 Yesterday we highlighted that we've all evolved to optimize our capacity to sense and respond to our environment. But the challenge that we face in the 21st century is a growing evolutionary mismatch with the pace and cultural technological changes that are outpacing our adaptive ability. A long standing challenge in the advancement of any field, but particularly in the fields that we've been talking about over the last day, is an ever increasing requirement for specialization, along with a need for what has been termed by EO Wilson and others as a resilience, or a jumping together of knowledge by the linking of facts and fact based theory across disciplines to create a common groundwork of explanation. Waiters not working. There's a challenge. What we learned yesterday is there's a challenge in linking metabolic epigenetic interactions with great brain interactions to predict cellular behavior. But as was so nicely highlighted by our speakers, was that computational tools provide an opportunity along with commonalities across species to help us understand these linkages across scales. And I think for our workshop, importantly examines opportunities and challenges in leveraging our growing understanding of the gut microbiota immune neurological access. All electronic medicine we learned yesterday has spanned disciplines over two decades of research and has led us to understand that the stimulation of the vagus nerve can attenuate the severity of a variety of inflammatory disorders including rheumatoid arthritis. And intriguingly that specific neurons are responsive to very specific unique inflammatory mediators. Decoding these connections can be a challenge. But machine learning approaches provide new opportunities to allow us to decode these signals that have been previously considered intractable and appreciation of readily available technologies. That spans multiple different disciplines from biology to engineering to computer sciences are now offering a significant amount of hope for patients with inflammatory bowel disease rheumatoid arthritis and other acute and chronic inflammatory diseases. We also learned that while behavior can be modulated by genetic or environmental factors, we now have an increasing understanding that microbes can modulate behavior independently. A variety of certain different findings have led to the discovery, as was highlighted yesterday by our speakers that lactobacillus ruteri specifically can rescue social deficits in mouse models of autism. There've also been six preclinical trials and a variety of animal models and two small independent human trials, but human trials nonetheless that suggests that lactobacillus ruteri is safe and that there's a potential for this new field of precision microbial intervention. With all that said, we need to understand the mechanisms that underlie the effects of bacteria on behavior if we're really to advance and optimize this field. Microbes can regulate the food that we eat and the neurochemicals that are generated, both directly by bacteria and through a variety of metabolic intermediaries to protect against seizure disorders. And we learned from our speakers yesterday that the opportunity exists to identify specific microbes, microbial products and nerve targets and to determine whether the applications of specific microbes can replace a variety of dietary interventions that have been identified for patients with seizure disorders and a variety of other ailments. Studies of Parkinson's disease have taught us that motor symptoms are often preceded by non motor GI symptoms, sometimes for by many years and even decades, and that treating GI symptoms have been described to affect the development of motor symptoms and Parkinson's disease. That has led to the hypothesis that this neurological disorder might actually originate in the gut. We learned that alpha-synuclein when misfolded can promote neuronal dysfunction and we also learned that mycoglia are potential target for prebiotics. Dietary interventions may limit neuroinflammation and potentially have an impact in either the prevention or mitigating the symptoms of Parkinson's disease. I think the biggest challenge that we highlighted yesterday were the potential roadblocks to the translation of the biology of the gut brain access to therapeutic interventions. The personalization of microbes across different people with different genetics and different backgrounds is clearly a barrier to intervention. The development of effective medicines is ultimately guided by a fundamental understanding of the mechanisms of action of these medicines. But how do we balance the real world immediate need for therapeutic interventions for patients and families who are staring into the abyss with the need to spend time and resources to understand the mechanisms required to ultimately optimize the therapeutics that we introduce. And fundamentally somebody has to pay for all of this. And someone, some entity or organization is asking what will be the return for this investment. And in a resource constrained environment, we're always challenged by the question of how do we prioritize investments for effective therapeutics and interventions. This research space also spans so many different fields that there are challenges to the traditional structures that have evolved in science and academia. And how can we think about building the research teams that we actually need to drive this highly interdisciplinary field forward. There's a critical need to identify talented people with varying different varying perspectives, but also who bring unique technical technical skills to bear to solve these problems. Critically we need increased levels of conversation and communication across specialty areas to become great problem finders and problem solvers. Ultimately we need to help trainees develop the applicable knowledge so that they can address these problems in a truly cross disciplinary approach. And consider changing the structures to incentivize multi disciplinary teams to tackle these very difficult areas. The most important question is how can we make the greatest advances in the fields and where do we hope to find ourselves 10 years from now. Anticipating that multiple interventions may be needed to address a disorder, rather than a simple monotherapeutic intervention is probably a very complex challenge that needs to to be recognized and appreciated. There is a need that has been emphasized throughout the first day that fundamental mechanistic studies regarding deciphering how circuits in the brain and the gut work individually and together will clearly be needed to accelerate discovery and translation. We will also need to change our framework and our culture to expand clinical trials and particularly community based clinical trials, all the while reducing the time for these trials and their cost. And clearly, there will be a need to choose carefully and wisely to mobilize all the resources. And there's nothing like success that generates success. And those early successes will clearly be huge motivators for the field as a whole. In the context of all this the question is, is our regulatory environment, fundamentally structured to promote and accelerate these changes and to accept the inherent risks, and the important ethical questions that will ultimately arise. All too often, culture and technical change outpaces the regulatory environment in which we work. I think it's been a wonderful first day of this symposium. And I think we look very much forward to the outstanding speakers that we have for the second day. It's a pleasure to introduce Kavita Berger who will be a co-moderator with me for this very first session. And Kavita, if you can come on up, and I will introduce the first speaker that we have this morning, who is Professor Mark Light. Dr. Light is the Eugene Lloyd Chair in Toxicology and Professor of Veterinary Microbiology and Preventive Medicine at Iowa State. Dr. Light completed his undergraduate degree at Farley Dickinson and his master's and PhD at the Weitzman Institute of Technology. He pursued postdoctoral fellowships in immunology at the Medical College of Virginia and the University of Pittsburgh. Dr. Light is the recipient of the 2017 Leo Bustad Distinguished Lecture Series Award and has been a pioneer in the field of microbial endocrinology that represents the union of microbiology and neurobiology. We focus on identifying neurochemicals that serve as an evolutionary based language between the host and the microbiota environment to enhance health and wellness and understand mechanisms of disease. Dr. Light, the virtual podium is yours. Well, thank you very much. And I want to thank the organizers for this invitation. And I want to specifically thank Dr. Tricolfer for introducing us to evolution and mentioning evolution because I want to really emphasize evolution throughout my talk because evolution is telling us something. What you see here on the screen is social conflict stress between two animals where one animal bites another animal and introduces into that animal an infectious insult from the bacteria in its mouth. And according to what we all always thought about stress of this animal that is being bitten and highly stressed is that its immunity is going to go down. And so what I want to ask you does that make evolutionary sense for an animal's immune system when faced with an infectious insult to just say I give up and die. Well, that's not the case. And so what we did over three years ago was look at this in the laboratory where we had mice fight each other under social conflict. And when we examined the stress animal, we found that actually the innate arm of the immune system. Those cells concerned with phagocytosis and destruction of infectious bacteria on first pass were actually increased by over 500% in activity. It was the first time that the innate immune system in immunology was going to be increased upon stress and that it was actually not unchangeable. It was actually changeable. So we were really excited about those results and we progressed on to give them an infection and we gave them oral your city enter politic infection which causes a GI based infection. And to our astonishment, we found that those animals which had increased immunity as you can see here where this defeated they were defeated animals. They actually died faster because they got the infection. So you were faced with a paradox how can increase immunity of over 500% in terms of what the cells that are normally concerned with phagocytosis of bacteria actually lead to increased death of these animals. So let me ask you, does this make evolutionary sense? And the question of is for whom? Who are we talking about? We're talking about usually the host and its survival, but we also need to talk about the poor bacteria that was in the mouth of one of the animals and now has been introduced into the system of the other animal or the bacteria that was in the food and you ate it and now it is your gut and now has to find out what to do. So what are bacteria doing and what what are bacteria sensing? So what I'm going to want you to consider for today is that these bacteria are recognizing neurochemicals and neurochemicals as such are an evolutionary concerned language and evolutionary based language of neurochemistry between the microbiota and the host. And that this is one of the ways by which the microbiota gut brain axis is going to function. So with that in mind, let me ask the question again from evolution, where are these neurochemicals and neurochemicals are actually omnipresent in nature. So if you look at plants, I often bring up bananas because bananas in their peel have dopamine and norepinephrine. And that, like us dopamine like or norepinephrine like it's exactly the same as the structure as what we have as animals have. And as a banana actually browns, if you're going to make banana bread, and you leave it on your table as that skin actually browns, it actually increases this amount of dopamine and norepinephrine in the peel. Now, back in the fifties, if this is a cardiac meeting back in the fifties and we were all the cardiologists were meeting. We were telling our patients you cannot eat bananas because you're going to start off a cardiac episode. And that was not changed until this was published in 1958 in science. When to get into science in those days, all it took you to do was to peel a banana. And that's all this person did. They peel the banana and they showed that there was a division of catecholamine between the peel and the pulp. In the pulp of the banana, there was actually little amount of catecholamine. It was all in the peel. So then it became okay for cardiac patients to again eat bananas. Other common food stuff. If there are neuro neurologists in the audience, they know that if you prescribe a psychoactive drug regimen, you put your patients on a restricted diet. So they don't eat tomatoes and other plant-based foods that have neurochemicals that may interfere with the treatment. Now that's neurochemicals and plants. Obviously, before they, we ever as animals appeared on the planet and that extends to potatoes in the ground. Everything has neurochemicals like we do. But what about bacteria? In bacteria, we know that probiotics have GABA. So when lactobacillus ruderi was mentioned yesterday, the first thing I was thinking of, do they know what the GABA potential in these bacteria are? Because GABA, gamma-unipotaric acid, primary neuro-inhibitory transmitter in our brains, was described decades ago in lactobacilli. Lactobacilli makes other types of neurochemicals such as acetylcholine. What are they doing it? Why are they making it? We really don't know the answer to that. So metastatin is found in Bacillus subtlet. The catecholene biosynthetic pathway that animals have of aldopa, dopamine, norepinephrine, and epinephrine, that exact biosynthetic pathway is found in E. coli. Not like our pathway, the same as our pathway with the same cofactors that are used in the catecholene biosynthetic mechanism. Actually, in E. coli, also high affinity, 100% homology with the environment being has been found for opioid-binding sites in E. coli. So there's a lot of data out there to show that there is a reason why these things that have been there and that neurochemicals are omnipresent. So there's a common evolutionary thread here that there's this relationship between microorganisms and the host. And in fact that the evolution of cell-to-cell signaling that we use in animals for our neurophysiological system may actually be due to late-horizontal gene transfer from bacteria. There's an excellent paper by Iyer in Transigenetics going back now nearly 20 years that discusses this. So microorganisms in the gut, such as those that are really anywhere, skin everywhere, do not simply rely on traditional nutritive energy sources. So that's one thing I want you to, in a way, unlearn from your microbiology experience that you have to grow them in this broth culture and they need nutritive energy sources to do what they do. They actually are relying on a lot of non-nutritive energy sources for their survival and behavior. And this means that there is a direct interaction between the neurochemistry and bacteria and the host neurophysiology and that bacteria are not dumb bugs. They simply do not divide 1 to 2, 2 to 4, 4 to 8 and so on, but they are interactive and they will change their behavior and physiology based on what they sense and what they possess in the community that they're around. So they're an interactive player in health and nutrition. So let me jump now to say what this comes together as. This comes together as the field of microbial endocrinology. So when people were talking about yesterday about that now a lot of this is accepted, you could imagine over 30 years ago when this was first proposed, what type of reticence was greeted when the proposal was put forward that you have to join the fields of microbiology and neurobiology. The neurobiology has a reason to interact with microbiology and vice versa. And that form back then I put forward the proposal, the hypothesis that that forms microbial endocrinology and that this interaction of bacteria and neurochemistry, whether the neurochemistry originates in the bacteria itself because the neurochemicals are present in the bacteria. Are they using it for community structure and regulating community structure? Just as Dr. Elaine Shaw mentioned yesterday about microbial communities, what's happening, there is a reason that they have them. But for today we're talking about host physiology so that this interaction between the host physiological neurophysiological system and the bacteria means this is microbial endocrinology and that can affect disease and behavior, the gut to brain access. And that stress and nutrition because you have neurochemicals in your foods all interact in this complex interaction and that this will ultimately affect human and animal health and evolution is being throughout. So let me show you what the relevance of this is. So a number of years ago when I was in the Department of Surgery, one of the things that I learned was of course one of the big problems in the surgical field was the fact that there are indwelling medical device infections and those indwelling medical devices such as pacifist for example that we use in patients often get infected with stachococcus epidermis. So that is the most common bacteria that will affect pacifist and other indwelling medical devices and this will of course cause huge issues for the patients and removal of this device and replacement. And it was not understood why these bacteria are not normal skin bacteria would infect these devices after because the obviously the point of insertion of the catheter is cleaned by the nurses, Pavadone, iodine, how does this happen? So we looked into this from a microbial endocrinology standpoint and as we published in Lansing in 2003, we finally explained over nearly a half a century of not understanding why this happens, why this happens. And what you see here is a is a catheter the instead of a catheter, which we have put a medium I won't go into now, which is reflective of what's in the patient. And we have put on to it, six, four to six, that's happy very low amount four to six clinically relevant amount. There is nothing there there is no growth that is simply background noise. But now, if instead we subject these four to six bacteria to the same thing that is going down the tube of the catheter, what is the physician giving this surgical patient, they're giving them catecholamine in atrope to prevent low blood pressure hypotensive effects following surgery and another medical condition. And what we show is this that these catecholamine in atropes will stimulate these four to six bacteria within a matter of hours to do this. They make this huge biofilm and as you can see here, here are all the hallmarks of the biofilm on this device. So we were able to show this that this catecholamine in atrope, which could be do butamine, or as I learned after the first time I gave this talk at a surgical affection to these meeting, the chief of surgery, who was actually the session chair came up to me and first question I was asked was, do you know what we call norupinephrine bit heart rate that we give to patients as a catecholamine in atrope. And I said, I don't know the name. And he said, well, it's called Leva Fed. And then he asked me, what do you know what the nickname for Leva Fed is. And I said, no, and he says, well, give him Leva Fed, leave him dead. And is this an explanation of why patients may develop sepsis afterwards and stimulation of low amounts of bacteria. And as you can see here, the answer is yes, because there is a powerful stimulant and this is not nutritive. This is micro molar amounts of a catecholamine can do this to staff at the. Since then, a number of labs have repeated that. And the rest so I won't go into that into a number of organisms E. coli salmonella which are similarly can be affected and change your bacteria physiology. But what we wanted to do then was to ask the question of, well, if that worked for infection, what about the microbiome gut brain access? Can the brain actually see what's in the gut? And as you see here in the title, I say the modern era demonstration of the axis, because as I will touch on briefly a little later, this microbiome gut brain access, although we assume it's rather new in how we're describing it, goes back to the 1910 over 100 years. People were doing the same things that we're doing now in terms of study. If you talk about lack of a still giving to patients with neurological issues, they were doing that in 1914, and using some of the same bacteria we're using now, we need to be aware of that literature. So back in 1998, we gave teaching to unite to mine, as a novel bacterium that did not induce an immune re activation, obviously, can't live after in humans is an infectious agent, but in mice, it's not, it just simply passes through the system, replicates, but doesn't cause immune activation. And when we did this, we found out that it actually was it gave induced anxiety like behavior and animals, which we gave it to something was going on. And a few years later, when we actually sectioned the brain and did double blind studies on this, we found that we were able to show that here is defoss activation, you see here by the black dot, that you actually activated centers in the brain, following this, these bacteria, these novel replicating bacteria going through the gut. And that if you did a bigotomy of these animals, you completely ablated this result, you cut out all of the guts to brain axis that was going on. So there is neurochemical potential and responsive civility within the microbiome, and that studies which have shown the ability to microbiota to produce a wide range of neurochemicals. So if you go in the literature, you will see that the microbiota can produce a enormous plethora of neurochemicals. And again, I want to emphasize these are not simply like ours, but exactly the same structure. And that human microbiome studies what I found fascinating have shown that after antibiotic, the drug class that most influences the microbial diversity in the gut are the antipsychotics, which to the authors of the paper Nature a few years back was quite a surprise that antipsychotic would influence the microbiome. What we did and when we began looking at this was to ask the question that we chose a model system to say, if we look at fluoxetine, which is many of you know isn't an SSI and is one of the most prescribed antidepressant, we chose to follow up on that in a way by using fluoxetine and saying, can bacteria recognize fluoxetine and do they have receptors for this. And one of the reasons is because the use of fluoxetine in clinical practice often leads to patient noncompliance because they have tremendous shifts in their weight, and they don't like having that so they stop using the drug. So what we were able to show in 2018 was that bacteria actually possess biogenic amine transporters, the plasma membrane monoamine transporter and the organic cation transporter that are involved with SSRI in mammalian cells in its mode of action. And that these transporters like ones that were present in bacteria actually were more active than what they were on mammalian cells, and they were most found in lactobacilli were most responsive to this. And when we did subsequent in vivo dosing of mice with fluoxetine, we were able to demonstrate dramatic shifts in the microbiome, microbiome diversity, and that lactobacilli were the most effective of these organisms of the microbiota that was there and that was So there was a microbial endocrinology demonstration here. We don't know what the function of these biogenic amine transporters on our own lactobacilli, but they are there and they interact with the drugs that were given, which speaks to what was discussed yesterday about the effects of drugs on the patients and how do they work. So taken as a whole it strongly suggests that in depth focus studies on the neurochemical potential of the microbiome to influence the gut brain access are really warranted in the future. And so I'll finish up by saying where will we come from. So as I mentioned before, when I said the modern gut brain access, the emergence of the microbiome and the microbiota gut brain access goes back at least to the 1910, where they were using the same probiotics that we use now for identical purposes. There's an excellent article by Stowe in the Journal of Medicine and Surgery of 1914, where it concludes that the use of lactobacilli should remain on the battleground forever to prevent what he called melancholia and other neurological issues. There are other studies where they basically before there was an IRB, they gave patients whole broth cultures of lactose and other bacteria to drink and see if they could change their behavioral states. Many of these probiotics, as they said, produced neurochemicals. So there is a huge knowledge base today that lacks information concerning the capacity of microbes to make neurochemicals. And if that's one of the challenges we discussed today, I want to emphasize that to the audience to say our databases that we utilize to do machine learning and all our bioinformatic pipelines, we have to look at our databases and say, are they complete? Have they been curated? Do they tell us what they think they tell us? And for example, no database could have predicted that enterococcus fission makes dopamine. So the common bug in our gut, enterococcus fission, can convert adopa to dopamine with nearly 100% efficiency. We described that discovery back in 2018. And again, it could not be predicted from the databases. But there are huge bioinformatic issues with the databases. And I put up here last though versus bio bakery. For those who don't know, those are bioinformatic pipelines to use predict function in the microbiome. And I was on a conference call with collaborators concerning this where they said they were going to use last though. And I said, well, why not bio bakery? And their answer to me was, well, last though gives us significant results. And I said, are you telling me that if you put the same exact date in these different pipelines, which I kind of already knew the answer, you're going to get different answers. And the answer is, of course, yes. And the thing is, we have to stop searching for the magical p value of less than 0.05 and start to ask, why don't our bioinformatic pipelines give us the same answers. And we need to resolve that issue, which is a big issue, I think in the community, which brings up one of my final points, which is there's a need for what I call old school microbiology. We have to get back to growing our bugs in the lab and seeing what they do and not try to predict what they are going to do and use nutritionally relevant media. What I mean by that is, we almost all of our database information now comes from bacteria that are grown in microbiological rich media. I mean, Luria broth, LB broth, BHI, Brain Heart Infusion broth, MRS medium for lactose. These are nutritionally rich. These are media which are bacteria are going to stay plantonic. They're not going to grow as biofilms in these medias because you're shaking them in a fermenter at high speed. You're growing them. That's not the way they act when they are faced with nutritionally relevant media. What I mean that if you're going to eat food for what you eat at lunch and at breakfast is not what you give the bacteria. When you give them food that we eat and make a medium approach in your lab that is constructed of the food that you actually eat, that you or your animal eat, you will get completely different results on what your bacteria are going to do. So if your bioinformatics databases are constructed of results from microbiology rich medias that don't reflect what's actually in the gut, how do you know what you have? And our discovery of dopamine back in 2018 was made in nutritionally relevant media where we actually took animal food, we ground it down and made media out of it by passing it through phases such as the mouth, the stomach, and the small intestinal fate. And we made a media very much like what the pharmacological community does to test drug stability. So I think we need to get back to that old school microbiology. And let me finish by saying to keep in mind that microbial endocrinology is an evolutionary based framework linking the components of the microbiota post nutrition which which we can interrogate the mechanisms by which the microbiome gut brain access influences health and disease. It is a common evolutionary language by which all elements can interact. And regarding the industry and our translational approaches we're discussing today, you can actually adopt this to the design of probiotics by asking what is the inflammatory condition we're addressing in the gut? What do we know about it? What are the neurochemicals that affect that axis, such as dopamine and inflammation or other conditions? And so can you design your probiotics that way? And this has been done actually now and we're doing studies now with this in animals. So I'm in vet school and we like working with pigs and all the rest which obviously their GI tract is more like humans. It's been done with the discovery of dopamine producing probiotics. So we're putting this into action now by doing this directed mechanistic approach. And I'll finish up by finally by saying this, it's fully recognized that microbial endocrinology is only one of the possible and I'll accentuate that possible mechanism that that can affect the microbiome gut brain access. And there is a vast array of other possibilities that exist and need to be explored and that we are discussing at this meeting. So with that, I'll finish and again, thank the committee for the invitation. Thank you. With your indulgence, if you wouldn't mind staying on, we're going to go through our speakers and then open it up for questions and comments. So our next speaker is Melody Zhang and she is an associate assistant professor of immunology in the Drickier Institute of Children's Health and the Department of Pediatrics at the Well Cornell Medical College. Her laboratory focuses on the crosstalk between immune cells and gut bacteria that underlie immune regulation in the brain, placenta and lung during early development. Hi, I'm just reading for the slide. Good morning, everyone. First, thank you to the organizers for the opportunity to introduce our research here today. It's been a wonderful enlightening workshop, especially for me as an immunologist. So learn about all the exciting work going on on the gut brain access. See if I can advance. All right. So the research in my labs really focus on elucidating the role of the gut microbiome in facilitating the immune system developments in early life. And within this context, the maternal gut microbiome is a really critical factor. We know the mom's gut bacteria can shape the composition of the nutrients in breast milk. And at the same time, some bacterial metabolites from the mom can be transferred to the fetus in neutral during early developments and that can impact immune imprinting in neutral. Right after birth, the mom's gut bacteria can be transferred vertically to the newborn, and that really sees the early colonizers of the bacterial microbes in the newborn's intestine that can have a very long lasting impact on the immune system developments in the newborn, as well as neurodevelopment. And that collectively can have a huge influence on the susceptibility to pediatric diseases, including asthma, obesity, as well as some neurodevelopmental disorders such as autism, which was described by some of the talks from yesterday. In the gut environment, the gut commensal bacteria are actually compartmentalized within the gut lumen. This is really under the control by multiple barriers to really keep the gut bacteria in check. We have a variety of immune cells in the laminopropia that are really very critical to keep the gut bacteria in check. In particular, IGA antibodies to gut bacteria has been well studied. We know IGA can transitose to the lumen of the intestine where it can target the invading bacteria and really prevent the invading bacteria from crossing the gut epidural barrier. At the time when I was a postdoc at the University of Michigan, it was still relatively unknown whether under homeostatic conditions if there would be systemic IgG response to the gut bacteria. And what my work demonstrated was that under homeostatic conditions there might be some small number of grand passive bacteria that somehow are able to bypass the gut barrier and able to induce systemic IgG response. So we do see circulating IgG in naïve mice and humans. And the IgG antibodies can respond very, very quickly when there's translocation of some gut bacteria into the bloodstream, as well as bacterial pathogens such as seminera, which shares some of the IgG conserved antigens with our gut commensal bacteria. And this might be a very important mechanism to really facilitating very quick removal of commensal bacteria in the bloodstream or pathogens during the acute phase of IgG response. And this may be also an important mechanism to really maintain the symbiosis between the gut bacteria and the host. So when I started my lab in 2019, we generated an IgG deficient mouse model in our lab. And to really further explore the role of IgG or IgG gut microbiome interplay in a variety of diseases including inflammatory diseases and infection. So what we found was that when there's a lack of commensal specific IgG in the mom, now we see alterations in the gut microbiome in the neonatal mice in both the colon and the small intestine. And this appears to drive the IL-17 cells in the intestines of the IgG Nugout Nu1 just 14 days after birth. We know this was driven by the alterations of the gut bacteria in the IgG Nugout Nu1 because now when we re-derive the IgG Nugout germ-free, we no longer see the increased IL-17 cells in the intestines of the IgG Nugout neonatal mice. And to take it one step further, we asked whether the gut bacteria-specific IgG may play a role in regulating the maternal immune response to the fetus during in-utero development. So here we're looking at the presenter at E16.5 during the third trimester pregnancy. We're actually seeing increased numbers of the IL-17 producing cells in the IgG Nugout presenters. Particularly the cells were gamma delta T cells compared to the cells from the wild type presenter. And to really ask what's really driving the IL-17 cells in the IgG Nugout presenter because now we're really looking at a site that's distant from the intestine. So to figure that out, we took a step back to really understand what's going on in the intestine during pregnancy. So we see that pregnancy can actually induce increased gut barrier defects and increasing the gut leakiness. And that's really associated with drastic changes in the gut microbiome in pregnant mice during E1005, so just the middle of the pregnancy. Basically, interestingly, we saw an over bloom of carob bacturum rodentium in the wild type mice. So normally in non-pregnant mice, this bacteria are very, very rare, low abundant. But you can see it blooms pretty drastically during pregnancy and then it comes back to the pregnancy level when we measure at just P4, four days after pregnancy. And interestingly, when we took the amniotic fluid from IgG Nugout mice and we were actually able to detect some carob bacturum rodentium in the IgG Nugout amniotic fluid. But that's not the case when we measure the amniotic fluid from wild type mice or germ free wild type mice. So this really suggests that maybe the homeostatic IgG that are specific for gut bacteria may be critical to really restrict the translocation of carob bacturum rodentium to the presenter. And when you are lacking the IgG now, that may allow more carob bacturum rodentium to either translocate or somehow the DNA of that bacterium somehow made it there. And that could be a driver of the increased IL-17 cells that we saw in the IgG Nugout presenter. We were interested in IL-17 or maternal IL-17 during pregnancy and there's actually a lot of attention now on maternal IL-17 during pregnancy in part due to some seminal studies from the past few years, including this one that's shown here from Drunher and Grace Joyce lab. So here they did a mouse model of maternal immune activation using parli-IC injection to make make viral infection in pregnant mice. They showed that that could actually lead to increased TH-17 cells in the pregnant mice and the increased maternal IL-17 signaling to the fetal brain can actually trigger deficits in neural development. And as a result, they saw behavioral deficits in the offspring. So that's really interesting and the study really underlined the critical role for maternal IL-17 during pregnancy and some possible link to impaired neural development in the fetus in utero. So we do have evidence of increased IL-17 signaling to the fetal brain of IgG Nugout fetus. So we see an increased IL-17 receptor expression in the IgG Nugout fetus, the fetal brains as well as an increase in IL-6. IL-6 is a no key cytokine to induce the differentiation of TH-17 cells. So far we do have evidence suggesting there might be increased IL-17 signaling to the IgG fetal brain and possibly coming from the mom, but obviously we have a lot more work to do to really demonstrate that. So this is where we are with this project. We're really seeing a role for the bacteria-specific IgG in regulating the immune response during pregnancy at the maternal fetal interface. And lacking that IgG may allow some commensal bacteria somehow to translocate to the placenta and possibly driving IL-17 signaling in the mom and that may have some impact on the fetal brain. Obviously I'm not a neuroscientist, so in order to explore more about the impact of that on the IgG Nugout fetal brain, luckily recently I just hired a neuroscientist postdoc. Hopefully he'll take on the project in exciting directions, so we'll see how the project's taking us in the next few months. We already know the newborn's gut environment is actually very, very vastly different than that in adults. So right after birth, the nutrients actually come mainly from breast milk, and at the same time there's relatively high oxygen content. And so we know the gut bacterial species are different in babies compared to that in adults, but so far there's not a lot to know about the neurotransmitters in babies intestine. So this is where we were when my postdoc, Catherine Senator, joined my lab. So here we profiled the metabolites in the small intestine of P14 neonatal mice compared to that in adult mice. As you can see, a lot of the, sorry, the metabolites were differentially abundance in the neonatal intestine. And interestingly, we found that the metabolites that were highly elevated in the neonatal intestine were mostly neurotransmitters including acetylcholine, serotonin, and that's a little surprising to us because I thought the neonatal gut environment might be relatively less developed. And somehow they were able to make a lot more neurotransmitters. And I think having the neurotransmitters at such high levels during that developmental period might be critical. There's obviously a reason for that. So we wanted to see how what's really facilitating the high amounts of serotonin in the neonatal intestine. And just very, very simple, simplistically. This is derived from dietary tryptophan through TPH1, and then it can be further broken down into 5-H-I-A-A to monoamine a MOA. And so here, if you compare the expression of MOA in the small intestine of the neonatal mice, in the adult mice we saw no difference between germ-free and non-germ-free mice. However, now when we only compare that in the neonates, we see that there's high levels of MOA expression in the germ-free small intestine neonatal mice, but very, very reduced levels in the normal neonatal mice with gut microbiome. So just think that gut bacteria may be inhibiting the expression of MOA to break down serotonin. And we see that at the protein level as well. So this might be one mechanism by which the gut bacteria in the neonatal intestine help facilitate higher abundance of serotonin. As I mentioned, TPH1 is very important to convert tryptophan to serotonin, and it's mainly done by entercomaphan cells. So the TPH1 from entercomaphan cells will facilitate that process, and that's really what we know in the adult. So here, if you remove TPH1 from entercomaphan cells using the variant CRE, we do see a reduction in serotonin in the adult small intestine. And that's really consistent with the prior study by Dr. Leigh Shou's lab. However, now when we remove TPH1 in the entercomaphan cells in the neonatal intestine, not only that, we're not really seeing a reduction in serotonin. It's actually slightly higher when we remove the TPH1 in the entercomaphan cells in the neonatal intestine. So this really demonstrates that there's a very distinct mechanism to facilitate serotonin biosynthesis in the neonatal gut, which is really different than what we know about that process in the adult intestine. And we actually found that there are some specific bacteria that might be unique in the neonatal intestine that directly makes serotonin. So here we isolated a huge library of bacteria that isolates from the neonatal intestine, and then by ELISA we were able to demonstrate that they directly make serotonin. We found the same thing in human bacteria isolates from infants as well. As an immunologist, I am most interested in the immune response, and serotonin is most well described for its impact on neurons. So as an immunologist, I wanted to see now what's the reason for having such high levels of serotonin in the neonatal brain. So we know during early developments being able to develop immune tolerance during that critical time developmental period is really important. So you don't want the babies to develop immune reactions to dietary antigens when they start eating solid foods. Or you don't want them to develop immune reactions to commensal bacteria when they're just having the gut microbiome developing. And so during that time when they're able to develop immune tolerance, that will help really facilitate the developments of the gut microbiome and help them to really up your nutrients that are critical for that development as well. So here to see if serotonin may actually impact the immune cells. We did a very simple essay, we did a Seahawks essay, adding 5-HT serotonin to the T cells. We were able to see a reduction in the oxygen consumption of the T cells, suggesting maybe there's some changes in the T cell metabolic state. And then we further by using the M-Tor activation marker, PRPS6, we showed that adding serotonin to the T cells can actually reduce M-Tor activation. We further really understand what's going on. So here now we treated the T cells with serotonin and then profiled into cellular metabolites within the T cells. And we saw a variety of metabolites that were elevated in the serotonin treated T cells. In particular, we saw endoesteroidaldehyde, I3A, that in particular were not really high in the serotonin treated T cells. And to see if that may be mediating the changes in the M-Tor activation. Now we only treated the T cells with I3A and we were able to see an inhibition of M-Tor1 activation. And now to come back to see the T cell response. So M-Tor activation has been pretty well studied now in T cells. So in general, it promotes the differentiation of pro-inflammatory TH1, TH2, and TH17 cells. At the same time, it restricts the differentiation of regulatory T cells and regulatory T cells are the cells that can suppress information. So to see whether now serotonin can actually change the T cell response. We found that when you add serotonin to the T cells ex vivo, now after 48 hours, now you see increased abundance of T-Rex. At the same time, a reduction in interferon gamma producing or L17 producing T cells. Suggesting serotonin can actually promote the T-Rex while suppressing the differentiation of other T cells subsets. And to see that may have some impact on immune tolerance during early life in vivo. So here with the NSA, we already go watch the germ free neonates with serotonin just twice. And then we sensitize them to over antigen, which here would represent a dietary antigen. And then later on, when they became adults, we challenged them with the same dietary antigens. We found that when the mice were treated with serotonin already, when they were just neonate. But later on when they were exposed to the same antigens in adults, as in adults, we actually saw that the serotonin exposed new mice had lower over specific IgG and IgE. So it really suggests that there might be some critical role for serotonin in the neonatal gut during early developments to really shape the T cell response such that it will suppress the T cell activation when they encounter dietary antigens. So later on in life, when they encounter the same dietary antigens, they're not really developing T cell reaction to the dietary antigens. And to really understand whether that's really facilitated by the T-Rex that we know are generated by the serotonin exposed T cells. So now we did a similar experiment. We treated the neonatal mice with serotonin very earlier at day A and day nine when they were neonates and then sensitized them with the over antigen. And now we isolated the T-Rex from the intestines of those mice and adaptively transferred into a new set of mice. And now we challenged the new mice with the over antigen. And we here demonstrated that the adaptive transfer of the T-Rex from the serotonin treated mice was able to reduce the immune reaction to the over antigen in the neon mice. So really suggesting the T-Rex as a mediators of serotonin mediated immune tolerance in early life. And here we wanted to also see maybe serotonin is doing something to the gut commensal bacteria. And here we treated the germ free mice with serotonin and then we transplanted the gut commensal bacteria into the mice after like a week. And later on after two weeks we saw increase in T-Rex in the colon of the intestine and at the same time reduce T17 cells in the intestine. And we also saw drastic changes in the mice treated with serotonin or PBS. So just think serotonin either directly signaling to the gut bacteria and changing the gut commensal composition or by changing the T cell response they change the gut bacterial community. So what I just shown today is really showing the the immune migratory effects of serotonin. Specifically during that early developmental period we found that somehow babies have more serotonin in the intestine and and besides possibly affecting the entire nervous system the serotonin that can directly signal to the T cells in in a way to facilitate regulatory T cell developments in the in the gut and that may be critical for developing immune tolerance to dietary antigens and commensal bacteria in the in the babies. And so that's what we have so far. Thank you so much for your attention and I'll be happy to take questions later on. It's a real pleasure to have the opportunity to introduce Dr. Michael Fischbach, who is associate professor of bioengineering at Stanford University and a Chan Zuckerberg biohub investigator. Professor Fischbach is a chemist, microbiologist and geneticist. He pursued his undergraduate education at Harvard College and subsequently a PhD in chemistry and chemical biology from Harvard University in the laboratory of Chris Walsh. Professor Fischbach's laboratory focuses on discovering and characterizing small molecules from microorganisms with an emphasis on the human microbiome. In addition, he develops computational tools that identify small molecule producing genes in the bacterial genome. In addition to his ongoing work on the genetics and ecology of complex microbial communities. And with that introduction, I will turn the podium over to Professor Fischbach. Thank you so much for the introduction. Let me just confirm that you can that you can hear me. And then I'm also going to confirm that I think it's Jessica has been communicating with me on AV and as a gem and I want to make sure that when I click the slide advances and it does that's amazing. Thank you for that. Okay, so I'm really sorry I couldn't be there with you in person but I'm delighted to be speaking here. Many things for having me and I wanted to, I want to jump right in on a story that is work from Kazuki is a very talented scientist in the lab. I won't belabor this because I want to, I want to show you this technology that we've been working on that I think could be helpful to some people in picking things apart mechanistically. As they relate to gut brain circuits and other aspects of physiology that are impacted by the microbiome. So let me queue it up by explaining the interest that we had and the problem we encountered. So we were very interested in studying a variety of ways that the microbiome impacts the host here I'm going to focus on immunology. So the studies that got us the most interested were ones with the format shown on the left where a fecal sample often from a human would be transplanted into a germ free mouse. And then some interesting phenotype would come along for the ride is a really interesting experiments in demonstrating that the microbiome is involved somehow physiologically and a phenotype of interest but they're, they're frustrating in the sense that it's difficult to figure out how it worked, which bugs were responsible and by what mechanism. There are other immunologic experiments that that we're also interested in where where folks would transplant a single organism or a small community, and then see some impact on on immune modulation. These are interesting as well we've done experiments like this but then the problem we've always run into is that the same bug in the context of a complex community, sort of a physiologic complexity doesn't do the same thing. And so we think that the data are accurate in some sense but they're not reporting on something that's going to teach us much about the real biology and so we envisioned a different format in which we could transplant a community that had the advantages of being defined and we know exactly who's in there, but also complex enough to capture the salient biology of the microbiome or enough of it anyway that we would learn something that could stand the test of time about what one organism or another contributes to to a phenotype like immune modulation so in, in thinking about doing this we had the good fortune of work that another member of the lab Alice Chang was was doing she's a physician scientist who had built a complex community that is meant to be a model system for the human microbiome and it done a lot of work to show that that it engraved stably in mice. This is a community at this point of 119 organisms, and I'm not going to go into detail on it because I want to show you about the immune modulation stuff but suffice it to say that when we colonize mice, it colonizes very reproducibly and stably and, and that the organisms in the community distribute themselves across roughly six orders of magnitude of relative abundance, much as we think you would find in a typical human gut. So, so we think this is a reasonable model system to move forward with even as we improve it in the background and hopefully make better versions of it. So here's the kind of experiment that was really the reason for the for wanting to do the work of putting together a model system like that this is a good example of the kind of thing that I think is possible to do now that wouldn't have been easy to do beforehand where Kazuki propagated the community as individual strains, mixed them together, colonized a germ free mouse, waited two weeks, and then isolated T cells from the gut. Then at this point, he could take this pool of T cells from the gut and incubate them one at a time with every bug in the community to figure out what each individual organism is contributing to the overall phenotype of immune modulation by the gut microbiome. And so the data from that experiment look like this I'm showing you a phylogenetic tree of the organisms in our community on the left side of the screen this is the top third of that phylogenetic tree. And then the single dot that the that this dotted line is coming off of indicates that this organism in testinobacter Bartlett EI, restimulates roughly 10% of the T regs in the gut, and, and so then the rest of the dots correspond to what each individual organism is contributing to the pool of T 17 cells and T regs in these T follicular helper cells that we call FR four positive T H cells. And so this is sort of a a plot that shows you what each organism is contributing this is the most of the termicutes if I click forward you'll see the contribution that the remaining firm acutes in the actinobacteria in the community have, and then the last click forward is is the Bactria deadies. And so we had initially intended in this story to find potent inducers of T reg in T H 17 cells that didn't just work when they were at home but they worked in the context of a complex community so that we were going to characterize them in detail that was going to be like the second half of our story, but instead we got captivated by a different observation which I think was one of the joys of working with the system like this. The unexpected result was that if you add up the total percentages of the of the entire pool of T cells that are re stimulated by a number of these organisms, you realize very quickly that it adds up to much more than 100%. And so an old model that we had had in our mind from pioneering work that people like yes mean bell Cade and Dan Lippman had done was that each individual organism was going to elicit its own pool of T cells that were entirely specific for that organism and the host and so this one to one relationship is what we had expected to see in this data. We realized that that couldn't be possible in this experiment because the total number would have added up to much more than 100%. And so we wanted to go deeper in terms of resolution. And then I'll show you what we did and I think this also highlights sort of the capabilities of a defined system like this. So, clicking forward, or attempting to click forward. There we go. Yeah, I think that that's displayed visually is that the old the old model we had had in our mind is that the blue strain induces a pool of T cells on the host side that are specific for the blue bug. So to the dark red strain induces a pool of T cells on the host side that are specific for it but we began adding up the percentages of the pool represented by the cells on the right and realize it added up to much more than 100%. So what's going on here. So Kazuki did an experiment that starts out the same way but ends ends up in sort of a bit of a more interesting place where he he propagates the individual strains in the community mixes them together, waits for two weeks and isolates T cells just as before. But now he does a single cell RNA sequencing experiment combined with TCR sequencing. So here we have for about 10,000 cells a tiny slice off the top of the T cell repertoire. Single cell RNA sequencing data and the sequence of the T cell receptor and through a process that I'd be happy to answer questions about but I but I'm going to I'm going to move through quickly here. He picked a subset of these T's not all of them are going to be specific for the microbiome he picked a subset that had that had expanded recently we could see that there are multiple cells in the pool that express the same T cell receptor indicating that they had divided recently and he hypothesized that those were going to be enriched in cells that were specific for one of the T cells. So we had one or another bug in the community and then took 92 such cells and we just had twist synthesized the the whole sequence of the T cell receptor and then Kazuki used a retrovirus to integrate the that that artificial gene including the T cell receptor into a stable cell line so then we had 92 different cell lines each making one TCR that we were crossing our fingers and hoping at least some of them would be specific for a bug in the community. So every one of those 92 cell lines by the hundred or so bugs in our community that that matrix was a big Eliza experiment and then that enabled us to draw the map for this tiny little portion of the T cell repertoire of exactly which cells were re stimulated by which bugs and so the portion of that keep map that has data is shown here. Not every bug in our community re stimulated one of those cell lines. Not every cell line was re stimulated by one of those bugs. But to make a long story short, there were these little chunks of the heat map which represent sets of T cell receptors. These are each each one of these rows is a cell line that expresses a single T cell receptor from that mouse and then each one of these columns is one of the bugs in our community and you can see that there's a whole set of of TCR clone types that are being re stimulated by a bunch of the gram positive bugs in our community and so because he went back in with another kind of high resolution experiment that would have been difficult to do. I should say graphically just to kind of close the loop here out what we what we began to think from this is that there's a pool of T cells in the host that are that seem to be re stimulated by a bunch of different bugs at the same time so because you wanted to figure out what's going on there. And to do that he made genomic DNA libraries from three of those gram positive organisms from that feature in the heat map to figure out like what gene in their genome. What was the was being recognized what was the epitope and so he made shock and genomic libraries from each of those strains formatted them in a big kind of pooled library experiment and then mix them together with that pool of 13 stable cell lines that were all re stimulated by the gram positive organisms in the community so that rather large matrix became a screen the screen delightfully had only three hits and those hits overlapped around a single region of the genome that is conserved, not only in these three organisms that he had made shock genomic libraries from but all of the gram positive organisms that that that were stimulatory and that feature in the upper left hand corner of our heat map. What is this thing. It's a system that sits on the outside of some gram positive organisms it's widely conserved in our community where they're the epitopic or the epitope that we eventually identified is in this lipo protein called a substrate binding protein that's very boring it just sits on the outside of the cell and grabs a sugar and dumps it down in ABC transporter. We now know that this protein, we think that one of the reasons it's, it's recognized by the by T cells in the repertoire is that it's widely conserved and also very highly expressed, both when you grow one of the organisms from our community culture, and if you tile reads from human metatranscriptomic data from from human subjects in a study. This is one of the most highly expressed genes in the genome of this organism, and now others that we've looked at in vivo in the human guts so we think that that's, that's probably a lot of the reason why the host is recognizing it as a T cell epitope and we've even been able to narrow down to the specific conserved peptide at the C terminus of this protein that seems to be recognized. So all of that is to say that I think the use of a defined community gives us some added resolution on the microbial side and makes it possible to go well beyond where we've been able to go when we use undefined communities in experiments where we colonize germ free mice and so we're excited to hear ideas from other people and what ways in which this could be a resource that would help other people make discoveries where we would are looking for ways to propagate to promulgate I should say this technology and send our defined communities out into the world. Thank you for listening. I'd be delighted to take questions an hour later. Thank you so much. Dr fish box. Our next speaker will be and I, let me just interrupt and say we are running way over so if you'll indulge us will go past the end period of this with questions. But our next speaker is a Diego for her class, who is an associate professor of medicine and a research track associate professor of neurobiology at Duke University, and he studies the gut brain sensory transduction. What what is sort of a new fields called sensory biology. Please, Dr. Dr Diego. Thank you very much. Can you hear me. Can you hear me now. Perfect. Good morning everyone. So I don't think that I can control the slides yet. Are the slides moving. Perfect. It's about time for a launch. And you will not let me lie that next time that you go to the grocery store you realize that how about 95% of our decisions are actually visceral nature in specifically when it comes to food. And by the time that we get to the parking lot we have already eaten half of it and we don't even realize what it has. How it is that we came to that decision. But if you ask any behavioral economist, they actually know that there are hidden forces that shape our decisions. In fact, Dana really who has in this very influential book predictably rational shows that at least behaviorally that about 9 in 10 of our decisions are influenced by the visceral but the neural basis of it has large large largely being unknown. So as early as the 1900s. We already already had a very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very. We had a very good understanding of the chemistry of how is he that we break down food into small molecules that we can digest up and absorb. So we take that piece of chocolate. And we break it down into pieces when it goes down through the sofa was into the stomach. It gets bait baited by hydrochloric acid pepsin raining, And then the gatekeeper, Mr. Pilaros, will let the time flow into the intestine where truly the magic happens. The intestine is where these basic building blocks of life will be not only further digested, but also absorbed and we'll go into circulation. And from that is that new cells are replicated. So that's how we acquire our proteins, fats, and our sugars. So even though we have known quite a bit about the chemistry of the digestion for a long time, how is it that the intestine recognizes those nutrients that we just ate to be able to guide our repetitive decisions? It has been largely unknown. In fact, here is one of my favorite papers. It's a classic paper from Professor Gibbs, who was at Columbia, I believe, at that time in 1973. And the first sentence is a classic sentence. It says that within 10 minutes of starting to eat, a rat stops eating grooms for a short period of time, then usually sleeps. We define this behavioral sequence as a tidy and its physiological basis is unknown. And this is the first time that they go on to propose that the gut, after sensing nutrients, it is capable of releasing these endocrine factors, like colcistokinin, that after several minutes, 5, 10, in fact, after 30 minutes, peak in the bloodstream. And those signaling molecules are partially responsible for inducing satiety in the animal. So they signal fullness from the intestine to the brain, and then the animal feels full. But if the hormone peaks after 30 minutes, here very clearly said that within 10 minutes of starting to eat, the rat has already not only brushed her teeth, put on PJs and get inside of her bed, so what happens in those first 10 minutes is really, it was really a black box for a long time. And here I'm going to, and the idea is that the nutrients will get into the intestine and then will stimulate the cells that are known as endocrine cells or cells that release hormones. And those hormones will either go into the bloodstream or will activate some nerve interference by diffusion. That ultimately will carry the signals. But in recent years, we have had access to new technology that has really helped us to visualize what is it that is happening well before the signals reach the threshold of consciousness and we are able to actually articulate that we feel full. So this is a beautiful example that comes out of a technology that was optimized, developed and optimized by Professor Teresa Lieber and her laboratory at the University of Missouri. She's a collaborator as, and when I saw this video, I was fascinated by the speed at which nutrients actually get to the intestine. So what you're seeing here is a mouse that has been sated and is about to eat a semi-liquid diet on the right. And it is being imaged with this X-ray microsystem. So the mouse goes to the right and it starts to eat. And then quickly, within a couple of seconds, you can see the food already arrives in the stomach. And within a couple of seconds more, it starts to diffuse in intestine and really expel into the intestine. And if you see in high contrast, this video, you can see that it goes throughout the entire intestine. So within only a few seconds, the food has already arrived into the stomach. The food has already arrived into the intestine. So how is either those signals get to the brain and what is the consequence of those signals reaching the brain? So one experiment that perhaps my laboratory, it's well known for, and it was really a catalyzing moment in this understanding of sensory transduction or the ability of nutrients to be converted into an electrical pulse that ultimately will guide repetitive decisions. It was this experiment that I often call as two brains in a dish. On the right-hand side, you're gonna see the sensory neuron from the brain. And on the left-hand side, you're gonna see this green cell from the gut. And when they're put together in culture, you will see what happens after a few hours in culture. So again, on the left, you have the gut and on the right, you have the brain. So not only is fascinating that a cranial neuron is able to regenerate an axon after being culture, but also that these cells, these two cells are able to find each other and the nutrients in cell in the gut is able to connect directly. So this is not by the diffusion of hormones, but actually a physical connection with that sensory neuron as a consequence. In 2018, we coined a new term and we call these cells neuropot cells because they have the ability that in space and time to transduce signals from the luminoid intestine into the nervous system that ultimately guide behavior. So here we have neuropot cell, our gut sensory petylacel, that's synapses with nerves to transduce stimuli in milliseconds. So this is the first part of the talk and I wanted to summarize with a series of experiments that ultimately have helped us to document the neural basis of sugar preference that goes all the way from the specific receptor in the gastrointestinal tract to the cells, the neurotransmitters and the circuitry and these video summarizes that work that my laboratory has done over the last 12 years. The well-oiled sugar cravings come from scientists have long known that our preference for sugar does not depend on its sweet taste alone. Instead, our preference for sugar depends on the sensation arising from the gut. But how the gut guides our preference for sugar through the artificial sweeteners was obscure until now. When sugar molecules enter the intestine, they are recognized by sensory cells. One of these is the neuropot cell. Neuropod cells synapse with the vagus nerve to tell the brain about sugars entering the intestine in milliseconds. We wondered if these cells are necessary to discern nutritive sugars from non-caloric sweeteners. We found that neuropod cells sense sweeteners using the sweet taste receptor T1R3. This causes the release of the neurotransmitter ATP, but sugars are sensed differently. Neuropod cells sense sugars using SGLT1, an electrogenic sodium glucose transporter. When glucose enters through SGLT1, neuropod cells release the neurotransmitter glutamate. As such, neuropod cells convey signals from sweeteners using ATP and from sugars using glutamate. But does the animal preference for sugars depend on neuropod cells? To answer this question, we used optogenetics. This technique allows researchers to silence or excite neuropod cells while a mouse is presented with a choice of sugars versus sweeteners. We discovered that when neuropods are silenced, the mouse cannot distinguish the sugar from the sweetener. It becomes blind to sugar. A mouse presented with a bottle containing sweetener consumes only small amounts of the liquid, but exciting their neuropod cells causes the mouse to double its intake of the sweetener. It drinks the sweetener as if it were sugar. By sensing nutrients, neuropod cells convey rapid, subliminal sensations to convert food into feelings. So what is fascinating is that the gut has the wisdom to sort very rapidly chemical properties, and for that matter, physical properties of a specific chemical molecule, in this case, a nutrient, and then being able to inform in real time the brain, so behavior can be adjusted. And as we know, the gut in a human is a large organ. In fact, for that matter, in terms of proportion, it's a very large organ in all organisms. In the human, in an average human, adult human, the gut is about eight meters long. So here what I show you is what happens in the proximalismal intestine. But what about in other segments of the intestine? And here's where several years ago, we began working on this project to be able to not only track the circuitry well, also see what happens in different regions of the gastrointestinal tract. And this project began with tracking the circuitry of the gut using a monosynaptic rabies virus. So we discovered that in one type animal, these neuro-pot cells, they actually get infected by rabies. As you can see here, this is one of the first images of our data. And then if we enable that rabies to be spread by Crelop-Luxperia recombination, then the rabies is spread onto a neurons, including a vagal notice neurons. And depending on where those neurons are projecting, obviously the function of those circuitry are different. So one of the, so this work had two very clear implications. One was that the gut, like the nose, like the tongue, like the skin has physical circuitry to be able to transduce the essence. So two neurons are an epithelial cell, and a sensor neuron connect to each other to bring that sensory information from the luminovine testing into the brain. And in this case, some of the terminals from the vagal notice neurons will end in the nucleus tract of solitaire in the brain. And the second implication is that as Illa Meshnikov in 1908 put it, death begins in the colon that through the circuitry, if there are some pathogens that could be able to access the circuitry, they will have access to the brain. And as there's quite a bit of interest today in neurodegenerative diseases, these actually provides a path for pathogens to be able to access the central nervous system. But what about the interaction with microbes? So we became interested in this topic, building on the pioneering work of Sarkis, Mauro, Elaine, and others. And began wondering whether or not the circuitry in different types or in different parts of the testing will actually convey very rapidly sensory information to the brain to regulate behavior. So as we know, microbes can regulate behavior and this has been done by doing these microbial transplants and also in germ free mice. So here we have this role of the entire intestine and colon on the left and in a transgenic mouse in which the promoter for the hormone peptide, the neuropeptide peptide, YY drives the expression of GFP in this way, we can label these cells. And one of the things that we quickly realized is that these cells in the colon specifically, they are highly enriched in this toy like receptor five. And we realized this like over six years ago. So one of the first experiments that we run is what would happen if we knock out these receptors specifically in these cells on the colon. And one of the things that we found is that these animals not only become overweight over time, but also like their meal pattern, including the meal duration and the meal size is altered. And what is fascinating is that they actually do not develop metabolic syndrome. They do not show like an inflammatory response, but rather they just simply eat more. And working through the pathway, always let me just share in here a very short summary. These neuro pod cells in the colon, they are capable of recognizing flat gelling. And when flat gelling is delivered into the lumen of the colon, it rapidly activates vagal notice neurons that inner rate the colon. So when we use optogenetics in this case, halo-rhodopsin using halo-rhodopsin and we provide with the right type of the light and we silence these neuro pod cells, then we get rid of that vagal activation, demonstrating that these neuro pod cells in the colon are necessary for the excitation of the vagus nerve and for the transmission of the flat gelling stimulus onto the vagus nerve very rapidly. And you can see here that also, as I mentioned, we put the flat gelling directly into the lumen of the colon, we get a very rapid vagal activation, as you see in that green bar, but in the knockout, in the PyTLR5 knockout mouse, that activation does not occur. And as you can see, that the food intake if we deliver directly the flat gelling directly into the lumen of the colon, over 20, 40 and 60 minutes actually reduces the amount of food intake in the mouse, but in the PyY3TLR5 knockout mouse, that effect does not happen, showing that the circuitry from gut to brain from the lumen of the colon actually is capable of recognizing the stimulus flat gelling and through the activation of TLR5 on these PyY neuro pod cells activates gut brain and neural circuit that ultimately modulates food intake. So finally, I just want to share with you that next year, Astronauts is gonna be organizing our third global symposium. It will be in Galápagos, the famous islands on June 1st to the third of 2023. So if you are interested, please take a look at it. And then obviously the music comes from the team. I'm here to represent them. And I wanna thank them for the work that they do. And thank you all for your attention and thank you, Kanya, for the invitation. Fantastic presentations this morning. And the last speaker for this morning is Vanessa Rudura. I hope that's the correct pronunciation. Who's a senior program officer at the Gates Foundation and leading the microbiome products thrust of all that the Bill and Melinda Gates Foundation is involved in. Dr. Rudura received her undergraduate degree at the University, Simon Bolivar and pursued her doctoral studies at Washington University in St. Louis in molecular genetics in the laboratory of Professor Jeff Gordon. She subsequently pursued a postdoctoral fellowship with Yasmeen Belcade at the NIH. And prior to her position at the Gates Foundation led project baseline at Verily Life Sciences. Dr. Rudura, the podium is yours. Thank you so much. In Spanish we say Ridaura, but I know it's like always really confusing. Thank you everyone for being here and thank you. I know we're running over time for taking the time to listen to what we're trying to do at the foundation. And I think I can move the slides. Yeah, I wanted to start with a quick overview of where our team sits and that will provide a little bit of context for the type of things that we're funding and the interests that we have. So the Bill and Melinda Gates Foundation, BMGF program strategies are divided across six divisions and our team, the maternal newborn child health discovery and tools team, I know it's a mouthful, falls under the gender equality team. This is really important for us because of the type of investments and proposals that we're trying to fund. Specifically our team here in red is responsible for funding drugs, food and microbes, devices and software, epidemiology insights and discovery insights all aimed at improving the health of mother and children. And the reason that I wanted to bring up a little bit of the GE division and the gender equality division is because we just recently moved to it and I think it's going to underlie a lot of like the way that we're thinking of funding for the topics at hand. And the gender equality division is to achieve gender equality and empower all women and girls, which is a sustainable development goal for UNICEF and the UN. And the division goals specifically is to accelerate that trajectory in Sub-Saharan, African and South Asian countries. And you'll see that the way that we're trying to do that is by really focusing and removing barriers to gender equality by not only what we fund but who we fund. And I wanted to bring that up in these four different sub-goals. The first one will be to understand that when women and children have good health and autonomy over their bodies, they thrive, other areas of their life thrives and everyone associated with them thrives as well. Women's economic opportunity and decision-making power really grows in the household income and keeps more children in school. So if we help women, we will help their families as well. And something that I hold near and dear to my heart when thinking about investments and how we do them is that changing the people in the decision-making room with more women in leadership roles will really continue the cycle of empowerment. And I think about this a lot when considering investment and our grantees and our partners. Finally, we really are thriving to strengthen positive social norms about women and girls, fostering an ecosystem that will support gender equality by using the strategic goal of accelerating these declines in neonatal and maternal mortality while supporting growth and resilience. And you'll see that exemplified by what we're trying to do with the gut microbiome portfolio. The strategic principles for our team is that we're gonna be addressing underlying biological risk rather than focusing on a specific indication or single syndrome, which is a little bit different to other teams at the foundation. We try to intervene as early as possible in the life course with a prevention approach and then also fund treatment-based interventions that can help reduce the morbidity associated with longer life in some of these scenarios. We also want to improve on the ground data and then push the global guidelines that will impact all of the population at hand. And finally, what we try and do is work through local. So from the optimal to the possible, and that's what we're focusing in with this portfolio, it's really optimizing these learnings to accelerate translation. And the goal that we wanna have in the next four year period is to really have a body of work that will answer the questions about the role of the maternal microbiome during pregnancy and help us inform a target product profile, the population and the mechanisms of actions that we should be targeted with these interventions. And specifically, we wanna hone in into using these experimental learning capacity, going back and forth with preclinical models to design an intervention that it's going to impact the biggest population possible by improving not only the final biomarkers which will not be powered to see, the final outcomes which will not be powered to see, but transitional biomarkers that we will continue to refine and define in these studies. I wanted to end there and then just quote something that one of our co-founders Melinda said, is that the beauty of our fight for gender equality, which interestingly goes beyond just fighting for this rights but really focusing in scientific questions that will improve the life of mother and children, is that every human being will gain from it, not only women and children. The points and areas that we wanted to go next is really understanding microbiome and pregnancy by targeting immune adaptations through pregnancy, understanding the interplay between the vaginal and the gut microbiome during pregnancy and finding ways to study how the gut microbiome can impact placental development. And with that, I just wanna thank you and I hope that you guys have a great lunch. We're gonna try and spend maybe 10 to 15 minutes in question and answer. We know we've cut into your lunchtime so we thank you for allowing us to spend some extra time with this. Wanna remind you if you have questions, please put them in Slido and if you're on the Zoom, please raise your electronic hand and we will call on you. But maybe we can get started with one question if you don't mind me asking, which is what is really interesting about all of these talks, not only is this idea that the gut microbiome and the immune system and external sort of stimuli are all affecting not only each other, but how our behavior, our behavior, our sort of neurological state and development. And in this last talk, Vanessa talked about experimental learning models and thinking about translation of some of that basic research. And I'm just sort of interested in understanding how where along the spectrum are some of the research that we talked about and how can we think about better translating into actual sort of practice? Are there things to sort of help with that ongoing learning from basic research to its applications and then back again for anybody who would like to ask the question. Anyone like to weigh in? I guess maybe a complimentary way of framing this is strengthen weaknesses of our pre-clinical models that allow us to translate with predictable results. And I think that's been alluded to by a number of the speakers, sort of benched to bedside and back again. And so I'm wondering if a few could talk about some of these strengths and weaknesses both of using murine models with inherent murine microbiota, alternative models which were alluded to by Dr. Fischbach which is defined communities of bacteria but then the challenges of a high level of diversity from one patient population to another or one geography from another. And sort of where our pre-clinical models have succeeded and where they failed or is it just too difficult to tell at this very early stage of the career? Michael, maybe you could start off and then Melody, Vanessa, Diego could weigh in as well. Yeah, this is a really interesting set of questions and I don't know if I can offer a panoramic perspective. I think that as far as the defined communities go I think it's exactly what you would expect from something that's new. It's a bit young and unrefined. And so far we only have one community that we've published on and another one that we're building in-house. I think the bright side is that they're gonna be shared widely and without any constraints that everybody can do whatever they want with them and even beyond that we are delighted to build communities or people that are missing certain organisms or groups of organisms in case that would be helpful. So we want to send this out as far and as wide as possible. The constraint that you noted about that all we have is one community that represents some kind of average at one point in time is right on. I mean, I think that we'll learn as much as we can from that, just some basic fundamental things that would apply hopefully some of them to other communities. But then in cases where we're interested in a particular malady, we have begun reconstructing communities from patients who have a disease. And I think that that too is gonna be something that we can learn a lot from. In the early cases that we've been working on we can capture the phenotype that we're after by building a new community from scratch. It's a difficult process right now to take a couple of people, a couple of months to make one community. Hopefully the technology behind that will change and make it possible to do that faster. But right now it's still a serial one-off process so we'll have to choose carefully. Others should comment on like the quality of the animal models and where I have much less depth. Yeah, can I jump in? I would just add this. Can we take a step back? Because you go to pre-clinical, clinical ask a simple question physiologically. Where in the gut are the microbiota interacting with elements of the enteric nervous system? Because throughout the gut it's not homogeneous. The ENS is differentially innervated, I'm sure Diego could speak to this much more than I, throughout the gut. So where are the bacteria? So as you eat food, they actually exist on the food and they will grow as biofilms as the food moves through. So do we know and how they communicate with the villi remembering that in the colon we have two types of mucus flowing through, any surgeon will tell you that operates on the patients for chronic surgery. You have a static layer that coats basically the villi. You have a moving layer that goes through and moves down the intestinal tract. The whole purpose of that is to keep the bacteria away from the villi and the innervated portions of the villi that will then hook up to the ENS vagus brain. My question is asking, do we know where the communication pathways from a anatomical geographical version actually where they are? Because if you're going to design only strategies about a microbiota, whether it's a defined community, do you know that the bacteria will proliferate, will survive to that region, will interact and proliferate in the region, deliver what they need to do? I think these are basic questions that need to be answered. Just my two cents. That's an excellent point, Mark. And if I can, I'm sorry. I know that we had an order interjecting. We'll just say like both Michael and Mark, I think that both of your perspectives are complementary, right? There are techniques to isolate small intestinal bacteria to make sure that then you can build communities that will colonize differentially throughout the gut. And I believe Michael's complex communities, if you just have all the members, they will take over wherever they are most able to thrive. I think that one of the challenges that we do have, both in the cognitive world and in others, is that in some cases we don't even understand the etiology of the disease that we're studying. That is definitely the case with the ED and certainly for IBD. So then what ends up happening is that even if you have a microbiome from a disease individual, you don't have a model that really replicates the whole side of the disease. That doesn't mean that all models aren't perfect. It doesn't mean that you can't learn from them, right? But that's I think where the power of this erasure with humans comes from. Because if you don't understand the disease, you can't mimic the disease in the Muran model. Even if you have the right bacteria, you don't have the right host piece. And I think that that becomes really challenging but that's kind of my two cents. But you're right, both Michael and Mark, like all diseases, EED for example, is a disease of the small intestine. So if we try to build mouse models with colonic or fecal bacteria, we may be missing key players that are responsible for driving some aspects of the disease. Definitely, I have more experience working with pre-clinical mouse models. And I think one nice thing about using mouse models for studies of immune response during pregnancy is that it's just 21 days. So you can just look at first trimester, second trimester, third trimester every week. And you can look at a variety of things and easily manipulate the gut microbiome by putting in the fine sets of bacteria. So that's really nice for us. We're able to isolate the immune cells from the center and do a single cell. And as you can see, unfortunately, I didn't get to show that data today. And I do agree with Vanessa and Dr. Mark, like that it's very difficult to recapitulate the niche that we see in humans then in mice. For example, my lab, we actually transplant stool samples from babies into germ-free mice, trying to really rebuild that community that we see in babies in mice. The challenging part is we know a lot of the bacteria never survived the gut micro-environment. So what we end up seeing in the mice two weeks later is not really definitely not representative of what we see in babies. The micro-environments difference, the bacteria are not able to colonize well, stability. So we've lost a lot of information during that process. So that's a huge limitation with really translating what we see in humans with the clinical models. Diego, any thoughts? I think that in terms of priorities, I would be to give as a perspective, I think that we have made tremendous progress in understanding some of the colonies and their interaction and the responses of the host. I believe that there is a strong opportunity to understand how is it that the microbes actually guide our repetitive decisions? Because ultimately the building block of the entire system is food, and not only like the basic food that acts as a nutrient, but also like additives. And now we see very common in the food landscape, artificial sweeteners, flavors and others. A couple of months ago I attended a conference in which I saw some very interesting and intriguing data on how microbes just by a supplementation of fiber could supplement the basic protein needs or essential amino acid needs of the host. And I think that that brings an entire new dimension in how is it that we are gonna use the microbes to be able to supplement dietary needs of populations worldwide, because depending on where we are located, different microbes will be able to supplement those diet or affect those dietary needs of the population and outcomes. How to get to it? I think that certainly like the epithelial interactions are very important. And the other one that Mark actually highlighted is that it is very important to recognize that the gut is not a common gene use organ, but rather it is very diverse, long ecosystem and depending on where the location is, the functions are very different. And that's something that several laboratories are actually beginning to take into from the micro-sat, from the epithelial side, and also from the diversity of the entire unit, so on. That was in my perspective. I think it's a unifying theme, not unique to this particular field, but that we can certainly learn a tremendous amount from mouse models, but there are huge limitations and a big gap between creating truly humanized versions of mouse models, particularly in this area. And then the risk is the tremendous amount of resources that are being deployed in the hope that they're predictive only to find out that there are many different ways they fall short. I was wondering if I could follow up on two other themes that you alluded to, Diego, but all the other speakers spoke to as well as the opportunities and the challenges with probiotics. And I'm just wondering if a number of you would weigh in on some of the challenges that you see in translating probiotics, whether it's scalability, reproducibility, QA, QC, cost. Where do you see it succeeding and where are the dirty little secrets that may hamper the movement of this field forward? Vanessa, you look like you're smiling. You want to weigh in? Sure, my two cents as well. I think that the field of probiotic really advanced things to food-derived bacteria that have been leading the way into how we manufacture and produce these taxa. The challenge there in developing those EMCs is that food-derived bacteria, like a lactobacilli strain that you find in yogurt, are usually more stable, easier to manufacture and scale up. So I think that when we're thinking about that next generation probiotics or live biotherapeutics, the challenge will be really the culturing piece, thinking beyond a single bacteria similar to what Michael is doing and thinking more about communities. How do you culture these communities where each of them and Michael can talk about this, of course, way better than I can, but how do you culture them with where they have each of them specialized needs? The other piece of information that will be important is that specifically for more vulnerable populations, the WHO has already given guidelines where they ask for these products to be manufactured in GMP or clinical GMP facilities. That is not what's happening with the majority of probiotics in the market right now. And this is really important because if you're giving a product to a baby, a vulnerable baby, a pregnant mother, or even a child, you want these products to be manufactured in clinical GMP or GMP facilities. And I think that those two are going to be really important changes and shifts while also considering keeping costs down because the moment that you add GMP for manufacturing, the moment that you start specializing, using specialized culturing techniques, costs start going up and then the product no longer becomes available for larger segments of the population. And let me build on what Vanessa said and kind of reoriented for better you about the microbiology here. Remember, you're growing these probiotics, you mentioned the dirty little secrets out here. So I'll give you one which is not really a dirty little secret, but think about it, you're growing these bugs up in a fermenter which is maybe 10,000 liters big. As you give in that inoculum into that media that's in 10,000 liters, those bugs are growing, they're going through more iterations, more evolutionary development and divisions than the human evolution easily over a 24-hour period. All you know at the end of the day with that probiotic is the genus and species of it. You have no idea if functionally that bug has changed going through all those divisions to get a nice full fermenter. So what I would say is needed to be done are functional assays of the bugs before you give them to the patient, not just knowing the genus and the species, because one bug made in the Netherlands, one lacto for example, fermented in, for example, the Midwest over here will be different physiologically and this is a huge problem in the veterinary sphere where we give probiotics to millions of chickens a year for inflammation and other things. I've had at meetings people from companies have come up and say, what did my probiotic work for inflammation in these chickens and we're to have millions of birds. One time then I got another lot in and it doesn't work. Then I got another lot in and it kind of works. I think we're gonna find the same thing with humans if we don't get to what Vanessa was talking about and go beyond genus and species and say functionally when we get this bug out of the fermenter and wanna use it in a patient, do we know functionally it is doing what it will need to do? Are you gonna answer this question? Yes, please. About probiotics. So Justin Sonneberg has done beautiful work on promoting probiotics, not probiotics. So you're adding in fibers that you know will promote the beneficial bacteria to bloom and to reach a level that will confer protection. So a lot of people, they actually, they're not entirely lacking the beneficial bacteria, it's just that they don't have enough of those. So we already know what kind of nutrients better to promote the expansion of those bacteria. So we can really look into certain probiotics such as fibers that can be an alternative to really promote the abundance of those bacteria. And Justin Sonneberg also had a beautiful paper illustrating that not just fibers, you can actually add fermenting bacterias at the same time that will help facilitate the digestion of the probiotics in the intestine. That will actually maximize the benefits of the probiotics. So that would be another thing to consider moving forward. Thank you. So we have one question on Slido which is directed to you, Melody. It says asks whether the, has the colonic 5-HT been measured? And another question is whether any toxic effects have been observed at 20 to 40 micromolar of 5-HT? So we go watch the neonatal mice with that. I believe that those of the serotonin we so far have not seen any side effects as far as for the immune cell response. Actually, we saw increased T-rex and reduced T17 cells. That's actually what we wanna accomplish in the newborn's intestine so that they develop immune tolerance so dietary antigens and commensal bacteria. We do have another study going on. That's really based on this work. We asked the question of now we know the bacteria, the unique bacteria in babies are actually suppressing the MLA which breaks down serotonin. So now what about the babies that have been exposed to antidepressants through the mom's breast milk? So the mom's taking antidepressants and then that can be transferred to the babies that would allow increased levels of the serotonin in the babies. But now we know babies actually having bacteria that would suppress the breakdown of serotonin. So that combination can actually increase the serotonin in babies so justly that can actually potentially be cytotoxic. So that's what we're exploring now. Thank you. So we have now got one more question. Sure. I'm sorry. I just can't help it, but I just the term food is medicine is taking on a whole new meeting. And perhaps this question is to Diego and to Mark or though others may want to weigh in as well. You know, with the work that's being done to identify novel chemo sensory signals that we end the ligands and agents that can activate them and drive a variety of desires and food signaling and with our capacity to engineer food in a variety of different ways, whether through genetically modified foods or tissue engineered foods. Do we have the appropriate regulatory framework to understand potential modifications that may be taking place to drive food desire as well as sometimes well intended effects to drive a certain benefits from food? I'm just wondering if you might be able to weigh on this. We always thought it was sugar and salt and maybe a few other sort of factors that would be in our food products, but as you alluded to, there are many others apart from the whole world of colorants, but Diego, is this something, a lot of GPCRs that are orphaned but are quickly gonna fall out of orphan status? I'll jump in. Can you all hear me? I cannot help but to think of the terminology that we colloquially use, like trust you've got now if it's taking a whole new meaning because we use all of these senses, the eyes, the nose, the tongue, the ears to find food. But once we put it in the mouth and we swallow it, then we literally have to trust the gut. And over the last 30, 40 years with the rise of food processing, we have been able to provide a type of skewed information to the gut. Like you will see broccoli with chicken flavor, buffalo sauce, Cheerios and so on and so forth. So essentially what we're providing is misinformation to the gut and the gut all it has to do is misinterpret that information and then drive certain desires, right? So I think that in terms of regulation there is a whole new arena in there. We just did not know what are the unintended consequences of that. As a consequence, it's very clearly seen since the appearance of artificial sweeteners, obesity actually has gone up instead of going down. So that's the first part of that point that I wanted to make. The second part is that now we are truly feeding two organisms that have a common purpose which is to eat but in both microbes and the host need to eat but the microbes cannot go to the grocery store. They have to rely on the human. So the microbes are gonna have to influence the human or the host to go and find that food. And we really need to understand how is either those specific nutrients drive the behavior of those bacteria. So they can guide our impulses. And specifically to the work that a little bit of the summary of the work that I presented to you. What is fascinating is that from meal to meal the amount of flood gelling that is shed by the microbes it seems to fluctuate. And they respond to the food in general and they shed the flood gelling and the flood gelling is what regulates the amount of food that we eat over time. But obviously one main question in there is different types of food perhaps will cause different types of shedding and that just for like those microbes that have like gelling, right? So certainly the microbes have the awareness of the nutrients that we are ingesting and they are able to influence the body. So I think that those two components are gonna go side by side to be able to tackle the soldiers of the gut and the soldiers of the brain because both are tightly linked to, you know, IBS and IBD, they have a strong component of mental health issues. And when you talk about your question really that you asked is really touches on the whole field of functional foods. Functional foods has been out there for many, many decades because even that's not in your head. So you can go to the market or go on the web and see you can buy Eldopa-infused bread, how to make it, that's well known. Treatments for Parkinson's in the third world since they can afford cinematics, they go after mucuna beans because they grind them up and there's recipes for doing that incorporating into your food and eating things like fermented foods for fermentation of foods goes back centuries that if you ferment what is sauerkraut good for sauerkraut is Koli, what do you make in sauerkraut, acetylcholine? And then you start eating it. So what are those effects of these functional foods as you eat them? And in my own experience, one thing I did because I was interested in all the bacteria that are in yogurts, for example, I once spun down a lot of yogurts from different manufacturers and I found a plethora of different neurochemicals and neurotransmitters at very large quantities in our common yogurts. So functional foods is a whole realm that can be exploited in a very much in a way that Diego was talking about. Well, in the interest of time, we wanna thank all of our speakers and for such an interesting discussion today and we'll certainly continue this on in the next two sessions. We are gonna take about a 25 minute break for lunch, not an hour. And we'll come back here at 1.45 Eastern time. Thank you. And it brings us back to some of the initial summary that Elliot made this morning about what we wanna look for in the next five to 10 years of the gut-brain access research and new technologies. And two of the things that were mentioned by Elliot in the summary will be touched on in this session. One of them is therapeutics and advances in therapeutics using the gut-brain access. And the other is on new methods to be able to understand and coordinate and measure connections between the gut and the brain. We have three speakers, Todd Coleman, who's the chair of this workshop as our first speaker from Stanford. Our second speaker is Hubert Lind from Minnesota and our third speaker is Stuart Campbell from Axial Therapeutics. When I was also thinking about the title of this session and its topic, it reminded me of work that was published in the history of neuroscience in the autobiographical form from a former professor of mine at Berkeley, Marion Diamond. She was a known for her work in neurobiology and neuroanatomy specifically and neuroplasticity. And she summarized her body of work very succinctly by describing that she identified, or broadly the field identified five different domains and key factors that are critical for brain health. And the first one she mentioned was diet. And of course, all the speakers in the sessions in yesterday and today have spoken a bit about its relevancy. And I wanna have everyone of us think about what the four others are. I think they're gonna come up in our discussion. The other thing that she mentioned in her autobiography was a summary of some work that's been done, had been done at the time in how the immune system affects the brain. And this topic will also come up in this discussion as well. What she didn't have, and she unfortunately passed away in 2017 at the age of 90s. But at the time that she was doing her research, she didn't have access to the technologies that will be shared. And the coming technologies were gonna be seen. And this next session kind of gives you a preview of what to expect in the coming years. So I'd like to really then turn it back to our speakers. And our first speaker again is Todd Coleman. Let me give you a little bit of background in context to where Todd is coming from. Todd is an associate professor in the Department of Bioengineering at Stanford University. And his research is extremely multidisciplinary. He uses tools from Applied Probability, Bioelectronics, and Synthetic Biology. And since 2012, he has served as a scientific advisor for the National Academies, specifically in the domains of science and entertainment and exchange. He has also been serving as a steering committee member for the National Academy's Keck Future Initiatives on collective behaviors from cells to societies. And he has been selected as a National Academy of Engineering, GoproFlucturer, a TEDMED speaker, and a fellow of the American Institute for Biomedical and Biological Engineering. He received his vouchers degree in electrical engineering and computer engineering in the University of Michigan. And then he went on to receive his master's and PhD at MIT in electrical engineering. He did his postdoc research at MIT in neuroscience. Todd, I'm turning it over to you. I look forward to your talk. Thanks so much. Can everyone hear me? I hope that's a yes. Can't hear you. Yes? Yes, and we see your first slide. Okay, great. And so, as was mentioned, I did my PhD in electrical engineering, and then I did a postdoctoral study in neuroscience. And so you see the perspective that I have. It comes from electrical engineering, neuroscience, and also technology development. So I got interested in the digestive system because I was doing brain science research, and then my dad passed away from pancreatic cancer in 2011. And it turns out he lost his mom to stomach cancer, so I began to develop an interest in the digestive system. At first glance, I didn't know much about it. One of the things I learned, though, is that the common symptoms of pain, nausea, bloating, constipation, can give rise to many of these different conditions. And it was not surprising why it's the second most common reason that someone in this school would work after a common cold with a huge annual burden and cost. So when I began to look at, what role can I play as an electrical engineer or slash brain scientist, I when I began to discover that we actually have pacemaker cells in our digestive system, just like we have in our heart, and they are connected electrically to our brain via things like the vagus nerve and the spinal cord. I began to be very intrigued. You know, a little bit more about this in young cases, people don't remember, is when you're in a very calm state where there's parasympathetic activity, your heart rate slows down, you breathe more calmly, and that actually activates your digestive system, and we're gonna stress out state with your sympathetic nervous system, the opposite occurs. This is a very high level sort of explanation. So one of the things I've become an interest with GI problems is that you can imagine that the engine of the car is not operating correctly, meaning maybe there's something wrong with your digestive system, or maybe the connection between the brain and the gut, like the gas pedal via the vagus, or the brake pedal via the sympathetic nervous system is off. We don't know that at first glance by just looking at symptoms. So I love the idea that I got interested in is, well, what if we can take a look at what the EKG has done for the heart and try to do something analogous for the digestive system? So unfortunately the technologies that have been developed to try to place electrodes over the abdomen that were first pioneered in the 1980s were subsequently deemed as having doubtful clinical usefulness, and so we call them the Rodney Dangerfield of electrophysiology. But I'm originally from Dallas, Texas, and in the spirit of the Dallas Cowboys and the Dallas Mavericks, being a Maverick and still beginning to pursue exploring this. So to make a long story short, we revisited what had been developed in the 1980s and used modern signal processing and modern technology development. We were able to place multi-electrode arrays over the abdomen to capture electrical activity from the digestive system. We can determine the direction that waves are propagating, for instance. And with that information, we're able to take a look at normal control patients versus patients with digestive disorders. And here we're focused on the stomach and in a normally operating digestive system, the contractions should propagate towards the small intestines, which we found basically in controls and patients with delayed gastric emptying, we saw a lot of sporadic types of patterns. And so we were able to demonstrate, those are the first demonstration of a not a basic measure of gastric function that basically correlates with the severity of symptoms. So on the horizontal axis, you have the percentage of slow waves that are deemed abnormal, not propagating towards the small intestines. And on the vertical axis, you have a clinically indicated symptom score about the severity of symptoms. We're able to go one step further and be able to showcase that when the symptoms exceed what we predict from a measurement of the engine, that was consistent with doctors using non-engine-based approaches to treat the condition, meaning suppose that the doctor subsequently went along and tried to treat the patient for depression or for something else, then in those situations, the symptoms resolved, further sort of confirming that this technique can actually be used to disambiguate the engine from the gas pedal, from the brake pedal. We got interested in going in bidirectional fashions and saying, well, suppose there is a problem with the engine, how can we correct the engine? And so we were able to develop some wireless pacing techniques, marring ideas from cardiology, and ensure we can place pacing devices right into the stomach and you can wirelessly transduce them cutaneously and we can actually entrain the slow wave of the stomach. And if you see where this red vertical line is, if you look at basically before the red vertical line, before we did the pacing versus afterwards, we are entraining the activity of the stomach toward the target direction. So we wanted to go one step further and actually trying to build, miniaturize the measurement aspect of this. And so we were able to build some portable devices that you can actually record for 24 hours a day. And with this opportunity, this enables us to track meals, responses to meals, track sleep, bowel movements, et cetera. And one of the things that we found sort of unexpected was that we know that one of the things at the heart of digestive system sharing common is that they're modulated by the autonomic nervous system. So we focused on looking at sleep versus wake. And one of the things that we found is that we saw very structured patterns and controls of increased parasympathetic activity during sleep as compared to normal. Whereas when we look at patients with, for instance, diabetic gastroparesis, if we look here at the balance between sympathetic and parasympathetic, or sapapovagal balance, you see almost no change in the diabetic gastroparesis patients in sleep versus wake. And so by combining sort of autonomic measures of sleep versus wake, along with GI intensity scores and looking at meal responses, we're able to separate controls versus patients as well as gastroparesis with sort of diabetic versus non diabetic patients who with gastroparesis or with not with remarkable accuracy. Now, going one step further in trying to really miniaturize the measurement modalities, we wanted to take it one step further to use these thin structural electronic technologies. And we were able to do that. And basically we're able to build some electrode arrays that you can locate the fish of the stomach with the portable ultrasound, then place almost imagine like a Band-Aid or the abdomen and be able to track the type of information that I mentioned previously. Another thing that we were able to do rather than indirectly looking at the autonomic nervous system in terms of heartbeat, dynamics and sleep versus wake, could we actually try to directly pick up autonomic neural recordings. And so to make a long story short, we're able to pick up on cervical autonomic neural structures such as the vagus, the sympathetic ganglia and the carotid body, both in tasks associated with the cold pressure test as well as the time respiratory challenge. And these are tasks that are well understood that initiate autonomic responses. And we were able to take a look at bio-type specific changes in the firing and looking at different cell types with cross strengths. So everything I've talked about so far basically measuring one thing versus measuring another and I had this epiphany that, well, there is this gut and brain access, why not try to measure both things at the same time? So part of my interest in this, the stymie body, there's a finding in 2017 by a group out of France that demonstrated that if you record the electrophysiology of the abdomen as I've been describing previously with the stomach, and in addition, you use the magnetic version of EEG, and you can actually find that there is a coupling between the stomach and the brain that can be measured noninvasively. So for people who don't know, MEG is basically the magnetic equivalent of EEG. However, it's a very expensive, million dollar types of system, magnets, cooling, all of these sorts of details. But the finding that was interesting was that the alpha band of the brain is actually coupled to the phase. So the apple to the alpha band is coupled to the phase of the slow signal of the stomach. So region specific matters. So to make a long story short, we got interested in saying, this is exciting, but if you take a look at where MEG locations are, there's two in California, for instance, one in San Diego and one at UC San Francisco, there's not even one in Stanford. In many states, as you can see here, wherever it's gray, there are basically no such scanners. So wouldn't it be nice if we could replace MEG with something like EEG, and could we still pick up this coupling? And to make a long story short in a recent publication of ours, we can. On the bottom left, this demonstrates where we saw phase amplitude coupling between the stomach and the brain and what regions using the magnetic. And on the right, basically, you can see an hour piece information with our study in MEG. So, one of the things that we got interested in is, well, what underlies this coupling? So starting to take a look at some of the anatomy and whatnot, well, we know that the stomach, for instance, in the enteric nervous system, it connected to the brain via areas in the brain cell, like the nucleus trapezoleterus and the dorsum motor nucleus of the vagus. And interestingly, the afferent information, the key cortical area that gets engaged is the insula. And then there's a recent finding in 2020 that showed that from an e-faren perspective or top down, the main cortical source of descending control to the stomach is, again, the insula. So, thinking about, well, what's the story on the insula? Well, it's a very, very interesting brain structure. Here's a 2007 story from the New York Times about it. And the high-level ideas that is tied to social emotions, lust and disgust, empathy and incombent, a very interesting brain structure. More importantly, it's implicated in this notion of, we have external assumption, which is like basically our ability to perceive external types of information. But what about our internal organs? Well, the notion of interoception, of perceiving the body's internal state, the main cortical area involved in that is the insula. And interoception has picked up a lot of attention clinically because of its going wrong or going a rise, implicated in things like drug addiction, alcoholism, anxiety, Parkinson's disease, eating disorders, et cetera. And one of the main measures to determine interoception is one of them is called the water load test, which is used clinically with eating disorders. And we learned about, you know, satiation before in the talk previously. And so the high-level ideas, you tell someone to drink until they feel satiated and you measure the amount of volume of water that they drink, then you tell them to drink again and drink more until they feel full. The high-level idea is that if we have good gastric interoception, then the exceeding amount associated with being full should be very small. Being full and being satiated should be similar, which means that the ratio of satiation volume as compared to satiation volume plus the additional amount for full, that ratio should be near one if you have good interoception versus if you don't. So this is a measure that's been validated clinically. And we got interested in determining, is there a physiologic measure of interoception? For instance, what if we did simultaneous gut-brain measurements where there was determined that there's this phase amplitude coupling between the stomach and brain? What happens if, is that at all associated with this behavioral measure? But just this past summer, we did some studies along those lines and we basically were able to, with a very small number of subjects, this is still very early on, that the behavioral measure of gastric interoception appears to associate with this physiologic measure of interoception. And you can envision how this could begin to get very interesting because of all sorts of ways of doing neuromodulation. So to basically conclude, I talked about first extracting electrical patterns of the stomach, not invasively with an emphasis on spatial patterns because they correlate with symptom severity. I demonstrated some work that we've done on actually a wireless pacing to entrain spatial patterns of the stomach. And one thing to remember that's interesting is that we know that the stomach and the brain are connected, so we envision in some situations, pacing the stomach to not only treat the stomach but possibly to treat brain structures. We demonstrated an ability to extract, both indirectly via heartbeat dynamics, information about autonomic function in sleep versus weight before after meals and to subtype GI disorders. But also we demonstrated with the advent of novel stretchable electronic technologies. So not only pick up gastric electrophysiology more intrusively, but to get direct measures of autonomic function in particular cervical neuronal structures. And then lastly, the thing I talked about is our recent works in developing approaches to simultaneously measure the stomach and brain and how they co-vary with one another and in particular focusing for the moment within the topic of interception. So sort of moving forward, what are interesting ways to think about how this relates to what other people are talking about? Well, we're gonna hear from Hubert soon about the immune axis and whatnot and a couple of things that I would like to highlight is that we know that the immune axis and the microbiome are interconnected and it turns out that the microbiome is connected to some of the stuff that I talked about with the gastric electrophysiology through short chain fatty acids and a lot of these regulatory mechanisms of how when glucose gets too high or too low it actually sends feedback signals from the stomach from the small intestines back to the stomach. So one can envision the future by doing things like measuring the gut microbiome, measuring hormonal signals, understanding the immune axis, along with electrophysiology, we can see how these different pieces of a puzzle all come together. So with that, I conclude my talk. Thank you. Wonderful talk, Todd. In the interest of time, we're gonna go on to the next speaker, which is Hubert and we'll save questions for Todd to the end with our discussion. Hubert Lim, it's a professor in biomedical engineering and auto technology departments at University of Minnesota and he was hired as an Institute of Technology Neuroscience Scholar. He currently hosts the Endowed Lions Professorship and he's also a co-director for the Center for Neuroengineering. He completed his bachelors in bachelors of science and bioengineering at UC San Diego, followed by a dual masters in biomedical engineering and electrical engineering and computer science and then a PhD in biomedical engineering at the University of Michigan. At the University of Michigan, his lab research focuses on neuroengineering, sensory neuroscience, neuroplasticity and neuroimmune physiology, which is gonna speak about at this meeting, including inflammatory conditions in collaboration with multiple clinicians and companies. He also is a chief scientific officer for two startup companies, NeuroMod in second way. Hubert. Thank you for that introduction and thank you for the organizers. Also the committee members for inviting me to speak here and organizing this meeting. I am not a brain, oh, did someone say? Oh, okay. I am not a brain gut scientist but I've learned a lot being at this meeting where I think I can fit in is I currently work on modulating through the Vegas nerve pathway to the spleen for immunomodulation. Originally I'm a hearing scientist so you can see I'm coming from a bit different topic but I think the technologies that we're using, particularly ultrasound, to modulate the nervous system in an organ could be potentially beneficial or it could be a unique tool that could be leveraged in humans but animals as well for probing the gut brain axis. So I think that's where I can help fit into these discussions. It was already mentioned, I have disclaimers, part of two companies. I won't speak about the first one which is the Tinnitus Hearing Company and the second one there is financial interest and equity. I will talk about that topic related to second wave systems and of course the funding sources that have been generous to support this. I just wanna start here. I know we're talking about the brain and we'll come back down to the end organs. It's just been an exciting time. I started in cochlear implant field but there's just so many novel technologies coming about with brain machine interfacing implants from Parkinson's tremors, depression treatments with deep brain stimulation. You have visual prosthetics, spinal cord stimulators and what you heard from Dr. Kevin Tracy, vagal nerve stimulation. And there's a lot of opportunities there in terms of being able to get specificity, that's great. You can interface with that nerve. One of the challenges though is that it is invasive and being able to probe or interface with a lot of patients or a lot of individuals especially for science investigations becomes a bit challenging. On the flip side, you have non-invasive technologies and you've got things like transcranial magnetic stimulation which creates magnetic fields can activate the nervous system not only the brain but in the body. You got old school current stimulation, transcranial or body stimulation like TENS devices. Those are great in terms of accessibility, large scale use, but then on the flip side they lack the focused ability. So this is why ultrasound has become such a hot topic in the last 10 years because it's kind of achieving the best of both worlds. You're gonna get ultrasound with beams that can be narrow but even using many transducers kind of like surround sound TV with your speakers you can focus in where you're sitting. You can do that also with ultrasound, cancel things out and beam form into regions into the skull and to the body and it's non-invasive. So this becomes very exciting. It can be for the nervous system in the brain, peripheral nerves potentially and organs. Now the big question is with all this excitement, does it work? And I spent about the last seven years discovering what can work, what can't work, a lot of other colleagues of mine also. And just the take home message is that with ultrasound at least of nerves and neurons it's not causing strong like action potentials but multiple groups are showing that you can modulate cells, depolarize cells and there's other literature showing that you can perturb non-neural cells as well. So this gives you an opportunity now that you can look into different ways of probing the system even in a modulatory way to control function, right? There's a lot of research has been done. This is just some animal, because I'm trying to tour the force here with ultrasound what's going on to bring it up to speed but I just put some of these in if you're interested you can look them up. Some of the early seminal studies in animals there's been a lot of studies in humans already. So humans ultrasound of the brain, ultrasound of the body and they're using parameters that are typically within the diagnostics safety limits. So we're talking about using parameters lower frequencies few hundred kilohertz to maybe low megahertz and pressures that you're looking at at few hundred kilopascals maybe up to mega Pascal. So if you look those up but generally speaking that they're within the safe limits and you can change the different parameters and pulse patterns, right? One thing though is with all this excitement and we didn't intend to find this we found out that there was a lot of confounding factors. You couldn't excite neurons directly with action potentials easily, you can modulate but we also found that the ultrasounds activating the skin and vibrating the skulls. There's three bospinal fluid in your brain is connected to the fluids in your cochlea. So when you vibrate the fluid you actually cause hearing with ultrasound which is interesting when you cause vibrations of your fluid. So there was a lot of confounding factors that we discovered but when you remove those for example we deafened the animals in these studies and then you could see what could be potentially modified or modulated. And this was not just our group the collaborators Dr. Michele Shapiro and Doris Chow over at Caltech independently found similar effects. Another paper that was interesting I'm gonna show you a couple of different studies very quickly Dr. Michele Shapiro at Caltech he did a beautiful study with his lab where we're trying to get the mechanism what's actually happening and it may not be causing action potentials per se if there are like a piece of electric or kind of mechanical channels, sensitive channels you can vibrate those and activate neurons or cells but it appears that what's going on is that when you cause mechanical perturbation of the membrane there is some calcium influx at one mechanism of it that then kind of amplifies other channels to open eventually becomes a voltage gated channels that cause not necessarily spiking but at least depolarization and maybe could lead to spiking depending on the ongoing activity. So this is one mechanism that has been shown by their group, there's a few others. The other thing I wanna point to just so you're aware is that there's already been 20 to 30 years of research before a lot of this hype and neuro where people were doing non-neuro cells, right? So chronicide stem cells endothelial cells and they had been showing similar parameters around one megahertz some of the similar pressures that you could activate integrin receptors or other mechanical sensitive receptors and so forth. So this shows that on the modulation excitation side there's a lot of potential. Now, if you increase the intensity a bit you can actually and in longer durations you can cause, right at the border of thermal you don't wanna do too much thermal because then you can actually cause lesions but if that fine region if you can use a thermal mechanism you can actually shut down neural activity or ongoing activity, not only in the brain but we also showed that this is possible with nerves. And I'm bringing this up because later as we get to end organ modulation it helps us to define what's actually being activated. It's not easy to cause actual potential nerves but potentially you can modulate neurotransmitter release from axon terminals by depolarizing them but also on receptors. So let's jump here, this is the part hopefully background was helpful for ultrasound neuromodulation of the brain and peripheral nerves but this is where I got quite excited. Really it stem from the seminal work Dr. Kevin Tracy already presented where you have this vagus nerve pathway to the spleen and you can drive an anti-inflammatory effect. And the question is, is that pathway can we access it non-invasively directly by stimulating the nerves and the end organ, the spleen to modulate or control the immune system. And this was already covered by Dr. Tracy. It's becoming more clear now that there is a bi-directional communication of the brain with the immune system. And particularly one of those pathways is this brain through the salic ganglion to the spleen. And one of the pathways here you see is release a norepinephrine activates the T cells a cholinergic pathway there that then causes the immune cells or macrophages and immune cells to kind of the breaks that Dr. Tracy was talking about that could shut down this hyper inflammation. What got me very excited was this 2016 paper that was already shown that in actual patients who are getting vagus nerve implants they're able to show this reduction in TNF, IL-6 and some other inflammatory cytokines or biomarkers. So we set out to do an experiment in rodents. What was interesting is this was DARPA funded and we were doing it independently, but GE, Dr. Chris Puglio and a team there also with the Feinstein group with Dr. Kevin Tracy's group there they were doing research independently. We didn't know about them. We were connected by Dr. Eric Vangiesen at DARPA who said, you got to share results because you're getting similar results which is encouraging. And so we decided to do companion paper. Ours is in a inflammatory mice model, a chronic or semi chronic and theirs is an LPS sepsis type of model Q and they showed more of mechanisms of action. The idea here is we took animals with inflammatory arthritis. We applied ultrasound, target ultrasound to the spleen and the idea is to access this pathway that I talked about before in the rodents and you could see hopefully the swelling of the ankles to represent the arthritis. Big team that was involved, Dr. Bryce Binsad is one of the MD PhD co-directors at a university has a beautiful model, mouse model for inflammatory arthritis. And then we have other colleagues there, Metronik was involved with this in the early stages as well. So this model, you basically take this KBXN model of inflammatory arthritis, you take the serum out and you inject it into animals and over very quickly in a period of days that arthritis forms in the animal it's not long lived, you got about 10 days. So you have to kind of time the experiments to track it within that maybe seven, eight day period. And you can measure the ankle thickness as well as some clinical scores. So what we did is we took the animals, this ultrasound transducer at the top and there's a cone and the cone basically for most purposes it's like a pointing device because we already characterized the beam and the beam due to the curvature of the ultrasound transducer helps the beam to be focused and the cone allows us to kind of pinpoint where that focus is. And we're able, the spleen is right underneath the skin and animals and the rodents in humans it's underneath your ribs. So there are some technologies that we have to develop for that. And we basically provide ultrasound over eight days we provide the serum on there you could see day zero a negative one, you know we shave the animal, do ultrasound already and then we daily stimulation in this case two minutes up to 20 minutes it depends on the protocol. And then we collect blood samples to do cytokine analysis and we track the animals. So just, you know, quick summary encouragingly what we found was that we could reduce the ankle thickness. You know, normally they would get swelling of their ankles we're able to reduce that substantially and also the clinical scores which is more visual assessment of the joints in these arthritic mice we're able to greatly reduce those using in this particular condition was one megahertz and very low, low intensity pressures 250 kilopascals. But what was important for us was a dose response. So we did two minutes and then we did six, 12 and 20 minutes and what happened was that we're able to further reduce the swelling, right? So change in ankle thickness you want that small because it means that you're not, you know experiencing the swelling and 20 minutes is what we then continue to use for clinical trials in the future. This is what I really appreciate about, you know Kateria at all, this is the GE paper with Dr. Kevin Tracy is that they did quite a bit of controls and mechanistic studies and blockers of the synapses here and what they found, is there a mouse? Oh, well, I'll just explain it. So what you see there is they basically narrowed it down to this synapse here in their model it wasn't all due to that but when they blocked that interface there most of the benefit disappeared. And so what could be happening is ultrasound as I mentioned, through our studies it doesn't excite easily the axons, right? The nerves we tried in every which way but based on the study you saw from Caltech Dr. Shapiro's group they just possible calcium influx or some kind of perturbation at the terminals that releases the neurotransmitters the other side is you saw with those papers non-neuro cell modulation you could basically apply ultrasound to the receptors of the immune cells, right? So that could be also what's happening could be one or both but the idea is then you can actually activate that synapse as one part of it and cause that cascade for anti-inflammatory effect. So does this work in humans? And this is where we spend quite a bit of time for the last three years pushing forward different human studies and because our device we didn't have a device ready yet the idea was potentially using a cart-based system that G had been using for different studies and this one actually was a study that was done by GE Caspulio and Feinstein group where they actually stimulate healthy subjects of the spleen to see if you could kind of their baseline cytokine levels, they could reduce it and also we had some RNA sequencing data in an ongoing rheumatoid arthritis study that we're able to pull that out while the study was still going to assess and we decided to try to pull this together into a single paper that's on Med Archive. This study basically what they did was they used a system that can image the spleen that's the beauty of it, I think 160K device but that's the beauty of it you can basically image the device or the spleen and then you can target and stimulate the spleen and they only had to do it for three minutes and when they did that for three minutes they had those stimulated and those not stimulated then they took the blood samples out because it's healthy subjects, the levels are low but they mixed the serum with LPS to kind of amplify the signal and they're able to show that you had significantly reduced inflammation, right? So less TNF release response for the stimulated case. So this was inhuman subjects. We ran or had run a rheumatoid arthritis study using off-the-shelf device due to COVID it got delayed but we just finished we didn't get to 20 but we got to all 20 but we got to 19 but what we were able to do was look at the cytokines that were modulated and what we found was that for the RNA sequencing pathways IL-1 beta, IL-8, NF, Kappa B and interfering gamma, gamma. What was interesting at the time was COVID was happening. This is when we got the data and they were showing the opposite, right? For SARS-CoV-2, those with inflammation were having an increase in those cytokines. So that then allowed us to go to DARPA and MCDC and different funding agencies to see that if we could run COVID studies in the hospitalized patients to see if we could reduce their inflammation. And so we have that study that was run. We started with a GE device. We enrolled 30 patients. We just are about to submit it. So I don't have it to share today but I promise it'll be up met archive pretty soon. So you'll be able to take a look at that and there's the clinicaltrials.gov. It was encouraging. So we're excited to share that with everyone when it's published. We have an ongoing rheumatoid arthritis study and this is where my interests are with second wave. We've developed a lower cost of device that can be taken home. Still need some smartness into it but you could see there the size of it about the size of your hand and it is placed over your ribs. It has some smarts to be able to find the ribs and then stimulate between the ribs to the spleen. And we've recruiting 15 patients. We've halfway done now. So hopefully in a few months I'll be able to present these results to you as well. The other thing where I wanna end on is just a few examples where it's been exciting to see that it's not just ultrasound of the spleen. There's opportunities for ultrasound of the spleen to affect other applications or organs or conditions but also ultrasound stimulation other and organs or targets. And here this is one of the seminal classic studies that happened early on was Dr. Mark Acusa and his team. They basically stimulated the spleen with ultrasound and they could actually see that they helped treat or improve the condition with acute kidney injury. They used the approach where they blocked the blood flow to the kidneys and then reperfusion and then they showed that you can reduce the inflammation but also the damage associated with ultrasound of the spleen. And they showed that the spleen was necessary or necessary for this to happen. This I have to give credit to Chris Pulio and GE team because they have stimulated with ultrasound a lot of different targets. They targeted the celiac ganglion or ganglia for looking at inflammatory bowel disease. And they showed that they could have improvements for condition related to IBD. They've stimulated the liver. They've stimulated the pancreas and they stimulated the intestine region. And all of those they've shown that you could modulate different biomarkers. Now they predominantly focused on glucose inflammatory markers but there isn't a reason why there couldn't be other biomarkers that are modulated if those are. And you could see there, if you're interested the two papers a nice overview of what's going on and opportunities. I just want to end with this one because I'm just always impressed with Chris and his team. They did a lot of animal studies and then they moved to large animals. It was in diabetic mice, rats and then swine but now they have a clinical trial running in diabetes patients using ultrasound of the liver. And from what I understand it's been encouraging and hopefully those results are going to be presented sometime soon as well. So with that, the last bit is there are technologies. We're developing technologies with the company second wave. We've got non-significant risk determination which is encouraging from FDA. So we can use them in humans. GE's been developing a portable, more miniaturized systems. And I think there's opportunities to leverage these technologies to probe the body. If you're interested, we did a podcast that's just here. Chris, myself and Miquel Shapiro with Arun and scraps Jojo, they have this thing. If you're interested in, we talk about all these different ultrasound areas. And with that, thank you for your attention. Thank you. Thank you, Herbert. Well, time for a few questions after the next speaker. Stuart Campbell is our next speaker. He is a CEO of Axel Therapeutics. He joined Axel Therapeutics in 2017 as VP of Research and Development. He has over 30 years or about 30 years experience in biopharmaceutical areas. And he obtained his bachelors and with honors in chemistry from St. Francis Xavier University, a PhD in organic chemistry from Queens University. And he did his postdoctoral work at Duke University. He's going to share some of his new business development for Axel Therapeutics and some of the new technologies they have focusing on the gut bearing access. Thank you, Stuart. Thank you very much. Thank you for the privilege to come and speak to everybody today. I think what maybe we'll introduce Axel Therapeutics to you just very briefly. And what our mission is, we've heard a lot these last two days about probiotics, live biotherapeutics, consortia and so on. And those are really promising approaches to addressing issues and disorders that lie in the gut brain access in terms of etiology. We have a different approach. It's not better or worse. It's just different in that we really want to take a more traditional drug discovery approach to really look at the gut microbiome as a source of drug targets that we can hit with small molecules in a variety of different disease indications. And so we're in the clinic. I'll tell you the autism story today. We are clinical stage. Our focus is in neurological indications, autism and Parkinson's disease principally. Again, the differentiator here is that all of our drugs are small molecules that target things in the gut, either microbes in the gut or something on the host side at the gut microbiome interface with the host. And so the key here is that our compounds need very little if any systemic absorption, they don't need to get to the brain. We think this imparts some advantages over traditional CNS drugs. And that seems to be bearing out in our clinical work. The way we go about this, we heard several really nice introductions to how different groups approach interrogating the gut microbiome and building model systems to ask certain questions or answer certain questions. In particular, from Michael Fischbach earlier today, went through a very nice sort of progression from a full fecal metal transplant to a very reductionist system and then to a very controlled consortium. We do the first two and that's really helped us get to therapeutics, but it's these reductionist systems that we think are particularly valuable for us because I think it does accelerate getting things into the clinic. And so one of the other benefits of small molecules is once we've identified a development candidate, it becomes very familiar territory to potential large pharma partners who are essentially going to put the bill for the large phase three studies and so on, but also regulators are very familiar with this. And so we can remove some of those, I guess in a way self-imposed barriers or hurdles on getting this type of approach into the mainstream. And so what I'll talk about today is the first two or three chapters of our journey on the autism front. And it starts more than a decade ago with work done by Elaine Schaue. We heard from a lot yesterday and Sarkis-Mosmanian at Caltech and then since developed further by Sarkis-Mosmanian's group at Caltech. And then more recently where we've collaborated with Sarkis to translate some of those basic science findings into the clinic. So we do have a pipeline that we're trying to build. So I won't talk about our Parkinson's work or oncology work today, but based on the same platform, which I think I'll use the autism program to illustrate is we're moving those closer to the clinic. But I'll talk today about our clinical story in autism with the molecule we call AB 2004, which I think it really, I hope will convince you that these types of basic science findings can be translated into the clinic. So I'll take a quick step back for a moment and just orient everybody on how we're approaching autism. I think autism touches a majority of people either directly in their family or their extended family. This is a very special population, incredibly diverse population. So it makes it really hard from a drug discovery point of view to know exactly how we can phenotype or categorize certain patients in order to develop a therapy. So we've taken a little bit of a tangential approach in that we've focused on irritability associated with autism. And this is a pretty significant problem. So most kids in their youth through to adulthood will struggle with things like irritability and anxiety. By irritability, I don't mean someone's in a bad mood or a little bit edgy. What I mean is these are self-destructive behaviors, self-harm, destruction of property, tantrums, things that become increasingly difficult to manage as kids get older, they hit puberty, they get physically stronger. And so this really leaves families in a predicament, but also the people themselves, they have trouble going to school on a regular basis and that obviously impacts them throughout their entire life. There are two drugs approved for irritability associated with autism. Ari Pippers-Ole or Abilify and Resperdone. Both atypical anti-psychotics, they work. They do calm the kids down, but they sedate the kids essentially. And so most families, when they put their child on an atypical anti-psychotic or irritability, it's really trading the irritability symptoms for another set of symptoms that are easier for them to cope with. It's not a really a fair choice that these families have to make. And so our motivation here is to provide something where that choice becomes much more agreeable. And so we go all the way back just to introduce you to how we got to where we are today is to a seminal publication by Elaine Schau and Sarkis-Mosmanian back in 2013 where we heard earlier as well about the maternal immune activation model in mice, which is from a face validity standpoint is representative of the risk associated with a child developing autism if their mother is infected at a certain time during their pregnancy. They modeled that in mice and thought that maybe the microbiome may be contributing to that because of the prevalence of GI symptoms in kids with autism. And what they found was by profiling all of the metabolomic composition in blood as well as what the microbiome was comprised of and the behaviors. And they found in these mice that had the so-called autism-like symptoms that there were a number of small molecule metabolites circulating in the blood that were dysregulated. And the one that was the most dysregulated was this molecule that we call 4EPS. It's for ethylphenol sulfate. It comes from tyrosine metabolism by the gut microbiome. So they did a very important experiment here. So on the face of it, maybe this 4EPS molecule could be a biomarker of some kind of clinical phenotype. And that was very exciting in and of itself. But the very important experiment that they did was to actually take synthetic 4EPS and inject it daily into wild-type mice. And they were able to induce a very profound anxiety phenotype in these mice. It didn't impact social interaction or repetitive behavior. Some of these other sort of core symptoms we associate with autism, but it did actually impact the anxiety phenotype. And so this was an example of how a dysregulated metabolite has its own pharmacology. Wasn't clear at that time, whether it was acting in the brain or somewhere else is systemically or what it was doing. And so that led to chapter two, which was to really understand first, is this 4EPS molecule a mouse curiosity or is it actually present in people? And so we published this paper last year in Biological Psychiatry, where we profiled about a hundred kids with autism against their match control peers. And we did full-blown metabolomics in blood and urine. And what we found was much to our surprise and delight was that the molecule 4EPS was one of the most dysregulated metabolites in the entire dataset. And in fact has a strong association with phenotype and also with GI symptoms as well. So this was the first indication that perhaps what was shown in mice might actually have some relevance in humans, but we're a long way, this is still just a correlation. And so to get to the bottom of this correlation versus causation, we go to this middle scenario that Michael Fischbach pointed out earlier, this really reductionist admittedly very contrived system. But our goal here is to use different technologies to isolate variables from a very complex ecosystem of the microbiome interfacing with the host and design a system where we can turn on or off a very specific feature and then look at the behavior, look at the tissue response to the presence or absence of a metabolite. So briefly what was done here was germ-free wild-type mice were colonized with two microbes and in the 4EP plus case that you see on the right-hand side, a little mouse, is the two microbes combined to constitutively produce 4EP fourth ophenol, which is the precursor to 4EPS. 4EPS is formed in the liver after it's absorbed. As a control group, same two microbes just with the last steps of the 4EP biosynthesis engineered out. So we show here is that the mice that have the competent microbes actually produce a lot of 4th ophenol sulfate and that 4th ophenol sulfate shows up in the brain. So this allows us to maybe now behavior test and look at all of the brain tissue and test the various functions in these animals to see exactly what 4EPS is doing. And so this was published in Nature in February of this year from Sarkis's group at Caltech with Brittany Needham who's now an independent researcher at University of Indiana. And what they showed was that chronic exposure to 4EPS by making 4EP via the gut microbiome led to a strong communication phenotype and a strong anxiety phenotype. So this was really the first demonstration that more natural exposure to these metabolites and then toggling them off with the control group we could actually see differences in behaviors. So the question here is what's going on and how might we build an intervention based on what we learn. So what was done in this study is a lot of work. I'll just show you one slide here on really what the, I think the take home message is is that 4EPS when it enters the brain it actually arrest oligodendrocyte maturation. It halts oligodendrocyte maturation at the precursor stage. And as you can see the graph on the bottom left that the total number of oligodendrocytes in the whole pool doesn't change from one group to the next but the number of mature myelinating oligodendrocytes is increased in the metabolite exposed group the 4EP plus group. And you see that in the fax images on the top right and also on the bar graph on the top left. And so what this suggests is that 4EPS is not inherently toxic. It simply has a regulatory function in impacting the ability of oligodendrocyte precursor cells to mature oligodendrocytes that can make myelin. Now what we've done is shown that 4EPS actually downstream impacts the myelination capacity of the brain and this is a very region specific. There's brain function and MRI images in the brain to show changes in connectivity and function in specific regions of the brain that are closely associated with emotional regulation. And so with that understanding, we said, well, how are we going to intervene here? What can we do? We could target the brain, but we still don't know exactly what 4EPS is doing. You don't know the target in the brain yet we really want to understand that. But if the culprit are these metabolites and they're being produced exclusively in the gut, why don't we just sequester them in the gut and have them ushered out in the stool? Taking that approach actually separates us from needing to know exactly which microbes are making it. And I think what we learned or saw earlier today and yesterday was that there's a lot of functional redundancy in the microbiome. So the idea of targeting the microbes that make it, it's not one. This is a very common biosynthetic pathway in clostridia. And so the idea of actually knocking that function out of the microbiome is a pretty tall order. So let's not bother with that. Let's just take the metabolites out once they're produced. And that formed the basis of AB 2004, which is a non-absorbed sequestrant molecule. Take it orally, travels through the GI, passes in your stool and it picks up these metabolites along the way. And so we went back into our mouse model. We put this sequestrant in the food and what we showed was that we could in fact rescue the anxiety phenotype and the repetitive behavior phenotype by treatment with the AB 2004, which reduces the four EPS or four EPS levels systemically by reducing the four EP levels in the gut. So we could have paused here and done a lot more preclinical work to really understand exactly what's happening in the brain. And what we thought was, we have a very safe molecule here. Let's put it in people. Let's go to where it really counts. And so we did an open-label dose escalation study in adolescence with autism. We were lucky enough to get this study completed in March of 2020. We were, I was literally coming back from Australia, being asked if I had been in China recently because they were worried about the coronavirus coming into the US. So we literally came in at the wire on this one. And one of the things that we did here was we were very cognizant and we had a discussion about it earlier about whether we can trust animal models, how translationally relevant are the end points we measure and we had that same concern. So going into this study, I talked about the anxiety phenotype and the repetitive behavior phenotype in the mice. We don't know how that's gonna translate into the human condition. So we measured a lot of things, a lot of different ways because if there was a signal there, we wanted to see it. And if we saw the signal, we could down select for future clinical trials and really hone in on an indication. And so as we expected, the drug was incredibly safe. It was a 30 patient or 30 subject trial. Done in three sites in Australia, New Zealand. And yeah, so no surprises there, that was great to see. And what was really exciting was that we did see significant reductions in a host of structurally related metabolites that share a lot of common features with 4EP and 4EPS. So paracresol sulfate, endoxyl sulfate, et cetera. And we expected that was really great to see, but what was really nice to see also was that we saw significant changes in several endpoints, most predominantly in irritability, which you can't really measure in a mouse, and anxiety. So we actually saw some translational value here between the mouse and the human, at least at a face level. And what we saw was that the irritability scores came down and between the end of treatment, the EOT and the recovery period four weeks later, the metabolites shot back up and the irritable behavior began to return. So the nice part here, another added feature is that the fact that the metabolites rebounded to baseline levels after removing the drug implies, we need to prove this, but it implies that the microbiome is not undergoing a significant alteration by virtue of being exposed to the drug. So the irritability signal was particularly interesting because there's a clear regulatory path. So we think in industry about how to get something approved, irritability isn't the end of the story, but it's maybe how we get this drug onto the market and helping kids because the FDA understands this indication, we understand how to get a drug through there. And so there's, again, it's about removing risks and barriers to getting a drug approved. So currently we are in the middle of a placebo-controlled phase two study, US, Australia, New Zealand with the primary endpoint being change in irritability at eight weeks of treatment. We plan to read that study out at the end of 2023. And so this is now, I guess, chapter four of this story and I hope it's a long book. So I'd like to thank all the great collaborators that we have. We have a very close relationship with Sarkis and his lab and a really dedicated and talented team at Axial along with a very strong investment group as well. So thank you for your attention. Thank you, Stuart. We've got a few minutes left for questions. I wanna start off with one that's broader that I hope each one of the speakers could address. One question is how translatable is your work that you're studying that relative to microbes perhaps in the gut or in the spleen, which Hubert mentioned, to other tissues and their effects on the brain. As Todd, would you be able to speak to the measurement tools you're using, how relevant those are? And Stuart, would you be able to speak towards whether there's known information about how AB 400 affects other tissues? I'm thinking more like the vasculature and skin and other, completely other tissues. Todd, please go ahead. Okay, great. So you were interested in understanding the translatability. So I talked about the stomach electrophysiology monitoring. You know, what we're picking up there is these pacemaker cells called the cahal cells and they actually are throughout the GI tract. So you could in principle pivot from the stomach to places like the small intestines or the colon. What's also interesting is that the inherent frequency changes in different parts of the GI tract. So in the stomach it's 0.05 hertz and small intestines it's 0.18 hertz. So in principle it's there. There are some anatomy issues that make it a bit more challenging because of how the direction the digestive system orients, but the principle is there. So that's one way this can be translated to other areas. You know, as I mentioned we're also measuring the electrophysiology, the brain's EEG doing these things in tandem. And we also demonstrated just recently the ability to pick up cervical neuronal structures indicative of things like the vagus and the sympathetic nerve. So I hope that's helpful. Does that address your question? Do you think? Stuart, would you be able to come in as well? Sure, thank you. As I mentioned, in our case, we believe that these microbial metabolites at least this class that we're investigating, since they seem to have an effect on myelin plasticity and myelination capacity, that's a pretty fundamental process. So we anticipate that these metabolites could be evaluated in other mood disorders, depression, schizophrenia, et cetera. Obviously we'd have to do all the basic science work to show that they're altered in those populations or at least that those people with those conditions are more sensitive to the certain levels of those metabolites. It doesn't mean that the AB 2004 approach, that sequestration approach is gonna work every time, but I think that there's, we're just at the tip, the front end of understanding the role of these circulating metabolites in neurological conditions. Next story, and Hubert, could you follow up briefly on the different tissues that you're studying with ultrasound under modulation? Yeah, and actually, Todd, as you're speaking, obviously ultrasound can potentially modulate different cells and different organs, but always it's kind of the targeting question. I mean, we can beam form and you gotta know that you're actually hitting the target. So I think there's opportunities where it can hit different cells and organs, but even a question back to you, Todd, is, is the approach of recording these neurophysiological signals, are they localized enough? Can you do like kind of EEG where you source localized the different organs? Because I think that's where getting to your question, I mean, is basically if we could track where we're pointing to and have a biomarker to say that we're actually modulating even the gut, I wonder if that could be something pretty exciting. So maybe that could be one option, Todd, and if you have any comments about how localized you can be on the sensing side. So it's great, you mentioned that, Hubert. So we actually have an active NIH grant on that and we already have one publication on that. So I've already demonstrated with my student, demonstrated with my student that we can perform the source localization using analogous methods to EEG. One thing that's easier is that there's not a skull there actually. That's the good news. The bad news is that the digestive organs move around a lot. And that's actually the question I have for you, Hubert, is that I saw in the GE work, they were using all those detailed imaging to stimulate and I guess that's in part because these organs are moving and I'm curious to what extent does that affect your approach given the dynamic movements that occur? Yeah, for the spleen, we've been fortunate because between the ribs, if you're sitting still, if you're not running around with the device on, but if you have the device on and you're sitting still, nearly all the movements, we've tracked in about 90 people now we've done imaging studies is mostly due to breathing and you can actually track the motion of the spleen with accelerometry for 20 bucks, right? So it actually works out well. Now, if they move, you could turn off the device but if you wanna track it more while they're moving or like the GI, then you're probably gonna have to move towards more sophisticated imaging, more to what GE can do. There's a company called Echo. They're using PESIL P-MUT technology to make a kind of more affordable wearable imaging side. So I think there's opportunities to integrate. Still costs are gonna be a bit higher but I think there's opportunities to have wearable devices that can image at high resolution and stimulate while tracking these biomarkers that you're creating. So I think there's opportunities. I think the gut test prize is a little bit challenging but the spleen, we're just fortunate that it's a large organ that's moving pretty consistently under breathing. As we have only one minute left, I was hoping that each of the speakers could share their perspective on what would be transformative technologies for their domains in the coming five to 10 years. Todd, do you wanna start off? You're doing the technologies now but what would also help you to accelerate your work? Well, when I first began working on this, I started actually working on this before the microbiome revolution took off, started working on this in 2011. And as I saw the microbiome revolution take off, I decided to just focus on the approach that we were developing and playing to my strengths. And as that matures, beginning to see the synergies and we've seen, I mean, look, there's axial therapeutics now, I mean, the microbiome stuff has evolved so much that I know there are these mechanisms of how they're linked, which I sort of alluded to at the end of my very last slide. So one of the things that I'm now intrigued about is beginning to envision taking very multimodal approaches where you're measuring the microbiome, measuring a variety of different things combined with physiologic-based approaches and identifying unique opportunities where two plus two equals five. So that's part of why I want to organize this workshop is to bring these different perspectives in. And hopefully that's something that we can all do together in the future. I'm from Hubert, but sorry for that pun. But I think for us in the space in which we operate, it's really, look are really good at collecting samples and collecting data. Now sequencing is getting very inexpensive, metabolomics are inexpensive and metabolomics is becoming much more comprehensive. It's really taking clinical metadata, metagenomics from the microbiome, metabolomics from different tissue sources. These all exist in silos today for the most part. It's really how do you bring those together and make those systems talk to each other? And we were talking a little bit about this. If we can get there, and that's gonna be a lot of algorithm development and so on to allow these different data sets to be integrated into being able to predict who's a likely candidate for a particular drug or who's at risk for a certain problem and what's their likelihood to respond. I think that to me would be transformative for the industry. Thank you, Stuart. Hubert. I could save us time and just say a double echo and then we could end there. But really, I mean, I totally agree with that. And the other piece of it is there's a lot of technologists. I'm interested in technology. I know Todd is as well. But there's so many things being developed now to non-invasive or minimally invasive modulate, the nervous system, cells in the body. The biggest challenge to be honest, a neural modulation field is kind of still a little bit like the Wild Wild West, right? Because we don't know what we're actually modulating sometimes and if we don't have those biomarkers. But I think the biomarkers are too simple and what you're saying, Stuart and Todd is like having a comprehensive view of what's going on and not treating each disease for each patient as one from the same sample, right? So I think that definitely on both ends, right? Therapeutic modulation side, but also on the sensing biomarker side. Thank you. And again, thank you Hubert and Stuart and Todd, who is also the co-organizer of this meeting. I will then conclude on that note and go on to the next session. Thank you. I just wanted to say one thing since Hubert and I are here and that's at Go Blue, Michigan second consecutive victory against Ohio State, big smiles to Kavita. I thought you're gonna have a slide on it, Todd, but yeah. Okay, we're gonna jump right into the last panel session. I have the honor of trying to synthesize everything we've done over the last two days. My name is June Acksep. I'm from E11 Bio where we're a nonprofit doing billing new tools for brain circuit mapping. And as we have learned over these last two days, there is a huge synergistic interplay between our microbial ecosystems, just as much as there is a similar interplay between all the different players in fostering innovation in the space. So we talked a lot about interdisciplinary collaboration from people from different disciplines, building the need for research tools such as next generation sequencing and mouse models in order to create the amazing scientific discoveries that many of our panelists have discussed today. Then that translates into drug discovery into medical devices. And of course, bringing clinicians to take that all the way through the clinical trials. And then coming in, of course, sending all that package to the FDA for approval to finally get it into the hands of patients. So we can see that there's many, many people involved in this process in order to take some of these discoveries and bring it to the world. So today, for this last session, we're gonna talk a little bit about this innovation ecosystem and then also envision what we need to advance this field. So we have one speaker first before we get into panels. And this is William Bambillion who is a lecturer at MIT and the senior director of special projects at MIT's Office of Digital Learning where he leads research projects on workforce education. For over a decade, he previously was director of MIT's Washington office supporting MIT's longstanding role in science policy at the national level. He has done a tremendous amount of work in bridging government policy, advancing manufacturing innovations and education, and has written many books on these topics, including innovation models like DARPA. So we will hear from him about the history and lessons learned from these different models of innovation and building an innovation ecosystem. June, thanks. I just wanna talk briefly about a few developments in the innovation ecosystem that affects some of the things we've been talking about yesterday and today. And I'm gonna need some help in moving my slides. Don't, there we go. So I wanna make kind of three points in my little 10 minute summary here. One is a little bit of background and convergence and what we're trying to do in that territory, kind of what it's about from a big picture point of view then look for some of the possibilities regarding convergence that come out of the proposed ARPA health, which is now being implemented. And also some lessons, third point, some lessons from Operation Warp Speed. So quickly, just background and convergence and as the folks in this room know that term can be applied to the kind of interdisciplinary approaches that we've been talking about for the last couple of days. Arguably it's the third kind of dramatic interdisciplinary revolution that's been going on. Arguably molecular biology brought to us by physicists like Max Del Brooke and Salvador Luria really brought physics into biology and really enabled molecular biology. That would be a first revolution. A second revolution would be around genomics. Again, very interdisciplinary. David Gallus, as many of you know really understood supercomputing at the Department of Energy and brought that into the possibility of genome sequencing. The genome project began under Watson and then Collins, Lee Hood, Craig Venter developed computerized synthesizing. Then the biotech revolution came along. Genotech was the first biotech around built around genetic engineering. Human genome project took off and with corresponding sequencing efforts. So we've got very high throughput genomes sequencing really in hours and minutes now as opposed to years that dramatically lowered cost. The molecular, the second revolution genomics really was still built around a biological model with computer science really as kind of a tool set. But I'd argue, and these are some of the leaders from some of those revolutions, Del Brooke and Luria on the left and Lee, Lira Hood and Venter and Human Genome Project Lander. Susan Hockfield who's now written about a lot of this. But the third revolution really is a combination of revolutions one and two. We're taking the methods and knowledge bases of those, incorporating a systems biology approach and adding a kind of new engineering design model and tool set plus integrating the physical sciences in this third revolution. So we bring in engineers and physical scientists not only for devices but really for the design of technology. So we're combining a biological model of complex dynamic interactive systems the traditional biological approach merging that with engineering design kind of prioritization and targeting of design that you can do in engineering with a merger of these talent bases from biology, engineering and physical sciences. So whole new fields are developed here synthetic biology, nanobio, systems biology, bioinformatics, computational biology, tissue engineering, regenerative medicine, AI now. All these fields are inherently convergence fields. So just to summarize briefly, we've got a history of other biomedical revolutions that are interdisciplinary at heart molecular biology and genomics and they have essentially built new knowledge bases but convergence, what are we trying to do? It could be different. So we get a new knowledge base plus a whole new suite of tools plus new therapies. And that's what makes this I think particularly interesting. And the new tools of course include many of the things we've been talking about today but imaging sensors, nanoscale work, simulation, modeling, big data, analytics and so on. So this convergence possibility can be accelerated and one, but it faces a whole series of barriers. An underlying issue is that NIH really is still organized around biology. It's still hard to consider engineering and physical science approaches in the context of biology. Within NIH peer review is difficult there. It's difficult for NIH to get engineering and physical science reviewers to understand these kind of multidisciplinary research proposals. And as we know, peer review itself has limits on its ability to select high risk, high reward research. There's no common language across these fields. We in effect kind of need a new kind of convergence creole when we start pouring in physical scientists and engineers into the life science benches. We don't have common interdisciplinary training. How do biologists learn these new tool sets? So we still have stove piped agencies. Interestingly, the Obama administration forced a series of really convergence collaborations. NIH had been reluctant and at NIH the bioengineering side is quite small. NCAS didn't really scale. So that administration really pushed the brain initiative, precision medicine, cancer, moonshot, all of which were organized around a convergence model and were interdisciplinary and cross agency as well. So there are barriers here that remain. ARPA-H potentially represents a new mechanism that could be an enabler for convergence research. And ARPA-H is being built, I think, in significant part on what DARPA was able to do is as everybody in this space knows, DARPA created its biological technologies office back in 2013, although it had a long history of biology related work prior to that. And that really helped foster the entry of engineering into biology. That was the approach DARPA fundamentally took. Our progress in mRNA vaccines a decade ago. DARPA jumped in a decade ago into mRNA at a time when NIH wasn't and it's funding to Moderna and other early researchers in the early days really helped advance those technologies to a point where they could be picked up when the pandemic hit. Their DARPA's goal was new vaccines within 60 days when it saw mRNA as a tool to do that. And that really laid some important foundations for our rapid COVID vaccine development in 2021. So the ARPA model is inherently a high risk, high reward approach. It's not peer review, it's strong program manager. Can ARPA-H, which is gonna pick up that DARPA model hopefully further convergence, further the whole convergence movement and an expansion of the ways that DARPA had. So there's challenges that ARPA-H faces, which we need to reckon with as we start to build it. Scale up is gonna be a challenge. And ARPA can take a technology only so far, but how does it scale? Obviously venture capital wants established pathways. DARPA has the advantage of being a connected model. It works within the Defense Department, within procurement agencies. How's ARPA-H gonna make those kind of connections? Island Bridge, another issue. You know, an ARPA entity needs a certain amount of protection from bureaucracy, but it also needs connection to decision makers who can further the scale up of technologies that are evolving from it. How's that scale up gonna work? What's the role of the NIH director? What's the role of the HHS secretary? We need to form hybrid models here. We're gonna need biotechs along with academics in kind of a portfolio approach. Building that is gonna be a challenge. And then overall, the culture of a new organization locks in very early. And you really have to get that early culture just right for it to work. So these are all some challenges ARPA health faces, but it potentially could take on a major role in convergence. And then to close, there's some lessons from Operation Warp Speed that may help us in these convergence approaches as well. So obviously Operation Warp Speed worked on technology to vaccine within an eight month period. It was a completely unique acceleration, as we all know. There were a set of tools that evolved in Warp Speed. Could we apply these to other health science challenges, including in this convergent space? What were some of the approaches that Warp Speed used? So it picked winners. Operation Warp Speed picked two leading companies in a series of four vaccine platform technology areas. It issued, and that's a risky process, right? Although the speed that it enabled turned out to be crucial. There's only so many approaches you can back if you wanna scale up. And Warp Speed had to limit the field for what it was gonna be able to do. Its critical tool was guaranteed contracts. In effect, if you came up with a viable approach, the government would guarantee a contract to purchase your vaccine. And that enabled parallel beginning and production at the same time vaccines were being developed. So you double track those two fundamental steps rather than having an undertaken in sequence. And that was enabled by this guaranteed contract approach. Technology certification. So obviously Warp Speed used FDA's emergency use approval and the advantage of an FDA certification that something works is that enables immediate market entry. That's a big advantage the life science field has over physical science fields. There's no comparable technology certification effort that's available on the kind of hard tech physical science world. The FDA rule, I mean, everybody hates FDA but we also love FDA because that approval assures markets and assures scale up. So how do we get that tool applied in the convergence space which merges physical sciences with life sciences? Flexible contracting. Extensive use of the Defense Production Act which enabled real supply chain scale up quickly for emergency needs. Also use of other transactions authority which enabled fast contracting outside of normal procurement. Operation Warp Speed mapped supply chains and really understood supply chains and how to scale them. It supported production scale up at factories. It wasn't just arms length it was deeply involved in the actual production process. Federal personnel were integrated into companies to speed regulatory compliance. The regulations were not sacrificed there was no less safety but the integration of personnel in the firms really enabled the speed up. And then obviously a national distribution effort which Warp Speed undertook to states and localities turned out to be quite important as well. The actual supply and distribution was pretty miraculous. So could some of these approaches not necessarily all but some of these maybe particularly the guaranteed contract idea which is the government will guarantee a contract if you come up with a solution. I think these are potential tools that could help enable the speed up of some of these convergence based approaches. Thank you. Let's turn to some discussion. Thank you so much, Bill. I'd like to welcome back our panelists. Mauro is on the line from Altos Labs who spoke yesterday, Hubert Lim here from University of Minnesota and Melody Zane for Cornell University, our panelists. Yeah, I think Operation Warp Speed was just amazing in how fast we can move when collaboration can come together. And I think one element of why that happened is because of the urgency. Obviously everyone was impacted by COVID and then the dire focus were multiple industries I would say globally, most industries had this push towards trying to fix this problem. And I guess one question, Bill, is what do you think how do we translate this when there are so many different diseases out there, right? Everyone is very rooting for their own disease. How can we provide enough resources, get enough attention from people to focus their energies for this kind of collaboration? Sure, I mean, obviously Warp Speed had the advantage that we were relatively close to actual vaccines in a series of platform areas. So that's one differentiator. If you want to apply some Warp Speed tools, you've really got to be in range of implementation. So that weeds out a lot of approaches. But on the other hand, this guaranteed contracts element behind Warp Speed, and this is BARDA authority that it was applying and BARDA has had this authority for some time, could be interesting. And one of the reasons why I like it is, when the government contracts for a defense technology or a space technology, they'll contract for all of the costs and expenses in developing the technology, right? Maybe it's a new rocket for NASA, right? They'll contract for the entire cost, right? And bear that cost all during the development side for a new aircraft, that's what they do. Guarantee contracts, that's a different kind of approach. That says the government pays if you've got a solution. In other words, if you come up with an approved FDA approved drug, then we will guarantee a purchase of that technology, right? The government doesn't have to bear the development risks, but it only buys, you know, something once the approval process in place. So it's potentially a much more efficient and manageable process for the government rather than a kind of complex burden of detailed contracting, that things like the defense department or NASA have to go through, it gets out of all of that. And yet it fits a venture capital model fairly neatly because venture capital is used to taking those kinds of risks. We're used to taking those kind of risks in the way in which we develop medical products. But the guaranteed contract piece enables you to kind of speed up the process because you've got a guaranteed buyer in effect a guaranteed marketplace at the end of the day. So I think that tool has some broader applicability. You know, again, you've got to make a selection based on, you know, what's within range, but the analysts, the economic analysts who have looked at Operation Warp Speed have concluded that there was a savings of at least a trillion dollars to the federal government by in effect reducing deaths in the United States by three million, right? And they translate that into at least a trillion dollars of savings of the healthcare costs the government would otherwise have to assume and other economic costs it would have to assume. So there are potentially huge savings to the government by creating a healthier population because the government bears so much of those health costs. So I think there may be ways to kind of apply some of those lessons to some really key emerging technologies that could have a big impact on societal health. That definitely changes the incentives for even VC-backable companies. Awesome. I have a question for each of the panelists because each of you are chosen to be on this panel because you guys have various types of collaborations that you do. So Mauro, as we know, you're at Altos which is a very new kind of way of collaborating scientifically. Hubert, your work as you've translated with industry now working with Second Wave and then Melody, you yourself have multiple disciplines within your lab. Can you each talk a little bit about the benefits of collaboration in your structures and then the challenges that you want to work on? Oh, sorry, we'll start with Mauro. Yeah. So we are very new. We are trying to establish this new model. And the only way that we will be able to answer that question is with facts. So all of us or many of us, we were in academia and we decided to take that opportunity to essentially try to answer questions that they are very difficult to answer and answer them in a way that we hope is the right way. Bill was discussing about the NIH funding and the system and sometimes you are precluded because you have a five years term and then you need to send your papers and so on and really deeply think about the process. Now, almost eight months in the company, I have to say that we all come with this mental setting and our brain is worried in a way that we are rewarded based on individual accolades. This is how all of us we grow up and we discussed with some of the people during dinner yesterday about this. And the goal is to leave all that aside. And the goal is to essentially, again, try to answer questions together. We are discussing the possibility of publishing papers without names or we just a list of names or random or so on and make the discovery to stand up but not the individual. Of course, discoveries are made by individuals, right? But I think if the focus is more on the discovery, I think we believe that we have a better chance to highlight which is relevant. But again, if you ask me in 10 years or in five years, I can let you know whether this is actually the right model. Now, the other point is we want to take the best of academia, the best of industry, you know, and this is a constant essentially battle of what is the best. But as I said, this is the biggest experiment in my life. I hope that it turned out well because this can create a different model, which might not be better than Genentech or might not be better than NIH or Howard Hughes. It might be a different alternative that other people might take. Experiment with more models in order to find out. Hubert? Yeah, so I have a few thoughts. Well, thank you for your talk. It was a nice summary and kind of putting in that form. So I'll watch it again. You know, one thing I found interesting about Operation Warp Speed, you had pointed out some other items, was two things on the more clinical side the human patient side. You know, there was a lot of people who, you know, because of the urgency, were volunteering for being involved with these clinical trials. Some were, I would say, high risk. There wasn't any corners cut, you know, but I have to imagine on the ethics side through review of these, there was probably a different benefit to risk assessment because of the urgency of COVID pandemic and enable things to move more smoothly forward. And it's not to say we reduce, you know, the importance of safety, but when you think of a lot of these health disorders, you know, the numbers are staggering. When you think about COVID pandemic, that was scary, but a lot of health conditions are scary too and the numbers are quite high. But when, you know, you look at local ethics and snow criticism or anything like that, it's just, it's hard to really weigh the benefits to risk from a much larger, bigger perspective of society and the risks and the things like happened with the pandemic. So I think there's some efforts there too to kind of help restructure our clinical trial system away from a liability model to an actual patient safety centric model. And I think that could speed things along. So I thought that was kind of interesting with the operation, you know, operation more sweet and vaccine. The other point, which I think is interesting that we shall consider is, I think it was great that they pushed forward, you know, those specific, like you said, to, you know, pick the winners. But one of the lessons I feel with the pandemic was also the therapeutics. You know, we were on the therapeutic side, we were trying to raise money from the government to push. And I feel, you know, I don't, maybe because of being further away from that, you know, the amount of money, you know, I don't know the numbers, but, you know, it seemed to me that there was less funds and probably because of less unknowns on the therapeutic side. And with the vaccines, the challenge has always been the variance. And so when you think about, you know, kind of the success on a long-term, you know, therapeutics becomes, you know, so it's kind of like, how do you determine what is the winners to select or where to invest the money if you're gonna go more aggressively through. So that's always a challenge that we'd have to work out. So those are my two thoughts there on that. The only thing I'll say is what you're saying tomorrow is I agree it's, it was refreshing to come out of academia and do these company things. I'm involved with two startup companies now. It was always about accolades and grants and R1s and kind of achievements. When you get out to industry, I mean, you've got investors, you've got milestones. I mean, it's kind of like growing up. And not to say I'm not grown up in academia, but academia sometimes can be a little bit comfortable. Not forgetting grants, that's always tough, but you realize very quickly that there are things that you have to achieve. You have to demonstrate at a very timely manner. And that was helpful for me to really kind of wake up and see that you got to move and you got to think practically in what you're gonna be doing and there's a lot of factors involved. So I think it's healthy to get away from the accolades. I agree with you, Maro, that basically kind of pushing that. So yeah, I echo a lot of what was said earlier. I just wanna just add to this and specifically what drove me is I was at a point in my career that if you think about big discoveries, right? I mean, discoveries that will last, let's say a couple of generations, a couple of generations. In reality, there's very few of those, right? And there's the factor of being incredibly talented but also incredibly lucky to do some of those discoveries. On the other hand, I think we are in this incredible time where we can really accelerate the movement from the bench to the clinic. And this is what Altos offered to me, offered the possibility to go for the big discoveries, go for essentially to answer those big questions but at the same time, if we really understand if we really understand the process, have the possibility to develop medicine. So, no everyone will have made the jump and not everyone made the jump. And by talking with all my colleagues, all of them, they have very different reasons. But if you were to do that and you were to have an impact in medicine and then Bill can add in addition of Genentech, maybe in 10, 15 years can add Altos as a revolution in medicine, I think that would be incredibly important just to be part of it. Well, compared to other speakers, I think my labs operating at a much smaller scale, just mostly in academia and collaborating with some other academic labs. I think for my stage right now, I just started my lab three years ago. So I think it's really important that I established my footing in academia before I really branch out and see if there might be industry partners that might be interested in taking our research to the next level. I do feel that there are definitely barriers to establish that bridge to industry partners. For example, recently we discovered some molecules from the bacteria that may moderate the coaguration response in SARS-CoV-2 infection and which was really interesting. We published the data and then I have no idea how to take it to the next step. Basically, I definitely don't want all our work to end just in the form of publications. But I've talked to the technology transfer office people while Cornell Medicine. So they basically told me you need more data in order to convince industry partners to be interested in collaborating with you. And that requires funding. So that's where we're stuck. Just my personal experience. And in terms of the multidisciplinary nature of gut microbiome research, on one hand, it's really exciting. On the other hand, it requires definitely things really beyond what I can do in my lab. I trained as immunologists. And now we're doing what we have in the field firing and we're doing single cell RNA-SIC. We're doing spacial transcriptomics. And this are really challenging for someone like me as a classic co-trained immunologist. So I am actually, for the most part, I think I'm doing a good job reaching out to a lot of collaborators. But it's definitely very challenging. And I think these platforms like today's workshop is really beneficial for me to get to know people that may have very complementary skill sets. I was just talking to Surin Ram this morning. We're excited to get to know each other and we'll talk more after today's workshop. Because we definitely have very complementary skill sets that potentially will be beneficial for both of our labs moving forward with our projects. Yeah, definitely getting people in the same room. It's very important. So we have several questions queued up, Andrew. Yeah, thanks. I think my question follows on quite nicely, perhaps with the theme of collaboration and Melody, your comments right at the end there are kind of of your own kind of academic journey. And I reflect a lot on my own as well. And I'm curious to kind of challenge our panelists. We heard a handful of our speakers over the last day or two kind of share their own journey of kind of how they reached this research in the very multi-disciplinary nature of it. And I'm curious to ask the question in thinking of how we currently progress through our traditional academic and scientific training and education pathways that it is very rote and kind of systematic. Even thinking all the way back to high school where you have one year of biology and a year of chemistry and a year of physics, then you go down a kind of deeper and deeper area of expertise and specialization. Does any panelists have kind of comments or thoughts around how we might re-envision those traditional systems to help enable not just this collaboration, but also this multi-disciplinary learning earlier to help facilitate this research moving forward faster? Oh, good. I mean, I can just add a few thoughts, Andrew. I mean, obviously there is a great barrier to these deep interdisciplinary approaches, these convergence kind of approaches because we're educated within stovepipes, right? We're educated within disciplinary stovepipes and across the disciplines means learning the whole new set of tools and learning a new language. And we really need to develop to further these interdisciplinary approaches, a training system by which, sure, take your deep dive in your discipline and that's probably very important, but we've got to create these T-shaped learners that are able to cut across disciplines as well. And really within schools, trying to develop an exchange process by which you can pick up, you know, rudiments and basics and the lingo of these different fields. Someone said that we really need a convergence creole to enable these interdisciplinary kind of discussions and to some extent that's necessary and that's gonna come out, I think, if schools and departments can collaborate on some real cross-disciplinary training to make sure that graduate students and particularly younger researchers are picking up tool sets from other fields so that they can understand how to advance their territory by taking pieces from others. Yeah, I'll add to that. You know, I'm in the Department of Biomedical Engineering and we, you know, we do our best to cover a lot of topics in one department. And, you know, you start to see kind of different shades of that. If you go too extreme, then you have people who know a little bit about everything and it causes some challenges there. If you go to the other side where it's too siloed, then you're very specific but almost too rigid, maybe too trained in a rigid way to easily navigate. So I like your Tee comment because you do need those specialties, right? You do need those people who can get into the, just the ultimate specialty and understand that. And then you need some individuals that can kind of see across fields. But ultimately, even for those who are very specific, there's just got to be a culture created where, you know, they're just reminded each way along the way, still learning in a specialty, but being nudged to kind of open up and try different things, but don't lose their specialty. And then when it comes up to be a faculty member or a team, you know, then you bring these people together, but as long as the culture's there and incentives are there to do that, I think then you can build around it. I think the problem with academia right now is incentives are, you know, it's not aligned with team science. I mean, they keep saying team science, but like you said right now for your career, if you go do that, you know, you got to fund your lab, you got to get grants, you got to get your publications. These are the metrics. How do you define a team science metric? You know, it's a tough one. How do you define a commercialization metric that you're actually helping get something out there or even a clinical translation metric? These are, you know, is it 30,000 pages of FDA documents that you wrote and then you get promotion, 50% of promotion based on that. I mean, it's not in there. I know, because I try to put my FDA documents in there, but you know, these are things that I think we have to work out if we're going to converge more on the team science, you know, direction. Any other additional comments? Otherwise we can move on. Okay, cool, Elliot. Yeah, Elliot, shake off. I just want to thank all the speakers this afternoon and our moderator for really fantastic presentations. And Bill, I really want to thank you again for stepping in on relatively short notice for really giving a brilliant presentation. It's fantastic. I have two questions. I think maybe one is for all the speakers and one probably a little bit more directed towards Bill. You know, one is again, for all the speakers who really straddled academia and industry, thought about it or living in it right now. Do we have a real gap in this science of scale-up? You know, both in terms of our educational systems, which really aren't focused on that and really in the funding mechanisms that aren't focused on that. It's really sort of left, you know, to the companies to step in that breach, but it is a substantial science and a lot of work and a lot of failures can go into that. So that's one basic question. I think the other question perhaps to Bill is there's certainly nothing like a guaranteed market that would spur industry to make a lot of investments. But do you think we have the right checks and balances to avoid a command economy and avoid sort of producing more yougos or laddas that really don't sort of fulfill the need that we really want to achieve? So perhaps maybe I'll start with the first question about this sort of, you know, a chasm of death, which is often related to the challenge of scalability. Yeah, Hubert. I have a lot to say, obviously, those things. So I'll start. I do want to give credit to NIH and DARPA and DOD. You know, they really did see this gap. DARPA, especially. So when they started, you know, Doug Weber, Justin Sanchez in the neuro space at least and Eric Van Giesen and individuals, you know, that B2O office was realizing this gap where, you know, you've got to do the science, but then, you know, you're not going to easily get investors until you get past enough data and some kind of prototype or technology. So that ultrasound project I showed you with second wave, that's actually because of DARPA. They took that high risk gamble on, you know, using ultrasound to modulate, but you need to get clinical data, you need to build a prototype. Those are all funds provided by DARPA to get us through. So now we're at a stage where we can go to investors and they created that pathway. So, you know, that's an example where I think the funding agencies are seeing this, at least in the neuro tech space. I don't know about the other, because I'm more in the neuro tech, but even NIH, you know, we have this other grant, this you mechanism where they, you know, brain implants are difficult to translate and they have this program where, you know, it's not a million dollars, you know, the grant I have is $10 million they invested in a single project and they're willing to push that forward and that project is bringing partners together. It was a large enough that the companies were willing to pair up and they could see that we can actually get it through the preclinical regulatory, you know, process and run a pilot study and they broke it up into two stages to de-risk it for them where they have a UG3 phase, which is a preclinical, get it to where it's the study and then a UH3 phase. So they had some kind of creative ways to do this, to de-risk it for themselves, but also enable us to kind of push it to the level where after the study is over, investors could come in. So whoever, you know, all the people at NIH give them credit for kind of thinking of these creative ways. So I think they are working hard at that, but still with that said, there's still a lot to, because, you know, even when you get out of that, we're still running into some of these gaps that are dictated by, you know, ultimately business model, which some of these factors, you know, unfortunately conditions that we know need to be targeted aren't really coming to the top because of the return that's available for the investors. So that part somehow, I don't know how we fixed that, but Kevin was getting to that, you know, yesterday. Mauro, do you have any thoughts now that you're in the middle of industry? So before I joined Altos, I had a big grant, a $7 million grant from Welcome Trust, this Welcome Leap program, which is actually a very interesting one, I'm sure Bill is aware of it. The review process is incredibly fast. They were looking for transformative, I think is, you know, similar to DARPA, right? The review process is extremely fast. They were looking for innovative ideas. The grants are given, so, you know, very good amount of resources they're given for free years. There is a constant reporting system, you know, it's like every two, three months, you know, they call, okay, what about the data? What about the data, right? So the, which these are disadvantaged because eventually we hope that we can predict the outcome in three months, in six months, and I will have prefer much more just to let me alone and I will let you know when I have something, right? So I think that there are models where, you know, we have the very traditional model from NIH, but we have models like DARPA and Welcome Leap and so on that they are providing sort of fresh air and the possibility to advance science, but unfortunately all of them are all short leave and if we are truly looking for that revolution, instead of having an aspirin for the brain, because if we are looking for the viagra from the brain, invariably we need a long, you know, a more long lasting system of funding. We discussed you and I earlier, what about the Bell Labs of biology? We have the Bell Labs, you know, can we do a system that now we can really push truly, truly innovation? And I think, you know, NIH and other funding agencies, maybe it's about time to sort of revisit the model where we are at and compare it to, you know, industry and how can essentially we can accelerate your progress? From the world of manufacturing, for the world of biology, oh, you're on mute, I think. Right. So, you know, Moral, I'm glad you raised Welcome Leap because, you know, as you know, it's led by two, a former director of DARPA and a deputy director of DARPA, Regina Dugan and Ken Gabriel and they have very purposely attempted to apply a DARPA model there with essentially a very DARPA-like approach. And, you know, I think an ARPA Health, as well as DARPA's own Biological Technologies Office, as well as Welcome Leap and a handful other comparable kind of organizations, they can really play a role, getting to your question, Elliot, in helping manage this kind of scale up kind of problem. And something which those organizations do, they create communities of thinkers, right? So it's not just a program manager. The program manager is very much influenced by a community that a DARPA-like organization, an ARPA-like organization helps build and it tries to bring in the very best ideas. The program manager in the end makes the decision on whether to fund a particular project but it is a very collaborative thinking community kind of effort. And that thinking community provides a certain amount of discipline to avoiding picking the Lottas, Elliot. You really have a certain amount of, I wouldn't say consensus-building because you've got strong program managers but you definitely have an idea space where ideas are being generated and shared that create a certain kind of discipline in the selection process. And then there is a, in effect, a technology road mapping effort that goes into this, right? So when DARPA or an ARPA-like entity decides to pick a technology for support and movement towards scale-up, there is a technology road mapping effort that accompanies that, that also puts some certain amount of discipline in the process. Now, look, overall on the scale-up side, the life science world is much better shape than the physical science world. And that's because venture capital, using the FDA approval and the FDA phases of approval, that enables them to do benchmarking and risk management at three different levels. So venture capital is willing to put money in in a series of tranches into a life science technology for a 10 or 15 year process. On the physical science side, with no such technology certification or benchmarking, you know, you're lucky to get a venture capital firm that's willing to take a risk of a technology that's two years away from actual implementation and production. So there's a lot of positive things going on the life science side. In the end, yes, there is definitely a scale-up problem. It's not nearly as severe, I don't think, because in the physical science side, like in the manufacturing scale-up space, but it's still there. And a DARPA-like entity, I think can help be an intermediary organization stretching between academic research, bringing in a hybrid model companies, and then helping move on a technology development, road mapping pathway towards an implementable technology, at least getting to that initial prototype. So that may be a tool set that we're gonna be able to use at more scale with the advent of ARPA-H than we've been able to do in the past. Very fascinating discussion. Unfortunately, we are at time, and please help me in thanking everyone on the panel. And with that, we turn it over to Elliot for closing remarks. Well, first I'd like to thank all of the speakers, moderators, the participants in this two-day workshop. I'd like to thank all the members of the planning committee and for Kavita Berger and Andrew Bremmer for, is that a signal? And from Andrew for their leadership, we were calling that our first really focused meeting on this workshop occurred on November 9th, so it was five weeks ago. And to be able to bring this together on such short notice and have so many spectacular presentations really reflects an incredible commitment from all of our speakers and participants. And certainly also wanna thank all the members of the Standing Committee on Biotechnology and National Security Needs for their commitment and support of this effort. I guess just to close out, maybe it's a bit of a verbal word cloud of many of the themes that came through over this last two-day period. This really important increasing understanding of sensing and signaling and the ability to modulate these pathways between the gut and the brain and the microbiota. The fundamental challenges of understanding mechanisms in order to optimize therapeutic effects. The real notion of the challenges of data integration and the need for new tools. And the ability to build bridges which ultimately will be based upon data sciences and likely machine learning and artificial intelligence to give us the convergence that we ultimately need. I think we learned a lot about the importance of food as medicine and microbes as food and that microbes are medicines. And all of this holds a tremendous amount of problems tremendous amount of opportunities for a variety of very debilitating conditions for patients and families who are in need. We learned especially during this last session about the challenges of the marketplace and of crossing this last mile. But the potential mechanisms that are being established that offer a tremendous amount of hope for accelerating these really exciting technologies at this important interface of the gut, the brain, the immune system and the microbial world in which we live. Again, I think this is a really exciting area that's on a tipping point that clearly reflects decades of work by a variety of leaders in the field who toiled away in darkness surrounded by a lot of skeptics who probably looked at this field much as how many in the community looked two decades ago at gene therapies at how we looked 15 years ago at RNA medicines and even just a decade ago at cell-based therapies for oncology. You know, it's always a myth before it's a reality. And so I don't have any doubt that a decade from now or potentially sooner, a lot of what we talked about today that we thought is just a dream and a notion will really be having real world effects in ways that may be difficult for us to believe at the moment. So again, thanks everybody for participating. I would like to turn the podium over to Andrew who would like to make a few brief remarks as well. Thanks so much, Elliot, and I will be very, very brief. I just wanna acknowledge everyone again who was involved in putting this together and first tell the speakers and participants for the really exciting discussions and exciting science that is really taking place. So first, thanks to the planning committee, as Elliot mentioned, and really wanna give quite a hand to Elliot and spearheading this effort as our chair. I think sometimes I wonder if I should go back to medical school so I can spend more time learning from you. But I also wanna acknowledge Dr. Yasmin Belke who was not able to attend last minute this week and all of her efforts and we're grateful for her and her efforts on the planning committee along with others putting together this workshop on relatively short notice. I also wanna send out a great thanks to our standing committee under which this workshop, under the auspices of which this workshop is being organized. As you can see from the next slide, it is quite a breadth of specialty and area of expertise. Very grateful for all of you and your input. And I think when our committee identified this topic over the last two days as one that we should focus our efforts around a workshop, I think the discussions over the past two days really showed just how much of value a workshop around this topic would be. So thank you all so much on the standing committee for all of your continued work. It's a really exciting group of folks to be able to interact with. Last but not least, I think Michael Fischbach used the word, some of the staff are gems and I wanna give such a shout out to all of my colleagues here at the National Academies who have helped put this workshop on first and foremost, Kanya, who's virtual, who spearheaded some efforts with the workshop planning committee, along with our fearless leader, Kavita Berger, as our board director on the board on life sciences, but just to name everyone, Crystal Saunders and Jess Moy who have been fundamental and critical to our work along with Trisha Tikolsky and Nancy Connell. I also wanna acknowledge Stephen Moss who is spending some dear time with his newborn and we miss him dearly and are grateful that he's able to be with her. And last but not least, Eric Edkin who's been such a fundamental asset to us for help facilitating our new world of hybrid meetings. Last but not least, as I mentioned yesterday, a proceedings in brief will result from the discussion. So look forward to that in the early spring and look forward for the announcement when that comes out. Besides that, have a wonderful, wonderful weekend and a warmest holiday season to you all and thank you all so much again for the active discussion. Thanks.