Gene therapy for β-thalassaemia: targeted modification of the human β-globin locus

© 2018 Dr. Astrid GlaserThe β-haemoglobinopathies, caused by insufficient synthesis (β-thalassaemia) or structural defects (sickle cell disease) of the β-globin protein, are the most prevalent inherited blood disorders worldwide. Due to severe haemolytic anaemia, most patients depend on regular blood transfusions throughout life. The high morbidity and mortality from these disorders constitutes a severe burden for global health care systems and affected families. Recent clinical trials using lentiviral gene addition therapy have demonstrated remarkable success in patients with β-haemoglobinopathies. However, current lentiviral constructs fall short of physiological β-globin transgene expression due to size constraints of this vector type. Genome editing, using programmable endonucleases, could be used to overcome this limitation. With particular focus on β-thalassaemia, this PhD project explored the potential of different genome editing strategies for the therapy of β-haemoglobinopathies using the CRISPR/Cas9 genome editing platform. Restoration of physiological β-globin gene expression through homology-directed repair (HDR)-mediated gene correction is the optimal outcome for therapeutic genome editing. However, gene correction is often overshadowed by disruptive mutations created through non-homologous end-joining (NHEJ). To facilitate the identification of interventions that bias genome editing towards HDR, a simple assay for the quantification of HDR and NHEJ frequencies was developed. Substitution of two adjacent amino acids in the commonly used fluorescent reporter EGFP via HDR converts EGFP to BFP. Conversely, disruptive mutations resulting from NHEJ lead to loss of fluorescence. HDR and NHEJ can thus be quantified using flow cytometry as blue fluorescence and loss of fluorescence, respectively. A small pilot screen performed in EGFP-modified K562 and HEK293T cells demonstrated the feasibility of this assay for a high-throughput format. Next, genome editing was applied to the creation of a novel cellular model of β-thalassaemia via NHEJ-mediated knockout of the β-globin gene in human erythroid HUDEP-2 cells. Five clonal β0-HUDEP-2 cell lines were created. Characterisation of these cell lines via morphological analysis and flow cytometry revealed differentiation defects characteristic for β-thalassaemia, which were corrected using a clinical gene therapy vector. Notably, treatment with pharmacological stimulators of γ-globin expression showed that β0-HUDEP-2 cells have an increased sensitivity to the reactivation of fetal γ-globin, a known disease modifier for β-haemoglobinopathies. Lastly, the recreation of the 7.2 kb Corfu deletion via NHEJ was attempted in β0-HUDEP-2 cells as an HDR-independent therapeutic genome editing strategy for the reactivation of γ-globin expression. Deletions were successfully introduced in at least one chromosome in ~30% of cells, indicating that deletions of this size can be generated with high efficiency inerythroid cells. However, random reactivation of γ-globin expression upon clonal expansion prevented further assessment of potential therapeutic benefits of the Corfu deletion in β0-HUDEP-2 cells. While no indication for the suitability of the Corfu deletion as a target for therapeutic genome editing was found in this study, this data further supports the notion that β0-HUDEP-2 cells have an increased γ-globin reactivation potential. We therefore propose that adult erythroid cells lacking β-globin expression may be primed for γ-globin reactivation. β0 erythroid model systems such as β0-HUDEP-2 cells could therefore aid in the identification of novel γ-globin inducers