Regulated Expression of GATA1 As a Gene Therapy Cure for Diamond-Blackfan Anemia

Diamond-Blackfan anemia (DBA) is a bone marrow failure disorder characterized by impaired red blood cell production leading to severe anemia. The only treatment options are regular transfusions, corticosteroids, or bone marrow transplant, and each has variable effectiveness and considerable morbidity. More effective, safer, and permanent cures for DBA are lacking, but few drugs are in active development. DBA is genetically heterogenous, characterized by loss-of-function mutations in one copy of at least 26 different ribosomal protein (RP) genes that alters the mRNA translation of a select group of genes, most notably the hematopoietic master regulator GATA1. Overexpression of GATA1 in hematopoietic stem and progenitor cells (HSPCs) from patients with DBA rescues the erythroid differentiation defects in vitro, but unregulated GATA1 expression in hematopoietic stem cells (HSCs) promotes premature erythroid differentiation at the expense of long-term (LT-) HSC maintenance. Gene therapy is an attractive strategy to achieve a cure for DBA, but a traditional gene therapy approach that overexpresses a functional copy of a mutated gene is not the ideal option, as it would require the development and validation of dozens of gene therapy vectors, each with a copy of one of the mutated DBA genes. Here, we report a unified gene therapy strategy, using developmentally-regulated and highly restricted expression of GATA1 that would be curative in most, if not all, patients with DBA, regardless of the underlying disease-causing mutation.

To be a cure for DBA, a GATA1 gene therapy vector must incorporate into the genome of undifferentiated LT-HSCs but only drive robust GATA1 expression in early erythroid progenitors after lineage commitment. To design a vector with lineage-restricted expression, we identified accessible chromatin regions upstream of GATA1 that only gain accessibility in differentiating erythroid cells, but not in upstream HSPCs. Using these selectively accessible regions concatenated together, we constructed a human GATA1 enhancer (hG1E) element and used it to drive GATA1 and/or GFP expression from a lentiviral backbone. To ensure erythroid-restricted expression of the hG1E-GATA1-GFP vector, we infected CD34⁺ HSCs from healthy human donors and cultured them in conditions that either allowed HSC maintenance or promoted erythrocyte differentiation. We observed very low GFP expression in bulk HSCs and essentially no expression in the subpopulation of immunophenotypic LT-HSCs. However, more than 80% of committed erythroid cells demonstrated high levels of GFP expression, confirming the fidelity of lineage-restricted expression from the hG1E-GATA1 vector. To examine whether lineage-restricted GATA1 expression confers a functional correction of the erythroid maturation block in DBA, we used CRISPR editing to recapitulate RPS19 haploinsufficiency in human HSPCs. In RPS19 edited samples that were infected with a control GFP virus, we observed a progressive depletion over time of RPS19 edits as the cells underwent erythroid differentiation, consistent with the notion that RPS19 editing leads to impaired erythroid differentiation and are selected against. In contrast, RPS19-edited and hG1E-GATA1 treated samples had a more than 10-fold higher preservation of RPS19 edits in the bulk population and in erythroid colonies derived from single progenitor clones, indicating that regulated GATA1 expression could overcome RPS19 haploinsufficiency.

To validate the results from the CRISPR model of DBA, we sought to correct the erythroid maturation block in primary HSCs from a DBA patient with a pathogenic mutation in RPL5. There was skewed differentiation of GFP-infected RPL5^+/- HSCs away from the erythroid lineage and toward myeloid and megakaryocyte lineages, but regulated GATA1 expression from the hG1E-GATA1 vector reversed that skewing. There was a 10-fold increase in the ratio of erythroid to non-erythroid cells, and a significant expansion of erythroid cells in the hG1E-GATA1 treated patient cells. Preclinical assessment in additional patient samples both in vitro and in vivo is ongoing.

Collectively, our preclinical data demonstrate that lineage-restricted expression of GATA1 is sufficient to rescue the impaired erythroid differentiation in DBA regardless of the underlying genetic lesion. These results will set the stage for the first in-human gene therapy trial for DBA.

Williams:Novartis: Consultancy, Other: Steering Committee (fees donated to NAPAAC); Insertion Site Analysis Advisory Board, Bluebird Bio: Consultancy; Orchard Therapeutics: Other: Provides vector; Bluebird: Consultancy, Other: Provides Vector; Chief Scientific Chair, Emerging Therapy Solutions: Consultancy; Scientific Advisory Board, Skyline Therapeutics: Consultancy; Scientific Advisory Board, Beam Therapeutics: Consultancy; Insertion Site Advisory Board, Biomarin: Consultancy; Novartis: Other: Provision of study materials, medical writing. Sankaran:Novartis: Consultancy; Branch Biosciences: Consultancy; Sana Biotechnology: Consultancy; Cellarity: Consultancy; Ensoma: Consultancy; Forma Therapeutics: Consultancy.