• No se han encontrado resultados

Distribución de la Tenencia de la Tierra

2.2 La reforma agraria y la ocupación habitacional de los ejidos

TCRs  specific  for  cancer  epitopes  are  generally  characterised  by  low  binding  affinities  (binding  KDs  high   micromolar  range)  (Bridgeman  et  al.,  2011).  This  lower  binding  affinity  is  thought  to  be  a  result  of   negative  selection  of  T-­‐cells  that  bear  TCRs  with  higher  affinity  for  self-­‐ligands  in  the  thymus.  Since   TCR  affinity  plays  an  important  role  in  T-­‐cell  activation,  the  TCR  affinity  gap  between  anti-­‐pathogen   and  anti-­‐cancer  T-­‐cells  leaves  the  latter  at  a  distinct  disadvantage  and  makes  it  more  difficult  to  break   self-­‐tolerance   to   such   antigens.   One   approach   to   enhance   the   T-­‐cell   response   to   tumour   antigen-­‐ derived   peptides   has   been   to   immunize   patients   with   altered   peptide   ligands   that   differ   from   the   native  sequence  by  a  single,  or  multiple,  amino  acid  residues  (Lesterhuis  et  al.,  2011).  However,  such   ‘heteroclitic’  peptides  with  even  single  amino  acid  substitutions  that  are  predicted  to  only  contact  the   HLA   can   have   unpredictable,   yet   important,   effects   on   TCR   engagement.   Despite   their   extensive   application  in  clinical  trials  as  cancer  vaccines,  to  date  only  a  few  X-­‐ray  structures  of  TCRs  bound  to   cognate  tumour  antigens  have  been  determined  (Borbulevych  et  al.,  2011;  Chen,  2005;  Deng  et  al.,   2007;   Madura   et   al.,   2015).   I   have   solved   the   first   crystal   structure   of   a   HLA-­‐A2   restricted   gp100   peptide  antigen  bound  to  a  cognate  αβ  TCR.  In  this  chapter,  I  show  how  the  PMEL17  TCR  bound  with   a  typical  diagonal  orientation  over  the  central  peptide  residues,  and  mainly  contacted  residues  4,  7   and  8  of  the  YLE-­‐9V  peptide  which  protruded  out  of  the  HLA-­‐A2  binding  groove.  Interestingly,  the   PMEL17  TCR  was  characterized  by  a  binding  affinity  (KD)  of  7.6  μM,  a  value  that  falls  in  the  very  high   end  of  affinity  ranges  described  so  far  for  cancer  TCR  and  pHLA  interactions  (Bridgeman  et  al.,  2011;   Aleksic   et   al.,   2012).   These   observations   suggest   that   healthy   donors   or   melanoma   patients   may   harbour  T-­‐cells  bearing  TCRs  with  reasonable  affinity  for  some  tumour  associated  antigens,  which  can   be  preferentially  chosen  for  TCR-­‐based  applications.  For  example,  gp100-­‐specific  ImmTACs  (Immune-­‐ mobilising  monoclonal  TCRs  against  cancer)  are  a  new  class  of  soluble  bi-­‐specific  anti-­‐tumour  agents   that   combine   a   high-­‐affinity   TCR-­‐based   gp100   recognition   domain   with   a   T   cell   activation   domain   (Liddy  et  al.,  2012;  Bossi  et  al.,  2014).  IMCgp100  is  being  tested  as  a  soluble  drug  and  is  showing  partial   or  complete  durable  responses  in  Phase  I/IIa  trial  in  patients  with  advanced  melanoma  (Middleton  et   al.,  2015).  

 

In   this   chapter   I   also   provide   insight   into   YLE  single   amino   acid   contribution   to   TCR   binding   by   performing  an  alanine  scan  mutagenesis  across  the  peptide  backbone  with  two  different  YLE-­‐specific   αβTCRs.  Interestingly,  both  PMEL17  TCR  and  gp100  TCR  were  most  sensitive  to  mutations  at  position   3   or   5   of   the   native   YLE   peptide   sequence   despite   these   TCRs   being   constructed   from   completely   different  Vα  and  Vβ  genes.  These  results  are  supported  by  a  recent  study  of  YLE  altered  peptide  ligands   which  described  YLE-­‐3A  as  a  null  agonist  for  a  different  TCR  (Shaft  et  al.,  2003;  2013).  Overall,  along   with  the  two  TCRs  studied  here,  the  sequences  of  further  two  distinct  YLE-­‐specific  TCRs  have  been   published,  demonstrating  diverse  gene  usage  and  different  CDR3  loop  sequences  (Table  3.3).     No  structural  data  supporting  these  observations  have  been  published  to  date.  

             

Table  3.3.  Alignment  of  TCR  CDR3  regions  of  four  gp100-­‐specific  TCRs  

PMEL17,  gp100,  MPD  (Schaft  et  al.,  2003)  and  296  (Schaft  et  al.,  2003)  gp100-­‐specific  TCR.  

TCR   CDR1α   CDR2α   CDR3α   CDR1β   CDR1β   CDR1β  

PMEL17   DSAIYN IQSSQRE CAVLSSGGSNYKLTFG SGHTA FQGTGA CASSFIGGTDTQYFG

gp100   TSINN IRSNERE CATDGDTPLVFG LNHDA SQIVND CASSIGGPYEQYFG

MPD   KALYS LLKGGEQ CGTETNTGNQFYFG SGHDY FNNNVP CASSLGRYNEQFFG

296   DSASNY IRSNVGE CAASTSGGTSYGKLTFG MNHEY SMNVEV CASSLGSSYEQYFG

     

Interestingly,  mutation  in  position  3  in  the  YLE  peptide  did  not  alter  the  conformation  of  the  peptide   backbone  itself,  but  resulted  in  a  ‘knock-­‐on’  effect  on  the  neighbouring  residue  Pro4  that  completely   abolished  TCR  binding  and  T-­‐cell  recognition.  This  can  be  explained  by  the  fact  that  Pro4  was  at  the   centre  of  a  sizeable  network  of  interactions  (both  vdW  and  hydrogen  bonds)  in  the  PMEL17-­‐A2-­‐YLE-­‐ 9V  structure.  In  addition,  Position  3  in  HLA-­‐A2  restricted  peptides  is  known  to  be  a  secondary  anchor   residue  (Ruppert  et  al.,  1993),  in  that  it  supports  the  exposed  peptide  bulge  that  is  normally  involved   in  TCR  binding.  By  mutating  the  residue  in  position  3  with  a  smaller  side  chain,  this  support  is  lost   causing  a  ‘molecular  switch  in  the  neighbour  Pro4.  A  similar  mechanism  in  an  HIV-­‐1  derived  peptide,   has  recently  been  described  by  our  group,  with  important  implications  for  the  immune  control  of  HIV   infection  and  patterns  of  viral  escape  mutants  (Kløverpris  et  al.,  2015).  Additionally,  the  existence  of  a   novel  mode  of  flexible  peptide  presentation  in  a  diabetes  model  has  been  demonstrated,  showing  the   potential   dynamic   nature   of   the   region   surrounding   the   HLA   F-­‐pocket   (Motozono   et   al.,   2015;   Borbulevych   et   al.,   2009).   Taken   together,   these   studies   support   the   notion   that   peptide-­‐HLA   interactions  are  more  plastic  and  dynamic  than  previously  appreciated,  with  obvious  implications  for   immune  recognition,  epitope  prediction  and  structural  modelling.  

   

Overall,  the  results  presented  here  represent  the  first  structural  insight  into  TCR  recognition  of  an   important  tumour  antigen,  targeted  by  many  clinical  therapies.  They  reveal  that  two  very  different   TCRs  share  a  similar  pattern  of  specificity,  demonstrated  by  their  near  identical  sensitivity  to  different   peptide  modifications.  Finally,  I’ve  shown  that  modification  to  peptide  residues  outside  of  the  TCR   binding  motif  can  have  unpredictable  knock-­‐on  effects  on  adjacent  peptide  residues  that  abrogate  TCR   binding   and   T-­‐cell   recognition,   highlighting   that   even   conservative   peptide   substitutions   can   have   unexpected  consequences  for  T-­‐cell  recognition  by  different  antigen-­‐specific  TCRs  due  to  ‘knock-­‐on’   structural  changes  in  the  HLA-­‐bound  peptide.  Such  ‘transmitted’  structural  changes  need  to  be  taken   into   consideration   when   designing   improved   peptides   for   cancer   vaccination.   Given   the   growing   evidence  that  plasticity  at  the  TCR-­‐pHLA  interface  can  influence  immune  recognition,  structural  and   biophysical  studies  of  binding  should  be  taken  into  account  when  attempting  to  design  altered  peptide   ligands  with  improved  immunogenicity.  

             

4   Dissection   of   T-­‐cell   responses   in   an   HLA-­‐A2+   remission