Advanced personalized gene therapy of B-thalassaemia

Student thesis: Doctoral ThesisDoctor of Philosophy

Abstract

Thalassaemias are amongst the commonest single‐gene disorders worldwide, but offer only limited curative treatment choices. Autologous transplantation of gene‐therapy‐corrected cells is therefore investigated  by  numerous  groups  and  is  already  under  clinical  trials  for  β‐thalassaemia,  based  on  gene addition by lentiviral vectors (LVs). The  main  aim  of  this  project  was  the  development  of  personalised  gene  therapy  (GT)  for  β‐thalassaemia caused by the common and severe HBBIVSI‐110 mutation, which results in missplicing of intron 1. The resulting aberrant mRNA interferes with protein expression from normal endogenous or vector‐encoded HBB. The two therapy approaches taken were a) modification or co‐transduction of the LV ΜΑ821_ΗΒΒ  (“GLOBE”)  to achieve  concurrent  vector‐derived  HBB  expression  and  RNAi‐based suppression of aberrant RNA and b) development of designer nucleases for genome editing and functional correction of the HBBIVSI‐110 mutation. To  establish  proof  of  principle  for  both  approaches  we  developed  transgenic  murine  erythroleukaemia cell lines MEL MA821‐HBBIVSI‐110 and MEL MA821‐HBBNormal with on average two vector copies per genome (VCN) and an additional, clonal cell line of VCN=1 for MEL MA821 HBBIVSI‐110. Approach a) Depletion  of  aberrant  HBBIVSI‐110  mRNA  might  enhance  translation  of  vector‐derived  or  residual  endogenous HBB mRNA and could be achieved by HBBIVSI‐110‐specific RNAi, mediated by expression of lentivirally delivered shRNAs. To this end, four different HBBIVSI‐110 mRNA‐specific shRNAs were cloned into the U6‐promoter‐driven pLKO.I lentiviral vector (LV). In clonal MEL HBBIVSI‐110 cells, two constructs, pLKO.I shIVSI‐110 Mid & Mid2,  exhibited no  discernible change  of normal:aberrant HBB  mRNA  but  a  significant  increase of  HBB chains relative to untransduced samples. Importantly, our approach was then validated by transduction  of  HBB IVSI‐110‐patient‐derived  hCD34+ cells,  in which we  observed a ~30% increase of  HBB:HBA chain ratios in samples transduced with shIVSI‐110 MID, regardless of co‐transduction with other vectors. These data indicate that RNAi‐targeting of aberrant HBB mRNAs, if not therapeutic in its own right, could substantially improve efficiency of GT by HBB gene augmentation and could thus lower conditioning, VCN and gene‐expression requirements for LV‐based gene therapies. Approach b) As an alternative to gene augmentation we established a novel genome‐editing approach in which permanent functional correction the HBBIVSI‐110 mutation was achieved with the targeted disruption of  the  aberrant  SA  site  and/or  its  sequence  context.  This  approach,  based  on  the  efficient  non‐homologous end‐joining (NHEJ) repair pathway, used HBBIVSI‐110 ‐ specific designer nucleases, transcription  activator‐like  effector  nucleases  (TALENs)  and  clustered  regularly  interspaced  short  palindromic repeats (CRISPR)/Cas9 RNA‐guided endonuclease (RGENs). Assessment  of  targeted  disruption  efficiency  of  our  designer  nucleases  on  gDNA  or  episomal  reporter plasmids indicated the superiority of the HBB TALEN over RGEN for HBBIVSI‐110. Plasmid transfection of HBBIVSI‐110 specific ‐ designer nucleases into MEL MA821‐HBBIVSI‐110 transgenic cells  resulted  in  functional  correction  at  the  RNA  (RT‐qPCR)  and  protein  level  (Immunoblots).  In  addition, full characterisation (type, frequency and context) of induced insertions/deletions (INDELs) was  achieved  by  T7  Endonuclease  1  assay,  decomposition  of  sequence  traces  (TIDE)  and  Sanger  sequencing of TOPO clones from edited bulk cells. Using edited MEL MA821‐HBBIVSI‐110 clones we moreover correlated specific  INDELs  with  HBB expression at  the  RNA  and protein  level, which  confirmed our hypothesis that functional correction of splicing could be achieved by disruption of upstream sequences of the aberrant SA site, leaving intact the HBBIVSI‐110 mutation. Subsequent  validation  of  our  nucleases  in  patient‐derived  CD34+  cells  by  electroporation  of  HBB  TALEN L1/R1 plasmid DNA was impaired by low targeted disruption (12%) despite high‐transfection efficiencies. Even though a marginal increase in the ΗΒΒ:ΗΒΑ chain ratio was detected, analyses of mRNA levels were inconclusive, calling  for additional experimentation  in  CD34+  cells  with  more  appropriate expression systems. Overall,  proof  of  principle  was  established  for  NHEJ‐based  functional  repair  of  HBBIVSI‐110,  and  we  suggest that the approach can be adapted for the functional correction of other of disease‐causing gain‐of‐function mutations in non‐coding regulatory regions.
Date of Award2018
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorMichael Antoniou (Supervisor)

Cite this

'