IGBMC, Strasbourg, France
My lab is interested in how transcription of DNA to RNA, the first step in gene expression, is regulated. In all domains of life, transcription is carried out by an enzyme called RNA polymerase. During transcription, RNA polymerase tends to pause regularly and this is used to modulate transcriptional output. One way to pause is through a process called backtracking, where RNA polymerase goes into reverse gear and extrudes the newly synthesised RNA transcript out of the active site, halting the catalytic cycle.
RNA polymerase tends to incorporate a wrong base every now and then, and this also results in backtracking. To resume transcription, RNA polymerase can catalyse RNA hydrolysis to cleave the backtracked portion and align the newly generated, shorter transcript with the enzyme’s active site. In the case of a misincorporation, this has the added benefit of removing the erroneous base. However, this cleavage reaction is not very fast. Protein transcription factors such as bacterial Gre factors or eukaryotic TFIIS accelerate the reaction.
The aim of our project was to use single particle cryo-EM and take snapshots along the entire reaction pathway beginning with a backtracked complex through Gre factor-assisted cleavage and followed by substrate binding to a reactivated transcription machinery. Getting microscope time on high-end instruments is difficult because demand is high. We heard great things about the setup at EMBL and applied for microscope time through iNext. We visited EMBL to collect data on the reactivated complex. We formed a complex using a functional RNA polymerase but with a modified RNA transcript lacking the 3’-OH. This allowed us to add the next cognate substrate as well as one of the Gre factors. But, since the RNA cannot perform the chemical attack, catalysis cannot occur and we trap the reactivated state before RNA extension.
The experience at EMBL was as good as it can get. It was obvious that Wim Hagen and Felix Weis know what they’re doing and it’s a pleasure to work with them. The setup is geared towards efficiency. They both try to look at each step, no matter how small, and try to figure out if it can be done in a better and more efficient way without sacrificing quality. Wim is also in touch with software developers to get the most out of SerialEM, the program used to control the microscope. Finally, he made changes to the hardware of the microscope to really maximise the amount of data we could collect. My understanding is that these changes will now be implemented in the next generation of Titan Krios microscopes.
We were able to obtain two reconstructions from the dataset collected at EMBL after in silico classification. The first one showed the bound rNTP substrate in the active site. This causes a conformational change incompatible with Gre factor binding. As a result, we observed loosely bound Gre on the surface of RNA polymerase. The second reconstruction showed Gre factor accessing the active site, but this is incompatible with the bound substrate. In other words, from a single dataset, we could directly visualise RNA polymerase in two states. For me this was another selling point for single particle cryo-EM and proof of the quality of data we obtained.
The biggest impact is clearly the quality of the data. The insights we got from the reconstruction were an important aspect of the story we told in our paper in Molecular Cell. The resolution was high enough to visualise the bound substrate in the active site, and a single dataset allowed us to directly visualise two mutually exclusive states – the reactivated, substrate-bound state of RNA polymerase and the Gre factor-bound state.
There are several reasons for the uniqueness of the setup. First of all, the software and hardware tweaks that I have mentioned earlier. Software tweaks are carried out at other sites as well, but the changes to the microscope hardware itself are pretty unique as far as I know. At least I have not seen them anywhere else so far.
The microscopes also appear to be in a superb state. Only minimal alignments were necessary to get going and we actually spent most of our time discussing collection strategies with Wim and Felix while the microscope was working through scripts to collect a grid atlas or take images of regions of interest.
Another aspect I really like is the fact that you can use SerialEM offline. You can select a small set of holes and the microscope can be started to collect data on those positions. You can then use SerialEM again, take your time and really try to select the best areas on your grid without having to worry about losing valuable microscope time. When you’re finished, data collection gets stopped briefly to merge your list of holes with the initial, short one and that’s it.
All of that affects the throughput, the amount, and the quality of data you can collect. Use of these instruments is expensive and finding ways to maximise data collection speed is a big win for all of us.
Finally, I also appreciate the lack of bureaucracy. Access is super easy and very well organised. You can stay overnight, there are parking spots on site that you can use, you can drop your storage dewar in front of the building that houses the microscope, and so on. These may seem like small details but it contributes to the overall positive user experience.
I would say, “Don’t think twice, bring the best and most interesting sample you have, and go for it!”
Absolutely – in fact we have come back since the time I’ve talked about. The quality and amount of data we obtained is difficult to match at any of the sites I know.