Using genome editing to introduce naturally occurring mutations associated with elevated foetal haemoglobin

Abstract

Beta-haemoglobinopathies are amongst the most common inherited diseases in the world with devastating prospects for the affected individuals if untreated. The discovery that high foetal haemoglobin (HbF) levels are beneficial for patients leading to less severe symptoms has been one of the key drivers of haemoglobin research. Naturally occurring mutations in the promoter region of foetal gamma-globin result in the continued expression of foetal haemoglobin into adulthood - a benign condition known as Hereditary Persistence of Foetal Haemoglobin (HPFH). Individuals with HPFH have foetal haemoglobin levels between 3 % and 40 % whilst the normal adult only produces about 1 % of HbF. The high foetal haemoglobin levels in individuals with HPFH are sufficient to ameliorate the symptoms in individuals with beta-haemoglobinopathies such as beta-thalassaemia and sickle cell anaemia. The purpose of our research is to explore reactivation of foetal globin expression in adult life as a therapeutic strategy by developing mechanistic understanding and by introducing these advantageous HPFH mutations in cell models. Here we introduced three different naturally occurring HPFH mutations into various erythroid cell models by TALEN- and CRISPR/Cas9-mediated genome editing and found that this resulted in elevated levels of HbF. Thus, we propose that introducing these mutations into patients with beta-haemoglobinopathies could represent a possible gene therapeutic approach to ameliorate symptoms. Furthermore, we were able to uncover the molecular mechanisms underlying these HPFH mutations through in vitro and in vivo binding studies. We demonstrated that the -175 T>C and the -198 T>C mutations create de novo binding sites for the erythroid specific activators TAL1 and KLF1, respectively. Chromatin conformation capture experiments revealed that TAL1 mediates looping of the LCR to the gamma-globin promoter through recruitment of LMO2 and LDB1 to activate foetal globin expression. We also provide evidence that a cluster of HPFH mutations around 200 bp upstream of the gamma-globin transcription start site decreases binding of the foetal globin repressor ZBTB7A. Overall, we deliver three different mechanistic explanations for non-deletional HPFH in humans. By uncovering the molecular basis underlying these mutations we made a significant contribution to better understanding the foetal to adult haemoglobin switch.