Gene Therapy for Germline Predisposition to Myeloid Malignancies: Insights from a Nonhuman Primate RUNX1 Model

Pathogenic germline variants in RUNX1 cause familial platelet disorder with predisposition to myeloid malignancy (RUNX1-FPDMM), which is associated with thrombocytopenia, abnormal platelet function, and the development of myeloid malignancies.^1,2  RUNX1 variants, both somatic and germline, are common in acute myeloid leukemia and myelodysplastic syndrome, and these variants may be germline in up to 30% of AML cases.³  Recent advances in gene therapy have raised the question of whether gene-editing technology could be used to “correct” germline pathogenic variants that predispose to myeloid malignancies to prevent the cancer from developing.⁴ 

Dr. Byung-Chul Lee and colleagues recently developed a nonhuman primate model to study the potential impact of gene editing for RUNX1-FPDMM. Following the mobilization of CD34+ hematopoietic stem and progenitor cells (HSPCs) from two rhesus macaques, the investigators performed gene editing using CRISPR/Cas9 to introduce loss-of-function variants in RUNX1, as well as in AAVS1 for the controls. The RUNX1- and AAVS1-edited HSPCs were then pooled together. The rhesus macaques received conditioning with two days of total body irradiation and underwent autologous hematopoietic cell transplantation with these pooled HSPCs. After engraftment, the variant allele frequencies of the RUNX1 and AAVS1 variants were tracked in the peripheral blood cells and the bone marrow for more than 18 months to assess the relative competitive advantage of RUNX1-mutated versus AAVS1 control HSPCs.

Following autologous transplantation and engraftment, peripheral blood analysis demonstrated that both rhesus macaques initially had a higher proportion of AAVS1-edited granulocytes than RUNX1-edited cells. However, in both animals the proportion of RUNX1-edited cells gradually increased over time relative to the AAVS1-edited cells, suggesting a possible competitive advantage for the HSPCs with loss of function in RUNX1. The frequency of RUNX1-edited CD34+ bone marrow cells remained stable or increased over time. Additional experiments confirmed that the rhesus macaque model recapitulates at least some aspects of the human RUNX1-FPDMM phenotype, including impaired in vitro megakaryocytic differentiation of HSPCs, megakaryocytic dysplasia, and a decrease in adenosine diphosphate–induced platelet aggregation.