Multiplex CRISPR/Cas9 Genome Editing Targeting the BCL11A/HBG Axis Maximizes Fetal Hemoglobin Reinduction but Generates Chromosomal Translocations Which Persist In Vivo

Sickle cell disease and β-thalassemia are among the most common monogenic disorders globally, causing significant morbidity and early mortality. The only curative option available is allogeneic hematopoietic stem cell transplantation, which is limited by a lack of matched donors and risks, including graft versus host disease and secondary malignancy. Retroviral gene transfer is being explored in clinical trials, but an alternative approach is more targeted CRISPR/Cas9 genome editing, to recapitulate naturally occurring hereditary persistence of fetal hemoglobin (HPFH) which ameliorates disease. HbF reinduction is achieved by disrupting either transcription factor binding sites within the HBG promoter, or an erythroid enhancer sequence within the HbF repressor, BCL11A. Results from preclinical studies have suggested that HbF levels may remain suboptimal when each locus is targeted individually. We thus investigated the feasibility of a dual editing approach, targeting both loci simultaneously or sequentially, to maximize HbF production.

G-CSF-mobilized human peripheral blood hematopoietic stem and progenitor cells (HSPCs) underwent CRISPR/Cas9 genome editing at either HBG promoter, BCL11A erythroid enhancer, or both loci targeted either simultaneously or sequentially. Immunodeficient mice were transplanted with edited cells.

HSPCs cultured in differentiation media demonstrated comparable levels of editing at each locus in the single edited arms (by TIDE).The dual editing approach did not impair editing efficiency at each site when conducted sequentially and only slightly reduced efficiency in the simultaneously-edited reactions. HbF reinduction was also greatest in sequentially double-edited reactions, with HbF/HbA ratio as high as 3.9 times that seen in mock reactions by flow cytometry and 4.1 times by HPLC. In single-edited arms, HbF reinduction was slightly greater with HBG promotor than BCL11A erythroid enhancer editing. Two separate chromosomal translocation events encompassing both loci were detected in each double-edited arm and quantified using digital droplet PCR. Both were more frequent in simultaneous (mean of 1.0% and 0.6%) than sequential reactions (means <0.2%, n=3, p=0.0420 and p=0.0296).

Analysis of single BFU-E colonies grown on methocult media revealed markedly different indel patterns at the 2 target loci. Concurrent hemoglobin fraction analysis of these clonal populations by HPLC demonstrated greatest HbF proportion in double-edited colonies. Where ≥60% editing was reported at both loci mean HbF was 71.3%, compared to 47.1% with ≥60% editing at HBG only, and 29.6% with ≥60% editing at BCL11A only (p=0.0373).

Following transplant into mice, engraftment and lineage differentiation were comparable among all experimental arms. At necropsy, bone marrow populations of human CD45⁺ cells, CD34⁺CD38^low HSPCs and, within these, HSC-enriched CD90⁺CD45RA^- subpopulation, were present at comparable levels indicating that the dual editing approach did not impair engraftment. Translocation events were detected, albeit at frequencies of less than 0.25%, in the bone marrows of all 4 mice transplanted with simultaneously dual-edited cells and in 5 of 6 mice transplanted with sequentially dual-edited cells. Bone marrow cells cultured ex vivo demonstrated greatest HbF from mice transplanted with sequentially double-edited cells.

In summary, we present evidence of maximized HbF reinduction with sequentially applied multiplex genome editing at BCL11A erythroid enhancer and HBG promoter, with no impairment of engraftment or differentiation. However, chromosomal translocation events were consistently seen in double-edited reactions, even when edits at each locus were applied sequentially, and persisted in vivo after engraftment, thus rendering this approach inappropriate for clinical applications. However, multiplex editing and targeting both the BCL11A erythroid enhancer and HBG promoter to maximize fetal hemoglobin induction may be a promising strategy for alternative platforms such as base editors which are expected to greatly reduce or completely eliminate the occurrence of translocations.