Mechanism of transcription and coupling to protein homeostasis in embryonic stem cells
Final Report Abstract
ES cells have been proposed to harbor especially high concentrations of general transcription factors leading to elevated transcription activity inside cells. In this study we developed in vitro systems for the analysis of ES cells, primary mouse embryonic fibroblasts and compared it with transformed cells. Against the expectation ES cell-derived systems display intermediate activities. They have significant higher activity especially in RNAP-I but also with respect to RNAP-II and RNAP-III activities compared with primary MEFs, but are clearly outperformed by many tumor cell lines at virtually all levels. We characterize ESCs as specifically dependent on CDK7 and propose that this reflects an exceptional dependency on transcription initiation. Primary MEF cells in turn are subject to severe limitations in transcription initiation and show moderately reduced mRNA synthesis rates in vivo. Overall, transcription activity is correlated to the self-renewal capacity of ESCs and to the transformed state of cells. Broadly, however, the data suggest that the homeostasis of RNA produced by all three RNA polymerases is largely preserved during differentiation and transformation. We have further studied RNAP-II transcription mechanisms, investigated the role of specific general factors and studied the pausing process related to RNAP-II in vitro. We show that pause complexes form rapidly, within few minutes after preinitiation complex formation, followed by a lag phase before RNAP-II resumes elongation. This requires the transcription kinases CDK7 and CDK9 and involves among others DSIF, a highly conserved protein complex (SPT4 and SPT5) in eukaryotes and functional homologue of bacterial NusG antitermination factor. We show, for the first time, that DSIF prevents disassembly of elongating RNAP-II complexes and facilitates readthrough through secondary DNA structures. Related to this topic our partner lab in Shanghai (Prof. Ying Jin) investigated the function of an ubiquitin ligase WWP2 that was predicted to control both the master regulator Oct4 as well as RNAP-II in pluripotent cells. The lab eventually developed mouse models for this enzyme. At present, ES cells derived from these mice remained without phenotype probably because of redundancy within the family of WWP2 proteins. However, they did display an unexpected adult phenotype. Some of our originally planned collaborative efforts thus await further combinatorial knockout/knockdown studies on WWP genes.
