2026 ASH®
Annual Meeting & Exposition December 12-15, 2026 ▪ New Orleans, Louisiana

Paper Details

Abstracts, authors, affiliations, and disclosures.

Saturday, December 12, 2026
04:00 p.m. - 05:30 p.m.
MCCNO - R02-R05
Technology and Procedures, Assays, Imaging, Omics Technologies, Bioinformatics, Computational Biology, Emerging Technologies, Gene Editing, Profiling, Treatment Considerations, Combination Therapy, Drug Development, Biological Processes, Pathogenesis, Genomics, Hematopoiesis, Molecular Biology, Diseases, Genetic Disorders, Myeloid Malignancies, Acute Myeloid Malignancies, Research, Fundamental Science, AML, Translational Research, Study Population, Human, Animal Model
PAPER

359 Functional genomics and targeted protein degradation unravel the transcriptional circuitry of CBFA2T3::GLIS2 acute megakaryoblastic leukemia

Cory D Rillahan, MD, PhD, Fanny Gonzales, MD, PhD2, Nicholas Aboreden, PhD3, Rachel Brooke Mersfelder, MD4, Giulia DiGiovanni5, Fabio Boniolo, PhD6, Nicholas K O'Neill, PhD7, Vivian Tong8, Allen Basanthakumar9, Delan Khalid10, Maxim Pimkin, MD, PhD11, Julie Riviere, PhD12, Alexia Le Bourgeois13, Thomas Mercher, PhD14, Volker Hovestadt, PhD15, Kimberly Stegmaier, MD16
1 Boston Children's Hospital, Boston, Massachusetts, United States, Broad Institute of MIT & Harvard, Cambridge, Massachusetts, United States, Dana Farber Cancer Institute, Boston, Massachusetts, United States
2 Fondation Cansearch, Genève, Switzerland
3 Boston Children's Hospital, Boston, Massachusetts, United States, Broad Institute of MIT & Harvard, Cambridge, Massachusetts, United States, Dana Farber Cancer Institute, Boston, Massachusetts, United States
4 Boston Children's Hospital, Boston, Massachusetts, United States, Broad Institute of MIT & Harvard, Cambridge, Massachusetts, United States, Dana Farber Cancer Institute, Boston, Massachusetts, United States
5 Boston Children's Hospital, Boston, Massachusetts, United States, Broad Institute of MIT & Harvard, Cambridge, Massachusetts, United States, Dana Farber Cancer Institute, Boston, Massachusetts, United States
6 Boston Children's Hospital, Boston, Massachusetts, United States, Broad Institute of MIT & Harvard, Cambridge, Massachusetts, United States, Dana Farber Cancer Institute, Boston, Massachusetts, United States
7 Boston Children's Hospital, Boston, Massachusetts, United States, Broad Institute of MIT & Harvard, Cambridge, Massachusetts, United States, Dana Farber Cancer Institute, Boston, Massachusetts, United States
8 Boston Children's Hospital, Boston, Massachusetts, United States, Broad Institute of MIT & Harvard, Cambridge, Massachusetts, United States, Dana Farber Cancer Institute, Boston, Massachusetts, United States
9 Boston Children's Hospital, Boston, Massachusetts, United States, Broad Institute of MIT & Harvard, Cambridge, Massachusetts, United States, Dana Farber Cancer Institute, Boston, Massachusetts, United States
10 Boston Children's Hospital, Boston, Massachusetts, United States, Broad Institute of MIT & Harvard, Cambridge, Massachusetts, United States, Dana Farber Cancer Institute, Boston, Massachusetts, United States
11 Boston Children's Hospital, Boston, Massachusetts, United States, Broad Institute of MIT & Harvard, Cambridge, Massachusetts, United States, Dana Farber Cancer Institute, Boston, Massachusetts, United States
12 Université Paris-Saclay, Villejuif, France
13 Université Paris-Saclay, Villejuif, France
14 Université Paris-Saclay, Villejuif, France
15 Boston Children's Hospital, Boston, Massachusetts, United States, Broad Institute of MIT & Harvard, Cambridge, Massachusetts, United States, Dana Farber Cancer Institute, Boston, Massachusetts, United States
16 Boston Children's Hospital, Boston, Massachusetts, United States, Broad Institute of MIT & Harvard, Cambridge, Massachusetts, United States, Dana Farber Cancer Institute, Boston, Massachusetts, United States

Acute megakaryoblastic leukemia (AMKL) is a rare form of AML that largely affects infants and toddlers. Among the most aggressive subtypes is that driven by the oncogenic fusion transcription factor CBFA2T3::GLIS2, with a 5-year overall survival of roughly 20%. Despite its discovery over 10 years ago, our mechanistic understanding of the fusion oncoprotein, which may hold the key to new rational therapies, is incomplete. Here we use functional genomics and targeted protein degradation to dissect how CBFA2T3::GLIS2, as the sole oncogenic driver, can maintain leukemogenesis.

 

By creating Cas9-expressing, fusion-positive AMKL cell lines and introducing gRNAs to genetically ablate the fusion transcription factor, we first confirmed the requirement of the fusion for disease maintenance in vitro and in vivo. Interestingly, loss of the fusion led to differentiation towards megakaryocytic/erythroid or myeloid lineages, revealing a capacity for blasts to differentiate into two distinct fates. RNA-sequencing of these cell lines mirrored these findings with pathways involved in megakaryocytic, erythroid, and myeloid differentiation being highly enriched upon fusion knockout. Importantly, the transcriptional profiles of our model systems were highly correlated with primary patient samples, thus faithfully recapitulating human disease.

 

To further define these differentiation trajectories, we coupled single cell RNA- and ATAC-sequencing to map cell states at various timepoints after fusion loss. Key myeloid transcription factors, CEBPA and SPI1, became highly upregulated in cells undergoing monocytic differentiation and this correlated with increased chromatin accessibility of their downstream targets. On the contrary, GATA1 and NFE2 accessible sites were more highly enriched in megakaryocytic differentiation, despite an upregulation of their expression in both lineages.

 

While powerful, an inherent limitation of CRISPR-Cas9 is the lack of kinetic control over protein target loss; protein disruption takes on the order of days, making it difficult to uncouple direct vs. indirect effects. To overcome this problem, AMKL cell lines with an FKBP12F36V protein tag engineered onto the C-terminus of the endogenous CBFA2T3::GLIS2 locus were created, allowing for rapid proteolytic degradation of the fusion on the timescale of minutes after treatment with the small molecule PROTAC dTAG-V1. This system recapitulated our CRISPR-based models including transcriptional programs that correlate well with primary patient samples.

 

With the ability to employ an acute degradation strategy, we carried out nascent transcriptomics with Pro-seq to identify the most direct transcriptional targets of CBFA2T3::GLIS2. Strikingly, this revealed that the fusion is predominantly repressive with 10-fold more genes becoming up- vs downregulated after fusion degradation. Further, we found that key master transcription factors involved in megakaryocytic (GATA1 and NFE2) and myelomonocytic (CEBPA and SPI1) differentiation were among the genes that became most highly expressed after fusion loss. We validated that this upregulation is responsible for the bipotent differentiation trajectories observed. On the contrary, among the few genes which appear to be activated by the fusion were the key transcription factors ERG and GATA3.

 

To understand how the fusion co-opts the intrinsic epigenetic and transcriptional circuitry to sustain leukemogenesis, we assessed the chromatin state after acute degradation with ATAC- and ChIP-seq. We found that CBFA2T3::GLIS2 binds to areas of open chromatin that are sustained in its absence and that the fusion was colocalized with ERG, GATA2, and RUNX1. Areas with higher density of the fusion were correlated with gene-repression and, specifically, high fusion occupancy appeared to block intrinsic GATA2-driven megakaryocytic and RUNX1-driven myeloid differentiation pathways. On the contrary, low levels of the fusion and high levels of ERG correlated with gene activation of fusion targets. We thus propose a model wherein the fusion is recruited to GATA2 and RUNX1 target sites to block differentiation, while its colocalization with ERG mediates self-renewal, thus single-handedly maintaining leukemogenesis.

Cory D Rillahan, MD, PhD: No relevant conflicts of interest to declare.