Regulation of Gene Transcription

类核结构：从我们之前的遗传学研究中，我们提出了大肠杆菌类核有一个确定的结构，这决定了它的转录模式。这是基于我们的发现，核蛋白HU的突变极大地改变了转录模式。为了遵循这个想法，我们想知道类核的结构。我们知道(1)HU参与染色体折叠；(2)长期以来，我们知道一个或多个未知的RNA参与染色体的形成；(3)我们也知道HU与RNA结合。我们测定了HU在大肠杆菌中的RNA结合谱。我们设计了一种Rip-Chip方法来鉴定与Hu结合的RNA物种。它们是：80个tRNAs、全部rRNA、5个非编码RNAs和29个mRNAs片段。其中一个非编码RNA(NON5)是染色体上数百个DNA序列的同源物。非5个DNA序列(有一个或两个错配)分布在染色体周围，该序列经常以重复单位出现在染色体的每个座位上。我们认为，安虎-non-5RNA复合体与非5个基因座中的每一个结合，然后聚集形成一个多结构域的类核结构(电子显微镜下曾在染色体上看到)。每个结构域都有不同的超螺旋，这影响了其组成启动子的转录。为了测试Hu-RNA介导的DNA结构域的形成，我们目前正在构建一个9kb的质粒DNA，其中有四个非5的DNA位点放置在不同的标记位置。我们建议在AFM下，在HU和Non-5 RNA存在的情况下观察该质粒，根据我们的模型，可能是三叶草的叶结构。GalR的类核折叠：通过使用3C(染色体构象捕获)分析，我们已经证明GalR是一种被鉴定为调节子特异性转录因子的靶标，它定位于大肠杆菌染色体周围的数百个结合位点。我们已经证明GalR与这些位点结合并与DNA结合有助于在染色体上形成三级结构。染色体的这种特殊折叠有助于DNA在类核中凝聚。我们认为其他类核蛋白(HNS、HU等)通过类似的机制帮助染色体凝聚。类核结构和转录：我们之前已经证明，在类核蛋白HU的突变体中，大肠杆菌中的基因表达模式与野生型模式有很大的不同。我们通过DNA拼接阵列实验发现，这种变化起源于转录水平。这些结果与HU参与折叠染色体的模型一致，一种折叠给出了一种转录图谱，另一种折叠给出了不同的转录图谱。腺苷环化酶-C反应蛋白相互作用：C反应蛋白与环磷酸腺苷结合，在大肠杆菌中激活许多基因的转录。CAMP是由来自三磷酸腺苷的腺苷环化酶合成的。虽然细胞中只存在100微摩尔的cAMP，但它负责野生型细胞中cAMP-CRP介导的最大量转录，需要将510 mM的cAMP添加到腺苷环化酶缺陷突变细胞中才能实现高水平转录。CAMP对细胞没有渗透屏障。我们解决了这个悖论，证明了腺苷环化酶与CRP结合，并通过隧道直接传递产物(CAMP)，而不是扩散到介质中。这使得cAMP在低得多的cAMP浓度下激活转录成为可能。我们通过(I)两种蛋白质相互交联的能力和(Ii)分离和鉴定CRP的非接触性突变体来证明环酶CRP的接触。用腺苷环化酶和三磷酸腺苷取代cAMP，在体外偶联转录系统中重现高效转录。噬菌体lambda转录调控中的DNA环：我们和其他人之前已经证明，在噬菌体状态下，噬菌体lambda基因的调控中存在DNA环。噬菌体蛋白CI抑制其裂解基因转录的两个启动子PL和PR的转录，这两个启动子相距3kb。CI通过分别与其同源运算符元件OL1、OL2、OL3和OR1、OR2、OR3结合而发挥作用。CI通过与OR2、OR3和OL3基因座结合来调节自身的合成。这一调控要求通过连接由操作者结合的CI分子介导的两组操纵子来形成DNA环。我们已经通过在纯化系统中使用不同的操作符突变体以不同的组合进行DNA环化来确认这一规则。单DNA分子研究：通过拴系粒子运动(TPM)分析确定了Lambda DNA环、其不同几何形状及其稳定性。

Nucleoid Structure: From our previous genetic study, we proposed that E. coli nucleoid has a defined structure, which dictates its transcription profile. This was based on our finding that a mutation in the nucleoid protein, HU, altered the transcription pattern dramatically. To follow this idea, we wanted to know the nucleoid structure. We knew that (i) HU participates chromosome folding; (ii) It was known for a long time that one or more unknown RNA participates in chromosome formation; (iii) It was also known that HU binds to RNA. We determined the RNA binding profile of HU in E. coli. We devised a Rip-chip assay to identify the RNA species that bind to HU. They are: 80 tRNAs, all rRNA, 5 non-coding RNAs, and segments of 29 mRNA. One of the non-coding RNA (non5) is homologues to hundreds of DNA sequences in the chromosome. The non5 DNA sequences (with one or two mismatch) are distributed around the chromosome, and the sequence is present frequently in repeating units at each locus in the chromosome. We proposed that an HU-non5 RNA complex binds to each of the non5 DNA loci and then aggregated to give rise to a multi domain nucleoid structure (previously seen in the chromosome by Electron Microscope). Each domain has different superhelecity, which influences the transcription of its constituent promoters. To test the HU-RNA mediated DNA domain formation, we are currently constructing a 9KB plasmid DNA, in which there are four non5 DNA sites placed at different marked positions. We propose to look at the plasmid under AFM in the presence of HU and non5 RNA to see, according to our model, a putative clover leaf structure. Nucleoid folding by GalR: By the use of 3C (chromosome conformation capture) assays, we have shown that GalR, identified as a regulon specific transcription factor targets several hundred binding sites around the E. coli chromosome. We have shown that GalR bind to these sites and associate while DNA-bound to help formation of a tertiary structure in the chromosome. This kind of specific folding of the chromosome helps DNA condensation in the nucleoid. We believe other nucleoid proteins (HNS, HU, etc) help chromosome condensation by similar mechanisms. Nucleoid structure and transcription: We have previously shown that gene expression patterns in E. coli dramatically changes from the wild type pattern in mutants of the nucleoid protein, HU. We have found by DNA tiling array experiments that such changes originate at the level of transcription. These results are consistent with the model that HU participates in folding chromosome; one kind of fold gives one kind of transcription profile, and another kind of fold a different profile. Adenylate cyclase CRP interaction: CRP protein which binds to cAMP, activates transcription of numerous genes in E. coli. cAMP is made by the enzyme adenylate cyclase from ATP. Although only 100 micro molar cAMP, present in the cell, it is responsible for maximal amount of cAMP-CRP mediated transcription in the wild type cell, 510 mM cAMP needs to be added to an adenylate cyclase defective mutant cell for high level transcription. There is no permeability barrier to cAMP to cell. We resolved the paradox by showing that adenylate cyclase binds to CRP and deliver the product (cAMP) directly by tunneling without diffusion to the medium. This makes it possible for cAMP to activate transcription at a much lower cAMP concentration. We demonstrated the cyclase CRP contact by (i) cross-linking ability of the two proteins to each other, and (ii) isolating and characterizing non-contacting mutants of CRP. The high efficiency of transcription was re-produced in an in vitro coupled transcription system by replacing cAMP by adenylate cyclase and ATP. DNA looping in phage lambda transcription regulation: We and others have previously shown the existence of DNA loops in the regulation of phage lambda genes in a prophage state. The phage protein CI represses transcription from its two promoters for lytic gene transcription, PL and PR, located 3 KB apart. CI acts by binding to their cognate operator elements, OL1, OL2, OL3 and OR1, OR2, OR3 respectively. CI regulates its own synthesis by binding to OR2, OR3 and OL3 loci. This regulation requires formation of a DNA loop by linking two sets of operators mediated by the operators-bound CI molecules. We have confirmed the regulations by DNA looping in a purified system by the use of different operator mutants in various combinations. Single DNA molecule studies: The lambda DNA loop, its different geometric forms and their stability were determined by tethered particle motion (TPM) analysis.