Aaron D. Viny, M.D.
Cohesin complex mutations decouple chromatin structure and transcriptional output in myelodysplastic syndrome

Aaron D. Viny, M.D.

Memorial Sloan Kettering Cancer Center

2018 Funding recipient

Cohesin complex mutations decouple chromatin structure and transcriptional output in myelodysplastic syndrome

EvansMDS Young Investigator Award


Mutations in all cohesin complex members, including Smc3 and Rad21, are observed in myelodysplastic syndrome (MDS), particularly in patients at high risk for transformation to acute myeloid leukemia (AML). MDS is inherently a bone marrow failure disorder where bone marrow stem cells are unable to effectively undergo coordinated differentiation and maturation into the red blood cells that carry oxygen, the platelets that clot the blood, and the white blood cells that fight infections. We previously demonstrated that deletion of the cohesin ring component, Smc3 increases hematopoietic stem/progenitor cell (HSPC) self-renewal but with defective differentiation and maturation due to structural changes in the intracellular DNA. However, the unique role of Stag2 mutations in MDS and in other oncogenic contexts remained elusive. Stag2 is the most commonly mutated cohesin component in MDS and is the only cohesin mutant recurrently in solid tumors. We therefore generated conditional knockout mouse models for Stag2 and its paralog Stag1. Deletion of Stag2 in hematopoietic cells induced a myelodysplasia phenotype with thrombocytopenia, myeloid skewed leukopenia, dyserythropoiesis, and expansion of HSPC. Moreover, we found that Stag2 deficient stem cells serially replate in vitro and had increased competitive transplantation capacity in vivo, consistent with increased self-renewal; yet, these cells do not effectively respond to the transcription factors that drive hematopoietic maturation. ChIP-seq analysis of primary HSPC revealed that a subset of sites with Stag2 occupancy had compensatory occupancy by Stag1 in the absence of Stag2 whereas other sites which lost Stag2 binding did not, consistent with site-specific differences in Stag2/Stag1 across the genome. Consistent with this, Hi-C revealed decreased short-range interactions with Stag2 loss. Importantly, co-deletion of Stag2 and Stag1 in resulted in lethal bone marrow aplasia and PCC similar to that observed in Smc3 null mice. Our data illustrate a key, novel role for Stag2 loss in hematopoietic transformation and transcriptional dysregulation, distinct from its redundant role with Stag1 in normal hematopoiesis. Moreover, our data suggest that Stag1 is a non-essential gene in hematopoietic cells but represents a mutant-specific dependency in Stag2 mutant cells, credentialing Stag1 as a novel molecular target in Stag2 mutant MDS.


The aims of this project are to 1) Establish the molecular mechanism of cohesin-dependent tumorigenesis 2) Determine the effect of cohesin loss of function on cis-regulatory elements leading to altered transcriptional programming by key cohesin-targets such as PU.1/Runx1, and 3) Exploit cohesin-specific sensitivities to inhibition of STAG1 and HDAC8 with preclinical compounds. These aims will answer three important questions and will shape the focus of my future independent lab:

1) Is the enhancer/cohesin landscape of MDS blasts different than normal cells from the same lineage?

2) Is overexpression of repressed transcription factors sufficient to overcome the transcriptional dysregulation resulting from cohesin loss of function?

3) How are tissue-specific promoter-enhancer interactions controlled by cohesin and how does gene-specific conformation change transcriptional output?