Cryogenic electron microscopy (cryo-EM) provides impressive near-atomic resolution for determining biomolecular structures, earning its creators — Jacques Dubochet, Joachim Frank and Richard Henderson — the Nobel Prize in Chemistry in 2017. Cryo-EM has become an increasingly popular as a complementary method to X-ray crystallography and nuclear magnetic resonsnce (NMR) spectroscopy, having the benefits of smaller sample size requirements and the ability to observe structures in their native state, without the need for crystallization.
Stephen Brohawn, an assistant professor at UC Berkeley's Department of Molecular and Cell Biology and Helen Wills Neuroscience Institute, recently served as senior author on a study into a unique membrane channel protein expressed by the SARS-CoV-2 virus, using cryo-EM to determine its 3D structure and investigate it as a potential drug target. Speaking with Editor-in-Chief Michelle Taylor, Brohawn discusses the advantages of this technique in analyzing membrane proteins, such as the ion channels encoded by SARS-CoV-2.
Q: I can imagine you and your lab had to pivot very quickly once SARS-CoV-2 became an obvious problem. What were your first few action steps in order to switch gears to COVID-19 research last year?
A: Scientifically, right at the beginning of the pandemic when it became clear that this was a new and concerning virus, we started to look at its genome and read work on related coronaviruses. A great postdoctoral scientist in the lab, Dr. Ben Sorum, became interested in three proteins with suspected ion channel activity. We thought these were interesting proteins to study that could potentially serve as vaccine or therapeutic targets, so we ordered genes encoding the channels and started to work on characterizing their function and structure in parallel.
Logistically, we spent a lot of time thinking about whether the potential benefit of our work was going to outweigh the risks of coming into the lab early on in the pandemic. Once we decided we could contribute something important, we worked with the University to make sure we could execute the project safely.
Q: What has come of your pandemic research? Can you detail your work on the SARS-CoV-2 membrane protein?
The structure of the SARS-CoV-2 protein called 3a, as determined by cryo-EM. Credit: UC Berkeley image courtesy of David Kern
A: From the literature on other coronaviruses, we saw that there were two or three putative ion channels that the virus encodes. One of these channels is called ORF3a, or just 3a, and it has been shown that deleting it from the related SARS-CoV-1 virus reduces viral maturation and morbidity in animal models. It has recently been shown to be similarly important in SARS-CoV-2. This looked like a promising and understudied target, so we investigated 3a in two ways, studying both its function and structure.
In just a few short weeks of very intensive work, another very talented postdoc in our lab, Dr. David Kern, worked out a way to purify the SARS-CoV-2 3a protein, reconstituted it into lipid nanodiscs, and determined its structure using cryo-EM. We were already able to generate a fairly high-resolution reconstruction on our equipment here at Berkeley (2.9 Å). After posting a preprint describing the structure, Dr. Abhay Kotecha from Thermo Fisher Scientific reached out to a graduate student in my lab, Christopher Hoel, who had worked on the cryo-EM data processing with Dr. Kern, to see if we would be interested in collaborating to push this resolution even further. Needless to say, we were very excited to try.
We took our extra grids, shipped them to the Netherlands, and crossed our fingers. Using the data collected at Thermo Fisher, we were then able to generate reconstructions to 2.1 Å resolution, which is really quite remarkable, especially given the small size of this target (62 kDa).
Ultimately, the work we, and so many others in the scientific community, did studying these channels contributed to our fundamental understanding of the virus, helping future vaccine or therapeutic development.
Q: Regardless of the virus, membrane proteins tend to be a challenging target for traditional structural analysis. Why is that?
A: Membrane proteins not only dictate how our cells communicate and function in unison but are also the key that viruses and bacteria use to gain entry to our cells. This class of protein, however, has been a challenging target for traditional structural analysis for a number of reasons. Membrane proteins are typically expressed at low levels, which makes purifying quantities required for structural studies difficult. Purifying membrane proteins presents extra challenges compared to soluble proteins because their hydrophobic surfaces that interact with cellular membranes need to be shielded during extraction and purification. One typically uses a detergent for this, but finding the right combination of detergents that keep membrane proteins stable often requires a lot of screening. Even in the best cases, membrane proteins are often relatively unstable and require time-consuming and difficult engineering to improve stability. Finally, many membrane proteins are dynamic in interesting ways related to their function, but conformational heterogeneity makes structure determination hard, for example by lowering the chances molecules can make ordered regular contacts with their neighbors in a crystal lattice.
Q: Given these difficulties, what makes cryo-EM an ideal analysis technique?
A: Cryo-EM offers a number of advantages over other structure determination methods like X-ray crystallography or NMR. One typically requires much less sample, so the low yields of membrane proteins present less of a barrier. Cryo-EM is very well suited to analyzing flexible proteins because particles can be computationally sorted into groups with similar shapes and analyzed separately or subjected to more sophisticated computational analysis of conformational heterogeneity. Aside from the challenges mentioned previously, we also prefer to look at membrane-protein structures in conditions that are closer to their native cellular environment. This makes cryo-EM very appealing because we can reconstitute membrane proteins into lipid nanodiscs that mimic the cell membrane much more closely than a detergent micelle ever could. Rapid technical advances in hardware and software for cryo-EM have recently allowed us to see even small and very flexible membrane proteins at high resolution – meaning we can analyze more difficult samples with single particle analysis.
Q: Since winning the Nobel Prize in 2017, I feel like cryo-EM has definitely become more mainstream. Have you experienced that among your colleagues and others, as well?
A: Absolutely. Work in my group is certainly part of that shift towards using cryo-EM as the preferred method for structural biology. I had used X-ray crystallography in my training in graduate school and as a postdoc, but watching the field develop made it clear that when I started my lab in 2016, we would focus largely on cryo-EM instead. I think it is a real testament to leaders in the field who have prioritized open software, communication, and community building that we have been able to have success making this shift.
Q: What do you see as the future of cryo-EM?
A: The field is developing rapidly, and it’s very exciting to be a part of it. Today, we’re able to see small membrane proteins and very flexible proteins at high resolution. But just a few years ago, we would have never imagined that these kinds of projects would be possible. We can now analyze more difficult samples with single particle analysis. We can even think about doing drug discovery with cryo-EM, where you need high-resolution structures to interpret the chemical interactions between small molecule drugs and proteins. Moving in the other direction in terms of scale, there are a lot of exciting developments using tomography of intact cells to study the structures of proteins and their interactions in situ. This will certainly be an important future direction in field as well.
Q: What research projects are you focusing on right on?
A: My lab studies the molecular basis for sensory transduction and electrical signaling in the nervous system. We are very interested in figuring out how ion channels in neurons that generate electrical signals work at a molecular level.
One of our recent projects was focused on a potassium ion channel called TASK-2. This work has also benefitted from new cryo-EM instrumentation from Thermo Fisher including a Falcon 4 Camera, the Selectris X Imaging Filter, and the E-CFEG. TASK-2 is important for a number of physiological processes and one of these is in regulating breathing rate in response to changing CO2 levels in the blood. CO2 levels impact blood pH, with higher CO2 levels meaning more acidic blood just like carbonating drinks makes them acidic, and TASK-2 activity is regulated by pH. We wanted to use a structural approach to understand how these changes in pH open and close the channel, down to the atomic-scale rearrangements that make it happen.
To do this, we solved cryo-EM structures of TASK-2 at different pH values, and what we saw is that, compared to other ion channels, protons inhibit TASK-2 in two totally new ways. We tried doing this with crystallography previously because TASK-2 is a very small membrane protein (~65 kDa), which puts it at the low end of what is feasible for cryo-EM. Even a few years ago, I would have said that it’s probably not going to be possible to analyze proteins this small anytime soon. However, another exceptional postdoc in my group, Dr. Baobin Li, in collaboration with a graduate student Robert Rietmeijer, decided to give it a shot. We worked out ways to use lipid nanodiscs to determine TASK-2 structures at ~3.5 Å resolution and have since improved the resolution of these structures to ~2.5 Å resolution using data collected by Dr. Abhay Kotecha at Thermo Fisher to reveal additional insight into channel function. Cryo-EM was essential to this project in which structures of the channel in different states provided the key insights into the mechanistic underpinnings of how this channel is regulated.