Ectopic BCL11B disrupts the activity of early hematopoietic transcription factors to skew lineage commitment Free

Enhancer hijacking events that aberrantly activate the T-lineage transcription factor (TF) gene, BCL11B, define a subtype of lineage ambiguous leukemia with T- and/or myeloid immunophenotype (“BCL11B-a” leukemia). Despite this lineage ambiguity, the gene expression profile of BCL11B-a leukemia resembles CD34+ hematopoietic stem and progenitor cells (HSPCs), suggesting an extra-thymic HSPC cell of origin wherein BCL11B exerts ectopic oncogenic activity. How ectopic BCL11B perturbs hematopoietic development to drive lineage skewing and transformation remains unknown. Here, we investigated mechanisms of ectopic BCL11B gene regulatory activity using viral expression in non-BCL11B expressing HSPCs, putative cells of origin for BCL11B-a leukemia.

We first investigated how developmental stage impacts ectopic BCL11B activity. We examined lineage commitment and T-lineage skewing as indicators of oncogenic BCL11B transcriptional activity using an in vitro single cell lineage-output assay in subsets of human cord blood HSPCs: hematopoietic stem cells (HSC), multi-lymphoid progenitors (MLP), common myeloid progenitors (CMP) and granulocyte-monocyte progenitors (GMP). Ectopic BCL11B preferentially induced T-lineage skewing (cytoplasmic CD3 positivity) in CMPs, to a lesser extent in HSCs and MLPs, and not in GMPs. Moreover, ectopic BCL11B promoted mixed erythroid-myeloid single cell-derived colonies at the expense of erythroid-only colonies in CMPs, compared to control CMPs, thus identifying a developmental window amenable to ectopic BCL11B lineage skewing.

To interrogate how ectopic BCL11B perturbs developmental gene programs, we used CUT&RUN to profile ectopic BCL11B chromatin binding in bulk HSPCs. We first examined whether ectopic BCL11B would bind its canonical T-lineage sites in an ectopic context. Compared to published thymic progenitor data, 54% of ectopic BCL11B sites were unique to HSPCs, indicating that cell-intrinsic factors direct BCL11B occupancy in different cellular contexts. To explore potential factors, we analyzed published binding data for key HSPC TFs (RUNX1, FLI1, LMO2, ERG, TAL1, GATA2, and LYL1) and identified RUNX1 as most strongly correlated with BCL11B binding in HSPCs. We confirmed RUNX1 at 74% of BCL11B sites in HSPCs and 87% of BCL11B sites in primary BCL11B-a leukemia cells, thus nominating RUNX1 as a molecular determinant of ectopic BCL11B binding.

Although ectopic BCL11B was recruited to RUNX1-occupied sites in HSPCs, CUT&RUN showed globally reduced RUNX1 binding following enforced BCL11B expression. To assess whether BCL11B directly interfered with RUNX1 binding at these sites, we employed a dTAG approach to acutely degrade BCL11B and then re-measured RUNX1 occupancy. RUNX1 returned to a subset of sites following acute (6 hr) BCL11B degradation, most (76%) of which were promoter-distal and enriched near genes involved in HSC differentiation (e.g. CD34, LMO2, IL1B), demonstrating that ectopic BCL11B directly interferes with RUNX1 binding in HSPCs as one mechanism to disrupt gene regulation.

To assess how ectopic BCL11B perturbs chromatin regulation more globally, we performed ATAC-seq in HSPCs 4 days post BCL11B transduction. Nearly one-quarter of open chromatin sites changed accessibility: most (77%) lost accessibility and were enriched for PU.1 binding, while regions that gained accessibility were enriched for GATA2 binding. BCL11B was previously reported to associate with SWI/SNF chromatin remodeling complexes, and we identified the core SWI/SNF member BRG1 as a chromatin-associated BCL11B interacting partner in BCL11B-a PDX cells, suggesting that ectopic BCL11B regulates chromatin accessibility by directly modulating SWI/SNF targeting. To test this, we performed ATAC-seq following acute BCL11B degradation and observed increased recruitment of BRG1 to newly accessible sites, confirming BCL11B's ability to selectively direct SWI/SNF to chromatin and regulate access of opposing TFs.

Collectively, these data demonstrate that ectopic BCL11B can disrupt the activity of at least 3 TFs in HSPCs, including blocking RUNX1 chromatin binding and potentiating a GATA2 regulated chromatin network at the expense of PU.1. These perturbations differ from BCL11B's normal gene-regulatory activity in thymocytes and are consistent with arrested hematopoietic differentiation and lineage skewing observed in BCL11B-a leukemia. Thus, targeting the ectopic BCL11B-chromatin remodeling axis could be of therapeutic benefit.