synthase   (NOS)   and   the   production   of   NO. 33,38     A   schematic   of   the  proposed  mechanism  can  be  seen  in  figure  1.2.     Recently,  receptors  for  ATP  have  been  found  to  have  a  role  in  MS. 39,40    Several  studies   investigated  the  role  that  the  ATP  receptor,  P2X7,  plays  in  MS.  The  P2X7  receptor  is  a   purinergic  ligand-­‐gated  cation  channel  receptor  that  is  activated  by  ATP.    This  receptor  is   found   mainly   on   immune   cells   both   in   the   peripheral   and   central   nervous   systems,   39 though  it  has  been  found  on  neurons  as  well  as  oligodendrocytes.    Low  doses  of  ATP   + 2+ reversibly  open  the  channel  to  small  cations,  such  as  Na  and  Ca ,  while  prolonged  or   repeated   stimulation   with   higher   levels   of   ATP   will   increase   the   pore   size   to   accommodate  larger  molecules  with  masses  in  the  range  of  100s  of  Daltons.    As  a  result   of  the  increased  permeability,  high  extracellular  concentrations  of  ATP  have  a  cytotoxic   41 effect  on  some  cells.       Matute   et   al.   showed   that   the   activation   of   P2X7   has   cytotoxic   effects   on   oligodendrocytes  both  in  vivo  and  in  vitro,  and  the  lesions  formed  in  mice  as  a  result  are   39 very  similar  to  those  found  in  MS.    In  addition,  they  found  that  there  was  a  significant   increase  in  the  levels  of  P2X7  in  both  the  optic  nerve  and  axonal  tract  of  MS  patients,   suggesting   an   increase   vulnerability   of   the   oligodendrocytes.     Feinstein   et   al.   have   investigated   the   effects   of   a   P2X7   deficiency   on   EAE.     P2X7   deficient   mice   were   four     13       Figure   1.2   –   Proposed   mechanism   of   ATP   release   from   RBCs   and   subsequent   NO   35 production   –   It   has   been   proposed   that   G-­‐protein   coupled   receptor   (GPCR),   cyclic   36 adenosine   monophosphate   (cAMP),   and   the   cystic   fibrosis   transmembrane   67   conductance   regulation   (CFTR)   protein are   all   required   for   the   release   of   ATP   by   mechanical   deformation,   though   how   the   ATP   leaves   the   cell   remains   in   question.     In   this  proposed  mechanism,  the  binding  of  ATP  to  the  P2Y  receptor  on  the  endothelial  cell   68 results  in  the  activation  NOS  and  the  production  of  NO.           14   40 times   less   likely   to   suffer   from   the   effects   of   EAE   than   normal   control   mice.     Interestingly,  the  P2X7  deficient  mice  still  had  increased  levels  of  cytokine  production,   resulting   from   an   immune   response.     This   suggests   that   these   mice   lack   an   initiation   event  in  the  development  of  EAE,  which  points  to  a  role  of  P2X7   in  this  disease,  as  well   as  in  MS.       RBCs  lack  mitochondria,  and  therefore  the  only  source  of  ATP  is  glycolysis.    Because  of   this,   it   is   expected   that   the   altered   ATP   release   in   the   RBCs   of   MS   patients   would   be   accompanied  by  abnormal  glucose  metabolism.    In  1966,  Raczkiewicz  and  Leyko  showed   that   MS   patients   had   a   significant   increase   in   the   ATP   content   of   their   RBCs   after   an   42 intake  of  50  g  of  glucose,  in  comparison  to  their  ATP  content  while  fasting.    Such  an   increase   was   not   seen   in   healthy   controls.     More   recently,   Regenold   et   al.   noted   an   43 increase  in  glucose  metabolism  outside  of  the  mitochondria  in  the  CSF  of  MS  patients.     This   is   significant   when   considering   the   lack   of   mitochondria   in   RBCs   and   the   increase   in   ATP   release   seen   from   the   RBCs   of   MS   patients.     Mitochondrial   dysfunction   in   MS   patients  may  be  leading  to  increase  glucose  metabolism  in  the  RBC.     1.5  Pregnancy  and  Estrogens  in  MS       While   there   are   treatments   for   MS,   there   are   still   no   concrete   explanations   for   the   exacerbations  and  ameliorations  of  the  disease.    Additionally,  MS  is  more  prevalent  in   44 females   than   males   in   a   ratio   between   2:1   and   3:1,   varying   by   region.     The   sex     15   45 discrepancy   is   seen   in   the   development   of   EAE   in   certain   strains   of   mice   as   well.     While  women  are  affected  more  than  twice  as  often  as  men,  they  tend  to  have  slower   disease  courses,  resulting  in  a  longer  time  between  disease  onset  and  certain  disability   levels.     Men   with   MS,   on   average,   have   a   shorter   time   before   relapse-­‐remitting   MS   becomes  secondary  progressive  MS  and  a  greater  amount  of  neurological  damage,  seen   46 in  the  form  of  brain  cell  death.         Research  has  been  conducted  in  several  areas,  such  as  genetics,  environmental  factors,   and   hormones,   in   an   attempt   to   explain   the   sex   differences.     Genetically,   it   has   been   suggested   that   there   may   be   a   protective   effect   found   on   the   Y   chromosome   or   a   44 promoting  effect  on  the  X  gene,  where  females  would  be  doubly  exposed.    In  the  past   60  years,  there  has  been  an  increase  in  the  ratio  of  female  to  male  diagnoses,  with  an   increase   in   female   diagnoses   being   credited   with   the   change.     Because   of   the   social   changes  that  have  taken  place  over  the  past  six  decades  affecting  the  role  of  women  in   the  home  and  workplace,  it  has  been  proposed  that  the  increase  may  be  the  result  of   environmental   factors.     Some   factors   that   have   been   suggested   for   further   epidemiological   study   include:     occupation,   obesity,   dietary   habits,   birth   control   and   later  childbirth. 47,48    The  sex  differences  in  MS,  as  well  as  anecdotal  patient  experiences   led  to  the  study  of  pregnancy  and  sex  hormones,  the  role  they  play  in  the  disease,  and   their  potential  use  as  therapeutics.         16   Up   until   the   1960s,   women   with   MS   were   advised   not   to   conceive,   as   pregnancy   was   thought  to  worsen  the  disease.    In  1998,  a  pregnancy  in  MS  (PRIMS)  study  investigated   49 the  number  of  relapses  during  each  trimester  before,  during  and  after  pregnancy.    It   was   found   that   the   number   of   relapses   was   decreased   during   pregnancy,   especially   in   the   third   trimester,   and   a   spike   in   relapses   was   often   seen   in   the   first   trimester   post-­‐ partum.         It  has  been  suggested  that  this  trend  corresponds  to  the  levels  of  sex  steroids,  such  as   estradiol   and   estriol,   that   are   present   during   pregnancy.     When   the   levels   are   at   their   highest,   the   relapse   rates   are   at   their   lowest.     Levels   of   these   hormones   dramatically   3 decrease   after   delivery,   and   a   corresponding   spike   in   relapse   rates   is   seen.     There   is   currently   an   on-­‐going   clinical   trail,   the   Prevention   of   Post-­‐Partum   Relapses   with   Progestin  and  Estradiol  in  Multiple  Sclerosis  (POPART’MUS)  trial,  aimed  at  ameliorating   these   relapses.     After   delivery,   women   with   MS   are   treated   with   estradiol   and   high   doses  of  progestin  and  are  monitored  for  three  months. 44,50     Since  PRIMS,  several  studies  have  looked  at  the  effects  of  estrogens  on  MS,  as  well  as  on   the   animal   model   EAE. 51-­‐53     Jansson   et   al.   investigated   the   effects   of   castration   of   female   mice   on   disease   onset,   as   well   as   the   effect   of   long   term   treatment   with   51 pregnancy   levels   of   estradiol   and   estriol.     It   was   found   that   both   estrogens   delayed   the  onset  of  EAE,  though  estriol  did  so  at  a  more  physiologically  relevant  concentration     17   and   for   a   longer   time.     Results   from   studies   looking   at   the   MS   relapse   rates   in   pregnancy,  as  well  as  the  study  done  by  Jansson  et  al.  have  lead  to  a  Phase  I  clinical  trial   53 treating   women   with   MS   with   the   pregnancy   hormone   estriol.     Although   this   was   a   small  scale  trial  involving  only  10  women,  it  showed  promising  results.    All  10  patients   had   a   decrease   in   size   and   number   of   lesions   while   being   treated   with   estriol.     When   treatment   ceased,   lesion   levels   returned   to   their   pretreatment   state   within   three   months.     Once   treatment   was   reinstated,   there   was   again   a   decrease   in   the   size   and   number  of  lesions.         The  results  of  this  study  led  Palaszynski  et  al.  to  investigate  the  effectiveness  of  estriol   52 treatment  on  EAE  in  male  mice.    This  study  showed  that  both  male  and  female  mice   pretreated  with  estriol  showed  no  clinical  signs  of  EAE  for  up  to  thirty  days  after  MOG   injections,   suggesting   that   estriol   may   be   a   beneficial   treatment   for   both   sexes.     In   addition,   there   are   currently   large   scale   clinical   trials   looking   at   the   effects   of   estriol   alone,  as  well  as  the  hormone  used  in  conjunction  with  Copaxone. 54,55     It   is   currently   thought   that   estrogens   may   ameliorate   MS   through   a   shift   in   immune   response  from  T  helper  1  (Th1)  cells  to  T  helper  2  (Th2)  cells. 52,53    However,  as  seen  in   figure   1.3,   estradiol   and   estriol   have   similar   structure   to   dehydroxyepiandosterone   (DHEA),   a   precursor   for   both   molecules.     It   has   previously   been   shown   by   the   Spence   56 group  that  DHEA  attenuates  ATP  release  from  healthy  rabbit  RBCs.    This  suggests  the       18     Figure   1.3   –   Steroid   structures   –   Dehydroxyepiandosterone   (DHEA)   is   a   precursor   of   both   estradiol   and   estriol,   and   all   three   steroids   have   similar   structures   leading   to   the   hypothesis  that  the  estrogens  will  have  a  similar  effect  on  the  ATP  release  from  the  RBC   as  that  of  DHEA.     19   potential  for  a  similar  effect  on  the  RBC  with  both  estrogen  and  estriol.    A  reduction  in   ATP  release  from  RBCs  would  lead  to  a  decrease  in  NO  production  of  endothelial  cells.     57,58 This  is  contrary  to  much  of  the  current  literature,  which  reports  stimulation  of  NO   production   in   endothelial   cells   in   response   to   estrogens.     It   is   interesting   to   note,   however,  that  previous  studies  showing  the  stimulation  of  NO  production  by  estrogens   58 were  not  performed  in  the  presence  of  RBCs.     1.6  Zinc  and  MS       59 The  idea  of  altered  zinc  status  in  MS  has  been  proposed  since  the  late  1970s,  but  the   data  associating  zinc  and  MS  has  been  generally  overlooked  by  MS  researchers  with  only   a  handful  of  reports  on  the  subject  in  the  past  four  decades.    Interestingly,  there  have   been   several   studies   done   on   clusters   of   MS   patients,   meaning   areas   where   there   are   higher   than   normal   rates   of   MS   diagnosis.     In   many   of   these   areas,   it   was   found   that   there  was  an  issue  with  zinc  contamination  of  some  type,  either  in  the  soil  and  water,  or   from  zinc  smelters. 60-­‐62       In   addition,   the   RBCs   of   MS   patients   have   been   found   to   have   increases   in   the   zinc   levels,   even   without   being   part   of   an   MS   cluster.     The   level   of   zinc   in   the   RBC   was   found   to  be  significantly  higher  in  MS  patients  than  that  of  healthy  controls,  as  well  as  being   higher   than   the   level   in   patients   with   other   neurological   impairments   or   inflammatory   diseases.    It  is  interesting  to  note  that  the  serum  zinc  levels  did  not  differ  between  these     20   groups,   suggesting   the   increase   in   RBC   zinc   levels   is   not   a   result   of   defective   zinc   59 absorption  or  individual  nutrition.         In   1983,   Dore-­‐Duffy   et   al.   proposed   that   the   increase   in   RBC   zinc   levels   may   be   an   indication  of  an  increase  affinity  of  zinc-­‐carrier  protein  complexes  to  surface  receptors   on  the  RBCs,  or  the  RBCs  from  MS  patients  have  an  increased  number  of  receptors  for   these  proteins.    Thirty  years  later,  the  work  presented  here  furthers  those  proposals  and   answers   some   questions   to   the   underlying   mechanism   of   the   etiology   of   MS.     Data   here   will  show  support  for  the  hypothesis  that  an  increase  of  zinc  delivery  to  the  RBC  results   in   an   increase   in   glucose   uptake   into   the   cell.     From   this,   the   release   of   ATP   at   three   times   the   level   of   controls   results   in   an   increase   in   the   amount   of   NO   produced   in   endothelial  cells,  potentially  leading  to  BBB  breakdown  and  disease  progression  in  MS.     21   REFERENCES   22       1     2     3     4     5     6     7     8     9     10     11   REFERENCES   Cook,  S.  D.          (Taylor  &  Francis,  New  York,  2006).   Gorman,   M.,   Healy,   B.,   Polgar-­‐Turcsanyi,   M.   &   Chitnis,   T.   Relapses   are   more   frequent   in   pediatric-­‐onset   than   adult-­‐onset   multiple   sclerosis.   Multiple   Sclerosis   14,  S66-­‐S66  (2008).   Vukusic,   S.   &   Confavreux,   C.   Pregnancy   and   multiple   sclerosis:   the   children   of   PRIMS.  Clinical  neurology  and  neurosurgery  108,  266-­‐270  (2006).   Handel,   A.,   Giovannoni,   G.,   Ebers,   G.   &   Ramagopalan,   S.   Environmental   factors   and   their   timing   in   adult-­‐onset   multiple   sclerosis.   Nature   Reviews   Neurology   6,   156-­‐166  (2010).   Pugliatti,   M.,   Sotgiu,   S.   &   Rosati,   G.   The   worldwide   prevalence   of   multiple   sclerosis.  Clinical  Neurology  and  Neurosurgery  104,  182-­‐191  (2002).   Sadovnick,  A.  D.,  Dyment,  D.  A.,  Ebers,  G.  C.  &  Risch,  N.  J.  Evidence  for  genetic   basis  of  multiple  sclerosis.  Lancet  347,  1728-­‐1730  (1996).   Minagar,  A.  &  Alexander,  J.  S.  Blood-­‐brain  barrier  disruption  in  multiple  sclerosis.   Multiple  Sclerosis  9,  540-­‐549  (2003).   Smith,  K.  J.  &  Lassmann,  H.  The  role  of  nitric  oxide  in  multiple  sclerosis.  Lancet   Neurology  1,  232-­‐241  (2002).   Boje,   K.   M.   K.   &   Lakhman,   S.   S.   Nitric   oxide   redox   species   exert   differential   permeability   effects   on   the   blood-­‐brain   barrier.   Journal   of   Pharmacology   and   Experimental  Therapeutics  293,  545-­‐550  (2000).   Dalton,  C.  M.,  Brex,  P.  A.,  Miszkiel,  K.  A.,  Hickman,  S.  J.,  MacManus,  D.  G.,  Plant   G.  T.,  Thompson  A.  J.  &  Miller  D.  H.  Application  of  the  new  McDonald  criteria  to   patients   with   clinically   isolated   syndromes   suggestive   of   multiple   sclerosis.   Annals  of  Neurology  52,  47-­‐53  (2002).   McDonald,  W.  I.,  Compston,  A.,  Edan,  G.,  Goodkin,  D.,  Hartung,  H.  P.,  Lublin,  F.   D.,   McFarland,   H.   F.,   Paty,   D.   W.,   Polman,   C.   H.,   Reingold,   S.   C.,   Sandberg-­‐ Wollheim,   M.,   Sibley,   W.,   Thompson,   A.,   van   den   Noort,   S.,   Weinshenker,   B.   Y.   &     Wolinsky,  J.  S.  Recommended  diagnostic  criteria  for  multiple  sclerosis:  Guidelines   from   the   International   Panel   on   the   Diagnosis   of   Multiple   Sclerosis.   Annals   of   Neurology  50,  121-­‐127  (2001).       23   12     13     14     15     16     17     18     19     20     21   Polman,   C.   H.,   Reingold,   S.   C.,   Edan,   G.,   Filippi,   M.,   Hartung,   H.   P.,   Kappos,   L.,   Lublin,   F.   D.,   Metz,   L.   M.,   McFarland,   H.   F.,   O'Connor,   P.   W.,   Sandberg-­‐Wollheim,   M.,  Thompson,  A.  J.,  Weinshenker,  B.  G.  &  Wolinsky,  J.  S.  Diagnostic  criteria  for   multiple   sclerosis:   2005   Revisions   to   the   "McDonald   Criteria".   Annals   of   Neurology  58,  840-­‐846  (2005).   Polman,   C.   H.,   Reingold,   S.   C.,   Banwell,   B.,   Clanet,   M.,   Cohen,   J.   A.,   Filippi,   M.,   Fujihara,   K.,   Havrdova,   E.,   Hutchinson,   M.,   Kappos,   L.,   Lublin,   F.   D.,   Montalban,   X.,   O'Connor,   P.,   Sandberg-­‐Wollheim,   M.,   Thompson,   A.   J.,   Waubant,   E.,   Weinshenker,  B.  &  Wolinsky,  J.  S.  Diagnostic  Criteria  for  Multiple  Sclerosis:  2010   Revisions  to  the  McDonald  Criteria.  Annals  of  Neurology  69,  292-­‐302  (2011).   Kermode,   A.   G.,   Thompson,   A.   J.,   Tofts,   P.,   MacManus,   D.   G.,   Kendall,   B.   E.,   Kingsley,   D.   P.,   Moseley,   I.   F.,   Rudge,   P.   &   McDonald,   W.   I.   Breakdown   of   the   blood-­‐brain-­‐barrier   precedes   symptoms   and   other   MRI   signs   of   new   lesions   in   multiple   sclerosis   -­‐   pathogeneic   and   clinical   implications.   Brain   113,   1477-­‐1489   (1990).   Awad,   A.,   Hemmer,   B.,   Hartung,   H.   P.,   Kieseier,   B.,   Bennett,   J.   L.   &   Stuve,   O.   Analyses   of   cerebrospinal   fluid   in   the   diagnosis   and   monitoring   of   multiple   sclerosis.  Journal  of  Neuroimmunology  219,  1-­‐7  (2010).   Avsar,  T.,  Korkmaz,  D.,  Tütüncü,  M.,  Demirci,  N.  O.,  Saip,  S.,  Kamasak,  M.,  Siva,  A.   &   Turanli,   E.   T.   Protein   biomarkers   for   multiple   sclerosis:   semi-­‐quantitative   analysis  of  cerebrospinal  fluid  candidate  protein  biomarkers  in  different  forms  of   multiple  sclerosis.  Multiple  Sclerosis  Journal  18,  1081-­‐1091  (2012).   Biomarkers   Definitions   Working   Group.   Biomarkers   and   surrogate   endpoints:   Preferred   definitions   and   conceptual   framework.   Clinical   Pharmacology   &   Therapeutics  69,  89-­‐95  (2001).   Comabella,  M.  &  Khoury,  S.  J.  in  Multiple  Sclerosis:  Diagnosis  and  Therapy      (ed   Howard  L.  Weiner)  Ch.  2,  26-­‐55  (John  Wiley&  Sons,  2012).   Bielekova,  B.  &  Martin,  R.  Development  of  biomarkers  in  multiple  sclerosis.  Brain   127  (2004).   Stankiewicz,   J.   M.   &   Khoury,   S.   J.   in   Multiple   Sclerosis:     Diagnosis   and   Therapy       (ed  Howard  L.  Weiner)  Ch.  8,  183-­‐212  (John  Wiley  &  Sons,  2012).   Zhang,   J.,   Hutton,   G.   &   Zang,   Y.   A   comparison   of   the   mechanisms   of   action   of   interferon   beta   and   glatiramer   acetate   in   the   treatment   of   multiple   sclerosis.   Clinical  Therapeutics  24,  1998-­‐2021  (2002).       24   22     23     24     25     26     27     28     29     30     31     32     Tremlett,   H.,   Yoshida,   E.   &   Oger,   J.   Liver   injury   associated   with   the   beta-­‐ interferons   for   MS   -­‐   A   comparison   between   the   three   products.   Neurology   62,   628-­‐631  (2004).   Yong,   V.,   Chabot,   S.,   Stuve,   O.   &   Williams,   G.   Interferon   beta   in   the   treatment   of   multiple  sclerosis  -­‐  Mechanisms  of  action.  Neurology  51,  682-­‐689  (1998).   Farina,   C.,   Weber,   M.,   Meinl,   E.,   Wekerle,   H.   &   Hohlfeld,   R.   Glatiramer   acetate   in   multiple  sclerosis:  update  on  potential  mechanisms  of  action.  Lancet  Neurology   4,  567-­‐575  (2005).   Kappos,   L.,   Radue,   E.   W.,   O'Connor,   P.,   Polman,   C.,   Hohlfeld,   R.,   Calabresi,   P.,   Selmaj,   K.,   Agoropoulou,   C.,   Leyk,   M.,   Zhang-­‐Auberson,   L.   &   Burtin,   P.   A   Placebo-­‐ Controlled  Trial  of  Oral  Fingolimod  in  Relapsing  Multiple  Sclerosis.  New  England   Journal  of  Medicine  362,  387-­‐401  (2010).   O'Connor,  P.  W.,  Li,  D.,  Freedman,  M.  S.,  Bar-­‐Or,  A.,  Rice,  G.  P.,  Confavreux,  C.,   Paty,  D.  W.,  Stewart,  J.  A.  &  Scheyer  R.  A  Phase  II  study  of  the  safety  and  efficacy   of   teriflunomide   in   multiple   sclerosis   with   relapses.   Neurology   66,   894-­‐900   (2006).   Burton,   J.,   O'Connor,   P.,   Hohol,   M.   &   Beyene,   J.   Oral   versus   Intravenous   Steroids   for  Treatment  of  Relapses  in  Multiple  Sclerosis.  Cochrane  Database  of  Systematic   Reviews  8,  No.  CD006921  (2009).   Andersson,  P.  &  Goodkin,  D.  Glucocorticosteroid  therapy  for  multiple  sclerosis:  A   critical  review.  Journal  of  the  Neurological  Sciences  160,  16-­‐25  (1998).   Gold,  R.,  Linington,  C.  &  Lassmann,  H.  Understanding  pathogenesis  and  therapy   of   multiple   sclerosis   via   animal   models:   70   years   of   merits   and   culprits   in   experimental   autoimmune   encephalomyelitis   research.   Brain   129,   1953-­‐1971   (2006).   Lavi,   E.   &   Constantinescu,   C.   S.   Experimental   models   of   multiple   sclerosis.     (Springer,  2005).   Hjelmstrom,   P.,   Juedes,   A.   E.,   Fjell,   J.   &   Ruddle,   N.   H.   B-­‐cell-­‐deficient   mice   develop  experimental  allergic  encephalomyelitis  with  demyelination  after  myelin   oligodendrocyte   glycoprotein   sensitization.   Journal   of   immunology   161,   4480-­‐ 4483  (1998).   Letourneau,   S.,   Hernandez,   L.,   Faris,   A.   N.   &   Spence,   D.   M.   Evaluating   the   effects   of   estradiol   on   endothelial   nitric   oxide   stimulated   by   erythrocyte-­‐derived   ATP   25     33     34     35     36     37     38     39     40     41     using   a   microfluidic   approach.   Analytical   &   Bioanalytical   Chemistry   379,   3369-­‐ 3375  (2010).   Sprague,  R.   S.,   Ellsworth,  M.  L.,  Stephenson,  A.  H.  &  Lonigro,  A.  J.  ATP:  the  red   blood   cell   link   to   NO   and   local   control   of   the   pulmonary   circulation.   American   Journal  of  Physiology  271,  H2717-­‐2722  (1996).   Oblak,   T.   D.,   Root,   P.   &   Spence,   D.   M.   Fluorescence   monitoring   of   ATP-­‐ stimulated,   endothelium-­‐derived   nitric   oxide   production   in   channels   of   a   poly(dimethylsiloxane)-­‐based  microfluidic  device.  Analytical  Chemistry  78,  3193-­‐ 3197  (2006).   Olearczyk,  J.  J.,  Stephenson,  A.  H.,  Lonigro,  A.  J.  &  Sprague,  R.  S.  Heterotrimeric  G   protein   G(i)   is   involved   in   a   signal   transduction   pathway   for   ATP   release   from   erythrocytes.   American   Journal   of   Physiology-­‐Heart   and   Circulatory   Physiology   286,  H940-­‐H945  (2004).   Sprague,  R.  S.,  Ellsworth,  M.  L.,  Stephenson,  A.  H.  &  Lonigro,  A.  J.  Participation  of   cAMP  in  a  signal-­‐transduction  pathway  relating  erythrocyte  deformation  to  ATP   release.   American   Journal   of   Physiology   -­‐   Cell   Physiology   281,   C1158-­‐C1164   (2001).   Sprague,   R.   S.,   Ellsworth,   M.   L.,   Stephenson,   A.   H.,   Kleinhenz,   M.   E.   &   Lonigro,   A.   J.  Deformation-­‐induced  ATP  release  from  red  blood  cells  requires  CFTR  activity.   American   Journal   of   Physiology   -­‐   Heart   and   Circulatory   Physiology   44,   H1726-­‐ H1732  (1998).   Sprague,   R.   S.,   Olearczyk,   J.   J.,   Spence,   D.   M.,   Stephenson,   A.   H.,   Sprung,   R.   W.   &   Lonigro,   A.   J.   Extracellular   ATP   signaling   in   the   rabbit   lung:   erythrocytes   as   determinants   of   vascular   resistance.   American   Journal   of   Physiology-­‐Heart   and   Circulatory  Physiology  285,  H693-­‐H700  (2003).   Matute,  C.,  Torre,  I.,  Pérez-­‐Cerdá,  F.,  Pérez-­‐Samartín,  A.,  Alberdi,  E.,  Etxebarria,   E.,   Arranz,   A.   M.,   Ravid,   R.,   Rodríguez-­‐Antigüedad,   A.,   Sánchez-­‐Gómez,   M.   &   Domercq,   M.   P2X7   receptor   blockade   prevents   ATP   excitotoxicity   in   oligodendrocytes   and   ameliorates   experimental   autoimmune   encephalomyelitis.   Journal  of  Neuroscience  27,  9525-­‐9533  (2007).   Sharp,  A.  J.,  Polak,  P.  E.,  Simonini,  V.,  Lin,  S.  X.,  Richardson,  J.  C.,  Bongarzone,  E.   R.   &   Feinstein,   D.   L.   P2X7   deficiency   suppresses   development   of   experimental   autoimmune  encephalomyelitis.  Journal  of  Neuroinflammation  5,  (2008).   Adinolfi,   E.,   Pizzirani,   C.,   Idzko,   M.,   Panther,   E.,   Norgauer,   J.,   Di   Virgilio,   F.   &   Ferrari,  D.  P2X7  receptor:  Death  or  life?  Purinergic  Signal  1,  219-­‐227  (2005).   26     42     43     44     45     46     47     48     49     50     51     52     Raczkiew,  J.  &  Leyko,  W.  Adenine  Nucleotide  Content  of  Erythrocytes  of  Multiple   Sclerosis  Patients.  Clinica  Chimica  Acta  14,  408-­‐409  (1966).   Regenold,   W.   T.,   Phatak,   P.,   Makley,   M.   J.,   Stone,   R.   D.   &   Kling,   M.   A.   Cerebrospinal   fluid   evidence   of   increased   extra-­‐mitochondrial   glucose   metabolism   implicates   mitochondrial   dysfunction   in   multiple   sclerosis   disease   progression.  Journal  of  the  Neurological  Sciences  275,  106-­‐112  (2008).   Voskuhl,  R.  &  Gold,  S.  Sex-­‐related  factors  in  multiple  sclerosis  susceptibility  and   progression.  Nature  Reviews  Neurology  8,  255-­‐263  (2012).   Papenfuss,  T.  L.,  Rogers,  C.  J.,  Gienapp,  I.,  Yurrita,  M.,  McClain,  M.,  Damico,  N.,   Valo,  J.,  Song,  F.  &  Whitacre,  C.  C.  Sex  differences  in  experimental  autoimmune   encephalomyelitis  in  multiple  murine  strains.  Journal  of  Neuroimmunology  150,   59-­‐69  (2004).   Confavreux,  C.,  Vukusic,  S.  &  Adeleine,  P.  Early  clinical  predictors  and  progression   of  irreversible  disability  in  multiple  sclerosis:  an  amnesic  process.  Brain  126,  770-­‐ 782  (2003).   Koch-­‐Henriksen,   N.   &   Sorensen,   P.   The   changing   demographic   pattern   of   multiple  sclerosis  epidemiology.  Lancet  Neurology  9,  520-­‐532  (2010).   Orton,   S.   M.,   Herrera,   B.   M.,   Yee,   I.   M.,   Valdar,   W.,   Ramagopalan,   S.   V.,   Sadovnick,   A.   D.   &   Ebers,   G.   C.   Sex   ratio   of   multiple   sclerosis   in   Canada:   a   longitudinal  study.  Lancet  Neurology  5,  932-­‐936  (2006).   Confavreux,   C.,   Hutchinson,   M.,   Hours,   M.   M.,   Cortinovis-­‐Tourniaire,   P.   &   Moreau,  T.  Rate  of  pregnancy-­‐related  relapse  in  multiple  sclerosis.  New  England   Journal  of  Medicine  339,  285-­‐291  (1998).   Vukusic,   S.,   Ionescu,   I.,   El-­‐Etr,   M.,   Schumacher,   M.,   Baulieu,   E.   E.,   Cornu,   C.   &   Confavreux,   C.   The   Prevention   of   Post-­‐Partum   Relapses   with   Progestin   and   Estradiol   in   Multiple   Sclerosis   (POPART'MUS)   trial:   Rationale,   objectives   and   state  of  advancement.  Journal  of  the  Neurological  Sciences  286,  114-­‐118  (2009).   Jansson,  L.,  Olsson,  T.  &  Holmdahl,  R.  Estrogen  induces  a  potent  suppression  of   experimental   autoimmune   encephalomyelitis   and   collagen-­‐induced   arthritis   in   mice.  Journal  of  Neuroimmunology  53,  203-­‐207  (1994).   Palaszynski,   K.   M.,   Liu,   H.,   Loo,   K.   K.   &   Voskuhl,   R.   R.   Estriol   treatment   ameliorates  disease  in  males  with  experimental  autoimmune  encephalomyelitis:   27     53     54     55     56     57     58     59     60     61     62     63     implications   for   multiple   sclerosis.   Journal   of   Neuroimmunology   149,   84-­‐89   (2004).   Sicotte,  N.  L.,  Liva,  S.  M.,  Klutch,  R.,  Pfeiffer,  P.,  Bouvier,  S.,  Odesa,  S.,  Wu,  T.  C.  &   Voskuhl,   R.   R.   Treatment   of   multiple   sclerosis   with   the   pregnancy   hormone   estriol.  Annals  of  Neurology  52,  421-­‐428  (2002).   http://clinicaltrials.gov/ct2/show/NCT00451204   Medicine,  2012).   (US   National   Library   of   http://clinicaltrials.gov/show/NCT01466114  (US  National  Library  of  Medicine).   Subasinghe,   W.   &   Spence,   D.   M.   Simultaneous   determination   of   cell   aging   and   ATP  release  from  erythrocytes  and  its  implications  in  type  2  diabetes.  Analytica   Chimica  Acta  618,  227-­‐233  (2008).   Jaubert,   A.   M.,   Mehebik-­‐Mojaat,   N.,   Lacasa,   D.,   Sabourault,   D.,   Giudicelli,   Y.   &   Ribière,   C.   Nongenomic   estrogen   effects   on   nitric   oxide   synthase   activity   in   rat   adipocytes.  Endocrinology  148,  2444-­‐2452  (2007).   Scott,   P.-­‐A.,   Tremblay,   A.,   Brochu,   M.   &   St-­‐Louis,   J.   Vasorelaxant   action   of   17-­‐ beta-­‐estradiol  in  rat  uterine  arteries:  Role  of  nitric  oxide  synthases  and  estrogen   receptors.  American  Journal  of  Physiology  293,  H3713-­‐H3719  (2007).   Dore-­‐Duffy,   P.,   Catalanotto,   F.,   Donaldson,   J.   O.,   Ostrom,   K.   M.   &   Testa,   M.   A.   Zinc  in  Multiple-­‐Sclerosis.  Annals  of  Neurology  14,  450-­‐454  (1983).   Henry,  J.  P.,  Williamson,  D.  M.,  Schiffer,  R.,  Wagner,  L.,  Shire,  J.  &  Garabedian,  M.   Investigation  of  a  cluster  of  multiple  sclerosis  in  two  elementary  school  cohorts.   Journal  of  Environmental  Health  69,  34-­‐38  (2007).   Schiffer,   R.   B.,   McDermott,   M.   P.   &   Copley,   C.   A   multiple   sclerosis   cluster   associated   with   a   small,   north-­‐central   Illinois   community.   Archives   of   Environmental  Health  56,  389-­‐395  (2001).   Stein,   E.   C.,   Schiffer,   R.   B.,   Hall,   W.   J.   &   Young,   N.   Multiple   sclerosis   and   the   workplace   -­‐   Report   of   an   industry-­‐based   cluster.   Neurology   37,   1672-­‐1677   (1987).   Goodin,  D.  S.,  Frohman,  E.  M.,  Garmany,  G.  P.,  Halper,  J.,  Likosky,  W.  H.,  Lublin,  F.   D.,   Silberberg,   D.   H.,   Stuart,   W.   H.   &   van   den   Noort,   S.   Disease   modifying   therapies   in   multiple   sclerosis   -­‐   Report   of   the   Therapeutics   and   Technology   Assessment  Subcommittee  of  the  American  Academy  of  Neurology  and  the  MS   Council  for  Clinical  Practice  Guidelines.  Neurology  58,  169-­‐178  (2002).   28     64     65     66     67     68     Marriott,   J.,   Miyasaki,   J.,   Gronseth,   G.   &   O'Connor,   P.   Evidence   Report:   The   efficacy   and   safety   of   mitoxantrone   (Novantrone)   in   the   treatment   of   multiple   sclerosis  Report  of  the  Therapeutics  and  Technology  Assessment  Subcommittee   of  the  American  Academy  of  Neurology.  Neurology  74,  1463-­‐1470  (2010).   Miller,  D.  H.,  Khan,  O.  A.,  Sheremata,  W.  A.,  Blumhardt,  L.  D.,  Rice,  G.  P.,  Libonati,   M.   A.,   Willmer-­‐Hulme,   A.   J.,   Dalton,   C.   M.,   Miszkiel,   K.   A.   &   O'Connor,   P.   W.   A   controlled   trial   of   natalizumab   for   relapsing   multiple   sclerosis.   New   England   Journal  of  Medicine  348,  15-­‐23,  (2003).   Kleinschmidt-­‐DeMasters,   B.   &   Tyler,   K.   Progressive   multifocal   leukoencephalopathy,  natalizumab,  and  multiple  sclerosis  -­‐  Reply.  New  England   Journal  of  Medicine  353,  1745-­‐1745  (2005).   Sprague,   R.   S.,   Ellsworth,   M.   L.,   Stephenson,   A.   H.,   Kleinhenz,   M.   E.   &   Lonigro,   A.   J.  Deformation-­‐induced  ATP  release  from  red  blood  cells  requires  CFTR  activity.   American  Journal  of  Physiology  275,  H1726-­‐1732  (1998).   Sprague,  R.  S.,  Ellsworth,  M.  L.,  Stephenson,  A.  H.  &  Lonigro,  A.  J.  ATP:  The  red   blood   cell   link   to   NO   and   local   control   of   the   pulmonary   circulation.   American   Journal   of   Physiology-­‐Heart   and   Circulatory   Physiology   271,   H2717-­‐H2722   (1996).   29   Chapter  2  –  The  Effect  on  Estrogen  on  Endothelial  Nitric  Oxide  Stimulation  via  Red   Blood  Cell  Derived  Adenosine  Triphosphate       2.1  Red  Blood  Cell  Derived  ATP  and  NO  Stimulation       2.1.1  Red  Blood  Cell  Glycolysis       Human   blood   is   made   up   of   plasma   and   three   different   cell   types:     leukocytes,   also   known   as   white   blood   cells,   platelets   and   erythrocytes,   also   known   as   red   blood   cells   (RBCs).    RBCs  are  the  most  common  type  of  blood  cell  and  account  for  35-­‐45%  of  total   blood   volume.     The   cells   are   produced   in   bone   marrow   through   the   process   of   erythropoiesis  and  have  an  average  life  span  of  120  days.    Oxygen  from  the  lungs  binds   1 to   hemoglobin   in   the   RBC   and   is   delivered   throughout   the   body.     This   was   originally   thought  to  be  the  sole  purpose  of  these  cells;  however,  in  1976,  Dean  et  al.  found  that   2 RBCs   contain   large   pools   of   adenosine   triphosphate   (ATP).     In   1992,   Bergfeld   et   al.   3 showed  that  the  ATP  is  released  from  RBCs  in  response  to  hypoxia,  and  Ellsworth  et  al.   4 later  suggested  this  mechanism  is  a  contributor  to  the  regulation  of  vascular  tone.         Stemming  from  these  discoveries,  in  1996,  Sprague  et  al.  found  that  RBCs  release  ATP  in   response   to   mechanical   deformation,   such   as   that   experienced   by   the   cells   as   they   travel   through   the   vascular   system   and   are   exposed   to   shear   stress.     Additionally,   it   was   shown   that   the   ATP   released   could   then   stimulate   nitric   oxide   synthesis   in   other   cell   5 types.    RBCs  are  anucleated  cells,  containing  neither  a  nucleus  nor  cellular  organelles.       30   This   allows   for   their   small   disk   shape.     The   lack   of   mitochondria   causes   the   RBC   to   produce  ATP  through  the  process  of  glycolysis,  making  glucose  the  only  energy  source   1 for  the  cell.     The  first  step  in  glycolysis  is  glucose  uptake  into  the  RBC.    This  occurs  through  facilitated   diffusion.     Glucose   is   a   polar   molecule   and   therefore   does   not   diffuse   through   the   hydrophobic   membranes   of   RBCs.     As   a   result,   it   is   necessary   for   carrier   molecules   to   transport   glucose   across   the   membrane   and   into   the   cell.     The   glucose   transport   proteins   (GLUT)   were   numbered   in   chronological   order   as   they   were   discovered,   with   the   major   protein   on   RBCs   being   the   first   of   the   13   discovered   thus   far. 6-­‐8     GLUT1   is   10-­‐ 20%  of  the  total  RBC  membrane  protein,  but  is  also  found  in  astrocytes,  cardiac  muscle   and  blood-­‐tissue  barriers,  such  as  the  BBB.    GLUT1  spans  the  plasma  membrane  with  12   transmembrane   domains.     These   hydrophobic   domains   are   α-­‐helical   in   nature   and   in   8 total  are  comprised  of  nearly  500  amino  acids.     In   a   healthy   individual,   RBCs,   in   total,   consume   around   20   grams   of   glucose   per   day,   resulting  in  about  10%  of  the  total  glucose  metabolism  in  the  body.    Once  the  glucose   has   entered   the   cell,   90%   is   metabolized   through   the   process   of   glycolysis.     Glycolysis   literally  translates  to  “carbohydrate  splitting”  and  is  the  way  in  which  glucose  is  broken   down  into  lactate.    This  10-­‐step  process  can  be  seen  in  figure  2.1.    Glycolysis  occurs  in   two  phases,  the  preparatory  phase  and  the  payoff  phase.    During  the  preparatory  phase,   there   is   an   investment   of   two   ATP   molecules   for   every   glucose   molecule.     In   the   pay   off       31   Figure  2.1  –  Glycolysis  –  This  process,  which  occurs  in  RBCs,  is  the  only  way  for  the  cells   to   metabolize   glucose.     Glycolysis   occurs   in   two   phases.     In   the   preparatory   phase,   on   the  left,  two  ATP  molecules  are  consumed.    During  the  payoff  phase,  on  the  right,  four   ATP  molecules  are  produced,  resulting  in  a  net  gain  of  two  ATP  molecules  per  glucose   molecule  consumed.             32   phase,  four  molecules  of  adenosine  diphosphate  (ADP)  are  converted  to  ATP,  resulting   in   a   net   gain   of   two   ATP   for   the   breakdown   of   each   glucose   molecule.     The   ATP   conserves   some   of   the   energy   from   the   glucose   molecule.     This   method   of   glucose   metabolism   occurs   under   anaerobic   conditions,   despite   the   RBC   carrying   oxygen,   and   9 results  in  far  less  ATP  than  glucose  breakdown  until  aerobic  conditions.     2.1.2  ATP  Release  and  NO  Stimulation       As   discussed   previously,   ATP   is   released   from   the   RBC   in   response   to   several   different   4 3 stimuli   in   the   local   environment,   including   lowering   of   pH,   hypoxia,   changes   in   10 osmotic   pressure,   and   mechanical   deformation. 5,11     ATP   is   believed   to   cross   the   plasma   membrane   with   the   aid   of   an   ion   channel,   assumed   to   be   the   cystic   fibrosis   transmembrane   conductance   regulator   (CFTR). 11,12     It   has   not   yet   been   determined   if   ATP  is  released  through  the  channel  of  CFTR  itself,  or  if  CFTR  is  regulating  another  ion   channel   in   the   cell   membrane   through   which   ATP   is   released.     Abnormal   ATP   release   11 13 from   RBCs   has   been   associated   with   cystic   fibrosis,   pulmonary   hypertension,   14,15 diabetes, 16  and  multiple  sclerosis  (MS).         While   at   Wayne   State   University,   the   Spence   group   preformed   a   small   scale   study   using   RBCs  from  18  MS  patients  and  11  controls.    The  RBC  ATP  release  was  measured  using  a   17 capillary-­‐based  flow  system  described  previously.    Briefly,  as  RBCs  pass  through  50  µm     33   internal   diameter   tubing,   they   experience   shear   stress   and   release   ATP,   which   reacts   with  a  luciferin/luciferase  mixture.    The  resulting  chemiluminescence  from  the  samples   and  standards  can  be  measured  with  a  photomultiplier  tube  (PMT),  and  the  amount  of   ATP  can  be  quantified.    The  resulting  data  are  shown  in  figure  2.2.    The  average  release   of  ATP  from  the  RBCs  of  MS  patients  was  375  ±  51  nM,  almost  three  times  higher  than   16 the  138  ±  21  nM  from  the  RBCs  of  healthy  controls.     When   released   from   the   RBC,   ATP   can   stimulate   nitric   oxide   (NO)   production   in   endothelial   cells   through   binding   to   the   purinergic   receptor,   P2Y.     This   stimulates   endothelial   nitric   oxide   synthase   (eNOS)   which   converts   L-­‐arginine   to   L-­‐citrulline   with   NO   as   the   byproduct,   as   shown   in   figure   2.3.     In   1987,   NO   was   identified   as   the   endothelium-­‐derived  relaxing  factor  and  it  is  now  well  established  as  a  determinant  in   the  control  of  vasodilation. 5,18    Because  of  this,  NO  production  is  often  thought  of  as  a   positive   event.     However,   in   many   autoimmune   disease,   NO   is   believed   to   contribute   to   19 immune  response  and  tissue  destruction.    NO  and  its  metabolites,  such  as  nitrite  and   nitrate,  are  found  at  higher  than  normal  levels  in  the  cerebral  spinal  fluid  (CSF)  and  urine   of  MS  patients,  as  compared  to  healthy  controls.    In  addition,  NO  and  its  metabolites  are   20 found  in  the  lesions  formed  from  the  characteristic  demyelination.             34   500 Control RBCs MS RBCs ATP Release (nM) 400 300 200 100 0   Figure  2.2  –  ATP  Release  From  MS  RBCs  (From  Letourneau  et  al.)  –   ATP   release   was   measured  from  RBCs  subjected  to  flow.    The  average  ATP  release  from  the  healthy  RBCs   was   138   ±   21   nM.     An   increase   to   375   ±   51   nM   was   seen   for   RBCs   obtained   from   MS   patients.    The  error  is  reported  as  standard  error  of  the  mean  for  11  controls  and  18  MS   samples.    The  values  are  statistically  different  at  a  value  of  p  <  0.001.   16   35       Figure   2.3   –   Proposed   mechanism   of   ATP   release   from   RBCs   and   subsequent   NO   49 production   –   It   has   been   proposed   that   G-­‐protein   coupled   receptor   (GPCR),   cyclic   50 adenosine   monophosphate   (cAMP),   and   the   cystic   fibrosis   transmembrane   11   conductance   regulation   (CFTR)   protein are   all   required   for   the   release   of   ATP   by   mechanical   deformation,   though   how   the   ATP   leaves   the   cell   remains   in   question.     In   this  proposed  mechanism,  the  binding  of  ATP  to  the  P2Y  receptor  on  the  endothelial  cell   5 results  in  the  activation  NOS  and  the  production  of  NO.             36   2.1.3  Estrogen  and  MS       As  previously  discussed,  reports  in  the  literature  have  described  a  decrease  in  relapses   21 during   the   second   and   third   trimesters   of   pregnancy   in   women   with   MS.     This   decrease  in  relapse  rate  correlates  with  the  increase  of  estrogens  in  pregnant  women.   The   mechanism   for   this   is   not   completely   understood,   but   as   a   result   of   these   observations,   estrogens,   specifically   estradiol   and   estriol   have   been   extensively   studied.     Voskuhl   et   al.   have   reported   that   increased   estrogen   levels   in   murine   EAE,   due   to   either   pregnancy  or  subcutaneous  estrogen-­‐containing  pellets,  have  a  protective  effect  on  the   22 severity   of   the   EAE.     These   results   are   intriguing   in   light   of   reports   that   substances   such  as  estradiol  and  estriol  have  the  ability  to  stimulate  NO  production  in  certain  cell   types.    It  would  seem  that  an  increase  in  estrogens  would  be  detrimental,  rather  than   protective,  to  MS-­‐like  complications. 23,24         Additionally,   as   mentioned   above,   CFTR   is   necessary   for   the   deformation-­‐induced   release   of   ATP   from   the   RBC.     There   have   been   several   reports   of   the   ability   of   estrogen   25 to   inhibit   CFTR   in   various   cell   types,   including   pancreatic   epithelial   cells   and   lung   26 epithelial  cells.    Singh  et  al.  showed  that  estrogens  interact  directly  with  CFTR  and  the   27 acute  effect  of  this  interaction  is  the  inhibition  of  CFTR.    Since  CFTR  is  required  for  ATP   release,   RBC-­‐derived   ATP   is   elevated   in   MS   patients,   and   estrogen   has   been   shown   to     37   ameliorate   the   disease,   it   is   hypothesized   that   estrogen   may   have   its   effect   on   MS   through  this  inhibition  pathway.     In  order  to  investigate  the  interaction  between  the  RBCs,  releasing  ATP,  and  endothelial   cells,   producing   NO,   these   cells   needed   to   be   in   close   contact   to   each   other   as   they   would  be  in  vivo.    A  microfluidic  device  has  previously  been  used  in  the  Spence  group   28 for   this   purpose.     This   device   uses   microfluidic   channels,   through   which   RBCs   are   pumped,   below   a   membrane   that   has   endothelial   cells   immobilized   on   it.     Having   the   RBCs   able   to   interact   with   an   endothelial   layer,   as   it   would   in   vivo,   enables   for   more   accurate  measurements  that  take  into  account  cell-­‐to-­‐cell  communication.    This  device   will  be  further  described  in  the  next  section.         2.2  Microfluidic  Devices             Microfluidics  is  the  science  of  systems  able  to  process  and  manipulate  low  volumes  of   fluids   using   devices   having   channels   with   dimensions   of   up   to   hundreds   micrometers.     Many   devices   handle   volumes   in   the   low   microliter   range,   with   some   able   to   handle   volumes   as   low   as   the   single   digit   nanoliter   range.     The   field   of   microfluidics   began   with   applications   in   analysis,   however,   over   time,   the   technology   has   spread   to   areas   of   29 synthesis,  biophysics  and  fuel  cells  as  well.         Although   microfluidic   devices   can   be   fabricated   from   many   different   materials,   one   of   the  most  common  and  well  established  for  rapid  prototyping  is  poly(dimethylsiloxane)     38   (PDMS).     Through   the   use   of   soft   lithography,   these   devices   can   be   made   quickly   and   repeatedly   through   the   use   of   a   master.     Typically   made   through   photolithography,   a   30 master  has  raised  features  that  will  create  recesses  in  the  PDMS  for  the  channels.    The   use  of  a  variety  of  software  programs  to  create  the  mask  for  the  master  results  in  the   ability   to   make   microfluidic   devices   in   a   variety   of   different   configurations.     To   create   the  master,  photoresist  is  coated  onto  a  silicon  wafer.    The  mask  is  then  placed  over  the   wafer   and   exposed   to   ultraviolet   light.     The   photoresist   polymerizes,   creating   the   31 features.    The  excess  is  then  washed  away.    This  process  is  shown  in  figure  2.4.         In  2007,  Genes  et  al.  reported  an  in  vitro  model  of  the  microvasculature  using  a  PDMS   28 microfluidic   device.     By   using   a   two-­‐layer   device   separated   by   a   membrane,   RBCs   flowing  in  a  channel  on  the  bottom  can  be  in  contact  with  endothelial  cells  cultured  on   the   membrane   in   wells   of   the   top   layer.     This   enables   the   device   to   closely   mimic   the   in   vivo   conditions   where   cells   are   able   to   affect   one   another   through   cell   signaling   methods.     Although   devices   enabling   cell-­‐to-­‐cell   communication   had   been   previously   32,33 reported,  this  device  was  the  first  to  incorporate  separated  components  of  blood   circulation   on   a   microfluidic   device.     This   device   enabled   ATP   released   from   RBCs   as   a   result   of   shear   stress   to   interact   with   the   endothelial   cells   immobilized   above   the   membrane.    The  resultant  NO  production  can  be  measured  with  a  fluorescent  probe.    By   increasing  the  ATP  release  from  RBCs  by  using  a  chemical  stimulus,  the  NO  production   was  seen  to  increase  as  well.    Therefore,  the  device  was  able  to  show  production  of  NO,       39       Figure   2.4   –   Photolithography   –   In   order   to   perform   rapid   fabrication   using   PDMS,   a   reusable   master   is   needed.     This   can   be   prepared   through   photolithography.     In   this   process,   a   clean   silicon   wafer   is   coated   with   SU-­‐8   photoresist.     After   pre-­‐baking   the   wafer,   a   mask   is   placed   over   the   photoresist   with   the   desired   features.     The   wafer   is   then   exposed   to   UV   light.     After   a   post-­‐bake,   the   wafer   is   soaked   in   developer   to   removed   unpolymerized   photoresist.     The   result   is   a   wafer   with   raised   features   that   can   be  repeatedly  used  as  a  mold  for  PDMS.           40   in   this   case,   occurs   as   a   result   of   cell-­‐to-­‐cell   communication   between   the   RBCs   and   endothelial  cells. 28,34       Knowing   that   the   RBCs   of   MS   patients   release   an   increased   amount   of   ATP,   and   the   protective   influence   of   estrogens   in   the   disease,   the   first   step   in   this   research   was   to   investigate  the  effect  of  estrogens,  specifically  estradiol  and  estriol,  on  the  ATP  release   from   healthy   RBCs.     The   previously   mentioned   microfluidic   device   was   used   to   determine   the   effect   of   the   estrogen-­‐induced   decrease   in   RBC-­‐derived   ATP   on   the   NO   production  of  endothelial  cells.    The  proximity  of  the  RBCs  to  the  endothelial  cells  in  a   microvascular   mimic   is   key   to   the   understanding   the   effects   of   estrogen   in   the   blood   stream.     2.3  Experimental       2.3.1  Collection  and  Purification  of  Rabbit  RBCs       RBCs   used   in   these   studies   were   obtained   from   male   New   Zealand   white   rabbits   between   2.0   and   2.5   kg.     The   rabbits   were   anesthetized   using   ketamine,   8   mg/kg,   intramuscular,  and  xylazine,  1  mg/kg,  intramuscular,  followed  by  pentobarbital  sodium,   15  mg/kg,  intravenously.    A  cannula  was  placed  in  the  rabbit  trachea  for  ventilation  with   room   air   at   a   rate   of   20   breaths/min.       Prior   to   exsanguination,   500   units   of   heparin   were   administered   using   a   catheter   placed   in   the   carotid   artery.     The   exsanguination   was   performed   through   the   same   catheter,   and   approximately   80   mL   of   whole   blood   were  collected  from  each  rabbit.         41     The  whole  blood  was  then  centrifuged  at  500g  for  10  min  at  25°C,  and  the  plasma  was   removed   and   retained   for   other   experimentation.     The   remaining   RBCs   were   then   resuspended   and   washed   three   times   in   physiological   salt   solution   (PSS)   containing,   in   mM,   4.7   KCl   (Fisher   Scientific,   Fair   Lawn,   NJ),   2.0   CaCl2   (Fisher   Scientific),   140.5   NaCl   (Columbus   Chemical   Industries,   Columbus,   WI),   12   MgSO4   (Fisher   Scientific),   21.0   tris(hydroxymethylaminomethane)   (Invitrogen,   Carlsbad,   CA),   5.6   glucose   (Sigma,   St.   Louis,   MO),   and   5%   bovine   serum   solution   (Sigma)   at   a   final   pH   of   7.4,   adjusted   with   hydrochloric   acid.     The   hematocrit   of   the   RBCs   was   measured   using   a   CritSpin©   analyzer.     2.3.2  Preparation  of  Regents       Purified  water  with  18.2  MΩ  resistance  was  used  for  all  experiments.  Estradiol  (Sigma)   stock   solutions   were   prepared   by   first   dissolving   2   mg   in   1   mL   dimethyl   sulfoxide   (DMSO)  (EMD  Chemicals,  Gibbstown,  NJ)  to  make  a  7.3  mM  solution.    Two  1:100  serial   dilutions  were  performed  in  PSS  to  make  73  µM  and  0.73  µM  solutions.    Estriol  (Sigma)   stock  solutions  were  prepared  by  dissolving  2  mg  in  1  mL  DMSO,  resulting  in  a  3.5  mM   solution.    Two  1:100  serial  dilutions  were  performed  in  PSS  to  make  35  µM  and  0.35  µM   2+ solutions.     Zn   stock   solution   was   prepared   by   dissolving   Zinc   (II)   chloride   (Jade   Scientific,  Canton,  MI)  in  purified  water  and  diluted  to  400  nM.           42   Crude   C-­‐peptide   (Genscript,   Piscataway,   NJ)   was   purified   using   reverse-­‐phase   high   performance   liquid   chromatography   (RP-­‐HPLC)   and   dried.     The   purified   C-­‐peptide   was   2+ dissolved   in   purified   water   and   diluted   to   a   400   nM   working   solution.     When   Zn ,   2+ bound  to  C-­‐peptide  was  added  to  samples,  the  Zn  and  C-­‐peptide  solutions  were  mixed   in  water  before  the  addition  of  any  other  components  or  buffers.    5  mL  samples  of  7%   RBCs   were   prepared   with   varying   concentrations   of   estradiol   or   estriol   by   adding   the   hormone  to  the  appropriate  volume  of  PSS,  followed  by  the  addition  of  the  RBCs  for  a   30  minute  incubation.    For  some  experiments,  these  samples  were  then  centrifuged  at   500g   at   25°C   for   5   minutes,   then   washed   twice   by   resuspending   the   cells   in   PSS   and   centrifuging  at  500g  at  25°C  for  3  minutes.       2.3.3  Preparation  of  Microfluidic  Device       The   microfluidic   device   used   in   these   studies   was   comprised   of   two   individual   PDMS   (Dow   Corning,   Midland,   MI)   layers   sealed   around   a   0.6   micrometer   pore   size   polycarbonate  membrane  (GE  Water  &  Process  Technologies,  Feasterville-­‐Trevose,  PA).     The  channel  layer  was  fabricated  by  pouring  a  degassed,  20:1  mixture  of  Sylguard  184   PDMS   elastomer   and   curing   agent   onto   a   silicon   master   with   12   raised-­‐feature   channels   wth  the  dimensions:    200  µm  wide  x  100  µm  deep  x  3  cm  long.    After  curing  for  12  min   at   75°C,   a   degassed   5:1   mixture   of   the   elastomer   and   curing   agent   was   poured   over   the   master.    This  layer  was  then  cured  an  additional  12  min  at  75°C  prior  to  removal  from   the   master.     Inlet   holes   were   punched   at   one   end   of   each   channel   using   a   20   gauge   luer     43   stub  adapter,  and  the  PDMS  at  other  end  of  the  channel  was  cut  and  removed  to  allow   for  the  exit  of  waste.         A  degassed  20:1  PDMS  mixture  was  also  poured  onto  an  unpatterned  silicon  wafer  and   cured  for  15  minutes  at  75°C.    This  can  be  seen  in  figure  2.5.    After  removal  from  the   th wafer,   36   holes   were   punched   in   a   3   x   12   array,   using   an   1/8   inch   hole   punch.     The   layers   were   then   aligned   around   the   polycarbonate   membrane,   with   three   wells   lined   up   on   each   channel,   and   cured   for   an   additional   45   min   to   seal   them.     This   design   is   34,35 similar  to  that  described  previously  by  the  Spence  group  and  can  be  seen  in  figure   2.6.     2.3.4  Chemiluminesce  Detection  of  ATP  Release  from  RBCs       In   these   experiments,   the   ATP   release   from   the   RBC   was   measured   in   a   static   system.     This   was   performed   using   the   chemiluminescent   reaction   of   ATP   and   a   solution   of   luciferin/luciferase   was   prepared   by   dissolving   2   mg   luciferin   (Gold   Biotechology,   St.   Louis,  MO)  in  5  mL  DDW,  and  transferring  this  to  a  vial  containing  100  mg  of  luciferase   (Sigma).     To   detect   the   chemiluminescence,   a   photomultiplier   tube   (PMT)   was   used   as   a   transducer.     100   µL   of   sample,   prepared   as   described   in   section   2.4.2,   and   100   µL   of   luciferin/luciferase   were   pipetted   into   a   plastic   cuvette   and   mixed.     The   measurement   was   taken   15   seconds   after   the   addition   of   the   luciferin/luciferase   mixture   by   placing   the  cuvette  over  the  PMT.    All  samples  were  measured  in  triplicate.    A  diagram  of  this   set  up  can  be  seen  in  figure  2.7.         44     Figure   2.5   –   Soft   Lithography   –   Through   this   process,   the   layers   of   the   microfluidic   device  were  created.    The  polymer  and  curing  agent  were  mixed  in  20:1  and  5:1  ratios  in   separate  plastic  cups  and  degassed  by  vacuum.    The  20:1  mixture  was  poured  over  the   channel  portion  of  the  master  and  baked  at  75°C  for  12  minutes.    The  5:1  mixture  was   then   pour   wafer   the   entire   master   and   bake   for   another   12   minutes   at   the   same   temperature.    The  PDMS  was  then  removed  from  the  master.    The  second  layer  of  the   microfluidic   device   was   prepared   using   the   same   steps   with   an   unpatterned   silicon   master.           45     Figure   2.6   –   Microfluidic   Device   –   A)  The  microfluidic  device  layers  are  diagramed.    A   polycarbonate  membrane  separated  the  bottom  layer,  containing  the  channels,  and  the   top  layer,  patterned  with  wells.    B)  A  photograph  of  a  completed  microfluidic  device.    C)   A  diagram  of  the  cross  section  of  a  well.  bPAECs  are  adhered  to  the  membrane  in  the   well.     The   porous   membrane   separates   the   bPAECs   from   the   channel   through   which   the   RBCs  are  pumped.           46     Figure  2.7  –  Measurement  of  ATP  Release  –  Estradiol  (E2)  or  estriol  (E3)  were  added  to   2+ PSS,  immediately  followed  by  RBCs  to  make  a  7%  RBC  solution.    When  Zn  bound  to  C-­‐ 2+ peptide  was  used,  the  Zn  and  C-­‐peptide  were  added  to  the  vial  first  and  allowed  to   incubate  for  2-­‐3  minutes  before  the  addition  of  PSS.    After  a  two  hour  incubation,  a  200   μL  sample  was  transferred  into  a  cuvette.    100  μL  of  the  luciferin/luciferase  solution  was   added  and  the  cuvette  was  lightly  shaken.    The  cuvette  was  placed  in  a  dark  box  over  a   PMT  and  at  15  seconds  the  luminescence  was  measured.         47   2.3.5  Adhering  Cells  to  a  Microfluidic  Device  and  Fluorescence  Determination  of  NO       For   the   purpose   of   these   studies,   a   confluent   layer   of   bovine   pulmonary   artery   endothelial   cells   (bPAECs)   was   required   in   each   well.     For   the   cells   to   adhere   to   the   membrane,  10  µL  of  a  100  µg/mL  fibronectin  solution  were  pipetted  into  each  well  and   allowed   to   dry.     The   device   was   then   sterilized   under   ultraviolet   (UV)   light   for   15   minutes.   The   bPAECs   were   harvested   from   a   T-­‐75   culture   flask   containing   a   confluent   monolayer  of  these  cells.    The  flask  was  first  washed  with  7  mL  of  4-­‐(2-­‐hydroxyethyl)-­‐1-­‐ piperazineethanesufonic   acid   (HEPES)   solution   for   2   minutes.     After   aspirating   off   the   HEPES,   5   mL   of   a   trypsin   solution   was   allowed   to   incubate   on   the   cells   for   2   minutes,   followed   by   the   addition   of   9   mL   of   trypsin   neutralizing   solution.     Pipetting   and   scraping   were   used   to   fully   remove   the   cells   from   the   surface   of   the   flask   so   they   were   suspended  in  solution.         The   suspension   was   removed   to   a   15   mL   centrifuge   tube   and   was   then   centrifuged   at   1500g  for  5  minutes  at  25°C.    Following  centrifugation,  the  supernatant  was  aspirated   off,  leaving  a  pellet  of  cells  that  were  then  resuspended  in  1  mL  of  endothelial  growth   media   to   ensure   a   homogeneous   solution.     Next,   10   µL   of   this   solution   were   then   pipetted  into  each  well  of  the  device  and  allowed  to  incubate  for  1  hour  at  37°C  and  5%   CO2.     After   incubation,   the   media   solution   on   the   cells   was   changed   by   carefully   removing   the   solution   and   adding   10   µL   of   media   to   each   well.     This   was   followed   by   an   additional  2  hour  incubation.       48   After   the   cells   had   become   immobilized   on   the   membrane,   the   media   was   removed   from  the  wells  and  replaced  with  10  µL  of  a  5  mM  L-­‐arginine  solution  in  Hank’s  Balanced   Salt  Solution  (HBSS)  to  ensure  sufficient  substrate  for  the  production  of  NO.    Following  a   30   minute   incubation   at   37°C   and   5%   CO2,   the   solution   was   removed   and   replaced   with   10   µL   of   100   µM   4-­‐amino-­‐5-­‐methylamino-­‐2’,7’-­‐difluorofluoroescien   diacetate   (DAF-­‐FM   DA),  a  fluorescent  probe  for  detecting  intracellular  NO,  pictured  in  figure  2.8,  for  a  30   minute   incubation.     DAF-­‐FM   DA   is   able   to   cross   the   cell   membrane.     Once   inside,   esterases  transform  the  probe  to  DAF-­‐FM,  which  loses  its  cell  permeability.    The  DAF-­‐FM   then   fluorescently   reacts   with   NO.     From   the   time   DAF-­‐FM   DA   was   introduced   to   the   cells   until   the   end   of   the   experiment,   the   device   was   kept   in   the   dark.     After   the   30   minute  incubation,  the  excess  DAF-­‐FM  DA  solution  was  removed  and  replaced  with  10   µL  of  HBSS  per  well.     In  these  experiments,  the  device  was  used  to  monitor  the  intracellular  NO  production  of   the  bPAECs  as  ATP  diffused  through  the  membrane  from  samples  with  RBCs  and  varying   concentrations   of   estradiol   that   were   pumped   at   the   rate   of   0.1   µL/min   through   the   underlying   microfluidic   channels   for   30   minutes.     Each   well   was   imaged   immediately   before   and   after   using   an   Olympus   MVX   fluorescence   macroscope,   fitted   with   a   mercury-­‐arc   lamp   and   a   fluoroscien   isothiocyanate   (FITC)   filter   cube   having   excitation   and  emission  wavelengths  of  470  and  525  nm,  respectively.    This  set  up  can  be  seen  in   figure  2.9.    The  pixel  intensity  of  each  well  was  measured  and  background  subtracted.         49       Figure   2.8   –   DAF-­‐FM   DA   –   DAF-­‐FM   DA   crosses   the   cell   membrane.   Esterases   then   transform   the   probe   to   DAF-­‐FM,   which   loses   its   cell   permeability.     The   DAF-­‐FM   then   reacts   with   NO,   producing   a   benzotiazol   derivative   that   is   fluorescent   at   the   above   excitation  and  emission.     50       Figure  2.9  –  Flowing  RBCs  Through  the  Device  –  Pictured  above,  the  RBC  solutions  were   pumped  at  the  rate  of  0.1  µL/min  through  the  underlying  microfluidic  channels  for  30   minutes.    Each  well  was  imaged  before  and  after  using  an  Olympus  MVX  fluorescence   macroscope,  fitted  with  a  mercury-­‐arc  lamp  and  a  fluoroscien  isothiocyanate  (FITC)  filter   cube  having  excitation  and  emission  wavelengths  of  470  and  525  nm,  respectively.           51   This   value   is   directly   dependent   on   the   concentration   of   NO   produced   within   the   bPAECs.     2.4  Results       In   figure   2.10,   it   is   shown   that   there   is   a   significant   decrease   in   the   ATP   release   from   RBCs  when  the  cells  are  pre-­‐incubated  with  estradiol.    When  compared  to  RBCs  with  the   absence   of   estradiol,   the   RBC-­‐derived   ATP   for   RBCs   treated   with   0.5   μM   estradiol   was   reduced   to   76   ±   7%   the   value   of   the   untreated   cells.     The   amount   of   ATP   released   decreased  with  increasing  concentration  of  estradiol;  for  the  1  µM  and  1.5  µM  estradiol   solutions,  the  RBC-­‐derived  ATP  was  reduced  to  62  ±  7%  and  56  ±  6%,  respectively.         Accounting  for  the  number  of  RBCs  in  the  sample,  the  levels  of  estradiol  originally  tested   were  higher  than  those  found  in  vivo.    Adjusting  for  the  reduced  concentration  of  RBCs   in  the  samples,  the  average  equivalent  levels  of  estradiol  in  healthy  women  are  32  pM   36 normally   and   up   to   90   nM   in   late   pregnancy.     For   estriol   these   values   are   7   nM   37 normally,   30   nM   in   early   pregnancy,   and   up   to   0.5   µM   in   late   pregnancy.   With   this   knowledge,   the   study   was   repeated   using   lower   concentrations   of   estradiol   and   also   with  estriol.         As   shown   in   figure   2.11,   incubating   the   RBCs   with   30   nM   estradiol   reduces   the   chemiluminescence   in   response   to   ATP   release   to   74   ±   4%,   which   is   statistically   equivalent    to    the    decrease    in    the    amount    of    chemiluminescence    seen    with    0.5  µM       52   1.0 * 0.8 * * 0.6 0.4 0.2 0. 5 µ M E2 E2 75 µ M E2 5 µ M 0. R BC s 0.0 1. Normalizalized Chemiluminescence Intensity 1.2   Figure  2.10  –  The  Effect  of  Estradiol  on  RBC  ATP  Release  -­‐   Compared   to   RBCs   with   the   absence   of   estradiol,   the   RBC-­‐derived   ATP   for   RBCs   treated   with   0.5   μM   estradiol   was   reduced   to   76   ±   7%   of   the   value   of   the   untreated   cells.     The   amount   of   ATP   released   decreased  with  increasing  concentration  of  estradiol;  for  the  1  µM  and  1.5  µM  estradiol   solutions,  the  RBC-­‐derived  ATP  was  reduced  to  62  ±  7%  and  56  ±  6%,  respectively.    Error   is   shown   as   standard   deviation   for   N   =   4   rabbits.   The   asterisks   denote   a   statistically   significant  difference  from  untreated  RBCs  at  p  <  0.05.           53   Normalized Chemiluminescence Intensity 1.2 1.0 * 0.8 * 0.6 0.4 0.2 0.0 RBCs 30 nM E2 0.5 µM E2 Figure   2.11   –   Estradiol   (E2)   Decreases   ATP   Release   at   Physiological   Concentrations   –   incubating  the  RBCs  with  30  nM  estradiol  reduces  the  chemiluminescence  in  response   to  ATP  release  to  74  ±  4%,  which  is  statistically  equivalent  to  the  decrease  in  the  amount   of  chemiluminescence  seen  with  0.5  µM  estradiol,  but  at  a  more  physiologically  relevant   concentration   of   the   hormone.     Error   is   standard   deviation   for   N   =   4   rabbits.     The   asterisk   denotes   the   decrease,   as   compared   to   untreated   RBCs,   is   statistically   significant   at  p  <  0.02.     54   estradiol,  but  at  a  more  physiologically  relevant  concentration  of  the  hormone.    In  figure   2.12,   the   data   shows   the   ability   of   estriol   to   decrease   the   chemiluminescence   due   to   ATP  release  to  an  even  greater  degree  at  physiologically  relevant  concentrations.    30  nM   of   estriol   reduced   the   ATP   release   to   70   ±   11%   of   that   of   RBCs   alone.     The   RBCs   incubated   with   0.5   µM   and   1.0   µM   had   their   ATP   release   reduced   to   69   ±   13%   and   62   ±   11%,  respectively.         2+ Additionally,   Zn   bound   to   C-­‐peptide   was   used   as   a   stimulus   of   ATP   release.     The   estrogens   were   then   used   to   attenuate   this   increase.     In   figure   2.13,   the   increase   in   ATP   2+ release  as  a  result  of  the  Zn  bound  to  C-­‐peptide  and  the  subsequent  decrease  with  the   addition  of  both  estrogens  can  be  seen.    It  is  important  to  note  that  the  concentrations   of  estradiol  and  estriol  are  not  the  same  looking  at  the  bars  left  to  right  in  an  attempt  to   2+ compare  the  data  sets.    In  both  sets  of  results  the  Zn  bound  to  C-­‐peptide  resulted  in   an  increase  in  chemiluminescence  to  more  than  double  that  seen  for  RBCs  alone.      For   estradiol,   a   significant   decrease   in   the   ATP   release,   as   compared   to   RBCs   and   10   nM   2+ Zn  bound  to  C-­‐peptide,  to  76  ±  15%  and  61  ±  24%  was  observed  for  the  0.75  µM  and   1.5   µM   samples,   respectively.     For   estriol,   the   statistically   significant   decreases   were   seen  for  the  0.5  µM  and  1.0  µM  samples.    They  were  decreased  to  73  ±  17%  and  61  ±   13%,   respectively.     While   neither   hormone   was   able   to   bring   the   ATP   release   back   to   basal  levels,  it  is  significant  that  decreases  were  seen,  even  when  an  ATP  stimulus  was   present.       55   Normalized Chemiluninescence Intensity 1.2 1.0 * 0.8 * * 0.6 0.4 0.2 0.0 RBCs 30 nM E3 0.5 µM E3 1.0 µM E3 Figure   2.12   –   The   Effect   of   Estriol   (E3)   on   RBC-­‐derived   ATP   –   Estriol   decreases   the   chemiluminescence   due   to   ATP   at   physiologically   relevant   concentrations.     30   nM   of   estriol  reduced  the  ATP  release  to  70  ±  11%  of  that  of  RBCs  alone.    The  RBCs  incubated   with   0.5   µM   and   1.0   µM   had   their   ATP   release   reduced   to   69   ±   13%   and   62   ±   11%,   respectively.     Error   is   shown   as   standard   deviation   for   N   =   4   rabbits,   and   the   asterisk   denotes  a  statistically  significant  decrease  at  p  <  0.05           56       2+ Figure   2.13   –   The   Effect   of   Estrogen   on   ATP   Release   Stimulated   By   Zn   bound   to   C-­‐ peptide  –  Note  that  the  concentrations  of  estradiol  and  estriol  are  not  same  looking  at   the   bars   left   to   right   in   an   attempt   to   compare   the   data   sets.     In   both   sets   of   results   the   2+ Zn   bound   to   C-­‐peptide   resulted   in   an   increase   in   chemiluminescence   to   more   than   double  that  seen  for  RBCs  alone.      For  estradiol  (E2),  a  significant  decrease  in  the  ATP   2+ release,  as  compared  to  RBCs  and  10  nM  Zn -­‐C-­‐peptide,  to  76  ±  15%  and  61  ±  24%  was   observed   for   the   0.75   µM   and   1.5   µM   samples,   respectively.     For   estriol   (E3),   the   statistically  significant  decreases  were  seen  for  the  0.5  µM  and  1.0  µM  samples.    They   were   decreased   to   73   ±   17%   and   61   ±   13%,   respectively.     Error   is   standard   deviation   for   N   =   5   rabbits   for   estradiol   and   N   =   4   rabbits   for   estriol.     The   asterisk   denotes   the   2+ significant  increase  of  ATP  release  with  10  nM  Zn  bound  to  C-­‐peptide  at  p  <  0.0005.   2+ The  pound  sign  denotes  a  decrease  as  compared  to  10  nM  Zn -­‐C-­‐peptide  at  p  <  0.05.     57   In   the   cell-­‐containing   microfluidic   device,   RBC   solutions   containing   RBCs   with   and   without  C-­‐peptide  and  estradiol  are  pumped  through  the  channels  in  the  bottom  layer   of  PDMS.    As  previously  discussed,  the  flow-­‐induced  shear  stress  results  in  ATP  release   from  the  RBCs.    This  ATP  interacts  with  the  bPAECs  lining  the  wells  in  the  top  layer  of   PDMS   resulting   in   NO   production   that   can   be   measured   using   DAF-­‐FM   DA   as   a   fluorescent  probe.    Photos  were  taken  with  a  charge-­‐coupled  device  (CCD)  camera,  such   as  those  seen  in  figure  2.14.    The  pixel  intensity  of  these  images  was  measured  and  then   background   subtracted   to   account   for   differences   in   the   number   of   bPAECs.     The   numbers  were  then  normalized.         The   data   in   figure   2.15   shows   the   normalized,   background   subtracted   emission.     As   shown,  the  emission  decreases  with  RBCs  incubated  in  buffer  containing  estradiol  in  as   low  a  concentration  as  10  nM.    In  this  case,  the  estradiol  is  actually  able  to  reduce  the   NO  production  to  43  ±  0.1%  of  that  of  untreated  RBCs.    However,  the  emission  intensity   then  begins  to  increase  with  again  with  increasing  concentration  of  estradiol,  although  a   significant  increase  in  emission,  in  comparison  to  RBCs  alone,  is  not  measured  until  the   estradiol   levels   reach   approximately   2   µM,   a   value   that   is   physiologically   high,   even   during  pregnancy.    At  this  value,  the  fluorescence  intensity  is  13  ±  4%  higher  than  the   untreated   RBCs.     The   next   highest   concentration,   1.5   µM,   is   not   significantly   different   from  the  untreated  RBCs.     Estrogen   has   been   found   to   have   nongenomic   effects,   increasing   NO   production   from   38 eNOS  in  endothelial    cells.    Because  of  this,  a  second  set  of  data  was  collected.    Before     58     RBCs   RBCs  +   10  nM  E2       Figure   2.14   –   Fluorescence   Intensity   Images   –   DAF-­‐FM   reacts   with   NO   in   bPAECs   resulting   in   fluorescence   than   can   be   photographed   with   a   CCD   camera   and   the   pixel   intensity  can  be  measured.    As  seen  above,  when  RBCs  incubated  with  estradiol  (E2)  are   flowed  beneath  wells  containing  bPAECs,  the  NO  production  of  those  cells  is  lower.    This   is  due  to  the  decrease  in  RBC-­‐derived  ATP,  which  stimulates  NO  production  in  bPAECs.                 59   * 1.2 1.0 * E2 * * E2 * 0.8 0.6 * 0.4 0.2 E2 µ M .0 +2 +1 .5 µ M E2 E2 µ M .0 +1 5 µ M +0 . nM +5 0 nM +2 5 nM BC +1 0 R E2 E2 0.0 s Normalized Fluorescence Intensity 1.4 Figure   2.15   –   Normalize   Fluorescence   Intensity   –   The   data   in   this   figure   shows   the   normalized,   background   subtracted   emission.     As   shown,   the   emission   decreases   with   RBCs  incubated  in  buffer  containing  estradiol.    In  this  case,  the  estradiol  is  actually  able   to   reduce   the   NO   production   to   43   ±   0.1%   of   that   of   untreated   RBCs.     The   emission   intensity   begins   to   increase   with   again   with   increasing   concentration   of   estradiol,   although  a  significant  increase  in  emission,  in  comparison  to  RBCs  alone,  is  not  measure   until   the   estradiol   levels   reach   approximately   2   µM.     At   this   value,   the   fluorescence   intensity  is  13  ±  4%  higher  than  the  untreated  RBCs.    The  next  highest  concentration,  1.5   µM,  is  not  significantly  different  from  the  untreated  RBCs.    Error  is  shown  as  standard   deviation   for   N   =   4   rabbits   and   the   asterisk   denotes   a   significant   difference   from   untreated  RBCs  at  p  <  0.05.     60   the  RBC  samples  were  pumped  through  the  microfluidic  device,  they  were  centrifuged   and  washed  twice  with  PSS  and  then  resuspended  in  PSS.    As  shown  in  figure  2.16,  there   was   a   marked   decrease   in   NO   production   in   the   bPAECs,   denoted   by   the   decrease   in   fluorescent  emission,  when  the  RBCs  were  pre-­‐incubated  with  estradiol.    However,  the   RBCs  that  have  been  washed  to  remove  excess  estradiol  do  not  result  in  a  statistically   significant   increase   in   NO   production.     The   black   bars   represent   the   data   from   the   previous  figure  and  are  included  for  comparison.    The  grey  bars  show  that  the  estradiol   is   able   to   decrease   the   NO   production   in   bPAECs,   as   detected   by   fluorescence   intensity.     This   decrease   is   the   result   of   a   decrease   in   RBC-­‐derived   ATP.     The   NO   production   was   decreased   to   59   ±   7%   with   10   nM   estradiol   and   this   decrease   remained   statistically   constant  across  the  range  of  concentrations,  with  the  largest  decrease,  down  to  51  ±  5%   that  of  the  NO  production  of  bPAECs  exposed  to  untreated  RBCs,  seen  at  2.0  µM.         Together,  these  data  show  that  estrogens  have  the  ability  to  inhibit  the  release  of  ATP   from  RBCs,  as  well  as  the  subsequent  ATP-­‐stimulated  NO  production  in  endothelial  cells.     The   estrogens   were   also   able   to   attenuate   the   ATP   release   from   RBCs   that   were   2+ previously   incubated   with   Zn -­‐activated   C-­‐peptide,   an   ATP   stimulus.     Knowing   that   patients  with  MS  have  higher  than  normal  levels  of  ATP  released  from  their  RBCs,  and   the   ameliorating   effects   of   estrogens   on   the   disease,   these   data   could   provide   a   missing   link.         61   Normalized Fluorescence Intensity 1.4 RBCs Washed RBCs 1.2 * 1.0 * 0.8 * * * # # # # 0.6 # # # * 0.4 0.2 E2 +2 .0 µ M E2 .5 µ M E2 +1 .0 µ M E2 +1 .5 µ M E2 +0 nM E2 +5 0 nM E2 +2 5 nM +1 0 R BC s 0.0 Figure   2.16   –   Washed   and   Unwashed   RBCs   –   There   was   a   marked   decrease   in   NO   production  in  the  bPAECs,  denoted  by  the  decrease  in  fluorescent  emission,  when  the   RBCs  are  pre-­‐incubated  with  estradiol.    RBCs  that  have  been  washed  to  remove  excess   estradiol  do  not  result  in  a  statistically  significant  increase  in  NO  production.    The  black   bars  represent  the  data  from  the  previous  figure  and  are  included  for  comparison.    The   grey  bars  show  that  the  estradiol  is  able  to  decrease  the  NO  production  in  bPAECs,  as   detected  by  fluorescence  intensity.    The  NO  production  was  decreased  to  59  ±  7%  with   10   nM   estradiol,   and   this   decrease   remained   statistically   constant   across   the   range   of   concentrations,   down   to   51   ±   5%   that   of   the   NO   production   of   bPAECs   exposed   to   untreated   RBCs,   seen   at   2.0   µM.     Error   is   reported   as   standard   deviation   for   N   =   4   rabbits.   The   asterisks   denote   a   significant   difference   from   untreated   RBCs   (black   bar)   and  the  pound  sounds  denote  a  significant  decrease  from  untreated  RBCs  (grey  bar).     62   2.5  Discussion       The  interaction  of  RBCs  and  the  availability  of  NO  has  proved  to  be  difficult  to  discern   and   is   heavily   debated.     There   are   multiple   mechanisms   by   which   the   RBCs   are   involved   in  the  physiological  control  of  NO  levels.    RBCs  have  the  ability  to  act  as  a  scavenger  and   NO   carrier   through   the   use   of   nitroso-­‐thiol   groups,   releasing   this   NO   in   response   to   stimulus,  such  as  hypoxia. 39-­‐41    It  has  also  been  suggested  that  the  hemoglobin  in  the   RBC  reduces  nitrate  to  NO  as  a  nitrite  reductase  to  contribute  to  NO  levels  in  the  blood   42 stream.     Like   the   research   presented   here,   others   have   demonstrated   that   RBC-­‐ derived   ATP   contributes   to   an   increase   in   NO   levels   in   response   to   various   stimuli. 5,28,43,44     ATP   has   been   shown   to   increase   NO   production   in   a   variety   of   cells   types,  including  endothelial  cells.       The   results   in   figures   2.10,   2.11,   and   2.12,   show   that   estrogens   have   the   ability   to   decrease   the   ATP   release   from   RBCs.     This   was   anticipated   based   on   previous   studies   45 showing  that  DHEA  attenuated  RBC-­‐derived  ATP,  and  the  inhibitory  effect  of  estrogen   25-­‐27 on  CFTR  that  has  been  seen  in  other  cell  types.    In  addition  to  lowering  the  levels  of   ATP  release  from  RBC  at  a  basal  state,  the  estrogens  were  also  effective  at  decreasing   2+ the  release  from  RBCs  that  had  been  previously  treated  with  Zn  bound  to  C-­‐peptide,   stimulus  of  ATP  release.    As  previously  mentioned,  estrogens  have  been  shown  to  have   a  protective  effect  against  EAE,  the  animal  model  of  MS,  and  as  well  as  on  the  disease  in     63   humans. 22,46-­‐48     Coupled   with   the   data   from   the   preliminary   study   showing   that   patients  with  MS  have  higher  than  normal  amounts  of  ATP  released  from  their  RBCs  as   16 compared   to   healthy   controls,   this   may   provide   a   link   to   the   protective   effects   of   estrogens  on  MS.    Additionally,  the  increase  in  ATP  release  from  the  RBCs  of  MS  patients   may  contribute  to  the  increased  levels  of  NO  and  its  products  nitrate  and  nitrite.     RBCs  pumped  through  the  channel  of  the  microfluidic  device  release  ATP  as  a  result  of   shear  stress  that  stimulates  the  production  of  NO  in  bPAECs.    The  data  shown  in  figure   2.16   shows   this   NO   production   was   decreased   when   the   RBCs   were   incubated   with   estrogens  prior  to  their  being  flowed  through  the  system.    Increasing  concentrations  of   estrogens   subsequently   increase   the   amount   of   NO   produced   by   bBAECs.     When   the   RBCs  were  washed  with  PSS  to  remove  excess  estrogen  prior  to  their  introduction  to  the   microfluidic   device,   the   NO   production   did   not   increase   significantly   with   increasing   concentrations  of  estrogens  in  the  RBCs.         Interestingly,   other   reports   have   suggested   that   estrogens   stimulate   NO   production   in   23,24 cells.     If   this   was   the   only   factor,   it   would   seem   that   estrogens   should   not   have   the   protective  effect  that  is  seem  in  both  EAE  and  MS.    The  original  data,  seen  in  figure  2.15,   shows  that  this  may  be  the  case,  though  the  NO  production  did  not  increase  back  to  the   levels  seen  in  RBCs  alone  until  a  superphysiological  amount  of  estradiol  was  incubated   with   the   RBCs.   Previous   studies   investing   the   effect   of   estrogens   on   NO   production   in   38 endothelial  cells  have  applied  these  hormones  directly  to  the  cells.  This  discrepancy  in     64   data   shows   how   important   it   is   to   have   and   use   devices   that   enable   cell-­‐to-­‐cell   communication.    Applied  alone  to  endothelial  cells  estrogens  have  different  effects  than   they   do   in   vivo.     The   interaction   of   chemicals   with   blood   components   and   their   subsequent   effects   on   other   cells   is   an   important   part   of   understanding   what   is   truly   happening.     Here,   the   microfluidic   device   was   able   to   discern   between   endothelium-­‐ derived  NO  due  directly  to  estradiol  and  the  inhibition  of  NO  production  as  a  result  of   the  decrease  in  RBC-­‐derived  ATP  caused  by  the  hormone.     In   conclusion,   ATP   is   released   from   RBCs   in   response   to   shear   stress.     This   ATP   stimulates   NO   production   in   the   endothelial   cells   lining   the   blood   vessels.     RBCs   from   patients   with   MS   release   three   times   the   amount   of   ATP   then   RBCs   from   healthy   controls.    This  increase  in  ATP  release  may  be  over-­‐stimulating  the  NO  production  in  the   endothelial  cells,  and  NO  is  known  to  be  toxic  to  the  BBB.    It  has  been  shown  that  high   levels  of  estrogens,  due  to  pregnancy  or  drug  therapy,  can  ameliorate  MS.  The  findings   here  may  provide  an  explanation  for  the  therapeutic  effects  of  estrogens.    If  estrogens   are   able   to   attenuate   the   excessive   release   of   ATP   from   the   RBCs   of   MS   patients,   the   amount  of  NO  production  would  also  be  decreased,  potentially  leading  to  less  damage   from  the  disease.           65   REFERENCES     66       1     2     3     4     5     6     7     8     9     10       REFERENCES   Harmening,  D.  Clinical  Hematology  and  Fundamentals  of  Hemostasis.    (F.  A.  Davis   Co,  2009).   Dean,   B.   &   Perrett,   D.   Studies   on   adenine   and   adenosine   metabolims   by   intact   human  erythrocytes  using  high-­‐performance  liquid  chromatography.  Biochimica   Et  Biophysica  Acta  437,  1-­‐15  (1976).   Bergfeld,  G.  &  Forrester,  T.  Release  of  ATP  from  human  erythrocytes  in  response   to  a  brief  period  of  mypoxia  and  hypercapnia.  Cardiovascular  Research  26,  40-­‐47   (1992).   Ellsworth,   M.   L.,   Forrester,   T.,   Ellis,   C.   G.   &   Dietrich,   H.   H.   The   erythrocyte   as   a   regulator  of  vascular  tone.  American  Journal  of  Physiology-­‐Heart  and  Circulatory   Physiology  269,  H2155-­‐H2161  (1995).   Sprague,  R.  S.,  Ellsworth,  M.  L.,  Stephenson,  A.  H.  &  Lonigro,  A.  J.  ATP:  The  red   blood   cell   link   to   NO   and   local   control   of   the   pulmonary   circulation.   American   Journal   of   Physiology-­‐Heart   and   Circulatory   Physiology   271,   H2717-­‐H2722   (1996).   Olson,   A.   &   Pessin,   J.   Structure,   function,   and   regulation   of   the   mammalian   facilitative  glucose  transporter  gene  family.  Annual  Review  of  Nutrition  16,  235-­‐ 256  (1996).   Kasahara,  M.  &  Hinkle,  P.  Reconsitution  and  purification  of  D-­‐glucose  transporter   from   human   erythrocytes.   Journal   of   Biological   Chemistry   252,   7384-­‐7390   (1977).   Carruthers,  A.,  DeZutter,  J.,  Ganguly,  A.  &  Devaskar,  S.  Will  the  original  glucose   transporter   isoform   please   stand   up!   American   Journal   of   Physiology-­‐ Endocrinology  and  Metabolism  297,  E836-­‐E848  (2009).   van  Wijk,  R.  &  van  Solinge,  W.  The  energy-­‐less  red  blood  cell  is  lost:  erythrocyte   enzyme  abnormalities  of  glycolysis.  Blood  106,  4034-­‐4042  (2005).   Light,  D.,  Capes,  T.,  Gronau,  R.  &  Adler,  M.  Extracellular  ATP  stimulates  volume   decrease   in   Necturus   red   blood   cells.   American   Journal   of   Physiology-­‐Cell   Physiology  277,  C480-­‐C491  (1999).   67   11     12     13     14     15     16     17     18     19     20     21   Sprague,   R.   S.,   Ellsworth,   M.   L.,   Stephenson,   A.   H.,   Kleinhenz,   M.   E.   &   Lonigro,   A.   J.  Deformation-­‐induced  ATP  release  from  red  blood  cells  requires  CFTR  activity.   American  Journal  of  Physiology  275,  H1726-­‐1732  (1998).   Reisin,   I.   L.,   Prat,   A.   G.,   Abraham,   E.   H.,   Amara,   J.   F.,   Gregory,   R.   J.,   Ausiello,   D.   A.   &  Cantiello,  H.  F.  The  cystic  fibrosis  transmembrane  conductance  regulator  is  a   duel  ATP  and  chloride  channel.  Journal  of  Biological  Chemistry  269,  20584-­‐20591   (1994).   Sprague,   R.,   Stephenson,   A.,   Ellsworth,   M.,   Keller,   C.   &   Lonigro,   A.   Impaired   release   of   ATP   from   red   blood   cells   of   humans   with   primary   pulmonary   hypertension.  Experimental  Biology  and  Medicine  226,  434-­‐439  (2001).   Sprague,   R.,   Stephenson,   A.,   Bowles,   E.,   Stumpf,   M.,   Ricketts,   G.   &   Lonigro,   A.   Expression   of   the   heterotrimeric   G   protein   Gi   and   ATP   release   are   impaired   in   erythrocytes   of   humans   with   diabetes   mellitus.   Hypoxia   and   Exercise   588,   207-­‐ 216  (2006).   Subasinghe,   W.   &   Spence,   D.   M.   Simultaneous   determination   of   cell   aging   and   ATP  release  from  erythrocytes  and  its  implications  in  type  2  diabetes.  Analytica   Chimica  Acta  618,  227-­‐233  (2008).   Letourneau,   S.,   Hernandez,   L.,   Faris,   A.   N.   &   Spence,   D.   M.   Evaluating   the   effects   of   estradiol   on   endothelial   nitric   oxide   stimulated   by   erythrocyte-­‐derived   ATP   using  a  microfluidic  approach.  Analytical  and  Bioanalytical  Chemistry  397,  3369-­‐ 3375  (2010).   Sprung,   R.,   Sprague,   R.   &   Spence,   D.   Determination   of   ATP   release   from   erythrocytes   using   microbore   tubing   as   a   model   of   resistance   vessels   in   vivo.   Analytical  Chemistry  74,  2274-­‐2278  (2002).   Palmer,  R.  M.,  Ferrige,  A.  G.  &  Moncada,  S.  Nitric  oxide  release  accounts  for  the   biological   activity   of   endothelium-­‐derived   relaxing   factor.   Nature   327,   524-­‐526   (1987).   Kolb,   H.   &   Kolb-­‐Bachofen,   V.   Nitric   oxide   in   autoimmune   disease:   cytotoxic   or   regulatory  mediator?  Immunology  Today  19,  556-­‐561  (1998).   Giovannoni,  G.,  Heales,  S.  J.,  Land,  J.  M.  &  Thompson,  E.  J.  The  potential  role  of   nitric  oxide  in  multiple  sclerosis.  Multiple  Sclerosis  4,  212-­‐216  (1998).   Confavreux,  C.  et  al.  Rate  of  pregnancy-­‐related  relapse  in  multiple  sclerosis.  New   England  Journal  of  Medicine  339,  285-­‐291  (1998).       68   22     23     24     25     26     27     28     29     30     31     32     Voskuhl,   R.   R.,   Kim,   S.,   Liva,   S.   M.,   Dalal,   M.   A.   &   Verity,   M.   A.   Estriol   ameliorates   autoimmune   demyelinating   disease   -­‐   Implications   for   multiple   sclerosis.   Neurology  52,  1230-­‐1238  (1999).   Jaubert,   A.   M.,   Mehebik-­‐Mojaat,   N.,   Lacasa,   D.,   Sabourault,   D.,   Giudicelli,   Y.   &   Ribière   C.   Nongenomic   estrogen   effects   on   nitric   oxide   synthase   activity   in   rat   adipocytes.  Endocrinology  148,  2444-­‐2452  (2007).   Scott,  P.  A.,  Tremblay,  A.,  Brochu,  M.  &  St-­‐Louis,  J.  Vasorelaxant  action  of  17β-­‐ estradiol   in   rat   uterine   arteries:   role   of   nitric   oxide   synthases   and   estrogen   receptors.  American  Journal  of  Physiology  293,  H3713-­‐H3719  (2007).   Sweezey,  N.,  Gauthier,  C.,  Gagnon,  S.,  Ferretti,  E.  &  Kopelman,  H.  Progesterone   and   estradiol   inhibit   CFTR-­‐mediated   ion   transport   by   pancreatic   epithelial   cells.   American  Journal  of  Physiology-­‐Gastrointestinal  and  Liver  Physiology  271,  G747-­‐ G754  (1996).   Sweezey,  N.,  Ghibu,  F.  &  Gagnon,  S.  Sex  hormones  regulate  CFTR  in  developing   fetal   rat   lung   epithelial   cells.   American   Journal   of   Physiology-­‐Lung   Cellular   and   Molecular  Physiology  272,  L844-­‐L851  (1997).   Singh,  A.  K.,  Schultz,  B.  D.,  Katzenellenbogen,  J.  A.,  Price,  E.  M.,  Bridges,  R.  J.  &   Bradbury,   N.   A.   Estrogen   inhibition   of   cystic   fibrosis   transmembrane   conductance   regulator-­‐mediated   chloride   secretion.   Journal   of   Pharmacology   and  Experimental  Therapeutics  295,  195-­‐204  (2000).   Genes,   L.   I.,   N,   V.   T.,   Hulvey,   M.   K.,   Martin,   R.   S.   &   Spence,   D.   M.   Addressing   a   vascular  endothelium  array  with  blood  components  using  underlying  microfluidic   channels.  Lab  on  a  Chip  7,  1256-­‐1259  (2007).   Whitesides,  G.  M.  The  origins  and  the  future  of  microfluidics.  Nature  442,  368-­‐ 373  (2006).   Duffy,   D.,   McDonald,   J.,   Schueller,   O.   &   Whitesides,   G.   Rapid   prototyping   of   microfluidic   systems   in   poly(dimethylsiloxane).   Analytical   Chemistry   70,   4974-­‐ 4984  (1998).   Zhang,   J.,   Tan,   K.   L.,   Hong,   G.   D.,   Yang,   L.   J.   &   Gong,   H.   Q.   Polymerization   optimization  of  SU-­‐8  photoresist  and  its  applications  in  microfluidic  systems  and   MEMS.  Journal  of  Micromechanics  and  Microengineering  11,  20-­‐26  (2001).   Lee,   P.,   Hung,   P.,   Shaw,   R.,   Jan,   L.   &   Lee,   L.   Microfluidic   application-­‐specific   integrated   device   for   monitoring   direct   cell-­‐cell   communication   via   gap   junctions   between  individual  cell  pairs.  Applied  Physics  Letters  86,  No.  223902  (2005).   69     33     34     35     36     37     38     39     40     41     42     43     Abhyankar,   V.,   Lokuta,   M.,   Huttenlocher,   A.   &   Beebe,   D.   Characterization   of   a   membrane-­‐based   gradient   generator   for   use   in   cell-­‐signaling   studies.   Lab   on   a   Chip  6,  389-­‐393  (2006).   Tolan,   N.   V.,   Genes,   L.   I.,   Subasinghe,   W.,   Raththagala,   M.   &   Spence,   D.   M.   Personalized   Metabolic   Assessment   of   Erythrocytes   Using   Microfluidic   Delivery   to  an  Array  of  Luminescent  Wells.  Analytical  Chemistry  81,  3102-­‐3108  (2009).   Genes,   L.   I.,   Tolan,   N.   V.,   Hulvey,   M.   K.,   Martin,   R.   S.   &   Spence,   D.   M.   Addressing   a   vascular   endothelium   array   with   blood   components   using   underlying   microfluidic  channels.  Lab  on  a  Chip  7,  1256-­‐1259  (2007).   Peck,   J.   D.,   Hulka,   B.   S.,   Poole,   C.,   Savitz,   D.   A.,   Baird,   D.,   Richardson   &   B.   E.   Steroid   hormone   levels   during   pregnancy   and   incidence   of   maternal   breast   cancer.  Cancer  Epidemiology  Biomarkers  &  Prevention  11,  361-­‐368  (2002).   Sicotte,  N.  L.,  Liva,  S.  M.,  Klutch,  R.,  Pfeiffer,  P.,  Bouvier,  S.,  Odesa,  S.,  Wu,  T.  C.  &   Voskuhl,   R.   R.   Treatment   of   multiple   sclerosis   with   the   pregnancy   hormone   estriol.  Annals  of  Neurology  52,  421-­‐428  (2002).   Chambliss,   K.   L.   &   Shaul,   P.   W.   Estrogen   modulation   of   endothelial   nitric   oxide   synthase.  Endocrinology  Reviews  23,  665-­‐686  (2002).   Jia,  L.,  Bonaventura,  C.,  Bonaventura,  J.  &  Stamler,  J.  S.  S-­‐Nitrosohemoglobin:  a   dynamic   activity   of   blood   involved   in   vascular   control.   Nature   380,   221-­‐226   (1996).   Stamler,  J.  S.,  Jia,  L.,  Eu,  J.  P.,  McMahon,  T.  J.,  Demchenko,  I.  T.,  Bonaventura,  J.,   Gernert,  K.  &  Piantadosi,  C.  A.  Blood  flow  regulation  by  S-­‐nitrosohemoglobin  in   the  physiological  oxygen  gradient.  Science  276,  2034-­‐2037  (1997).   Gow,  A.  J.  &  Stamler,  J.  S.  Reactions  between  nitric  oxide  and  haemoglobin  under   physiological  conditions.  Nature  391,  169-­‐173  (1998).   Minneci,  P.  C.,  Deans,  K.  J.,  Shiva,  S.,  Zhi,  H.,  Banks,  S.  M.,  Kern,  S.,  Natanson,  C.,   Solomon,   S.   B.   &   Gladwin,   M.   T.   Nitrite   reductase   activity   of   hemoglobin   as   a   systemic   nitric   oxide   generator   mechanism   to   detoxify   plasma   hemoglobin   produced   during   hemolysis.   American   Journal   of   Physiology   295,   H743-­‐H754   (2008).   Sprague,   R.   S.,   Olearczyk,   J.   J.,   Spence,   D.   M.,   Stephenson,   A.   H.,   Sprung,   R.   W.   &   Lonigro,   A.   J.   Extracellular   ATP   signaling   in   the   rabbit   lung:   erythrocytes   as   70     44     45     46     47     48     49     50           determinants   of   vascular   resistance.   American   Journal   of   Physiology-­‐Heart   and   Circulatory  Physiology  285,  H693-­‐H700  (2003).   Spence,   D.   M.,   Torrence,   N.   J.,   Kovarik,   M.   L.   &   Martin,   R.   S.   Amperometric   determination   of   nitric   oxide   derived   from   pulmonary   artery   endothelial   cells   immobilized  in  a  microchip  channel.  Analyst  129,  995-­‐1000  (2004).   Spence,   D.   M.   &   Subasinghe,   W.   Simultaneous   determination   of   cell   aging   and   ATP  release  from  erythrocytes  and  its  implications  in  type  2  diabetes.  Analytica   Chimica  Acta  618,  227-­‐233  (2008).   Jansson,  L.,  Olsson,  T.  &  Holmdahl,  R.  Estrogen  Induces  a  Potent  Suppression  of   Experimental   Autoimmune   Encephalomyelitis   and   Collagen-­‐Induced   Arthritis   in   Mice.  Journal  of  Neuroimmunology  53,  203-­‐207  (1994).   Palaszynski,   K.   M.,   Liu,   H.,   Loo,   K.   K.   &   Voskuhl,   R.   R.   Estriol   treatment   ameliorates  disease  in  males  with  experimental  autoimmune  encephalomyelitis:   implications   for   multiple   sclerosis.   Journal   of   Neuroimmunology   149,   84-­‐89   (2004).   Gold,  S.  M.  &  Voskuhl,  R.  R.  Estrogen  treatment  in  multiple  sclerosis.  Journal  of   the  Neurological  Sciences  286,  99-­‐103  (2009).   Olearczyk,  J.  J.,  Stephenson,  A.  H.,  Lonigro,  A.  J.  &  Sprague,  R.  S.  Heterotrimeric  G   protein   G(i)   is   involved   in   a   signal   transduction   pathway   for   ATP   release   from   erythrocytes.   American   Journal   of   Physiology-­‐Heart   and   Circulatory   Physiology   286,  H940-­‐H945  (2004).   Sprague,  R.  S.,  Ellsworth,  M.  L.,  Stephenson,  A.  H.  &  Lonigro,  A.  J.  Participation  of   cAMP  in  a  signal-­‐transduction  pathway  relating  erythrocyte  deformation  to  ATP   release.   American   Journal   of   Physiology-­‐Cell   Physiology   281,   C1158-­‐C1164   (2001).     71   2+ Chapter  3  –  Delivery  of  Zn  to  the  Red  Blood  Cell  by  C-­‐peptide       3.1  Diabetes  Mellitus       3.1.1  Classifications       The   term   diabetes   mellitus,   commonly   referred   to   as   diabetes,   denotes   a   group   of   diseases   that   result   in   hyperglycemia   through   various   mechanisms.     Hyperglycemia   is   the  term  used  to  describe  higher  than  normal  amounts  of  glucose  in  the  blood  plasma.     Normally,  the  blood  glucose  level  falls  between  70  and  100  mg/dL,  or  3.8  and  5.5  mM.     A  fasting  glucose  level  above  100  mg/dL  is  referred  to  as  impaired  fasting  glucose,  and   1 can   be   a   precursor   to   diabetes.     If   this   level   is   greater   than   or   equal   to   126   mg/dL,   or   7   mM,  a  patient  is  diagnosed  as  diabetic.    Another  way  to  diagnose  diabetes  is  by  using  a   glucose  tolerance  test,  where  75  grams  of  glucose  are  ingested  and  the  blood  is  tested  2   hours   later.     If   the   glucose   level   is   at   or   above   200   mg/dL,   or   11   mM,   the   diagnosis   is   2 confirmed.     There   are   four   recognized   classifications   of   diabetes:     type   1   and   type   2   diabetes,   gestational   diabetes,   and   other.     The   other   types   of   diabetes   include   any   mechanism   by   which   the   pancreas   becomes   damaged   or   destroyed   and   hormonal   imbalances   or   genetic   defects   that   alter   insulin   levels.     Type   1   and   type   2   diabetes   differ   by   the   role   of   insulin.    Type  1  diabetes,  also  known  as  juvenile  diabetes,  is  characterized  by  impaired   insulin   secretion.     Insulin   is   the   hormone   necessary   for   regulating   glucose   uptake   into     72   cells.    As  a  result,  if  its  secretion  is  impaired,  elevated  levels  of  blood  glucose  are  seen.     In  type  1  diabetes,  the  beta  cells  in  the  pancreas,  responsible  for  the  secretion  of  insulin   2 are  destroyed  in  an  autoimmune  process.    The  cause  of  this  destruction  has  not  been   fully  elucidated,  but  is  believed  to  have  both  environmental  and  genetic  factors  and  may   be  caused,  in  part,  by  a  virus. 3,4       Patients   with   type   2   diabetes   make   up   about   90%   of   the   diabetic   population   in   the   5 United   States.     Type   2   diabetics   are   insulin   resistant,   meaning   that,   as   compared   to   normal   controls,   the   same   amount   of   insulin   will   not   be   as   effective   in   stimulating   6 glucose  uptake  and  utilization.    This  resistance  eventually  causes  the  beta  cells  of  the   pancreas  to  decrease  and  eventually  stop  the  production  of  insulin,  although  the  reason   7 for  this  is  unclear.    Type  2  diabetes  often  develops  later  in  life,  generally  after  the  age   of  40.    While  there  is  a  genetic  component  to  the  disease  susceptibility,  environmental   factors   also   play   a   role.     Being   overweight   or   obese   has   been   significantly   associated   with  several  aspects  of  poor  health  status,  including  type  2  diabetes.    When  compared   to   adults   of   normal   weight,   adults   who   are   considered   to   be   extremely   obese,   with   a   8 body  mass  index  of  40  or  above,  are  more  than  7  times  more  likely  to  have  diabetes.     Gestational   diabetes   mellitus   (GDM)   is   defined   as   glucose   intolerance   first   recognized   during   pregnancy.     Approximately   4%   of   pregnancies   are   complicated   by   GDM   by   increasing   the   risk   of   high   blood   pressure   and   preeclampsia,   as   well   as   higher   birth     73   9 weights   and   an   increase   in   the   number   of   caesarian   delivery.     Woman   who   experience   GDM   are   at   a   higher   risk   for   the   development   of   type   2   diabetes,   with   approximately   5 70%  developing  the  disease  within  ten  years  of  delivery.     The  hyperglycemia  seen  in  diabetic  patients  is  associated  with  a  large  variety  of  diabetic   complications.     The   major   cause   of   disability   and   death   in   diabetic   patients   is   artherosclerosis,   a   vascular   disease   where   substances,   such   as   fat   and   cholesterol,   build   10 up   on   the   walls   of   arteries,   forming   plaques,   which   harden   them.     Some   of   these   plaques   are   vulnerable,   meaning   the   fibrous   cap   that   separates   the   plaque   from   the   lining  of  the  artery  is  likely  to  rupture.  When  this  happens,  a  cascade  of  events  occurs   and  a  thrombus  is  formed  that  can  be  lodged  in  an  artery,  slowing  or  preventing  blood   11 flow   to   tissue   in   a   certain   area   of   the   body.     Additionally,   hyperglycemia   has   been   shown  to  increase  oxidative  stress  in  endothelial  cells  and  as  a  result,  nitric  oxide  (NO)   release   is   decreased.     Furthermore,   it   has   been   shown   that   the   red   blood   cells   (RBCs)   of   patients   with   type   2   diabetes   release   significantly   less   adenosine   triphosphate   (ATP)   than  those  of  healthy  controls. 12,13    ATP  is  a  known  stimulator  of  NO,  as  described  in   chapter   1.     NO   is  a   known   vasodilator,   so   the   lack   of   ATP   release   and   vessel   dilation   add   14 to  the  vascular  complications  seen  in  the  disease.           74   3.1.2  Complications       Complications  in  the  microvascular  system  lead  to  a  variety  of  other  common  diabetic   complications,   such   as   retinopathy,   neuropathy   and   nephropathy.     Retinopathy   is   damage  to  the  retina  that  can  cause  blindness.    In  fact,  12%  of  all  new  cases  of  blindness   15 each   year   are   the   result   of   diabetic   retinopathy.     One   of   the   major   factors   involved   in   the   decline   of   retinal   function   is   vascular   damage   leading   to   a   decrease   in   blood   flow   to   the  area  of  the  eye,  which  results  in  a  lack  of  oxygen  to  the  tissue.    As  a  result,  there  is   an   increase   in   neovascular   growth   factors   to   promote   the   proliferation   of   new   blood   vessels.    While  it  would  seem  that  new  blood  vessels  would  ameliorate  the  issues,  the   vessel  growth  results  in  an  accumulation  of  fibrous  tissue  that  can  distort  the  retina  or   16 even  detach  the  retina,  leading  to  vision  loss.     In  addition  to  an  increase  risk  of  blindness,  nearly  all  patients  with  diabetes  have  at  least   some  level  of  neuropathy,  or  damage  to  the  nervous  system.    Although  it  can  occur  at   any   time,   the   prevalence   of   neuropathy   positively   correlates   to   disease   duration   and   17 18 severity,  with  a  more  rapid  progression  seen  in  patients  with  type  1  diabetes.    Like   retinopathy,  it  is  believed  that  neuropathy  is  caused  in  part  by  abnormal  blood  flow  in   19 the   neurovasculature,   in   addition   to   direct   damage   due   to   hyperglycemia.     The   previously   mentioned   decrease   in   endothelial   NO   production   may   also   play   a   role   in   diabetic   neuropathy.     The   lack   of   vasodilation   may   make   the   occlusion   of   vessels   an   even  more  common  event.    Neuropathy  leads  to  the  loss  of  sensation  and  weakness  in     75   the   extremities   and   when   seen   in   combination   with   other   microvascular   issues,   is   the   20 leading  cause  of  non-­‐traumatic  amputations  in  the  United  States.     Nephropathy  affects  the  kidneys  of  both  type  1  and  type  2  diabetic  patients,  with  the   21 former  being  at  a  great  risk  for  total  renal  failure.    Combining  the  types,  diabetes  is   22 the   most   common   cause   of   end-­‐stage   kidney   failure   in   the   United   States.     In   the   kidney,   there   are   systems   of   ducts,   tubules,   and   arteries   that   make   up   nephrons.     These   systems  are  responsible  for  filtering  the  blood  and  removing  the  waste  as  urine.    In  each   nephron  there  is  a  capillary  referred  to  as  the  glomerulus  in  which  the  blood  pressure  is   very   high.     This   pressure   forces   water   and   solutes   out   of   the   blood   into   the   surrounding   glomerular   capsule,   where   it   is   further   processed   and   eventually   becomes   urine   to   be   excreted   from   the   body.     In   diabetic   patients,   nephropathy   is   characterized   by   glomerular   hyperfiltration.     Because   of   the   difficulties   in   the   microvasculature,   the   pressure  in  the  glomerulus  is  increased  and  as  a  result,  albumin,  a  common  protein  in   23 blood   plasma,   is   forced   out   into   the   urine,   causing   a   condition   called   albuminuria.   Additionally,  in  the  early  stages  of  nephropathy,  the  glomerular  basement  membranes   surrounding   the   glomerulus,   become   thicker   as   a   result   of   upregulated   production   of   the  macromolecules,  such  as  collagen,  that  make  up  this  membrane.    The  thickening  of   the   glomerular   basement   membranes   also   leads   to   glomerular   dysfunction   and,   eventually,  to  kidney  failure. 24,25       76   It   is   important   to   note   that   retinopathy,   neuropathy,   and   nephropathy   are   all   at   least   partially  caused  by  the  issues  that  develop  in  the  microvasculature  of  diabetic  patients.     While  none  of  these  conditions  can  be  avoided  completely,  the  risk  can  be  lessened  by   controlling   blood   sugar   through   diet,   insulin   injections,   and   other   pharmaceuticals,   26 depending   on   the   type   and   severity   of   the   diabetes.     If   the   complications   in   the   microvasculature   can   be   minimized   the   quality   of   life   for   diabetic   patients   would   be   much  improved.           3.2  Insulin  and  C-­‐Peptide       3.2.1  Discovery  of  Insulin       It   is   believed   that   diabetes   was   first   described   by   the   ancient   Egyptians   over   3000   years   ago.     In   the   first   century,   Aretaeus,   a   physician   in   ancient   Greece   noted   that   people   afflicted  with  the  disease  urinated  often  and  therefore  used  the  Greek  word  diabetes,   meaning  siphon,  to  refer  to  it.    Patients  of  the  time  did  not  live  long,  as  there  was  no   th treatment.     The   term   diabetes   mellitus   was   not   used   until   the   17   century   when   a   British   physician,   Thomas   Willis,   noted   a   sweet   taste   in   the   urine   of   patients   with   diabetes.     The   method   of   diagnosis   became   tasting   the   urine   of   a   patient;   if   it   was   th sweet,   the   diagnosis   was   diabetes   mellitus.   This   remained   unchanged   until   the   20   century.     Despite   physician   efforts,   there   remained   no   treatment   for   diabetic   patients   th until  the  early  20  century  when  physicians  began  using  starvation  diets,  limiting  food     77   to  less  than  500  calories  per  day,  to  combat  the  disease.    Unfortunately,  this  treatment   left   patients   in   very   weakened   states   and   it   was   not   uncommon   for   them   to   die   of   starvation. 27,28     In  1889  Oskar  Minkowski  and  Joseph  von  Mering  removed  the  pancreas  from  a  dog  and   the   animal   developed   severe   diabetes.     Through   further   testing,   they   determined   that   the   pancreas   produces   a   substance   that   is   involved   in   the   regulation   of   blood   glucose   levels.    In  1921,  Fredrick  Banting  and  Charles  Best  discovered  that  by  injecting  a  crude   extract  of  the  pancreas  into  a  dog  that  had  previously  had  his  pancreas  removed  they   were   able   to   ameliorate   the   symptoms.     The   dog   became   healthier   and   his   blood   glucose  levels  dropped.    The  following  year,  after  Bertram  Collip  had  purified  insulin  out   of  the  extract,  the  team  had  their  first  human  subject.    A  14-­‐year-­‐old  boy,  near  death,   28 was  successfully  treated  and  lived  for  an  additional  13  years.     3.2.2  Insulin  and  C-­‐Peptide  Production  and  Release     Since   the   work   of   Banting   and   Best,   much   has   been   learned   about   insulin   and   its   production  in  the  pancreas.    The  pancreas  contains  the  islets  of  Langerhans,  scattered   throughout   the   organ,   comprising   1-­‐2%   of   total   pancreatic   mass.     There   are   four   different  types  of  islet  cells:  α-­‐cells,  β-­‐cells,  δ-­‐cells,  and  pancreatic  polypeptide  cells.    It  is   28 the   β-­‐cells   that   are   responsible   for   the   synthesis,   storage,   and   secretion   of   insulin.     Insulin   is   a   51   amino   acid   peptide   hormone   that   is   made   up   two   chains,   A   and   B,   connected  by  two  disulfide  bonds.    It  is  first  synthesized  as  preproinsulin  in  the  cytosol     78   before  being  taken  up  into  the  rough  endoplasmic  reticulum.    After  the  signal  peptide   has   resulted   in   the   correct   relocation   of   preproinsulin,   it   is   cleaved   off   to   form   proinsulin.         Proinsulin   consists   of   the   A   and   B   chains   that   are   connected   by   a   third   chain,   C-­‐peptide,   as   shown   in   figure   3.1.     Proinsulin   is   then   transported   to   the   Golgi   apparatus   where   it   is   subsequently   packaged   into   vesicles.   C-­‐peptide   is   cleaved   from   proinsulin   during   this   29 process,   which   is   completed   in   the   vesicles.   Three   enzymes   are   involved   in   this   process;   two   endopeptidases   and   carboxypeptidease   H   remove   C-­‐peptide   from   30 proinsulin  to  create  insulin.     2+ At   the   same   time   the   C-­‐peptide   is   being   cleaved   from   the   proinsulin,   Zn   ions   are   2+ 31 entering  the  vesicle  through  ZnT8,  a  Zn  transporter  in  the  vesicle  membrane.    As  a   2+ 32 result,   the   concentration   of   Zn   in   the   vesicles   approaches   millimolar   levels.     This   2+ Zn   is   involved   in   the   creation   of   proinsulin   hexamers,   consisting   of   six   proinsulin   molecules   and   two   Zn 2+   2+ atoms.     After   C-­‐peptide   is   removed,   the   Zn -­‐insulin   complex   33 becomes  less  soluble  and  is  stored  in  the  vesicles  in  a  crystalline  form.     Insulin   is   released   from   the   β-­‐cells   in   response   to   blood   glucose   levels.     When   these   levels  exceed  5.5  mM,  as  is  the  case  after  a  healthy  individual  eats  a  meal,  there  is  an   increase  in    glucose  uptake  into  the  β-­‐cell    through  a  glucose  transporter,  GLUT2.      Inside       79     Figure  3.1  –  Proinsulin  –  This  basic  diagram  shows  the  role  of  C-­‐peptide  in  the  structure   of   insulin,   as   well   as   the   amino   acid   sequence   of   C-­‐peptide.     The   acidic   residues   have   been  colored  purple.    When  the  C-­‐peptide  is  cleaved  from  proinsulin,  the  A  and  B  chains   remain,  forming  insulin.           80   34 the  β-­‐cell,  glucose-­‐6-­‐phosphate  is  formed  through  a  phosphorylation  by  glucokinase.     As   a   result   of   the   phosphorylation,   the   concentration   of   ATP   decreases,   as   the   concentration  of  adenosine  diphosphate  (ADP)  increases.    The  change  in  this  ratio  closes   a   potassium   ion   channel   and   opens   a   calcium   ion   channel;   in   response   to   in   influx   of   calcium  ions,  the  vesicles  in  the  β-­‐cells  undergo  exocytosis. 35-­‐37    A  diagram  of  this  can   be  seen  in  figure  3.2.    It  is  important  to  note  that  when  this  occurs,  insulin  and  C-­‐peptide   2+ are  released  in  equimolar  amounts  and  Zn  is  present  in  as  well.             3.2.3  Structure  of  C-­‐peptide       As  mentioned  in  the  previous  section,  C-­‐peptide  is  produced  as  a  by-­‐product  of  insulin.     C-­‐peptide   is   a   31   amino   acid   peptide;   the   sequence   of   this   peptide   has   not   been   38 conserved   evolutionarily,   nor   is   it   conserved   between   species.     However,   at   the   carboxyl  end  of  the  peptide,  the  glutamine  in  the  first  position  and  the  glutamic  acid  in   the   fifth   position   are   conserved   in   almost   70%   of   species,   and   the   pentapeptide   from   this  end,  EGSLQ,  has  been  shown  to  elicit  75%  of  the  activity  that  is  seen  with  the  intact   39 peptide.     Pure  C-­‐peptide  does  not  have  a  stable  secondary  structure  in  aqueous  solution.    While  it   40 is   not   physiologically   relevant,   there   is   evidence   of   α-­‐helices   in   trifluoroethanol.     Circular   dichroism   (CD)   spectroscopy   data   from   the   Spence   lab,   collected   by   Wathsala   Medawala,    confirmed  that    there    was    no    observable    secondary  structure    for  pure    C-­‐     81       Figure  3.2  –  Insulin  Packaging  and  Release  –    Insulin  is  first  synthesize  as  preproinsulin   in   the   cytosol   before   being   uptaken   into   the   rough   endoplasmic   reticulum   where   the   signal   peptide   is   cleaved   to   form   proinsulin.     It   is   then   transported   to   the   Golgi   apparatus   where   it   is   subsequently   packaged   into   vesicles.   C-­‐peptide   is   cleaved   from   29 2+ proinsulin   during   this   process.     At   the   same   time,   Zn   ions   are   entering   the   vesicle   2+ 31 2+ through   ZnT8,   a   Zn   transporter   in   the   vesicle   membrane.     This   Zn   is   involved   in   2+ the  creation  of  proinsulin  hexamers,  consisting  of  six  proinsulin  and  two  Zn .    After  C-­‐ 2+ peptide  is  removed,  the  Zn -­‐insulin  complex  becomes  less  soluble  and  is  stored  in  the   33 vesicles   in   a   crystalline   form.     Insulin   is   released   from   the   β-­‐cells   in   response   to   blood   glucose   levels,   and   the   vesicles   undergo   exocytosis. 35-­‐37     It   is   important   to   note   that   2+ when  this  occurs,  insulin  and  C-­‐peptide  are  released  in  equimolar  amounts  and  Zn  is   present  as  well.           82   2+ peptide  in  aqueous  solution.    However,  when  Zn  is  added  in  a  1:1  mole  ratio  with  C-­‐ peptide,   there   is   a   decrease   in   the   minimum   on   the   CD   spectra,   denoting   less   2+ randomness  and  potential  folding.    Interestingly,  when  a  higher  concentration  of  Zn  is   41 added,  the  minimum  returns  to  the  value  for  C-­‐peptide  alone.     To   date,   no   receptor   for   C-­‐peptide   has   been   found,   though   studies   have   shown   it   42 binding   to   several   different   human   cell   membranes.     While   the   mechanism   of   the   bioactivity   of   C-­‐peptide   is   incompletely   understood,   there   have   been   physiological   + + 43 44,45 effects  measured  on  N /K -­‐ATPase  activity,  endothelial  NO  synthase  stimuation,   46 mitogen  activated  protein  kinase  (MAPK)  pathway  activation,  phosphatidylinositol-­‐3-­‐ 47 48 kinase  activation,  and  downstream  activation  of  transcription  factors.     3.2.4  Biological  Effects  of  C-­‐Peptide     Originally   discovered   in   1967,   C-­‐peptide   was   long   considered   to   be   biologically   49 inactive.    It  was  widely  accepted  that  it  was  only  required  for  the  proper  arrangement   50 of  the  insulin  A  and  B  chains,  as  shown  in  figure  3.1.    In  fact,  clinically,  C-­‐peptide  has   only   been   used   as   a   diagnostic.     Because   of   its   longer   half-­‐life   in   blood   and   its   direct   correlation  to  insulin  release,  C-­‐peptide  is  used  to  determine  insulin  release  in  diabetic   patients.     This   information   is   used   to   discern   a   diagnosis   of   diabetes,   including   type   1   versus  type  2,  as  well  as  to  monitor  patient  response  to  pharmaceuticals  that  stimulate     83   51 the   secretion   of   insulin.   In   the   past   20   years,   however,   there   has   been   emerging   52,53 research   showing   the   C-­‐peptide   has   other   physiological   effects,   particularly   when   2+ 54 bound  to  Zn .     C-­‐peptide  lacks  basic  residues,  but  contains  five  acidic  residues:  four  glutamic  acids  and   one   aspartic   acid.     The   result   of   this   is   a   negatively   charged   peptide   that   result   in   the   2+ 55 ability   to   bind   cations,   specifically,   Zn .     While   studies   have   emerged   finding   56-­‐58 physiological   activity   of   C-­‐peptide,   there   have   been   reports   that   this   activity   is   only   59 seen   when   C-­‐peptide   is   co-­‐administered   with   insulin,   and   the   results   were   often   difficult   to   reproduce.     In   2008,   Meyer   et   al.   were   also   able   to   show   that   C-­‐peptide   increases   the   ATP   release   from   RBCs;   however,   a   metal   ion   was   necessary   for   this   to   occur.    During  the  course  of  their  investigation,  it  was  noted  that  the  C-­‐peptide  that  had   been   believed   to   be   pure   actually   contained   metal   ions.     This   impure   C-­‐peptide   was   active   when   first   dissolved   in   water,   but   lost   activity   over   the   course   of   hours.     The   2+ 3+ metal   ions   originally   used   to   produce   C-­‐peptide   activity   were   Fe   and   Cr ,   however,   2+ based  on  the  high  concentration  of  Zn  in  the  β-­‐cell  where  C-­‐peptide  is  produced,  it  is   60 predicted  that  this  is  the  metal  activating  C-­‐peptide  in  vivo.         It   is   the   position   of   this   thesis   that   as   a   result   of   presumable   pure   C-­‐peptide   being   contaminated   with   metals,   the   reproducibility   issues   with   previous   results   are     84   explainable.     Additionally,   in   cases   where   activity   is   seen   only   when   insulin   is   co-­‐ 2+ administered,   it   is   possible   that   the   insulin   still   contains   high   amounts   of   Zn   able   to   interact   with   C-­‐peptide.     In   an   in   vivo   study   looking   at   the   effects   of   C-­‐peptide   on   glucose   metabolism   in   type   1   diabetics,   C-­‐peptide   alone   increased   glucose   utilization   by   61 25%.     A   separate   study   found   that   the   co-­‐administration   of   C-­‐peptide   and   insulin   62 2+ increased  glucose  metabolism  by  66%.    Zn  bound  to  C-­‐peptide  has  also  been  shown   to  increase  ATP  release  from  RBCs  obtained  from  a  type  1  diabetic  rat  model  back  to  the   level  of  healthy  controls.    This  effect  was  not  seen  in  the  RBCs  from  the  type  2  diabetic   rats   unless   they   were   first   incubated   with   metformin,   a   common   pharmaceutical   to   treat  the  disease.    This  has  lead  to  the  suggestion  that  the  RBCs  of  patients  with  type  2   54 diabetes  are  both  C-­‐peptide  and  insulin  resistant,  and  therefore,  there  is  an  increased   need  for  the  study  of  the  biological  effects  of  C-­‐peptide.     The   research   presented   here   uses   65 2+ Zn   and   enzyme-­‐linked   immunosorbant   assays   2+ (ELISA)   to   investigate   the   transport   of   Zn   and   C-­‐peptide   to   healthy   human   RBCs.     A   mutant   form   of   C-­‐peptide   was   employed   to   investigate   the   importance   of   the   acidic   th 2+ residue  in  the  27  position  in  the  binding  of  Zn  to  the  peptide  and  then  subsequent   2+ delivery   to   the   RBC.     These   studies   help   to   more   fully   understand   the   effect   the   Zn   bound  to  C-­‐peptide  has  on  the  metabolism  of  RBCs.       85   3.3  Experimental       3.3.1  Preparation  of  Reagents       Purified   water   with   18.2   MΩ   resistance   was   used   for   all   experiments   to   eliminate   metal   ion   contamination.     Crude   C-­‐peptide   (Genscript,   Piscataway,   NJ)   and   its   E27A   mutant   were  purified  using  reverse  phase  high  performance  liquid  chromatography  (RP-­‐HPLC).     Fractions  were  analyzed  for  purity  using  liquid  chromatography-­‐mass  spectrometry  (LC-­‐ MS)  and  MS/MS.    The  fractions  containing  the  pure  peptides  were  lyophilized  overnight.     The  dried  product  was  then  weighed  to  0.25  mg/vial  and  stored  at  -­‐20°C  until  use.    To   prepare  a  stock  solution  of  8.3  μM  peptide,  one  vial  of  0.25  mg  peptide  was  thawed  and   dissolved  in  10  mL  of  deionized  water.  Usually,  this  solution  can  be  stored  for  a  month   at  4°C.    On  the  day  of  use,  for  either  C-­‐peptide  or  the  mutant,  a  working  solution  of  332   nM  was  prepared  in  pure  water.     Several   of   these   experiments   involved   the   radioligand   65 2+ Zn   (Perkin   Elmer,   Boston,   MA).     Because   of   the   relatively   short   half-­‐life   of   this   isotope,   the   concentration   of   the   stock   solution   was   calculated   regularly.     For   use,   the   stock   solution   was   diluted   in   water   2+ to   prepare   an   800   nM   working   solution.     Non-­‐radioactive   Zn   stock   solution   was   2+ prepared   by   dissolving   Zn   (II)   chloride   (Jade   Scientific,   Canton,   MI)   in   purified   water   and  diluted  to  800  nM.           86   RBCs   used   in   these   experiments   were   washed   and   incubated   in   physiological   salt   solution   (PSS)   containing,   in   mM,   4.7   KCl   (Fisher   Scientific,   Fair   Lawn,   NJ),   2.0   CaCl2   (Fisher   Scientific),   140.5   NaCl   (Columbus   Chemical   Industries,   Columbus,   WI),   12   MgSO4   (Fisher  Scientific),  21.0  tris(hydroxymethylaminomethane)  (Invitrogen,  Carlsbad,  CA),  5.6   glucose   (Sigma,   St.   Louis,   MO),   and   5%   bovine   serum   solution   (Sigma)   at   a   final   pH   of   7.4,  adjusted  with  hydrochloric  acid.     3.3.2  Collection  and  Preparation  of  Human  RBCs       Following  human  venipuncture  and  collection  into  heparinized  vacutainers  (BD,  Franklin   Lakes,   NJ),   whole   blood   was   centrifuged   for   10   minutes   at   500g   to   separate   the   RBCs   from  the  other  blood  components.    Following  the  removal  of  the  plasma  and  buffy  coat,   the  RBCs  were  washed  three  times  with  PSS  and  centrifuged.    After  the  final  washing,   the   supernatant   was   removed   and   the   hematocrit   of   the   RBCs   was   measured   using   a   CritSpin©  analyzer.     2+ 2+ 3.3.3   Radiolabeled   Zn   Assays   for   Determination   of   the   Amount   of   Zn   Interacting   with  the  RBC       2+ The  amount  of  Zn  interacting  with  the  RBC  in  the  presence  of  C-­‐peptide  or  the  mutant   was   determined   using   the   radioligand   65 2+ Zn .     First,   65 2+ Zn   and   the   peptide   were   incubated  in  pure  water  for  three  minutes  to  ensure  peptide  activation.    PSS  was  then   added  to  the  mixture,  immediately  followed  by  the  addition  of  the  appropriate  volume     87   of   RBCs   to   prepare   1   mL   samples   with   a   7%   hematocrit.     Following   a   one   hour   incubation   at   37°C,   samples   were   centrifuged   at   500g   for   four   minutes   and   the   supernatant   was   collected.     In   some   of   the   experiments,   the   RBCs   were   then   washed   three   times   and   lysed   with   bleach.     For   the   saturation   experiments,   the   RBCs   were   2+ incubated   a   second   time   in   PSS   containing   100   μM   Zn   to   displace   the   specifically   65 2+ bound   Zn .    After  a  one-­‐hour  incubation,  the  samples  were  centrifuged  at  500g  for  4   minutes   and   the   supernatant   was   removed   and   the   amount   of   radioactivity   in   the   sample  was  measured.    In  a  96-­‐well  plate,  200  μL  of  supernatant  or  lysate  solution  were   combined   with   100   μL   of   scintillation   cocktail   (optiphase   supermix,   Perkin   Elmer,   Waltham,   MA).     The   amount   of   radioactivity   in   each   sample   was   then   determined   using   a   1450   Microbeta   Plus   liquid   scintillation   counter   (Wallac,   Turku,   Finland).     A   diagram   of   the   experimental   set-­‐up   can   be   seen   in   figure   3.3.   Comparing   to   a   set   of   65 65 2+ Zn   2+ standards   prepared   in   RBC   supernatant   or   lysate,   the   concentration   of   Zn   in   each   65 2+ sample  was  quantified.    In  the  case  of  the  supernatant,  knowing  the  amount  of   Zn   added   to   the   original   RBC   sample   and   how   much   was   left   in   the   supernatant,   the   65 2+ amount  of   Zn  interacting  with  the  RBCs  was  determined  by  subtraction.     3.4  Results       65 2+ Figure  3.4  shows  the  results  of  the  first  study  involving   Zn  interacting  with  human   RBCs  in  the  presence  and  absence  of  C-­‐peptide.    Because  of  previous  studies  showing  an       88       Figure   3.3   –   Radiolabeled   Zn   Assays   –   Zn   and   the   peptide   were   incubated   together   in   pure   water   for   three   minutes   to   ensure   peptide   activation.     PSS   was   then   added  to  the  mixture,  immediately  followed  by  the  addition  of  the  appropriate  volume   of   RBCs   to   prepare   1   mL   samples   with   a   7%   hematocrit.     Following   a   one   hour   incubation  samples  were  centrifuged  and  the  supernatant  was  collected.    In  some  of  the   experiments,  the  RBCs  were  then  washed  three  times  and  lysed  with  bleach.    In  a  96-­‐ well   plate,   200   μL   of   supernatant   or   lysate   solution   was   combined   with   100   μL   of   scintillation  cocktail.    The  amount  of  radioactivity  in  each  sample  was  then  determined   using  a  1450  Microbeta  Plus  liquid  scintillation  counter.                 2+   65 89   2+ 65     2+ Figure  3.4  –   Zn  Remaining  in  the  Supernatent  after  Incubation  with  RBCs  –  For   each  set  of  bars,  when  C-­‐peptide  was  also  added,  it  was  added  in  a  1:1  ratio  with   65 2+ 65 2+ Zn .    As  is  evident  from  the  data,   Zn  alone,  black  bars,  does  not  interact  with  the   65 2+ RBC.    However,  with  the  exception  of  the  1  nM  concentration,  when   Zn  bound  to  C-­‐ 65 2+ peptide  is  incubated  with  the  RBCs,  grey  bars,  significantly  less   Zn  is  found  in  the   supernatant.    Error  is  represented  as  standard  deviation  for  N  =  7  humans.  The  asterisk   represents  p  <  0.05.   90   2+ 60 increase   in   ATP   release   from   RBCs   incubated   with   Zn   bound   to   C-­‐peptide,   it   was   thought   that   there   may   be   an   increase   interaction   of   65 2+ Zn   with   the   RBC.     Prior   to   2+ these   studies,   by   investigating   the   amount   of   Zn   measured   in   the   supernatant   and   65 2+ that  remaining  on  the  RBCs,  it  was  determined  that  all  of  the   Zn  could  be  accounted   for.     Using   the   supernatant   saves   time   and   increases   sensitivity,   so   for   the   majority   of   the   studies,   the   65 2+ Zn   was   measured   only   in   the   supernatant.     For   each   set   of   bars,   65 2+ when  C-­‐peptide  was  also  added,  it  was  added  in  a  1:1  ratio  with   Zn .    As  is  evident   from   the   data,   65 2+ Zn   alone   does   not   interact   with   the   RBC.     However,   with   the   65 2+ exception  of  the  1  nM  concentration,  when   Zn  bound  to  C-­‐peptide  is  incubated  with   65 2+ the  RBCs,  significantly  less   Zn  is  found  in  the  supernatant.     65 2+ For  comparison  with  the  amount  of  C-­‐peptide  interacting  with  the  RBC,   Zn  was  also   measured  on  the  RBC  directly.    Figure  3.5  shows  a  correlation  between  the  amount  of   65 2+ Zn   and   the   amount   of   C-­‐peptide   interacting   with   the   RBC.     On   the   left,   data   collected  by  Wathsala  Medawala  using  an  enzyme-­‐linked  immunosorbent  assay  (ELISA)   41 to   find   the   amount   of   C-­‐peptide   interaction   with   the   RBC   is   shown.     Comparing   this   65 2+ data   to   the   data   on   the   right   showing   the   amount   of   Zn   recovered   from   RBC   lysate,   2+ it  is  evident  that  C-­‐peptide  and  Zn  interaction  with  the  RBC  in  a  1:1  ratio.    For  both       91     Figure   3.5   –   C-­‐Peptide   and   Zn   on   the   RBC  –  The  data  on  the  left  shows  the  amount  of   2+ 41 C-­‐peptide  interaction  with  the  RBC.    Comparing  to  the  data  on  the  right,  showing  the   65 2+ 2+ amount   of   Zn   recovered   from   RBC   lysate,   it   is   evident   that   C-­‐peptide   and   Zn   interaction   with   the   RBC   in   a   1:1   ratio.     For   both   binding   curves,   saturation   seems   to   2+ begin  around  10  picomoles  of  added  Zn  bound  to  C-­‐peptide  and  results  in  a  maximum   65 2+ of  2.64  ±  1.03  picomoles  of  C-­‐peptide  and  2.96  ±  0.17  picomoles  of   Zn  interacting   2+ with  the  RBC  when  20  picomoles  of  Zn -­‐activated  C-­‐pepide  have  been  added.    Within   2+ error,  this  shows  that  Zn  and  C-­‐peptide  interact  with  the  RBC  in  a  1:1  manner.    Error  is   2+ shown  as  standard  deviation  for  N  =  4  humans  for  C-­‐peptide  and  N  =  5  for    Zn .     92   2+ binding   curves,   saturation   seems   to   begin   around   10   picomoles   of   added   Zn   bound   to   C-­‐peptide  and  results  in  a  maximum  of  2.64  ±  1.03  picomoles  of  C-­‐peptide  and  2.96  ±   65 2+ 2+ 0.17   picomoles   of   Zn   interacting   with   the   RBC   when   20   picomoles   of   Zn -­‐activated   2+ C-­‐pepide   have   been   added.     Within   error,   this   shows   that   Zn   and   C-­‐peptide   interact   2+ with   the   RBC   in   a   1:1   ratio.     Experiments   examining   different   Zn   to   C-­‐peptide   ratios   will  be  discussed  later.     2+ It  was  noted  during  experimentation  that  with  increased  amounts  of  Zn -­‐activated  C-­‐ 65 2+ peptide,  up  to  100  picomoles,  an  increased  amount  of   Zn  was  observed  interacting   with   the   RBCs,   while   the   amount   of   C-­‐peptide   remained   relatively   stable,   under   3   65 picomole.    To  discern  if  this  increased  amount  of   Zn 2+   was  the  result  of  non-­‐specific   binding,   where   the   binding   is   not   the   result   of   a   unique   arrangement   of   the   binding   partner,   a   saturation   study   was   completed.     RBCs   were   first   incubated   with   various   65 2+ concentrations   of   Zn -­‐activated   C-­‐peptide,   after   the   supernatant   was   removed   and   65 2+ 2+ measured   for   excess   Zn ,   the   RBCs   were   incubated   a   second   time   with   normal   Zn -­‐ activated  C-­‐peptide  at  a  concentration  10  times  greater  than  the  highest  concentration   65 2+ 65 2+ of   Zn  bound  to  C-­‐peptide.  The  non-­‐radioactive  competitor  displaces  only  the   Zn   that   had   been   specifically   bound.   The   non-­‐specific   binding   can   then   be   calculated   through  simple  subtraction.    The  results  of  this  can  be  seen  in  figure  3.6.    From  this,  it       93   Zinc involved in binding (pmol) 10 8 6 4 2 0 0 20 40 60 80 100 120 Zinc initially incubated with RBCs (pmol)     Figure   3.6   –   Saturation   Curve   –   After   total   and   specific   binding   (circles   and   triangle,   respectively)   were   found,   the   non-­‐specific   binding   (squares)   was   calculated.     2+ Comparable   to   the   C-­‐peptide   binding,   the   amount   of   specifically   bound   Zn   is   1.85   ±   65 2+ 0.53  picomoles  for  20  picomoles  of  added   Zn  bound  to  C-­‐peptide.    This  data  shows   2+ 2+ that  Zn ,  or  Zn  bound  to  C-­‐peptide,  is  interacting  with  a  receptor  on  the  RBC.    The   error  is  shown  as  standard  deviation  for  N  =  7  humans  for  2,  5,  10,  20  and  40  picomoles,   N  =  4  humans  for  80  picomoles,  and  N  =  5  humans  for  100  picomoles.     94   can   be   seen   that   there   was   non-­‐specific   binding   occurring,   and   comparable   to   the   C-­‐ 2+ peptide  binding,  the  amount  of  specifically  bound  Zn  is  1.85  ±  0.53  picomoles  for  20   65 2+ 2+ 2+ picomoles  of  added   Zn  bound  to  C-­‐peptide.    This  data  also  shows  that  Zn ,  or  Zn   bound  to  C-­‐peptide,  is  interacting  with  a  receptor  on  the  RBC.     Additionally,  Wathsala  Medawala  investigated  the  amount  of  C-­‐peptide  interacting  with   2+ the  RBC  with  and  without  Zn ,  to  compare  to  data  collected  showing  the  interaction  of   2+ Zn   with   the   RBC   with   and   without   C-­‐peptide.     In   figure   3.7,   the   data   shows   that   2+ whether   or   not   C-­‐peptide   is   activated   by   Zn ,   statistically   the   same   amount   is   found   to   41 2+ be  interacting  with  the  RBC.    However,  when  Zn  alone  is  added  to  an  RBC  solution,   2+ all   of   it   is   found   in   the   supernatant   after   incubation.     When   Zn   is   introduced   to   the   2+ 2+ sample  as  Zn  bound  to  C-­‐peptide,  significantly  less  Zn  was  found  in  the  supernatant,   2+ meaning   it   is   interacting   with   the   RBC,   but   only   when   added   to   the   sample   as   Zn   2+ bound  to  C-­‐peptide.  This  is  significant  as  it  was  previously  thought  that  Zn  was  helping   2+ C-­‐peptide  interact  with  the  RBC,  when  in  fact,  C-­‐peptide  is  carrying  Zn  to  the  RBC.     2+ The   binding   of   Zn   to   C-­‐peptide   and   several   mutants   been   studied   using   mass   66 41 spectrometry   and   CD   spectroscopy.     One   of   the   mutants,   called   E27A,   had   the   glutamic  acid  in    position  27  switched  for  an  alanine.      Wild  type  C-­‐peptide    was  found  to       95       2+ Figure  3.7  –  C-­‐Peptide  and  Zn  in  the  Supernatent  –    The  data  on  the  left  shows  that   2+ whether   or   not   20   nM   C-­‐peptide   is   bound   to   20   nM   Zn ,   statistically   the   same   amount   41 2+ is   found   to   be   interacting   with   the   RBC.     However,   when   Zn   alone   is   added   to   an   2+ RBC   solution,   all   of   it   is   found   in   the   supernatant   after   incubation.     When   Zn   is   2+ 2+ introduced   to   the   sample   as   20   nM   Zn   bound   to   C-­‐peptide,   significantly   less   Zn   was   found  in  the  supernatant,  meaning  it  is  interacting  with  the  RBC,  but  only  when  added   to  the  sample  bound  to  C-­‐peptide.    Error  is  standard  deviation  for  N  =  4  humans  for  C-­‐ 2+ 2+ peptide   and   N   =   5   humans   for   Zn .     The   asterisk   denotes   p   <   0.01   as   compared   to   Zn   added  alone.     96   2+ 2+ bind   to   Zn   in   a   1:1   ratio,   while   E27A   was   able   to   bind   four   Zn   ions   per   molecule.     2+ This  suggests  that  C-­‐peptide  would  have  a  more  closed  structure  when  bound  to  Zn ,   and  E27A  would  have  an  open  structure.    Figure  3.8  shows  the  effect  this  would  have  on   2+ 65 2+ Zn   delivery   to   the   RBC.     The   mutation   in   E27A   prevents   Zn   interaction   with   the   RBC   when   the   experimental   conditions   are   kept   the   same.     65 2+ Zn   and   the   peptides   were  added  in  a  1:1  ratio.    Because  binding  studies  suggested  that  E27A  has  the  ability   2+ 65 to  bind  Zn ,  other  ratios  of   Zn 2+   2+   to  peptide  were  investigated  for  Zn 65 delivery  to  the   2+ RBC.    In  normal  PSS,  shown  by  the  black  bars,  only  the  1:1  ratio  of   Zn  to  C-­‐peptide   65 2+ ratio  was  able  to  delivery  any   Zn  to  the  RBC.    One  of  the  components  of  PSS,  bovine   2+ 67 65 2+ serum  albumin  (BSA),  is  known  to  bind  Zn ,  and  if  the   Zn  ions  bound  to  the  E27A   are   not   bound   tightly,   they   would   be   easily   removed   by   the   BSA.     Therefore,   the   experiments   were   repeated   using   a   BSA-­‐free   PSS.     They   grey   bars   of   figure   3.8   show   2+ both   peptides   are   able   to   deliver   Zn   to   the   RBC,   regardless   of   ratio.     However,   C-­‐ 2+ peptide  is  able  to  deliver  rough  double  the  amount  of  Zn  as  compared  to  the  mutant.     3.5  Discussion       2+ The   major   finding   of   this   research   is   the   evidence   of   C-­‐Peptide   being   a   Zn   carrier   to   the  RBC.    It  was  already  known  that  C-­‐peptide  will  increase  the  ATP  release  from  RBCs       97     Zinc on the RBCs (pmol) 4 3 2 1 # # # 0 # # A Zn /E 27 A 3: 1 Zn 2: 1 Zn /E 27 A /E 27 P 1 1: 3: 1 Zn /C P /C Zn 2: 1 1: 1 Zn /C P -1   2+ Figure   3.8   –   Effect   of   Mutant,   Ratios,   and   BSA   on   Zn   Interaction   with   RBCs   –   The   mutation   in   E27A   prevents   65 2+ Zn   interaction   with   the   RBC   in   the   presences   of   BSA   (black   bars).     Also,   in   normal   PSS,   only   the   1:1   65 2+ 65 2+ Zn   to   C-­‐peptide   ratio   was   able   to   delivery   any   Zn   to   the   RBC.     When   the   experiments   were   repeated   using   a   BSA-­‐free   2+ PSS,   (grey   bars)   both   peptides   are   able   to   deliver   Zn   to   the   RBC,   regardless   of   ratio.     2+ However,  C-­‐peptide  is  able  to  deliver  roughly  double  the  amount  of  Zn  as  compared   to   the   mutant.     The   error   is   standard   deviation   for   N   =   4   humans.   The   asterisk   represents  p  <  0.05  and  the  pound  sign  represents  p  <  0.001  as  compared  to  1:1  ratio  of   2+ Zn  bound  to  C-­‐peptide.     98   2+ 60 only   when   previously   activated   by   Zn .   Studies   completed   by   Watshsala   Medawala   2+ 41 have   previously   shown   that   C-­‐peptide   does   in   fact   bind   to   Zn   in   a   1:1   manner.     It   2+ was  the  expectation  that  C-­‐peptide  would  need  Zn  to  interact  with  the  RBC,  however,   2+ it  was  found  that  it  is  Zn  that  needs  C-­‐peptide  for  RBC  interaction.    Figure  3.4  shows   2+ the   interaction   of   Zn   with   the   RBC   at   different   concentrations,   and   from   this,   it   is   2+ evident   that   Zn   will   not   interact   with   the   RBC   without   the   presence   of   C-­‐peptide.     Mimicking   in   vivo   conditions,   the   RBCs   were   suspended   in   PSS   containing   BSA.     Because   2+ 2+ of   known   BSA   and   Zn   binding,   it   is   not   surprising   that   Zn   in   the   solution   is   unable   to   reach  the  RBC  on  its  own.     2+ Upon   further   investigation,   it   was   found   that   after   incubation   with   Zn   bound   to   C-­‐ 2+ peptide,   the   amount   of   Zn   interacting   with   the   RBC   and   the   amount   of   C-­‐peptide   interaction  with  the  RBC  were  relatively  equivalent  with  approximately  2  picomoles  of   2+ each  interacting  with  the  RBC,  as  seen  in  figure  3.5.    Since  the  C-­‐peptide  and  Zn  are   acting   together,   it   is   reasonable   for   the   saturation   of   the   RBCs   to   occur   in   a   10   nM   2+ solution  of  Zn  bound  to  C-­‐peptide,  as  the  levels  of  C-­‐peptide  are  typically  in  the  single   digit  nanomolar  level  in  vivo.           99   Because  of  the  specific  and  non-­‐specific  binding  seen  in  figure  3.6,  it  is  believed  that  the   2+ Zn  is  binding  to  a  receptor  on  the  RBC.    While  there  is  currently  no  known  receptor  for   2+ C-­‐peptide,  it  is  known  that  Zn  has  an  effect  on  glucose  transport  in  certain  cell  types   68 that   is   believed   to   be   mediated   through   GLUT1,   the   major   glucose   transporter   in   2+ RBCs.     With   the   increase   in   RBC-­‐derived   ATP   after   incubation   with   Zn   bound   to   C-­‐ peptide,  it  is  believe  that  there  will  also  be  an  increase  in  glucose  uptake  into  the  RBC,   69 as   glycolysis   is   the   only   means   the   cell   has   to   produce   ATP.     Therefore,   it   would   be   2+ logical  for  the  mechanism  of  increase  to  begin  with  GLUT1.    Additionally,  acute  oral  Zn   supplementation   improved   glucose   use   in   vivo   in   response   to   an   intravenous   glucose   70 test   when   no   changes   in   insulin   secretion   on   sensitivity   were   seen.     This   suggests   that   2+ Zn  alone  was  responsible  for  the  increased  in  glucose  tolerance.    Further  studies  need   to  be  completed  to  investigate  the  mechanism  of  RBC  glucose  uptake  and  increase  ATP   2+ release  after  incubation  with  Zn -­‐activated  C-­‐peptide.         2+ When   starting   this   area   of   research,   it   was   hypothesized   that   C-­‐peptide   required   Zn   for  transport  to  the  RBC  or  for  its  activity  in  some  way.    However,  through  the  course  of   experimentation,  it  was  discovered  that  while  the  C-­‐peptide  can  interact  with  the  RBC   2+ 2+ without  Zn ,  the  Zn  cannot  interact  with  the  RBC  without  C-­‐peptide,  as  seen  in  figure   3.7.     While   it   has   been   shown   that   C-­‐peptide   interacts   with   the   RBC   without   the     100   2+ presence   of   Zn ,   it   is   known   that   the   increase   ATP   release   will   not   occur   unless   the   2+ RBCs  have  been  incubated  with  Zn  bound  to  C-­‐peptide.    This  has  led  to  the  idea  that   2+ C-­‐peptide  is  a  Zn  carrier  to  the  RBC.         2+ Previous   work   done   by   Wathsala   Medawala   shows   that   C-­‐peptide   binds   to   Zn   in   a   1:1   41 ratio.     Results   shown   in   figure   3.8   support   this   finding.     In   normal   PSS,   only   a   1:1   ratio   2+ 2+ of   Zn   to   C-­‐peptide   exhibits   Zn   interaction   with   the   RBC.     C-­‐peptide,   having   no   2+ histidine  residues,  is  assumed  to  bind  to  Zn  through  the  acidic  amino  acids,  glutamic   acid  and  aspartate.    These  amino  acids  bind  monodentate  through  one  carboxyl  oxygen.     2+ Because  of  this,  it  is  hypothesized  that  when  Zn  and  C-­‐peptide  are  combined  in  a  1:1   2+ ratio,  folding  of  the  C-­‐peptide  occurs,  binding  to  the  Zn  with  multiple  amino  acids.    A   minimalistic  diagram  of  this  theory  can  be  seen  in  figure  3.9.    Results  showing  the  lost  of   2+ 2+ Zn   interaction   with   the   RBC   with   higher   Zn   to   C-­‐peptide   ratios   led   to   the   hypothesis   2+ that   multiple   Zn   ions   may   be   binding   to   the   C-­‐peptide   in   those   cases.     Then,   upon   2+ addition  of  serum  albumin  containing  PSS,  the  Zn  may  be  removed  from  the  peptide   by  the  albumin  because  of  its  high  binding  affinity.    Further  proof  of  this  is  seen  in  figure   3.8,  when  the  E27A  mutant  of  the  peptide  was  used,  along  with  the  use  of  BSA-­‐free  PSS.     2+ 2+ When  E27A  is  incubated  with  Zn  prior  to  the  addition  of  normal  PSS  and  RBCs,  no  Zn       101       2+ Figure  3.9  –  Hypothesis  of  Zn  Binding  to  C-­‐peptide  and  a  Mutant  –     Results   showing   2+ 2+ the   loss   of   Zn   interaction   with   the   RBC   with   higher   Zn   to   C-­‐peptide   ratios   led   to   the   2+ hypothesis   that   multiple   Zn   ions   may   be   binding   to   the   C-­‐peptide   in   those   cases.     2+ Then,  upon  addition  of  serum  albumin  containing  PSS,  the  Zn  may  be  removed  from   the   peptide   by   the   albumin   because   of   its   high   binding   affinity.   It   is   believed   that   the   2+ 2+ Zn  may  bind  to  the  E27A  peptide,  however  there  is  no  folding,  and  the  Zn  ions  are   easily  removed  by  the  BSA.    This  hypothesis  is  related  to  the  data  in  figure  3.8.           102   th is   seen   on   the   RBCs.     This   was   expected,   as   the   glutamic   acid   in   the   27   position   has   been   shown   to   be   necessary   for   biological   effects   on   RBC-­‐derived   ATP   release.     It   is   2+ believed  that  the  Zn  may  bind  to  the  E27A  peptide,  however  there  is  no  folding,  and   2+ the   Zn   ions   are   easily   removed   by   the   BSA.     Further   proof   of   this   is   seen   when   the   2+ Zn -­‐activated  peptides  are  incubated  with  the  RBCs  in  a  BSA-­‐free  PSS.    Here,  the  data   2+ shows  that  for  C-­‐peptide,  approximately  2.4  picomoles  of  Zn  are  interacting  with  the   2+ 2+ RBC,   regardless   of   the   Zn   to   C-­‐peptide   ratio.     Additionally,   the   Zn   is   also   able   to   interact   with   the   RBC   after   incubation   with   the   mutant,   though   in   a   statistically   lower   quantity   of   approximately   1.3   picomoles.     Interestingly,   this   data   matches   up   with   binding  affinities  found  by  Wathsala  Medawala,  who  found  that  E27A  has  a  decrease  in   2+ 2+ Zn   binding   affinity   of   about   50%   as   compared   to   the   Zn   binding   affinity   for   C-­‐ 41 2+ peptide.    Because  of  these  results,  it  is  concluded  that  Zn  cannot  interact  with  the   RBC   without   the   presence   of   C-­‐peptide,   and   that   C-­‐peptide   and   the   mutant   can   carry   2+ 2+ Zn   to   the   RBC,   regardless   of   the   Zn   to   peptide   ratio   without   the   presence   of   BSA.     2+ However,   in   in   vivo   conditions,   with   serum   present,   C-­‐peptide   will   bind   to   Zn   ions   in   a   2+ 1:1  ratio  for  the  most  effective  delivery  of  Zn  to  the  RBC  for  biological  effects.     Unpublished   data   from   the   Spence   lab   shows   that   purified   C-­‐peptide   has   biological   2+ effects   only   when   co-­‐administered   with   Zn -­‐containing   insulin   and   not   when   co-­‐   103   2+ administered  with  purified  insulin  alone.    Specifically,  Zn  bound  to  C-­‐peptide  increases   RBC-­‐derived   ATP   and   subsequently,   NO   release   from   endothelial   cells   lining   the   blood   vessels   causing   vasodilation. 63-­‐65 2+     Therefore,   Zn   bound   to   C-­‐peptide   has   great   potential  in  the  relief  of  the  vascular  complications  of  diabetes.    While  this  has  obvious   2+ implications   for   diabetes   research   and   potential   treatment,   Zn   is   also   implicated   in   other  diseases,  such  as  multiple  sclerosis  (MS),  as  discussed  in  chapter  1.6.    Because  of   2+ the   results   presented   here,   Zn   bound   to   C-­‐peptide   was   used   in   future   studies   to   2+ investigate   the   differences   in   Zn   interaction   with   the   RBC   between   MS   patients   and   healthy  controls.           104   REFERENCES   105       1     2     3     4     5     6     7     8     9       REFERENCES   Gavin,   J.   R.,   Alberti,   K.   G.   M.   M.,   Davidson   M.   B.,   DeFronzo,   R.   A.,   Drash,   A.,   Gabbe,   S.   G.,   Genuth,   S.,   Harris,   M.   I.,   Kahn,   R.,   Keen,   H.,   Knowler,   W.   C.,   Lebovitz,  H.,  Maclaren,  N.  K.,  Palmer,  J.  P.,  Raskin,  P.,  Rizza,  R.  A.  &  Stern,  M.  P.   Report   of   the   expert   committee   on   the   diagnosis   and   classification   of   diabetes   mellitus.  Diabetes  Care  22,  S5-­‐S19  (1999).   Alberti,  K.,  Zimmet,  P.,  Consultation,  W.  &  Consultation,  W.  Definition,  diagnosis   and  classification  of  diabetes  mellitus  and  its  complications  part  1:  Diagnosis  and   classification   of   diabetes   mellitus   -­‐   Provisional   report   of   a   WHO   consultation.   Diabetic  Medicine  15,  539-­‐553  (1998).   Vaarala,  O.,  Knip,  M.,  Paronen,  J.,  Hämäläinen,  A.  M.,  Muona,  P.,  Väätäinen,  M.,   Ilonen,   J.,   Simell,   O.   &   Akerblom,   H.   K.   Cow's   milk   formula   feeding   induces   primary   immunization   to   insulin   in   infants   at   genetic   risk   for   type   1   diabetes.   Diabetes  48,  1389-­‐1394  (1999).   Akerblom,   H.   K.,   Virtanen,   S.   M.,   Ilonen,   J.,   Savilahti,   E.,   Vaarala,   O.,   Reunanen,   A.,   Teramo,   K.,   Hämäläinen,   A.   M.,   Paronen,   J.,   Riikjärv,   M.   A.,   Ormisson,   A.,   Ludvigsson,   J.,   Dosch,   H.   M.,   Hakulinen,   T.   &   Knip,   M.   Dietary   manipulation   of   beta   cell   autoimmunity   in   infants   at   increased   risk   of   type   1   diabetes:   a   pilot   study  Diabetologia  48,  1676-­‐1676  (2005).   Fowler,   M.   J.   Diabetes:   Magnitude   and   Mechanism.   Clinical   Diabetes   25,   25-­‐28   (2007).   Lebovitz,   H.   Insulin   resistance:   definition   and   consequences.   Experimental   and   Clinical  Endocrinology  &  Diabetes  109,  S135-­‐S148  (2001).   Polansky,   K.   S.   Dynamics   of   Insulin   Secretion   in   Obesity   and   Diabetics.   International   Journal   of   Obesity   and   Related   Metabolic   Disorders   24,   S29-­‐31   (2000).   Mokdad,  A.  H.,  Bowman,  B.  A.,  Ford,  E.  S.,  Vinicor,  F.,  Marks,  J.  S.  &  Koplan,  J.  P.   The  continuing  epidemics  of  obesity  and  diabetes  in  the  United  States.  Journal  of   the  American  Medical  Association  286,  1195-­‐1200  (2001).   Associat,  A.  D.  &  Associat,  A.  D.  Diagnosis  and  classification  of  diabetes  mellitus.   Diabetes  Care  31,  S55-­‐S60  (2008).   106   10     11     12     13     14     15     16     17     18     19     20       Creager,   M.,   Luscher,   T.,   Cosentino,   F.   &   Beckman,   J.   Diabetes   and   vascular   disease   -­‐   Pathophysiology,   clinical   consequences,   and   medical   therapy:   Part   I.   Circulation  108,  1527-­‐1532  (2003).   Libby,  P.  Coronary  artery  injury  and  the  biology  at  atherosclerosis:  Inflammation,   thrombosis,  and  stabilization.  American  Journal  of  Cardiology  86,  3J-­‐8J  (2000).   Carroll,  J.,  Raththagala,  M.,  Subasinghe,  W.,  Baguzis,  S.,  D'amico  Oblak,  T.,  Root,   P.  &  Spence,  D.  An  altered  oxidant  defense  system  in  red  blood  cells  affects  their   ability   to   release   nitric   oxide-­‐stimulating   ATP.   Molecular   Biosystems   2,   305-­‐311   (2006).   Sprague,   R.,   Stephenson,   A.,   Bowles.   E.,   Stumpf,   M.,   Ricketts,   G.   &   Lonigro,   A.   Expression   of   the   heterotrimeric   G   protein   Gi   and   ATP   release   are   impaired   in   erythrocytes   of   humans   with   diabetes   mellitus.   Hypoxia   and   Exercise   588,   207-­‐ 216  (2006).   Johnstone,  M.  T.,  Creager,  S.  J.,  Scales,  K.  M.,  Cusco,  J.  A.,  Lee,  B.  K.  &  Creager,  M.   A.   Impaired   endothelium-­‐dependent   vasodilation   in   patients   with   insulin-­‐ dependent  diabetes  mellitus.  Circulation  88,  2510-­‐2516  (1993).   Icks,   A.,   Trautner,   C.,   Haastert,   B.,   Berger,   M.   &   Giani,   G.   Blindness   due   to   diabetes:   Population-­‐based   age-­‐   and   sex-­‐specific   incidence   rates.   Diabetic   Medicine  14,  571-­‐575  (1997).   Sheetz,   M.   &   King,   G.   Molecular   understanding   of   hyperglycemia's   adverse   effects   for   diabetic   complications.   Journal   of   the   American   Medical   Association   288,  2579-­‐2588  (2002).   The   Diabetes   Control   and   Complications   Trial   Research   Group.   The   effect   of   intensive   treatment   of   diabetes   on   the   development   and   progression   of   long-­‐ term  complications  in  insulin-­‐dependent  diabetes  mellitus.  New  England  Journal   of  Medicine  329,  977-­‐986  (1993).   Sima,   A.   The   heterogeneity   of   diabetic   neuropathy.   Frontiers   in   Bioscience   13,   4808-­‐4816  (2008).   Sugimoto,   K.,   Murakawa,   Y.   &   Sima,   A.   Diabetic   neuropathy   -­‐   a   continuing   enigma.  Diabetes-­‐Metabolism  Research  and  Reviews  16,  408-­‐433  (2000).   Boulton,  A.  Neuropathy  and  diabetes  -­‐  Introduction.  Diabetes  Reviews  7,  235-­‐236   (1999).   107   21     22     23     24     25     26     27     28     29     30     31     32       Aoun,  S.,  Blacher,  J.,  Safar,  M.  &  Mourad,  J.  Diabetes  mellitus  and  renal  failure:   effects   on   large   artery   stiffness.   Journal   of   Human   Hypertension   15,   693-­‐700   (2001).   Ritz,   E.,   Rychlik,   I.,   Locatelli,   F.   &   Halimi,   S.   End-­‐stage   renal   failure   in   type   2   diabetes:   A   medical   catastrophe   of   worldwide   dimensions.   American   Journal   of   Kidney  Diseases  34,  795-­‐808  (1999).   Helal,   I.,   Fick-­‐Brosnahan,   G.,   Reed-­‐Gitomer,   B.   &   Schrier,   R.   Glomerular   hyperfiltration:  definitions,  mechanisms  and  clinical  implications.  Nature  Reviews   Nephrology  8,  293-­‐300  (2012).   Tervaert,   T.   W.,   Mooyaart,   A.   L.,   Amann.   K.,   Cohen,   A.   H.,   Cook,   H.   T.,   Drachenberg,  C.  B.,  Ferrario,  F.,  Fogo,  A.  B.,  Haas,  M.,  de  Heer,  E.,  Joh,  K.,  Noël,  L.   H.,   Radhakrishnan,   J.,   Seshan,   S.   V.,   Bajema,   I.   M.   &   Bruijn,   J.   A.   Pathologic   Classification   of   Diabetic   Nephropathy.   Journal   of   the   American   Society   of   Nephrology  21,  556-­‐563  (2010).   Schleicher,  E.,  Kolm,  V.,  Ceol,  M.  &  Nerlich,  A.  Structural  and  functional  changes   in   diabetic   glomerulopathy.   Kidney   &   Blood   Pressure   Research   19,   305-­‐315   (1996).   Cryer,   P.   Hypoglycemia   is   the   limiting   factor   in   the   management   of   diabetes.   Diabetes-­‐Metabolism  Research  and  Reviews  15,  42-­‐46  (1999).   Ahmed,   A.   History   of   diabetes   mellitus.   Saudi   Medical   Journal   23,   373-­‐378   (2002).   Poretsky,  L.  Principles  of  diabetes  mellitus.  2nd  ed.,    (Springer,  2010).   Orci,   L.   Patterns   of   cellular   and   subcellular   organization   in   the   endocrine   pancreas  -­‐  The  Henry  Dale  Lecture  for  1983.  Journal  of  Endocrinology  102,  3-­‐4,   (1984).   Hutton,   J.   Insulin   secretory   granule   biogenesis   and   the   proinsulin-­‐processing   endopeptidases.  Diabetologia  37,  S48-­‐S56  (1994).   Chimienti,   F.,   Favier,   A.   &   Seve,   M.   ZnT-­‐8,   a   pancreatic   beta-­‐cell-­‐specific   zinc   transporter.  Biometals  18,  313-­‐317  (2005).   Hutton,  J.  The  insulin  secretory  granule.  Diabetologia  32,  271-­‐281  (1989).   108   33     34     35     36     37     38     39     40     41     42       Emdin,  S.,  Dodson,  G.,  Cutfield,  J.  &  Cutfield,  S.  Role  of  zinc  in  insulin-­‐biosynthesis   -­‐  some  possible  zinc-­‐insulin  interactions  in  the  pancreatic  beta-­‐cell.   Diabetologia   19,  174-­‐182  (1980).   Vanschaftingen,   E.,   Detheux,   M.   &   Dacunha,   M.   Short-­‐term   control   of   glucokinase   activity   -­‐   role   of   a   regulatory   protein.   Faseb   Journal   8,   414-­‐419   (1994).   + Cook,   D.   &   Hales,   C.   Intracellular   ATP   directly   blocks   K   channels   in   pancreatic   beta-­‐cells.  Nature  311,  271-­‐273  (1984).   Wollheim,  C.  &  Sharp,  G.  Regulation  of  insulin  release  by  calcium.  Physiological   Reviews  61,  914-­‐973  (1981).   2+ Jones,  P.,  Persaud,  S.  &  Howell,  S.  Time-­‐course  of  Ca -­‐induced  insulin  secretion   from   perifused,   elecrically   permeablised   Islets   of   Langerhans   -­‐   effects   of   cAMP   and  a  phorbol  ester.  Biochemical  and  Biophysical  Research  Communications  162,   998-­‐1003  (1989).   Permutt,  M.  A.  The  Islets  of  Langerhans.    (Academic  Press,  1981).   Henriksson,   M.,   Nordling,   E.,   Melles,   E.,   Shafqat,   J.,   Ståhlberg,   M.,   Ekberg,   K.,   Persson,   B.,   Bergman,   T.,   Wahren,   J.,   Johansson,   J.   &   Jörnvall,   H.   Separate   functional  features  of  proinsulin  C-­‐peptide.  Cellular  and  Molecular  Life  Sciences   62,  1772-­‐1778  (2005).   Henriksson,   M.,   Shafqat,   J.,   Liepinsh,   E.,   Tally,   M.,   Wahren,   J.,   Jörnvall,   H.   &   Johansson,   J.   Unordered   structure   of   proinsulin   C-­‐peptide   in   aqueous   solution   and  in  the  presence  of  lipid  vesicles.  Cellular  and  Molecular  Life  Sciences  57,  337-­‐ 342  (2000).   Medawala,  W.  Mechanistic  and  Functional  Studies  of  Zinc  Activation  of  C-­‐Peptide   and   its   Effect   on   Red   Blood   Cell   Metabolism   PhD   thesis,   Michigan   State   University,  (2011).   Rigler,  R.,  Pramanik,  A.,  Jonasson,  P.,  Kratz,  G.,  Jansson,  O.  T.,  Nygren,  P.  A.,  Stahl,   S.,   Ekberg,   K.,   Johansson,   B.   L.,   Uhlen,   S.,   Uhlen,   M.,   Jornvall,   H.   &   Wahren,   J.   Specific  binding  of  proinsulin  C-­‐peptide  to  human  cell  membranes.  Proceedings   of  the  National  Academy  of  Sciences  of  the  United  States  of  America  96,  13318-­‐ 13323  (1999).   109   43     44     45     46     47     48     49     50     51     52       Vague,   P.,   Dufayet,   D.,   Coste,   T.,   Moriscot,   C.,   Jannot,   M.   F.   &   Raccah,   D.   Association   of   diabetic   neuropathy   with   Na/K   ATPase   gene   polymorphism.   Diabetologia  40,  506-­‐511  (1997).   Pieper,  G.,  Siebeneich,  W.,  MooreHilton,  G.  &  Roza,  A.  Reversal  by  L-­‐arginine  of  a   dysfunctional   arginine   nitric   oxide   pathway   in   the   endothelium   of   the   genetic   diabetic  BB  rat.  Diabetologia  40,  910-­‐915  (1997).   McNally,   P.   G.,   Watt,   P.   A.,   Rimmer,   T.,   Burden,   A.   C.,   Hearnshaw,   J.   R.   &   Thurston,   H.   Impaired   contraction   and   endothelium-­‐dependent   relaxation   in   isolated   resistance   vessels   from   patients   with   insulin-­‐dependent   diabetes   mellitus.  Clinical  Science  87,  31-­‐36  (1994).   Kitamura,   T.,   Kimura,   K.,   Makondo,   K.,   Furuya,   D.   T.,   Suzuki,   M.,   Yoshida,   T.   &   Saito,   M.   Proinsulin   C-­‐peptide   increases   nitric   oxide   production   by   enhancing   mitogen-­‐activated   protein-­‐kinase-­‐dependent   transcription   of   endothelial   nitric   oxide   synthase   in   aortic   endothelial   cells   of   Wistar   rats.   Diabetologia   46,   1698-­‐ 1705  (2003).   Al-­‐Rasheed,   N.   M.,   Meakin,   F.,   Royal,   E.   L.,   Lewington,   A.   J.,   Brown,   J.,   Willars,   G.   B.  &  Brunskill,  N.  J.  Potent  activation  of  multiple  signalling  pathways  by  C-­‐peptide   in  opossum  kidney  proximal  tubular  cells.  Diabetologia  47,  987-­‐997  (2004).   Kitamura,  T.,  Kimura,  K.,  Jung,  B.  D.,  Makondo,  K.,  Sakane,  N.,  Yoshida,  T.  &  Saito,   M.   Proinsulin   C-­‐peptide   activates   cAMP   response   element-­‐binding   proteins   through   the   p38   mitogen-­‐activated   protein   kinase   pathway   in   mouse   lung   capillary  endothelial  cells.  Biochemical  Journal  366,  737-­‐744  (2002).   Steiner,   D.,   Cunningham,   D.,   Spigelman,   L.   &   Aten,   B.   Insuling   biosynthesis   -­‐   evidence  for  a  precursor.  Science  157,  697-­‐700  (1967).   Ko,  A.,  Smyth,  D.,  Markussen,  J.  &  Sundby,  F.  Amino  acid  sequence  of  C-­‐peptide   of  human  proinsulin.  European  Journal  of  Biochemistry  20,  190-­‐199  (1971).   Horwitz,  D.,  Starr,  J.,  Mako,  M.,  Blackard,  W.  &  Rubenstein,  A.  Proinsulin,  insulin,   and   C-­‐peptide   concentrations  in   human  portal  and  peripheral  blood.  Journal  of   Clinical  Investigation  55,  1278-­‐1283  (1975).   Pramanik,  A.,  Ekberg,  K.,  Zhong,  Z.,  Shafqat,  J.,  Henriksson,  M.,  Jansson,  O.,  Tibell,   A.,  Tally,  M.,  Wahren,  J.,  Jörnvall,  H.,  Rigler,  R.  &  Johansson,  J.  C-­‐peptide  binding   to   human   cell   membranes:   Importance   of   Glu27.   Biochemical   and   Biophysical   Research  Communications  284,  94-­‐98  (2001).   110   53   Forst,   T.,   De   La   Tour,   D.   D.,   Kunt,   T.,   Pfützner,   A.,   Goitom,   K.,   Pohlmann,   T.,   Schneider,  S.,  Johansson,  B.  L.,  Wahren,  J.,  Löbig,  M.,  Engelbach,  M.,  Beyer,  J.  &   Vague,   P.   Effects   of   proinsulin   C-­‐peptide   on   nitric   oxide,   microvascular   blood   + + flow   and   erythrocyte   Na ,K -­‐ATPase   activity   in   diabetes   mellitus   type   I.   Clinical   Science  98,  283-­‐290  (2000).     54     55     56     57     58     59     60     61     62       Meyer,   J.   A.,   Subasinghe,   W.,   Sima,   A.   A.,   Keltner,   Z.,   Reid,   G.   E.,   Daleke,   D.   &   Spence,   D.   M.   Zinc-­‐activated   C-­‐peptide   resistance   to   the   type   2   diabetic   erythrocyte   is   associated   with   hyperglycemia-­‐induced   phosphatidylserine   externalization  and  reversed  by  metformin.  Molecular  Biosystems  5,  1157-­‐1162   (2009).   Steiner,   D.   F.   in   Diabetes   &   C-­‐Peptide       (ed   Anders   A.F.   Sima)     (Humana   Press,   2012).   Wahren,   J.,   Johansson,   B.   &   Wallberg-­‐Henriksson,   H.   Does   C-­‐peptide   have   a   physiological  role?  Diabetologia  37,  S99-­‐S107  (1994).   Wahren,   J.   C-­‐peptide:   new   findings   and   therapeutic   implications   in   diabetes.   Clinical  Physiology  and  Functional  Imaging  24,  180-­‐189  (2004).   Forst,  T.  &  Kunt,  T.  Effects  of  C-­‐peptide  on  microvascular  blood  flow  and  blood   hemorheology.  Experimental  Diabesity  Research  5,  51-­‐64  (2004).   Jensen,  M.  &  Messina,  E.  C-­‐peptide  induces  a  concentration-­‐dependent  dilation   of   isolated   skeletal   muscle   arterioles   only   in   the   presence   of   insulin.   Faseb   Journal  13,  A88-­‐A88  (1999).   Meyer,   J.   A.,   Froelich,   J.   M.,   Reid,   G.   E.,   Karunarathne,   W.   K.   &   Spence,   D.   M.   Metal-­‐activated  C-­‐peptide  facilitates  glucose  clearance  and  the  release  of  a  nitric   oxide  stimulus  via  the  GLUT1  transporter.  Diabetologia  51,  175-­‐182  (2008).   Johansson,   B.,   Kernell,   A.,   Sjoberg,   S.   &   Wahren,   J.   Influence   of   combined   C-­‐ peptide   and   insulin   administration   on   renal   function   and   metabolic   control   in   diabetes   type   1.   Journal   of   Clinical   Endocrinology   &   Metabolism   77,   976-­‐981   (1993).   Shafqat,  J.,  Melles,  E.,  Sigmundsson,  K.,  Johansson,  B.  L.,  Ekberg,  K.,  Alvelius,  G.,   Henriksson,   M.,   Johansson,   J.,   Wahren,   J.   &   Jörnvall,   H.   Proinsulin   C-­‐peptide   elicits   disaggregation   of   insulin   resulting   in   enhanced   physiological   insulin   effects.  Cellular  and  Molecular  Life  Sciences  63,  1805-­‐1811  (2006).   111   63     64     65     66     67     68     69     70           Sprague,  R.  S.,  Ellsworth,  M.  L.,  Stephenson,  A.  H.  &  Lonigro,  A.  J.  ATP:  The  red   blood   cell   link   to   NO   and   local   control   of   the   pulmonary   circulation.   American   Journal   of   Physiology-­‐Heart   and   Circulatory   Physiology   271,   H2717-­‐H2722   (1996).   Oblak,   T.   D.   A.,   Root,   P.   &   Spence,   D.   M.   Fluorescence   Monitoring   of   ATP-­‐ Stimulated,   Endothelium-­‐Derived   Nitric   Oxide   Production   in   Channels   of   a   Poly(dimethylsiloxane)-­‐Based   Microfluidic   Device.   Analytical   Chemistry   78,   3193-­‐ 3197  (2006).   Sprague,   R.   S.,   Olearczyk,   J.   J.,   Spence,   D.   M.,   Stephenson,   A.   H.,   Sprung.   R.   W.   &   Lonigro,   A.   J.   Extracellular   ATP   signaling   in   the   rabbit   lung:   erythrocytes   as   determinants   of   vascular   resistance.   American   Journal   of   Physiology-­‐Heart   and   Circulatory  Physiology  285,  H693-­‐H700  (2003).   Keltner,  Z.,  Meyer,  J.  A.,  Johnson,  E.  M.,  Palumbo,  A.  M.,  Spence,  D.  M.  &  Reid,  G.   E.  Mass  spectrometric  characterization  and  activity  of  zinc-­‐activated  proinsulin  C-­‐ peptide  and  C-­‐peptide  mutants.  Analyst  135,  278-­‐288  (2010).   Masuoka,  J.  &  Saltman,  P.  Zinc  (II)  and  copper  (II)  binding  to  serum  albumin  -­‐  a   comparative   study   of   dog,   bovine,   and   human   albumin.   Journal   of   Biological   Chemistry  269,  25557-­‐25561  (1994).   Shay,   N.   F.   &   Tang,   X.   H.   Zinc   has   an   insulin-­‐like   effect   on   glucose   transport   mediated   by   phosphoinositol-­‐3-­‐kinase   and   AKT   in   3T3-­‐L1   fibroblasts   and   adipocytes.  Journal  of  Nutrition  131,  1414-­‐1420  (2001).   Harmening,  D.  Clinical  Hematology  and  Fundamentals  of  Hemostasis.    (F.  A.  Davis   Co,  2009).   Brun,   J.   F.,   Guintrand-­‐Hugret,   R.,   Fons,   C.,   Carvajal,   J.,   Fedou,   C.,   Fussellier,   M.,   Bardet,  L.  &  Orsetti,  A.  Effects  of  oral  zinc  gluconate  on  glucose  effectiveness  and   insulin   sensitivity   in   humans.   Biological   Trace   Element   Research   47,   385-­‐391   (1995).     112   Chapter  4  –  Potential  Multiple  Sclerosis  Biomarkers       4.1  MS  Diagnosis  and  the  Need  for  Biomarkers       Because   of   the   difficulty   of   diagnosis,   the   search   for   biomarkers   for   MS   has   been   ongoing   since   1922,   when   an   abnormality   was   first   discovered   in   the   CSF   of   MS   1 patients.    Even  after  almost  100  years  of  investigation,  there  is  still  no  biomarker  used   2 clinically.    Biomarkers  are  defined  as  characteristics  that  can  be  objectively  measured   to  be  evaluated  as  an  indicator  of  a  biological  process,  normal  or  pathogenic,  or  as  an   3 indicator   of   response   to   pharmaceuticals.     In   other   words,   biomarkers   should   both   correlate   with   how   a   patient   feels,   functions,   and   survives   and   with   the   effects   of   4 treatment   on   these   areas.     Because   of   the   complexity   of   MS,   it   is   unlikely   that   one   biomarker  will  be  able  to  encompass  the  full  effect  of  the  disease,  but  rather  one  of  the   many  ongoing  processes.      As  a  result,  there  are  four  proposed  biomarker  categories  for   MS:     diagnostic   biomarkers,   predictive   biomarkers,   process-­‐specific   biomarkers   and   treatment-­‐related  biomarkers.         A   reliable   diagnostic   biomarker   for   MS   would   discern   the   difference   between   healthy   persons,   MS   patients,   and   patients   with   other   neurological,   autoimmune   or   5 inflammatory   diseases.   Currently,   IgG   oligoclonal   bands   are   used   for   diagnostics,   as   described   in   chapter   1,   however   they   are   not   specific   enough   to   be   considered   as   a   biomarker  for  the  disease.    Evidence  of  IgG  oligoclonal  bands  in  the  CSF,  but  not  in  blood     113   serum,  denotes  a  local  immune  response.    While  this  is  often  seen  in  MS  patients,  it  can   also  be  seen  in  a  variety  of  both  inflammatory  and  non-­‐inflammatory  disorders  as  well.     Additionally,  it  is  unknown  if  oligoclonal  bands  are  involved  in  the  pathogenesis  of  the   disease,  or  if  they  are  a  product  of  the  disease,  and  therefore  not  a  reflection  of  disease   2 prognosis.     Predictive  biomarkers  in  MS  would  be  used  to  determine  if  a  patient  with  CIS  would  go   on   to   develop   the   disease   during   the   first   clinical   manifestation   or   early   phase   of   the   5 disease.     In   addition   to   being   diagnostic,   IgG   has   also   been   suggested   to   have   the   potential   to   be   a   predictive   biomarker   as   well.     The   presence   of   IgG   in   the   CSF   has   been   considered  as  an  indicator  of  the  development  of  MS  in  CIS  patients.    While  an  increased   risk  of  disease  development  has  been  noted  in  CIS  patients  with  IgG  presence,  it  is  not  a   6 strong   enough   correlation   to   be   considered   a   biomarker.     In   2012,   Avsar   et   al.   investigated  five  proteins  that  have  been  suggested  as  potential  predictive  biomarkers   in  the  CSF:  myelin  oligodendrocyte  glycoprotein  (MOG),  myelin  basic  protein  (MBP),  tau,   glial   fibrillary   acidic   protein   (GFAP),   and   neurofilament   light   chain   (NFL).     MOG   and   MBP   are  both  involved  in  the  myelination  of  nerves  and  tau,  GFAP  and  NFL  are  all  known  to   be   associated   with   neuronal   damage.     The   investigators   were   seeking   to   predict   if   patients   with   CIS   would   develop   MS   and   to   be   able   to   diagnose   between   the   types   of   MS   and   determine   prognosis.     It   was   found   that   MOG,   tau,   GFAP   and   NFL   were   only   successful  predictors  of  disease  classification  and  prognosis  when  considered  together,     114   and  further  investigation  is  needed  in  larger  scale  studies  to  attempt  to  confirm  these   2 findings.     Process-­‐specific   biomarkers   are   not   disease   specific,   but   can   be   indicative   of   different   processes  occurring  in  the  disease.    In  MS,  the  processes  that  have  been  measured  with   biomarkers   include:   inflammation,   demyelination,   oxidative   stress,   remyelination,   and   neuroaxonal  damage.    While  these  biomarkers  do  not  help  with  diagnostics,  they  can  be   useful  to  measure  disease  course.         The   final   type   of   biomarker   used   to   study   MS   is   the   treatment   response   biomarker.     Genes,   mRNA,   and   protein   levels   can   be   used   to   test   the   effectiveness   of   drug   5 therapies.     Most   of   the   research   in   this   area   to   date   has   focused   on   IFNs   and   the   response   to   treatment   as   a   function   of   genetic   variants.     These   studies   are   aimed   at   determining  the  difference  in  effectiveness  of  the  treatment  in  individual  patients.    If  it   can   be   determined   that   a   certain   gene   allele   makes   patients   more   responsive   to   IFN   treatment,   that   information   can   be   used   to   aid   in   the   decision   of   the   course   of   5,7 treatment,  helping  the  patient  sooner,  as  well  as  saving  time  and  money.     While  recent  biomarker  findings  are  promising,  most  use  CSF  for  testing  and  this  is  an   invasive   procedure.     CSF   is   traditionally   believed   to   be   the   most   likely   source   of   a   MS   biomarker   because   of   the   proximity   to   the   inflammation   and   lesions   in   the   CNS.     This   may  not  be  the  case,  as  CSF  is  collected  through  a  lumbar  puncture  in  the  lower  back   and  may  not  accurately  reflect  inflammatory  markers  in  the  brain  regions  where  most     115   MS  lesions  occur.    Because  of  the  invasiveness  of  CSF  collection,  sampling  can  only  be   performed   for   a   limited   number   of   time   points.     Additionally,   the   production   of   CSF   8 varies  with  the  time  of  day  and  any  results  would  therefore  need  to  be  standardized.     As   previously   discussed,   MS   patients   have   been   shown   to   have   an   increase   in   RBC-­‐ derived   ATP   release.     The   studies   presented   here   will   confirm   those   studies   as   well   as   show   that   the   increase   is,   in   fact,   due   to   mechanical   deformation   as   a   result   of   shear   stress.    Based  on  previous  research  investigating  the  role  of  zinc  in  MS,  the  basal  level  of   2+ Zn   was   measured   in   order   to   show   that   it   is   higher   in   MS   patients   than   in   healthy   controls,   with   controls   to   ensure   the   same   number   of   cells   was   being   studied   in   each   65 2+ sample.  In  addition,  the  amount  of   Zn  and  the  amount  of  C-­‐peptide  interaction  with   2+ the   RBC   were   also   measure   to   demonstrate   the   effect   of   Zn   delivery   to   the   RBC.     Finally,   because   of   the   increased   ATP   release   from   RBCs   and   the   potential   for   altered   glucose   metabolism   in   MS   patients,   the   amount   of   glucose   uptake   into   the   RBC   was   measured  and  was  expected  to  be  higher  than  that  of  normal  controls  to  keep  up  with   the  demand  for  ATP  release.           116   4.2  Experimental       4.2.1  Preparation  of  Reagents       2+ Purified   water   with   18.2   MΩ   resistance   was   used   for   all   experiments.     Zn   stock   solution   was   prepared   by   dissolving   zinc   (II)   chloride   (Jade   Scientific,   Canton,   MI)   in   purified   water   and   diluting   to   400   nM.     Crude   C-­‐peptide   (Genscript,   Piscataway,   NJ)   was   purified   using   reverse   phase   high   performance   liquid   chromatography   (RP-­‐HPLC)   and   dried.    The  purified  C-­‐peptide  was  dissolved  in  purified  water  and  diluted  to  a  400  nM   2+ 2+ working  solution.    When  Zn  bound  to  C-­‐peptide  was  added  to  samples,  the  Zn  and   C-­‐peptide  solutions  were  mixed  in  water  before  the  addition  of  any  other  components   or   buffers.     Physiological   salt   solution   (PSS)   was   prepared,   containing,   in   mM,   4.7   KCl   (Fisher   Scientific,   Fair   Lawn,   NJ),   2.0   CaCl2   (Fisher   Scientific),   140.5   NaCl   (Columbus   Chemical   Industries,   Columbus,   WI),   12   MgSO4   (Fisher   Scientific),   21.0   tris(hydroxymethylaminomethane)   (Invitrogen,   Carlsbad,   CA),   5.6   glucose   (Sigma,   St.   Louis,   MO),   and   5%   bovine   serum   solution   (Sigma)   at   a   final   pH   of   7.4,   adjusted   with   hydrochloric   acid.     ATP   solutions   were   prepared   by   first   dissolving   ATP   (Sigma)   in   purified  water  to  a  concentration  of  100  µM.    Standards  were  then  prepared  by  dilution   of   the   stock   in   buffer   to   concentrations   of   0   to   1   μM.     For   the   incubations   involving   glybenclamide,   a   10   mM   working   solution   was   prepared   dissolving   48   mg   of   glybenclamide   (Sigma)   in   a   0.1   M   sodium   hydroxide   (Fisher   Scientific)   solution   also   containing  50  mg/mL  dextrose  (Sigma).    The  glybenclamide  was  diluted  to  100  µM  in  the     117   RBC   samples.   A   solution   of   luciferin/luciferase   was   prepared   by   dissolving   2   mg   of   luciferin  (Gold  Biotechology,  St.  Louis,  MO)  in  5  mL  DDW,  and  transferring  this  to  a  vial   containing   100   mg   of   luciferase   (Sigma,   St.   Louis,   MO).     Saline   was   prepared   by   dissolving   9   g   of   NaCl   in   1   L   of   purified   water.     The   Drabkin’s   solution   used   for   the   determination  of  hemoglobin  was  prepared  by  reconstituting  a  vial  of  Drabkin’s  reagent   (Sigma)  in  1  L  of  purified  water,  adding  0.5  mL  of  Brij  35  solution,  product  code  B  4184   (Sigma)   which   contains   sodium   bicarbonate,   potassium   ferricyanide   and   potassium   cyanide.     4.2.2  Collection  and  Preparation  of  Patient  and  Control  Samples       Patient  samples  were  collected  at  the  Neurology  and  Ophthalmology  Clinic  at  Michigan   State   University,   following   their   informed   consent.     The   patients   also   filled   out   questionnaires   pertaining   to   their   sex,   age,   disease   type,   duration,   and   medications.     Following  venipunture,  the  samples  were  refrigerated  briefly  before  being  transported   to   the   Spence   Lab.       For   control   and   patient   samples,   following   venipuncture   and   collection   into   heparinized   vacutainers   (BD,   Franklin   Lakes,   NJ),   whole   blood   was   centrifuged   for   10   minutes   at   500g   to   separate   the   RBCs   from   the   other   blood   components.     Following   the   removal   of   the   plasma   and   buffy   coat,   by   aspiration,   the   RBCs   were   washed   three   times   with   PSS   and   centrifuged   as   previously   stated.     After   the   final   washing,   the   supernatant   was   removed   and   the   hematocrit   of   the   RBCs   was   measured  using  a  CritSpin©  analyzer.       118   2+ 4.2.3  Determination  of  Radiolabelled  Zn  Interaction  with  MS  Patient  RBCs       2+ The   amount   of   Zn   interacting   with   MS   and   healthy   RBCs   was   determined   using   the   65 2+ 65 2+ radioligand   Zn  and  C-­‐peptide.    First,   Zn  and  C-­‐peptide  were  incubated  together   in   pure   water   for   three   minutes   to   ensure   peptide   activation.     PSS   was   then   added   to   the  mixture,  immediately  followed  by  the  addition  of  the  appropriate  volume  of  RBCs  to   prepare  1  mL  samples  with  a  7%  hematocrit.    Following  a  two  hour  incubation  at  37°C,   samples  were  centrifuged  at  500g  for  4  minutes  and  the  supernatant  was  collected.    In  a   96-­‐well   plate,   200   μL   of   supernatant   were   combined   with   100   μL   of   scintillation   cocktail   (Optiphase  Supermix,  Perkin  Elmer,  Waltham,  MA).    An  appropriate  range  of  standards   was  also  prepared  and  measured.    The  amount  of  radioactivity  in  each  sample  was  then   determined   using   a   1450   Microbeta   Plus   liquid   scintillation   counter   (Wallac,   Turku,   65 2+ Finland).     The   concentration   of   Zn   in   each   sample   was   quantified   by   comparing   to   a   65 2+ set  of  standards  prepared  in  RBC  supernatant.    Knowing  the  amount  of   Zn  added  to   the   original   RBC   sample,   and   how   much   was   left   in   the   supernatant,   the   amount   of   65 2+ Zn  interacting  with  the  RBCs  can  be  determined  by  subtraction.     4.2.4  Determination  of  Radiolabelled  Glucose  Uptake  of  MS  Patient  RBCs       14 Radiolabeled   C-­‐glucose  was  used  to  determine  the  amount  of  glucose  taken  into  the   14 14 RBC,.    After  the  RBCs  were  purified,  they  were  incubated  with   C-­‐glucose  alone  or   C-­‐   119   2+ 2+ glucose  and  Zn  bound  to  C-­‐peptide.    For  the  samples  incubated  with  Zn  bound  to  C-­‐ 2+ peptide,  Zn  and  C-­‐peptide  were  first  combined  in  a  1:1  ratio  and  allowed  to  interact   for   three   minutes   before   the   addition   of   36   µL   of   5   millicurie   per   50   mL   sterile   water   14 C-­‐glucose   (Moravek   Biochemicals,   Brea,   CA)   and   the   correct   volumes   of   PSS   and   RBCs   2+ to  make  a  1  mL  sample  of  7%  RBCs.    For  the  samples  without  Zn  bound  to  C-­‐peptide,   2+ equivalent   volumes   of   purified   water   were   used   in   place   of   the   Zn   and   C-­‐peptide   solutions.    Following  the  addition  of  the  RBCs,  the  samples  were  incubated  for  4  hours   at   37°C   before   being   centrifuged   at   500g   for   4   minutes.   The   supernatant   was   then   removed   and   discarded   and   the   cells   were   resuspended   in   PSS   and   centrifuged   again.     14 This  process  was  repeated  three  times  to  ensure  the  removal  of  excess   C-­‐glucose  from   the  RBCs.    Once  washed,  the  RBCs  were  then  lysed  with  1  mL  of  bleach.    After  waiting  30   minutes,  the  samples  were  centrifuged  at  500g  for  4  minutes  for  any  debris  to  settle.    A   14 14 standard  curve  of   C-­‐glucose    was  prepared  by  adding  5  µL  of   C-­‐glucose  to  1  mL  of   bleach   and   making   three   1:1   serial   dilutions.     A   96-­‐well   plate   was   then   prepared   with   200   μL   of   supernatant,   combined   with   100   μL   of   scintillation   cocktail   (optiphase   supermix,  Perkin  Elmer,  Waltham,  MA),  and  the  amount  of  radioactivity  in  each  sample   was   then   determined   using   a   1450   Microbeta   Plus   liquid   scintillation   counter   (Wallac,   Turku,   Finland).     While   the   radioactivity   of   the   14 C-­‐glucose   is   known,   the   number   of   radiolabeled   carbons   on   each   glucose   molecule   can   vary.     Therefore,   the   amount   of     120   radioactivity   cannot   be   directly   calibrated   to   the   amount   of   glucose   uptake   into   the   14 RBCs.     For   this   reason,   the   amount   of   C-­‐glucose   was   normalized   by   counts   per   minute   to  the  measurement  for  healthy  RBCs.     4.2.5  Determination  of  ATP  Release  from  MS  Patient  RBCs  in  Response  to  Flow       9 ATP   release   was   determined   in   a   manner   than   has   been   previously   reported.     After   purification,   RBCs   are   diluted   to   a   7%   hematocrit   and   pumped   through   tubing   with   dimensions  similar  to  arterioles.  In  response  to  mechanical  deformation,  ATP  is  released   from   RBCs.     To   accomplish   this,   two   syringes,   one   containing   the   sample   solution   and   the   other   containing   a   luciferin/luciferase   mixture,   were   used   in   a   dual   syringe   pump.     Solutions   were   pumped,   at   6.7   µL/min,   through   50   µm   internal   diameter   microbore   tubing.    The  solutions  met  at  a  mixing  T-­‐junction  and  the  combined  stream  was  pumped   through   an   additional   segment   of   75   µm   internal   diameter   tubing,   with   a   portion   of   the   polyimide  coating  removed,  over  a  PMT  in  a  light  excluding  box,  as  shown  in  figure  4.1.     This   allowed   for   the   chemiluminescence   resulting   from   the   reaction   of   ATP   with   luciferin/luciferase   to   be   detected.   Comparing   to   a   calibration   curve   of   ATP   standards   with  known  concentrations,  the  amount  of  ATP  released  from  the  RBCs  was  calculated.     To  ensure  the  ATP  was  not  the  result  of  lysis,  samples  of  the  RBCs  were  incubated  with   glybenclamide  for  one  hour  before  measurement.           121       Figure  4.1  –  Determination  of  ATP  Release  from  MS  Patient  RBCs  in  Response  to  Flow   –   RBCs   were   diluted   to   a   7%   hematocrit   and   pumped   through   tubing,   one   syringe   containing   the   sample   and   the   other   containing   luciferin/luciferase.   Solutions   were   pumped,   at   6.7   µL/min,   through   50   µm   internal   diameter   microbore   tubing.     The   solutions  met  at  a  mixing  T-­‐junction  and  the  combined  stream  was  pumped  through  an   additional  segment  of  75  µm  internal  diameter  tubing,  with  a  portion  of  the  polyimide   coating   removed,   over   a   PMT   in   a   light   excluding   box.     This   allowed   for   the   chemiluminescence   resulting   from   the   reaction   of   ATP   with   luciferin/luciferase   to   be   detected.           122   4.2.6  Determination  of  C-­‐peptide  Interaction  with  MS  Patient  RBCs  and  the  Plasma   Concentration  of  C-­‐Peptide  by  Enzyme  Linked  Immunosorbent  Assay       2+ The   interaction   between   Zn   bound   to   C-­‐peptide   and   RBCs   was   studied   using   solutions   2+ containing  7%  RBCs    and  20  nM  each  of  Zn  and  C-­‐peptide.    Samples  were  prepared  by   2+ mixing   the   appropriate   volumes   of   Zn   and   C-­‐peptide   in   water   for   several   minutes   before   adding   PSS,   immediately   followed   by   the   RBCs   from   either   MS   patients   or   healthy   controls.     Following   a   three   hour   incubation   at   37°C,   the   samples   were   centrifuged  at  500g  for  4  minutes  and  the  C-­‐peptide  remaining  in  the  supernatant  was   measured  using  an  enzyme-­‐link  immunosorbent  assay  (ELISA)  (Millipore,  Billerica,  MA).     The   process   of   an   ELISA   is   depicted   in   figure   4.2.     The   amount   of   C-­‐peptide   in   the   supernatant   was   calculated   using   C-­‐peptide   standards   that   were   prepared   in   the   supernatant   of   a   7%   RBC   sample   in   order   to   account   for   any   matrix   effects.     Knowing   this,  the  amount  of  C-­‐peptide  interaction  with  the  RBC  was  calculated.     2+ 4.2.7  Determination  of  Basal  Zn  Levels  in  MS  Patient  RBCs  by  Atomic  Absorption   Spectroscopy       2+ To  compare  the  basal  levels  of  Zn  on  the  RBCs  from  MS  patients  and  healthy  controls,   2+ the   amount   of   Zn ,   relative   to   the   amount   of   hemoglobin,   was   determined   using   10 atomic  absorption  spectroscopy.    After  the  initial  centrifugation  of  the  RBCs,  200  µL  of   the  cells  were  removed  and  set  aside.    These  RBCs  were  washed  three  times  with  1  mL       123     Figure   4.2   –   Determining   C-­‐peptide   Interaction   with   the   RBCs   of   MS   Patients   using   2+ Enzyme-­‐Linked   Immunosorbent   Assay   –   The   interaction   between   Zn   bound   to   C-­‐ peptide  and  RBCs  was  studying  using  solutions  containing  7%  RBCs    and  20  nM  each  of   2+ 2+ Zn  and  C-­‐peptide.    Samples  were  prepared  by  mixing  the  appropriate  volumes  of  Zn   and  C-­‐peptide  in  water  for  several  minutes  before  adding  PSS,  immediately  followed   by   the   RBC   of   either   MS   patients   or   healthy   controls.     Following   a   three   hour   incubation   at   37°C,   the   samples   were   centrifuged   at   500g   for   four   minutes   and   the   C-­‐peptide   remaining  in  the  supernatant  was  measured  using  an  ELISA.           124   of  saline,  centrifuged  at  500g  for  4  minutes  with  removal  of  the  supernatant  after  each   washing.     Following   the   washings,   100   µL   of   the   RBCs   were   lysed   with   200   µL   of   cold,   purified   water.     Because   of   the   complex   matrix,   samples   were   prepared   using   the   standard   addition   method.     In   this   method,   an   aliquot   of   lysed   RBCs   and   a   known   2+ amount  of  Zn  were  added  to  each  sample,  allowing  for  the  calculation  of  the  original   2+ amount  of  Zn  in  the  solution  using  the  equation:     𝑏𝑐!   𝑚𝑉 ! 𝑐!   =   where   b   and   m   are   the   intercept   and   slope,   respectively,   of   the   resulting   calibration   2+ curve,   cx   is   the   unknown   concentration   of   Zn   in   the   initial   solution,   cs   is   the   known   concentration  of  the  standard  solution  added  to  each  sample,  and  Vx  is  the  volume  of   identical  aliquots  of  the  lysed  RBC  solution  in  each  sample.    A  blank  with  only  an  aliquot   of  lysed  RBCs  was  also  measured.    To  account  for  differences  in  hematocrit  and  number   of   cells   in   the   lysed   RBC   solutions,   the   amount   of   hemoglobin   (Hb)   in   each   lysate   was   also   measured.     A   stock   solution   of   hemoglobin   was   prepared   by   dissolving   solid   hemoglobin  (Sigma)  in  purified  water  before  being  diluted  to  0.72  mg/mL  in  Drabkin’s   Reagent.     Several   further   dilutions   were   made   in   Drabkin’s   Reagent   to   create   a   calibration  curve.    The  lysed  RBC  solutions  were  diluted  2:1000.    200  µL  of  each  solution   were  pipetted  into  a  96-­‐well  plate,  and  after  15  minutes,  the  absorbance  was  measured   on   a   plate   reader   (SpectraMax   M4,   Molecular   Devices,   Sunnyvale,   CA)   at   540   nm.   The     125   2+ 2+ amount  of  Zn  in  the  RBC  samples  was  then  reported  as  micrograms  of  Zn  per  gram   of  hemoglobin.     4.3  Results       The  amount  of  ATP  release  from  the  RBCs  of  MS  patients  was  found  to  be  an  average  of   344.7   ±   46.8   nM   where   the   average   release   from   healthy   controls   was   132.1   ±   14.1   nM,   11 as  seen  in  figure  4.3.    Reaffirming  the  preliminary  studies,  the  ATP  release  of  the  RBCs   from   the   MS   patients   was   nearly   three   times   the   amount   of   that   from   the   healthy   controls.     When   the   RBCs   of   the   MS   patients   were   incubated   with   a   CFTR   inhibitor,   glybenclamide,  the  ATP  release  is  blocked  back  down  below  the  amount  of  the  healthy   controls,  to  a  level  of  65.3  ±  11.6  nM,  showing  that  the  increase  in  ATP  release  of  the   flowing  RBCs  of  MS  patients  is  not  the  result  of  RBC  lysis.     2+ Figure   4.4   shows   the   basal   amount   of   Zn   in   the   RBCs   of   the   MS   patients   was   found   to   2+ 2+ be   41.8   ±   1.7   µg   of   Zn /g   Hb,   which   is   a   27%   increase   of   the   basal   levels   of   Zn   of   the   2+ 65 2+ RBCs  of  healthy  controls,  32.9  ±  2.2  µg  of  Zn /g  Hb.    Additionally,  the  amount  of   Zn   that  is  able  to  interact  with  the  RBCs  of  MS  patients  is  significantly  higher,  at  a  value  of   3.61  ±  0.22  picomoles,  than  that  of  healthy  controls,  at  a  value  of  2.26  ±  0.24  picomoles,   as  seen  in  figure  4.5.    Similarly,  it  can  be  seen  in  figure  4.6  that  the  amount  of  C-­‐peptide   interacting   with   the   RBCs   of   MS   patients,   3.61   ±   0.18   picomoles,   is   very   similar   to   the           126       Figure  4.3  –  ATP  Release  from  RBCs  of  MS  Patients  –   The   results   from   the   ATP   release   studies   show   that   the   ATP   release   from   the   RBCs   of   MS   patients   was   found   to   be   an   average   of   344.7   ±   46.8   nM   where   the   average   release   from   healthy   controls   was   132.1   ±   14.1   nM.     When   the   RBCs   of   the   MS   patients   were   incubated   with   a   CFTR   inhibitor,   glybenclamide,   the   ATP   release   is   decreased   back   down   below   the   amount   of   the   healthy  controls,  to  a  level  of  65.3  ±  11.6  nM,  showing  that  the  increase  in  ATP  release   of  the  flowing  RBCs  of  MS  patients  is  not  the  result  of  RBC  lysis.    The  error  is  reported  as   standard   error   of   the   mean,   for   N   =   19   MS   patients,   10   healthy   controls   and   12   glybenclamide  inhibitions.    The  asterisk  represents  p  <  0.001.           127   50 * µg Zn/g Hb 40 30 20 10 0 Control MS     2+ 2+ Figure  4.4  –  Basal  Levels  of  Zn  in  the  RBCs  of  MS  Patients  –  The  basal  amount  of  Zn   2+ in  the  RBCs  of  the  MS  patients  was  found  to  be  41.8  ±  1.7  µg  of  Zn /g  Hb,  which  is  a   2+ 27%  increase  of  the  basal  levels  of  Zn  of  the  RBCs  of  healthy  controls,  32.9  ±  2.2  µg  of   2+ Zn /g  Hb.    The  error  is  reported  as  standard  error  of  the  mean  for  N  =  21  MS  patients   and   11   healthy   controls.   The   asterisk   represents   p   <   0.01   as   compared   to   the   control   sample.           128   pmoles 65Zn2+ on RBCs 5 4 * 3 2 1 0 Control   MS   65 2+ 65 2+ Figure  4.5  –   Zn  interaction  with  the  RBCs  of  MS  Patients  –  The  amount  of   Zn   that  is  able  to  interact  with  the  RBCs  of  MS  patients  is  significantly  higher,  at  a  value  of   3.61  ±  0.22  picomoles,  than  that  of  healthy  controls,  at  a  value  of  2.26  ±  0.24  picomoles.     The   amount   of   C-­‐peptide   interaction   with   the   RBC   correlates   to   this   very   well,   as   shown   in  figure  4.6.    The  error  is  reported  as  standard  error  of  the  mean  for  N  =  22  MS  patients   and  11  healthy  controls.      The  asterisk  represents  p  <  0.001  as  compared  to  the  control   samples.             129   pmole C-Peptide on RBCs 4 * 3 2 1 0 Control MS     Figure   4.6   –   C-­‐peptide   Interaction   with   the   RBCs   of   MS   Patients   –   Correlating   to   the   data  in  figure  4.5,  it  can  be  seen  that  the  amount  of  C-­‐peptide  interacting  with  the  RBCs   65 2+ is  very  similar  to  the  amount  of   Zn .    3.61  ±  0.18  picomoles  of  C-­‐peptide  interacted   with  the  RBCs  of  MS  patients,  while  only  2.43  ±  0.20  picomoles  interacted  with  the  RBCs   of  healthy  controls.    The  error  is  reported  as  standard  error  of  the  mean  for  N  =  12  MS   patients  and  6  healthy  controls.    The  asterisk  represents  p  <  0.001  as  compared  to  the   control  sample.     130   65 2+ amount  of   Zn  interacting  with  these  cells.    Also  shown  in  these  results,  only  2.43  ±   65 2+ 0.20  picomoles  of   Zn  interacted  with  the  RBCs  of  healthy  controls.     The  final  set  of  experiments  completed  with  the  MS  patient  samples  was  to  investigate   the   glucose   uptake   into   the   RBCs   of   both   the   patient   and   healthy   control   groups.   Figure   4.7  shows  the  results,  and  it  should  be  noted  that  the  x-­‐axis  does  not  start  at  zero,  so   while  there  are  differences,  they  are  not  as  extreme  as  they  seem.    Comparing  the  RBCs   of  the  MS  patients  and  those  of  the  healthy  control,  no  significant  difference  was  seen   2+ in   the   amount   of   glucose   uptake.     However,   when   the   RBCs   were   stimulated   by   Zn   bound  to  C-­‐peptide,  those  from  the  MS  patients  took  in  6.0  ±  0.1%  more  glucose  than   those  of  the  healthy  controls.    While  this  seems  like  a  small  increase,  it  was  a  statistically   significant  one.     4.4  Discussion       As  expected,  the  ATP  release  from  the  RBCs  of  MS  patients  was  nearly  three  times  that   of   the   RBC   from   healthy   controls.     As   shown   in   figure   4.3,   while   there   is   a   statistical   difference,  the  standard  deviation  for  the  average  from  the  MS  patients  is  quite  large.     This  may  be  due  to  differences  in  the  individual  patients,  such  as  age,  sex  disease  type   and   duration,   or   even   medications.     While   data   was   collected   on   these   areas   for   each   patient,   the   sample   set   is   not   yet   large   enough   to   separate   by   those   factors.     Additionally,  for  this  data  set,  the  MS  samples  were  also  incubated  with  glybenclamide.     This      pharmaceutical      has      been      shown      to      be      an      inhibitor      of      the      cystic      fibrosis     131   Normalized 14C-glucose uptake 1.2 * * * 1.0 Control MS   Figure   4.7   –   Glucose   Uptake   into   the   RBCs   of   MS   Patients   –   It  should  be  noted  that  the   x-­‐axis  does  not  start  at  zero,  so  while  there  are  differences,  they  are  not  as  extreme  as   they  seem.    Comparing  the  untreated  RBCs  (black)  of  the  MS  patients  and  those  of  the   healthy   control,   no   significant   difference   was   seen   in   the   amount   of   glucose   uptake.     2+ However,   when   the   RBCs   were   stimulated   by   Zn   bound   to   C-­‐peptide   (grey),   those   from   the   MS   patients   took   in   6.0   ±   0.01%   more   glucose   than   those   of   the   healthy   controls.    While  this  seems  like  a  small  increase,  it  is  a  statistically  significant  one.    The   error  is  reported  as  standard  error  of  the  mean  for  N  =  22  MS  patient  RBC  samples  and   11  healthy  control  RBC  samples.  The  asterisk  represents  p  <  0.001  and  the  pound  sign   represents  p  <  0.01.     132     12 transmembrane   conductance   regulator   (CFTR).     CFTR   has   also   been   implicated   as   13 necessary   in   RBC-­‐derived   ATP   in   response   to   mechanical   deformation.     Therefore,   the   subsequent  decrease  of  ATP  release  from  RBC  following  incubation  with  glybenclamide   shows  that  the  increase  in  ATP  in  the  samples  for  the  MS  patients  is  not  due  to  lysis  of   the  cells,  but  is,  in  fact,  due  to  an  increase  in  ATP  release  from  the  RBCs.    It  is  known   that   RBC-­‐derived   ATP   increases   the   production   of   nitric   oxide   (NO)   in   the   endothelial   14 cells  lining  the  blood  vessels.         This   is   significant   due   to   previous   reports   showing   the   increased   levels   of   NO   and   its   metabolites   in   the   lesions   that   occur   as   a   result   of   demyelination   and   in   the   cerebral   15 spinal  fluid  (CSF),  serum  and  urine  of  MS  patients.    Knowing  that  there  is  an  increase   in  RBC-­‐derived  ATP  in  MS  patients,  it  logically  follows  that  there  would  be  an  increase  in   NO  production  in  these  patients  as  well.    NO  has  been  suggested  as  one  of  the  factors   involved   in   the   change   in   permeability   of   the   blood   brain   barrier   leading   to   its   16 breakdown.    The  research  presented  here  has  potentially  found  another  link  further   upstream  in  this  process  and  may  help  the  overall  understanding  of  the  etiology  of  MS.     While  the  increase  in  glucose  uptake  in  the  RBCs  of  MS  patients  as  compared  to  healthy   controls,   as   shown   in   figure   4.7,   is   not   significant,   the   RBCs   of   both   groups   had   a   2+ significant  increase  in  glucose  uptake  when  incubated  with  Zn  bound  to  C-­‐peptide  and   the  increase  glucose  uptake  of  the  RBCs  of  MS  patents  was  significantly  higher  than  that     133   of  the  RBCs  of  the  healthy  controls.    These  finding  were  expected,  as  glucose  is  used  in   glycolysis  to  produce  ATP  in  the  RBC.    The  increase  in  ATP  release  from  the  RBC  would   likely   require   an   increase   in   ATP   production,   resulting   in   the   need   for   an   increase   in   glucose  uptake.    While  glucose  metabolism  has  only  been  investigated  in  MS  in  a  very   limited   capacity,   it   had   previously   been   shown   that   MS   patients   had   a   significant   increase   in   the   ATP   content   of   their   RBCs   after   an   intake   of   50   g   of   glucose   that   was   not   17 seen  in  healthy  controls.    It  has  also  been  noted  that  there  is  an  increase  in  glucose   18 metabolism  outside  of  the  mitochondria  in  the  CSF  of  MS  patients.    This  suggests  that   2+ there   may   be   some   dysfunction   of   glucose   metabolism   in   MS   patients.     Zn   has   19 previously  been  implicated  in  increased  glucose  uptake  in  fibroblasts  and  adipocytes.     Until  now,  this  had  never  been  investigated  in  RBCs.    When  taken  together,  the  results   2+ 2+ from  chapter  3  showing  the  ability  of  Zn  bound  to  C-­‐peptide  to  deliver  Zn  to  the  RBC   and   data   presented   here   showing   an   increase   in   glucose   uptake   in   the   RBCs   of   both   MS   2+ patients   and   healthy   controls   when   incubated   with   Zn   bound   to   C-­‐peptide   lead   to   the   2+ conclusion   that   Zn   also   causes   an   increase   in   glucose   uptake   in   human   RBCs.     It   is   interesting  to  note  that  while  the  idea  of  altered  glucose  metabolism  in  MS  patients  has   17,20 only   been   limitedly   explored,   there   is   genetic   overlap   between   type   1   diabetes   and   MS,  with  a  significant  increase  in  prevalence  of  MS  in  the  first-­‐degree  relatives  of  type  1   21 diabetics.     Genetic   investigation   is   still   ongoing,   however,   there   has   been   some     134   evidence  of  overlap  between  single  neucleotide  polymorphisms  associated  with  type  1   22 diabetes  and  those  associated  with  MS.     2+ As  seen  in  figure  4.4,  the  basal  level  of  Zn  in  the  RBCs  of  MS  patients  is  higher  than   that   of   healthy   controls.     Because   of   the   nature   of   the   samples,   this   level   was   measured   in   relation   to   the   amount   of   hemoglobin   to   ensure   that   the   same   number   of   RBCs   were   2+ being   analyzed.     There   has   been   some   previous   discussion   about   the   role   of   Zn   in   MS,   though   it   has   not   been   the   focus   of   major   MS   research.     It   has   been   noted   that   in   areas   2+ 2+ of   Zn   contamination,   either   in   the   soil   and   water,   or   from   Zn   smelters,   clusters   of   MS  have  been  found. 23-­‐25    Other  reports  have  shown  that  the  RBCs  of  MS  patients  have   2+ increased  levels  of  Zn ,  even  when  they  are  not  associated  with  an  MS  cluster.    While   2+ 2+ the   amount   of   Zn   in   or   on   the   RBC   was   increased,   the   serum   Zn   levels   did   not   differ   2+ between   these   groups,   suggesting   the   increase   in   RBC   Zn   levels   is   not   a   result   of   2+ 26 defective  Zn  absorption  or  individual  nutrition.    The  previous  research  in  this  area   2+ has   hypothesized   that   the   amount   of   Zn   in   the   RBC   was   correlated   with   a   decrease   in   the  amount  of  cholesterol  in  the  cells  as  well.    Here  it  is  concluded  that  regardless  of  the   2+ amount   of   cholesterol,   the   amount   of   basal   Zn   in   the   RBCs   of   MS   patients   is   higher   than  that  of  healthy  controls.       135   2+ While   there   has   been   previous   research   into   the   amount   of   Zn   bound   to   the   RBC,   there   has   not   been   research   performed   investigating   the   ability   of   the   RBCs   of   MS   2+ 65 2+ patients   to   bind   Zn .     Looking   at   the   amounts   of   Zn   and   C-­‐peptide   interacting   with   the   RBC   in   figures   4.5   and   4.6,   respectively,   it   is   clear   that   for   both   molecules   the   amount   was   higher   with   the   RBCs   of   MS   patients   as   compared   to   healthy   controls.     Comparing   the   data   presented   in   figures   4.5   and   4.6,   it   can   be   observed   that   the   65 2+ average  amounts  of   Zn  and  C-­‐peptide  interacting  with  the  RBC  are  nearly  identical.     2+ This   confirms   the   data   described   in   chapter   2   showing   the   interaction   of   Zn   and   C-­‐ peptide  with  the  RBC  occurs  at  a  1:1  ratio.    While  C-­‐peptide  has  never  been  implicated   in   the   etiology   of   MS,   one   of   the   following   possibilities   may   serve   to   explain   the   increased  interaction.    The  β-­‐cells  of  MS  patients  may  contain  an  increased  amount  of   2+ 2+ Zn ,   allowing   for   more   of   the   C-­‐peptide   become   Zn   bound   to   C-­‐peptide.     Another   possibility  may  be  an  increase  in  number  of  binding  sites  on  each  RBC  in  MS  patients.     While   the   identity   of   this   binding   site   is   not   currently   known,   work   is   currently   being   done  in  the  Spence  group  investigating  GLUT1  as  a  potential  target,  hypothesizing  that   2+ the  larger  increase  in  glucose  uptake  by  the  RBCs  may  be  the  result  of  Zn  binding  to   the  protein  in  the  membrane.    This  will  be  discussed  further  in  the  future  work  section   of  chapter  5.       The   collection   of   information   from   MS   patients   regarding   age,   sex,   disease   type   and   duration,   and   the   usage   of   steroids   and   other   medications   was   done   in   a   hope   to     136   discover   a   link   between   these   areas   and   the   biological   effects   of   the   disease   on   the   RBC   as   studied.     The   vast   majority   of   the   pool   of   patients   to   date   has   been   relapse-­‐remitting   MS  patients,  and  the  pool  is  not  yet  large  enough  to  discern  differences  in  many  of  the   areas  listed  above.    However,  as  seen  in  figure  4.8,  there  is  a  significant  difference  in  the   amount   of   65 2+ Zn   interaction   with   the   RBC   between   males   and   females   patients   with   65 2+ MS  that  is  not  seen  in  healthy  controls.    The  highest  amount  of   Zn  interacting  with   the   RBC   s   is   seen   in   male   MS   patients.     While   females   are   twice   as   likely   as   males   to   27 develop  the  disease,  men  with  MS  tend  to  progress  to  disability  at  a  more  rapid  rate.     65 2+ With  a  larger  sample  size  it  will  be  interesting  to  see  if  the  higher  levels  of   Zn  will   have  down  stream  effects  of  increase  glucose  uptake  and  ATP  release  from  the  RBCs  of   male   MS   patients.     This   could   potentially   lead   to   discoveries   into   the   etiology   of   MS   and   the  sex  differences  seen  in  the  disease.     While   there   needs   to   be   more   conclusive   evidence   before   any   one   of   these   elements   of   MS   etiology   can   be   used   as   a   clinical   biomarker,   the   research   presented   here   shows   several  potential  targets.    As  further  data  is  collected  on  the  various  types  of  MS,  along   with   CIS   patients,   and   immune   disease   and   neurological   controls,   hopefully,   one   of   these   targets   will   prove   to   be   a   usable   biomarker   for   the   detection   and   monitoring   of   MS.         137   * 4 3 pmoles 65 Zn 2+ on RBC 5 2 1 0 Control   65 MS   2+ Figure  4.8  –  The  Difference  in   Zn  interaction  with  the  RBCs  of  Male  and  Female  MS   65 2+ Patients  –  There  is  a  significant  difference  in  the  amount  of   Zn  interaction  with  the   RBC   between   male   (black   bars)   and   female   (white   bars)   patients   with   MS   that   is   not   65 2+ seen   in   healthy   controls.     The   highest   amount   of   Zn   interacting   with   the   RBC   s   is   seen   in   male   MS   patients   at   a   level   of   4.16   ±   0.18   picomoles.     The   error   is   shown   as   standard   error   of   the   mean   for   N   =   6   male   MS   patients,   3   male   healthy   controls,   16   female  MS  patients,  and  8  female  healthy  controls.    The  asterisk  represents  p  <  0.001.     138   REFERENCES     139       1     2     3     4     5     6     7     8     9     10       REFERENCES   Awad,   A.,   Hemmer,   B.,   Hartung,   H.   P.,   Kieseier,   B.,   Bennett,   J.   L.   &   Stuve,   O.   Analyses   of   cerebrospinal   fluid   in   the   diagnosis   and   monitoring   of   multiple   sclerosis.  Journal  of  Neuroimmunology  219,  1-­‐7  (2010).   Avsar,  T.,  Korkmaz,  D.,  Tütüncü,  M.,  Demirci,  N.  O.,  Saip,  S.,  Kamasak,  M.,  Siva,  A.   &   Turanli,   E.   T.   Protein   biomarkers   for   multiple   sclerosis:   semi-­‐quantitative   analysis  of  cerebrospinal  fluid  candidate  protein  biomarkers  in  different  forms  of   multiple  sclerosis.  Multiple  Sclerosis  Journal  18,  1081-­‐1091  (2012).   Biomarkers   Definitions   Working   Group.   Biomarkers   and   surrogate   endpoints:   Preferred   definitions   and   conceptual   framework.   Clinical   Pharmacology   &   Therapeutics  69,  89-­‐95  (2001).   Prentice,   R.   Surrogate   endpoints   in   clinical   trials   -­‐   definition   and   operational   criteria.  Statistics  in  Medicine  8,  431-­‐440  (1989).   Comabella,  M.  &  Khoury,  S.  J.  in  Multiple  Sclerosis:  Diagnosis  and  Therapy      (ed   Howard  L.  Weiner)  Ch.  2,  26-­‐55  (2012).   Villar,  L.  M.,  Masjuan,  J.,  González-­‐Porqué,  P.,  Plaza,  J.,  Sádaba,  M.  C.,  Roldán,  E.,   Bootello,  A.  &  Alvarez-­‐Cermeño,  J.  C.  Intrathecal  IgM  synthesis  predicts  the  onset   of   new   relapses   and   a   worse   disease   course   in   MS.   Neurology   59,   555-­‐559   (2002).   Cénit,  M.  D.,  Blanco-­‐Kelly,  F.,  de  las  Heras,  V.,  Bartolomé,  M.,  de  la  Concha,  E.  G.,   Urcelay,  E.,  Arroyo,  R.  &  Martínez,  A.  Glypican  5  is  an  interferon-­‐beta  response   gene:  a  replication  study.  Multiple  Sclerosis  15,  913-­‐917  (2009).   Bielekova,  B.  &  Martin,  R.  Development  of  biomarkers  in  multiple  sclerosis.  Brain   127,  1463-­‐1478  (2004).   Subasinghe,   W.   &   Spence,   D.   M.   Simultaneous   determination   of   cell   aging   and   ATP  release  from  erythrocytes  and  its  implications  in  type  2  diabetes.  Analytica   Chimica  Acta  618,  227-­‐233  (2008).   2+ Whitehouse,  R.,  Prasad,  A.,  Rabbani,  P.  &  Cossack,  Z.  Zn  in  plasma,  neutrophils,   lymphocytes,   and   erythrocytes   as   determined   by   flameless   atomic   absorption   spectrophotometry.  Clinical  Chemistry  28,  475-­‐480  (1982).   140   11     12     13     14     15     16     17     18     19     20       Spence,   D.   M.,   Letourneau,   S.,   Hernandez,   L.   &   Faris,   A.   N.   Evaluating   the   effects   of   estradiol   on   endothelial   nitric   oxide   stimulated   by   erythrocyte-­‐derived   ATP   using  a  microfluidic  approach.  Analytical  and  Bioanalytical  Chemistry  397,  3369-­‐ 3375  (2010).   Schultz,  B.  D.,  DeRoos,  A.  D.,  Venglarik,  C.  J.,  Singh,  A.  K.,  Frizzell,  R.  A.  &  Bridges,   R.   J.   Glibenclamide   blockade   of   CFTR   chloride   channels.   American   Journal   of   Physiology-­‐Lung  Cellular  and  Molecular  Physiology  271,  L192-­‐L200  (1996).   Sprague,   R.   S.,   Ellsworth,   M.   L.,   Stephenson,   A.   H.,   Kleinhenz,   M.   E.   &   Lonigro,   A.   J.  Deformation-­‐induced  ATP  release  from  red  blood  cells  requires  CFTR  activity.   American  Journal  of  Physiology  275,  H1726-­‐1732  (1998).   Sprague,  R.  S.,  Ellsworth,  M.  L.,  Stephenson,  A.  H.  &  Lonigro,  A.  J.  ATP:  The  red   blood   cell   link   to   NO   and   local   control   of   the   pulmonary   circulation.   American   Journal   of   Physiology-­‐Heart   and   Circulatory   Physiology   271,   H2717-­‐H2722   (1996).   Smith,  K.  J.  &  Lassmann,  H.  The  role  of  nitric  oxide  in  multiple  sclerosis.  Lancet   Neurology  1,  232-­‐241  (2002).   Boje,   K.   M.   K.   &   Lakhman,   S.   S.   Nitric   oxide   redox   species   exert   differential   permeability   effects   on   the   blood-­‐brain   barrier.   Journal   of   Pharmacology   and   Experimental  Therapeutics  293,  545-­‐550  (2000).   Raczkiewicz,   J.   &   Leyko,   W.   Adenine   Nucleotide   Content   of   Erythrocytes   of   Multiple  Sclerosis  Patients.  Clinica  Chimica  Acta  14,  408-­‐409  (1966).   Regenold,   W.   T.,   Phatak,   P.,   Makley,   M.   J.,   Stone,   R.   D.   &   Kling,   M.   A.   Cerebrospinal   fluid   evidence   of   increased   extra-­‐mitochondrial   glucose   metabolism   implicates   mitochondrial   dysfunction   in   multiple   sclerosis   disease   progression.  Journal  of  the  Neurological  Sciences  275,  106-­‐112  (2008).   2+ Shay,   N.   F.   &   Tang,   X.   H.   Zn   has   an   insulin-­‐like   effect   on   glucose   transport   mediated   by   phosphoinositol-­‐3-­‐kinase   and   AKT   in   3T3-­‐L1   fibroblasts   and   adipocytes.  Journal  of  Nutrition  131,  1414-­‐1420  (2001).   Regenold,   W.   T.,   Phatak,   P.,   Makley,   M.   J.,   Stone,   R.   D.   &   Kling,   M.   A.   Cerebrospinal   fluid   evidence   of   increased   extra-­‐mitochondrial   glucose   metabolism   implicates   mitochondrial   dysfunction   in   multiple   sclerosis   disease   progression.  Journal  of  the  Neurological  Sciences  275,  (2008).   141   21     22     23     24     25     26   Dorman,   J.   S.,   Steenkiste,   A.   R.,   Burke,   J.   P.   &   Songini,   M.   Type   1   diabetes   and   multiple  sclerosis.  Diabetes  Care  26,  3192-­‐3193  (2003).   Ban,  M.,  Co,  I.  M.  S.  G.  &  Co,  I.  M.  S.  G.  The  expanding  genetic  overlap  between   multiple  sclerosis  and  type  1  diabetes.  Multiple  Sclerosis  14,  S296-­‐S296  (2008).   Henry,  J.  P.,  Williamson,  D.  M.,  Schiffer,  R.,  Wagner,  L.,  Shire,  J.  &  Garabedian,  M.   Investigation  of  a  cluster  of  multiple  sclerosis  in  two  elementary  school  cohorts.   Journal  of  Environmental  Health  69,  34-­‐38  (2007).   Schiffer,   R.   B.,   McDermott,   M.   P.   &   Copley,   C.   A   multiple   sclerosis   cluster   associated  with  a  small,  north-­‐central  Illinois  community.  Archives  Environmental   Health  56,  389-­‐395  (2001).   Stein,   E.   C.,   Schiffer,   R.   B.,   Hall,   W.   J.   &   Young,   N.   Multiple-­‐Sclerosis   and   the   Workplace   -­‐   Report   of   an   Industry-­‐Based   Cluster.   Neurology   37,   1672-­‐1677   (1987).   Dore-­‐Duffy,   P.,   Catalanotto,   F.,   Donaldson,   J.   O.,   Ostrom,   K.   M.   &   Testa,   M.   A.   2+ Zn  in  Multiple-­‐Sclerosis.  Annals  of  Neurology  14,  450-­‐454  (1983).     27     Vukusic,   S.   &   Confavreux,   C.   Pregnancy   and   multiple   sclerosis:   the   children   of   PRIMS.  Clin  Neurol  Neurosurg  108,  266-­‐270  (2006).   142   Chapter  5  –  Overall  Conclusions  and  Future  Directions     5.1  Overall  Conclusions       The  results  presented  in  this  thesis  further  the  understanding  of  the  etiology  of  MS.    In   addition,  the  reason  behind  ameliorations  of  the  disease  seen  with  the  use  of  estrogen   begins   to   be   explained.     It   was   previously   known   that   adenosine   triphosphate   (ATP)   is   released   from   RBCs   in   response   to   several   stimuli   and   this   ATP   stimulates   nitric   oxide   (NO)   production   in   the   endothelial   cells   lining   the   blood   vessels. 1,2     The   early   studies   presented  in  this  thesis  show  that  the  estrogens,  in  the  form  of  estriol  and  estradiol  can   3 decrease  the  amount  of  ATP  released  from  the  RBC,  even  in  a  non-­‐flow  system.    This  is   of   note   taking   into   consideration   the   knowledge   that   RBCs   from   patients   with   MS   release   three   times   the   amount   of   ATP   then   RBCs   from   healthy   controls,   and   that   it   has   been   shown   that   high   levels   of   estrogens,   due   to   pregnancy   or   drug   therapy,   can   ameliorate   MS. 4,5     It   is   believed   that   the   increase   in   ATP   release   may   be   over-­‐ stimulating  the  NO  production  in  the  endothelial  cells,  and  NO  is  known  to  be  toxic  to   6 the   blood   brain   barrier   (BBB).     If   the   overproduction   of   ATP   from   the   RBCs   can   be   returned  to  normal  levels,  the  leakage  in  the  BBB  may  be  able  to  be  stopped.         As   discussed   in   chapter   1   and   shown   in   figure   1.1,   cystic   fibrosis   transmembrane   conductance  regulator  (CFTR)  activity  is  necessary  for  the  release  of  ATP  from  the  RBC  in   7 response   to   mechanical   deformation.     There   have   been   several   reports   showing   the     143   effect   of   estrogen   on   the   CFTR.     Estrogens   have   been   show   to   inhibit   the   chloride   channel   activity   and   have   been   called   “glybenclamide-­‐like   inhibitors”   in   other   cell   types. 8,9     Some   very   preliminary   results   that   will   be   discussed   shortly   and   shown   in   figure   5.1,   show   this   to   be   the   case   with   ATP   release   from   RBCs.     The   data   suggests   that   estrogens  attenuate  the  excessive  release  of  ATP  from  the  RBCs  of  MS  patients.    If  this  is   the  case,  the  amount  of  NO  production  would  also  be  decreased,  potentially  leading  to   less   damage   from   the   disease.     The   use   of   the   microfluidic   device   to   investigate   the   interactions   occurring   in   the   microvasculature   showed   that   RBCs   flowing   through   the   channel   of   the   microfluidic   device   release   ATP   as   a   result   of   sheer   stress   that   stimulates   the  production  of  NO  in  bovine  pulmonary  artery  endothelial  cells  (bPAECs)  that  are  in   proximity  to  the  ATP.         Using   this   device,   the   role   of   estrogen   in   this   process   was   determined,   as   shown   in   figure  2.16.    Specifically,  when  acting  on  the  RBCs,  estrogen  has  a  negative  effect  on  the   ATP  release,  subsequently  decreasing  the  NO  production  in  the  bPAECs.    However,  the   excess   estrogen   in   the   system   interacted   with   the   bPAECs,   increasing   the   amount   of   NO   production,   though   not   above   the   levels   seen   in   RBCs   alone   until   a   superphysiological   amount  of  estrogen  was  incubated  with  the  RBCs.      The  data  presented  here  looks  at  the   role   of   estrogen   in   this   process   in   a   distinctly   different   way   from   past   studies   in   this   10 area,  where  the  hormone  was  applied  directly  to  the  cell  to  measure  NO  production.     This  reaffirms  that  it  is  essential  to  study  drug  interactions  not  only  on  a  single  cell  type,   but        to      also      investigate      the      influence      of      cell-­‐to-­‐cell      communication.        Here,      the       144   250 ATP Release (nM) 200 150 100 50 E3 50 + M S M S + 30 0 nM nM E3 M S C on tro l 0   Figure  5.1  –  The  Effect  of  Estriol  (E3)  on  ATP  Release  from  the  RBCs  of  MS  Patients  –     Flow  based  studies  were  performed  on  one  sample  of  RBCs  from  an  MS  patient  with  the   addition   of   30   and   500   nM   estriol.     Without   estriol,   the   RBCs   from   the   MS   patient   released  208.1  nM  ATP,  while  the  healthy  control  released  103.9  nM.    When  incubated   for  a  half  hour  with  30  or  500  nM  estriol,  the  ATP  release  dropped  to  59.1  and  21.6  nM,   respectively.  N  =  1.     145   microfluidic  device  was  able  to  discern  between  endothelium-­‐derived  NO  due  directly  to   estradiol   and   the   inhibition   of   NO   production   as   a   result   of   the   decrease   in   RBC-­‐derived   ATP  caused  by  the  hormone.     2+ Another  major  finding  of  this  research  is  the  evidence  that  C-­‐peptide  is  the  Zn  carrier   2+ to  the  RBC.    It  was  previously  known  that  Zn  bound  to  C-­‐peptide  will  increase  the  ATP   11 release  from  RBCs,  and  studies  completed  by  Watshsala  Medawala  have  shown  that   2+ 12 C-­‐peptide   binds   to   Zn   in   a   1:1   ratio.     It   was   the   expectation   that   C-­‐peptide   would   2+ 2+ need  Zn  to  interact  with  the  RBC,  however,  it  was  found  that  it  is  Zn  that  needs  C-­‐ peptide  for  RBC  interaction.    Through  the  course  of  experimentation,  it  was  discovered   2+ 2+ that   while   the   C-­‐peptide   can   interact   with   the   RBC   without   Zn ,   the   Zn   cannot   interact  with  the  RBC  without  C-­‐peptide,  as  seen  in  figure  3.7.      Because  C-­‐peptide  only   causes  an  increase  in  ATP  release  from  the  RBC  when  it  has  been  previously  incubated   2+ 2+ with  Zn ,  it  was  concluded  that  C-­‐peptide  is  a  Zn  carrier  to  the  RBC.         This   knowledge   was   used   in   experimentations   on   the   RBCs   of   MS   patients,   and   may   2+ have  significance  in  diabetes.    Using  Zn  as  the  metal  to  bind  C-­‐peptide  comes  from  the   in  vivo  process  in  which  C-­‐peptide  is  cleaved  and  released  into  the  blood  stream.    This   2+ occurs  in  the  β-­‐cells  of  the  pancreas,  where  Zn  is  readily  available  to  the  C-­‐peptide  for   binding  before  release  into  the  body  where  there  are  many  competitors  for  the  metal.         146   2+ As  previously  discussed,  Zn  bound  to  C-­‐peptide  has  been  shown  to  have  a  biological   effect;  it  has  also  been  shown  to  increase  ATP  release  from  RBCs  obtained  from  a  type  1   diabetic  rat  model  back  to  the  level  of  healthy  controls.    This  effect  is  also  seen  in  the   RBCs   from   the   type   2   diabetic   rats,   though   only   after   incubation   with   metformin,   a   13 common  pharmaceutical  to  treat  the  disease.           2+ It  as  been  repeated  shown  that  Zn  bound  to  C-­‐peptide  increases  RBC-­‐derived  ATP  and   subsequently   NO   release   from   endothelial   cells   lining   the   blood   vessels   causing   vasodilation. 2,14,15    With  further  understanding  of  the  structure,  potential  receptor  and   2+ mechanism   of   the   bioactivity   of   Zn   bound   to   C-­‐peptide,   there   is   pharmaceutical   potential  for  this  molecule.    Improving  the  blood  flow  of  diabetic  patients  would  help  to   ameliorate   the   complications   seen   with   the   disease,   such   as   retinopathy,   neuropathy,   and  nephropathy,  and  could  vastly  improve  the  health  and  quality  of  living  of  diabetic   patients.     Working  with  the  Neurology  and  Ophthalmology  Clinic  at  Michigan  Sate  University  and   evaluating   the   RBCs   of   MS   patients   yielded   a   wealth   of   new   information   about   the   potential   role   of   these   cells   in   the   disease.     It   was   reaffirmed   that   the   amount   of   ATP   release  from  the  RBCs  of  MS  patients  is  significantly  higher  than  that  from  the  RBCs  of   healthy  controls.    The  studies  here  went  further  than  previous  studies  and,  through  the   use  of  glybenclamide,  it  is  now  clear  that  the  increase  in  ATP  seen  in  the  samples  is  the     147   result   of   an   increased   ATP   release   from   the   RBC,   rather   than   from   cell   lysis.     These   findings   are   especially   significant   when   discussed   in   the   context   of   previous   research   showing   the   increased   levels   of   NO   and   its   metabolites   in   the   lesions   that   occur   as   a   result   of   demyelination   and   in   the   cerebral   spinal   fluid   (CSF),   serum   and   urine   of   MS   16 patients.    With  the  proven  increase  in  ATP  release  from  the  RBCs  of  MS  patients,  an   increase   in   NO   production   in   these   patients   is   expected.     An   increase   in   ATP-­‐derived   NO   would   be   a   significant   discovery   for   the   reasons   mentioned   above,   as   well   as   the   knowledge  that  NO  is  one  of  the  factors  involved  in  the  change  in  permeability  of  the   17 blood   brain   barrier   leading   to   its   breakdown.     Future   work   in   this   area   will   be   discussed  in  section  5.2.2.         As  discussed  in  chapter  1,  high  levels  of  ATP  increases  the  pore  size  of  the  P2X7  receptor   to   transport   molecules   with   masses   in   the   range   of   100s   of   Daltons   rather   than   just   18 small   cations.     This   results   in   cytotoxic   effects   on   oligodendrocytes   forming   lesions,   in   19 mice,   that   are   similar   to   those   seen   in   MS.     Interestingly,   P2X7   deficient   mice   were   four   times   less   likely   to   suffer   from   EAE   than   control   mice,   despite   having   the   same   20 increase   in   cytokine   production.     Because   of   this,   it   seems   that   the   increase   in   ATP   release   from   the   RBCs   of   MS   patients   may   be   having   a   direct   cytotoxic   effect   on   the   oligodendrocytes,  causing  the  lesions.         148   Early  work  evaluating  the  level  of  ATP  release  from  healthy  RBCs  after  incubation  with   estrogen   showed   a   decrease   in   the   concentration   of   ATP   in   these   samples.     Based   on   this,   and   reports   in   literature   on   the   ability   of   this   hormone   to   ameliorate   the   lesions   of   5,21,22 MS   patients   and   in   the   animal   model,   flow   based   studies   were   performed,   as   described   in   section   4.2.5,   on   one   sample   with   the   addition   of   30   and   500   nM   estriol.     Although   further   work   needs   to   be   performed   in   this   area,   the   early   results   are   promising.    As  seen  in  figure  5.1,  without  estriol,  the  RBCs  from  the  MS  patient  released   208.1   nM   ATP,   while   the   healthy   control   released   103.9   nM.     When   incubated   for   a   half   hour   with   30   or   500   nM   estriol,   the   ATP   release   dropped   to   59.1   and   21.6   nM,   respectively.   This   encouraging   data   shows   that   estriol   does   have   the   ability   to   inhibit   RBC-­‐derived  ATP  and  may  lend  to  the  explanation  of  the  effects  of  this  hormone.    It  is   believed   by   some   that   estrogens   help   to   ameliorate   MS   by   correcting   an   immune   23 shift.    However,  the  data  presented  in  chapter  2,  combined  with  what  was  seen  in  this   preliminary   study   and   the   previously   mentioned   ability   of   estrogen   to   interact   directly   with  and  inhibit  CFTR,  show  that  the  effect  of  estrogen  on  MS  may  not  be  related  to  the   immune  system,  but  rather  to  its  ability  to  decrease  ATP  release  from  the  RBC.     Because  ATP  in  the  RBC  is  produced  from  glycolysis,  the  increase  in  ATP  release  from  the   RBC   of   MS   patients   suggests   that   there   is   an   increase   need   for   glucose   in   the   cell   and   therefore   it   was   expected   that   the   glucose   uptake   into   these   cells   would   also   be   increased.     While   this   was   not   the   case   when   comparing   the   glucose   uptake   in   the   RBCs   of   MS   patients   to   healthy   controls,   the   RBCs   of   both   groups   had   a   significant   increase   in     149   2+ glucose   uptake   when   incubated   with   Zn   bound   to   C-­‐peptide,   and   the   increased   glucose   uptake   of   the   RBCs   of   MS   patents   was   significantly   higher   than   that   of   the   RBCs   of   the   healthy   controls.     Glucose   metabolism   has   only   been   investigated   in   MS   in   a   very   limited   capacity;   however,   as   previously   discussed,   it   has   been   noted   that   there   may   be   some  dysfunction  of  glucose  metabolism  in  MS  patients. 24,25    While  the  data  presented   here  is  not  the  first  to  suggest  that  there  is  more  than  an  autoimmune  aspect  to  MS,  nor   the  first  to  suggest  altered  glucose  metabolism,  it  is  the  first  to  suggest  a  mechanism  by   which  this  abnormality  can  influence  MS  etiology.    The  increase  in  glucose  uptake  into   the  RBCs  of  MS  patients  likely  results  in  an  increase  in  ATP  production  and  release.    This   ATP   subsequently   activated   NO   production   in   the   endothelium,   and   this   free   radical   has   a   damaging   effect   on   both   BBB,   causing   leaking   and   eventually   leading   to   the   formation   of  lesions.     While  it  is  hypothesized  that  the  increase  in  RBC-­‐derived  ATP  results  from  an  increase  in   glucose  uptake,  the  reason  for  this  increase  remains  unknown.    One  possibility  for  this   may   be   related   to   some   sort   of   mitochondrial   dysfunction,   though   little   research   has   been   completed   in   this   area.     The   belief   of   this   thesis   is   that   the   increase   in   glucose   2+ uptake   is   related   to   the   increased   levels   of   Zn   on   the   RBC.     As   previously   discussed,     2+ the   role   of   Zn   in   MS   has   not   been   the   focus   of   major   MS   research,   although   it   has   2+ been  noted  that  clusters  of  MS  have  been  found  in  areas  of  Zn  contamination,  either   2+ in  the  soil  and  water,  or  from  Zn  smelters.   26-­‐28 150      RBCs  of  MS  patients  have  also  been   2+ shown   to   have   an   increased   level   of   Zn ,   even   when   not   associated   with   an   MS   29 cluster.    The  previous  research  in  this  area  was  in  support  of  the  hypothesis  that  the   2+ amount   of   Zn   in   the   RBC   was   correlated   with   a   decrease   in   the   amount   of   cholesterol   in  the  cells  as  well.    The  results  presented  here  show  that  regardless  of  the  amount  of   2+ cholesterol,  the  amount  of  basal  Zn  in  the  RBCs  of  MS  patients  is  higher  than  that  of   healthy  controls.         2+ Glucose   uptake   stimulated   by   Zn   has   been   seen   previously   in   fibroblasts   and   30 adipocytes;  however,  this  is  the  first  time  it  was  investigated  in  RBCs.    The  ability  of   2+ 2+ Zn  bound  to  C-­‐peptide  to  deliver  Zn  to  the  RBC  and  the  increase  in  glucose  uptake  in   2+ the  RBCs  of  both  MS  patients  and  healthy  controls  when  incubated  with  Zn  bound  to   2+ C-­‐peptide  lead  to  the  conclusion  that  Zn  also  causes  an  increase  in  glucose  uptake  in   human  RBCs.    It  is  interesting  to  note  that  there  has  been  research  conducted  looking   into  the  genetic  overlap  between  type  1  diabetes  and  MS,  showing  a  significant  increase   31 in  prevalence  of  MS  in  the  first-­‐degree  relatives  of  type  1  diabetics.     65 2+ Using   Zn  bound  to  C-­‐peptide,  it  was  shown  that  for  both  molecules  that  the  amount   of  interaction  was  higher  with  the  RBCs  of  MS  patients  as  compared  to  healthy  controls.     Comparing  the  data  presented  in  figures  4.5  and  4.6,  it  can  be  noted  that  the  average     151   amounts   of   65 2+ Zn   and   C-­‐peptide   interacting   with   the   RBC   are   nearly   identical.     In   2+ addition   to   showing   the   increased   Zn   interaction   with   the   RBCs   of   MS   patients,   this   2+ data  also  reaffirms  that  the  interaction  of  Zn  and  C-­‐peptide  with  the  RBC  occurs  at  a   1:1  ratio.    C-­‐peptide  has  never  been  investigated  in  the  etiology  of  MS,  and  there  could   be  a  number  of  different  reasons  for  this  increased  binding.    The  β-­‐cells  of  MS  patients   2+ 2+ may  contain  an  increased  amount  of  Zn ,  either  through  an  increased  number  of  Zn   transporters   on   the   β-­‐cell   or   because   of   an   increased   transporter   efficiency,   either   of   2+ which   would   allow   for   more   of   the   C-­‐peptide   to   become   Zn   bound   to   C-­‐peptide.     Another  possibility  may  be  an  increase  in  the  number  of  binding  sites  on  each  RBC  in  MS   2+ patients  for  interaction  with  Zn  bound  to  C-­‐peptide.    While  the  identity  of  this  binding   site   is   not   currently   known,   work   is   currently   being   done   in   the   Spence   group   investigating   GLUT1   as   a   potential   target,   hypothesizing   that   the   larger   increase   in   2+ glucose   uptake   by   the   RBCs   may   be   the   result   of   Zn   binding   to   the   protein   in   the   membrane.         This   research   presented   here   is   a   leap   forward   in   the   understanding   of   MS   etiology.     More   conclusive   evidence,   along   with   validation   by   several   controls   is   still   needed   before   any   one   of   these   elements   can   be   used   as   a   clinical   biomarker.     However,   potential   targets   have   been   identified.     As   further   data   is   collected   on   the   various   types   of   MS,   along   with   CIS   patients,   and   immune   disease   and   neurological   controls,     152   hopefully,  one  of  these  targets  will  prove  to  be  a  usable  biomarker  for  the  detection  and   monitoring   of   MS.     Regardless,   the   data   presented   here   shows   that   along   with   the   autoimmune  component  in  MS,  there  is  clearly  also  some  dysfunction  with  the  glucose   metabolism  in  these  patients.    Research  into  this  area  needs  to  continue  until  both  the   cause  of  the  altered  ATP  release  and  the  effect  it  has  on  the  etiology  of  the  disease  are   determined.         5.2  Future  Directions     2+ 5.2.1  Zn  and  C-­‐Peptide  Binding     2+ While  work  has  previously  been  completed  looking  at  the  binding  of  Zn  to  C-­‐peptide,   there  is  still  more  to  be  learned  in  this  area.    Currently,  isothermal  calorimetry  is  being   used  to  look  at  the  interaction  between  these  two  molecules,  as  well  as  the  interaction   between  each  of  them  and  albumin,  a  metal  binding  protein  commonly  found  in  blood   plasma.         2+ Work  is  on-­‐going  in  the  Spence  Lab  to  further  investigate  the  interactions  of  Zn  and  C-­‐ peptide.    Since  it  was  first  shown  that  C-­‐peptide  needed  metal  activation  for  biological   2+ activity,  it  has  been  assumed  that  C-­‐peptide  is  somehow  being  activated  by  Zn  in  the   β-­‐cell,  as  both  molecules  are  present  in  high  concentration.    It  is  also  the  hypothesis  that   2+ Zn  is  binding  to  C-­‐peptide  in  the  β-­‐cell,  since  once  in  the  blood  stream  there  would  be     153   2+ many  strong  competitors  for  Zn  binding.    There  is  preliminary  work  in  the  creation  of  a   circulation   microfluidic   device   where   the   β-­‐cell   will   be   suspended   in   a   well   above   a   2+ membrane   over   a   long   channel   of   circulating   RBCs.     After   stimulation   of   the   β-­‐cell,   Zn   and  C-­‐peptide  will  be  release  into  the  blood  stream.    There  will  be  a  measurement  well   in   order   to   look   at   the   ATP   release   from   the   RBCs.     By   blocking   different   parts   of   the   secretion,   the   process   of   β-­‐cell   secretion   and   the   interaction   with   RBCs   and   the   subsequent  ATP  release  will  be  investigated,  to  further  prove  that  the  effects  seen  in  the   lab  also  occur  in  vivo.     Despite  decades  of  finding  biological  relevance,  there  has  no  been  receptor  found  for  C-­‐ 2+ peptide   at   the   time   of   writing.     Because   of   the   effect   of   Zn   bound   to   C-­‐peptide   on   glucose   uptake   and   ATP   release,   GLUT1,   the   major   glucose   transporter,   is   currently   2+ being   investigated   as   the   receptor   on   RBCs.     It   is   known   that   Zn   has   an   effect   on   30 glucose   transport   in   certain   cell   types   that   is   believed   to   be   mediated   through   GLUT1.     2+ As   previously   mentioned,   C-­‐peptide   will   interact   with   the   RBC   whether   or   not   Zn   is   2+ present,  however  Zn  is  necessary  for  biological  activity.    GLUT1  can  be  purified  from   2+ the  membranes  of  RBCs  and  can  be  used  for  binding  studies  with  C-­‐peptide,  Zn ,  and   2+ Zn   bound   to   C-­‐peptide.     Work   using   gel   electrophoresis   has   shown   a   potential   increase   in   GLUT1   in   the   membrane   of   RBCs   of   MS   patients   as   compared   to   those   of   healthy   controls.     To   discern   the   actual   amount   of   this   protein   present,   future   work   will     154   use   antibodies   for   GLUT1   to   examine   the   concentration   using   either   ELISA   or   flow   cytometry.     5.2.2  Future  MS  Studies       While   investigating   blood   samples   from   22   MS   patients   has   given   a   wealth   of   preliminary   data,   the   large   majority   of   these   patients   were   diagnosed   with   relapsing-­‐ remitting  MS  and  16  of  the  20  were  female.    Running  the  same  tests  of  patients  with  the   other   types   of   MS,   as   well   as   clinically   isolated   syndrome   (CIS)   will   result   in   further   information   about   the   disease,   as   well   as   show   when   these   abnormalities   begin   to   occur,  and  to  what  extent.    With  a  large  sample  group,  the  effect  of  disease  duration  on   these  factors  can  be  further  investigated.    It  is  the  hope  that  the  ATP  release,  glucose  or   2+ Zn   uptake   may   correlate   to   disease   state   or   severity,   as   well   as   being   a   specific   biomarker   for   MS.     To   ensure   this   is   the   case,   the   RBCs   of   several   control   groups   also   need   to   be   investigated.     To   prove   that   the   effects   seen   are   not   simple   the   result   of   any   neurological   or   inflammatory   disease,   control   samples   from   these   groups   will   be   tested.     For   example   RBCs   from   patients   with   Parkinson’s   disease   and   patients   with   arthritis   will   be  collected  and  tested  in  the  same  manner  as  those  from  the  MS  patients.     Very  preliminary  results  from  MS  patients,  along  with  the  results  discussed  in  chapter  2,   suggest   that   estriol,   as   well   as   estradiol,   will   be   effective   in   reducing   the   ATP   release   from   the   RBCs   of   MS   patients.     Future   MS   patient   RBC   samples   and   controls   will   be   incubated  with  various  physiological  concentrations  of  these  estrogens  and  along  with     155   2+ the  ATP  release,  glucose,  Zn  and  C-­‐peptide  uptake  will  also  be  evaluated.    Because  of   the  increased  cancer  risk  associated  with  estrogens,  the  effects  of  estrogen  mimics,  such   as  xenoestrogens,  should  be  investigated  for  their  effectiveness  as  well.    Alternatively,  if   the  reason  for  the  ameliorating  effects  of  estrogen  can  be  determined  and  ameliorated   with  a  different  compound,  that  compound  can  also  be  tested.     5.2.3  Microfluidics  and  MS  Studies       From   the   data   presented   in   this   thesis,   it   is   hypothesized   that   the   increase   in   ATP   release  from  the  RBCs  of  MS  patients  will  result  in  an  increase  in  NO  production  in  the   endothelial   cells   lining   the   blood   vessels.     In   order   to   explore   these   interactions,   the   microfluidic  device  described  and  employed  in  chapter  2.    Because  it  is  known  that  RBC-­‐ derived   ATP   stimulates   NO   production   in   endothelial   cells   and   that   the   RBCs   of   MS   patients   release   extremely   high   amount   of   ATP   in   response   to   sheer   stress,   it   is   anticipated  that  an  large  increase  in  NO  production  will  be  seen  using  this  method.     Because  blood  brain  barrier  (BBB)  breakdown  is  a  hallmark  feature  of  MS,  microfluidic   studies   should   also   be   completed   investigating   the   permeability   of   this   barrier.     Work   has   been   completed   in   the   Spence   Lab   on   the   measurement   of   cell   confluency   in   a   microfluidic   device.     Vogel,   et   al.   described   a   system   that   employs   transendothelial   32 electrical   resistance   measurement   (TEER)   on   a   flow-­‐based   microfluidic   device.     As   the   confluency   of   cells   cultured   on   the   device   increased,   so   did   the   electrical   resistance,   while   the   amount   of   charge   that   passed   through   the   cells   decreased.     This   allows   for     156   objective   measurement,   rather   than   relying   on   visual   inspection   to   determine   confluency.    The  cells  used  in  the  preliminary  experimentation  were  bovine  pulmonary   artery   endothelial   cells,   though   the   work   could   easily   be   expanded   to   use   brain   microvascular  endothelial  cells,  which  form  much  tighter  junctions  in  the  BBB.    The  used   of   these   cells   would   allow   for   the   study   of   the   effects   for   RBC-­‐derived   ATP   and   subsequent   endothelial   NO   production   on   the   integrity   of   BBB.     These   experiments   would  aid  in  the  further  understanding  of  the  role  of  the  RBC  in  the  etiology  of  MS.           157   REFERENCES     158       1     2     3     4     5     6     7     8     9     10   REFERENCES   Bergfeld,  G.  &  Forrester,  T.  Release  of  ATP  from  Human  Erythrocytes  in  Response   to   a   Brief   Period   of   Hypoxia   and   Hypercapnia.   Cardiovascular   Research   26,   40-­‐47   (1992).   Sprague,  R.  S.,  Ellsworth,  M.  L.,  Stephenson,  A.  H.  &  Lonigro,  A.  J.  ATP:  The  red   blood   cell   link   to   NO   and   local   control   of   the   pulmonary   circulation.   American   Journal   of   Physiology-­‐Heart   and   Circulatory   Physiology   271,   H2717-­‐H2722   (1996).   Letourneau,   S.,   Hernandez,   L.,   Faris,   A.   N.   &   Spence,   D.   M.   Evaluating   the   effects   of   estradiol   on   endothelial   nitric   oxide   stimulated   by   erythrocyte-­‐derived   ATP   using  a  microfluidic  approach.  Analytical  and  Bioanalytical  Chemistry  397,  3369-­‐ 3375  (2010).   Confavreux,   C.,   Hutchinson,   M.,   Hours,   M.   M.,   Cortinovis-­‐Tourniaire,   P.   &   Moreau,  T.  Rate  of  pregnancy-­‐related  relapse  in  multiple  sclerosis.  New  England   Journal  of  Medicine  339,  285-­‐291  (1998).   Sicotte,  N.  L.,  Liva,  S.  M.,  Klutch,  R.,  Pfeiffer,  P.,  Bouvier,  S.,  Odesa,  S.,  Wu,  T.  C.  &   Voskuhl,   R.   R.   Treatment   of   multiple   sclerosis   with   the   pregnancy   hormone   estriol.  Annals  of  Neurology  52,  421-­‐428  (2002).   Thiel,  V.  &  Audus,  K.  Nitric  oxide  and  blood-­‐brain  barrier  integrity.  Antioxidants  &   Redox  Signaling  3,  273-­‐278  (2001).   Sprague,   R.   S.,   Ellsworth,   M.   L.,   Stephenson,   A.   H.,   Kleinhenz,   M.   E.   &   Lonigro,   A.   J.  Deformation-­‐induced  ATP  release  from  red  blood  cells  requires  CFTR  activity.   American  Journal  of  Physiology  275,  H1726-­‐1732  (1998).   Singh,  A.  K.,  Schultz,  B.  D.,  Katzenellenbogen,  J.  A.,  Price,  E.  M.,  Bridges,  R.  J.  &   Bradbury,   N.   A.   Estrogen   inhibition   of   cystic   fibrosis   transmembrane   conductance   regulator-­‐mediated   chloride   secretion.   Journal   of   Pharmacology   and  Experimental  Therapeutics  295,  195-­‐204  (2000).   Sweezey,  N.,  Ghibu,  F.  &  Gagnon,  S.  Sex  hormones  regulate  CFTR  in  developing   fetal   rat   lung   epithelial   cells.   American   Journal   of   Physiology-­‐Lung   Cellular   and   Molecular  Physiology  272,  L844-­‐L851  (1997).   Chambliss,   K.   L.   &   Shaul,   P.   W.   Estrogen   modulation   of   endothelial   nitric   oxide   synthase.  Endocrine  Reviews  23,  665-­‐686  (2002).       159   11     12     13     14     15     16     17     18     19     20     Meyer,   J.   A.,   Froelich,   J.   M.,   Reid,   G.   E.,   Karunarathne,   W.   K.   &   Spence,   D.   M.   Metal-­‐activated  C-­‐peptide  facilitates  glucose  clearance  and  the  release  of  a  nitric   oxide  stimulus  via  the  GLUT1  transporter.  Diabetologia  51,  175-­‐182  (2008).   2+ Medawala,  W.  Mechanistic  and  Functional  Studies  of  Zn  Activation  of  C-­‐Peptide   and   its   Effect   on   Red   Blood   Cell   Metabolism   PhD   thesis,   Michigan   State   University,  (2011).   Meyer,  J.  A.,  Subasinghe,  W.,  Sima,  A.  A.  F.,  Keltner  Z.,  Reid,  G.  E.,  Daleke,  D.  &   2+ Spence,   D.   M.   Zn -­‐activated   C-­‐peptide   resistance   to   the   type   2   diabetic   erythrocyte   is   associated   with   hyperglycemia-­‐induced   phosphatidylserine   externalization  and  reversed  by  metformin.  Molecular  Biosystems  5,  1157-­‐1162   (2009).   Oblak,   T.   D.   A.,   Root,   P.   &   Spence,   D.   M.   Fluorescence   Monitoring   of   ATP-­‐ Stimulated,   Endothelium-­‐Derived   Nitric   Oxide   Production   in   Channels   of   a   Poly(dimethylsiloxane)-­‐Based   Microfluidic   Device.   Analytical   Chemistry.   78,   3193-­‐3197  (2006).   Sprague,   R.   S.,   Olearczyk,   J.   J.,   Spence,   D.   M.,   Stephenson,   A.   H.,   Sprung,   R.   W.   &   Lonigro,   A.   J.   Extracellular   ATP   signaling   in   the   rabbit   lung:   erythrocytes   as   determinants   of   vascular   resistance.   American   Journal   of   Physiology-­‐Heart   and   Circulatory  Physiology  285,  H693-­‐H700  (2003).   Smith,  K.  J.  &  Lassmann,  H.  The  role  of  nitric  oxide  in  multiple  sclerosis.  Lancet   Neurology  1,  232-­‐241  (2002).   Boje,   K.   M.   K.   &   Lakhman,   S.   S.   Nitric   oxide   redox   species   exert   differential   permeability   effects   on   the   blood-­‐brain   barrier.   Journal   of   Pharmacology   and   Experimental  Therapeutics  293,  545-­‐550  (2000).   Adinolfi,   E.,   Pizzirani,   C.,   Idzko,   M.,   Panther,   E.,   Norgauer,   J.,   Di   Virgilio,   F.   &   Ferrari,  D.  P2X(7)  receptor:  Death  or  life?  Purinergic  Signal  1,  219-­‐227  (2005).   Matute,  C.,  Torre,  I.,  Pérez-­‐Cerdá,  F.,  Pérez-­‐Samartín,  A.,  Alberdi,  E.,  Etxebarria,   E.,   Arranz,   A.   M.,   Ravid,   R.,   Rodríguez-­‐Antigüedad,   A.,   Sánchez-­‐Gómez,   M.   &   Domercq,   M.   P2X(7)   receptor   blockade   prevents   ATP   excitotoxicity   in   oligodendrocytes   and   ameliorates   experimental   autoimmune   encephalomyelitis.   Journal  of  Neuroscience  27,  9525-­‐9533  (2007).   Sharp,  A.  J.,  Polak,  P.  E.,  Simonini,  V.,  Lin,  S.  X.,  Richardson,  J.  C.,  Bongarzone,  E.   R.   &   Feinstein,   D.   L.   P2x7   deficiency   suppresses   development   of   experimental   autoimmune  encephalomyelitis.  Journal  of  Neuroinflammation  5,  33  (2008).   160     21     22     23     24     25     26     27     28     29   Jansson,  L.,  Olsson,  T.  &  Holmdahl,  R.  Estrogen  induces  a  potent  suppression  of   experimental   autoimmune   encephalomyelitis   and   collagen-­‐induced   arthritis   in   mice.  Journal  of  Neuroimmunology  53,  203-­‐207  (1994).   Palaszynski,   K.   M.,   Liu,   H.,   Loo,   K.   K.   &   Voskuhl,   R.   R.   Estriol   treatment   ameliorates  disease  in  males  with  experimental  autoimmune  encephalomyelitis:   implications   for   multiple   sclerosis.   Journal   of   Neuroimmunology   149,   84-­‐89   (2004).   Sicotte,  N.  L.,  Liva,  S.  M.,  Klutch,  R.,  Pfeiffer,  P.,  Bouvier,  S.,  Odesa,  S.,  Wu,  T.  C.  &   Voskuhl,   R.   R.   Treatment   of   multiple   sclerosis   with   the   pregnancy   hormone   estriol.  Annals  of  Neurology  52,  421-­‐428  (2002).   Raczkiewica,   J.   &   Leyko,   W.   Adenine   Nucleotide   Content   of   Erythrocytes   of   Multiple  Sclerosis  Patients.  Clinica  Chimica  Acta  14,  408-­‐409  (1966).   Regenold,   W.   T.,   Phatak,   P.,   Makley,   M.   J.,   Stone,   R.   D.   &   Kling,   M.   A.   Cerebrospinal   fluid   evidence   of   increased   extra-­‐mitochondrial   glucose   metabolism   implicates   mitochondrial   dysfunction   in   multiple   sclerosis   disease   progression.  Journal  of  the  Neurological  Sciences  275,  106-­‐112  (2008).   Henry,  J.  P.,  Williamson,  D.  M.,  Schiffer,  R.,  Wagner,  L.,  Shire,  J.  &  Garabedian,  M.     Investigation  of  a  cluster  of  multiple  sclerosis  in  two  elementary  school  cohorts.   Journal  of  Environmental  Health  69,  34-­‐38  (2007).   Schiffer,   R.   B.,   McDermott,   M.   P.   &   Copley,   C.   A   multiple   sclerosis   cluster   associated   with   a   small,   north-­‐central   Illinois   community.   Archives   of   Environmental  Health  56,  389-­‐395  (2001).   Stein,   E.   C.,   Schiffer,   R.   B.,   Hall,   W.   J.   &   Young,   N.   Multiple-­‐Sclerosis   and   the   Workplace   -­‐   Report   of   an   Industry-­‐Based   Cluster.   Neurology   37,   1672-­‐1677   (1987).   Dore-­‐Duffy,   P.,   Catalanotto,   F.,   Donaldson,   J.   O.,   Ostrom,   K.   M.   &   Testa,   M.   A.   2+ Zn  in  Multiple-­‐Sclerosis.  Annals  of  Neurology  14,  450-­‐454  (1983).     30     31     2+ Shay,   N.   F.   &   Tang,   X.   H.   Zn   has   an   insulin-­‐like   effect   on   glucose   transport   mediated   by   phosphoinositol-­‐3-­‐kinase   and   AKT   in   3T3-­‐L1   fibroblasts   and   adipocytes.  Journal  of  Nutrition  131,  1414-­‐1420  (2001).   Dorman,   J.   S.,   Steenkiste,   A.   R.,   Burke,   J.   P.   &   Songini,   M.   Type   1   diabetes   and   multiple  sclerosis.  Diabetes  Care  26,  3192-­‐3193  (2003).   161   32         Vogel,   P.,   Halpin,   S.,   Martin,   R.   &   Spence,   D.   Microfluidic   Transendothelial   Electrical   Resistance   Measurement   Device   that   Enables   Blood   Flow   and   Postgrowth  Experiments.  Analytical  Chemistry  83,  4296-­‐4301  (2011).   162