在 x 染色体失活过程中， 早期染色体变化的影响

The Implication of Early Chromatin Changes in X Chromosome Inactivation

题目：在 x 染色体失活过程中， 早期染色体变化的影响

作者及单位：

Edith Heard

Affiliations

Institut Curie, PSL Research University, CNRS UMR3215, INSERM U934, UPMC Paris-Sorbonne, 75005 Paris, France

发表期刊及时间

Cell. Published: December 27, 2018

Highlights

• •An epigenomic roadmap for initiation of X chromosome inactivation (XCI)
• •Histone deacetylation and H2A ubiquitination are among the earliest XCI events
• •HDAC3-mediated histone deacetylation is required for efficient XCI
• •PcG marks are first deposited intergenically and spread when gene silencing occurs

亮点

• 启动X染色体失活（XCI）的表观基因组学路线图
• 组蛋白去乙酰化和H2A泛素化是最早的XCI事件之一
• HDAC3介导的组蛋白脱乙酰化是有效XCI所必需的
• PcG标记首先在遗传上存储并在发生基因沉默时扩散

Summary

During development, the precise relationships between transcription and chromatin modifications often remain unclear. We use the X chromosome inactivation (XCI) paradigm to explore the implication of chromatin changes in gene silencing. Using female mouse embryonic stem cells, we initiate XCI by inducing Xist and then monitor the temporal changes in transcription and chromatin by allele-specific profiling. This reveals histone deacetylation and H2AK119 ubiquitination as the earliest chromatin alterations during XCI. We show that HDAC3 is pre-bound on the X chromosome and that, upon Xist coating, its activity is required for efficient gene silencing. We also reveal that first PRC1-associated H2AK119Ub and then PRC2-associated H3K27me3 accumulate initially at large intergenic domains that can then spread into genes only in the context of histone deacetylation and gene silencing. Our results reveal the hierarchy of chromatin events during the initiation of XCI and identify key roles for chromatin in the early steps of transcriptional silencing.

摘要

在发育过程中，转录和染色质修饰之间的精确关系仍然不清楚。我们使用 X 染色体失活(XCI) 程序来探讨染色质改变在基因沉默中的影响。我们利用雌性小鼠胚胎干细胞，通过诱导 Xist 来启动 XCI，然后通过等位基因特异性谱来监测转录和染色质的时间变化。该实验揭示了组蛋白去 乙酰化和 H2AK119 泛素化是 XCI 过程中最早的染色质改变。我们发现 HDAC 3 在 X 染色体上 是预结合的，在 Xist 包裹时，它的活性是有效的基因沉默所必需的。我们还发现，第一个 PRC 1 相关的 H2AK119Ub，以及 PRC 2 相关的 H3K27me3 先在大的基因间区域积累，然后在 组蛋白去乙酰化和基因沉默后才能扩散到基因中。我们的结果揭示了在 XCI 启动过程中染色 质事件的层次性，并确定了染色质在转录沉默早期的关键作用

图表选析

image.png

Figure 1Histone Deacetylation Is among the First Events of XCI

(A) Schematic representation of the experimental design. The hybrid TX1072 mouse ESC line was used, in which Xist can be induced from the endogenous B6 allele.

(B) H3K27ac (blue) deacetylation dynamics in comparison with the loss of other active histone marks. Shown are average d-scores (shading is the interquartile range) of all peaks at the X chromosome.

(C) Paired comparison of the dynamics in H3K27ac and H3K4 methylation loss. Plotted are IC35 parameters of peaks at active promoters (left) and putative active enhancers (right) and within the quantitative range of the experiment (IC35 < 24 hr). The p value is from a paired Wilcoxon rank-sum test.

(D and E) Genome browser tracks showing H3K27ac and H3K4me3 at the Pdk3 promoter (D) or H3K27ac and H3K4me1 at a putative Pdk3 enhancer (E). Allele-specific tracks are overlaid (Cast reads in blue; B6 reads in red). The plots show the quantification of the inactivation dynamics with fitted sigmoidal curves.

图 1.组蛋白去乙酰化是 XCI 发生的第一批事件之一。 (A)实验设计原理图。本研究采用 ESC 系 杂种 TX1072 小鼠，利用内源性 B6 等位基因诱导 Xist。 (B)H3K27ac(蓝色)去乙酰化动力学与 其他活性组蛋白标记的丢失相比较。显示的是 X 染色体上所有峰的平均 d 分数(阴影是四分 位数区间)。 (C)H3K27ac 和 H3K4 甲基化缺失动态的配对比较。绘制了活性启动子(左)、活性 增强子(右)和实验定量范围内(IC 35<24 小时)峰的 IC 35 参数。 p 值来自配对 Wilcoxon 秩和检 验.(d 和 E)基因组示踪标记在 Pdk 3 启动子(D)上显示 H3K27ac 和 H3K4me3，或在假定的 Pdk 3 增强子(E)上显示 H3K27ac 和 H3K4me1。等位基因特异性踪迹被覆盖了(Cast 是蓝色的， B6 是红色的)。用拟合的反曲线表示失活动力学的定量。

image.png

Figure 2Histone Deacetylation Is Tightly Correlated with Transcriptional Silencing

图2. 组蛋白脱乙酰化与转录沉默密切相关

(A) Comparison of IC35 parameters for gene silencing (TT-seq) and histone modifications at associated promoters. The p values are from a paired Wilcoxon rank-sum test.

(B) Pairwise comparison of IC35 from gene silencing and histone mark dynamics as in (A). Linear regression fitting (red) and perfect correlation with slope = 1 (black) is shown. Pearson’s correlation (ρ) and p value are shown for each set.

(C) Heatmap showing gene silencing and associated promoter H3K27 deacetylation dynamics. Rows were ordered by their genomic position from centromere to telomere (the green triangle indicates the Xist location). The averages from biological duplicates are shown.

(D) Quantification of features associated with early and late deacetylated promoters (H3K27ac). Shown are gene density, LINE density, proximity to LADs, and distance from the Xist locus. The p values are from a Wilcoxon rank-sum test with Benjamini Hochberg correction.

（A）相关启动子基因沉默（TT-seq）和组蛋白修饰的IC35参数的比较。 配对 Wilcoxon 秩和检验计算 p 值。 （B） 如（a） 中基因沉默和组蛋白动态标记所示， 失活标准参数为 0.35（IC35） 时的配对比较。 线性回归拟合（红色） 和与斜率=1（黑 色） 的相关线。 每一组都显示了皮尔逊相关性（ρ） 和 P 值。 （C） 显示了基因沉默和相关启动子 H3K27 去乙酰化的热图。 行按其 基因组位置从着丝粒到端粒排列（绿色三角形表示 Xist 的位置） 。 （D） 早期和晚期去乙酰化启动子相关特征（H3K27AC） 的量化。 显 示了基因密度、 长散在重复序列（LINE） 密度、 与层粘连相关结构 域（LADS）的接近程度以及与现有位点的距离。 p 值来自于 Wilcoxon 秩和检验和 Benjamini Hochberg 校正。

image.png

Figure S2Histone Deacetylation Correlates with Transcriptional Silencing and Some Genomic Features, Related to Figure 2

图S2Histone去乙酰化与转录沉默和一些基因组特征相关，与图2相关

(A) Violin plots comparing the dynamics of transcriptional silencing and the loss of active histone modifications from associated promoters. Plots compare the IC35 parameter calculated from TT-seq nascent transcription profiling to the IC35 of different active histone marks (H3K27ac, H4ac, H3K9ac, H3K4me3). p values were calculated using paired Wilcoxon rank sum test. (B) Heatmap of H3K27ac allelic dynamics at all promoters clustered using k-means. (C) An array showing features enriched (gray font) or depleted (black font) at promoters efficiently inactivated (low IC35) compared to late inactivated (higher IC35) using a panel of histone modifications. Shown are p values calculated using Wilcoxon rank sum test with Benjamini Hochberg correction. Red cells are reaching statistical significance of p < 0.05.

（A）小提琴图比较转录沉默的动态过程和相关启动子的活性组蛋白修饰的丧失。图比较了从TT-seq新生转录分析计算了不同活性组蛋白标记（H3K27ac，H4ac，H3K9ac，H3K4me3）的IC35参数。使用配对的Wilcoxon秩和检验计算p值。 （B）使用k-means聚类的所有启动子的H3K27ac等位基因动力学的热图。 （C）与使用一组组蛋白修饰的晚期灭活（更高的IC35）相比，在启动子处有效地灭活（低IC35）的特征丰富（灰色字体）或耗尽（黑色字体）的阵列。显示了使用Wilcoxon秩和检验与Benjamini Hochberg校正计算的p值。红细胞达到p <0.05的统计学显着性。

image.png

Figure 4HDAC3 Promotes Silencing by Rapidly Deacetylating Histones

图四、HDAC3 通过快速去乙酰化组蛋白来促进沉默

(A) IF/RNA FISH on TX1072 (Hdac3+/+) and two Hdac3−/− ESC clones induced with DOX for 24 hr. Cells were probed using anti-H3K27ac (red) antibodies and a Xist intronic probe (green). The white dotted line encircles the Xist domain. Because of technical variability, contrast settings were individually adjusted to reflect relative enrichment. Scale bar, 10 μm.

(A)在 TX1072 (Hdac3+/+)和两个 Hdac3-/- ESC 克隆上用 DOX 诱导 24h 的 IF/RNA FISH。 使用抗 h3k27ac(红色)抗体和 Xist 内含子探针(绿色)探测细胞。白色虚线围绕 Xist 领域。 由于技术上的差异，对比度设置被单独调整以反映相对富集。 比例尺为 10 微米

(B) Quantification of (A) over at least 100 Xist clouds. Shown is the average signal for Xistand H3K27ac when profiles were aligned to the Xist cloud boundary (for more details, refer to Figure S4A and D).

对至少 100 个 Xist 云进行量化。所示 当 谱比对到Xist 云边界时的Xist 和H3K27ac 的平均信号 (更多细节请参见图 S4A 和 D)

(C) Relationship between gene repression (RNA-seq) and promoter deacetylation (H3K27ac) at 24 hr in TX1072 (Hdac3+/+, blue) and Hdac3−/− (red). On the scatterplot, the black dotted line represents fitted linear regression. ρ = 0.860. Density plots show the distribution of the normalized d-score for H3K27ac (horizontal) and RNA-seq (vertical); note that d-scores in mutant cells are shifted toward biallelic levels (0.5, dotted lines on density plots). Each dot represents the average normalized d-score from biological duplicates.

基因抑制(RNA-seq)和启动子脱乙酰作用之间的关系(H3K27ac)24 hr TX1072(Hdac3 + / + 蓝)和 Hdac3--/(红色)。 在散点图上， 黑色虚线表示拟合的线性回归。 ρ= 0.860。 密度图显 示 H3K27ac(水平)和 RNA-seq(垂直)的归一化 d-score 分布;注意，突变体细胞的 d 分数向 双等位基因水平转移(密度图上的点线为 0.5)。每个点代表来自生物复制的平均标准化 d 得 分。

(D) Comparison of promoter deacetylation dynamics and transcriptional silencing after 24 hr of DOX treatment in TX1072 (Hdac3+/+) and Hdac3−/− ESCs. Examples of genes sensitive (top) and more resistant (bottom) to HDAC3 loss are shown. For RNA-seq, shown is the average from biological duplicates (± SD).

在 DOX 处理 TX1072 (Hdac3+/+) 和 Hdac3-/- ESCs 24 小时后子脱乙酰作用动力学和转录沉 默的比较。 显示基因敏感(上)和更耐受(下)HDAC3 丢失的例子。 RNA-seq,显示生物复制的平 均值(±SD)。

翻译小组：

陈凯星、王俊豪、黄敬潼、黄子亮、邓峻玮、郑凌伶