 I'm really honored and pleased to welcome you all to Ben's awardee lecture. And you can see the title there, and Ben's picture, and today's date. And go to the next slide. Ben, good. So let me give you a little background on Ben, and then I'm going to talk a little bit about Paul Boyer, for those of you who aren't aware of the funding or why it's called the Boyer Award. So Ben got his undergraduate at the University of Pittsburgh in 2010, and then his PhD with me at UW in 2016. He's received numerous awards, including a campus training grant, the Genomic Sciences Training Grant through NIH. Since 2016, he's been the scientist in my lab in charge of our mass spec based analysis of proteins in plasma membrane robidosis. Next slide, Ben. So my lab shift direction about probably a decade ago. We had already been doing a lot of quantitative proteomics, mainly phosphoproteomics, and we were learning a lot about the protein kinase cascades and plants, which is enormous. There's 1,000 protein kinases and 300 phosphatases, and Ben did his thesis on that. And then at a faculty meeting, Ivan Raymond made this statement. And I know Brian has probably passed out if he's watching this, because this actually means I actually listened at the faculty meetings, or I did that one day. But basically, Ivan said ultimately all, well, I can't use his accent, but he said ultimately all biologists wind up analyzing a single amino acid side chain in a protein three structure. And then Ben made, why the hell am I studying the dead proteins? We should be studying the confirmations that allow proteins to do what they do. So we started a big program on that using mass spec based methods. Some of you are familiar with it. I'm not. And Ben will be talking about one using hydroxy radicals. Next slide. So basically, Ben has led the charge in converting my lab from the post-PTM work to basically fulfilling Ivan's statement and identifying amino acids residues involved in either maintaining or altering a protein 3D structure. And his talk will be about a new system we developed in collaboration with an electrical engineering professor who was a plasma physicist for analyzing conformational changes in proteins using mass spectrometry. I won't go into great detail on how that arose, but perhaps Ben wants to share that. It involves a trip to Washington University. And yeah, OK. Next slide. So let's talk about Boyer. So Paul Boyer won the Nobel Prize, 97, I think. And he got his PhD at UW in 1943. I think it came from Brigham Young. And it was work time. And he went from UW to Stanford for a postdoc. And then he spent, I believe, a couple of decades climbing the 10-year ladder at the University of Minnesota doing some really beautiful work in the field of bio-angetics. And that is a field that I grew up in as a postdoc myself all the way starting with Peter Mitchell. At the time when I met him, the thing that stood out to me was he was wearing an earring. And he was one of the first male scientists I met who did that. This was even before the Nobel Prize, though. He was pretty amazing. But chemist Moses, hypothesis wasn't the end. We still have to figure out how the damn proteins operated. And Paul got the award for doing some really insightful work on O-18 labeled water and the synthesis of ATP. And I would like to remark that possibly. I know the University of Minnesota had a very strong mass spec program. And I suspect at the time. So it was kind of a nice wedding between a brilliant biologist and chemist who had a technology. Something that can always work magic when the two get together. Next slide, Ben. So I got real curious and tried to track down his thesis. And it turns out there is a hard copy. And it's being stored in the storage facility in Verona. Kind of like at the end of Indiana Jones, that big giant storage warehouse. So Paul's thesis is buried somewhere in Verona. It's not a really big thesis. It's only got 57 pages. And he didn't win the Nobel Prize from his thesis. That's the point to make on next slide. So again, he didn't win it. It was a good thesis. But he didn't win the Nobel Prize. He wanted for insights involving how the protomotive force was coupled to ATP synthesis. And this article I found, I found it fascinating. It's titled To Error and Win a Nobel Prize, Poor Boyer, ATP Synthesis and the Emergence of Bioenergetics. And some of you may know it was a very bumpy road, both for him personally and for the field. There were a lot of false leads. I think they were all looking for this high energy intermediate that passed by his thing. But it turns out to be in a different enzyme, not in the ATP synthase. So it wasn't a smooth journey. Some of you may think that's what happens. It doesn't happen like that. Next slide. So the only scientist I know whose PhD thesis won him or her in the Nobel Prize was Josh Leiterberg. But he's a god or a demigod. So for the rest of us earthlings, you must realize the PhD thesis is just a ticket to begin the research journey, not the destination. And I just want to say if any of you PhD candidates are watching this seminar, you may also wind up winning the Nobel Prize down the road. It may not be for your thesis, but down the road, if you live long enough, it can happen. So I'm done with the introduction. I want to thank Paul for the gift that funds this annual competition. And I'm most important of all, I'm very honored to be working with Ben. And I'm also very honored and pleased to turn the Zoom, whatever this podium. It's not podium, but you know what I mean. I'm turning it over to Ben so he can give his 2020 four year awardee lecture. Go ahead, Ben. OK, let's OK. Let me make sure I can advance. OK, great. Can you hear me? Right now I can see Mike's picture. Mike, can you give me a thumbs up? Can you hear me just fine? All right. Yeah, gotcha. Cool. So thanks for the introduction, Mike. Unlike the end of Indiana Jones, I hope that if I ever find that thesis and open it, reading it doesn't melt my face from my skull. So let's hope that doesn't happen. So yeah, thanks for that intro. Thank you very much to the Department of Biochemistry for selecting me for this award. Thanks to Mike for nominating me. And moreover, thanks to the Department of Biochemistry for enabling science like this, not only for our lab, but for labs throughout the history of the department. It's a great department. So today I'm going to talk about a technique that we invented to analyze protein structure using microsecond pulses of hydroxyl radicals. And we do this to ascertain structural features of protein and solution. And we do this using the fourth state of matter, plasma, which is why I have these little lightning bolts on the title slide. So to give you an idea of what I'm going to talk about today, here's a quick little outline. I'm going to first talk about the field of protein footprinting in general, why we selected hydroxyl radical footprinting to do structural mass spec. I'll talk about our technique, kind of the development and a little bit of the pitfalls and getting that instrument built. I'll show some example data and some benchmarking experiments. And I'll talk about some preliminary new methods that we're developing, how we're developing them, and where we're going in the future. So let's get started with some protein footprinting. Protein footprinting at its base is a structural tool. It's a tool for doing protein structure analysis. The general principle looks something like this. So if we have a protein folded in a form in solution and we react it with some sort of chemical label, and this can be non-covalent or covalent, we'll modify that protein all about it in a solvent accessible fashion, whereas the hydrophobic interior that's buried won't be modified. If we digest that protein down to peptides using standard mass spec proteases and analyze the peptide mixture with mass spectrometry, we can identify where and to what extent the protein was modified with that chemical label and in a backwards fashion infer which regions of the protein were accessible to solvent and which were buried and ascertain higher order structural features. The real power in this technique comes from comparing two states of a protein, two conformers, I guess, so to speak. So if you add a ligand or a binding partner or some sort of external stimulus that unfold a portion of the protein, do the same type of experiment, you can look for differences in modification in the mass spec data, and that will give you some idea about how and where the protein has changed its shape. And there are a number of ways that you can footprint a protein, especially when we started getting interested in doing structural mass spectrometry, I don't know, maybe eight years ago or so, the field was growing and it still is. The most, I guess, prevalent historic way to do it is with hydrogen deuterium exchange. This is a non-covalent labeling technique in which backbone hydrogens along the peptide backbone are exchanged with deuterium by immersing your protein in a solution of D2O instead of H2O. Regions that are quite solvent accessible, more flexible or dynamic, will have the backbone hydrogens exchange with deuterium very quickly, whereas more ordered or more structured regions will exchange in a much more slow fashion. And this is an example, as I said, of non-covalent labeling. One of the technical difficulties with this technique is that as soon as you stop the labeling process and start processing your sample for mass spec, generally in aqueous buffers with H2O instead of D2O, you have exchange of those deuteriums you've labeled your protein with back with the bulk solvent, so time is a huge factor in these experiments. So that's one of the, I guess, pitfalls or I guess, detriments of doing non-covalent labeling specifically in this fashion. Another example of a way to footprint a protein is with residue-specific chemistry-covalent labeling. So one such example that we in the Sussman Lab have a little bit of experience with is carbodynamic labeling. So you basically activate the carboxylic acid group on aspartates and glutamates, react it with a secondary molecule and you get solvent accessible labeling, but only of the aspartates and glutamates. And this is covalent, so it doesn't suffer the same back exchange issues that you have with HDX. And I will point you all to a wonderful review from Mike Gross' lab that details a number of residue-specific labels you could use if you wanted to do protein footprinting in your lab, and they're all commercially available. The technique that we ultimately settled on was hydroxyl radical protein footprinting, and we thought that this sort of wetted the two other techniques that I just mentioned in a very nice way. Hydroxyl radicals will covalently modify amino acid side chains in your protein, and in theory, they will react with all residues. And we now know that this is more like the most reactive, I guess, subsection of residues, methionines, aromatics, and so on, but nonetheless, you get specificity for more residues than just a single chemical group or a single residue. So when we decided to do structural mass spec, we picked hydroxyl radical footprinting as a technology that we wanted to adapt within the lab. And when we entered the field, there were two ways to do it. The first was kind of pioneered and pushed forward in the late 90s by Mark Chance at Case Western Reserve. And to do this first technique, which is called synchrotron radiolysis, you put your protein in a vessel with an aqueous buffer and you hit it with an enormous amount of synchrotron radiation generated energy. And what that causes is radiolysis of water molecules in your sample, a sort of nuclear blast of hydroxyl radicals that label your protein very, very quickly. So again, this is an example of radiolysis of water to produce hydroxyl radicals. And to do this, it's not feasible for every lab, you need access to a synchrotron beam line. The other technique that existed at the time was as Mike mentioned, invented by Mike Gross at Wash U in St. Louis. And this is called fast photochemical oxidation of proteins or F-POP. This involves mixing your protein with millimolar amounts of hydroxyl radicals and a number of radical scavengers and flowing that mixture through a capillary at which you're pulsing a 248 nanometer laser. And when you pulse that laser, the volumes there that get hit with this laser actually photolyze the hydrogen peroxide to create little controlled bomb blasts of radicals in a volume dependent fashion. And this was more compelling to us because if we wanted to explore this technique, we could build it on a bench top and actually implement it in our lab on a regular basis. So in, I don't know, 2013, 2014, maybe 2012, Mike Susman and I went down to Wash U in St. Louis, spent a few days at Mike Gross's lab learning how to do this. We got some promising results with the technique. So we wanted to bring it back here to UW Madison for our lab and to offer it to other labs through collaboration. So what we didn't know when we started that was that to buy a laser to do this, it would cost us something like 40 to 50 K at the time. And we weren't ready to throw down that amount of cash until we knew we had a working setup here in our lab. So Mike Susman talked to a number of colleagues around campus. One of them was Professor Leon Chauhet in the Department of Electrical and Computer Engineering. And Mike said, I want to borrow a laser so we can make hydroxyl radicals so we can label proteins. And Leon said, I don't know about proteins, but I can tell you that we make hydroxyl radicals all the time in our lab from water with plasma. You can do it really easily. And thus the idea for our technique, Plym, plasma-induced modification of biomolecules kind of was seeded in our minds. And in initial discussions between our lab groups, we realized that if we could get a system that would produce hydroxyl radicals from plasma that we create by electrical discharges, I should say, we may have some selective advantages over both synchrotron radiolysis and FPOP. So we decided to pursue this. Specifically in our initial talks, what came out was that plasma can be generated in microsecond bursts. And this in theory would lead to microsecond bursts of hydroxyl radicals, which is very important because you don't want to label a protein on any longer timeframe really, due to thermal motion, due to potential unfolding, things like that. Leon told us and assured us that we could create radicals from water. And this was important to us because it meant that if we get this up and running, we could have a system in which a protein was in any aqueous buffer that kept it happy. We could generate radicals from the aqueous buffer and label a protein in that fashion. And finally, Leon mentioned that he could build this into a benchtop device that someone like me, a biochemist with no electrical engineering background could safely and reproducibly use to actually do these experiments. So for all those reasons, the fact that we could generate radicals from water and the fact that we could do it on a benchtop, we thought this kind of wetted the two advantages of those other techniques and didn't have some of the detriments of those other techniques, synchrotron radiolysis that I mentioned. And I'll note here, in contrast to radiolysis or photolysis, this would be an example of plasmalysis of water. This is a term that was coined, I think, by wastewater sterilization treatment. Scientists, because they refer to treating wastewater with plasma to generate radicals to basically kill microbes as plasmalysis. So I guess we're doing plasmalysis, which is kind of cool. And plasma is very interesting. It's its own sort of beast here, come on now. Okay, if you give a solid some amount of energy, it will melt into a liquid. If you give that liquid some amount more energy, it will vaporize into a gas. And if you hit that gas with a significant amount more energy, and what we do is we use electrical energy, you'll create this kind of pseudo gaseous species that is incredibly reactive, and it's incredibly energized. And that mattered for our purposes because plasma contains and produces radical species. So what I'm showing you here is an idealized example that's not actually too far removed from our kind of final instrumentation. This plasma is generated in argon, and you can see the discharge here in purple. We are thus creating argon radicals, or the authors of this paper are thus creating argon radicals in the plasma itself. There's a number of negatively charged ions created in the water vapor above the liquid here. There's an interface where there may be diffusion between the vapor and the bulk liquid. And in that bulk liquid, when you hit an aqueous liquid with plasma, you're generating hydroxide ions. And as we now know, you're generating hydroxy radicals, hydroxyl radicals, which is very important. So in theory, we thought this could work. The first thing we did, why is it not letting go? Because I'm a co-host, okay. Okay, the first way that we decided to try this was with direct current voltage. And I think the PIs have probably never seen this video. It's really old. This might be the first video we have of trying to get something to work. I will note our setup now doesn't resemble this, as I'll show you in a few slides. But what I wanna point out here is this beaker, and this is where the plasma is gonna be generated. And you can see hopefully two wires immersed in there with maybe a centimeter of space between them. Over here, we have to close the circuit from a distance with a long wooden rod with a hanger on the end. So you're gonna see that come down, tap here, and you'll see the plasma created in a flash right in that beaker. And it's pretty violent, boom, right? So if I kind of slow this down, there is a huge flash that we see there. And that is the plasma, and it's a microsecond discharge of plasma in water. And it causes so much pressure that the water jumps out of the beaker all the way off screen. And so in order to do this direct current voltage setup, we had to charge up a capacitor and discharge it from a distance to do it safely. It was not sustainable in the long run, suffice it to say. After about a year and a half, maybe two years of banging your heads against the wall, getting some positive results here and there, we switched to alternating current voltage. Here's a schematic of the current setup. I will just say that alternating current voltage has some unique advantages over the direct current, the DC voltage that we used, in that it's got a lot more tunable parameters via computer-controlled function generation. It results in a gentler plasma. So when we use a computer to control a function generator, push that through a high voltage amplifier and apply about 10,000 volts through an electrode as shown here, what we get is a plasma that's a cold atmospheric plasma. It's a lot gentler than what you saw before. And it's directed down through our sample with a ground electrode underneath the tube. So another important thing here is that this is a dielectric barrier discharge. So the electrodes here do not contact our sample at all, which for various, for obvious reasons, I guess, is much better than having to throw wires down in our sample every time we wanna blast it with plasma. So this worked very well. What this looks like in practice is this. So yeah, so we hit it with 10,000 volts and we get a little bolt of lightning that comes out of the tip of that needle, sort of dances over the surface of our sample. And that's how we generate radicals. This is also hooked up to a Peltier cooling block underneath so we can cool our sample while we expose it to plasma. So if there is some minor heating due to the energy deposition, we can offset that. And what I think is kind of cool is we found this proprietary material ceramic that we machined into a single well here. This is a PCR tube. It is thermally very conductive and it's totally electrically non-conductive, which was super beneficial because if we didn't have that, we burned through a lot of tubes with the plasma getting pulled to the side of the tubes if we didn't have this insulation at first. So this is kind of the final setup. And while that may look like it's a constant bolt of lightning, the reality is that it's not. It's many, many, many, many microsecond bursts of kind of lightning in a tube that are separated by between five and 30 microseconds between each of them. So if we apply a voltage in this sine wave function that you see here, when enough energy is built up, each of these spikes is a single plasma discharge and the time scale, as I said, is microsecond. And this is of fundamental importance because in between each of these microsecond bursts of hydroxyl radicals, there is enough time for the radicals that we generate that have not modified our protein to recombine with bulk solvent. So in contrast to secretron radiolysis or F-POP, where you're basically hitting a protein with nuclear blasts of hydroxyl radicals, I guess the analogy here is we're hitting it with kind of little tiny POP gun bursts, very gentle doses of radicals. After each little burst, the radicals kind of disseminate, dissipate into the bulk solvent. So it's not seeing an over, the proteins are not seeing an overwhelming amount of radicals at any given point in time. So we thought that was workable. And okay, and so we wanted to benchmark the system with a number of biological systems and they're shown here and I'll go through these. So we started with methionine, expanded it to bovine serum albumin, looked at higher order structural feature, determination with EGFR dimerization, and then epitope mapping with thrombin and a monoclonal antibody against thrombin. So the general principle for PLIM is shown here. We've got a protein in aqueous buffer, we hit it with plasma and we should get an oxidized protein in aqueous buffer out the back end and that should happen in a solvent accessible fashion. The first thing I did to test this was to use the most readily oxidizable amino acid methionine. So it's unmodified form is 150 Dalton's, methionine sulfoxide, a very well-studied oxidation product of methionine is 166. I can easily resolve those on a mass spectrometer. So if I take methionine in buffer, hit it with increasing doses of plasma, I should be able to track evolution of methionine sulfoxide and hopefully we see it evolve in a dose dependent fashion. And that's exactly what happens. As we hit methionine in solution with increasing treatments of plasma, we get a linear response that demonstrates we are generating a constant flux of hydroxyl radicals. Towards the end here, variability may increase to a certain extent, but nonetheless, it really does maintain the linear trend that we see, which is great because it suggests we're producing radicals at a constant flux. We next moved to bovine serum albumin and we use this because it's cheap, it's pure and you can buy it in massive amounts. So we had some around our lab. It's also a well-studied protein. People use it as a standard in experiments all the time. And so we're taking BSA in PBS here, hitting it with plasma, it's phosphate buffer saline, hitting it with plasma and hopefully getting oxidized BSA out the back end that we can, the oxidation of which we can measure and quantify using mass spectrometry. And the way that that looks is a little bit more complicated than with pure methionine. So as I said before, protein footprinting does rely upon mass spectrometry. And you can use standard proteases. In our case, we generally use trypsin to digest a protein down into little pieces, peptides. And each of those peptide variants can be unmodified a string of amino acids here or it can pick up any number of oxidation events to become oxidized. And we can easily resolve that mass difference on a mass spectrometer. So when we get signals for the unmodified and modified forms, I can build that into a percent modification by basically looking at the signal intensity for modified forms over the signal intensity of all forms to give us kind of a percent occupancy of that peptide by oxidized forms. So it's a percent modification. And when I do this with BSA, here's what we find. What I'm showing you here are peptides from N-terminus to C-terminus. And I'm showing percent modification on the Y-axis and I'm showing increasing doses of plasma going from blue to red to green to purple. And we see there are regions that are not modified at all. We see there are many regions that are modified to a very low extent, but we see regions that jump out immediately, peptides that are very highly modified in a dose dependent fashion by plasma treatment. And I will tell you right now, these are areas that are solvent exposed that contain methionine side chains. So remember I mentioned before in the intro, there's a range of reactivities. Methionine is by and large the most reactive to hydroxyl radicals. And we see that again and again. That said, if I go down into the weeds here, so I've scaled the Y-axis to half a percent, we see dose dependence in peptides on a much lower scale and you can see that here. And this is a peptide that does not contain methionine, but contains some other residue composition that nonetheless yields dose dependence just on a totally different scale. If I scale the Y-axis even lower, we see as well a peptide up here towards the enterminus where we see dose dependence. And I'm not gonna go any lower than this, but we see hints of dose dependence at even lower oxidation rates. So this is cool and this confirms pretty much what we knew about differences in variability between the amino acid side chains with hydroxyl radicals. We wanted to make sure that we weren't unfolding our protein throughout exposure. So I did a second experiment where I took BSA and these are 10 peptides spaced relatively evenly apart throughout the primary sequence. And when I use a protein that's folded and native BSA and I hit them with up to 60 seconds of plasma like I just showed you, none of these peptides are highly modified and they're not modified in a strongly dose dependent fashion. If I take BSA and digest it down to peptides, kind of like the ultimate structural perturbation, remove total tertiary structure from this thing, and then I hit it with plasma. Every single one of these peptides is now much more highly modified and modified in a plasma dose dependent fashion. So this to us suggested that at least for BSA under these conditions, we're not unfolding our protein throughout this treatment since when we do, we get modification all over the protein in regions we didn't see it before. The next thing we wanted to do was test the ability of our system to identify structural changes on a higher order. And so to do this, and then that's what I mentioned in the intro, the real power in this technique is looking at modification differentials between different conformational states. And to do that, we used epidermal growth factor receptor plus or minus its native ligand hormone, epidermal growth factor. And we basically treated EGFR with and without plasma in the presence and absence of the growth factor. So the mechanism for how the growth factor affects the overall structure is shown here. EGFR is a monomer and this is the extracellular domain. And this is the closest construct. This is the closest crystal structure to the construct that we're using for these experiments. We actually have a C terminal tail down here that wasn't crystallized. When you add EGF to EGFR, there's a binding event shown here and it forces dimerization about this very clear interface that you can see here. So as I said, we mapped solvent accessibility of EGFR as a monomer and as a dimer after adding EGF using our system. And the oxidative differences that we saw are shown here. We get statistically significant decreases all over the protein. And I wanna note similar to BSA, we see a range of reactivities. So there's significantly reactive peptides, there's far less reactive peptides, but nonetheless, we can still resolve differences between these two conformational states due to that decreasing oxidation. And when I take those oxidation decreases shown in yellow here and I map them to that crystal structure, here's what we find. We clearly do get a point of interfacial interaction in that dimerization interface that decreases in solvent accessibility. There's a point down here where we see lower oxidation, that's a hinge region about which two of these domains are proposed to rotate when EGF binds, we get some other slight conformational shifts throughout the protein. And I wanna note that peptide 515 to 523 here is not in this crystal structure, that's part of our construct. It is down here, it's a juxtamembrane domain that's positive to interact as part of facilitating the transmembrane domain interaction here. So that was very cool to see. Now we did not, we were not able to map EGF binding, but this gave us confidence that we could at least see higher order structural changes using this technique. So one of the growing fields for protein footprinting is epitope mapping. And because of that, we wanted to test the ability of Klim to define an epitope. And I wanna note that this work and analysis was performed by Faraz Chatteri and Dan Benjamin. And I'm only gonna show you this one experiment that they did, but over the past few years, they've done so much more, a number of collaborations, a number of other experiments. And so credit and thanks to them for this dataset and for many others. So we picked this system, thrombin and a monoclonal antibody. Thrombin is a protein that's involved in the blood coagulation cascade in our bloodstream upon wounding. We picked it specifically though, because hydrogen deuterium exchange and F-POP have already performed this experiment and they've defined the existing epitope or they've defined the epitope. So we wanted to compare our protein footprint printing technique to others that are complimentary that have already determined the epitope. So here's where they called it. This is the primary sequence of thrombin I'm showing you. F-POP defined the epitope here in red, it's underlined, and HDX defined it in these sequences in black. And this is cool to see that these techniques produce overlapping data. So we wanted to see what ours told us. So similar to the EGFR, EGF experiment, I just showed you, we're doing thrombin plus or minus the monoclonal antibody raised against it that was used in those other experiments. And so when we map oxidative differences to thrombin, as compared with the thrombin monoclonal antibody complex, here's what we see. We only get four points across the protein that are changing in oxidation. And what we see here are two areas that decrease their solvent accessibility and two areas that increase their solvent accessibility. When I map those changes to the crystal structure, it looks something like this. So those two decreases peptide 114 to 119 and 143 to 159. And these correspond to residues within the primary sequence for what it's worth. These two decreases together produce a contiguous three-dimensional surface across one face of the protein, which is very cool. The amino terminus 1 to 17 increases in oxidation shown here and this loop right here was a disorder loop that couldn't be crystallized here. It's actually around back of the structure, but there is not electron density for it. So I could not map it here. Now, as we look at this crystal structure, I think it's very reasonable, as you guys are all probably thinking now, that the epitope is very likely the area where we see decreased oxidation. And it's not unreasonable given this is the end terminus that antibody binding may shift that end terminus out of the way and that could be the reason why it is more solvent accessible once we bind the antibody to thrombin. So we would call the epitope as a combination of residues 143 to 159 and 114 to 119. And so how does our data compare with the complementary footprinting techniques? It's spot on. We get overlapping regions almost identically. And so we are very happy to see that we corroborate the other data sets that were obtained using fundamentally slightly different technologies, but nonetheless are all protein footprinting techniques. So this gave us confidence that we could use Plym to map epitopes. So that was very cool. So those are just two experiments that we've performed. Again, there are many more, but I wanna use the remainder of the time to talk about some newer methods we're developing, how we're developing them and where we hope to go in the future. The preface here is that a few years ago, we started thinking very hard about some of our data sets and we realized that plasma contains and produces radicals specifically that stem from the media used in the system you're generating within where you're generating that plasma. And so in this intro slide, I mentioned that the authors here are using an argon plasma. So the gas medium that they're using to generate that electrical discharge is argon. And so they get argon radicals. We have never really used argon in our exposures. We've always used air, the air that we breathe, the air that's in our labs, that's kind of the baseline that we used. And air is mostly nitrogen. In fact, it's 80% nitrogen. It's a lot of nitrogen. It's 20% oxygen and there's a little bit of argon and there's trace other compounds. If we measure the air in our lab, just to be sure, and this was done by Josh Blatz, designed to analyze by Josh Blatz. Yeah, the air that we have in our lab is in fact 80% nitrogen. It's pretty spot on with reported literature values as is our water vapor composition. So this caused us to step back and think for a little bit about what we can actually use our system for. Because air is mostly nitrogen, we kind of had this hypothesis that we might be generating reactive nitrogen species modifications on our protein, given that we're making a plasma out of mostly nitrogen. And so I went back and I probed our EGFR data for kind of the most widely reported RNS modification. And that's on tyrosines and tryptophanes. And I'm showing you that here via peroxynitrate radicals or nitrogen dioxide radicals. You can get nitrogenation about the aromatic ring here. And this also happens on phenylalanine. So I went back and searched the EGFR data again. And yeah, we found reactive nitrogen modifications all over the protein. So I'm actually showing you peptide sequences here and highlighting tyrosines and tryptophanes that we saw as modified. And importantly, it's plasma dose dependent. In spite of some of the massive error bars and variability here, we're still seeing RNS modifications on our protein in a dose dependent plasma fashion, which was super cool. This had been hiding below the surface for a year or two and we hadn't even thought to search for it. So this also in our minds led us to the idea that we might be able to use, take advantage of this, take advantage of plasma composition and use different radical species to access solvent accessibility information about regions of the protein that hydroxyl radicals alone and their very specific subset of reactivities can't give us. So it caused us to ask the question, can we control what species we generate? And we would do that by controlling, as I said, before the medium in which we're generating this plasma and that, I guess media because we're referring to the whole system here. So I mentioned previously that we don't wanna mess with the bulk liquid. One of the advantages here is that we're creating, at least for now, radical species from aqueous solvents. So this isn't the thing we wanna add hydrogen peroxide or chemicals to at this stage. There's not much we can do about the interfacial region. So that was kind of out. Water vapor pressure wasn't something we were super interested in up front. So that kind of left the bulk plasma and the gas that we used to generate the bulk plasma as a point of modification. So we tested this. Josh Blatz in Leigh and Showhead's lab built this very cool setup. It's a rough setup, but very cool where we bought pure argon, pure nitrogen, pure helium, hooked them up to a mass flow controller that fed into a chamber around our little exposure system here. And what we could do here is kind of evacuate the air as best we could by closing this chamber off and flowing gases at a high rate through the back. And so what this would do in a, again, a crude fashion is it would enrich the airspace here with different gases and it would fundamentally change the composition of the plasma that we're creating. And hopefully change something about what we're doing to our peptides and our proteins. So the first thing I did here was exactly what I did when we got the original air plasma working and that was to examine what happened to methionine when I hit it with plasmas generating using these differential gases. And I wanna take an aside here to show you some videos helium breaks down far more easily than the rest. So if we were to, when we use helium and the voltages we normally use for our biological experiments with the air plasma we basically get breakdown all throughout the air of the chamber. So helium breaks down into plasma very easily. Path of least resistance is now even though it's a much farther path it's all the way basically to the side of the chamber down grounded into the LTA cooler, which was kind of a bummer. That said, we were able to turn down the voltage on the helium again, the tunable parameters of the AC voltage. And when we did that, I'm showing you a video in the dark we were able to get exposure that was directed down through the sample itself but we got this kind of waterfall cascade of discharges around it. And that's even with a much lower voltage than the other thing. So helium was a bit hard to work with but nonetheless we were able to get some information out of the helium samples. Okay, so again, I asked the question what happens to methionine when we hit it with plasmas generated by different gases. And the first thing I looked at was classical oxidation and what was very interesting to us here was that we create that 166 classic oxidative product but we created at different rates. So in blue you can see air, in yellow and orange you can see nitrogen and helium and all of these are mostly linear except for argon and gray which appears to get saturated or something after about 30 seconds. And this led us to hypothesize that we're getting some secondary products or something else that's happening here because this is not what we would expect. So I went in and manually pulled out all of the mass shifts that we saw on methionine with all of these doses of plasma and the ones that were dose dependent have plotted here. So basically we see a number of different mass shifts that occur with the different gases. I wanna note that helium itself generated fewer modifications overall really these only these two in the center that all the other gases shared so I'm not showing helium here but there's a couple of things to note here that are very promising. Some of the modifications that we're identifying that are dose dependent are unique to each gas used. Some of the modifications are shared between all the gases. So we can now see plasma gas dependent modifications and we can now see plasma gas independent but plasma specific modifications. So of course we wondered, do we know what any of these things are? And the answer is yes, but very few of them. Plus 16, I mentioned before it's classic hydroxylation. Plus 14 that we see in argon is classic carbonylation that's a C double bond O replacing two hydrogens. Minus 30 that we see in the air is very widely reported oxidative decarboxylation pretty much of the C terminus of peptides or in this case an amino acid. And that leads into the minus 14 that we see in all the gases as well which is the minus 30 plus 16 decarboxylation and hydroxylation of that thiol in the methionine. So this was very cool. And we wanted to push this forward because as I mentioned before what we really care about are proteins and protein modifications. So we wanted to ask if we see the same effect the same phenomena with protein. And so to do this, I used E. coli produced human P53 in PBS and the construct here was provided to us by the Anderson labs. Thank you to them for that. And this was produced for me by our lab manager Heather Burch. So using wonderfully pure wonderfully concentrated P53 and leaving out helium because of kind of the pain the pain in the asset was to deal with we're looking at the effect of air plasma, argon plasma and nitrogen plasma on P53. And so the first thing I did was look at the rates of modification of peptides for classical known oxidative products. And what I'm showing you here are two examples of that. So again, these are residues 66 to 101, 140, 156. What I'm showing you here is that when I try to quantify classical oxidation we see a different rate of reactivity with these peptides when comparing argon and nitrogen plasma versus air and nitrogen plasma versus argon. Argon yields far less reactivity on these peptides overall. The next thing that I asked similar to methionine was do we identify mass shifts on these peptides that are induced by the plasma in a dose dependent fashion and are there unique modifications that we can see to each of the gases? And the answer is overwhelmingly yes, there are more here than there are with free methionine which is perhaps unsurprising since, once you go from a free amino acid to a full protein with structure, the chemistry gets, it blows up pretty much. And I wanna note that I use, the bioinformatics here are a little bit more complicated and I wanna note I use two pieces of software. One is called MS Fraggar, the other is called Metamorphies. And I wanna thank the, this is from the Lloyd Smith group in chemistry. And I wanna thank Mike Shortread in particular for helping us get this set up in our lab. And so some things to note here, kind of analogous with methionine, the methionine data, we see fewer modifications overall in Argon, as you can see here. And this tracks with the lower reactivity that we saw on a peptide level with the classical modification for Argon. Interestingly here, something that we did not see with methionine is a significant overlap in modification. So mass shifts identified between nitrogen and air. And this is very cool, it makes total sense. Air is 80% nitrogen, as I said before. So this might be something we would expect when looking at a data set like this. Similar to methionine, we get modifications that are unique to each gas plasma use. And we get modifications that appear to be shared between all of them. If I go in and try to assign, elemental and product composition to these or figure out what these modifications might be, there's only a few that I can find literature citations for. And they are circled here. Minus 22 and minus 10 are oxidative histidine ring opening, pretty widely published. And minus 32 is an oxidative loss of the thiogrub on methionine. So we don't know what the composition of the rest of these modifications are, and this is pretty much where we are with that line of experimentation. We're excited to follow up on it. So to summarize that little section, both methionine and protein yield known modifications, known oxidative modifications I should say, both methionine and protein yield far more unknown modifications, things that we cannot find literature citations for, things that as far as we know haven't been reported. So this leaves us with a lot to pursue. What are these modifications? How do we go about figuring out what they might be, but really with the explicit purpose of pushing the technology forward and making it have more utility, can they be used to paint a larger picture of protein solvent accessibility given that we only get modification of a subset of amino acids with hydroxyl radicals, but we may be generating a plethora of other radical species with these differential gas plasmas. The last thing I wanna say is we have already built a better gas exchange setup to do this, and this is data from Josh Blatz in Leon Schohet's lab. So instead of the kind of semi crude system we have before where there's feed gas in the back and the top might not be super airtight, he has built a totally airtight cuvette cap with a gas flow system inside that feeds gas into the airspace up here. And he's designed an electrode that will generate plasma from the tip as with our other needle, but it's hollow and can continuously sample the airspace in that tube as we replace the gas. So we can actually analyze what the airspace composition is at any point. And Josh has already tested this and its ability to sample gases. And we've got 95% or greater enrichment for helium, nitrogen, argon when we use this new setup. We don't have any amino acid data or protein data with this yet, but we are excited to get started. And he's got a little bit of dosimeter, like little chemical dosimeters for radical generation that I'm not gonna show here. But at some point in the near future we're going to get to doing amino acid and protein work with this new setup. So I will finish and summarize by talking about the current and future states of Klim. So I talked about the idea for the method, how we built an optimized instrument for getting to the point where we could expose biological samples. I've given you some examples of those biological systems we've used this for. I've mentioned and I wanna mention again that there are a number of other academic and industry collaborations that Dan and Faraz have taken on that I didn't have time to talk about today but have yielded wonderful data as well. I showed some work that Josh and I did towards building a system to get a differential gas plasma modifications of proteins. Yeah, and so that is where we are. So with respect to the future, I can't remember if I disclosed this up front. So I'm gonna disclose it again. This technology, what I've told you about today has had a patent that's been assigned to Worf and is working towards being licensed by Amidoscientific Incorporated, a company that our team founded in 2018 to commercialize this technology. So that is something that's currently happening. I am very excited on the academic side to benchmark this new gas exchange system with a number of biological systems. Moreover, exploring some absolutely bonkers chemistry using the new gas system. I would say that perhaps unsurprisingly, the field of plasma interacting with biological molecules is not as large as you might think. And so we are, I guess we're heading into uncharted territory here, which is super cool. So we're very excited to explore that specifically with the goal of expanding modifications we can use to measure protein-solving accessibility and ascertain structural features of proteins and solution. So I'll finish by acknowledging the Susman Lab, current and past members, far too many to list. They have made everything possible. The Biotech Center Core Facility, the Greggs have given me wonderful mass-spec advice over the years. Lee and Showhead's lab, they took this idea and made it like a practical reality. So thank you very much to them for a fruitful collaboration over the years. I wanna mention specifically Dan Benjamin and Josh Blatz as they have defended their theses within the past month. So they are newly minted PhDs. So congratulations to them. Funding for this work came from NSF, Worf and the SRC. The company Protein Metrics provided us software to analyze some of these data sets. And I'll finish by again thanking the Department of Biochemistry for selecting me for this Boyer Award and for enabling the science that I've talked about today and a number of other great scientific endeavors in other labs. So thank you all for listening and I'm happy to take any questions. You're on mute yourself. Yeah, right. So first, I wanna thank Ben, I can't see if anyone else is applauding, but I can pull up loud enough for everybody. All right. Oh, there's Brian. Good. So- I have a question. Lloyd has a question. I have a question for Ben. Oh yeah. All right. Here I am. Hey Ben, good talk. Thanks. Yeah, thanks. Thanks for the software and all that. I'm a little confused about like two phases. There's a gas phase like above the G and a liquid phase in the G and the plasma like the most like visible plasma is generated in the gas phase. But then you're talking about how the nature of the gas phase affects the ions in the liquid phase. So I'm just sort of, does that mean that you're actually working with gases that are dissolved in the liquid phase and that is how you're getting those molecules? So there's a few things to talk about there. Number one, Mark Chance, who does the synchrotron rate radiolysis has shown multiple times over that it is not just the water that generates the species that we're seeing but it's actually dissolved oxygen in the water itself. So the mechanism is generally thought to be a combination of water molecules being broken apart into hydroxyl radicals and solubilized elemental oxygen jumping on to side chains as well. So yes, it is dependent upon the gas in the liquid itself. When I talk about the system that we've built we do not, we do not sparge the water at all. So we still have solubilized oxygen in our liquid and that could very well be why with all of the gases we've used we see a significant amount of classical hydroxylation. It's a very reasonable hypothesis. The other thing I will mention is that we know that when we hit our sample with plasma we are getting chemistry that's happening in that liquid. I don't have the data here but when I use an unbuffered water solution containing bromothymol blue, a pH indicator that as it acidifies turns yellow and we hit pH near neutral unbuffered water with the plasma, we see a significant amount of kind of spiraling down pH change that actually very likely happens by nitrate and nitrite ions in the water itself. So we know that there's interaction of that plasma with the bulk solvent itself in some capacity. Does that answer some of your question? Yeah. Ben, could you maybe unshare your screen? So there'll be... I do have supplemental slides in case, you know... Okay, we can reshare but I think the question is a... Brian, did you have a question or... Brian? Or let's see, in the list, there was a... Yeah, I did. I put it in the chat but my connection is just really bad. Concentration protein. Yeah, yep. So we're working... So that BSA work was with, I think, 10 micromolar, possibly 15 micromolar, admittedly relatively large and we also were treating it without to 60 seconds of plasma, which is very intense. The thrombo work... So do you see protein dimerization happening, radical, radical coupling of... We're not sure... That's not a thing that we've looked at. What I can tell you is that when I run a gel, when we hit the BSA, we don't see induced dimerization. So if anything, what we would very likely see is radical mediated cleavage of the protein. I think that's more likely to happen. We do, to a very small extent. When we treat it, however, when we treat proteins, as I showed you in the Thrombin experiment, so these are much lower concentrations. This is like high nanomolar but we're only treating it for a few seconds rather than 60, we don't see the same cleavage. So it's time dependent, perhaps unsurprisingly. And if we drop the concentration and we drop the treatment, such that we still get labeling, we don't see that cleavage. Thank you. Good talk. Good talk. Thanks, Mark. That was very cool. So who else, I think there was one other question. Mark Richards had a question, it looks like. Go for it, Mark. Okay. Yep, I think I'm unmuted. Can you hear me? Yep. All right, yeah. So with the, like a nitrogen atom radical, if that would add, that would be plus 14 as a guess. Is that something you've come across? I just always think, yeah. So the thing is we can't differentiate nitrogen from carbonylation because it's plus 14. So C double bond O dropping out to H's gives you the same mass adduct. So we can't differentiate that. And one of the things that we probed in those gas samples was, do we see something as simple as O and argon that modifies our protein? And the answer is no. Unfortunately, the chemistry seems to be a lot more complicated than just, oh, we have a radical species. Oh, it modifies our protein, which would be great. If that were the case, but guess not. Before we all leave, I do have one comment. While everyone's here, if you're interested, things we could use some help with. I hope you don't mind me saying this, Ben, but we're not well funded in this area at all. But we'd like to explore molecular dynamics. From my simplistic point of view, there's 25,000 proteins in a cell and they're each gonna fold and unfold at different rates. And since we can control the pulse with of the discharge, and since hydroxyl radicals are so short lived, there might be some interesting uses of this technology to work when one works with proteins that have very drastically different dynamic conformational changes. The other thing I would like to help with is deep scanning mutagenesis, if Watson or Phil Romero, I know you guys are listening. I would love to, P53, in response to the reviewers, we didn't get funded the first time, they wanted to see something more biologic relevant. The first time was just methionine and it was real chemistry, but they wanted more biology. So then we resubmitted with P53, but then they came up with other reasons they didn't want to fund it. So, but we're stuck on P53 and I think, and that's a good thing to be stuck on. This is a very important protein and it's easy to produce. Okay, I'm gonna shut up now. Does anybody have any other questions or comments or anything? There is one that just popped up in the chat. Is it feasible to tune, Plym and gas composition to introduce PTMs in your recombinant protein, study their effect on protein function? Man, if you could resolve them, right? If you could somehow generate a homogenous population and resolve them, then yeah, I don't see why not, but I think that the trick would be only modifying what you care about and getting a homogenous population out of that for actual study. I will say, so I don't have, man, I didn't put the data in here, but I looked at the effects of Plym on lysozyme activity and with longer exposures similar to cleavage, much longer exposures on the order of BSA, we do see a drop in lysozyme activity. So we are either unfolding or modifying if we go out to like 60 seconds, 90 seconds, as compared to the few second treatments we hit proteins with now. So that's an example of at least oxidation probably on protein function, yeah. Well, if there are no more questions, I wanna thank Ben again and also congratulate him. If there are more, feel free to email me. Someone just directly message me a question, but we can take it offline. DeBion, feel free to email me and we'll talk about it. Thank you again, Ben, congratulations. Yep, thanks everyone. Thanks again to the department. Bye. You're welcome.