Epigenetic and transcriptional control in hematopoietic development and lineage differentiation

Hematopoiesis is a well-established model system to study molecular mechanisms of lineage-specific differentiation. Key transcription factors (TFs), such as PU.1, NF-E2 and GATA1 are implicated in crucial aspects of distinct hematopoietic lineages. How TFs collaborate with histone modificatons and how they affect the chromatin status remains to be elucidated. Chromatin-Immunoprecipitation followed by next-generation sequencing (ChIP-seq) has been proven to be an excellent tool to study chromatin modifications genome-wide. In this study, ChIP-seq was used to investigate the H3K4me2 landscape at enhancers during hematopoietic differentiation, starting from the hematopoietic stem cell (HSC) up to the fully differentiated erythrocyte, megakaryocyte or granulocyte. Although cell morphology and gene expression profiles differ extensively in committed erythrocytes, megakaryocytes and granulocytes, the genomic landscape of H3K4me2 is surprisingly alike across the three cell types. Granulocytes, in particular, seem to display an ‘erythocyte-like’ chromatin pattern. Similar results were observed for another chromatin mark, H3K27ac. Unexpectedly, the common progenitors of erythrocytes and granulocytes do not display a similar chromatin pattern, excluding the idea that the H3K4me2 mark is placed early on during differentiation. These results also suggest that the chromatin state is probably not the determining factor for lineage differentiation, but that rather lineage specific TF binding is important. In agreement with this, mature megakaryocytes loose the H3K4me2 mark surrounding erythrocyte-specific genes. This might be explained because megakaryocytes and erythrocytes both express the lineage specific TFs GATA1 and NF-E2. Altogether, we show that committed granulocytes and erythrocytes display similar H3K4me2 and H3K27ac patterns, and that those marks alone cannot predict which genes will be expressed eventually.

造血作用（Hematopoiesis）是研究谱系特异性分化分子机制的成熟模型体系。关键转录因子（Transcription Factors，TFs）如PU.1、NF-E2与GATA1，参与调控不同造血谱系的核心生物学过程。目前，转录因子如何与组蛋白修饰协同作用，以及它们如何影响染色质状态，仍有待阐明。染色质免疫共沉淀结合高通量测序（Chromatin Immunoprecipitation followed by Next-Generation Sequencing，ChIP-seq）已被证实是全基因组范围内研究染色质修饰的优异工具。本研究采用ChIP-seq技术，探究了从造血干细胞（Hematopoietic Stem Cell，HSC）至终末分化的红细胞、巨核细胞或粒细胞的造血分化过程中，增强子区域的H3K4me2修饰图谱。尽管定型红细胞、巨核细胞与粒细胞在细胞形态与基因表达谱上存在显著差异，但三者的H3K4me2全基因组图谱却惊人地相似。尤为值得注意的是，粒细胞呈现出类红细胞的染色质模式。另一种染色质修饰标记H3K27ac也得到了相似的实验结果。出乎意料的是，红细胞与粒细胞的共同祖细胞并未呈现相似的染色质模式，这排除了H3K4me2标记在分化早期即被建立的假说。上述结果同时表明，染色质状态或许并非谱系分化的决定性因素，而谱系特异性转录因子结合才是关键。与此一致，成熟巨核细胞会丢失红细胞特异性基因周边区域的H3K4me2标记，这一现象可通过巨核细胞与红细胞均表达谱系特异性转录因子GATA1与NF-E2得以解释。综上，本研究证实定型粒细胞与红细胞具有相似的H3K4me2与H3K27ac修饰模式，且仅靠这类修饰标记无法最终预测哪些基因会被表达。