Welcome, everybody, to the final plenary session of ENDO 2021. Yesterday, we had a total of 7,323 attendees from 97 countries join our sessions. Once again, let me remind us all to visit and comment on our outstanding posters in the abstracts hall today. It has been a pleasure to directly connect you to a variety of esteemed speakers as part of ENDO 2021. In the weeks and months to come, we hope many more of you will take the opportunity to watch online any plenaries or other sessions you may have missed over the last four days. Thank you. I now have the pleasure to introduce our first speaker, Dr. Adrian Vela. Dr. Vela is a professor of medicine in the Mayo Clinic College of Medicine and has been on staff at Mayo Clinic Rochester since June 2001. Dr. Vela serves as the research chair for the Division of Endocrinology and Metabolism. He directs a program funded by the NIH aimed at understanding the role of common genetic variants in the pathogenesis of prediabetes and other factors affecting beta cell function. His clinical interests include the diagnosis and management of hypoglycemic disorders such as insulinomas and hypoglycemia post Roux-en-Y gastric bypass. Please join me in welcoming Dr. Vela as he now explores how insulin signaling and other defects interact with the liver in the pathogenesis of type 2 diabetes. Dr. Vela, the floor is yours. Thank you for that kind introduction, and above all else, thank you to the organizers of the Endocrine Society for inviting me to speak on a topic close to my heart. These are my disclosures. So I'm the recipient of grant support from Novo Nordisk, looking at the addition of liraglutide to patients after sleep dystrectomy, and I've consulted for VTV Therapeutics and Zeeland Pharmaceuticals. The defects in type 2 diabetes could probably be summarized by this slide, which I owe to my mentor, Dr. Rizza, who showed that people with type 2 diabetes, in contrast to people without diabetes, have high fasting glucose. Although their fasting insulin is higher than non-diabetic individuals, this fasting insulin is clearly inappropriate for the prevailing glucose concentrations. And then subsequently, in response to a standardized meal challenge, people with diabetes have blood sugars that go high and stay higher for a lot longer before returning to their high fasting baseline, compared to non-diabetic individuals. This is because insulin secretion is decreased as well as delayed, so that peak insulin concentrations occur at 120 minutes as opposed to 30 minutes in non-diabetic individuals, and you can see that there's a significant disparity in peak concentrations achieved. These defects are exacerbated by a lack of suppression of glucagon, and indeed, in people with type 2 diabetes, sometimes there is a paradoxical rise in glucagon concentrations after a meal, compared to non-diabetic individuals, and these defects together contribute to the hyperglycemia seen in people with type 2 diabetes. And in fact, early on in my training, my training was dominated by these two papers that I'll be showing you. The first was this published by Dr. Basu and Dr. Rizza in 1996, which showed that if you give an insulin profile to individuals that mimics that seen in people with diabetes, as opposed to a non-diabetic insulin profile mimicking that seen in non-diabetic individuals, the degree of glucose intolerance in both cases rises with the degree of insulin resistance. So people who are lean have near normal glucose concentrations, even if they are subjected to a diabetic insulin profile. But as insulin resistance increases, so people who are obese and then people who actually have frank type 2 diabetes, their blood sugars increase progressively, even in the presence of a non-diabetic insulin profile. Similarly, my friend and colleague Dr. Shah, working with Dr. Rizza, showed that if you gave people, if you studied people twice, on one occasion you suppressed glucagon, and on the other you kept glucagon elevated, as happens in people with type 2 diabetes. Even in the presence of a non-diabetic insulin profile, as opposed to a diabetic insulin profile, there was a rise in peripheral glucose concentrations. And this was likely explained by impaired suppression of endogenous glucose production. However, these defects were far more marked in people given a diabetic insulin profile, where glucose concentrations were markedly higher compared to the non-suppressed day. And again, this was explained by a lack of suppression, in fact a paradoxical rise in endogenous glucose production in this setting. And in fact, most of my career I have focused on not so much what happens to glucose once it's taken up, partially extracted by the liver, and then appears systemically, or on endogenous glucose production, which is the sum of gluconeogenesis and glycogenolysis. But basically I've followed my colleagues Michael Camilleri, as well as outside inspiration Dan Drucker in Toronto, Jens Holst in Copenhagen, and Dave D'Alessio at Duke. And I've always come to the conclusion that ultimately, primacy in glucose homeostasis is afforded by the alpha cell and the beta cell, and the liver, in a sense, could be said to be the prisoner of alpha cell and beta cell function. That has been more or less my modus operandi so far. However, over the years, there's been an increasing realization that that is not always the case. So this is one of our publications, Dr. Meera Shah, who studied people with type 2 diabetes before and after Roux-en-Y gastric bypass. And what is actually interesting in this setting is that fasting glucose is the main thing that changes quite dramatically in these subjects. These are studied four weeks after Roux-en-Y gastric bypass. In fact, and not surprisingly because of their upper GI anatomy, the incremental rise in glucose post-meal is actually higher post-Roux-en-Y gastric bypass compared to pre-gastric bypass. But what's also interesting is that although endogenous glucose production suppresses to more or less the same nadir as prior to Roux-en-Y gastric bypass, the fasting endogenous glucose production, which is what explains the decrease in fasting glucose, is decreased quite substantially compared to pre-Roux-en-Y gastric bypass. And this is clearly not explained by an increase in insulin secretion. If anything, insulin secretion in the fasting state, here measured by c-peptide, actually decreases slightly but significantly compared to prior to surgery, implying that there has been an improvement in hepatic insulin action, but also that endogenous glucose production has decreased quite substantially, 5 micromoles per kilogram per minute, despite these small changes in c-peptide. And this, of course, is reminiscent of the changes that, for example, Dave Kelly described in 1993, when he showed that one week of caloric restriction prior to any significant weight loss was accompanied by a substantial increase in glucose uptake as measured by CLAMP. And then post-weight loss, from 90 kilograms to about 78 kilograms, there is a subsequent further increase in glucose uptake. And in fact, we actually replicated this finding when we fed people with type 2 diabetes off of all medications. So they had a fasting glucose of about 180 milligrams per deciliter and peak blood sugars in the 240 to 250 milligram per deciliter range. We fed them, for six weeks, a diet of about 800 calories a day. And then subsequently, for the remaining six weeks, we loosened up a little bit to 900 calories a day. But as you can see, there was a very dramatic reduction in glucose concentrations. And despite ongoing weight loss, by 12 weeks, there was really no change compared to six weeks. The most marked change actually occurred in fasting blood sugars. And in fact, incremental area above basal of glucose wasn't really substantially different from that observed at baseline. So really, it was the fasting glucose which was the main change that occurred with caloric restriction. And indeed, like we saw with bariatric surgery, the decrease in fasting glucose was really what was most marked. And we saw this at six weeks, and it was essentially maintained at 12 weeks. Not statistically significantly different from six weeks, but significantly different from that observed at baseline. And again, the difference was approximately five micromoles per kilogram per minute in these circumstances. Again, note that there was no change in nadir suppression of endogenous glucose production. So that was the first set of observations. The second set of observations actually was prompted by this. So Gurlis Bok, working in Bob Rizza's lab, was studying beta cell function across the spectrum of prediabetes. And in fact, if you classified people based on having normal fasting glucose and normal glucose tolerance, and you took all comers with impaired fasting glucose, you could see there was a substantial decrease in disposition index. Disposition index is actually the expression of insulin secretion as a function of the prevailing insulin action, the power-to-weight ratio, if you will, of the beta cell. But as you can see, when you break down this subset according to their glucose tolerance status, there is a group of people who have impaired fasting glucose but normal glucose tolerance, whose disposition index is actually indistinguishable from people with normal fasting glucose and normal glucose tolerance. And then as glucose tolerance worsens, perhaps to some extent independently of whether they have impaired fasting glucose or otherwise, disposition index decreases till they are practically in the diabetic range, when they have impaired fasting glucose and diabetic levels of glucose at two hours after an OGTT. And this has subsequently been replicated in some of my own cross-sectional population studies. But also, it's interesting because here I am quoting two review articles. One was by Eric Stollerman and Jose Flores, and this was by Galina Smushkin and myself. And these are both outdated because since then there are several more loci associated with diabetes. But the important thing is to appreciate that there are genes that have been associated with type 2 diabetes but not with fasting glucose. And there are genes that have been associated with a higher fasting glucose but not necessarily with predisposition to type 2 diabetes. And then of course there are genes that are associated with both. But even the genetic evidence would support the prior observation that fasting glucose does not necessarily imply that beta cell function is abnormal. And in fact, glucokinase perhaps could be everybody's favorite expression of this. So heterozygous inactivating mutations alter the set point for insulin secretion. And in fact, MoD2, where this occurs, is characterized by having a high fasting glucose but then a normal response to an oral challenge. On the other hand, homozygous inactivating mutations present at birth with neonatal diabetes mellitus. And then also of interest to me, activating mutations of this gene are one recognized cause of neonatal hyperinsulinemic hypoglycemia. But glucokinase acts as a glucose sensor in islets and in hepatocytes. And it's not inhibited by the product of the reaction it catalyzes. That's why it is able to act as a glucose sensor. And MoD2, which I've already told you represents a heterozygous state for inactivating mutations, is characterized by impaired hepatic glycogen synthesis, as Dr. Schulman's group at Yale was able to demonstrate in this paper from 1996. So that after breakfast, people with MoD2 had far less change in hepatic glycogen compared to unaffected controls. And the same pattern was observed after each meal. This represents the same thing. Now, in the RISA lab at the time that I was in training, we were actually using the acetaminophen glucuronidation method, which essentially uses acetaminophen to glucuronate, which gets glucuronidated in the same way that glycogen gets glucuronidated and galactosed. And we can sample using a different label for glucose, the contribution of glucose to glycogen in the hepatocyte. And this would be the direct pathway, the so-called direct pathway of glycogen synthesis in the postprandial period. And in fact, my colleague Dr. Basu at the time showed that people with type 2 diabetes in the presence of low insulin had elevated glucose production compared to non-diabetic individuals. By the time they were exposed to high insulin, splanchnic glucose production suppressed to the same degree as people without diabetes. However, what was interesting is that even in the presence of hyperinsulinemia, splanchnic glucose uptake was significantly decreased in people with type 2 diabetes. So that net splanchnic glucose balance was in both cases significantly different compared to non-diabetic individuals. And this was actually explained by a decrease in direct synthesis of glycogen, which implied an acquired defect in the glucokinase gene, the glucokinase enzyme activity. So this, in a sense, was a replication of what Dr. Schulman showed in people who had a genetic defect in the glucokinase gene. Whereas here we were showing that people with type 2 diabetes were behaving as if they had developed an acquired defect in glucokinase activity, and this was subsequently replicated too. Now, this comes from my compatriot, Dr. Ajus, at the University of Newcastle-upon-Tyne. And it's a useful reminder of how glucokinase works. So glucokinase is regulated by the glucokinase regulator protein, which keeps it sequestered within the nucleus. However, in the presence of high glucose, this dissociates. Glucokinase makes its way into the cytoplasm and can catalyze the phosphorylation of glucose to G6, to glucose 6-phosphate, which can then either be stored as glycogen or else participate in lactate synthesis or conversion of pyruvate. Pyruvate can go to the TCA cycle, it can go to fatty acid synthesis. The pathways are reversed, so glucokinase is sequestered back into the nucleus in settings of low glucose, low starvation situations, such as starvation, etc. So you can imagine that glucokinase plays a key role as a gatekeeper and on-off switch for the modes of hepatic glucose metabolism. Now, unfortunately, hepatic glucokinase activators, so far, have been limited because of tachyphylaxis, there's been some association with hepatic steatosis, and hypertriglyceridemia. This novel glucokinase activator that I was privileged to participate in was different because it seems that it leaves the regulatory protein to continue its work of regulating glucokinase on the basis of prevailing glucose concentrations. So in this in vitro study, you can see that despite higher concentrations of the glucokinase activator, glucokinase remained sequestered within the nucleus. As glucose concentrations increased, glucokinase was able to leave the nucleus, and again, concentrations didn't really make any difference to the ability of glucose, moderate hyperglycemia, and marked hyperglycemia to cause GKRP to dissociate from GCK and allow glucokinase to act in the cytoplasm. And indeed, this activator was shown to have equivalent effects over a six-month period to that observed with a DPP-4 inhibitor, so approximately half a percentage point decrease in A1c. Whether this will become a viable therapy for type 2 diabetes remains to be ascertained, but again, it's exciting to demonstrate the principle that glucokinase activation can actually be associated with a clinically significant glucose lowering. One other aspect of hepatic metabolism is sympathetic and parasympathetic control. So in the case of sympathetic denervation, this comes from the lab of Dr. Charrington in dogs. Dogs underwent denervation of the hepatic artery, or a sham procedure, and the dogs were subsequently fed a high-fat diet. So as you can see, well, actually, the denervation came after the high-fat diet, so at baseline, this was their response to a glucose challenge. Subsequently, with high-fat diet alone, they had become markedly glucose intolerant, but with the procedure, they actually improved their glucose tolerance slightly but significantly. This was not the case with the sham procedure, where they remained glucose intolerant as they had been once they started the high-fat diet. And in fact, using a clamp in these subjects, they were able to demonstrate that despite matching glucose, insulin, and glucagon concentrations, the dogs who had the sham procedure had higher, their net hepatic glucose favored during the clamp, ongoing hepatic glucose output, whereas the base chow-fed dogs favored uptake in the setting of hyperinsulinemia and hyperglycemia, as would be expected, and the procedure of sympathetic denervation actually produced some neutralization of net hepatic glucose balance. So again, illustrating that factors other than insulin and glucagon can regulate hepatic metabolism. Now, this is something that actually my friend and collaborator showed when he was still an exercise physiologist beginning to dabble in true physiology. And Dr. Matveenko actually produced this experiment, which I actually found rather interesting, in dogs who underwent partial pancreatectomy but preserved their glucose tolerance. However, in the setting of a clamp, they exhibited insulin resistance. In this particular experiment, the partial pancreatectomy produced mainly peripheral insulin resistance, but in subsequent experiments, he was able to show an effect of partial pancreatectomy on hepatic insulin resistance. And this decrease in insulin action was attributable to a decrease in insulin pulsatility. And I thought that herein lay the explanation for the close connection between a decrease in insulin secretion and a decrease in insulin action observed in people with prediabetes. However, and this has just actually been accepted for publication, Marcello Laurenti, working in my lab, developed a way of deconvoluting hepatic vein insulin concentrations from C-peptide. And as you can see in the fasting state in non-diabetic individuals, there are discrete visible pulses. He actually was able to Fourier transform these and obtain a measure of all the different frequencies and all the different pulses contributing to this pattern that we see here. And he was able to measure disorderliness using Appen, basal insulin secretion, i.e. the non-pulsatile component of insulin secretion, amplitude of the pulses, and the pulse interval. So the pulse interval was actually taken as the frequency that contributed the most to net pulsatility. But we also came up with an index called the frequency dispersal index, which basically tried to account for all the different frequencies contributing to pulsatility. None of these pulse characteristics correlated with hepatic insulin sensitivity, nor with peripheral insulin sensitivity for that matter, if you look closely at the r-squared values for these correlations. So at least in humans, it doesn't appear, in non-diabetic humans perhaps, it doesn't appear that insulin pulse characteristics have anything to do with hepatic insulin sensitivity. One other important aspect is the presence of hepatic islet crosstalk. So this schematic is actually inspired by Daniel Dean's presentation and subsequent publication. Where Daniel showed that actually amino acid concentrations working through SLC38A5 may actually be responsible for changes in alpha cell mass. And this is mediated in part because glucagon can actually stimulate amino acid catabolism and therefore regulate amino acid concentrations in the periphery. In keeping with this, Philip Knopf's group has recently shown that people who have hepatic steatosis when exposed to basal and high glucagon concentrations have higher circulating amino acid concentrations compared to lean individuals without hepatic steatosis. We actually are in the process of working through a dataset aimed to develop a model for glucagon action and its contribution to oral glucose tolerance. And we actually replicated one of the experiments I showed you at the beginning, performed by Dr. Shah, where we studied people at different insulin concentrations in the presence of suppressed and non-suppressed glucagon. And what I want to illustrate by showing you this data is how much variability there is in endogenous glucose production in individuals in the suppressed glucagon state, but even more so when glucagon is not suppressed. There is a huge variation in response of endogenous glucose production to these changes in glucagon. So, I'll conclude by saying that hepatic glucose metabolism may be bound by islet function, but there is considerable variability in these boundaries. I think I've shown you that fasting and postprandial glucose metabolism are regulated quite independently, and the bonds of hepatic glucose metabolism, so to speak, become considerably wider as insulin resistance and islet dysfunction develop. Glucokinase activation may also be a viable adjunct for the prevention and treatment of type 2 diabetes in the future. I'll conclude by acknowledging my study subjects and the CTSA and CRU for their help. My current lab, for whose work I am greatly indebted. Also, my friends and colleagues who over the years have provided a lot of input and a lot of advice, Michael Camilleri, Bob Rizza, Mike Jensen, Alexey Medveenko, and Quinn Peterson, as well as Claudio and Chiara at the University of Padova for their help with mathematical modelling. Thank you very much. Adrian, thank you so much for such a stimulating talk. I'd like the audience to take a look at the Q&A box during our plenary session today under the open and answer tabs. While Dr. Vela has already responded to a couple of queries via written response, I'll take the prerogative to ask a few more questions now that are coming through and were sent to me via chat. Adrian, given the complexity of current diabetes pharmacotherapy, how can further combination pharmacotherapy be better adapted to specifically address the liver abnormalities that you've outlined today in diabetes? I think everybody acknowledges that current pharmacotherapy has limitations, although it's a dramatic improvement to what was available even 20 years ago. But I think it makes sense to try and combine hepatic sensitizers with insulin secretogogues or compounds which both secrete insulin and suppress glucagon. But I think some degree of hepatic sensitization is necessary. The other thing, of course, is that we always talk about lifestyle changes. And I think the experiments with caloric restrictions show very clearly, in my opinion, that unless you decrease caloric intake, the liver isn't going to be as responsive to changes in insulin and glucagon concentrations. Now, in the future, what becomes exciting is these selective glucokinase activators actually come through into clinical utility. We basically should actually focus on using these perhaps as preventives and perhaps as early intervention in people with type 2 diabetes, with established type 2 diabetes. And what are your thoughts about emerging therapies or directing therapy to treat hepatic steatosis? That's a difficult question. So I think Matthias is going to address that indirectly, because so far, the best therapy for hepatic steatosis is significant weight loss. So bariatric surgery is one of the best ways of improving or significantly decreasing hepatic steatosis. Whether we want to be directly fiddling with hepatic metabolism specifically for a condition which, to some extent, is ameliorated very significantly by caloric restriction, I think the jury is still out. I don't think T3 or high doses of vitamin A would be a reasonable solution in this setting. Thanks. Question coming in. With the model of partial pancreatectomy leading to insulin resistance, thus type 1 diabetes could have insulin resistance as well? Oh, definitely. I mean, the people who are newly diagnosed with type 1 diabetes are very insulin resistant. And in fact, the honeymoon period, to some extent, is a reversal of that so-called glucose toxicity that improves hepatic insulin action. Incidentally, those people have very high glucagon concentrations, which may reflect the catabolic state and the high amino acid flux that's going on at the time when they're catechetic. So it's an interesting scenario. And lastly, any new knowledge on the genetic basis of controlling fasting versus postprandial and hepatic suppression? Yeah, very insightful question. I mean, there are more than 200 loci now weekly associated with type 2 diabetes. The interesting thing, and with every passing day, the more I realize this, is how relatively poor we are at being able to measure insulin action and insulin secretion in the fasting state as opposed to the postprandial state. So in the model systems that we have used, we're very good at dynamic changes, detecting dynamic change and relative change. But in the fasting state, so far, we've not progressed much more than a ratio of insulin to glucose and glucagon to glucose and glucagon to insulin. So we are working through this. And the other difficulty in matching this with genetic variation is that I focused on TCF7L2 because TCF7L2 is the one with the strongest effect. But even that effect isn't very big. So you still need to study, you know, 60 people with each genotype. And that's the sort of things we do with traces is prohibitively expensive. Yes. Well, thank you so much. I mean, with that, I'd like to thank you again, Dr. Bella, for such a beautiful talk. Thank you. Thank you for having me. Our second speaker, Dr. Matthias Chip, received his MD from Ludwig Maximilian University of Munich, where he also trained as a physician in internal medicine. After a postdoctoral fellowship at Eli Lilly Research Laboratories in the US and establishing an independent research laboratory at the German Institute of Human Nutrition in Potsdam, Matthias returned to the United States to serve as the Director of Diabetes and Obesity Center and the Arthur Russell Morgan Endowed Chair of Medicine at the University of Cincinnati. In 2011, he took on a position of Scientific Director at the Helmholtz Diabetes Center at Helmholtz Zentrum München and was named Chair of the Division of Metabolic Diseases at the Technical University of Munich. He is the first German physician to receive the Alexander von Humboldt Professorship. And in 2018, he became CEO and Scientific Director of Helmholtz Zentrum München. Matthias Chip is a member of the Bavarian Academy of Sciences and the Academia Europa. He has received numerous awards for scientific achievements, including the Outstanding Scientific Achievement Awards by both the ADA and the Obesity Society, the Hanson Family Award, the Erwin Schrodinger Prize, the Paul Martini Prize, the Keres Medal of the German National Academy of Sciences, the Paul Langerhans Medal, and the Endocrine Society Outstanding Innovation Award. So please join me in welcoming Dr. Chip as he discusses novel single molecule approaches to polypharmaceutical therapeutics and their potential to impact the treatment of metabolic diseases. Dr. Chip. Well, thank you for the kind introduction. And it is a real pleasure to be invited by the Endocrine Society. I've been a lifelong member of the Endocrine Society and I'm really glad I have the opportunity tonight to speak to you about a topic that has been pursued in my lab for the last two decades. It's about hormonal hybrids for the treatment of diabetes and obesity. Some of you may have heard my lecture about almost exactly 10 years ago at the American Diabetes Association for the Outstanding Scientific Achievement Award. And this is about that same topic. It's just a jump of a decade now. And let's see where that's brought us in terms of going from a concept back then to what now is something that is emerging as a clinical candidate. I start with the disclosure statement though, which is rather lengthy. And it basically just says that I have a complicated life, but none of what you see here has to do with what I'm telling you today in terms of content and relevance. We are living in special times, certainly, not only because we're in the midst of a COVID pandemic and for a plenary lecture at the Endocrine Society, I would have loved to be personally there with all of you. We'll hopefully get to do that next year, but it's also a special year because this is 100 years after the insulin was discovered, after the discovery of insulin. These four gentlemen were all participating in this major breakthrough and they achieved something tremendous. They turned what was a clearly deadly disease into something that became a chronic, manageable disease. Now, much later, another disease, obesity, was something where we saw a big discovery, a big breakthrough. My friend, Jeff Friedman, now, what is it, 27 years ago, discovered that leptin is a hormone that comes from fat tissue that regulates foot intake and body weight. And that triggered a second wave of metabolism research. And it basically also was a trigger that jump-started my career as a scientist. I was a clinician when I saw that paper coming out and then started a postdoc and really went into research from there on. And one of the observations with leptin that really fascinated me was how well that translated from an observation made in mice to something that was then seen and reported by Steve O'Reilly, Sara Farouki, and others in humans. Not everything that we are working on in metabolism research has that example, that you can go from a mouse model to a human model and a lot of it replicates. And in this case, indeed, if you are a mouse or a human and you don't have leptin, there is massive obesity. But replacement of the hormone leptin can really correct that. Now, we learned a lot of lessons based on the leptin discovery and many years of research to follow in hundreds and thousands of labs. One of the key lessons in my mind is that leptin acts in the brain and most of what it does to systemic control of metabolism, it achieves by acting in the brain. So the brain is very, very important for the regulation of metabolism, of energy balance, of body weight, and of course, of appetite and hunger. Now, when inspired by Jeff's work and going into a postdoctoral fellowship, at that time, it was unusual, but I went into industry and joined Eli Lilly and company, Lilly Research Laboratories. I wanted to understand better what leptin does and was looking for maybe other factors that would participate in the regulation of body weight. And the one we ran into and reported in Nature in 2000 is the hormone ghrelin, which is secreted by the stomach and also targets the brain. Now, not dissimilar to what leptin does, but it's a leptin opponent, say hunger hormone, if you so will, that has weaker action profile than leptin in terms of regulating body weight, but seems to work at least overall on some of the same structures in the brain that leptin has as a target. So it was really fascinating to see at that time that there is more than one player involved in the regulation of systemic metabolism by acting in the brain. Now, ever since we have learned so much about the role of the brain in metabolism, I think some of the lessons learned come from genetics. So this is a slide that I think shows very clearly, but I've learned from Stephen O'Reilly and others that if you look at the overall collective data sets of genetics in obesity research, it really looks like obesity is a brain disease. Most of the genes are expressed in the nervous system or code for some molecule that functions in the nervous system. So you put that together, now two of the major hormones that are regulating body weight act in the brain and the fact that most of the genes associated with obesity in human population studies also seem to be located in the brain. Question then becomes, well, is obesity a brain disease? Now, at that point, I need to introduce my friend and mentor and collaborator, Richard DeMarchi, who for two decades was now the other lab that closely interacted with my lab, now in a truly interdisciplinary collaboration because Richard is a chemist. And as somebody who now didn't do well with chemistry at school, I had a steep learning process in understanding the mutual language between the two labs, but it really paid off that after we had met at Eli Lilly back then and both went into academia shortly after, we reconnected and between a physiology pharmacology lab of an ex-physician and a chemistry lab of a leading scientist in the peptide field, we said, well, is there something we can do together to come up with novel concepts, novel therapeutics that would have maybe breakthrough potential for these devastating diseases associated with obesity and diabetes and those pandemics? So we started working together and it was a lot of brainstorming and a lot of vision and a lot of trial and error, of course. And so here are three guiding hypotheses for what Richard and I embarked out on doing with our teams. One was, and I just shared some data in that direction, that obesity may just be a brain disease. Therefore, if we wanna really find novel therapeutics to prevent or cure obesity, we gotta target the brain somehow. Now, if you do that, it is complicated because the structures you wanna target have receptors and molecules and neuropeptides and neurotransmitters that are a little bit promiscuous that you find in other regions where it's more about memory or mood or depression or motor activity. So if you wanna really selectively find the multiple places where metabolism is regulated in the brain, it seemed to us we gotta find something that's indirect. Use mother nature's toolkit. Factors, afferent factors from the periphery that are designed to find places in the brain that then specifically react to control metabolism systemically. So we did that and based our overall vision on endocrine factors that would indirectly target hopefully the right structures in the brain. And the third part of what made up our overall guiding hypothesis was, no, it's not gonna be one. I mean, we had done so many experiments on leptin and ghrelin antagonists and so on, and just always appeared that what the brain really sees is a pattern of hormone signals that make up a certain combined signal that then the brain reacts to in terms of readjusting metabolism overall. All right, and with that, we said, okay, now where to start? And by then, this was an explosion of discoveries. If you just look at the gut-brain axis, there are so many different players and receptors and targets it was difficult to choose from. And frankly, there was some information coming from bariatric surgery trials, gastric bypass and so on, where we saw, okay, some of these gut hormones seem to be normalizing or changing some directions after a gastric bypass surgery, for example. So maybe they're contributing to that benefit that's clearly impressive, even though this is a highly invasive, irreversible procedure. Now, that was one guiding hypothesis. The other one was, you know, there are some hormones that seem to be doing some good to metabolism and we know about it. And, you know, there's GLP-1 that, there are drugs that are based on that, monoagamists. So can we find something that's pragmatic? It builds on something that's clinically proven, but goes beyond that. And sort of tries to, you know, move toward a version of a gastric bypass in an injection. All right, that brought us, you know, initially to deal with the glucagon peptide family. Now, GLP-1 is part of that, GIP is part of that, and of course, glucagon. And when you look at that family of hormones and their sequence, there are some similarities. And glucagon, you know, strikes everybody initially as, isn't that the wrong hormone, right? These are pre-diabetic obese patients we're talking about here, for example. You really want to give them glucagon because that drives up blood sugar. Now, it does more than that. If you look at older literature, and there is not a lot of that because there isn't, it isn't easy to work with glucagon in the lab. It needs to be, you know, soluble and long-acting and stable. And making versions of glucagon that fulfill these requirements is something that, you know, a fantastic peptide lab like Richard's can do, but not necessarily was something that was available to everybody. But still, there were some really good physiology studies done, and it shows that glucagon can increase thermogenesis, so help us burn calories. It can increase satiation, but it can also help with lipolysis, et cetera, energy expenditure. So we thought, well, you know, maybe we've got to stop thinking about glucagon receptor antagonism to block things that drive up blood glucose. But we atomize the glucagon receptor, we burn calories, and whatever bad thing glucagon may do to glucose metabolism, we use GLP-1 to override that. The idea is combining different endocrine factors, different hormones, and then have truly synergistic action profiles. Now, enhancing glucagon is something that, again, took a while, and it wasn't easy, but there was lessons learned from the chelomonster that you see here, in which, in the saliva of which, there was a lot of lessons learned to turn a short-acting to a long-acting GLP. And the long-acting GLP-Xendin has a tail that then can be also used to enhance other peptides. CE-X tail can be put on glucagon, and then glucagon becomes more long-acting. And if you do that, then you can better study glucagon, and you can also develop chemistry that goes into a single molecule that's a glucagon-GLP co-agonist, a true dual agonist. And you see some data here that are now, you know, more than a decade old, but still, I think, pretty mind-blowing that if you put these compounds once a week in a mouse that's obese and insulin-resistant, you will see that 10% of gluco-connection, or 100% of gluco-connection, in red and blue, are really making a difference in body weight loss on top of GLP-1, which is, of course, in that molecule as well, 100% in this case. So everybody was like, yeah, well, you just made a fantastic super-GLP. You just show us that there's no side effect on glucose and insulin. You know, if there's really glucagon in there, you're gonna be in for a surprise. Turns out, yeah, it was a surprise, but in the other direction, because we seem to be normalizing insulin levels, and we seem to be normalizing glucose tolerance in these mice, even though we gave glucagon agonism together with GLP agonism. Now, the next question was, well, is there really glucagon in there? Because this looks too good. This is probably just a really well-done GLP-1 receptor monoagonist. Well, it turns out, and this is a collaboration with Dan Drucker back then, that if you use mice that don't even have a GLP-1 receptor and you still give these dual agonists, you still got body weight change. And the more glucagon action you dial into the molecule chemically, the more you drive down body weight and drive down fat mass. Now, this was in mice, but by now, and there's a recent study that was just added this year, there are several trials that came out in academic journals that show that in humans, indeed, a glucagon GLP receptor agonism drives down body weight. It may be mainly something that is helpful in decreasing fat mass. And of course, it's all about finding the right balance between how much GLP do you need, how much glucagon do you wanna afford to put in there so that still GLP and the increased metabolic flexibility after the weight loss override the glucagon action. That was the dual agonist one, and we thought, well, a lot can happen in the coming years. And of course, a lot always happens when you go into developing and translating drugs. So we thought we need more shots at the goal. And the next we wanna focus on is something without glucagon. If glucagon turns out to be doing something funny in human trials or not be behaving the same way as it does in preclinical models, let's deal with GLP-GIP dual agonism. Now, that's much closer, both incretin hormones. We call it twin-cretin. And still, we thought there was a chance that these two would do something different to a sufficient extent to cause synergy. And that was indeed true. Actually, what was surprising to us, and here are some data from the first publication ever on a double-incretin single molecule that we published in Science Translational Medicine in 2013, it drives down body weight, and that seems to have to do with GIP on top of GLP. Again, a surprise when you were basing your hypothesis and conclusions on global knockout mice for GIP receptors, based on which one would have to be preferring to antagonize the GIP receptor. Now, that seemed to translate into humans as well. And this is a first human trial that we co-published with several colleagues from pharmaceutical companies showing that, in principle, this works in humans, a dual GIP-GLP to both decrease HbA1c to decrease body weight. But then, you know, fantastic studies came out about two or three years ago by Eli Lilly and Company with another version of this double-incretin tercepatide that is slightly different in the sense that there's even more GIP in there compared to GLP, so more weighted in the balance toward GIP. But it showed that, you know, there's very impressive body weight change, body weight loss dose-dependently, very impressive HbA1c change. And this may just go beyond of what we see with mono-agonism. Of course, you know, there's currently phase three trials, ongoing large trials, and we will see what that brings. My personal opinion is that if the dual agonists and triple agonists that I'm talking about today will be successful in terms of, you know, becoming best of class and being approved in a couple of years, it may have to do with an increased therapeutic window. Because as you see here, the dual agonists, especially the double-incretins, they seem to have, when you compare it relative to the action, relative to the efficacy, less side effects that last less long, that last for a shorter duration of time. So therefore, you don't have to dose as high, but you could dose higher without causing the same amount of side effects that you see with mono-agonism. So suddenly, there's more flexibility because since you're occupying, I figure, some GIP receptors and some GLP-1 receptors and never have to go to, you know, close to 100% of occupying receptors of each target. And therefore, you don't get to the place where maybe then side effects are overwhelming. That is, of course, still philosophy until it's proven. Why is this working, right? And this is a very difficult question because frankly, there are data indicating that, at least pre-clinically, if you block the GIP system, some good things happen as well. Well, it seems that the dual agonism may be the more impressive one based on what we hear from clinical trials and the final answer will come in the future. Now, what you see here are some of our studies. This just came out this week in Molecular Metabolism where Timo Müller and colleagues find that GLP-GIP shows faster recycling of the internalized GLP-1 receptor. As a matter of fact, there's also less internalization of the GLP-1 receptor, meaning that there's more GLP-1 receptor on the surface. This can be acted upon when there is GIP on board with a dual agonism. So there is a mutual enhancement effect on a cellular molecular level. But there's something else, and I think this is really important. Turns out, now this paper is coming out today. And when I say that, then it's February 10th because this is where I'm recording this today. So it's not at the day where you will be seeing this lecture. But it's recently published. And this is, again, work coming out of a lab by Timo Müller in collaboration with Richard DiMarchi and many others. Where I think there's clear proof here, as you can read in Cell Metabolism, that it's the brain GIP receptors that mediate the benefits of GIP agonism within a dual agonist that is GIP-GLP receptor target. So in other words, what you see here is that a chronic ICV infusion of a GIP, that's long-acting, indeed decreases body weight. And this has nothing to do with decreasing GIP receptor expression. It could have been sort of a paradox phenomenon. You give GIP, GIP-1 receptor is downregulated, and then it's sort of an antagonism that helps us with metabolism. No, that's not the case. GIP receptor expression stays the same, but GIP action, when you put it into the brain, helps with decreasing body weight. Now, this is a pharmacological, again, a function study. How about a genetic loss of function study? There are indeed superior benefits of GLP-GIP over GLP, as you see on the left side in wild-type nights here. There's a substantial additional phenomenon in driving down body weight. Well, it turns out, when you look at mice that are only lacking GIP receptors in the central nervous system, you don't have that benefit anymore. Well, that means that GIP does that job that it adds on to GLP action in the brain. This is a nesting premediated CNS-GIP receptor knockout that, of course, could be leaky to some extent, and there could be some GIP receptors that are missing in the periphery, but it was very carefully checked for these experiments, and you can read it in the paper that this has nothing to do with GIP receptors missing in the pancreas or elsewhere, because that didn't happen in this case. This is a targeting of the brain. So we believe that GIP-GLP double increments have real potential, and we will see, hopefully, next year or so, what data say from large clinical trials, and we believe this is the mechanism. It's a brain-mediated GIP receptor action, and that speaks to what I was telling you in the beginning, Jeff's discovery on leptin and actin in the brain, and our work on ghrelin acting in the brain, that the brain just is really important for a target when you wanna treat obesity. So, of course, if you can't do dual agonists for GLP and glucagon, and dual agonists for GLP and GIP, you can do a triple agonist, and that's what Richard Slack was able to do, and we went a lot of back and forth. This is how this looks. You take the natural hormone, and you just add that CEX tail, a few small changes that make it more long-acting, or albumin binding, et cetera, is increased. But really, it's very similar to an endogenous hormone, which we think is important. This isn't just sticking several peptides together, making it something non-natural. No, this looks very similar to a natural hormone, very similar to a mono-agonist, and therefore we think antigenicity may not become a huge problem. So when you look at these tri-agonists, they're even better than the dual agonists in pre-clinical models, really driving down body weight aggressively, clearing the liver of fat, et cetera. Of course, there are clinical trials for GLP-glucagon, for GLP-GIP, and for the tri-agonists, as we call them. I personally think tri-agonists may be the most interesting ones, because there you have the double-incretin, and you have a little bit of glucagon that is very nicely balanced by a double-incretin action. So based on our own data, this should really be the most efficacious drug, and we just need to see if it's safe, and there's clinical studies ongoing on all of these. Now, let me end with a view into the future, perspective for the future. We worked hard on coming up with better and better models to test these poly-agonists, and to even make them better. Something that took us literally years, several years now, is a mouse that has neither GIP, nor glucagon, nor GLP-1 receptors, a triple knockout mouse line. This was possible using CRISPR technology. Alberto in the lab here in Munich made this, and it's real magic, because you can at the same time knock out these three receptors simultaneously targeting deletions of XM4 and 5, and all the three genes via CRISPR at the cycle stage. Now, it turns out, believe it or not, that these mice are alive and well, and we just started phenotyping and comparing Y-type mice, single knockout mice, dual knockout mice for all of these combinations, and the triple knockouts that seem to do surprisingly well. But of course, it gets really interesting now when we can administer all these dual and triple agonists and future combinatorial single-molecule drugs we want to test in these models. Now, with that, I would like to state the obvious that I haven't done all of this work, not even close. This work has been more recently led in the lab by Timo Miller, important contributions by Aaron Novikoff, Alberto Serrano, Qian Zhang. Without Richard DiMarchi and Brian Finnan, this would have never been possible. Brian, I think, is a beautiful example of our interdisciplinary collaboration. He's a peptide chemist who trained as a graduate student with Richard and came to Munich and learned how to do physiology, how to do pharmacology, how to party at Oktoberfest. But he now is a director at Novo Nordisk where he does other great work, but he really was key in publishing a lot of these reports that are highly cited today. Dan Drucker, Randy Seeley, Darlene Sandoval are a great inspiration and wonderful collaborators that always keep us sharp and point out where things may not be as simple as we'd like them to be, but also have been giving us tools without which we wouldn't have been able to prove mechanistic work. Christian Bollfromm and David Calibiero have been helping us recently with the papers that were just published this week. And with that, of course, research isn't possible without lots of funding. There are lots of institutions and companies and funding bodies that we are grateful to. I want to mention the Helmholtz Association and the Helmholtz Center, European Research Council, but also Alexander von Humboldt Foundation and the National German Diabetes Center. With that, back to the conflict of interest statement. And I would like to thank you for the attention and I'm looking forward to a discussion. Well, thank you, Matthias, for just a wonderful presentation. I see a few questions in the chat, some of which you've answered, but let's get to it. And I'll ask you a few more in our closing two minutes. A question from Dr. Morata's Flyer. The emerging clinical data with tears appetite show increased nausea compared to GLP-1R alone. Do you think this is action of GIP directly on GIP receptor in the brain, or do you think the effect dual agonism to induce faster recycling and less internalization of GLP-1R to allow more time for GLP-1 agonist on the receptor? Well, that is of course an expert question from an expert. I think that indeed the fact that we have a combination of GIP and GLP will give us a larger therapeutic window, as I said in the talk. If the side effects that are probably something that will occur more in some subpopulations than in others or more on the side of GIP or GLP remains to be shown, we can only speculate. I still believe that overall we will likely have, especially then maybe even with the triagonist, a more flexible therapeutic window where we don't have to dose even as high as we have to with monoagonist, therefore prevent these side effects. But certainly some additional experiments are necessary to find out which component contributes which. Then some brief questions, short questions. What about the possibility of these polyagonists inducing an immune response? Yeah, Gary, that's always something we had on our radar, but based on the available data to date, it seems that there is no neutralizing or limiting process ongoing with some sort of antigenicity or immunogenicity. So it seems that the similarity between these polyagonists and endogenous gut hormones is so significant that we might get lucky there. And your data suggests that these polyagonists cross the blood-brain barrier, yes? In a sense, you would assume that. I personally think they might, nobody has really shown that, but I personally believe it's more likely that they will just act within the area, the vicinity of the circumventricular organs, like the median MNN stereopostrema, where then they have a significant effect that indirectly goes into the brain. They may cross the blood-brain barrier, but they may not even have to, I believe. Circumventricular organs where we have a lot of receptors for GIP and GLP in the vicinity may be just where it happens. And your data suggests that the di- and tri-polyagonists have benefit over using multiple single agonists, and it certainly suggests that there's additional emerging properties of your polyagonists. Can you kind of, can you comment on that? Well, I think the properties in terms of their biochemical properties are very similar. They're a little bit more long-acting, of course, and made soluble so that it become a better drug. But in terms of their action, we always look at three things with dual agonists. Now, what does GIP do? What does GLP do? And there is synergy between the two. And one aspect may be that, for example, the GIP action may lead to more GLP to be delivered into the cell or be active at the receptor. I showed some data related to that. I think there is much more. There is that third entity of synergy, not only downstream synergy in terms of what happens in systemic metabolism, but actually also on a cellular level based on receptor interaction. Well, thank you so much, Matthias. I think we've run out of time, although we could talk for another hour. But with that, I'd just like to thank you again for such an outstanding talk. Thank you so much. Thank you. My pleasure. And with that, this session is now concluded. And I want to thank all of you in the audience for participating in our final plenary of ENDO 2021. In a few minutes, we have several other sessions starting that you can join. So bye for now.

type 2 diabetes

defects

high fasting glucose

inappropriate fasting insulin levels

insulin secretion

hyperglycemia

glucagon

glucose production

hepatic metabolism

glucose homeostasis