This conversation with Dr. Loren Andreas, from the Max Planck Institute, delves into the growing use of NMR to study condensed systems and to complement X-ray crystallography in illuminating our understanding of structural biology. Paolo and Loren also talk about the expat experience and how science is truly an international field of study.
Since the elucidation of the DNA structure by James Watson and Francis Crick in 1951, the importance of understanding the three-dimensional structure of biomolecules has become obvious. Over the last few decades scientists have resolved the structure of thousands of complex biomolecules enabling incredible innovations in drug design, biological and medical sciences. X-Ray crystallography has been the key technique, but in recent years Nuclear Magnetic Resonance (NMR) has emerged as an additional, complementary approach. Dr. Loren Andreas explains to us how NMR has grown to be the technology of choice as it has expanded its field of application from liquid solutions to condensed systems. The discussion is a surprising discovery of how progress in engineering and instrument design has completely changed the landscape in structural biology. Modern NMR allows scientists to study molecules in complex systems, simulating more closely their natural environment, including interaction between them. This episode offers an exciting glimpse of the future, through a few examples from today’s science.
Visit https://thermofisher.com/bctl to register for your free Bringing Chemistry to Life T-shirt and https://www.alfa.com/en/chemistry-podcasts/ to access our episode summary sheet, which contains links to recent publications and additional content recommendations for our guest.
Dr. Loren Andreas 0:06
I think often, some of the more fascinating things happen almost by accident coming across new and unexpected and interesting things.
Paolo 0:16
Proteins are complicated molecules, and analyzing them can be incredibly challenging. It can help to think outside the box and be prepared for the unexpected. Lauren Andreas, a group leader at the Max Planck Institute for Biophysical Chemistry in Goettingen, Germany, has been doing just that, in its work with NMR. One of the few techniques that can characterize molecules at an atomic level. In this season one bonus episode of Bringing Chemistry to Life, we speak with another member of the Chemical and Engineering News' 2019, Talented 12 about their work and trends in their field. I'm your host, Paolo Braiuca, Senior Manager of Global Market Development at Thermo Fisher Scientific. We began by asking Dr. Andreas about his background, what took him to his current role, and how he's spending his time during COVID-19.
Dr. Loren Andreas 1:10
Goettingen in as a small town, it's 130,000 people or so. And it's a great place to have a family, I go everywhere on a bike. So this was also quite a good setup, when COVID hit, I can just get on my bike and be at the Institute in 15 minutes and come back by bike. So I'm not exposed to any anything on public transit. For NMR, specifically, Europe has made a much larger investment recently, in NMR equipment at several different centers, and big laboratories, which makes it a very attractive place to go for research. In the academic setting, there's really a now a global community, you see people at conferences from different countries all the time. And now with COVID, we see everybody online from all over the world.
Paolo 2:02
So let's jump into the science a little bit Loren. Of course, your personal development is fascinating, but so is your science. And so let's let's start speaking about NMR. So your research, the way I understand this focuses mostly on structural biology. So what modernd NMR is allowing you to do is looking into these big proteins, in general terms biological molecules, in in sort of different environments in something which is closer to their real native environment. So you can actually understand a bit better how they look like and how they function. Am I right?
Dr. Loren Andreas 2:45
Yeah, so my my own interest came originally more from the physics, an interest in Hamiltonians, and quantum physics. And then I made my way towards the biology and the biochemistry. And that's exactly right, that the technique is fairly complicated and fairly expensive. And so we are typically looking for places where we can have the most impact. And one of those is to to look at proteins in a more native, like environment. And what we see from membrane proteins is, in some cases, the use of detergents, which is fairly common with other techniques like crystallography, or cryo-EM. In many cases, these don't cause problems. But for some, the detergent really disrupts part of the structure. And it's not too surprising when you think of detergents in a everyday setting. They basically are stripping out grease, or cleaning things out and denaturing molecules. And so for some of these proteins, we really see a large difference between the structures. When they're done by crystallography by solution NMR, solid state NMR.
Paolo 3:59
for the sake of our listener, Lauren, you mentioned detergents, so I assume they are used to bring the protein in solution and to allow them to, you know, either probe being properly dissolved or creating a solution for then crystallizing the protein in a in a pure form. So to be able to run the X ray, X ray crystallography diffraction experiment, am I right?
Dr. Loren Andreas 4:20
Yeah, that's right. And we use detergents too. So we still extract the proteins, typically with detergents purify, but at the end, we have detergent free preparation with just the membrane. So we removed the detergents before we do the final analysis.
Paolo 4:36
So what you basically do is being able to run sophisticated NMR experiments of proteins in solution, but at the same time in our sort of condensed state, solid state of some kind.
Dr. Loren Andreas 4:52
That's why we call it solid state NMR. Although you're right, that it's more of a somewhat solid situation. In the case of membrane proteins, because there can still be quite a bit of fluidity in these membranes. So there can be what we call uniaxial diffusion, where the molecules might be moving in two of the dimensions, but stationary, relatively in third one, although most of our samples are much more towards the rigid side of things than the fluid or dynamic side.
Paolo 5:24
When I studied chemistry, that was half a century ago, you know, NMR, in all a solid material was unthinkable pretty much right? Maybe there were the first experiments, but the reality is that you really had to have everything in a deuterated solvent. And then, you know, you could run the experiments like that, are you guys do something which is much more complicated and sophisticated? And you know, and there's always when you read this paper, there's always the use of this, what they call the magic spinning angle, right? Which sounds like black magic to me. And, you know, I never really heard anyone offering a simple enough explanation, or why this angle, which is about 55 degrees, if I'm not wrong, you know, is is so magic, right? What why, what does he Why does it allow to, to obtain appropriate NMR spectrum out of condensed data? molecules? And can I can I try asking you, can you can you can you try and explain to non initiated, why this work and how time works?
Dr. Loren Andreas 6:24
I have two different explanations, maybe. So, one is that it just falls right out of the physics and the math, but maybe a more accessible explanation is that this is the angle, it's the long diagonal of a cube. And so this is the angle, which has equal representation of all three axes X, Y, and Z. This gives it this property that it averages the x, the Y, and Z. And that turns out to be quite useful to us in averaging the interactions that we have, which leads to a lot of more difficult issues in in the spectroscopy and solid state so that this leads to broader lines, and more difficult access to the information we're looking for.
Paolo 7:10
And how can you overcome the basic difficulties of these sort of broader lines and less resolved peaks in your in your spectrum?
Dr. Loren Andreas 7:21
A lot of so a lot of it comes to incremental improvements. And a lot of it comes to kind of clever tricks which people have been coming up with over the decades. And these days, we're excited to see another incremental improvement although the increment now it's quite a big one going to 1.2 gigahertz frequency. Whereas previously, 1 gigahertz magnets were the highest field magnets that existed. And so that's the kind of brute force way of getting both higher sensitivity, higher signal compared to noise, and also pulling apart the resonances, which in solid state then leads to narrower lines, so more ease in separating the lines and getting information out.
Paolo 8:10
And I have a question, which is probably a bit silly, right. So you really need to spin your sample very fast to be able to have results, all the different magnetic interactions that the nuclei see right? And while I can visualize that, for how we could work for a solution, and I can visualize that, for a proper solid material with a certain mechanical resistance, I'm struggling to see how you keep the morphology of sort of semi condensed state like lipidic bilayers or you know, sort of systems mean mimicking cellular membranes and stuff like that?
Dr. Loren Andreas 8:57
This can be an issue for some of the systems, some of the assemblies, for example, could be more sensitive to these kind of forces, which exists when we spend actually the spinning results in quite extreme G forces. If you calculate it gets into millions of G's. Although in pressures, which is probably more important thing than the G force itself, this is still a relatively mild pressure. So most proteins are not particularly perturbed the protein itself. But you're right, that the larger assemblies Can, can start to break apart under these forces. So one thing that can be done is to reduce the difference in density between the protein and other components, adding glycerol, for example, okay, use these kind of forces.
Paolo 9:49
So I'm assuming that it's not just a single experiment, and then you're confident that you've got what you need is probably running in multiple times, trying to control a lot of different variables. To ensure your system is macroscopically in the state you want it to be. And then it represents reality as much as possible. Am I right?
Dr. Loren Andreas 10:09
Yes, although screening by NMR is pretty slow, it's pretty time consuming to delicately pack the cylindrical rotor full of material and take it to the spectrometer and do a measurement. This takes quite a bit of time. So we can only really screen one or two samples a day on an instrument. And so when we do the initial preparation steps, we're making use of other analytical techniques to the best, best we can. So we might go to electron microscopy, negative stain electron microscopy to see the overall morphology of the samples. And we also make use of solution NMR in the earlier stages when we, for example, solubilized in detergent to track to track the situation. And for some of the membrane proteins, we can also see the difference between a folded active state of the protein and a denatured state just on a SDS PAGE gels very simple laboratory technique, which most people learn. And that's gel, gel shift, it's been noted for quite a while for some of these proteins.
Paolo 11:19
It's quite interesting what you're saying, you know, there's a number of things you can do. But the reality is that you depends on instruments that are becoming more and more sophisticated, and I'm assuming crazily expensive, you can only run a couple of experiments per day. That means the experiments are really expensive, and failure is probably almost not acceptable, even though you're still in a sort of semi academic environment. It sounds like there's a lot of responsibility for the people involved.
Dr. Loren Andreas 11:52
If we could reduce the cost of these instruments by a factor of 10, it would make absolutely an enormous difference in what we could do. That's true. And we're also using liquid helium for the magnets and liquid nitrogen. So the running costs are also fairly significant. What we'll typically do in that case, is start with a lower field magnet, which isn't quite as expensive, and start to characterize the samples, make sure that the samples are in shape, and then take them to the much more expensive magnet. So in the United States, there's some efforts now to get facilities going for the highest magnetic fields for the most expensive instruments. And then with this kind of model, it'd be possible to use the in house instrument, which isn't so expensive, characterize samples, and then once you're really sure you know, what's going on in sample, and you know which sample to take, you could go to a regional facility for, for final measurements. So I think that'll be a prospect. And it's true that the the cost really gets into the millions. And so it's substantial investment, although I should say, at the same time, comparable, well, competing techniques, let's say, like cryo EM are also in the millions. There's also a lot of opportunities, these techniques together, or for different applications. So cryo EM is fantastic for really large complexes, really rigid complexes, but it's very difficult to study much smaller proteins. And so that's a good balance I think we're having both techniques, where we can come in with the smaller membrane proteins, and then cryo EM comes with much larger complexes
Paolo 13:38
Is there any economy of scale or cost reduction driven by technology progress in the field? So an 800 megahertz instrument probably a few years ago was the top you could you could think of it was probably the most expensive things. Is it cheaper to buy one now? Or is it just getting more and more expensive as the technology and the complexity of things increase?
Dr. Loren Andreas 14:03
Well, actually, the last jump from 1 gigahertz to 1.2 gigahertz is substantial. 20% jump. Yeah, the incremental cost over a gigahertz is not as much as you might think. So it's per gigahertz, those last 200 are actually a pretty good deal. And the other component to the answer, I think, is essentially the the field was limited to around a gigahertz based on the material. And so there was a long lag, where we've been waiting for the next installment and this required going to a completely different material for the for the most inner coils of the magnet, these are high temperature superconductors. And that required a sustained engineering effort over more than a decade to get then this, this improvement. So now that now that the company has figured out how to make these magnets we're hoping to have further increases in the years come, it might help reduce costs, if we had a second instruments supplier in the market, who could who could provide some more competition,
Paolo 15:09
There's no competition in the market, interesting.
Dr. Loren Andreas 15:12
Well, there is, I always remind this company that there is still competition in the market, because each department, each Institute still has a choice to buy an NMR spectrometer, or to buy a cryo EM, apparatus, or to invest in other areas.
Paolo 15:33
I can tell from the way you speak about this, that you're fascinated by the technology and probably the physics behind the and that, that seems to be a very strong driver for you. On the other hand, I'm really fascinated by what you can do with these instruments and results you can obtain, like in your work, that you publish PLOS, which is what I think one, your consideration for the Chemical and Engineering New's Talented 12, in 2019. Was was, quite frankly, astonishing. I mean, it's, it was interesting how, you know, the power of this technique, and the progress has been made in the field, you know, could could allow you to manage a controversy that was in the field between the use of different technologies that were giving contradictory results, and how you could actually, with an improved application with an improved technology, you can actually shed light on, on your more reliable results. And you know, and actually learn a lot in terms of how the of the biomolecule looks like and the way it functions. So can you can you comment a little bit on on the importance of that work, and what was the the novelty of it?
Dr. Loren Andreas 16:47
Sure. So, the controversy essentially was regarding some structure or lack of structure in part of the protein. So, I should say also, that within the membrane, actually, the different results agreed. So the protein had basically the same shape within the membrane. But these loops, which extend outside the membrane, were either seen as quite flexible, using techniques of solution NMR. And they were seen as much more rigid in in crystals. So these were both detergent preparations, either detergent micelles in solution NMR or crystals would perform in the presence of detergents.
Paolo 17:30
And just for my understanding, they the flexible part is the one which is not supposed to be in the transmembrane part of the protein. So here we are speaking about a transmembrane protein that acts as a sort of transporter in the cell in in natively, how does it work?
Dr. Loren Andreas 17:49
Yeah, so I should say also that the the controversy is a among a class of related proteins. So not all the data came from an exactly comparable protein. But what you say is right that the transmembrane portion, it's like a rock, it's like an anchor in the membrane. And this is also part of the puzzle of how do these proteins really function in in transporting, when the membrane part looks quite rigid, and doesn't have a whole lot of space for transport. So we picked one of these proteins, which has a pretty good biochemical data to back up the fact that this protein allows very hydrophobic molecules to pass through the membrane, because some of the other examples, have some information. And maybe it's maybe these conduct amino acids, but it was a little bit less clear. Or for AlkL, the protein that we selected, there's really several different lines of evidence coming from more defined like liposomes polymer preparations, or from whole cell-based assays, you can see that these allowed hydrocarbon molecules into the cell. And in some of the crystal data, you could see already actually that some of the detergent molecules that have a long hydrocarbon chain on one end, were found with that hybrid hydrocarbon chain extending inside of the protein in the structured part outside of the membrane. And so this was maybe a first hint towards this being a pathway for conduction. So what we found was essentially, in this case, that the answer came a bit closer to the case of the crystal structure. And we can explain the transport for the related protein, ours is AlkL there. So it's either AlkW or OPRG. And we could explain then the transport that in the external part, the part which we describe, it's outside the membrane. This is where the molecule passes through the protein, and then it exits out the side and goes into the hydrophobic membrane. And of course, the next question asked is okay, why does it need, why do you need Protein outside the membrane. And that's because it's not really outside the membrane, it's, it's in the polysaccharide layer of the outer membrane of bacteria. And for hydrophobic molecules, this lipo-polysaccharide, which is full of ions, this is actually the most impermeable part of the membrane. So that's where it needs to have a protein to increase the permeability of that more charged and hydrophilic part. And once it gets to the membrane, it's it's perfectly able to transit the hydrophobic part of the membrane. No problem, since it's public molecules. after all.
Paolo 20:44
We hope you're enjoying this episode of Bringing Chemistry to Life. Stay tuned at the end of the episode for information on how to access content recommendations from our guest, as well as information on how to register for a free Bringing Chemistry to Life T-shirt. And now back to our conversation.
It sounds like typically, the structural biology experiments are ultra simplifying, some systems that are actually very, very complex, right, so you actually take a protein and you put it in a solution. And to do that, you need to add some additives like detergents. And, and you know, it's quite intuitive that there might be an effect there. And the protein is on an in an environment, which is not it's native one. On the other hand, if you put it in a crystal, you have another environment, which is very unusual for the protein in its functioning shape. But the very least that we know some of these proteins are actually very flexible, right? So again, it's a, it's fairly obvious to think that what you actually see in terms of structure is not exactly how the protein looks like. So you're finding a way to study complex methods that are more similar to the native ones, but they're probably still quite simplified. So how, what's the level of reliability of the system we can represent or simulate at the moment?
Dr. Loren Andreas 22:05
I think we probably need a lot more examples before we can be sure. So solid state NMR structural characterizations like these are not so many. So in the protein databank, they're I think under 150 still. But it's really not a huge number to draw on. And for membrane proteins, then it's a subset of those structures. But from our limited experience so far, with several of these membrane proteins, we can get into trouble with detergents, when these detergents break hydrophobic interactions which exists. So in the case of AlkL, this was outside the hydrophobic part of the bilayer, where then this part of the protein has something like a hydrophobic core. And this is where the detergents can go in and cause some trouble. And in addition, for another class of membrane proteins for helical proteins, the helix helix interactions, I have the impression these may be a bit more sensitive to the environment, as compared with the beta barrels. These beta barrels from bacteria, these are incredibly robust in the membrane, and they have to survive large swings in temperature as the bacteria experienced these swings in temperature changes in membrane as the bacteria make a different lipid membrane at a different temperature and so on.
Paolo 23:18
Yeah, making me think so if I if I think about what I know of the Protein Data Bank, there's there's a huge number of proteins there, the vast majority coming out of X ray diffraction experiments pretty much, right? And of course, you said there, there are not that many transmembrane proteins, for obvious reason, I think it's much more difficult to work with them for that type of techniques. But uh, you know, in a way, I'm leaning to think that a number of the things we have learned over the years, based mostly on X ray, and maybe more recently on on NMR in solution might be potentially wrong. Do we really according to this risk?
Dr. Loren Andreas 23:55
there's, there's always some risk of that, I think, with any of the techniques, including what we do. So I think it's still important to gather more experimental evidence and get closer and closer to real native situation. So probably the largest difference from to to native situation that we still have in membranes is that we really need to concentrate our samples to get enough signal. So a protein in a cell might have just a few copies, or certain proteins. And in order to get a signal, we need to produce many, many copies and crowd them very close together in the membrane. And so this could introduce some, some issues for certain proteins. That being said, taking it again, the example of bacterial outer membrane proteins, when they're needed, are expressed in pretty high numbers. So they they can sometimes be seen clustering on the membrane already. So for example, bacteriorhodopsin will will be seen clustering on the membrane. If you take the outer mitochondrial membrane from from potato, this has shown that you can see vdax forming arrays and clustering directly out of the membrane of these mitochondria. So that's going to be a totally case by case situation. And it's also often pretty unclear which other factors may be needed, which other proteins may be needed for, for some. So we were often studying proteins, which, at least we think function on their own. But of course, many proteins need the context of a larger complex as well. And that's, I think, something quite exciting to see with, with cryo EM for these larger contract complexes, that often, it's possible then to take older data coming from NMR from X ray crystallography, typically, and place these into the larger complex. So fitting the crystal structures or NMR structures into the density of a larger complex seen in single particle cryo EM.
Paolo 25:58
for the benefit of our audience. Can you comment a little more about cryo EM? How does it work?
Dr. Loren Andreas 26:03
I'm not an expert, but everybody in this field has been watching the tremendous revolution and what's possible to do with cryo electron microscopy. And essentially, the method is, involves taking many, many pictures, images of single particles, and then through sophisticated software, figuring out what is the angle that each of these particles was at when his picture was taken, and signal averaging and averaging many many of these particles together to then come up with a model for the density. And then fit Of course, the primary structure of the protein which is known, or proteins, which is at least known and then generating an atomic model, ideally, once the resolution reaches high enough.
Paolo 26:54
So being microscopy, Basie, obviously, the resolution is lower than NMR. So you need to look at bigger systems, right? And and this is a fascinating combination between, you know, different techniques, able to look at different scale and putting them together creating more and more complex models. That's, that's, that's really, that's really intriguing. And at the end of the day, what you're saying is that, regardless of the technique you use, you are able to build pretty sophisticated models, or more and more sophisticated models that somehow tend to try to represent reality. But, you know, we are still pretty far from being able to actually study reality, or study system as as they are in real life. So you know, how far are we from running an NMR, or a cryo EM, over a whole cell system or something like that?
Dr. Loren Andreas 27:47
I think this is an area A lot of people are interested in with cryo ET (electron tomography) to look at larger structures, in cells, for example. And there's also an interest in doing this in NMR, to do in cell NMR, or within the case of membrane proteins would be great if we could just harvest membranes, for example, off of off of the cell, it'd be good to get rid of most of the other components and just keep the membrane so that we can increase the number of molecules we can get inside of our sample holder. But as I mentioned, for structural studies, we're reaching to the limits constantly, just because of sensitivity reasons. And typically, if we take off even a whole membrane, the concentration is much lower, but the target protein sensitivity goes down. And the level of changes that can be detected then are, are somewhat reduced. So we can maybe take a two dimensional spectrum, but not a four dimensional spectrum anymore.
Paolo 28:52
What is the role of computational chemistry in all this? You know, there's been recent exciting results with DeepMind alpha fold, able to predict folding up quite impressive way. And this was things that scientists have tried to do for many years failing consistently. So, how do you see are you see this?
Dr. Loren Andreas 29:12
This is potentially a tool which can be applied to really benefit all of these techniques. First of all, shocked to see the results coming out with with AlphaFold, but then also excited to see what can be done potentially with more limited data. So as I as I just mentioned in the whole cell case, the amount of data that could be acquired will be more limited. But with larger compliment coming from computational methods, it may be possible now to better interpret such data. And I think that's true from both the crystallography but also from from NMR. And I also wanted to comment that been so I'm also not really an expert in how AlphaFold is working. I'm not sure they even publish the details yet. But the fundamental information is actually very similar to the information you get out of NMR. So, with AlphaFold, they're using sequence information looking for correlated mutations to establish which part of the protein is next to which other part of the protein, right? At least, that's my understanding. And in NMR, we're doing exactly the same thing. But experimentally, we're taking a direct measure of proximity, so we're directly measuring one atom in the protein is near another atom of the protein. So I think there's a real opportunity probably, if we get the right people together in the same room to combine this kind of information into a single approach to benefit from the sequence data and within AlphaFold, and also from experimental data.
Paolo 30:46
So for understanding right, what you're saying, the way complex NMR is, of these these big molecules work is that you try to fit the sequence that, you know, to the spectrum that you see, and you calculate basically from from your experiment. And, and potentially computational chemistry can give you some extra tools to accelerate or, you know, to make it better, and, or faster in a way you actually correlate the two things.
Dr. Loren Andreas 31:11
Absolutely, yeah. So a huge amount of time is still invested into the primary sequence assignments, which is what you just described, but we want to match each peak in the spectrum to unique atom in the sequence. And so that's a big sink in time that we have to invest in just getting the primary assignments. And then from there, the second step is to look what happens through space, and to measure which of those peaks is nearby and space to another one. And it's that if that step, which is fundamentally very similar to the kind of very local information, which is being used in AlphaFold, although the sorts of information coming from many sequences is different, of course.
Paolo 31:58
This is fascinating, and and it's also fascinating how, you know, there's almost an infinite choice on what to apply all these wonderful new techniques to, right, because there's so many different directions to look at. And that's what leads to my next question is, you know, if I look at your work, you obviously focus on a couple of different things, right? There's the methodology on one side, and there's the application or you know, the structure of biological significance of the application of your work. What part fascinates you the most, what, what drives you and, and how you pick your next project?
Dr. Loren Andreas 32:37
I think often, some of the more fascinating things happen almost by accident, coming across new and unexpected and interesting things. But of course, we also need a sustained effort into a lot of these projects to make progress. One example of something unexpected, which was quite interesting to look at, we wanted to use just the amino acid phenylalanine. So one of the basic building blocks of a protein. And so we made some crystals and phenylalanine, and we started to measure them. And then we realized, actually, that it wasn't until 2014, that the correct crystal structure for phenylalanine was reported. Before that it was the data was misinterpreted with a smaller number of molecules inside of the crystal. And actually, with the technique that we were using, we were able to then detect this this difference. Of course, we were a little bit later than 2014. So we couldn't publish it as a saying, Okay, this, this older description of the data isn't correct. But we can at least verify by NMR, that actually, we could detect the same features, which were were noted when there was really exceptionally high quality crystal X ray diffraction data that could detect this subtle difference, and expanded unit cell. And so we ended up describing by NMR, how we could detect the same feature. And that ended up with the cover of the journal you can see behind me,
Paolo 34:05
There's a little bit of serendipity in that, Loren.
Dr. Loren Andreas 34:08
Yeah, I think it's a matter of being kind of open for surprises.
Paolo 34:13
That's interesting. So you're, you're saying, keep your eyes open, be ready to get input and react to it. And you know, connecting the dots is really what makes the difference between a good scientist and a great scientist. So speaking about scientist and as we come to the to the to the final part of our of our discussion, you know, this is a question I always ask to close our interviews, it's that you're still a young scientist, but already quite accomplished, you know, with a bright future ahead of you. What is one single piece of advice you'd pass on to a young scientist just starting their career?
Dr. Loren Andreas 34:51
This is exactly the answer that we already discussed is to make sure to be looking out for the interesting surprises. Sometimes a sequence may not work the way you expect. And it can, of course, be a variety of reasons. But sometimes it's an interesting surprise. I also have the impression that a lot of things change faster these days than maybe they did before. So this situation doesn't really lend itself towards giving simple advice or clear advice for the future. And so to be a little bit humble, I should admit that I never saw this advanced in cryo electron microscopy coming back in 2008 or 2009, when the resolution was something 10 or 20, angstroms. Not sure exactly, but not nearly high enough to get atomic models. I never saw this, this kind of improvement coming. So that's again, I guess a willingness to be open to surprises.
Paolo 36:00
That was Dr. Loren Andreas, group leader at the Max Planck Institute for Biophysical Chemistry in Germany, and one of the Chemical Engineering News' Talented 12. Thanks for joining us for the season one bonus episode of Bringing Chemistry to Life, and keep an ear out for more. If you enjoy this conversation, you're sure to enjoy Dr. Andreas' book, video, podcast and other content recommendations. Look in the episode notes on the web page or the app you're using for a URL where you can access these recommendations and register for a free Bringing Chemistry to Life T-shirt. This episode was produced by Matt Ferris, Matthew Stock and Emma-Jean Weinstien.
Transcribed by https://otter.ai