Final Report Summary - POL2-CCR (Structure-function analysis of mRNA metabolism by Ccr4-Not)
1.1 Background
Gene expression is described by the central dogma of molecular biology: DNA is transcribed by RNA polymerase into mRNA, and mRNA is translated by the ribosome into protein. This process is highly regulated on several levels. The multifunctional, 9-subunit Ccr4-Not complex controls gene expression at all stages, from regulation of mRNA transcription over mRNA decay to repression of protein translation and protein quality control. A recent study showed a positive effect of Ccr4-Not on transcription elongation and suggested that Ccr4-Not may physically interact with RNA polymerase II (Pol II). We aimed at investigating the still elusive transcription promoting mechanism and thus proposed a structure-function analysis of Pol II-Ccr4-Not, using cryo electron microscopy (cryo EM) and interaction assays, complemented by a mRNA interaction studies.
1.2 Results
When I started this project, cryo EM of Pol II complexes typically yielded low to medium resolution structures of 15-20 Å, and high resolution structures could only be obtained by X-ray crystallography, which does not work for complexes where good quality crystals cannot be obtained. I started my project by optimizing the cryo EM workflow for Pol II samples on all levels; including specimen preparation, data collection, and computational analysis including 3D structure solution.
First, since Pol II is a rather small, low contrast molecule compared to typical “good” EM objects like viruses, I tried obtaining very thin ice layers on EM grids improve image quality. The best way to do this for our original microscope setting with a Falcon 2 detector was to use freshly prepared carbon-coated EM grids, along with longer glow-discharging (to improve surface wetting) and long blotting times at high air humidity. For a newer microscope setting with a K2 direct detector, carbon-free EM grids yielded even better results. Next, data collection was optimized to get a trade-off between highest possible number and highest possible quality of images. For this, semi-automatic data collection proved ideal. The user first determined the best grid areas suitable for data collection and the actual spots for later automated image acquisition. This yielded more than 90% good quality images, and typically 0.2 -1 million particles for one microscope session. Best results for particle image extraction from raw micrographs were produced with half-automated reference-based picking, followed by manual checking through all electron micrographs, and finally through all extracted particles.
For data processing, three different software packages were compared: IMAGIC, SPIDER and RELION. From all trials, RELION emerged as the by far best software for Pol II complexes, yielding well-defined, high-resolution Pol II structures without reconstruction artefacts to which the other programs were prone in case parameters were not known and not yet ideal. In order to speed up image processing, my host lab invested heavily into new computing hard ware so that I had access to an extremely powerful computing infrastructure.
Together with my colleague Dr. Martinez from the Cramer lab, I employed my new cryo EM workflow on the structure of the Pol II-capping enzyme complex and in parallel, on a high resolution structure of a transcribing yeast Pol II elongation complex. Capping enzyme catalyzes the first two co-transcriptional enzymatic steps leading to the 5’ cap in eukaryotic mRNAs, which is essential for mRNA splicing, export, translation, and stability. The yeast Pol II elongation complex primarily served as a test case for high resolution EM studies with Pol II, which at the time were not available. We were successful in solving both structures, and elucidated the mechanism capping enzyme uses on newly produced mRNA. This very successful work was published in 2015 in Molecular Cell (Martinez-Rucobo, Kohler, et. al., Mol Cell 58, 2015). It also proved the general feasibility of high-resolution EM structure determination for Pol II complexes using the workflow optimized for the Pol II-Ccr4 project.
To analyse the interaction between Ccr4-Not and Pol II in vitro, I first screened several interaction assays that would be applicable to a large number of different protein variants that I wanted to test. Using pull-down tests with biotinylated Pol II and Streptavidin coated magnetic beads produced best results. I tested various Ccr4-Not sub-complexes for Pol II binding and found that the binding interface is likely located in the nuclease module of Ccr4-Not consisting of the subunits Caf1, Ccr4, and the Not1 N-terminus. Constructs containing these subunits were then used for EM analysis. I also tested the effect of different potential substrate mRNAs on Ccr4-Not binding. Since Ccr4-Not can interact with mRNA, our aim was to find a good substrate mRNA that would lock the Pol II –Ccr4-Not complex into one defined three-dimensional structure, thus suitable for high-resolution structure determination by cryo-EM. However, none of the tested mRNAs improved binding between Ccr4N-Not and Pol II. Further research for better suited substrate mRNAs is thus planned.
The cryo-EM protocol that I generally established for Pol II complexes (see above) was also applied to Pol II-Ccr4-Not complexes. Unfortunately, none of the tested samples showed interpretable density for Ccr4-Not on the polymerase. Most likely, this is due to either sensitive binding that does not survive EM conditions, or to flexible binding, which would lead loss of signal for flexible parts of the structure in the EM reconstructions. Since fellow researchers working on Ccr4-Not complexes with cryo-EM reported similar difficulties, we decided to postpone further cryo-EM studies of Ccr4-Not complexes until more biochemical and functional data were available to allow the design of better suited, structurally more defined Pol II- substrate mRNA- Ccr4-Not reaction intermediates.
When the original plan to solve the high resolution structure of an active Pol II-Ccr4-Not complex turned out to require a longer time frame than feasible within the IEF time frame, another project opened up where again control of gene expression from transcription to translation would be in the focus. It had recently been shown that in bacteria, transcription and translation were coupled. An expert lab specializing on bacterial transcription (University of Wisconsin, USA) offered to work on the functional aspects of the coupling, as well as providing all necessary protein constructs, which would allow quick progression of the project. Considering my previous expertise on translation complexes, and my newly acquired knowledge about transcription complexes in mind, we decided to give this project idea a short trial period to decide if it had chances to lead to high-impact results and a publication within the remaining time frame of the IEF fellowship.
Indeed, in very short time I succeeded in reconstituting a transcription-translation complex and first EM images showed an RNAP-shaped density in a defined position on many particles. We were thus ensured that we had embarked upon a very promising, and highly competitive, project, that when completed, would contribute immensely to the understanding of the linkage between the two essential mechanisms of gene expression. We decided to proceed with the new study and for the moment, to postpone Pol II- Ccr4-Not studies. Our aim was to ensure best deliverables within the timeline of this IEF project, such as another high-profile publication.
By now, I have solved the cryo-EM structure of a transcription-translation complex. Using interaction assays, I identified regions of the RNAP that are central to the interaction, and our collaborators in the US analyzed the in vivo importance of the coupling. From our data, we could deduct a mechanism for direct gene expression control by coordinated transcription and translation. Furthermore, we discovered links to the control of transcriptional pausing and termination, and found that transcription-translation coupling is a conserved feature, which coordinates the two central mechanisms in gene expression, thereby adjusting both to environmental clues and cellular signals. A manuscript will soon be submitted for publication in one of the highest profile journals in the field.
1.3 Socio-economic impact
Establishing a high-resolution cryo EM protocol for Pol II complexes for the original project has paved the way for many other structural studies of multiprotein Pol II complexes. When I started the project, no high resolution EM structures were available for any transcriptional complex. By now, our lab alone has published several important studies featuring high resolution EM structures of RNA polymerase complexes. All those structures have been analysed by very similar protocols to the one I developed and which served as a proof of principle for the feasibility of such projects.
The work Dr. Martinez and I did on the active Pol II - capping enzyme complex represents an important mile stone in the field. The high profile open access publication allows fellow researchers to design future studies based on the molecular structure and mechanism presented by our lab. Since mRNA capping is a central step in the life cycle of most mRNAs, not only the field of transcription but many related scientific areas can profit from our data and our results are relevant to a wide audience.
Our most recent work on the coupled transcription-translation complex will be of huge impact to both the fields of transcription and translation. Since the link between those two crucial mechanisms seems to be evolutionarily conserved, our study will be the foundation for broader research. More functional studies characterizing the in vivo effects of this coupling can now be designed. Our research will be of exceptionally high relevance for a large scientific public and also of interest to the general public. See also: http://www.mpibpc.mpg.de/cramer
Gene expression is described by the central dogma of molecular biology: DNA is transcribed by RNA polymerase into mRNA, and mRNA is translated by the ribosome into protein. This process is highly regulated on several levels. The multifunctional, 9-subunit Ccr4-Not complex controls gene expression at all stages, from regulation of mRNA transcription over mRNA decay to repression of protein translation and protein quality control. A recent study showed a positive effect of Ccr4-Not on transcription elongation and suggested that Ccr4-Not may physically interact with RNA polymerase II (Pol II). We aimed at investigating the still elusive transcription promoting mechanism and thus proposed a structure-function analysis of Pol II-Ccr4-Not, using cryo electron microscopy (cryo EM) and interaction assays, complemented by a mRNA interaction studies.
1.2 Results
When I started this project, cryo EM of Pol II complexes typically yielded low to medium resolution structures of 15-20 Å, and high resolution structures could only be obtained by X-ray crystallography, which does not work for complexes where good quality crystals cannot be obtained. I started my project by optimizing the cryo EM workflow for Pol II samples on all levels; including specimen preparation, data collection, and computational analysis including 3D structure solution.
First, since Pol II is a rather small, low contrast molecule compared to typical “good” EM objects like viruses, I tried obtaining very thin ice layers on EM grids improve image quality. The best way to do this for our original microscope setting with a Falcon 2 detector was to use freshly prepared carbon-coated EM grids, along with longer glow-discharging (to improve surface wetting) and long blotting times at high air humidity. For a newer microscope setting with a K2 direct detector, carbon-free EM grids yielded even better results. Next, data collection was optimized to get a trade-off between highest possible number and highest possible quality of images. For this, semi-automatic data collection proved ideal. The user first determined the best grid areas suitable for data collection and the actual spots for later automated image acquisition. This yielded more than 90% good quality images, and typically 0.2 -1 million particles for one microscope session. Best results for particle image extraction from raw micrographs were produced with half-automated reference-based picking, followed by manual checking through all electron micrographs, and finally through all extracted particles.
For data processing, three different software packages were compared: IMAGIC, SPIDER and RELION. From all trials, RELION emerged as the by far best software for Pol II complexes, yielding well-defined, high-resolution Pol II structures without reconstruction artefacts to which the other programs were prone in case parameters were not known and not yet ideal. In order to speed up image processing, my host lab invested heavily into new computing hard ware so that I had access to an extremely powerful computing infrastructure.
Together with my colleague Dr. Martinez from the Cramer lab, I employed my new cryo EM workflow on the structure of the Pol II-capping enzyme complex and in parallel, on a high resolution structure of a transcribing yeast Pol II elongation complex. Capping enzyme catalyzes the first two co-transcriptional enzymatic steps leading to the 5’ cap in eukaryotic mRNAs, which is essential for mRNA splicing, export, translation, and stability. The yeast Pol II elongation complex primarily served as a test case for high resolution EM studies with Pol II, which at the time were not available. We were successful in solving both structures, and elucidated the mechanism capping enzyme uses on newly produced mRNA. This very successful work was published in 2015 in Molecular Cell (Martinez-Rucobo, Kohler, et. al., Mol Cell 58, 2015). It also proved the general feasibility of high-resolution EM structure determination for Pol II complexes using the workflow optimized for the Pol II-Ccr4 project.
To analyse the interaction between Ccr4-Not and Pol II in vitro, I first screened several interaction assays that would be applicable to a large number of different protein variants that I wanted to test. Using pull-down tests with biotinylated Pol II and Streptavidin coated magnetic beads produced best results. I tested various Ccr4-Not sub-complexes for Pol II binding and found that the binding interface is likely located in the nuclease module of Ccr4-Not consisting of the subunits Caf1, Ccr4, and the Not1 N-terminus. Constructs containing these subunits were then used for EM analysis. I also tested the effect of different potential substrate mRNAs on Ccr4-Not binding. Since Ccr4-Not can interact with mRNA, our aim was to find a good substrate mRNA that would lock the Pol II –Ccr4-Not complex into one defined three-dimensional structure, thus suitable for high-resolution structure determination by cryo-EM. However, none of the tested mRNAs improved binding between Ccr4N-Not and Pol II. Further research for better suited substrate mRNAs is thus planned.
The cryo-EM protocol that I generally established for Pol II complexes (see above) was also applied to Pol II-Ccr4-Not complexes. Unfortunately, none of the tested samples showed interpretable density for Ccr4-Not on the polymerase. Most likely, this is due to either sensitive binding that does not survive EM conditions, or to flexible binding, which would lead loss of signal for flexible parts of the structure in the EM reconstructions. Since fellow researchers working on Ccr4-Not complexes with cryo-EM reported similar difficulties, we decided to postpone further cryo-EM studies of Ccr4-Not complexes until more biochemical and functional data were available to allow the design of better suited, structurally more defined Pol II- substrate mRNA- Ccr4-Not reaction intermediates.
When the original plan to solve the high resolution structure of an active Pol II-Ccr4-Not complex turned out to require a longer time frame than feasible within the IEF time frame, another project opened up where again control of gene expression from transcription to translation would be in the focus. It had recently been shown that in bacteria, transcription and translation were coupled. An expert lab specializing on bacterial transcription (University of Wisconsin, USA) offered to work on the functional aspects of the coupling, as well as providing all necessary protein constructs, which would allow quick progression of the project. Considering my previous expertise on translation complexes, and my newly acquired knowledge about transcription complexes in mind, we decided to give this project idea a short trial period to decide if it had chances to lead to high-impact results and a publication within the remaining time frame of the IEF fellowship.
Indeed, in very short time I succeeded in reconstituting a transcription-translation complex and first EM images showed an RNAP-shaped density in a defined position on many particles. We were thus ensured that we had embarked upon a very promising, and highly competitive, project, that when completed, would contribute immensely to the understanding of the linkage between the two essential mechanisms of gene expression. We decided to proceed with the new study and for the moment, to postpone Pol II- Ccr4-Not studies. Our aim was to ensure best deliverables within the timeline of this IEF project, such as another high-profile publication.
By now, I have solved the cryo-EM structure of a transcription-translation complex. Using interaction assays, I identified regions of the RNAP that are central to the interaction, and our collaborators in the US analyzed the in vivo importance of the coupling. From our data, we could deduct a mechanism for direct gene expression control by coordinated transcription and translation. Furthermore, we discovered links to the control of transcriptional pausing and termination, and found that transcription-translation coupling is a conserved feature, which coordinates the two central mechanisms in gene expression, thereby adjusting both to environmental clues and cellular signals. A manuscript will soon be submitted for publication in one of the highest profile journals in the field.
1.3 Socio-economic impact
Establishing a high-resolution cryo EM protocol for Pol II complexes for the original project has paved the way for many other structural studies of multiprotein Pol II complexes. When I started the project, no high resolution EM structures were available for any transcriptional complex. By now, our lab alone has published several important studies featuring high resolution EM structures of RNA polymerase complexes. All those structures have been analysed by very similar protocols to the one I developed and which served as a proof of principle for the feasibility of such projects.
The work Dr. Martinez and I did on the active Pol II - capping enzyme complex represents an important mile stone in the field. The high profile open access publication allows fellow researchers to design future studies based on the molecular structure and mechanism presented by our lab. Since mRNA capping is a central step in the life cycle of most mRNAs, not only the field of transcription but many related scientific areas can profit from our data and our results are relevant to a wide audience.
Our most recent work on the coupled transcription-translation complex will be of huge impact to both the fields of transcription and translation. Since the link between those two crucial mechanisms seems to be evolutionarily conserved, our study will be the foundation for broader research. More functional studies characterizing the in vivo effects of this coupling can now be designed. Our research will be of exceptionally high relevance for a large scientific public and also of interest to the general public. See also: http://www.mpibpc.mpg.de/cramer