A new model for understanding a brain tumor epigenetic driver

摘要 弥漫性内在脑桥神经胶质瘤（DIPG）是一种侵袭性肿瘤的年轻儿童，浸润脑干， 尽管放射治疗可以减缓肿瘤生长，但大多数DIPG患儿在诊断后两年内死亡。 为了开发更有效的治疗方法，有必要了解其分子和细胞机制， 驱动DIPG。大约80%的DIPG肿瘤在组蛋白3（H3）基因中携带点突变， 导致赖氨酸27取代甲硫氨酸（H3 K27 M）。H3 K27 M突变体蛋白的表达是一个新的研究领域。 致癌驱动因子，增加细胞增殖，并导致异常基因表达。染色质是一种动态的 H3赖氨酸27（H3 K27-me 3）的三甲基化与DNA和组蛋白的聚合物相关。 转录抑制H3 K27-me 3由EZH 2催化，并且H3 K27 M突变体组蛋白蛋白抑制 EZH 2活性，导致整个基因组中H3 K27-me 3的广泛丢失。然而，这种H3 K27-me 3的损失 与转录去抑制并不密切相关。例如，在表达H3 K27 M的细胞中， 在H3 K27-me 3突变中，只有约4%的丢失H3 K27-me 3的基因表达增加。H3 K27 M是否代替 通过改变基因组组织的“高阶”方面来影响基因表达，例如 染色质与特定的核“隔间？”核内膜的层是转录的 压制性核隔间板层相关结构域（LADs）是大的、特异的基因组区域 位于核周边，并且LAD内的基因表现出低转录活性。H3 K27-me 3是一种高分子化合物， 在LAD和非LAD基因组区域之间的“边界”处富集，表明这种染色质的作用 在调节这方面的高阶基因组组织的修改。在初步研究中，我们 在携带H3 K27 M突变的DIPG细胞中绘制LAD，发现LAD结构是独特的， 与其他类型的癌细胞和正常细胞相比，鉴于我们对生物化学的理解， H3 K27 M突变蛋白，EZH 2的功能，以及我们对LAD边界结构的了解，我们提出 通过抑制EZH 2并引起基因组范围的变化， 在H3 K27-me 3水平上，H3 K27 M突变蛋白破坏正常的LAD结构，并且这种对H3 K27-me 3的破坏可能是一种潜在的机制。 核区室相关的基因组组织是DIPG异常基因表达的基础。这 这项工作挑战了目前对H3 K27 M致癌作用的理解。而不是建立在当前的基础上 在局部染色质状态变化水平上解释H3 K27 M致癌功能的模型，获得的结果 H3 K27 M测试了一种新范式，它对人类基因组的组织造成了更大规模的变化， 原子核组蛋白如H3 K27 M中的突变代表了一种新出现的致癌驱动因子， 各种各样的癌症，既罕见又普遍。因此，了解H3 K27 M如何影响核 区室相关的基因组组织可以提供基础知识，以更广泛的类 癌组蛋白突变

ABSTRACT Diffuse intrinsic pontine glioma (DIPG) is an aggressive tumor of young children that infiltrates the brainstem, and although radiation therapy can slow tumor growth, most children with DIPG die within two years of diagnosis. To develop more effective therapies, it is necessary to understand the molecular and cellular mechanisms that drive DIPG. Approximately 80% of DIPG tumors carry a point mutation in one of the histone 3 (H3) genes that causes a substitution of lysine 27 for methionine (H3K27M). Expression of H3K27M mutant protein is an oncogenic driver, increasing cell proliferation, and causes aberrant gene expression. Chromatin is a dynamic polymer of DNA and histone proteins, and the trimethylation of H3 lysine 27 (H3K27-me3) correlates with transcriptional repression. H3K27-me3 is catalyzed by EZH2, and the H3K27M mutant histone protein inhibits EZH2 activity, causing widespread loss of H3K27-me3 across the genome. However, this loss of H3K27-me3 does not closely correlate with transcriptional de-repression. For example, in cells expressing the H3K27M mutation, only about 4% of the genes that lose H3K27-me3 increase their expression. Does H3K27M instead affect gene expression by altering “higher-order” aspects of genome organization, such as the interaction of chromatin with specific nuclear “compartments?” The lamina of the inner nuclear membrane is a transcriptionally repressive nuclear compartment. Lamina associated domains (LADs) are large, specific genomic regions located at the nuclear periphery, and genes within LADs exhibit low transcriptional activity. H3K27-me3 is highly enriched at the “borders” between LADs and non-LAD genomic regions, suggesting a role for this chromatin modification in regulating this aspect of higher-order genome organization. In Preliminary Studies, we have mapped LADs in DIPG cells that carry the H3K27M mutation, finding that the LAD architecture is uniquely disrupted as compared to other types of cancer and normal cells. Given our biochemical understanding of H3K27M mutant protein, the function of EZH2, and our knowledge of LAD border structure, we propose experiments that test the following hypothesis: By inhibiting EZH2 and causing genome-wide changes in H3K27-me3 levels, the H3K27M mutant protein disrupts normal LAD structures, and that such disruption of nuclear compartment-associated genome organization underlies the abnormal gene expression of DIPG. This work challenges the current understanding of H3K27M oncogenic effects. Instead of building upon the current model that explains H3K27M oncogenic function at the level of local chromatin state changes, results obtained test a new paradigm, that H3K27M causes a much larger-scale change to the organization of the genome in the nucleus. Mutations in histone proteins such as H3K27M represent an emerging class of oncogenic drivers in a wide diversity of cancer, both rare and prevalent. Thus, understanding how H3K27M affects nuclear compartment-associated genome organization may provide foundational knowledge to the broader class of oncohistone mutations.