Chromatin Structure And Function

我们继续研究了表达基因附近的染色质结构。鸡红系细胞中的珠蛋白基因家族作为一个模型系统，其中有可能研究红系发育过程中与家族的簇和个体成员的调节相关的机制。我们将注意力集中在鸡β-珠蛋白基因座5'端的1.2kb绝缘子DNA序列及其上游元件上，该绝缘子既能阻断外部增强子的影响，又能防止可能关闭整个区域表达的浓缩染色质的侵入。我们之前已经证明，增强子阻断活性与单个蛋白质CTCF与增强子内位点的结合有关。绝缘子还具有保护稳定转染到细胞系或动物中的位置效应报告基因的单独能力，作为防止浓缩染色质侵入的边界。我们发现这种保护能力存在于一个？2kb绝缘子内的250 bp长的“核心”元件，并且核心内含有CTCF位点的亚区的缺失不影响活性。然而，其他四个亚区对应的结合位点的核蛋白是重要的位置效应保护（边界功能）。我们发现，这些结合位点之一是专门负责维持高水平的组蛋白乙酰化和甲基化的网站与基因激活。该位点结合蛋白USF 1和USF 2的异源二聚体。这些结果与我们提出的模型是一致的，在该模型中，屏障功能与绝缘体附近的多个组蛋白修饰有关。为了了解USF蛋白如何介导这一作用，我们研究了USF 1通过体内染色质免疫沉淀和体外共沉淀招募的辅因子。我们已经表明，精氨酸甲基转移酶，PRMT 1，被招募到绝缘子，并通过RNAi方法的PRMT 1的下调不仅导致组蛋白H4 arg 3甲基化的损失，而且在基因组中广泛的H3和H4乙酰化的损失。现在，我们已经通过与标记的USF 1的共免疫纯化直接鉴定了与USF 1相互作用的蛋白质。当与旨在检测稳定复合物的凝胶过滤研究相结合时，这揭示了两种不同复合物的存在，其中一种含有PRMT 1和HAT PCAF和SRC-1。与其与PRMT 1的相互作用一致，USF 1的siRNA下调导致绝缘子处H4 R3甲基化的局部丢失;与常染色质相关的其他组蛋白修饰也局部丢失。值得注意的是，当USF 1耗尽时，与异染色质化相关的H3 K27甲基化的增加扩散到绝缘子的下游。这与绝缘子通常通过维持活性染色质的局部环境作为屏障的模型一致。 我们在禽类红系细胞系中进行了组蛋白变体H3.3的研究。H3.3在间期掺入染色质，而主要变体H3仅在S期掺入。据报道H3.3主要沉积在活性启动子的下游。我们已经研究了H3.3在α-珠蛋白基因座内和附近的基因上的分布，以及基因组中的其他地方。观察到了各种各样的模式，特别是包括上游调控区如增强子和基因座控制区的显著掺入。在FR（叶酸受体）和VEGFD（血管内皮生长因子D）的情况下，其中掺入限于上游区域，外源H3的存在导致表达减少，而H3.3刺激表达。这表明H3.3可能在基因表达调控中起积极作用。目前正在将研究扩展到其他变体。 作为我们早期对Igf 2/H19位点的染色质结构和表观遗传调控的兴趣的延伸，我们在人类胰岛以及来自其他组织的细胞中对附近的胰岛素基因进行了类似的研究。我们也开始与Marvin C博士合作。Gershengorn和布鲁斯Raaka，NIDDK，对已被诱导经历上皮向间充质转化的人胰岛衍生的前体细胞进行类似的分析。 我们继续与大卫恐惧和汉娜古尔德博士（国王？s College，伦敦）和Martin Gellert博士（NIDDK）研究了人免疫球蛋白H链基因座中生殖系基因附近的染色质结构。我们早期的研究发现，单个细胞包含多个生殖系基因的转录本，表明所有基因的染色质结构都是开放的，结构变化本身不能成为类别转换重组中选择性的基础。我们现在开始测量在类别转换之前和转换期间的基因座上的染色质结构和DNA甲基化模式。

We have continued our studies of chromatin structure in the neighborhood of expressed genes. The globin gene family in chicken erythroid cells serves as a model system in which it is possible to study the mechanisms associated with regulation of the cluster and individual members of the family during erythroid development. We have focused attention on the 1.2 kb insulator DNA sequence at the 5' end of the chicken beta-globin locus, and elements upstream of it. This insulator is capable both of blocking the influence of outside enhancers and of preventing the encroachment of condensed chromatin that might shut down expression of the entire region. We have shown previously that enhancer blocking activity is associated with binding of a single protein, CTCF, to a site within the enhancer. The insulator also has the separate ability to protect against position effects reporter genes that are stably transfected into cell lines or animals, serving as a boundary against encroachment of condensed chromatin. We found that this protective ability is present in a ?core' element, 250 bp long, from within the 1.2 kb insulator, and that deletion of subregions within the core that contain the CTCF site do not affect activity. However four other subregions corresponding to binding sites for nuclear proteins are important for position effect protection (boundary function). We showed that one of these binding sites is specifically responsible for maintaining a high level of histone acetylation and methylation at sites associated with gene activation. This site binds a heterodimer of the proteins USF1and USF2. These results are consistent with a model we have proposed in which barrier function is connected with multiple histone modifications in the neighborhood of the insulator. In an effort to understand how the USF proteins mediate this action, we have investigated what co-factors are recruited by USF1, both by chromatin immunoprecipitation in vivo and by co-precipitation in vitro. We have shown that the arginine methyl transferase, PRMT1, is recruited to the insulator, and that down regulation of PRMT1 by RNAi methods results in loss not only of histone H4 arg 3 methylation, but also in widespread loss of H3 and H4 acetylation over the genome. Now we have directly identified proteins interacting with USF1 by co-immunopurification with tagged USF1. When combined with gel filtration studies designed to detect stable complexes, this revealed the presence of two distinct complexes, one of which contained PRMT1 and the HATs PCAF and SRC-1. Consistent with its interaction with PRMT1, siRNA downregulation of USF1 resulted in localized loss of H4R3 methylation at the insulator; other histone modifications associated with euchromatin are also lost locally. Significantly, increases in H3K27 methylation, associated with heterochromatinization, spread downstream of the insulator when USF1 is depleted. This is consistent with a model in which the insulator normally works as a barrier by maintaining a local environment of active chromatin. We have carried on studies of the histone variant H3.3 in an avian erythroid cell line. H3.3 is incorporated into chromatin during interphase, whereas the predominant variant, H3, is incorporated only during S phase. H3.3 has been reported to be deposited predominantly downstream of active promoters. We have studied the distribution of H3.3 over the genes in and near the a-globin locus, as well as elsewhere in the genome. A wide variety of patterns was observed, notably including marked incorporation over upstream regulatory regions such as enhancers and locus control regions. In the case of FR (Folate Receptor) and VEGFD (Vascular endothelial growth factor D) in which incorporation is confined to upstream regions, the presence of exogenous H3 results in reduced expression, while H3.3 stimulates expression. This suggests that H3.3 may play an active role in regulation of gene expression. Studies are now being extended to other variants. As an extension of our earlier interest in chromatin structure and epigenetic regulation at the Igf2/H19 locus, we have ubdertaken similar studies in human islets, as well as in cells derived from other tissues, of the nearby insulin gene. We have also begun a collaboration with Drs. Marvin C. Gershengorn and Bruce Raaka, NIDDK, to carry out similar analyses of human Islet-derived Precursor Cells that have been induced to undergo an epithelial to mesenchymal transition. We continue a collaboration with Drs. David Fear and Hannah Gould (King?s College, London) and Dr. Martin Gellert, NIDDK, to examine the chromatin structure in the neighborhood of the germ line genes in the human immunoglobulin H chain locus. Our earlier studies had found that individual cells contain transcripts from multiple germ line genes, showing that chromatin structure over all the genes is open, and that structural changes per se could not be the basis of selctivity in class switch recombination. We are now beginning to measure chromatin structure and DNA methylation patterns over the locus before and during class switching.