Specifically, significantly regulated genes were placed at the top or bottom of the list and ordered by descending or ascending fold changes, respectively. Less significantly regulated selleck chem Dorsomorphin genes were placed in the middle of the list and ordered by ascending or descending p values. GO terms with the top ten Normalized Enrichment Scores were sellckchem selected and combined from both the upregulated and downregulated GO terms in each time point, compared to the 0 min treatment condition. Heatmaps were generated using the heat map. 2 function in the gplots package in R sellectchem with default parameters. Inputs for the heatmaps included nor malized GRO seq signals and NESs. For the latter, heatmaps comparing later to earlier time points were generated in a similar manner, but using different edgeR outputs for the comparisons. Background Genome wide mutagenesis and subsequent phenotype driven screening has been pivotal to a complete under standing of how complex biological processes operate in classical model organisms including flies, nematodes, and plants. The level of saturation in mutagenesis has been shown to be a critical parameter for this approach to determine all relevant genes involved in a biological function, without prior knowledge of the gene products. In mammalian model systems, much effort has been expended to saturate, i. e. to disclose all the genes involved in some specific biological pathways. However, the relatively large scale and labor intensity of experi ments have hampered the achievement of actual sat uration mutagenesis, especially for recessive traits that require biallelic mutations to manifest detectable phe notypes. To overcome these drawbacks, the haploid mouse embryonic stem cell system, in which a single hit mutation can directly lead to phenotypic changes without being compensated by the second copy of the gene, has been recently developed, and reviewed in. Here, to address the issues of mammalian saturation mutagenesis, we mutagenized the haploid mouse ESCs with the chemical mutagen N ethyl N nitrosourea and subjected them to a phenotypic screening of mutants defective in various steps of the glycosylphosphatidylinositol anchor biosynthetic pathway. Results The GPI anchor biosynthetic pathway as a model target of screening We chose the GPI anchor biosynthetic pathway as a model target of phenotype driven screening for a number of reasons.
The GPI anchor is a glycolipid that tethers many proteins to the plasma membrane of eukaryotic cells, forming a diverse family of molecules including hydro lytic enzymes, receptors, adhesion molecules and com plement regulatory proteins. The biosynthetic pathway is mediated by sequential additions of sugars and other components to phosphatidylinositol. The cell surface expression of GPI anchored proteins involves a total of 26 genes, which, with the ex ception of the X chromosome linked phosphatidylinositol glycan anchor biosynthesis, class A gene, are widely distributed throughout the autosomes. Clos tridium septicum toxin selectively binds GPI anchored proteins at the cell surface and kills host cells. A defect in the GPI anchor biosynthetic pathway does not affect ESC proliferation per se, but a loss of function mutation in the pathway results in an toxin resist ant phenotype, providing a platform for positive selec tion screening.