Leukemia-Associated Cohesin Mutants Dominantly Enforce Stem Cell Programs and Impair Human Hematopoietic Progenitor Differentiation

Recurrent mutations in the components of the cohesin complex (RAD21, SMC1A, SMC3, and STAG2) have been identified in human AML and other myeloid malignancies, and have been shown to occur as pre-leukemic mutations in HSC. Cohesin functions to hold chromatin strands within a ring-like structure composed of the four core components, and although its best-established role is to maintain the polarity of sister chromatids during mitosis, cohesin is also involved in double-stranded DNA damage repair and regulation of transcription. As little is known about their contributions to leukemogenesis, we sought to investigate the effects of cohesin mutants on human hematopoiesis, particularly hematopoietic stem and progenitor cells (HSPC).

Introduction of mutant cohesin into AML cell lines and primary human HSPC resulted in a differentiation block with an increased frequency of CD34+ progenitor cells. A similar phenotype was observed with knockdown of core component RAD21 both in vitro and in vivo, indicating that mutant cohesin can act either through haploinsufficiency or dominant-negative mechanisms. Mutant cohesin increased the serial replating ability of HSPC in vitro and showed enrichment for HSC and leukemia stem cell gene expression programs, indicating an effect to enforce stem cell functions. Furthermore, we observed a skewing toward the myeloid lineage in cohesin mutant colonies cultured in methylcellulose, which was recapitulated by a strong myeloid skewing of human engrafted cells in vivo. Thus, mutant cohesin enforces stem cell programs and impairs human hematopoietic progenitor differentiation.

Since cohesin complex mutations were identified in pre-leukemic HSC in many of the cases we investigated, we hypothesized that they may impart their phenotype in a cell context-dependent manner. To investigate this hypothesis, six human HSPC subpopulations (HSC, MPP, LMPP, CMP, GMP, and MEP) were isolated from cord blood, and these cells were transduced with cohesin WT, cohesin mutant, RAD21 shRNA, or control lentivirus. Transduced cells were then cultured in either myeloid differentiation or erythroid differentiation-promoting conditions. Strikingly, a strong myeloid differentiation block was only observed with cohesin mutant-transduced HSC and MPP, but not GMP. Similarly, a strong erythroid differentiation block was also observed in HSC and MPP, but not MEP. These results indicate that the effect of mutant cohesin is context dependent and restricted to the most immature HSPC.

We next sought to elucidate the mechanism by which cohesin mutants exert their effects on human HSPC. Since the cohesin complex functions to establish and maintain DNA accessibility, and knockdown of cohesin can led to a decrease in chromatin accessibility at transcription factor (TF) clustered regions (Yan et al., 2013), we hypothesized that cohesin mutants impart their phenotypic effects through modulation of chromatin accessibility. To investigate this hypothesis, we used a newly developed method known as ATAC-Seq (Buenrostro et al., 2013) to assess genome-wide accessibility in cohesin WT and mutant HSPC. As expected, we found that cohesin mutants exhibited globally reduced chromatin accessibility at transcriptional regulatory elements. However, we detected increased chromatin accessibility at motifs for transcription factors known to be highly expressed in and critical regulators of HSPC including ERG, GATA2 and RUNX1. Further footprinting analysis, a proxy for ChIP-Seq experiments, showed a strong enrichment of binding of these factors in the mutant cells compared to WT cells.

Based on these results, we developed a model in which the functional effects of mutant cohesin on human HSPC are mediated by transcription factors exhibiting increased chromatin accessibility such as ERG, GATA2 and RUNX1. From this model, we hypothesized that knockdown of these transcription factors would prevent the enforcement of stem cell programs and increase in CD34-expressing cells observed with cohesin mutants. As predicted, knockdown of ERG, GATA2, or RUNX1, but not GATA1 or PU.1, in the presence of cohesin mutants completely prevented the increase in CD34-expressing cells. These results strongly support our proposed model that mutant cohesin impairs hematopoietic differentiation and enforces stem cell programs through the modulation of ERG, GATA2, and RUNX1 chromatin accessibility, expression, and activity.