The early stages of transcription are critical for regulation of gene expression in eukaryotes. Our mechanistic understanding of this process remains very limited due to the difficulty in obtaining structural information brought about by the size and complexity of the molecular players involved. We are using electron microscopy and single particle 3-D reconstruction to characterize the architecture, dynamics and interactions of large human transcription complexes involved in transcription initiation. In the previous funding period we determined the architecture of co-factors CRSP and ARC-L, the chromatin remodeling complex PBAF, and the general factors TFIID and RNAPII. For the last two complexes, our statistical analysis showed that the transition between conformations in solution, which we believe reflect mechanistic stages essential to their function. We will extend this work to characterize the repressive interaction of RNAPII with non-coding RNAs (ncRNAs), and to study the binding of TFIID to core promoter DNA, TFIIA and TFIIB, and human Mediator. Where does binding occur? What is the effect of binding on conformation and dynamics? How is this information integrated at the promoter? The goals of this proposal will shed light on the essential step of core promoter recognition, and on processes of transcriptional activation and transcriptional repression that act at the very early stages of gene transcription. By working with full, functional human complexes, our work will better reflect the extent and complexity of protein-protein and protein-nucleic acid interactions that occur in the cellular context. On the other hand, the availability of some relevant crystallographic structures (e.g. yeast RNAPII, TBP, TFIIA., TFIIB) will give added value to our reconstructions, by allowing us to generate pseudo-atomic models via docking of the crystallographic models into our 3-D densities. Finally, we hope to go beyond static pictures, by being able to exploit the solution state of our samples and new image reconstruction methods and statistical tools, to describe the conformational landscape intrinsic to these dynamic complexes, and to the dynamic processes they drive.
Transcriptional regulation in eukaryotes is a multifaceted process involving a very large number of protein complexes. Most of these complexes cannot be produced recombinantly and reconstituted, so that all in vitro analysis requires purification from endogenous material. Given that these complexes, by nature of their function, exist in very small amounts in the cell, biochemical and biophysical studies have been significantly limited. This in turn has lead to an almost complete lack of mechanistic understanding of how these complexes work at the molecular level. Using electron microscopy and single particle image analysis, the methodology of choice in the structural characterization of large, low-abundance complexes, we propose to shed new light on the essential step of core promoter recognition, and on processes of transcriptional activation and transcriptional repression that act at the very early stages of gene transcription.
Showing the most recent 10 out of 25 publications
