Scientist Stories: Edward Chouchani, Chemoproteomics and Targeted Therapeutics for Metabolic Disease

On behalf of the Department of Molecular Metabolism, who hosts the annual award, I'd like to welcome everyone to the 11th Annual Tashjian Award lecture.

We're here today for several reasons. The first is to honor the memory and legacy of Dr. Armen Tashjian, an extraordinary scientist and neuroendocrinologist and an outstanding mentor of many, many trainees, and you will hear more about Armen in a few minutes.

Second, we're here to award the Taschen Awards Recognition of Excellence to Dr. Ed Suchany.

Congratulations from me, Ed, for his outstanding work on mitochondria and thermogenesis, and we'll hear about Ed and we'll hear from Ed shortly.

It's now my pleasure to introduce the Dean of Academic Affairs, Jane Kim, who will make a few remarks on behalf of the Chan School.

Thank you, Bob. And thank you all for being here this afternoon. It's my great pleasure on behalf of the school to welcome you all to this 11th annual event honoring the many accomplishments of the late Dr. Armen Tostian, Professor of Toxicology in what was the Department of Molecular and Cellular Toxicology at the Harvard School of Public Health. The purpose of this event is to honor the recipient of the Armen H. Toshjian Jr. Excellence in Endocrine Research Award. The Armen Toshjian Award was established to help junior faculty members with a primary appointment at the Harvard T.H. Chan School of Public Health, Harvard Medical School, Harvard School of Dental Medicine, or other related departments at Harvard pursue innovative research ideas and basic biomedical sciences relevant to endocrinology. The award is made on an annual basis, and recipients are selected by a committee of distinguished faculty and academic luminaries. I'd like to acknowledge Suda Biddinger, Associate Professor of Pediatrics at Boston Children's Hospital and Harvard Medical School, and who was the 5th Tostian Award recipient; David Golan, Dean for Research Operations and Global Programs George R. Minot Professor of Medicine and Professor of Biological Chemistry and Molecular Pharmacology at Harvard Medical School; Gokhan Hotemishlagil, James Stevens Simmons Professor of Genetics and Metabolism and Director of the Sabri Ulker Center at the Harvard Chan School; Brendan Manning, Professor of Molecular Metabolism at the Harvard Chan School; and Yuhua Zheng, Associate Professor of Medicine at the Joslin Diabetes Center in Harvard Medical School, and who was the 7th Taschian Award recipient. We thank you so much for serving such an important role in selecting this year's award recipient. We'd also like to thank those whose generosity has made this annual event possible, particularly Armen's wife, Carol Taschian, his daughters, Amy and Elizabeth Taschian, and their husbands Donald Brown and Uri Lowenstein. We at the school are so pleased to be able to celebrate and honor Dr. Tasjian by acknowledging promising young scholars with an award in his name. Now it's my pleasure to pass the microphone to Dr. David Golan, who will share a few words about Dr. Tasjian.

Well, thank you kindly, Jane, for your generous introduction, and welcome to, to everyone. First, I want to offer my own heartiest congratulations to Ed Tschuciani, the winner of the 2021 Taschen Award for Excellence in Endocrine Research, and thank you for this opportunity to remember Armen. It's wonderful that the Taschen family are here with us again to celebrate this occasion. Warmest of greetings to Carol, Amy, Donald, and any other family members who are attending this virtual, unfortunately, event. Armin was my very dear friend, colleague, and mentor for more than 3 decades. We published papers together, taught courses together, advised companies together, and eventually wrote a textbook of pharmacology together. I believe that I was privileged to know just about every aspect of Armin's remarkable professional life. As many of you know, Armin had a long and distinguished career in biomedical science. With research and teaching accomplishments and recognition across the spectrum of pharmacology, toxicology, endocrinology, metabolism, and cell biology. His laboratory made a great many contributions to our fundamental understanding of pituitary hormone regulation and calcium homeostasis. Armin was also the prime motivating force behind the development of several important drugs used in the treatment of osteoporosis and breast cancer. This is an almost unique statement about an academic scientist as accomplished as Armen, virtually all of whom would be overjoyed for their laboratories to have been as productive as Armen's was, and to have even one of their discoveries be transformed into a clinically useful drug by someone else. But Armen not only directed a lab that produced academic work of the highest caliber, he also translated more than one discovery into a clinically useful therapy himself. This is a truly remarkable achievement. As Jane and Bob have already mentioned, Armand guided and mentored hundreds of young scientists and physicians over several generations, including some in this room. I'm sure all of us who knew him remember the first time we saw, with a bit of a shock, Armand's trademark edits and comments filling the margins in red ink and reluctantly we had to admit that he was virtually always right. For those of us who were lucky enough to know him and work with him, Armin had many qualities that defined his essence and in combination made him the wonderful friend, colleague, and mentor he was to so many of us. He had a love of science and medicine. He had an encyclopedic knowledge base. He had a voracious appetite for the literature. He had an infectious enthusiasm for drug discovery. A keen appreciation for analytic rigor, and a prodigious work ethic. And most importantly and endearingly, Armin had a genuine warmth and excitement for people. I know that he would have loved to have gotten to know Ed and all of you in this symposium and would have thrilled to listen to the science we're about to hear. I would now like to introduce my friend and colleague Gokhan Hotemishnaghl, who will introduce the Tashjian awardee. Gökhan is the James Stevens Professor of Genetics and Metabolism in the Department of Molecular Metabolism at the Chan School. He's also director of the Olga Sabri Okhla Center for Nutrient Genetic and Metabolic Research at the school and an associate member of the Broad Institute and the Harvard Stem Cell Institute. His laboratory studies the regulatory pathways controlling glucose and lipid metabolism and disorders in these pathways that lead to human metabolic and inflammatory diseases. Gökhan, please come to the virtual podium to introduce our awardee. Thank you very much.

Thank you, David. Thanks for reminding us the great, great legacy of Armen, and I also welcome the Tajian family, Carol, Amy, and all of our guests, colleagues, students, fellows, Dean Kim for kindly opening today's events. It's really, it's a tremendous pleasure to honor Armen Tajjan today on the occasion of this award. And this is actually even under the most challenging circumstances that we are living right now, it is still a very joyful thing. It brings a big smile on my face just to remember Armen, but also the chance to be able to celebrate our really most cherished treasures, even though they're not with us at the moment. And this, I mean, to me, it's very important for our younger generation students and fellows and rising stars to see that you're not just not a passing element or a vehicle in the advancement of the institution, but you're part of the fabric of this culture and you will be remembered. Your legacy will be remembered and your contributions will be treasured. And we hope to continue this tradition and really preserve some of these sacred treasures of academic life. And for that reason, I feel myself really privileged. I had the joy of actually starting this award program together with the help of the family and, of course, their generous contributions, as well as many friends Armen had from every walk of life. I mean, fortunately, in many of them in industry investment communities, all the way to writers and painters, and very, very diverse group of really intellectuals. As really Armen was in love with the activities that happened here, no matter in which framework they operated. And so I think, I mean, sadly, I will also, maybe this is an opportune time for me to say that I will not be actually involved with launching the ceremony. So the department and the school is the most appropriate place, of course, for this to reside. So this will be my last ceremony, so I will have a severe withdrawal, I think, experience come next year. And so I met Armen, of course, as an assistant professor at the time that I was very anxious and insecure and didn't quite know what to do, and really remained very close with him until his death. And what David said, of course, oftentimes we will say great things about those that are not with us right now, but it is really true. Armin's intelligence, I would say, it was so much that it sounded at times it could be really annoying. And which is the feeling that I had many times. And I am certain if he was alive, just like David said, he would have become an expert in COVID and start learning, you know, viral life cycles and how to tackle these problems and probably also epidemiology and other aspects of this pandemic. And I want to end by saying one really critical thing about Armen, or maybe two if you would allow me since this is my last session. The first one is, if as a basic scientist you are enjoying what you're enjoying in our school, and if a school of public health like ours is really leading the chart among all other schools of public health, there is a really important component underlying that, which is the presence of basic science, fundamental science integrated into all other disciplines of public health. And this is one of the greatest legacies that Armen really left to us. So if you're enjoying this current department, that effort came from Armen's vision. And because he really transformed, I think, the way we did business at the school and crystallized the fact that issues of extreme importance to public health are best modeled, understood, translated, and eventually resolved on a fundamental science-based platform. And perhaps, I mean, this extremely challenging experience that we are all going through right now really takes home this message. And I don't think that his vision has never been as pertinent as it has been this year. And the second is, I would mention that Armen also recognized at least 30 years ago that I am aware of the inequalities and problems in the academic structure, whether that is advancement of women in scientific careers, gender inequities, and presence of racial and ethnic minorities. And I remember this, that 30 years ago he was articulating this in a very vibrant and loud manner in every faculty meeting that we had. So if we have covered some distance in that area, we should also, I think, owe him a tremendous debt of gratitude. So with that, I am extremely happy that we are honoring Ed Tchouchani as the recipient of the Armen Tajjan Award. There could not have been something that will make Armen more proud to see the emergence of such a potent scholar in such a relatively rapid sense. In our community, and we would have been— of course, he would have been thrilled to see you as part of the Harvard community as well, Ed. So without really making this long, I will just say that Ed is currently an assistant professor in Harvard Medical School and in cell biology. He has achieved his doctoral degree at University of Cambridge and MRC, predominantly focusing on mitochondrial biology and biochemistry. And then he did his postdoctoral fellowship here at Dana-Farber Cancer Institute with Bruce Spiegelman, and soon after started his independent career. And so he is a very diverse and interesting program studying bioenergetic processes and some new really substrate-energy relationship, and also platforms that serve broadly to many interests in, in the community. And I mean, if I have one really regret from the time that I served as the chair of another department was to miss the opportunity to recruit you to school, Ed. So, but we accept our shortcomings perhaps to make up for our failure at that time, I'm extremely pleased to offer you this package now, which has an undisclosed amount of award money in it. And I think the medallion of Armen is already delivered to you. So it is an absolute pleasure to present you this award. Ed, and I look forward to your lectures. Congratulations.

Thank you, Gökhan. Thank you, everybody. Thank you for that very kind introduction. It's really an honor to receive this award, not least because of Armen's legacy that we just heard about, but also looking at the past winners of this award. These are all individuals whose work I hold in the highest regard, and it's really a privilege to to be considered alongside them. I also think this award highlights the really special research environment that the Harvard community has in terms of metabolic research especially. And it's a community I'm really excited to be a part of. And I think this award really highlights that. The talk I'm going to give today starts with a question, which is a very simple question— why succinate? And this is a question that has come to the forefront of our minds in my lab because of a, a general view we've been focusing on, which is the idea that the bioenergetic state of cells can be an important mode of eliciting cellular adaptation and physiological adaptation. And we've been interested in exploring that question generally because Built into that premise is the idea or realization that most adaptive processes, essentially every adaptive process in a given cell or given tissue, necessarily requires specific metabolic parameters to be fulfilled to elicit that adaptive process. So we've been curious in trying to understand whether specific bioenergetic parameters, specific metabolic intermediates play essential roles as regulatory species in physiological in pathophysiological contexts. And to address those questions, we've been interested in developing mass spectrometry-based technologies to understand mechanisms of metabolite regulation. And that's led us to focus in on this particular metabolite, succinate, mitochondrial TCA cycle intermediate, which has emerged as an apparently important general regulator of physiological adaptation. And what I'd like to talk about today is how our very recent work has led us to begin to think about the principles of regulation by succinate, the logic of why this metabolite in particular appears to be well-suited as a general regulator of physiology and pathophysiology. So by way of introduction, what I'm showing you on the first slide here on the left is a sampling of some of the physiological models that we and our colleagues and collaborators have been studying in recent years. What I hope to convey with this slide is that this is— these are all very distinct biological adaptive processes involving highly specialized cells or highly specialized tissues to elicit a highly specialized adaptation. So for example, we've been interested in studying how brown and beige adipocytes become activated to drive this process of thermogenesis, which is the process of rapidly catabolizing stored fats and circulating glucose, which can be potently protective against obesity and diabetes. Separate work we've been interested in understanding the mechanisms through which macrophages produce particular cytokines, both pro- and anti-inflammatory. And in separate work I'll be talking about today, we've also been interested in studying how muscle in the process of exercise adapts and remodels to become more glucose and insulin sensitive and stronger. And so these— we've been studying these processes using small molecule metabolite profiling approaches to basically systematically track how metabolism changes during the given adaptive process.. And what we've been surprised to find is that all of these very different adaptive processes have a common metabolic signature associated with them. And that metabolic signature involves the accumulation of this mitochondrial TCA cycle metabolite succinate. This deceptively simple 4-carbon, 4-oxygen dicarboxylate. And the signature we observe is something like what we see here on the right. This is just a representative example taken from some of our recent work on brown adipose tissue biology. This is an experiment where we take a very standard model of acute activation of this process of thermogenesis genesis in brown adipose tissue. You can achieve this just by putting a mouse— subjecting a mouse to cold temperatures for a short period of time. And then we profile the steady-state levels of mitochondrial metabolites and metabolites proximal to mitochondrial metabolism. And you can see there's this striking and selective increase in succinate. And this is the type of signature we observe in all of these settings. And the reason that this has interested us is not just because it occurs, but because, as is reported in all of these papers, if we or our colleagues, we use genetic or pharmacological tools to prevent the accumulation of succinate, this abrogates often, if not completely, the adaptive response we're studying, whether it's activation of thermogenesis, induction of pro-inflammatory cytokines, the capacity for muscle to remodel in response to exercise. And so these striking observations have led us to ask two sort of major and related questions. The first question is the, the why succinate question. So What is it first of all about succinate's positioning in cellular metabolism that renders it amenable to selective accumulation as opposed to some other metabolite? And another way of wording that question would be to ask what are the common bioenergetic parameters that control accumulation of succinate over other metabolites? What's the logic of that accumulation process? We think that by understanding that question in principle, we can begin to think about what succinate accumulation is a specific response to. What's it a signal for? And then the second related question that follows on from that first one is, once succinate accumulates, what are the mechanisms through which accumulated succinate can elicit such a broad range of highly context-dependent adaptive processes depending on the cellular tissue context in which it accumulates? And that's what I'm going to spend most of my time talking about, how we've been beginning to uncover the mechanisms through which accumulated succinate can regulate physiology generally. So I want to touch on the first question first, which is the question of what are the common bioenergetic parameters that control succinate accumulation. This is a question that we think we have a satisfying answer for, and I'm just going to briefly touch on. I mean, this is really the summary of some work from us, but a lot of work from other groups, especially Mike Murphy's group, who was my PhD supervisor at the University of Cambridge. And what we've come to understand in the field is is distinct about succinate comes down to its unique relationship or distinct relationship with the mitochondrial ubiquinone pool. So we all recognize and know that the mitochondrial Q pool is this major reservoir for electrons in the mitochondrial respiratory chain. It shuttles electrons from upstream dehydrogenases to downstream oxidoreductases complexes 3 and 4 in such a way as to generate a membrane potential that then goes on to drive ATP production in mitochondria. And what's interesting about the Q pool from the perspective of succinate metabolism is that it is the furthest upstream common electron reservoir for any reducing equivalent that's going to contribute to mitochondrial ATP generation. So if you think of all the reducing equivalents in the cell that could contribute to mitochondrial ATP generation—glucose-derived carbons, amino acids, fatty acids—those electrons are going to contribute to that process. There's a common reservoir where they all have to meet, and that common reservoir is the Q pool. And so that means that the Q pool redox state has some interesting features associated with it. It means that the, the number of electrons stored in the Q pool is actually subject to a small number of highly central bioenergetic parameters in any given cell. The first parameter is the aggregate electron supply parameters, so the aggregate electron supply agnostic of source— fats, pyruvate, amino acids, etc.— pushing into the Q pool. And then on the demand side, the number of electrons stored in the Q pool is subject to the two bioenergetic parameters that dictate the rate of CoQ oxidation, specifically oxygen tension, which is responsible for pulling the electrons through complexes III and IV. And because those reactions are coupled to proton translocation, the transmembrane gradient of protons also specifically affects the CoQ pool redox state. And of course, that's an indirect manifestation of the rate of ATP utilization. The less ATP demand there is, the higher membrane potential. You'll have in mitochondria. And so these three parameters converge on the CoQ pool redox state in a way that's rather interesting from the perspective of succinate metabolism. Because while this CoQ— while the CoQ pool is a central redox couple in the cell, it's also a highly hydrophobic species that's trapped in the mitochondrial inner membrane. So all this information about the redox state of mitochondrial respiration is trapped in the mitochondrial inner membrane. But distinctly, the CoQ pool redox state is directly connected to the succinate pool via the succinate dehydrogenase enzyme complex. In a way that for which the reaction has a delta G of 0. And so what that means is that the rate of succinate utilization by succinate dehydrogenase is extremely sensitive to changes in the coq pool redox state. And so we and others have observed is that small differences, small imbalances in supply and demand in the coq pool are really the only upstream requirement that will control levels of succinate accumulation. So for example, cases where there is decreased electron demand for the coq pool electrons, in the cases of, for example, transient hypoxia, where oxygen tension is low lowered, or alternatively, in cellular cases where ATP/ADP ratios are high because the ATP demand is low in the cell in the face of high electron supply, these imbalances are sufficient to drive highly selective succinate accumulation simply because of this delta G zero reaction. An additional feature that renders succinate amenable to accumulation in this way is that there aren't other major catabolic processes that can utilize succinate at appreciable rates in the cell. So when SDH oxidation is not favored, succinate just continues to accumulate and this way. That's how we understand the logic of succinate accumulation so far, in that it appears to be a manifestation of acute imbalances in electron supply and demand and doxidata phosphorylation, which in and of itself is a bioenergetic parameter that presumably all cells, all mammalian cells care a lot about. So once succinate accumulates, we then have this sort of more open-ended question of how does this simple mitochondrial metabolite elicit such a broad range of adaptive processes depending on the cellular and tissue context. This is something we've been working a lot on in recent years. We focused in on two major pathways I'd like to talk about today through which we understand how succinate regulates physiological adaptation and how we've been developing methods to try to understand the mechanisms through which that occurs. So the first mode we've come to appreciate about accumulated succinate with regards to how it can regulate physiological adaptation comes down to its ability to act as a direct metabolic control point for the controlled production of redox-active metabolites, so-called reactive oxygen species and related molecules. And so this is really— this cartoon is the culmination of several years' work from our lab and other labs, especially Mike Murphy's lab, showing that accumulated succinate in cellular contexts where additional bioenergetic parameters are filled, specifically high oxygen tension and high membrane potential, can allow the electrons stored in succinate to engage in an electron transfer reaction that is not normally favored, but it becomes favored under these conditions. And that is a process called reverse electron transport, wherein the electrons stored in succinate are pushed through succinate dehydrogenase, through the Q pool, and onto a second dehydrogenase complex called mitochondrial complex I, where the electrons that were once stored in succinate eventually become poised on this terminal flavin of mitochondrial complex I, where they're poised to perform a single electron reduction reaction with oxygen, to generate superoxide, this proximal reactive oxygen species that then goes on to generate hydrogen peroxide. And this has been an interesting observation because we found in many of the models I showed you on that first slide that this controlled production of redox-active metabolites appears to be essential for the signaling of succinate in many of the contexts, but not all, but many of the contexts in which succinate accumulation is regulating adaptation. And so it's been— this has been These discoveries have actually contributed to a changing view in the field of redox biology that I think has become the dominant view in the last several years, which is a refutation of the classical view of redox-active metabolites, reactive oxygen species in physiology. There's been this classic view that these molecules are unregulated and undesirable byproducts of dysfunctional mitochondrial metabolism and something that should be avoided at all costs. And what we and others have clearly observed is that The opposite is true. These molecules are being produced all the time in healthy tissues. We now have tools that allow us to actually study them reliably in healthy tissues. And they're perhaps most— more importantly, subject to strict upstream regulation by defined metabolic parameters like succinate accumulation and mitochondrial membrane potential. And moreover, the controlled production of these metabolites appears to be critical for eliciting a range of adaptive processes that are cell-type dependent. And so in some ways, this has been— this was interesting and exciting, but in other ways, it sort of kicked the can down the road because it's forced us to ask the question, well, what are the defined mechanisms through which the controlled production of these redox-active metabolites actually then go on to control effector processes or cellular adaptive processes. And this has been a question that's been an interesting paradox over the past several years because we've known the general signaling principles of redox-active metabolites for several years now, and specifically the fact that these molecules have the specific capacity to reversibly covalently modify a particular amino acid side chain, the cysteine, thiol sidechain on proteins and can affect protein function in that way. So we've had a general view on the importance of this for some time, and it's schematized on this slide here, wherein redox— the controlled production of redox-active metabolites can reversibly covalently modify certain cysteine residues, which can elicit a range of reversible covalent modification that has been shown in certain cases to affect the structure of the target and therefore in certain cases enzymatic activity, binding affinity, localization. There are examples of each. And importantly, as I mentioned, this process is completely reversible through the endogenous thyroidoxin and glutathione systems in the cell. So one way of thinking about this type of regulation is a way akin to protein phosphorylation. Instead of having a kinase and an ATP controlling the reversible covalent modification of some effector protein, what you have is the controlled production of a redox-active metabolite due to, for example, accumulated succinate, controlling the production of this metabolite that then goes on to reversibly covalently modify some set of effector proteins to affect biological adaptation. And so that's the principle. And we know that this class of post-translational modification is important because we have tools that allow us to manipulate the entire family of modifications. But what we have surprisingly, until recently, have had surprisingly little information on are the specific cysteine targets that are subject to modification this way across the proteome. We don't have the— we haven't had until recently the equivalent of phosphoproteomics, for example, for this cysteine proteome field. Field. And the reason for this has really been that cysteine residues, unlike other amino acids, their post-translational state cannot be directly monitored by standard mass spectrometry technologies. And so one major goal of ours in recent years has been to try to develop a mass spectrometry-based technology that would allow us to quantitatively map the landscape of these type of post-translational modification for the first time, to begin to try to think about general principles of this type of signaling, and also to begin to define the protein families that are subject to this type of modification to understand the mechanistic basis for this type of regulation in different physiological settings. And so the methodological requirements that we were hoping to achieve are summarized here, wherein we wanted to develop a method that would allow us to assess the stoichiometric modification of a particular cysteine in a biological sample, the absolute quantifications, the percent modification. We wanted a method that would allow us to do this not only for one protein or a small subset of the cysteine proteome, but a method that would allow us to simultaneously assess this information for tens of thousands of cysteines across the proteome in a single experiment. And we wanted a method that would allow us to report on these parameters in living tissues and in living systems, so we could have a view of the endogenous regulation landscape. So what I'm going to show you now is a method that fulfills all these criteria and has allowed us to begin to explore the general principles of this type of signaling. And so the method that I'm going to talk about today is based on a new reagent that we developed that we call a CPT tag. These are short for cysteine-reactive phosphate tag, and these new tags have 3 special bits to them. The first special bit is a cysteine-reactive warhead. This is basically an iodoacetamide group that, when introduced in— allows for these molecules, when introduced into a complex biological sample, to indiscriminately label every available cysteine that is not modified in the biological sample. We'll basically blanket label every available cysteine. We then include in this tag an isotopic linker, which allows us to barcode different versions of this tag so they can be differentiated by mass spectrometry. And the third and perhaps most important new part to the tag is this metal affinity chromatography moiety, this moiety that allows these tags to be purified by metal affinity chromatography methods. This is an extremely mature enrichment method that's been used for many years in the phosphoproteomics world and really is one of the major reasons why we're able to achieve deep phosphoproteomes using mass spectrometry today. And we reasoned that we could apply these same enrichment technologies to achieve better coverage of the cysteine proteome. And so what I'm showing you on the right part of this slide here are recent publications from our group and others that use the best, the at the time current best methods to map the cysteine proteome for endogenous modification. You can see that the best that we or others could do was to map a few hundred to maybe a couple of thousand cysteines across the proteome. Very small proportion of the cysteine proteome. And with these CPT tags, what we were able to do was obtain over an order of magnitude improvement in coverage of the cysteine proteome, mapping tens of thousands of cysteines in a single experiment. And for the first time, really, giving us the ability to map the modification status of very low-abundance proteins systematically. So transcription factors, transcriptional regulators, low-abundance kinases, proteins like that. So with this improvement in coverage, we devised a workflow that would allow us to map the quantitative modification state of cysteines across the proteome in living tissues. I'll just briefly walk through how that works. We start with a biological sample of interest. It can be anything, really. We used mouse tissues in this study. And we would use a chemical trick to stabilize the endogenous post-translational state of the cysteine proteome in those tissues and then split the biological sample into two parallel channels. That would be worked up in slightly different ways. So the first channel would report on the total cysteine signal in the biological sample. So what we would do in this, in this sample is we would take the proteome and we would treat it with a thiol-reducing agent, which would remove all the reversible cysteine modifications across the proteome. And we do this in the presence of our CPT tag. So in this sample, all the cysteines, modified or not, are labeled with a CPT tag. In the second sample, we do something slightly different. To the same biological sample, we first treat the sample with iodoacetamide, and that labels cysteine cysteines in such a way that they're rendered invisible to us. We can't see them. And then we come in in a second step with TCEP, the thiol reducing agent, that removes the reversible oxidation modifications, and we do that in the presence of CPT. And so in the first channel, every cysteine is labeled with CPT. In the second channel, only the reversibly modified cysteines are labeled with CPT. We then digest the samples in parallel with trypsin. We purify the cysteine-containing peptides using metal affinity chromatography because of the the CPT tag, and then we barcode the samples using TMT reagents. Many of you will be familiar with these isobaric tags that are chemically identical but can be differentiated at the point of fragmentation on a mass spectrometer. These are really powerful for our purposes because they allow us to recombine these two samples into a single tube and then shoot it on a mass spectrometer. This makes the peptides from sample A identical to the peptides in sample B until the point of fragmentation, where we can then differentiate the stoichiometry of contribution from sample A versus sample B. And so the way that this workflow— what this workflow allows us to do is it gives us two important pieces of information. We identify tens of thousands of cysteines in the experiment, but perhaps more importantly, by quantifying the relative abundance of the peak pairs for each of those cysteine residues, we get an absolute measurement of stoichiometry of modification for that given cysteine residue. And I'll just mention that TMT comes now in 16 different flavors, so in fact what we can do is 8 biological replicates in a single mass spec shot, so it actually it can be transformed into a relatively high-throughput analytic workflow. So for the sake of time, you'll have to take my word for the fact that we've done a lot of characterization of this method to convince ourselves that it reports accurately on the endogenous modification status of the cysteine proteome. And what I want to show you is one of the first datasets we generated using these tags that was published last year. Called the OxyMouse dataset. And so the OxyMouse dataset is really a reflection of us realizing that when we generated these tags, we had the opportunity to really just obtain the basal tissue-specific cysteine proteome modification status across tissues in mice, which we reasoned would be a good benchmark dataset for us to begin to understand some general principles of reOx signaling and physiology. And so what the OxyMouse dataset is, is effectively our application of the CPT tags to 10 different tissues in young and old mice. 5 biological replicates each, which eventually resulted in the generation of this very large dataset including several hundred— now thousand— cysteine quantification events. Importantly, the large, large majority of which had not been previously reported in the literature. So we think there's a huge amount of new biological information here that we've been using to try to understand specific modes of redox signaling in different tissues. And so what I'm going to do is just describe a couple of high-level insights we've gained from this dataset that hopefully will illustrate how this dataset can be used by others to not only understand, as I said, general principles of redox signaling, but also to dig in on particular biological processes that this dataset has uncovered as being potentially redox regulated. So the first insight that actually surprised us from this dataset was the insight that redox regulated protein networks exhibit a very high degree of tissue specificity. So one of the, one of the things one can do with a big data set like this one is to perform population-level analyses of the type I'm showing you on this slide. You can ask questions like, within a given tissue proteome, what's the average percent modification of the cysteine residue across the proteome? When you ask questions like that, you get very similar answers regardless of the tissue you're looking at. So you can see the average percent modification of a liver cysteine, it's about 10%, about the same in kidney, as high as you go is about 15%. Fat. So very similar. When you break the proteomes down by subcellular location, again, you see very much the similar features from tissue to tissue. You can see that the average percent modification of mitochondrial proteins in any tissue is very, very similar at the bulk level. The same is not true for secreted proteins, but that's something that I won't have time to talk about, but it's potentially interesting. Another question one can ask is, within a given tissue proteome, what proportion of proteins have at least one highly modified cysteine residue? In this case, we define highly modified as over 20% modified. And you can see again that the general answer appears to be very similar across tissues. You can see in almost every tissue the large majority of proteins don't have a highly modified cysteine residue. About 20 to 30% of proteins have at least one, and far fewer have two or even three. And so when you look at the bulk characteristics of cysteine proteomes from tissue to tissue, things look very similar. And one might think that those highly modified proteins in the liver would be those same highly modified proteins that are found in the brain. Maybe a A highly oxidized cysteine is just always highly oxidized. And what we found was actually the exact opposite was the case. And so when we actually dug into this data at the site level, what we found was that many of these highly modified proteins in a given tissue exhibited a very high degree of tissue specificity. And so one way of visualizing that is to take any two cysteine proteomes and score them as a comparative redox proteomics experiment. So in this plot, we're comparing the brain cysteine proteome to the liver cysteine proteome. Each dot is a cysteine residue, and they're scored on the x-axis in terms of difference in percent modification. And what you can immediately see is that there are clusters of highly modified cysteines in the liver that are distinct to the liver and a completely different set of cysteines that are highly modified in the brain that are distinct to the brain. And this is actually true for every pairwise comparison one performs in the dataset. You can visualize this at the global scale by just taking the whole dataset and putting it on a heat map and clustering it. And so in this case, we've got the entire dataset where each line is a cysteine residue. It's colored according to its percent oxidation. And then we cluster it and what's, I think, immediately visually obvious is that each tissue has a cluster of cysteines that are highly modified distinctly in that tissue. You can see heart, brain, spleen, subcutaneous fat, etc. One important thing I should note is that the way that the proteomic— the way that we perform the proteomics in these experiments completely controls for differences in protein abundance. So this is all normalized for protein abundance. And so these differences are fully attributable to actually differences in stoichiometry of modification. Modification, which is an important point. So this highly tissue-specific modification has been, first of all, surprising, and it's been, I think, a useful lens for us to begin to dig into the tissue-specifically modified cysteines on particular proteins to begin to think about what I alluded to before, this idea that the controlled production of redox-active metabolites is eliciting highly context and cell and tissue-dependent adaptive processes. Well, one clear explanation for that could be that the the proteins that are subject to this type of regulation exhibit a high degree of context dependency. And so I'm highlighting subcutaneous fat because we've been— one particular area of interest of ours has been to look at thermogenically competent fat, so-called beige and brown fat, wherein these tissue-specifically modified proteins are proteins that are known to be involved in, for example, thermogenic processes, many of them, or catabolic processes, where it stands to reason that some of these could be functionally contributing to that process. And that's a sort of principle one could generalize and apply that to any tissues. In the dataset. So when we look at individual targets, I'm just showing you a couple of targets here where we know what the biological consequence of the oxidative modification is, and we see a high degree of tissue specificity in terms of the degree of modification. So this is a cysteine residue that we've been interested in for some time in our lab. It's a very well-characterized cysteine on mitochondrial complex I, on a subunit in mitochondrial complex I, cysteine 39 on the ND3 subunit. It's well established that reversible oxidative modification of this site can reversibly inhibit electron transfer through the complex. It's a major control point, or can be a major control point, for flux of electrons through mitochondrial complex I. And what we see is that in our dataset that this cysteine is subject to highly divergent tissue-dependent modification. You can see in skeletal muscle and brain it's negligibly modified, and it can be highly modified in certain adipose depots, providing us a hint that this is likely to be a tissue-specific node of redox regulation of Complex I activity. A different cysteine residue that was characterized by Mike Murphy's lab is on PDK2. And this is the cysteine that when oxidized can affect flux through the TCA cycle. PDK2 is a kinase that affects PDH function. Oxidation of this cysteine is known to increase flux into the TCA cycle via PDH. And what we see again is this cysteine is subject to widely divergent oxidation profiles depending on the tissue. Going from fat all the way up to heart, you can see very divergent profiles. And so this is obviously a lens we can apply to say, well, what are the tissue-specific regulations of established nodes of redox modification on proteins? But of course, the large majority of proteins in the dataset, we don't know what the biological consequence is of oxidative modification. So an additional approach we've been taking is thinking about systematic analyses we can apply to the dataset to uncover biological— concerted biological activities or biological pathways that are subject to coordinated redox modification in a particular tissue context. And one fruitful application that's worked in our hands is by combining our dataset with another dataset that I think many of you will be familiar with called BioPlex. So BioPlex is a dataset that was generated by Steve Geeky's lab and Wade Harper's lab. It's a It's the largest experimentally validated protein interaction network database in the literature. Effectively, it's the product of, I think now, over 15,000 individual IP mass spec experiments, which has been used to build out this web of protein-protein interactions, which effectively reconstructs established biological complexes, biological pathways. And so we thought that an interesting thing to potentially do would be to combine combine this established protein network database with our cysteine oxidation networks from different tissues and ask the question, are there concerted biological complexes or pathways associated with particular biological functions that are subject to coordinated redox regulation? And this could be a starting point for hypothesis-driven studies as to the mechanistic consequences of these types of modifications. And so we applied these analyses and have identified dozens of examples of concerted biological networks that are subject to coordinated redox regulation. So to show you a couple of examples, this is one example of a so-called tissue ubiquitous network where it's coordinately redox modified in every tissue we looked at in mice. This is known as the tRNA aminoacylation complex. These are tRNA aminoacyl synthetases that are known to be dynamically incorporated into a complex, and this complex is known to be involved in rates of translational efficiency, mediating translational functional efficiency. And what we find is that these complexes are coordinately redox modified in every tissue. Most of the networks we identify exhibit some degree of tissue specificity. So this is an example of proteins that are known to be involved in glycogen catabolism that are coordinately redox modified only in brown adipose tissue, skeletal muscle, and heart. And so we've been using this as starting points to look at specific potential new mechanisms of redox regulation over effector protein function. I'll show you just a couple of brief examples of that, of ongoing studies in our lab. One is actually focusing on this tRNA aminoacylation complex I mentioned before, just a couple of seconds ago, where there's established literature on the fact that these proteins are dynamically incorporated to this complex and that differential incorporation of particular proteins into the complex is known to mediate differences in translational efficiency. And so we just had a very straightforward hypothesis that perhaps cysteine oxidation might be somehow regulating differential incorporation of proteins into this complex. And so we've begun to explore that and find that that appears to be the case for certain subunits. So this is just some data demonstrating if we perform native IP pull-downs of the tRNA aminoacylation complex with or without cysteine modification reagents, in this case we use DTT to remove all of the reversible modifications on cysteines, we find that that's sufficient to modulate the incorporation of two key aminoacyl synthetases, specifically the methionine and leucine synthetases, and rather interestingly in opposite directions. So, and we're now focusing in on a particular subunits that— or the particular residues that explain these effects. Another example we've been working on are hexokinases that came out of the network analysis as well. We actually specifically focused in on this particular cysteine— conserved cysteine on hexokinase 2, which we have subsequently found evidence appears to be sufficient— modification of it appears to be sufficient to modulate flux from glucose to glucose-6-phosphate, that first step in glycolysis. Just a couple of examples of that here where we find that that increased reduction of that cysteine is sufficient to increase flux from glucose to glucose-6-phosphate. I'm not showing you data here, but actually when we either pharmacologically oxidize or attack that cysteine with a targeted cysteine electrophile, we see the exact opposite. We see inhibition of flux. So this appears to be a redox control point for the first step of glycolysis that we're continuing to work on. And as I said, there are sort of dozens of examples of these. From this analysis. So that's a very brief overview of the OxyMouse dataset and just a couple of examples of how we've been using it in our lab. One thing I would like to mention is that this whole dataset is available on this OxyMouse website for anyone who's interested in the audience and taking a look at their favorite biological process and seeing if it's subject to redox regulation in a particular tissue or in a particular adaptive context. We're continuing to add to those datasets as we study different biological settings and different physiological adaptive processes. So what I've shown you so far is one of the major pathways we've been focusing on with regards to succinate regulation, which is specifically this capacity for accumulated succinate to act as a local signal by controlled production of redox-active metabolites that can then affect the post-translational modification state of effector proteins. And we think of this as sort of cell-autonomous mode of cellular regulation, something we continue to work on. What I'd like to talk about for the rest of my talk is a newfound and completely separate mode of biological regulation we've recently observed for accumulated succinate, which involves its ability to actually act as a paracrine and also endocrine regulator of biological adaptation. So I'm going to show you here is— I'm going to show you the summary first before going into the data. But what we found is that under certain biological contexts, that accumulated succinate can be rendered amenable to secretion, wherein it can have very interesting and strong effects in some cases on paracrine tissue adaptation and endocrine tissue adaptation. And the reason that this has surprised us is because the classical view of succinate as a bioenergetic entity, as a metabolite, is that it's thought to be trapped within the mammalian cell. So most mammalian cells do not have, or are not thought to have, plasma membrane transporters that can release dicarboxylates. And what we found is that under certain physiological sucinate which is normally trapped in the cell can be chemically transformed and transiently rendered amenable to secretion in a controlled way, and that this controlled release of sucinate can have very interesting physiological effects. We're showing this in a number of different areas, but I'm going to show you some data from a recent publication of ours where we've looked at this in the context of muscle adaptation to exercise. So just to take a quick step back and provide an overview as to why we're interested in exercise in the first place, I mean, it's in some ways it's self-evident. Exercise as an intervention is, I think, extraordinary just in terms of the, the protection it affords against metabolic disease. There's this well-appreciated, very well-appreciated in the clinical literature that vigorous exercise not only has the capacity to increase muscle size, muscle strength, but has very powerful effects on improving whole or on organismal glucose and insulin sensitivity and is therefore highly protective against metabolic disease. And it's now being appreciated that it's likely protective against other diseases as well, including of course cardiovascular disease and cancers. So one of the questions we were interested in with regards to how muscle adapts to exercise stemmed from the fact that we understand in the field that the process of vigorous muscle contraction that involves muscle myofibers requires absolutely the involvement of non-myofibrillar cells that are resident in muscle tissue— immune cells, stromal cells, neurons, endothelial cells— to somehow contribute to the remodeling process. These non-myofibrillar cells are essential to the muscle remodeling that renders it more glucose sensitive, more insulin sensitive, stronger post-exercise. And we were curious to understand the nature of potential metabolic signals that might emanate from contracting muscle and communicate to the paracrine tissue environment to initiate this adaptive response. And specifically, we had as our lens the question as to whether or not metabolites might fulfill this process, because vigorous exercise is, of course, inherently a bioenergetic process. So it might make sense that the nature of a signal that could be mediating paracrine adaptation would be metabolic in nature. And so, we designed a series of experiments to specifically focus in on— or to identify metabolites that were selectively released from exercising muscle in both mouse and humans. And so, to do this, I'm going to show you the mouse data first. We performed a series of parallel experiments that would define what we called a local exercise-responsive metabolite signature. We performed first an experiment where we We subjected mice to an established exercise protocol, an acute exercise protocol that's known to remodel muscle in the way I described on the last slide. And then immediately following that protocol, we isolated local extracellular fluids from muscle and compared them to extracellular fluids of sedentary muscle. And we were specifically interested in the post-exercise accumulated metabolites. Then in a separate experiment, we asked the question, what metabolites specifically increase in extracellular fluids post-exercise compared to the muscle tissue itself? 'cause we're specifically interested in the release signature. And then finally, we performed a third experiment where we compared the accumulation of metabolites in local extracellular fluids compared to bulk plasma, because we were specifically interested in metabolites that were likely to be released from muscle itself as opposed to some other tissue during exercise. And to be most interesting to us, the metabolite would have to fill all three of these criteria. Profiled several hundred metabolites, and several metabolites fulfilled any one of the three criteria, but only one metabolite actually fulfilled all three. And the metabolite was, was one that initially surprised us., and it was succinate. I'll briefly walk you through this data where we see that this is just the first experiment where we see succinate is one of the metabolites that accumulates in extracellular fluids post-exercise. It specifically accumulates in extracellular fluids and not in the tissue itself. And it also is much more accumulated in local extracellular fluids compared to bulk plasma. And so, we saw this data in mice. We also saw it in humans. We were fortunate to collaborate with Eric Richter at the University of Copenhagen Eric has this really fantastic experimental setup where he, he catheterizes the femoral artery and femoral vein of an exercising muscle in an individual and then has them use an exercise bicycle on an established exercise protocol or set VO2 max and then samples the blood entering the exercising muscle and draining from the exercising muscle during the intervention. We were able to use these datasets to effectively perform the same experiments I showed you on the last And strikingly again, we saw this highly selective release of succinate during exercise. You can see it peaks at a couple hundred micromolar in the femoral vein samples and then rapidly renormalizes post-exercise. You can see here, this is just a classic way of determining whether or not a tissue is likely to be a net producer or consumer of a given metabolite by doing what's called AV difference, arterial-vein differences. And this large negative effect tells us that the exercising muscle muscle is the actual source of the released succinate. And so we had this evidence, again, that surprised us that succinate was this— was being selectively released specifically during exercise. And as I alluded to before, this was a surprising observation because the literature on succinate predominantly treats it as an intracellular metabolite and not one that should be selectively released under any condition. But what we were clearly seeing here was there was some controlled mechanism for succinate release because we weren't seeing every intracellular metabolite released. In fact, most of the structurally similar metabolites to succinate were not being released in the context of exercise. So there was some controlled fashion through which this metabolite was being secreted by muscle during exercise. And we were curious to understand that. One of the lenses we applied to that question was to simply just compare the chemical properties of succinate to the other mitochondrial metabolites that were not released in the same context of exercise. And this is basically the TCA cycle metabolites that I'm showing you here. And what was interesting to us, one of the chemical features that struck us that was distinct of succinate— to succinate— is that it has an unusually high monocarboxylic pKa. And so what that means is that at unusually high pHs, at physiologically relevant pHs, a substantial proportion of the succinate pool can be protonated and transformed from a dicarboxylate into a monocarboxylate. And this is something that can be achieved at pHs that are known to occur during exercise. It's known that vigorous exercise transiently lowers intracellular pH into the mid-6s, and we can see that that would be sufficient on the basis of the chemical properties of succinate to render a substantial proportion of the pool a monocarboxylate. And that's not true for other mitochondrial metabolites. And the reason this chemical property of succinate interested us so much is because in parallel we were performing proteomics experiments of muscle tissue to try to understand what transporters might be facilitating succinate release.. And what we found were the most abundant plasma membrane transporters— there were no dicarboxylate transporters, but we found the most abundant carboxylate transporters in muscle tissue to be monocarboxylate transporters, specifically MCT1 and MCT4. These are transporters that are known to facilitate the transport of small molecule monocarboxylates like lactate, pyruvate. So on the basis of those two observations, we generated a hypothesis wherein maybe succinate that's normally trapped in a cell because it's a dicarboxylate, in the context of an acute perturbation such as exercise, where we know cytosolic pH is lowered, that a proportion of the succinate becomes transformed into a monocarboxylate, and in this transiently protonated state, it can be rendered a transport substrate for MCT, and that's how it's getting secreted. And so we tested that in a number of different ways. I'll just show you a few representative examples. We did this in cells and in tissues, but effectively we found that this was the case. So I'm just showing you some examples here where we use different chemical tricks tricks to acidify intracellular pH. In this case, I'm showing you in muscle cells. We can use mitochondrial inhibitors that basically force glycolysis and lower cytosolic pH, and that's sufficient to drive succinate release. And then if we use— and then if we use pharmacological inhibitors of MCT1, we see substantial inhibition of succinate release from muscle cells. Or if we genetically deplete MCT1, again, we substantially inhibit succinate release, and we see the same thing in muscle tissue. Issue as well. Moreover, we were able to show in a reductionist system that MCT1 can in fact be a succinate transporter. So, we expressed recombinant human MCT1 in an oocyte transport system and directly examined succinate transport, and we found that MCT1 can indeed transport succinate, and rather importantly, it specifically has the capacity to transport high amounts of succinate when pH is lowered into the range where where a substantial amount of succinate exists in the monocarboxylate form as opposed to the dicarboxylate form. And this transport activity is fully inhibitable by MCT1 inhibitors. So, we had evidence at this point for this model of what we call pH-gated succinate secretion that transiently occurs during exercise. And this encouraged us to then ask the next question, which is what might the role of transiently released succinate be in the context of paracrine adaptation to exercise? And the reason we were interested in paracrine adaptation or the potential for paracrine adaptation was based on parallel studies we were performing of muscle tissue that demonstrated that certain cell types in muscle appear to express high amounts of the G-protein-coupled receptor that's well established to specifically respond to extracellular succinate. This is GPR91, otherwise known as succinate receptor 1. But the relevance of these cells in muscle biology hadn't really been explored at all. So we first of all decided to better characterize which cell types are actually expressing the succinate receptor in muscle tissue. Tissue. As you can see, this is just a delta-delta CT where you can see substantial signal for succinate in whole muscle tissue. And then we purified major cell populations from muscle tissue and compared the signals. In this case, delta-delta, so a lower signal means increased abundance. So we could see is that there's substantial succinate receptor expression in endothelial cells, satellite cells, and stromal cells. And we found actually, I think rather interestingly, that muscle tissue or muscle cells themselves, the pure myotubes or differentiated C2C C12s expressed undetectable amounts of the receptor. So to the extent to which we could measure it, the receptor appears to be uniquely expressed in non-myofibrillar cells in muscle tissue. We corroborated this with in situ hybridization experiments where we probe for succinate receptor 1 signal in red here. This is in muscle tissue cross-sections. And then we cross-reference with stains that stain the myofibers in yellow or in this case stromal cells in blue. And you can see we see a very strong overlap with the stromal population and no antico-localization with the muscle fiber population. So, this encouraged us to ask the question, well, what might the relevance of succinate release and then subsequent ligation and activation of succinate receptor-dependent signaling be in the context of how a muscle remodels to exercise? And so, the first thing we did in this context was to first use succinate receptor knockout mice subject those mice and wild-type mice to an acute exercise protocol that drives succinate release based on our previous data. And then immediately following that, we purified the cell populations from muscle tissue that are known to express the receptor, and we performed RNA-seq analysis to define the transcripts that were, one, exercise responsive, and two, completely dependent on the receptor, genetically dependent on the receptor. And what we found from these analyses, the most striking differences were found in satellite and stromal cell populations were transcript families that were known to be involved in neurotrophic effects. So, mediating the production of neurotrophic factors and increasing muscle innervation, as well as factors that are known to be involved in muscle extracellular matrix remodeling. For both of these populations were exhibiting both exercise dependence and a genetic dependence on the succinate receptor. And so, that suggested to us that perhaps the succinate ligation and GPCR signaling might be relevant for muscle adaptation to exercise. And this encouraged us to perform a training study in mice where we subjected them to this established running protocol, wherein a mouse is basically given a weighted wheel that increases in resistance over 3 weeks and is known to improve muscle strength in wild-type mice. And then we performed a number of physiological and molecular analyses after these— after this intervention in both wild-type and and receptor knockout mice. I'll just show you some of the most striking results from this analysis. The first, I think, most striking physiological difference between these mice is shown on this slide, wherein interestingly the receptor knockout mice are very happy to run on these training protocols. They run at least as much as the wild-type mice. Actually, they run more than the wild-type mice. There's no aversion to training. But despite that, what you can see in the wild-type mice is that there's this expected increase in muscle strength that occurs over the 3 weeks of training, and it's effectively lost in the than mice that don't have the receptor. So there's this very peculiar phenotype where the mice train and train and train, but they don't receive the strength benefits to training in the same way that wild-type mice do. And so we performed a number of molecular analyses to try to understand the basis for this. I'll just show you one that I think is likely to be at least a major part of the story. And it's where we, on the basis of our transcriptomic studies or RNA-seq studies, we thought to focus in on muscle innervation. And what I'm showing you here are compressed 3-dimensional stacks of muscle innervation signal. This is TUB3 staining where we're staining the nerves in muscle tissue. You can see wild-type and receptor knockout mice from our— under sedentary conditions have similar signal. And then wild-type mice that are trained on the protocol for several weeks, you see this striking increase in muscle innervation. And you see that this is largely inhibited in the mice that don't have the succinate receptor. You don't see the same increase in innervation so we think that because innervation is known to be a major component to improvements in muscle strength, this could be at least part of the reason why these mice don't adapt in the same way. These are just the, uh, some representative 3-dimensional videos that I just like to look at sometimes because I think they're quite cool to look at. But you can see the striking difference between the wild-type trained muscle in terms of extent of muscle innervation versus the, the receptor knockout mice. So that work is really a summary of what we've begun to understand. What I'm showing you on the right here, this unexpected role for the same metabolite succinate to not act as a local regulator of cellular adaptation through its capacity to control post-translational modification of proteins, but also this newly appreciated role for succinate to be released from cells in a controlled fashion in a way that is highly dependent on intracellular pH. And this is a way of potentially intracellular and potentially also intratissue communication of metabolic status through controlled release of succinate, which can then go on to to affect paracrine and endocrine adaptation through receptor signaling. I mentioned endocrine. We actually have ongoing work now suggesting that in different physiological settings, that distal tissues to those that actually release succinate can also be affected by this type of regulation. And that's something we continue to explore. And of course, we showed— I showed you work from our lab in the context of acute exercise, but there's— it stands to reason that this is a general bioenergetic phenomenon where in any transient physiological state that would result in the lowering of cytosolic pH could in principle drive the same selective release of succinate. And in fact, work that was published last year from Mike Murphy's lab showed the context of hypoxic tissues— the hypoxic heart has the capacity to release succinate, and in through the exact same mechanism that we're describing here. And so this could likely be a generalizable principle. So some areas that we're looking at, for example, are the tumor microenvironment, which is known to be highly hypoxic and associated with high amounts of succinate accumulation or other metabolic states where transgene acidification occurs intracellularly. So, with that, it leaves me to thank the individuals whose work really— whose work I talked about today, who drove all of the studies I outlined for you today. The two papers I talked about in particular were driven by two exceptional fellows in my lab. Haocong Zhao is a postdoctoral fellow, drove the OxyMouse work, and Anita Reddy as a PhD student, and Luis Bossi, a postdoctoral fellow who drove the exercise work I talked about. And with help from many individuals in the lab, we benefited from really great collaborations on these two studies in particular with Steve Gigi's lab, Diane Mathis's lab, Nathaniel Gray's lab, and Eric Richter's lab. And a lot of the work I talked about, historical and ongoing, and through collaborations and inspiration, has really benefited from great work collaboration with my former mentors Bruce Spiegelman and Mike Murphy. Thank you all, and happy to take questions.

Thank you, Ed.

That was terrific. Thank you. Yeah, I'm just— there we go. Okay, thank you very much, Gokhan. So I'll start with the Q&A section here. So there's some questions in the chat and some in the Q&A. I'll start with Q&A first. I've got a A question from Valentin that asks, what is your opinion on the elusive NADPH oxidases NOX1 through 5, DUOX1 and 2 in the context of redox cysteine, reduced cysteine, oxidized cysteine signaling? I think that's a great question. I mean, I think it's, you know, what I— the short answer is we haven't specifically focused on those as sources of reactive oxygen species, but what I think the OxyMouse dataset clearly or is this context dependence that I described? And I think a big part of that context dependence almost certainly has to do with the nature of the source of the redox-active metabolite. So I described succinate metabolism as a potential source, wherein the actual source of the redox-active metabolite would be mitochondrial complex I. It stands to reason that NADPH oxidases as a source, which are obviously differentially localized in the cell, will likely result of the post-translational modification of completely different subsets of proteins. And we think that the spatial constraints of redox signaling, wherein the cysteines most likely to be modified are those that are most likely to be proximal to the production site, is likely to be a major way through which— a reason for which we're observing these differences in the particular cysteines that are modified. So I think NADPH oxidase probably explains some of that in our datasets, but we We haven't specifically looked at it. And so the next question I have is from Aeon Ibrahim who asks, do the succinate molecules stay as monocarboxylate ions in the interstitium? And if so, is it possible that some of the succinate is being taken by non-myofibrillar cells through their own MCTs? Okay. Yeah, that's a really cool question. So our default position on that question has been that it's almost it's likely that the succinate is immediately transformed back into a dicarboxylate just because extracellular fluids are so well pH controlled that it's unlikely that the bulk pH is going to change substantially such that succinate would remain in a monocarboxylate form. Now, having said that, we have unpublished work in the lab that suggests that MCT in certain cell types, actually brown adipocytes being a major one, which is actually— I didn't talk about it today, but we had work showing that brown adipocytes have this unusual capacity to sequester succinate from extracellular fluids. One of the major transporters that appears to be doing this is MCT1, which is confusing if you just consider the framework in the way I just described it, because succinate should not be a monocarboxylate in extracellular fluids. So this is really an open question where either through mechanisms of local pH control or local protonation, maybe succinate can be transformed in such a way as to to render it amenable to uptake by other cell types. And I think that that is probably an additional way through which succinate regulation can occur in tissues if they are able to take it up. And then the final question in the Q&A is from Matthew. Have you looked at whether any of the succinate-mediated processes can be regulated by lysine succinylation? Also, what would you say that metabolites are alternative signaling molecules. Okay. So, the— yeah, we haven't considered succinylation for one major reason, and that is that succinate itself actually does not promote succinylation. So succinyl-CoA is the major metabolite required to drive that post-translational modification. And so we don't tend to see under conditions where succinate specifically accumulates a coinciding increase in succinylation. I would also mention that actually my personal view on succinylation, at least from what I understand from what's been published, is that the extent to which stoichiometric technologies have been used to map succinyl modifications, we tend to find that they're relatively low stoichiometry. And so I think that that suggests to me that unless it's an activating modification, I'm not sure what the biological meaning of those modifications is at that standpoint, but at that stage. But I think stoichiometric methods would be useful to apply to those types of PTMs. And then, yeah, as to the sort of conceptual question of whether there are alternative signaling molecules, I don't have a strong view on alternative to what. I don't know. But I think in as far as one would treat, you know, a quote-unquote signal as some factor that's responding to a perturbation to elicit a response, I think that succinate and redox-active molecules like reactive oxygen species appear to fulfill that role.

Can I ask you, I may have asked you this before, but are all the cysteine modifications identical to each other? Or do you also, are you able to dissociate several different types of cysteine modifications?

Yeah. That's an excellent question. One of the— so, what we— the functional consequence of particular cysteine PTMs can be different, for sure. And so, our initial strategy with the OxyMouse dataset was to simply obtain the aggregate reversible modification signal, so that would be the net contribution of anything that is TCEP reversible, so glutathionylation, sulfenylation, S-nitrosylation, disulfide formation, et cetera, just to find hotspots, basically. But then as you say, I think it is imperative to now try to deconvolute the extent to which the contribution of those signals is, for example, an SNO or an SOH, which would presumably in many cases be functionally distinct from an interprotein disulfide. And actually, there's, you know, a lot of suggestion for particular protein targets that, for example, S-nitrosylation can sometimes act to prevent an otherwise dominant interprotein disulfide, which would have a very different functional consequence. And so what one can do with these types of proteomic workflows is effectively replace the TCEP reduction step with alternative workflows wherein, for example, you could use copper and ascorbate to selectively remove the SNO modifications and say, okay, what proportion of this landscape is attributable to nitric oxide signaling or different chemotypes that have been used to specifically label different oxoforms of cysteines. And I think that that's, yeah, that's certainly an important layer that we could look at.

I think you have an anonymous attendee question.

So I've got the anonymous attendee question about Gokhan about your bow tie. In honor of Armen. And I think, yes, that is, uh, I'm seeing yes, once a year it's worn, right? For exactly that reason. And then, um, I've got one more. Oh yeah, another question here from, from Hiran that, um, that asks, um, can you drive muscle innervation further with succinate, um, supplementation, uh, and training? Um, that's, that's something we're trying right now. Um, uh, so So Hiren is most likely referring to some other work we published a couple of years ago, wherein we were able to show that actually you can transiently elevate circulating succinate through simple dietary intervention. Succinate is readily taken up through the digestive organs— through the gut, I should say. And we were studying that in a very different context, in the context of brown adipocyte biology But as he suggests, it stands to reason that potentially agonism may be sufficient to elicit some of this signaling. That's something we're looking at now. And we're also thinking about this in terms of succinate receptor 1 agonists, which actually have not been well advanced as chemical entities. Most people have been working on antagonists of the succinate receptor, but we're hoping to try to develop some ways of drugging this receptor from the agonist point of view to see to see if it affects some of these processes. One thing I should mention, though, is that the beneficial effects we're seeing with succinate-dependent innervation appear to be coinciding with this pulsatile succinate release, where it rapidly elevates and then rapidly renormalizes after. We have ongoing studies in our lab, some of which will be published soon, wherein very different physiological settings where there's this chronic release of succinate from tissues and the steady-state levels in the circulation remain high for long periods of time we see very different adaptive processes and actually maladaptive processes. We see fibrotic signaling in tissues, which gives us pause before thinking about just, you know, drugging this receptor, at least, or at least trying to differentiate between the pulsatile nature of— or the consequences of pulsatile signaling of the receptor versus chronically elevated signaling. Oh, and I've got one more. Yes. Have you noticed whether your cysteine enrichment molecule— this is from Matthew again— have you noticed whether your cysteine enrichment molecule can enrich methionine as well? So it can't. The, the iodoacetamide chemistry cannot be used to label methionines. Actually, until recently, methionines were thought to be non-derivatizable by there was no established chemistry to do that. Some really nice work has come out of Chris Chang's group in Berkeley a few years ago now, where he identified these new warheads that can be used to target methionines and actually differentiate in some cases between oxidized and reduced methionine. Methionine can also be oxidized, but it's sort of been invisible to any real biological investigation because we've had no way of really comparing oxidized versus non-oxidized methionine. But in principle, one could could modify the types of tags we've developed for cysteine or close for methionine. That's something we've been thinking about doing and could potentially illuminate that aspect too.

Yeah, good. Thank you very much. I think this was a terrific, terrific lecture and very well fitting into what we really want to achieve in remembering Armen. So thank you so much and enjoy your award and the large cash once the restaurants and bars are open.

Absolutely, absolutely.

I'm looking forward to it. We will definitely take you out also for a dinner when our lives return to normal.

That's great, I look forward to it, absolutely, and seeing you all in person. Perfect. Thanks, Ed. Congratulations. Thanks, Matt. Thanks, David.

Thanks.