Profiling of H3K27ac in ETO2-GLIS2 positive AMKL cells and ETO2-GLIS2 binding

In the recent years, massively parallel sequencing approaches identified hundreds of mutated genes in cancer providing an unprecedented amount of information about mechanisms of cancer cell maintenance and progression. However, while (it is widely accepted that) transformation processes result from oncogenic cooperation between deregulated genes and pathways, the functional characterization of candidate key players is mostly performed at the single gene level which is generally inadequate to identify these oncogene circuitries. In addition, studies aimed at depicting oncogenic cooperation involve the generation of challenging mouse models or the deployment of tedious screening pipelines. Genome wide mapping of epigenomic modifications on histone tails or binding of factors such as MED1 and BRD4 allowed identification of clusters of regulatory elements, also termed Super-Enhancers (SE). Functional annotation of these regions revealed their high relevance during normal tissue development and cancer ontogeny. An interesting paradigm of the tumorigenic function of these SE regions comes from ETO2-GLIS2-driven acute megakaryoblastic leukemia (AMKL) in which the fusion protein ETO2-GLIS2 is sufficient to promote an aberrant transcriptional network by the rewiring of SE regions4. We thus hypothesized that important regulatory regions could control simultaneously expression of genes cooperating in functional modules to promote cancer development. In an effort to identify such modules, we deployed a genome-wide CRISPRi-based screening approach and nominated SE regions that are functionally linked to leukemia maintenance. In particular, we pinpointed a novel SE region regulating the expression of both tyrosine kinases KIT and PDGFRA. Whereas the inhibition of each kinase alone affected modestly cancer cell growth, combined inhibition of both receptors synergizes to impair leukemia cell growth and survival. Our results demonstrate that genome-wide screening of regulatory DNA elements can identify co-regulated genes collaborating to promote cancer and could open new avenues to the concept of combined gene inhibition upon single hit targeting. ChIPseq for H3K27ac in cord blood CD34 cells, patient derived AMKL cells and inducible HEL_5j20 cell line

近年来，大规模平行测序（massively parallel sequencing）技术已在癌症中鉴定出数百个突变基因，为解析癌细胞维持与进展的分子机制提供了前所未有的海量信息。然而，尽管学界普遍认为肿瘤转化过程源于失调基因与信号通路间的致癌协同作用，但目前对候选关键靶点的功能表征大多仍基于单基因水平，这通常不足以识别此类致癌基因调控环路。此外，旨在阐明致癌协同作用的相关研究往往需要构建难度较高的小鼠模型，或是部署繁琐的筛选流程。全基因组范围内对组蛋白尾端的表观遗传修饰或MED1、BRD4等因子的结合位点进行图谱绘制，可鉴定出调控元件簇，即超级增强子（Super-Enhancers, SE）。对这些区域的功能注释表明，它们在正常组织发育与癌症发生发展过程中均具有高度相关性。此类超级增强子区域的致瘤功能有一个极具代表性的范例：ETO2-GLIS2驱动的急性巨核细胞白血病（acute megakaryoblastic leukemia, AMKL），其中融合蛋白ETO2-GLIS2可通过重编程超级增强子区域⁴，诱导异常转录网络的形成。据此我们提出假说：关键调控区域可同时调控功能模块中协同作用的基因表达，进而促进癌症发生发展。为鉴定此类功能模块，我们采用了基于CRISPR干扰（CRISPR interference, CRISPRi）的全基因组筛选策略，筛选出与白血病维持功能相关的超级增强子区域。尤为重要的是，我们定位到一个全新的超级增强子区域，该区域可同时调控酪氨酸激酶KIT与PDGFRA的表达。尽管单独抑制任一激酶仅会对癌细胞增殖产生轻度影响，但联合抑制这两种受体可产生协同效应，显著削弱白血病细胞的增殖与存活能力。本研究结果表明，对调控性DNA元件开展全基因组筛选，可鉴定出协同促进癌症发生的共调控基因，也为“单靶点联合基因抑制”的治疗理念提供了全新的研究方向。本数据集包含脐带血CD34⁺细胞、患者来源的AMKL细胞以及诱导型HEL_5j20细胞系中H3K27ac的染色质免疫沉淀测序（Chromatin Immunoprecipitation sequencing, ChIP-seq）数据。