Uncovered: Record-setting laser boosts cryo-electron microscopy

In this episode of C&EN Uncovered, host Craig Bettenhausen speaks with C&EN senior editor Laurel Oldach about a laser phase plate, a device that boosts the contrast in electron microscopy using the most intense laser beam yet made, and the story of its path from an idea to an invention that could open new lines of inquiry in structural biology and biochemistry. You can find the link to the article here.

Subscribe to Stereo Chemistry now on Apple Podcasts, Spotify, or wherever you listen to podcasts.

Executive producer: David Anderson

Host: Craig Bettenhausen

Reporter: Laurel Oldach

Video and audio producers: David Anderson, Jeremy Barr

Episode artwork: Biohub

Music: Commercial Flow, Shutterstock

Contact Stereo Chemistry: Contact us on social media at @cenmag, or email [email protected].

The following is a transcript of the episode. Interviews have been edited for length and clarity.

Craig Bettenhausen: Welcome to C&EN Uncovered. I’m Craig Bettenhausen. C&EN Uncovered is a podcast series from Stereo Chemistry. In each episode, we’ll take another look at a recent story in Chemical & Engineering News and hear from C&EN reporters about striking moments from their reporting, their biggest takeaways, and what got left on the cutting-room floor. In this episode, we’ll be taking a deeper look into a recent breakthrough in the field of structural biology, specifically an ultrabright laser technology that helps shift the phase of electrons. We’re here with C&EN editor Laurel Oldach, who wrote the article. We’ll put a link in the show notes along with the episode credits. Hi, Laurel.

Laurel Oldach: Hi, Craig. Thanks for having me.

Craig: So for those of you that haven’t had a chance to read the article yet, what’s it about?

Laurel: It is about bringing a technology called “phase contrast from light microscopy” into electron microscopy so that researchers who are using this very powerful but often quite blurry, low-signal-to-noise-ratio technology to do characterization of biological materials, whether that’s individual proteins or whole sections of tissue, can get stronger images that they can then use to get a better picture of what’s going on in our complex molecular chemical environments.

Craig: So how did you find this story?

Laurel: It was pitched to our colleague Fionna Samuels. It’s a story that started out at UC [University of California] Berkeley and the national lab there, which I think is Lawrence Berkeley National Lab, but then was picked up over at Biohub, the Chan Zuckerberg Initiative–founded research institute in San Francisco. And Biohub has this crackerjack PR team that said, “This sounds like a C&EN story.” Fionna was not able to take it on at the time, and so she handed it off to me because I’ve also done some coverage of cryo-EM [cryo-electron microscopy] in the past.

Craig: Cryo-EM—how cold are we talking for this suite of technologies?

Laurel: That is a great question. I know that you’ve done some liquid-helium-shortage coverage in the past, and so I was briefly interested in, Is this a thing that uses helium too? It’s not quite that cold. It’s all under vacuum and uses liquid nitrogen. So I do not remember exactly how cold that is. It’s cold, but it’s not as cold as liquid helium or as NMR.

Craig: A bit of inside baseball, but I’m aware that the shape and the scope of this story changed while you were reporting it. Tell us about what happened there.

Laurel: Initially this was supposed to be a short news story. And then I got on the phone with the physicist who had initially theorized back in 2010 the way to introduce phase contrast to a beam of electrons without also introducing a lot of variable charge, which is basically, you hit the beam of electrons with a beam of photons, and you let quantum interactions do the work, and no charge accumulates. However, there’s a lot of reasons that that’s very difficult to achieve. And so he had spent basically the past 15 years (1) trying to show that it could work, and (2) trying to get funding to get it to work, which is where Biohub comes in. After spending about 45 min hearing from him about the work that he had done and just all the effort that had gone into it, and after realizing how little was out there in the general science press about this project, which really just came to fruition fairly recently, it didn’t feel like a news story.

It felt like a feature story. And so I pitched to my fabulous editor, Laura Howes, that we should see if we could make it a longer treatment of the same subject but with more of a historic view. And she was on board fortunately for me. So that’s what we did.

Craig: So in this device, we’re shooting a laser at an electron beam to get better images of biomolecules.

Laurel: Exactly. You can get better images of all kinds of stuff, but biomolecules are relatively low contrast because the biomolecules don’t interact very much with a beam of electrons compared to other stuff that you might image, like amorphous carbon. That has a really high-contrast image. A biomolecule doesn’t really interact that much. So you get this kind of grainy image. It looks like you’re kind of peering through a blizzard at maybe your protein or whatever. And so for the longest time, cryo-EM was not practical at all for biological imaging. People would do electron microscopy of fixed samples that had some kind of higher-contrast thing put over them, like gold. About a decade ago, there was something called the resolution revolution where they combine many, many images of the same thing from multiple angles, whether that’s many images of protein structures that you expect would be identical, or you can take a section of a cell and sort of tilt it and collect many images that way.

People combine many different images and sort of average them together to get a higher-resolution image.

Craig: Now this laser phase plate, did this start with a biochemistry problem that needed solving or a physics phenomenon in search of a use case?

Laurel: That’s a great question. What I can tell you is that phase contrast is taking—I think I mentioned before that biological samples don’t interact all that strongly with electron beams. However, to the extent that they do interact with electron beams, the electrons that do interact get a little bit scattered, and that means that they are a little bit out of phase with the majority of the beam, which passes through as if the sample were transparent. This is something that also happens with visible light. However, back in the 1940s in some Nobel-winning research at that time, people figured out a way to fix it so that, you can imagine, you have your sample, it’s flat, you have your beam go through, whether it’s an electron or a photon imaging beam. And some of the beam is in phase, some of the beam is out of phase.

Phase-contrast microscopy basically introduces what’s called a phase plate, something that interacts with the in-phase and the out-of-phase light a little bit differently, or that interacts with light according to what phase it’s in, I guess is a way of saying it. And by doing that, it introduces interference, something . . . something quantum, and you end up with a higher-contrast image, where it basically converts the phase difference into a number of photons difference. So that was solved in the 1940s for light microscopy. However, it’s much, much easier to figure out something that interacts stably and without getting completely destroyed with a beam of photons compared to with a beam of electrons. So people have known for decades that electrons, like photons, have this sort of phase change after they pass through a biological sample, but they have this blurriness and this low-signal-to-noise-ratio problem.

People have not, until now—There have been several different kind-of-interesting but ultimately dead-end efforts to introduce some physical material into the electron beam after it passes through the sample. Apparently somebody figured out a way to do it with spider silk back in the ’80s or ’90s. There’s an earlier-generation laser phase plate that is called the Volta that was an actual physical thing that you stick into the electron beam. However, the problem with physical materials is, like I said, (1) they get sort of degraded over time by being constantly bombarded with electrons, and (2) they can accumulate charge themselves, and then that changes. So they’ll interact in a nonstable way. They have to calibrate. And then in the process of calibration, you’ve introduced some charge to whatever your object is, has some static on it, and now it is different than what you calibrated.

So as a result, you end up with better contrast but not for a long-enough period that you can actually use the images, if that makes sense. So instead, what Holger Müller, the physicist at UC Berkeley, kind of suggested about 15 years ago was, “Well, what if you use this property of how photons and electrons interact to, instead of shifting the phase with a physical material, you could use a light material.” And that sounds really great, but you need the brightest laser in the world. It needed to be so bright that no one had ever built such a bright laser. And that was what he concluded at the time. And he was like, “Well, I think it ought to be able to work” and started working on it back then. The way that Müller concluded you could make this ultra-super-duper-bright laser was by shooting a laser into a spherical mirror, and the laser goes in, and it bounces off the spherical mirror, and then it bounces back, and then it bounces back.

So there’s some amount of bouncing back, back, back, back, back. But with each time that it comes back, it’s collecting more incoming light such that you’re essentially adding intensity until it gets to a really, really bright equilibrium.

Craig: How big of an object then is this laser-phase-plate device?

Laurel: I’m told it’s not enormous. It’s maybe an inch square or something like that. I mean, the mirrors are tiny but flawless because to keep the photons all vibing together, the reflective index of the mirrors has to be near perfect. So they have this wild . . . three different companies, that this one makes the mirrors, and then this one finishes the mirrors, and then this one applies this fine reflective coating. It was really a material story. They have this complex supply chain on the mirrors, and if the mirrors get a grain of dust on them, the dust of course incinerates because it’s being blasted by a laser that gets up to brighter than most things in the solar system except certain parts of the inside of the sun, and it just ruins the mirrors as well. So it’s a very, very sensitive system.

Craig: A lot of scientific lasers fire in pulses, and then there’s some continuous beams. Which type of laser is this, and what difference would that make for this kind of work?

Laurel: That is such a great question, and I’m really glad you asked because this technology, the laser phase plate that we’re talking about, is a continuous laser beam. And so that’s part of what makes it so difficult is because it’s hard to get a really bright continuous laser. And so that’s why it has to be bounced back and forth a bazillion times before it reaches equilibrium.

There is also a group at, I think, Columbia [University] led by Anthony Fitzpatrick that’s working on a different way of doing this, which uses a femtosecond laser, which is much easier to make really, really bright. And there are some advantages to this, but they haven’t yet installed it in a cryo-electron microscope. The disadvantage about having a really, really bright femtosecond laser pulse is then, in order to get the benefit in terms of contrast, your electron beam also needs to be a femtosecond.

And one of the folks I spoke to, David Agard, described it as trying to hit a bullet with a bullet, getting your electron beam and your laser beam to hit both at the same time so that you’re not wasting a lot of essentially electron images of your sample without adding the contrast.

Craig: So you mentioned that this laser is one of the brightest things in our solar system, the brightest laser ever. I have several questions about this laser beam. Okay.

Laurel: I’ll answer to the best of my ability.

Craig: What color is it? What wavelength did they end up using for this system?

Laurel: Oh, good question. I meant to say that. Cyan.

Craig: Cyan. Okay.

Laurel: Yeah, because it’s a narrower wavelength, so it makes it easier to focus it in on the appropriate part of the electron beam.

Craig: And then can you give us a sense of how much energy that kind of beam would contain? What would happen if you stuck your finger in it?

Laurel: Oh, nothing good. Don’t do that. I actually had to ask my colleagues for some help on this because it’s just so much energy. And we ended up comparing it to—and this didn’t, I don’t think, make it into the story, but we ended up comparing it to—it’s gigawatts of energy. So it’s comparable to a few medium-sized power plants, but it’s a very, very small area. So it’s gigawatts of energy per, I think, square centimeter, maybe per square millimeter. So it’s not necessarily the case that the laser is producing a power plant worth of energy. It’s just that if you stuck your finger in there, the intensity is that bright.

Craig: So the point of this is it’s an imaging-type science. What do these images look like, and what do you do with them? I assume it’s not just a pretty picture.

Laurel: It is not just a pretty picture, but you do also get a very pretty picture. So what the images look like is essentially—It depends on what kind of imaging you’re doing. If it’s single-particle analysis, you end up with a field that kind of looks like a polka-dot field, but each of the polka dots is a little bit different. And because each polka dot is, say, a protein or in some cases something bigger than a protein, like a protein complex that is in a slightly different orientation, like we talked about before, then you can look at these different orientations, and you can sort them, and then you can average across many different iterations of the same thing to get a much-finer-grained image than the initial raw material.

Craig: It’s an image you do math on more than an image you look at.

Laurel: Yes. Although with the laser on in a laser-phase-plate microscope, you do get a better image. It’s still an image that you’re going to want to do some math on. So again, this is for the sort of single-particle analysis. You can use these images in the aggregate; you collect particles by the thousand. Each image will have a good number of particles in it. You can use these images to determine the 3D structure of a protein. The reason that cryo-EM has had sort of a revolution in the past decade or so in biology is because you can use it to get the structure of a protein that is not crystallizable. So back in the day, protein structural biologists were really limited. You had to either have a protein that you could get to form a crystal structure, or you had to do NMR.

And NMR is great. I know that you’re very into NMR grades, so I don’t want to say anything bad about it, but for proteins in solution, it gives you a vague image of the outlines that your protein might be occupying. These days, for proteins bigger than a certain cutoff size, that’s like a mid to large protein; you can use a cryo-electron microscope to get a fabulous structure. Below that size, people are still using NMR for now. And for certain types of experiment, like for drug discovery, for small proteins that are binding to a small molecule, you want NMR for that still. However, with this advance, researchers should be able to get better cryo-EM structures of much smaller proteins than they could before. And that’s because if a protein is really, really small, it’s not scattering as many electrons because it’s just a smaller target. So you need better signal, and the laser phase plate gives you a better signal.

So that’s an exciting potential application for drug development. But the other thing that most of the researchers that I spoke to are really, really excited about is not single-particle analysis, but you can do more-complex structural analyses. Everybody’s seen in high school biology textbooks, like, an electron micrograph of something in a cell, and that’s usually something that is fixed and then potentially either coated with metals or labeled with antibodies with gold particles, or otherwise somehow there’s some contrast kind of layered in on top of the biology. The expectation here is that with the stronger signal that you can get from laser-phase-plate-induced cryo-EM, you can get better contrast, and therefore you can look without having to put extra stuff in on top of your biological sample. You might be able to even do proteomics of the inside of a cell. How is this protein that’s adjacent to the mitochondria different than a different copy of the same protein that’s over there, far from the mitochondria, and maybe adjacent to the cell membrane or something like that?

Craig: So zooming out a little bit, what does this invention, this laser phase plate, change for biochemistry and molecular biology?

Laurel: So I think thing number 1 is the ability to get a noncrystallized solution structure for smaller proteins. I think that’s a big deal. I think taking that ability, the solution structure—by which I mean it’s in solution, it’s not in a crystal—taking that into cells is the really, really exciting thing. Being able to look at a cell and say, “Okay, it used to be that I could only pick out really big stuff like ribosomes or cytoskeletal elements or whatever.” Being able to instead say, “Okay, I can reasonably estimate more than half of the proteins that I’m looking at here and say something about what shape they’ve adopted as well,” that’s going to be huge. It really fuels into this wave of spatial biology—that is, a lot of different groups are pushing on a lot of different technologies to look at chemical analysis of biology through imaging in a way that is closer to the unaltered, original biological state in which it’s found.

And the closer we can get to unaltered tissues, the more biologically relevant our findings are going to be. But at the same time, with techniques like this, the more chemically precise we’re going to be able to be. So I think it’s a really, really exciting time for imaging-based approaches to chemically explaining biology.

Craig: You mentioned there are two of these laser phase plates exist today, but there are a lot of electron microscopes out there. How broadly useful could this technology become? Is it going to be in every electron microscope, a chicken in every pot?

Laurel: Great question. The people that I spoke to, well— So one thing is that the second iteration of it—So Holger Müller built his one at UC Berkeley. The second one is at Biohub, and it uses a somewhat different scheme. It has two less-bright lasers that cross beams so that you get the same amount of intensity but without needing quite such impeccably perfect optics. David Agard, who’s at Biohub, said to me, “Sometimes people invent something, and only the inventor can get the thing to work perfectly that has this exquisite technical requirement. It only works in one person’s hands or one lab’s hands.” That’s demonstrated to be not the case here. There’s a second one up and running, and people are working on it and trying to see how it works. Will it be widespread? I think the question is, Will labs be able to afford it?

Because the cost is going to be astronomical. It might double the cost of a $6 million microscope to build a laser phase plate into it. And that’s from someone that I spoke to for the story. I don’t have an estimate on the cost of the newer cross-laser phase plate, and some other versions of the same basic principle, like the femtosecond pulse laser is less expensive, but also that one hasn’t been demonstrated to work. So will it be installed in every new electron microscope that ships? Probably not. How widespread will it be? I mean, I think that’s to be determined, and there’s a possibility that it will be something akin to a synchrotron, where you book your time on the laser-phase-plate-enabled cryo-electron microscope. But I think there were also questions, like I said, 15 years ago when the resolution revolution happened, and structural biologists at the time that cryo-EM was first rolled out were saying, “How can we possibly afford this?”

And Anthony Fitzpatrick, who’s working on this other type of cross-phased laser plate, sort of said, “Well yes, we all asked each other that, but then the other question was, How could we afford not to do it, because it’s such a potentially transformative technology?” And so yeah, I think the field is wide open. I think everyone I spoke to agrees this is going to be a big deal, but how it gets rolled out exactly, particularly in this funding environment here in the US, is a really big question.

Craig: And so you talked to Holger Müller for this article. How are they feeling about having somebody else replicate it? Because I can see both being very proud and also feeling some competition or jealousy if someone else is building another version of your invention. How are they taking that news?

Laurel: I think that he was a coauthor on the paper, the preprint from Biohub. I think he’s worked really closely with the Biohub team. And I think, like I said earlier, that they kind of provided the funding that he needed to make it clear that this could work in a state-of-the-art microscope. So I don’t think that there’s any animosity there. I think that it’s a really productive partnership. He’s also working with—I mean, he’s a physicist, so more focused on working out the physics of how this thing works. And actually something that’s kind of interesting that I heard while I was reporting this story is that there are some people who are much more interested in doing the physics and saying, “Okay, well, this would work if you had a bright enough laser.” There’s some people who are interested in doing the engineering and saying, “Okay, what do we need to take this bright laser to the level that it actually achieves what we want it to achieve?”

And then there’s some people who are interested in doing the applications and saying, “Okay, what biological questions can we ask using this really bright laser?” But we don’t know really. Like, we don’t know exactly how the interaction between the electrons and the laser is happening, but we know that it makes for these amazing images. And so one of the cool things about reporting this story was just getting to speak to people at every phase of that research ecosystem about what has turned out to be a really interdisciplinary kind of a project.

Craig: So what’s next for this line of work?

Laurel: They are pushing hard to get it into cryo-electron tomography, which is the way that the visual proteomics work would be done. I think there’s questions about—So this was something crazy that I learned. Because of the low signal-to-noise ratio of cryo-EM, everyone collects their images a little bit out of focus. So they’re working on getting closer to true focus, because you should be able to do that better with better phase contrast. There’s a few different technical things that they’re kind of pushing on. And then Biohub just announced, or I forget if it was 400 or $500 million in funding for these kinds of spatial-biology projects toward a virtual cell, collecting data to train AI [artificial intelligence] to predict a cell.

Craig: Well, Laurel, thank you for diving deep into this with us.

Laurel: Yeah, thank you so much for letting me nerd out about optics. Yeah, I’ve had a great time reporting this story, and all the people involved seem to be really nice, which is always lovely.

Craig: So listeners can find me on social media as @craigofwaffles. Laurel, how can our listeners get in touch with you?

Laurel: You can find me on Bluesky at laureloldach, L-a-u-r-e-l dot O-l-d-a-c-h. I’m on LinkedIn too, if you’re more active there. And if you have any hot cryo-electron microscopy tips, or cold ones, and I also cover drug discovery and policy and a few other beats, then I’m on Signal, @laurel_oldach.07.

Craig: You can find Laurel’s story about this laser edge technology on C&EN’s website. We put a link in the show notes along with the episode credits. We’d love to know what you think of C&EN Uncovered. You can share your feedback with us by emailing [email protected]. This has been C&EN Uncovered, a series from C&EN’s Stereo Chemistry. Chemical & Engineering News is an independent news outlet published by the American Chemical Society. Thanks for listening.