Systematic analysis of gene regulatory networks

我们一直在开发工具和资源，使其能够在各种实验条件下分析大量基因。在我们早期的工作中，我们1）构建了来自早期小鼠胚胎和干细胞的cDNA文库，并产生了大量的表达序列标签（EST），2）开发了一个载玻片微阵列平台，该平台包含原位合成的60-mer寡核苷酸探针，代表了大约44，000个独特的小鼠转录本，3）制作了基于网络的ANOVA-FDR软件，以提供用户友好的微阵列数据分析，和4）开发了一种算法和全自动计算管道，用于从与小鼠基因组比对的表达序列组装转录物。此外，我们最近开发了一个全面的数据库和网络浏览器的结合位点的转录因子（TF）和顺式调控模块（CRM）的小鼠基因组。这些资源和工具现在被应用于小鼠胚胎干细胞基因调控网络的系统分析。在我们的试点项目中，我们已经证明，通过监测小鼠ES细胞系的全局基因表达模式，可以分析和鉴定下游靶基因;当编码特定TF的基因（在这种情况下为Pou 5 f1或Oct 4）被操纵时，该基因可以过表达或抑制。 为了进一步扩展我们的策略，到目前为止，我们产生了137个ES细胞系，在每个细胞系中，总共137种不同TF中的一种可以以四环素诱导的方式过表达。我们已经对这些ES细胞系进行了如下表征：（i）通过免疫组织化学对Flag标记的转录因子进行亚细胞定位;（ii）通过定量RT-PCR对操纵的转录因子进行诱导水平;（iii）在诱导转录因子之前和之后基于DNA微阵列的表达谱分析;（iv）蛋白质印迹法;以及（v）核型分析。总之，这些结果表明，我们已经产生了可靠的TF-可操纵的ES细胞系。我们对前50个ES细胞系进行了详细的分析，发现在50个TF中，Cdx 2在ES细胞中引起了最广泛的转录组扰动，其次是Esx 1，Sox 9，Tcf 3，Klf 4和Gata 3。ChIP-Seq显示CDX 2与上调的靶基因的启动子结合。相比之下，被CDX 2下调的基因不显示CDX 2结合，但富含POU 5 F1、SOX 2和NANOG的结合位点。与这些核心TF的结合位点的基因也下调诱导至少15个其他TF，这表明一个共同的初始步骤的ES细胞分化介导的干扰核心TF的结合，其靶基因。对其他TF可操作的小鼠ES细胞系的进一步分析表明，单个TF的过表达确实足以启动小鼠ES细胞分化为特定的细胞谱系。 一般来说，人们认为功能丧失研究比功能获得研究更有助于描绘基因网络和揭示基因的功能。因此，除了TF的过表达（即，功能获得研究），我们用shRNA系统地抑制了100个TF中的每一个，并在小鼠胚胎干（ES）细胞中进行了全局基因表达谱分析。出乎意料的是，只有少数TF的抑制显著影响了转录组，其在两个方向/轨迹上发生变化：一个轨迹是Pou 5 f1或Sox 2的抑制;另一个轨迹是Esrrb，Sall 4，Nanog或Tcfap 4的抑制。这些数据表明，基因表达变化的轨迹已经由基因调控网络预先配置，并分别大致对应于细胞分化的胚外和胚胎命运。这些数据也表明了多能性基因网络的稳健性，因为大多数TF的瞬时抑制并没有改变转录组。

We have been developing tools and resources that make it possible to analyze a large number of genes in various experimental conditions. In our earlier work, we 1) constructed cDNA libraries from early mouse embryos and stem cells and generated a large number of expressed sequence tags (ESTs), 2) developed a glass-slide microarray platform containing in situ-synthesized 60-mer oligonucleotide probes representing approximately 44,000 unique mouse transcripts, 3) produced the web-based ANOVA-FDR software to provide user-friendly microarray data analysis, and 4) developed an algorithm and a fully-automated computational pipeline for transcript assembly from expressed sequences aligned to the mouse genome. In addition, we recently developed a comprehensive database and web browser of the binding sites of transcription factors (TFs) and cis-regulatory modules (CRMs) on the mouse genome. These resources and tools are now applied to the systematic analysis of gene regulatory networks in mouse embryonic stem cells. In our pilot project, we have demonstrated that it is possible to analyze and identify downstream target genes by monitoring the global gene expression patterns of mouse ES cell lines; when a gene encoding a specific TF (Pou5f1 or Oct4 in this case) is manipulated so that the gene can be overexpressed or repressed. To extend our strategy further, we generated 137 ES cell lines thus far, in each of which one of a total of 137 different TFs can be overexpressed in a tetracycline-inducible manner. We have been characterizing these ES cell lines as follows: (i) subcellular localization of Flag-tagged transcription factors by immunohistochemistry; (ii) induction levels of the manipulated transcription factors by quantitative RT-PCR, (iii) DNA microarray-based expression profiling before and after the induction of transcription factors; (iv) western blotting, and (v) karyotyping. Together, these results indicate that we have generated reliable TF-manipulable ES cell lines. We carried out detailed analyses of the first 50 ES cell lines and found that among the 50 TFs, Cdx2 provoked the most extensive transcriptome perturbation in ES cells, followed by Esx1, Sox9, Tcf3, Klf4, and Gata3. ChIP-Seq revealed that CDX2 binds to promoters of up-regulated target genes. By contrast, genes down-regulated by CDX2 did not show CDX2 binding, but were enriched with binding sites for POU5F1, SOX2, and NANOG. Genes with binding sites for these core TFs were also down-regulated by the induction of at least 15 other TFs, suggesting a common initial step for ES cell differentiation mediated by interference with the binding of core TFs to their target genes. Further analyses of additional TF-manipulable mouse ES cell lines demonstrated that indeed overexpression of a single TF is sufficient to initiate the differentiation of mouse ES cells into specific cell lineages. In general, it has been thought that loss-of-function studies are more useful for delineating a gene network and revealing the function of a gene than the gain-of-function studies. Therefore, in addition to the overexpression of TFs (i.e., gain-of-function study), we systematically repressed each of 100 TFs with shRNA and carried out global gene expression profiling in mouse embryonic stem (ES) cells. Unexpectedly, only the repression of a handful of TFs significantly affected transcriptomes, which changed in two directions/trajectories: one trajectory by the repression of either Pou5f1 or Sox2; the other trajectory by the repression of either Esrrb, Sall4, Nanog, or Tcfap4. The data suggest that the trajectories of gene expression change are already preconfigured by the gene regulatory network and roughly correspond to extraembryonic and embryonic fates of cell differentiation, respectively. These data also indicate the robustness of the pluripotency gene network, as the transient repression of most TFs did not alter the transcriptomes.