Logo Logo
Switch language to English
Boeing, Stefan (2008): Functions of Mediator and the RNA Polymerase II C-terminal Domain in Transcription Initiation. Dissertation, LMU München: Fakultät für Chemie und Pharmazie



RNA polymerase II (RNAPII) has been identified almost 40 years ago, but the molecular details of its regulation and fine tuning during messenger RNA (mRNA) synthesis are still far from understood. Subsequently to RNAPII six general transcription factors (GTFs; TFIIA, TFIIB, TFIID, TFIIE, TFIIF, TFIIH) were discovered of which all except TFIIA are necessary and sufficient for promoter-dependent basal transcription initiation. In addition to the GTFs activator-dependent transcription requires the presence of a transcription cofactor, the Mediator complex. Mediator serves as a link between transcription activators, enhancers and the general transcription machinery. Initial studies revealed that Mediator stimulates the activity of the TFIIH associated kinase CDK7 and thereby facilitates RNAPII C-terminal domain (CTD) phosphorylation. Furthermore the Mediator complex interacts functionally with several signal transduction pathways and serves as an signal integration platform. In order to dissect the process of transcription initiation, early studies made use of in vitro transcription systems reconstituted from recombinant or highly purified GTFs and RNAPII. In this system basal, activator-independent transcription does not require the presence of the Mediator complex. If however a more physiological nuclear extract transcription system is used, our laboratory and others have established previously that basal transcription becomes critically dependent on Mediator. Another difference between both transcription systems is that the first is insensitive to the kinase inhibitor H8 whereas in the second transcription can be inhibited by H8. This suggests that only the second transcription system is regulated by RNAPII CTD phosphorylation. In this thesis the interplay between Mediator, RNAPII, GTFs and transcription cofactors was studied using immobilized promoter template assays in combination with various immunodepleted nuclear extracts and recombinant factors. Negative cofactor 2 (NC2) is an evolutionary conserved general cofactor that binds to many active genes in vivo. Previous studies in our laboratory had shown with recombinant proteins that NC2 competes with TFIIA and TFIIB for binding to TATA-binding protein (TBP) at a promoter in vitro. Genetic studies in yeast provided evidence that Mediator acts antagonistically to NC2. Here I have studied the role of NC2 on preinitiation complex (PIC) formation and transcription in nuclear extracts. I observed rapid association of TFIID with promoters whereas NC2 enters PICs with a slow kinetic which is similar to that of TFIIB recruitment. My data indirectly suggest that TBP binds to DNA in a yet to be defined inactive form (perhaps as a TFIID complex) which is then slowly converted into an active TBP-TATA complex that is rapidly recognized by GTFs or NC2. My data support the notion that NC2 and TFIIB compete for binding to a PIC also in immobilized promoter assays under physiological conditions. NC2 concentrations in nuclear extracts appears to be tightly controlled. Doubling the NC2 concentration in a nuclear extract by adding recombinant NC2 (rNC2) abolished functional PIC formation and transcription. However, the in vitro analysis also showed that upstream of NC2 PIC formation is fully dependent on Mediator. Hence, TFIID binds to a promoter in a nuclear extract in vitro transcription system but we have no indication that a transcription competent PIC is formed in the absence of Mediator. In yeast studies it was reported that upon transcription initiation in vitro several GTFs dissociate from the promoter DNA template whereas the Mediator complex is retained in a reinitiation complex. In the human system I recapitulate this observation for TFIIB and CDK7. In addition I provide evidence that Mediator partially dissociated from the promoter template upon transcription initiation. Upon transcription initiation the middle module subunit MED7 was retained on a promoter template, whereas the tail module subunit MED15 and CDK8 did dissociate. This data suggest that upon transcription initiation a head/middle module Mediator subcomplex is retained at the promoter whereas the tail and CDK8 modules dissociate. Previous studies have established that Mediator promotes CDK7-dependent phosphorylation of the RNAPII CTD at serine-5 (ser-5). Various studies found that CTD ser-5 phosphorylation does coincide with transcription initiation. Using new monoclonal antibodies I observed two functionally distinct modes of CTD ser-5 phosphorylation in vitro: Hypo- and hyperphosphorylation of the largest RNAPII subunit Rpb1. I observed that CTD ser-5 hypophosphorylation is established already before complex opening by TFIIH. I found CTD ser-5 hypophosphorylation to be critically dependent on TBP, Mediator, TFIIB and CDK7. In addition I noted that CTD ser-5 hypophosphorylation correlates with the transcription potential of a PIC. CTD ser-5 hyperphosphorylation was established in a Mediator-dependent fashion but independent of productive transcription. Immunodepletion of CDK7 did not led to a reduction in CTD ser-5 hyperphosphorylation. However, immunodepletion of CDK8 caused a reduction but not a loss of CTD ser-5 hyperphosphorylation upon transcription initiation indicating that another yet to be identified kinase might be involved in this process. These data suggest that CTD ser-5 hypophosphorylation is established only in the PIC context on RNAPII located at bona fide promoter regions but not on RNAPII complexes bound to DNA outside of promoter regions, e.g. in an open reading frame. Recently phosphorylation of the RNAPII CTD at serine-7 (ser-7) was reported. In that study the entire coding region of the TCRβ locus was found to be associated with RNAPII CTD phosphorylated at ser-7. Starting from there I found that establishment of CTD ser-7 phosphorylation in the process of transcription initiation can be recapitulated in an immobilized template assay system in vitro. I confirmed the in vitro finding that establishment of CTD ser-7 phosphorylation correlates with transcription initiation with chromatin immunoprecipitation experiments on an inducible model gene system in vivo. Similar to CTD ser-5 phosphorylation, I observed two modes of CTD ser-7 phosphorylation: CTD ser-7 hypo- and hyperphosphorylation. In contrast to CTD ser-5 hypophosphorylation, which was established before complex opening, I observed establishment of CTD ser-7 hypophosphorylation predominantly after complex opening by TFIIH. Both, CTD ser-7 hypo- and hyperphosphorylation were found to be Mediator-dependent. A mass spectrometric screen for PIC associated kinases (in collaboration with the laboratory of Gerhard Mittler) yielded 13 kinases. Seven of the identified kinases were further tested for their potential to phosphorylate the RNAPII at ser-7 in an immobilized template assay.