Supplementary Figures S1-S8 with legends.

Phosphoinositide 3-kinase gamma (PI3Kγ) is implicated as a target to repolarize tumor-associated macrophages and promote anti-tumor immune responses in solid cancers. However, cancer cell-intrinsic roles of PI3Kγ are unclear. Here, by integrating unbiased genome-wide CRISPR interference screening with functional analyses across acute leukemias, we define a selective dependency on the PI3Kγ complex in a high-risk subset that includes myeloid, lymphoid, and dendritic lineages. This dependency is characterized by innate inflammatory signaling and activation of phosphoinositide 3-kinase regulatory subunit 5 ( PIK3R5 ), which encodes a regulatory subunit of PI3Kγ and stabilizes the active enzymatic complex. Mechanistically, we identify p21 (RAC1) activated kinase 1 (PAK1) as a noncanonical substrate of PI3Kγ that mediates this cell-intrinsic dependency independently of Akt kinase. PI3Kγ inhibition dephosphorylates PAK1, activates a transcriptional network of NFκB-related tumor suppressor genes, and impairs mitochondrial oxidative phosphorylation. We find that treatment with the selective PI3Kγ inhibitor eganelisib is effective in leukemias with activated PIK3R5 , either at baseline or by exogenous inflammatory stimulation. Notably, the combination of eganelisib and cytarabine prolongs survival over either agent alone, even in patient-derived leukemia xenografts with low baseline PIK3R5 expression, as residual leukemia cells after cytarabine treatment have elevated G protein-coupled purinergic receptor activity and PAK1 phosphorylation. Taken together, our study reveals a targetable dependency on PI3Kγ/PAK1 signaling that is amenable to near-term evaluation in patients with acute leukemia.

MYB fusions are recurrently found in select cancers, including blastic plasmacytoid DC neoplasm (BPDCN), an acute leukemia with poor prognosis. They are markedly enriched in BPDCN compared with other blood cancers and, in some patients, are the only obvious somatic mutation detected. This suggests that they may alone be sufficient to drive DC transformation. MYB fusions are hypothesized to alter the normal transcription factor activity of MYB, but, mechanistically, how they promote leukemogenesis is poorly understood. Using CUT&RUN chromatin profiling, we found that, in BPDCN leukemogenesis, MYB switches from being a regulator of DC lineage genes to aberrantly regulating G2/M cell cycle control genes. MYB fusions found in patients with BPDCN increased the magnitude of DNA binding at these locations, and this was linked to BPDCN-associated gene expression changes. Furthermore, expression of MYB fusions in vivo impaired DC differentiation and induced transformation to generate a mouse model of myeloid-dendritic acute leukemia. Therapeutically, we present evidence that all-trans retinoic acid (ATRA) may cause loss of MYB protein and cell death in BPDCN.

Tumours most often arise from progression of precursor clones within a single anatomical niche. In the bone marrow, clonal progenitors can undergo malignant transformation to acute leukaemia, or differentiate into immune cells that contribute to disease pathology in peripheral tissues(1-4). Outside the marrow, these clones are potentially exposed to a variety of tissue-specific mutational processes, although the consequences of this are unclear. Here we investigate the development of blastic plasmacytoid dendritic cell neoplasm (BPDCN)-an unusual form of acute leukaemia that often presents with malignant cells isolated to the skin(5). Using tumour phylogenomics and single-cell transcriptomics with genotyping, we find that BPDCN arises from clonal (premalignant) haematopoietic precursors in the bone marrow. We observe that BPDCN skin tumours first develop at sun-exposed anatomical sites and are distinguished by clonally expanded mutations induced by ultraviolet (UV) radiation. A reconstruction of tumour phylogenies reveals that UV damage can precede the acquisition of alterations associated with malignant transformation, implicating sun exposure of plasmacytoid dendritic cells or committed precursors during BPDCN pathogenesis. Functionally, we find that loss-of-function mutations in Tet2, the most common premalignant alteration in BPDCN, confer resistance to UV-induced cell death in plasmacytoid, but not conventional, dendritic cells, suggesting a context-dependent tumour-suppressive role for TET2. These findings demonstrate how tissue-specific environmental exposures at distant anatomical sites can shape the evolution of premalignant clones to disseminated cancer.

Dendritic cells play an important role in anticancer immunity by exposing T cells to tumor-associated antigens. In a recent study, Zhao et al. show that BCL2 inhibition improves the ability of dendritic cells to present antigen to T cells and activate their antitumor cytotoxicity.

Blastic plasmacytoid dendritic cell neoplasm (BPDCN) is a rare acute leukemia with poor prognosis arising from the plasmacytoid dendritic cell (pDC) lineage. MYB is a hematopoietic transcription factor overexpressed in many leukemias. Genomic rearrangements of MYB were identified in a significant fraction of BPDCNs, including in cases without any other clear driver mutations, and result in its fusion to one of several partner genes (Suzuki et al., Leukemia 2017). These fusions are rare or absent in other acute leukemias and the mechanisms by which they contribute to BPDCN oncogenesis are not understood. To examine the impact of Myb fusions on hematopoietic differentiation and leukemogenesis, we used mouse bone marrow progenitors immortalized by estradiol (E2)-dependent activation of HoxB8 and culture with Flt3 ligand (HoxB8-FL cells). We used HoxB8-FL cells deficient in Cdkn2a, since CDKN2A deletion is present in up to ~67% of BPDCNs (Lucioni et al., Blood 2011) but is rare in other myeloid leukemias. HoxB8-FL cells differentiate into myeloid, lymphoid and dendritic cells in vitro upon withdrawal of E2, or in vivo upon transplantation. They therefore represent lymphoid-primed multipotent progenitor (LMPP)-like cells that are unable to self-renew in the absence of E2. We co-expressed a dTom reporter with V5-tagged Myb constructs in HoxB8-FL cells: wild-type full-length Myb (Myb-FL), truncated Myb (Myb-TR) or Myb-PLEKHO1. MYB-PLEKHO1 is the most frequent fusion in BPDCN patients, while truncated Myb is analogous to recurrent BPDCN MYB fusions in which the partner is out-of-frame. Empty vector was used as a control. HoxB8-FL cells expressing Myb-FL, Myb-TR or Myb-PLEKHO1 displayed arrested differentiation upon withdrawal of E2 in vitro (Figure 1). While all empty vector-transduced cells had upregulated CD11b and/or CD11c at day 7 post-E2 withdrawal, Myb overexpression caused an accumulation of CD11b-CD11c- undifferentiated cells. In vivo, a similar effect was observed at day 7 post-transplantation: dTom+ CD11b-CD11c-CD19-B220- undifferentiated cells were still present in the bone marrow of Myb-FL, Myb-TR and Myb-PLEKHO1 recipient mice, while in empty vector recipients all dTom+ cells expressed at least one of these differentiation markers. Surprisingly, we observed long-term persistence of dTom+ cells in the peripheral blood of Myb-TR and Myb-PLEKHO1, but not Myb-FL, recipients. This developed into lethal malignancy in 100% of Myb-TR and Myb-PLEKHO1 recipients by 28 weeks (Figure 2). Malignant cells showed blast-like morphology of acute leukemia and expressed markers of a myeloid/dendritic progenitor (Kit+CD11b+Cx3cr1+). dTom+ disease was never observed in Myb-FL or empty vector recipients (log-rank P<0.001). Together, while expression of either full-length Myb or Myb fusions impaired myeloid/dendritic differentiation, only Myb rearrangements associated with BPDCN generated a fully penetrant leukemia. To investigate chromatin-level differences between physiologically expressed wild-type MYB and MYB fusions, we performed Cut&Run in K562 cells with V5-tagged MYB-FL, MYB-TR or MYB-PLEKHO1 knocked in to the endogenous MYB locus using CRISPR/Cas9 and homology-directed repair. We observed increased binding of MYB-TR and MYB-PLEKHO1 relative to MYB-FL at promoters of cell cycle genes that function at the G2/M checkpoint, including CDC20, CDCA3 and CKS2. These binding sites contained canonical MYB binding motifs, indicating that MYB fusions increase direct DNA binding at chromatin involved in control of specific cell cycle genes. Cut&Run for V5-Myb-PLEKHO1 in mouse leukemias revealed similar direct DNA binding at cell cycle gene promoters. Moreover, expression of G2/M cell cycle genes was enriched in patient BPDCNs relative to normal pDCs. Our data suggest that BPDCN MYB fusions exert distinct oncogenic effects to overexpression of wild-type MYB. In addition to arresting differentiation, BPDCN MYB fusions can reactivate self-renewal in progenitor cells, which may occur through increased direct DNA binding at cell cycle gene promoters. This may explain the simple genetics of MYB-rearranged BPDCN, which lack other recurrent myeloid-type mutations, if the fusions are sufficient to generate leukemia via these dual roles. These data also indicate that therapeutic strategies targeting MYB and/or specific cell cycle checkpoints may be active in BPDCN. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Background Expansion of blood cells derived from a single mutated blood stem cell - clonal hematopoiesis - often precedes leukemic transformation. However, the circumstances that promote or prevent transformation are poorly understood. Here, we investigate the evolution of clonal hematopoiesis into blastic plasmacytoid dendritic cell neoplasm (BPDCN), an aggressive blood cancer that often presents with malignant cells isolated to the skin (referred to as skin-only disease). We gain a better understanding of the influence of site-specific selective pressures and the role of TET2 mutations, which may inform interventions to prevent disease progression. Methods We analyzed bone marrow and skin from 13 BPDCN patients using targeted, exome, and whole-genome sequencing, with germline controls. We performed multimodal single-cell sequencing (10x 3' v3.1 single-cell RNA-seq with genotyping of expressed variants) on bone marrow from 6 healthy donors (20,411 cells), 5 skin-only BPDCN patients with malignant cells in the skin but not conventionally detectable in marrow (36,018 cells), and 6 BPDCN patients with overt bone marrow involvement (30,582 cells). For functional studies, we targeted Tet2 using CRISPR/Cas9 in mouse blood progenitor cells immortalized by an estrogen receptor hormone binding domain fused to HoxB8, which undergo synchronized differentiation to myeloid and dendritic cells upon estrogen withdrawal. We used CellRanger, Seurat, RandomForest machine learning, and the R tidyverse package collection for bioinformatics. Results Genetics allowed us to reconstruct the evolution of BPDCN across bone marrow and skin (representative patient, panel a). Invariably, we detected founder mutations in marrow, including TET2 inactivation in 9/13 patients (two mutations, likely bi-allelic, in 5/9). Single-cell sequencing of marrow from 5 skin-only patients showed that pre-malignant mutant stem cells contributed to multiple lineages, and that TET2 mutated clones were myeloerythroid-biased. We generated a gene expression signature to identify rare malignant BPDCN cells in marrow of skin-only patients. In 3/5 patients, we identified rare cells (0.03-0.19%) with the transcriptional profile of fully transformed cells, and concurrent single-cell genetic analysis suggested that these cells originated from the skin tumor (panel b). To assess whether disseminated skin tumor cells establish bone marrow disease, we quantified mutational signatures: skin tumors and malignant BPDCN marrow cells in 5/5 patients harbored UV-induced mutations, likely acquired in a single skin-resident plasmacytoid dendritic cell (pDC) during malignant transformation. Finally, we edited Tet2 in mouse myelo-dendritic progenitors and tested the response to UV irradiation in vitro. In addition to observing that Tet2 inactivation increases dendritic cell commitment, the more striking result was a survival advantage of Tet2-mutated cells in the presence of UV (n = 6, fold change 1.803, P = 0.0012). These results highlight an unexpected function of TET2 that may predispose mutated pDCs to survive UV-induced DNA damage in the skin, leading to disease progression. Conclusion The complex evolution of BPDCN traverses multiple anatomic sites prior to full transformation. This allows for accurate sequencing of oncogenic events. Disease evolution frequently starts with acquisition of a leukemia-associated mutation in a multipotent blood stem cell, leading to clonal expansion. In the most frequently observed scenario, bi-allelic TET2 inactivation promotes myeloerythroid bias, increases dendritic cell commitment, and confers resistance to UV-mediated cell death. Subsequent evolution of skin-resident mutant pDCs leads to a high burden of UV-induced mutations and ultimate transformation. Dissemination of malignant cells, which can be detected by integrating single-cell gene expression and mutation profiles, initiates "retrograde" bone marrow disease. Our discovery that multiple anatomic sites - and local selective pressures - play a role in disease progression provides a template to study multi-site evolution of clonal hematopoiesis. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

AbstractBlastic plasmacytoid dendritic cell neoplasm (BPDCN) is an aggressive leukemia of plasmacytoid dendritic cells (pDC). BPDCN occurs at least three times more frequently in men than in women, but the reasons for this sex bias are unknown. Here, studying genomics of primary BPDCN and modeling disease-associated mutations, we link acquired alterations in RNA splicing to abnormal pDC development and inflammatory response through Toll-like receptors. Loss-of-function mutations in ZRSR2, an X chromosome gene encoding a splicing factor, are enriched in BPDCN, and nearly all mutations occur in males. ZRSR2 mutation impairs pDC activation and apoptosis after inflammatory stimuli, associated with intron retention and inability to upregulate the transcription factor IRF7. In vivo, BPDCN-associated mutations promote pDC expansion and signatures of decreased activation. These data support a model in which male-biased mutations in hematopoietic progenitors alter pDC function and confer protection from apoptosis, which may impair immunity and predispose to leukemic transformation.Significance:Sex bias in cancer is well recognized, but the underlying mechanisms are incompletely defined. We connect X chromosome mutations in ZRSR2 to an extremely male-predominant leukemia. Aberrant RNA splicing induced by ZRSR2 mutation impairs dendritic cell inflammatory signaling, interferon production, and apoptosis, revealing a sex- and lineage-related tumor suppressor pathway.This article is highlighted in the In This Issue feature, p. 275

Yolk sac (YS) hematopoiesis is critical for the survival of the embryo and a major source of tissue-resident macrophages that persist into adulthood. Yet, the transcriptional and epigenetic regulation of YS hematopoiesis remains poorly characterized. Here we report that the epigenetic regulator Ezh2 is essential for YS hematopoiesis but dispensable for subsequent aorta–gonad–mesonephros (AGM) blood development. Loss of EZH2 activity in hemogenic endothelium (HE) leads to the generation of phenotypically intact but functionally deficient erythro-myeloid progenitors (EMPs), while the generation of primitive erythroid cells is not affected. EZH2 activity is critical for the generation of functional EMPs at the onset of the endothelial-to-hematopoietic transition but subsequently dispensable. We identify a lack of Wnt signaling downregulation as the primary reason for the production of non-functional EMPs. Together, our findings demonstrate a critical and stage-specific role of Ezh2 in modulating Wnt signaling during the generation of EMPs from YS HE.

Juvenile Myelomonocytic Leukemia (JMML) is a poor prognosis childhood leukemia usually caused by germline or somatic RAS-activating mutations. The cellular hierarchy in JMML is poorly characterized, including the identity of leukemia stem cells (LSCs). FACS and single-cell RNA-sequencing reveal marked heterogeneity of JMML hematopoietic stem/progenitor cells (HSPCs), including an aberrant Lin-CD34+CD38-CD90+CD45RA+ population. Single-cell HSPC index-sorting and clonogenic assays show that (1) all somatic mutations can be backtracked to the phenotypic HSC compartment with RAS -activating mutations as a “first hit”, (2) mutations are acquired with both linear and branching patterns of clonal evolution and (3) mutant HSPCs are present after allogeneic HSC transplant before molecular/clinical evidence of relapse. Stem cell assays reveal inter-patient heterogeneity of JMML-LSCs which are present in, but not confined to, the phenotypic HSC compartment. RNA-sequencing of JMML-LSCs reveals upregulation of stem cell and fetal genes ( HLF, MEIS1, CNN3 , VNN2 , HMGA2 ) and candidate therapeutic targets/biomarkers ( MTOR, SLC2A1 , CD96 ) paving the way for LSC-directed disease monitoring and therapy in this disease.

Somatic mutations in acute myeloid leukemia are acquired sequentially and hierarchically. First, pre-leukemic mutations, such as t(8;21) that encodes AML1-ETO, are acquired within the hematopoietic stem cell (HSC) compartment, while signaling pathway mutations, including KRAS activating mutations, are late events acquired during transformation of leukemic progenitor cells and are rarely detectable in HSC. This raises the possibility that signaling pathway mutations are detrimental to clonal expansion of pre-leukemic HSC. To address this hypothesis, we used conditional genetics to introduce Aml1-ETO and K-RasG12D into murine HSC, either individually or in combination. In the absence of activated Ras, Aml1-ETO-expressing HSC conferred a competitive advantage. However, activated K-Ras had a marked detrimental effect on Aml1-ETO-expressing HSC, leading to loss of both phenotypic and functional HSC. Cell cycle analysis revealed a loss of quiescence in HSC co-expressing Aml1-ETO and K-RasG12D, accompanied by an enrichment in E2F and Myc target gene expression and depletion of HSC self-renewal-associated gene expression. These findings provide a mechanistic basis for the observed absence of KRAS signaling mutations in the pre-malignant HSC compartment.

Lympho-myeloid restricted early thymic progenitors (ETPs) are postulated to be the cell of origin for ETP leukemias, a therapy-resistant leukemia associated with frequent co-occurrence of EZH2 and RUNX1 inactivating mutations, and constitutively activating signaling pathway mutations. In a mouse model, we demonstrate that Ezh2 and Runx1 inactivation targeted to early lymphoid progenitors causes a marked expansion of pre-leukemic ETPs, showing transcriptional signatures characteristic of ETP leukemia. Addition of a RAS-signaling pathway mutation (Flt3-ITD) results in an aggressive leukemia co-expressing myeloid and lymphoid genes, which can be established and propagated in vivo by the expanded ETPs. Both mouse and human ETP leukemias show sensitivity to BET inhibition in vitro and in vivo, which reverses aberrant gene expression induced by Ezh2 inactivation.

Recent advances in single-cell transcriptomics are ideally placed to unravel intratumoral heterogeneity and selective resistance of cancer stem cell (SC) subpopulations to molecularly targeted cancer therapies. However, current single-cell RNA-sequencing approaches lack the sensitivity required to reliably detect somatic mutations. We developed a method that combines high-sensitivity mutation detection with whole-transcriptome analysis of the same single cell. We applied this technique to analyze more than 2,000 SCs from patients with chronic myeloid leukemia (CML) throughout the disease course, revealing heterogeneity of CML-SCs, including the identification of a subgroup of CML-SCs with a distinct molecular signature that selectively persisted during prolonged therapy. Analysis of nonleukemic SCs from patients with CML also provided new insights into cell-extrinsic disruption of hematopoiesis in CML associated with clinical outcome. Furthermore, we used this single-cell approach to identify a blast-crisis-specific SC population, which was also present in a subclone of CML-SCs during the chronic phase in a patient who subsequently developed blast crisis. This approach, which might be broadly applied to any malignancy, illustrates how single-cell analysis can identify subpopulations of therapy-resistant SCs that are not apparent through cell-population analysis.

Although previous studies suggested that the expression of FMS-like tyrosine kinase 3 (Flt3) initiates downstream of mouse hematopoietic stem cells (HSCs), FLT3 internal tandem duplications (FLT3 ITDs) have recently been suggested to intrinsically suppress HSCs. Herein, single-cell interrogation found Flt3 mRNA expression to be absent in the large majority of phenotypic HSCs, with a strong negative correlation between Flt3 and HSC-associated gene expression. Flt3-ITD knock-in mice showed reduced numbers of phenotypic HSCs, with an even more severe loss of long-term repopulating HSCs, likely reflecting the presence of non-HSCs within the phenotypic HSC compartment. Competitive transplantation experiments established that Flt3-ITD compromises HSCs through an extrinsically mediated mechanism of disrupting HSC-supporting bone marrow stromal cells, with reduced numbers of endothelial and mesenchymal stromal cells showing increased inflammation-associated gene expression. Tumor necrosis factor (TNF), a cell-extrinsic potent negative regulator of HSCs, was overexpressed in bone marrow niche cells from FLT3-ITD mice, and anti-TNF treatment partially rescued the HSC phenotype. These findings, which establish that Flt3-ITD–driven myeloproliferation results in cell-extrinsic suppression of the normal HSC reservoir, are of relevance for several aspects of acute myeloid leukemia biology.