SMARCD1 negatively regulates myeloid differentiation of leukemic cells via epigenetic mechanisms

Aberrations in transcription and epigenetic factors lead to neoplastic transformation such as acute myeloid leukemia (AML), which is characterized by accumulation of hyperproliferative blasts originating from leukemic stem cells. The use of therapeutic agents designed to lift the differentiation block and reinforce terminal cellular differentiation and growth arrest is 1 way to manage AML pathophysiology. Therefore, understanding these critical regulatory switches is essential for designing selective and effective drug targets for AML. The ATP-dependent SWItch/sucrose non-fermentable (SWI/SNF) chromatin remodeling complex has been implicated in 20% cancers, including AML.1 The catalytic subunit, SMARCA4, drives leukemogenesis by facilitating constitutive Myc expression via enhancer remodeling.2,3 The loss of auxiliary subunits such as ACTL6A leads to proliferation defects in stem cells and bone marrow failure, whereas defects in SMARCD2 affect neutrophil development.4-6 This prompted us to investigate the differential expression of SWI/SNF complex subunits across different stages of blood cell development. We mined several sequencing datasets for the expression of 21 subunits of SWI/SNF complex in various hematopoietic cells and found several of them to be differentially expressed in a cell type-specific manner (supplemental Figure 1). Interestingly, we found that the SMARCD isoforms showed distinct and opposing cell type-specific enrichment. SMARCD1 and SMARCD2 expression was high in CD34+ hematopoietic stem/progenitors (HSPCs), whereas SMARCD3 was specifically enriched in monocytes (Figure 1A; supplemental Figures 1 and 2). The role of SMARCD1 in myeloid differentiation and leukemia has not been investigated thus far, and hence, we validated the above findings using both ex vivo and in vitro hematopoietic models. SMARCD1 expression was high in cord blood-derived CD34+ HSPCs and was significantly reduced in macrophage colony stimulating factor-differentiated HSPCs (Figure 1B). Similarly, SMARCD1 expression was reduced in vitamin D3-differentiated HL-60 cells (Figure 1C). Interestingly, we observed concomitant increase in SMARCD3 expression in differentiated cord blood and promyelocytic human leukemia-60 (HL60) cells, indicating interplay of SMARCD isoforms in hematopoietic differentiation (Figure 1B-C). Next, we investigated the expression of SMARCD1 in patients with AML. Compared with the CD342 compartment, similar enrichment profiles of SMARCD1, SMARD2, and SMARCD3 were observed in the CD34+ AML HSPCs (Figure 1D). The French-American-British (FAB) system classifies AML based on their maturity, and as SMARCD1 expression is enriched in the HSPCs, we assessed its expression across subtypes. Interestingly, we found higher expression of SMARCD1 in undifferentiated AML (M0, M1, M2 subtypes) than in the more differentiated AML FAB subtypes (M3, M4, M5; Figure 1E). These observations indicate a strong correlation between SMARCD1 expression and undifferentiated cell state (both normal and leukemic). Furthermore, the leukemic cell lines show significant dependency on SMARCD1, highlighting its potential role in leukemic cells (supplemental Figure 2G-I). The preferential expression of SMARCD1 in normal and leukemic stem/progenitor cells is attributed to a high promoter accessibility of SMARCD1 in those cell types (supplemental Figure 2E-F).