Lineage-Specific RUNX2 Super-Enhancer Activates MYC Via Translocation (6;8) to Promote the Development of Blastic Plasmacytoid Dendritic Cell Neoplasm

Blastic plasmacytoid dendritic cell neoplasm (BPDCN) is a rare and aggressive hematological malignancy, characteristic of skin lesions followed by hematopoietic organ dissemination. The cell of origin of which is considered to be precursors of plasmacytoid dendritic cells (pDCs). BPDCN cells show high frequencies of mutations in TET2 and p53; however, the molecular mechanisms underlying the pathogenesis of BPDCN have not yet been elucidated. RUNX2 transcription factor, located on chromosome 6p21, is critical for the differentiation of pDCs and the enhancer of RUNX2 is activated in pDCs. Since translocation (6;8)(p21;q24), which is a rare, but specific anomaly for BPDCN, involves regions adjacent to RUNX2 and MYC, we demonstrate that the pDC-specific RUNX2 super-enhancer activates the expression of RUNX2, which functions as a lineage-survival transcription factor, but also is hijacked to activate expression of MYC via t(6;8) in BPDCN cells, and that RUNX2 and MYC promote the initiation and propagation of BPDCN by generating a novel mouse model.

In order to identify the breakpoint of t(6;8)(p21;q24), we first performed fluorescent in situ hybridization and whole genome sequencing of CAL-1 cells, a BPDCN cell line, and identified a fusion point of chromosome translocation between chromosome 8 in 69 kilobases (kb) downstream of MYC and chromosome 6, which was 58 kb centromeric to a long and clustered super-enhancer of RUNX2 (791 kb upstream of RUNX2) defined by chromatin immunoprecipitation sequencing using anti-H3K27ac or anti-BRD4 antibodies. As we observed the enhanced levels of MYC and RUNX2 expression in BPDCN cells in patients and CAL-1 cells, we knocked down expression of MYC or RUNX2 using distinct shRNA vectors in CAL-1 cells. We found that the knockdown of MYC and/or RUNX2 significantly impaired colony formation capacities. By performing microarray analysis, we found that RUNX2 knockdown significantly reduced expression of pDCs-signature genes in CAL-1 cells, accompanied with the enhanced apoptosis in CAL-1 cells, implying that RUNX2 is critical for the survival of BPDCN cells due to expressing pDCs-signature genes. To evaluate the function of RUNX2 super-enhancer, we examined how BRD4 inhibition affected the proliferative capacities of CAL-1 cells in vitro. Indeed, we found that JQ1-treated CAL-1 cells showed significantly lower H3K27ac modification levels at the RUNX2 super-enhancer and significantly decreased levels of MYC and RUNX2 expression, resulting in the impaired colony formation capacities, which were rescued by the ectopic expression of both RUNX2 and MYC. We also genetically deleted the mutant-allele super-enhancer of RUNX2 on der(8) (seRUNX2der8), but not that on chromosome 6, using CRISPR-Cas9 vectors. After establishing single cell clones, all seRUNX2der8-deleted clones showed markedly impaired colony formation capacities accompanied with the reduced expression of MYC. Taken together, the seRUNX2der8 directly activates the expression of MYC to promote the development of BPDCN, which is reversed by the inhibition of BRD4.

We finally examined whether the transduction of MYC and RUNX2 was sufficient for the initiation of BPDCN in vivo in the absence of Tet2 and p53. We purified Lineage^-Sca-1⁺c-Kit⁺ stem/progenitor cells from wild-type and Tet2/p53 double knockout (DKO) mice and infected them with MYC- and RUNX2-retrovirus vectors. After a 9-day culture promoting the differentiation of pDCs, we transplanted transduced cells into recipient mice together with wild-type competitor cells. MYC+RUNX2-DKO mice showed robust leukocytosis, anemia, and thrombocytopenia and died by two months post-transplantation following the expansion of immature leukemic blasts. A FACS analysis showed that these leukemic blasts were CD11b^-CD11c^mid/+B220⁺Bst2⁺, which was consistent with the murine pDCs immunophenotype, and massively infiltrated the spleen and liver tissues. MYC+RUNX2-DKO leukemic cells were transplantable in secondary recipient mice with the same immunophenotype. Thus, the transduction of MYC and RUNX2 is sufficient to initiate the transformation of lethal BPDCN-like disease in mice lacking Tet2 and p53. We are now exploring the molecular mechanism of how MYC and RUNX2 collaborate to initiate the formation of BPDCN by performing RNA-sequencing analysis and transplantation assay of BPDCN-initiating cells.