Abstract: S146

Type: Presidential Symposium

Presentation during EHA24: On Friday, June 14, 2019 from 16:00 - 16:15

Location: Hall 5

Background
Transient abnormal myelopoiesis (TAM) arises from fetal hematopoietic cells in 5-30% of neonates with Down syndrome (DS) and is characterized by the accumulation of immature megakaryoblasts with the triad of fetal origin, GATA1s mutation and trisomy 21 being necessary and sufficient for its induction. Yet, molecular mechanisms underlying this cooperation are incompletely understood.

Aims
We aimed to identify the oncogenic factors on human chromosome 21 (Hsa21) and unravel their molecular synergy with GATA1s during the development of TAM/ML-DS.

Methods
To study the mechanisms underlying the cooperativity between GATA1s and trisomy 21 in TAM/ML-DS pathogenesis, we performed a CRISPR-Cas9 screening targeting the 218 currently annotated coding genes on Hsa21 with 1090 sgRNAs in both a ML-DS and control cell line. The top ML-DS-specific candidates were functionally and molecularly validated in vitro and in vivo.

Results
Comprehensive CRISPR-Cas9 loss-of-function screening of the Hsa21 coding genes (n=218) revealed a strong and specific RUNX1 dependency indicated by the depletion of the ML-DS cell line CMK with minor effect on non-DS leukemia cells (K562). RNA-sequencing followed by isoform-specific qRT-PCR validation in leukemic blasts from ML-DS patients demonstrated deregulation of  the RUNX1 isoform equilibrium compared to other types of leukemia or hematopoietic stem and progenitor cells (HSPCs) from healthy donors. In an in vitro set up using Gata1s-mutated, pre-leukemic murine fetal HSPCs, we observed that ectopic deregulation of RUNX1 isoforms by overexpression synergized with Gata1s, resulting in enhanced proliferation and accumulation of immature megakaryocytic progenitors. Accordingly, ectopic expression of the ML-DS-dominant RUNX1 isoform in human CD34⁺ HSPCs led to a loss of mature megakaryocytes and increased monocytic differentiation. Inversely, shifting the expression towards the main hematopoietic RUNX1 isoform enhanced megakaryocytic differentiation and impaired proliferation of HSPCs. The same was observed in ML-DS patient blasts. Restoring the expression of the main hematopoietic RUNX1 isoform induced differentiation and cell cycle arrest of ML-DS blasts, while further elevating the expression of the ML-DS-dominant RUNX1 isoform enhanced proliferation and accumulation of a megakaryocytic CD41⁺CD117⁺ population. Importantly, overexpression of the RUNX1 isoform in Gata1s fetal HSPCs was sufficient to induce a leukemic phenotype upon transplantation into syngeneic C57BL/6J recipients after a short latency of 40 days and high penetrance (100%). The leukemic cells engrafted in secondary recipients and displayed a megakaryocytic progenitor-like phenotype (CD41⁺CD117⁺CD34^-CD16/32^low). Moreover, global gene expression profiling by RNA-sequencing confirmed a ML-DS-like RNA expression profile of the murine leukemias. Using ChIP-sequencing of tagged versions of GATA1 and GATA1s, we could determine differential chromatin occupancy at the RUNX1 locus implicating a feed-forward loop driving the expression of specific RUNX1 isoforms in GATA1s-mutated cells. Similarly, pulldown assays of tagged GATA1/GATA1s showed isoform specific interactions with RUNX1.

Conclusion
Our Hsa21-wide CRIPSR-Cas9 screening in combination with functional validation in vitro and in vivo places a specific RUNX1 isoform in the center of an interaction network with mutated GATA1s during the transformation of fetal HSPCs in the background of trisomy 21. Our data highlight the importance of alternative splicing in leukemias and will guide the development of truly specific and targeted therapies.

Session topic: 3. Acute myeloid leukemia - Biology & Translational Research

Keyword(s): AML, Down Syndrome, GATA-1, RUNX1