I'm pleased to say that every year, a distinguished leader in GI and/or hepatology is invited to address us. Last night, Gary Wu spoke to us at a dinner. And today, we're honored to really have him speak to us today at Grand Rounds. 

Let me just tell you a few words about Gary. I have three minutes, Gary, so we're good, but I'm not going to fill the full three minutes, although I could talk for 20 minutes about Gary's accomplishments. Gary is a product of Ann Arbor, Michigan. He grew up there and went to the public schools there. And an outstanding student at Pioneer High School. 

He went on to Cornell University where he received a degree in chemistry. He then went on to Northwestern where he was a medical student and chose to do his residency in Minnesota, one of the premier GI training programs-- I'm sorry, residency programs-- in the country. And his next step was to come back home to Ann Arbor where he spent three years at the University of Michigan at a very special time when research was really growing. 

And a very young former football player for the University of Michigan was his mentor, Peter Traber. And Bill Kelley made a move to the University of Pennsylvania in 1993 and took the finest members of the division with him and that included people like Peter and Gary. 

And Peter's career also reached varied heights at Pennsylvania and elsewhere, but Gary's has really blossomed enormously. And Gary now, as you saw in the invitation to this lecture recognizing him as the Penn CHOP-- for those of you who don't know what CHOP stands for-- the Children's Hospital of Pennsylvania. He runs the microbiome there, but that's only one of eight other programs that Gary runs at Pennsylvania. 

He has a remarkable CV. I encourage you to take a look at it where he regularly publishes in Science and Nature as well as JCI. He's a member of all the honorific associations you can imagine. And, of course, the American Gastroenterology Association was wise enough to include him on their governing board as a basic research counselor. 

So, Gary, I'm incredibly honored to have you come speak to us today. We have a fantastic turnout. I know you taught us a lot last night. Please. Thank you very much for coming. 

[APPLAUSE] 

Thanks so much. Let me see if I can get this to work. Can you hear me? Is this on? All right? All right, great. 

Thanks so much, Jim, for that really generous introduction and for inviting me to give this presentation. I'm deeply honored here to give this lectureship in the name of James Respess. It sounded like an absolutely wonderful [INAUDIBLE] addition here at UVA. So let me see if I can get my slides up here. Where were they? 

All right. So what I thought I would do in the next 40 minutes or so is give you a broad overview of the microbiome space-- not necessarily so much focused on our own research in this area, but just to give you an understanding of the possibilities of the microbiome world. But also to make it, I think, real in terms of what we currently do in the clinical practice and some of the limitations that we actually face. These are my disclosures here. 

And so these are things that I want to talk about. I'm going to talk about the gut microbiome, something that we're actually quite interested in-- early life events, the hygiene hypothesis, and association with disease. I'm going to use asthma and obesity as examples of this. I'll talk about diet, the gut microbiota, and the small molecules that they make in health and disease. And of course, I'll talk about FMT, because that's a practical consideration, with some challenges. And then I'll end up with maybe two slides on drug metabolism and the gut microbiota. 

It's actually important to know that in its current iteration, the microbiome field is only about 20 years old, so it's relatively new. Microbiologists have been studying bacteria for over three centuries. There are a lot of really fascinating things that we see in the microbiome space, but we have a long way to go. And there are a number of significant challenges that I will highlight in my presentation here. 

So the microbiome is exceedingly complex. It's not just bacteria, but there are viruses, archaea, microeukaryotes that live on our body surfaces. And each microbiome and each body surface is actually very different from each other. The gut microbiota is particularly unique because of its biomass-- enormous biomass of organisms in the gut with high-level of complexity. It's one of the most densely populated microbial communities on planet Earth. 

So we interact with our microbiota at this 30,000 foot view of it. So the microbiota lives in the external world topologically in the lumen of the gut. In the internal world, you have the mucosal immune system. What separates us in the internal world from the external world is only one cell layer thick-- the intestinal epithelium-- as well as intestinal mucus, which is actually a very important barrier. 

And in the colon, there are two types of mucus. There's sort of like a loose mucus layer in which the microbes actually live and feed off of, but there's actually a very dense sterile layer of mucus that completely separates us from the external environment. 

Now, we produce substances like mucin and anti-microbial peptides that are secreted into the lumen of the gut and have an effect on the composition and functionality of the microbiota. In return, the microbiota interacts with us. Certain types of bacterial taxa, as well as their products, lead to an anti-inflammatory response through T regulatory cells. 

Other types of microbes lead to more pro-inflammatory response. Bacteria, in general, can program plasma cells to secrete specific immunoglobulin A's that target specific bacterial taxa, then help to fortify a barrier function at the mucosal surface. 

For the purposes of my talk, it's important to know that what you eat-- diet, antibiotics, xenobiotics-- not only shape the composition of the microbiota, but diet in particular is a substrate. It allows the microbiota to produce a whole myriad of small molecules that we're actually not able to make, because our genomic representation doesn't contain those enzymes. 

And we absorb those metabolites that can circulate widely throughout the body that can activate various types of receptors and have a significant effect on mammalian physiology. So together with these metabolites that we absorb for the microbiota, as well its ability to program our immune system, we think in some way plays an important role in the various types of diseases that we associate with industrialization. 

And this is only a partial list, but all these diseases have in common inflammation. All right? So it's an important environmental factor. So the increasing incidence of these diseases suggests that maybe something associated with time and development is actually important in the development of predisposition to disease. 

So the first thing I want to talk about a little bit is this notion of the hygiene hypothesis. And so as an infant, you're born pretty much sterile. And if you're born into a normal microbial environment, you develop the normal microbiome. And because of that, for all the reasons I showed you on a previous slide, you develop a normal immune system, which means you have normal immune tolerance. So if you develop an infection, you develop up inflammation. You clear the infection, the inflammation goes away, and you're largely healthy. 

On the other hand, if you're an infant born into an abnormal or different microbial world, then we know based on studies in industrialized nations versus individuals living in isolated agrarian cultures and tribes that our microbial world is fundamentally different now than it was a long time ago. And partly, that's because we're living in much more sanitized environments. We're using antibiotics. Our diets are different. 

Did my microphone just go off here? You can still hear me? I don't think so. No. I think it just died. Sorry. Yeah, I think the battery just went out. Let's try this. Can you hear me now? No. How about now? All right. So you can hear me now. All right. 

So you're living in an abnormal microbial environment. Now you've failed to establish the normal microbiome. And perhaps now you'll begin to develop an inflammation-autoimmunity-prone immune system that somewhere later on in life, due to another environmental hit, you now begin to respond abnormally to microbial products and autoantigens. And you develop these unrestrained inflammatory processes leading to disease like Crohn's disease, asthma, metabolic syndrome, and other types of diseases. So implicit in this model is that, well, perhaps early life exposures may predispose you to development of disease later on in life. 

Now, we do know that the early gut microbiota is actually very dynamic. There's a very specific succession by which you are colonized early on at birth with various types of organisms that increase in density and complexity out to the age of about two or three where your microbiota at that age is pretty much as complex as an adult. And it's actually very dynamic. Things change very significantly based on the types of food that you eat, whether or not you were born by c-section or vaginal birth, et cetera, et cetera. 

So I personally think that one of the best examples of environmental effects on a development of disease is atopic disease. I know this, and as a gastroenterologist-- I used this as an example last night. So when I started practicing gastroenterology almost a quarter of a century ago, we found an individual with eosinophilic esophagitis in an adult. That was like a case report. 

But now we routinely biopsy the esophagus in patients that have dysphagia on a proton pump inhibitor. Make sure they don't have eosinophilic esophagitis. So that rapid rise in incidence of that particular atopic disease is emblematic of these types of diseases that are environmentally-driven, that are profoundly been impacted with increased incidence over the past couple of decades. 

So in this particular study, and there are a number of studies that support evidence shown in this particular study, investigators studied asthma in two populations-- the Amish and the Hutterites. So the Amish and the Hutterites are genetically, actually, very similar, living in Pennsylvania. But the Amish live in a very different environment. They're living in an agrarian environment where Hutterites live in much more of an urban setting. And it turned out that asthma is much lower in incidence in the Amish than in Hutterites. 

They went into the homes of these children, of Amish and Hutterite homes, and collected house dust, and did sequencing, and realized that the microbial composition was actually very different in the house dust of these two different environments. They then did immune profiling of peripheral blood mononuclear cells and found out that the Hutterite children had a more pro-inflammatory signature in circulating lymphocytes than the Amish children. 

They took the house dust that they had actually collected from those two environments and put it into a mouse model of allergic asthma and could show that the Amish house dust actually suppressed the development of asthma in a way that the Hutterite house dust did not. So it suggests that house dust, and perhaps the microbes in that house dust, are one of the different types of environmental factors that in an Amish setting actually helps to suppress the development of an allergic response. 

Now, I use as an example of this something personal to us, and that's peanut allergy. So we have a very strong history of atopic disease in my family. My daughter was born with peanut allergies, and we did exactly the wrong thing. 

We actually kept her away from peanuts, and now she has anaphylactoid-type reactions to peanuts. She needs to carry an EpiPen around with her. What we should've done is tolerize her immune system by giving her a low dose of peanut antigens early on in life in a safe way. And then she would not have developed this severe allergy. 

So the notion is that perhaps because we're living in a much more sanitized environment, our microbial exposures are much less now. We're not tolerizing our immune system in a way that actually prevents immune-related disease. 

Now, we can actually see this in animal models. So this is work done in the laboratory of Rick Blumberg. So germ-free mice have a lot of these pro-inflammatory innate immune cells called NKT cells in the colon and the lung. And for that reason, germ-free mice are very predisposed to develop this particular immune-mediated colitis as well as asthma. 

If you take a germ-free mouse as an adult and you colonize those mice, they continue to have these high levels of iNKT cells and still have higher levels of predilection to these types of immune-mediated disease compared to conventionally-housed animal. How if you take a germ-free mouse and colonize it earlier on in life, you now suppress the development of these iNKT cells, and now these mice look like conventional mice, and they have a lower predilection to these immune-mediated diseases. 

So that educated me is that the timing of exposure to antigens in a microbiota is actually fundamentally important and predisposing you to the development or protection against disease. And again, the notion that early life events may be very important. 

So one example of this are epidemiologic associations that we see in human biology. And there are a number of associations. For example, the use of antibiotics in children before the age of two years that's associated with a risk for childhood obesity with a modest but significant increase in relative risk. And there are a number of studies that actually have sort of demonstrated this. 

So the work of Marty Blaser has suggested that, well, you can actually demonstrate this in mice. So if you take a mouse and you expose it to antibiotics, it tends to become obese. And it becomes obese because of the change in the microbiota. 

And Marty could show this, because you can take the microbiota of this obese mouse and put it into a germ-free mouse as well as the microbiota from a mouse that never received antibiotics that is lean, put that into another germ-free mouse, and it turns out that the mouse that received the microbiota from the obese mouse becomes obese relative to the mouse that never was exposed to the antibiotics remains lean. So that transferability of the phenotype based on just the transfer of microbes to reconstitute that phenotype suggests that it really is the microbiota that's responsible for this in a mouse model. 

The remarkable thing is that Marty can treat these mice with one pulse of subtherapeutic antibiotics, and that mouse forever more has an increased risk for development of obesity, suggesting that early life exposures, again, could predispose you to development of, for example, obesity. And we know that the first four months of life, the trajectory of growth, rapid growth early on in life in the first four months, is a predictor of childhood obesity, which is then a predictor of adulthood obesity. 

But I have to show you this slide. This is a slide that I have yet to see be presented in microbiome [INAUDIBLE]. So this is a very nice study by [INAUDIBLE] Ellen Blaak. And so what Ellen did was she took 57 adult males that had prediabetes, and she treated them with two antibiotics, amoxicillin and vancomycin for a week. 

And these individuals had a deep disturbance in their microbiota, decreased short-chain fatty acids, altered bile acids. It changed, like, everything that we see in mice, but it's associated with alterations in metabolic function. She then did very deep metabolic phenotyping in these individuals, and despite the massive change in all these molecules as well as change in the microbiota, there was not a single metabolic alteration that actually occurred in these adults. All right? 

So one could say, well, maybe you didn't treat them with antibiotics long enough. One could say maybe the effect size is actually very small in humans in the microbiota. But perhaps when you're an adult, the impact of the microbiota is going to be much smaller than when you're young. These are some of the challenges that we actually face in the field where we're trying to translate what we see in animal models to human physiology. And I'll give you another example of this later on. 

All right. Diet, the gut microbiota, and the metabolome-- so we live in a mutualistic relationship with our gut microbiota. We provide a lot of benefits to bacteria, and bacteria provide a lot of benefits to us as the host. And I have a whole list of things here that I'm not going to read to you. 

But as a host, and as I mentioned, mucus is a source of nutrition for the microbiota. In turn, bacteria actually provide benefit, because they actually digest and ferment undigestable carbohydrates in plant cell wall products. These are carbohydrates that we don't have the genomic structure and the enzymes to actually digest, but the microbiota can actually do that. And we think it's meaningful, because they produce small molecules like short-chain fatty acids that propose to have beneficial effects. 

So why is the microbiota so good at digesting these plant products? Well, it's because they have a 10-fold increase representation of enzymes-- for example, glycoside hydrolases compared to our genome. And a lot of these glycoside hydrolases are really focused on plant cell wall glycan digestion. But there is a representation in the genome of bacteria that can actually encode enzymes that digest animal glycans. And in the gut, animal glycans are principally mucus. 

So again, the notion that the gut microbiota digesting complex plant polysaccharides leads to production of short-chain fatty acids. And a mammalian physiology butyrate, for example, could be actually very meaningful, because they can activate certain types of receptors, like G protein coupled receptors and induce immune intolerance. They can actually change the structure of chromatin through epigenetic effects. And butyrate is a primary source of energy for the colonic epithelium, again conferring some type of barrier function advantage. 

So I showed this diagram last night to the group. I really sort of like this picture. So this is work by Jeff Gordon and Justin Sonnenburg a number of years ago. And this is an electron micrograph of the small intestinal epithelium of a germ-free mouse that's colonized with a single bacterial organism-- a bacteroides organism. 

So this the villus. This is the epithelium. When you zoom in here, this is a food particle. And when you zoom in on this food particle, you can actually see the bacteroides sitting on top of the food particle. If you zoom into this picture here, you can see bacteroides actually embedded in mucus. 

So we have bacteroides, an organism embedded in mucus. And you have bacteroides embedded on a food particle. What's it actually doing in those two different environments? 

So what Justin did was he did a transcriptomic analysis. He's looking at gene activation of glycoside hydrolases that degrade mucus, which is group one, and glycoside hydrolases that digest plant products. 

And his transcriptomic analysis genes that are turned on are red. Genes that are turned off are green. And the intermediate is black. 

So if you take these mice that are mono-associated with that one organism, and you feed it a fiber diet-- a lot of fiber-- then you can see that the genes that are associated with plant cell degradation are turned on, because they're red, right? The genes that are involved in mucus degradation are less turned on, because it's essentially black. 

On the other hand, if you feed these mice a fiber-free diet and the source of carbohydrates is now simple carbohydrates like glucose and sucrose, the microbiota, these organisms, turn off the genes that are involved in plant degradation and now turn on genes that are associated with mucus degradation. So this means that the microbes in your gut can actually sense what you're eating, and they will actually take advantage of whatever is in the environment to provide them nutrition. So if you're not eating fiber, they're going to turn to your mucus and begin to digest mucus. 

Well, Eric Martens a couple of years ago published this paper with that same notion in mind. Again, an autobiotic mouse colonized with a number of organisms. Fed a fiber-rich diet, the organisms are digesting the plant fiber, and it's largely protecting the mucus. On the other hand, these mice fed a fiber-free diet, the mucus actually gets digested by the microbes, because there is no other source of carbohydrate for them. 

And this thinner mucus is a degradation of barrier function. And these microbes now actually come in acquisition with the epithelium. These mice are now prone to inflammation, infection, metabolic syndrome. 

So it's another way to think about dietary fiber and how it may afford advantages in preserving barrier functions, not just short-chain fatty acids. It may be by turning off genes that are prone to digest mucus, so you preserve your mucus barrier. 

From the personalization standpoint, I just want to point out this particular paper, because I think you're going to be seeing more of this. Remember, the microbiome is a high-dimensional analytic feature, right? And our microbiomes are actually very individualized. Each one of us in this room has a very different microbiota. 

Well, in the Weizmann Institute, they did 24-hour glucose monitoring-- continuous glucose monitoring over a substantial number of individuals. And they realized that there was actually a lot of intercepted variability in postprandial glycemic responses. And you know that postprandial glycemic responses are a good biomarker for metabolic syndrome and cardiovascular disease. 

Well, if you take a high-dimensional analytic data set like intersubject variability in the gut microbiota, match it with intersubject variability in postprandial glycemic responses, add a couple of blood tests and some dietary questionnaires to amp up the metrics, you can actually combine all that information in a systems biology approach to come up with an algorithm that would tell an individual what type of diet you should eat to maximize reduction in your postprandial glycemic response with the notion that you could come up with a dietary predictor on an individual basis that would be useful for the treatment of type 2 diabetes. 

So this is an algorithm that was actually validated in a separate cohort, and now there's a publication that came out recently from the Mayo Clinic that basically reproduced the results. So it's a notion that you can use high-dimensional analytics to actually come up with a notion of personalized dietary interventions. 

I've seen this [INAUDIBLE] a number of times. It's not intuitively obvious that this particular individual eating this diet would maximize a reduction of postprandial glycemic response. So I think we're going to be seeing more and more of this using microbiome data to sort of stratify risk as well as different types of interventions in individuals in a personalized way. 

So let me just touch upon something as a gastroenterologist we've always actually been quite interested in. And then this is the environment in inflammatory bowel disease. So our current strategies for the treatment of inflammatory bowel disease are largely focused on us, and that's deep immunosuppression. And if you don't respond, then you unfortunately have to have surgery. 

This doesn't take into effect of the environment. Now, we know based on several lines of evidence that the environment is the primary driver for the development of inflammatory bowel disease, because the genetic contribution is relatively modest. 

So let me just give you one example of this. So this is the increasing incidence of inflammatory bowel disease associated with industrialization. This began 150 years ago in Western civilization, primarily in Europe. And now we're seeing a very rapid rise in inflammatory bowel disease in Asia, and China, and India. 

It's obviously not a genetic effect on us. This is another example of a very powerful environmental factor that is driving increased incidence of disease. And this is not just simply detection bias. It's much more than that. 

It's important to know that the environment of the gut is actually very different in inflammatory bowel disease. The gut microbiota is dysbiotic. Its structure looks a lot different than it is in health. 

Why does it look like this? Well, this is a response of a complex microbial community to an environment of stress. And in inflammatory bowel disease, that environmental stress is inflammation of our own tissues. 

So colitis, for example, causes dysbiosis, because it's an oxidative process because it produces alternative electron acceptors that facilitates anaerobic respiration. There are many reasons for why you would lead to an outgrowth of certain organisms like enterobacteriaceae in inflammatory bowel disease. 

So we cause that dysbiosis, but the real question is does this dysbiosis play a role in the pathogenesis of disease. Well, we think that it does, because in animal models, for example, these organisms that outgrow in inflammatory bowel disease are disadvantageous. They perpetuate disease in animal models. 

And we even have some evidence in humans that this might be the case. So fecal transplantation-- taking feces from a healthy individual that's non-dysbiotic, inoculating individuals with ulcerative colitis-- in two pretty good placebo-controlled studies shows that it can actually provide a benefit leading to an increased remission in patients with ulcerative colitis. 

Again, this is human data, so suggesting that deeply changing the microbiota may be of benefit. And that's sort of the paradigm. If diseases associated with dysbiosis try to alter that dysbiosis in one of many different ways, to change the configuration of the microbiota through diet, for example, change the environment of the gut, that may promote health. 

We've been actually interested in this area for some period of time. And there is actually a diet that actually works in the treatment of Crohn's disease. These are defined formula diets. Or elemental diets are the more simplest form, but you can use polymeric diets too. 

And most of them are not palatable, but there are a few that you can actually drink. And they are actually very good at inducing remission in patients with inflammatory bowel disease, but they're difficult to use, which is why they're not a standard of care. 

The problem is we don't actually understand how these defined formula diets actually work. So our paradigm is shown on this slide. So we know that diet is epidemiologic associated with IBD. The gut microbiota, for many reasons, we think, is a driver for the development of IBD. And I showed you some information-- I'll show you a little bit more-- about the impact of diet on the microbiota. 

So our strategy was to use a diet that we already know works, exclusive enteral nutrition, as a therapeutic probe, because then we can figure out how this diet works. We could come up with better diets for our patients with inflammatory bowel disease. And that's, like, number one question the patients with IBD actually ask us as physicians-- what type of diet should I eat? 

So as a physician scientist, I believe that astute observations by clinicians are enormously valuable in helping us to think about, in the scientific world, hypotheses that are worth pursuing. And the clinical observation of the use of these enteral nutrition diets is that they work the best for Crohn's disease when they're consumed completely in place of a normal diet. That's why we call it exclusive enteral nutrition. If you drink this liquid diet but you start to eat whole food, it doesn't work as well, right? 

So implicit in this clinical observation, and this is really important information, because this is-- the human response are two fundamental questions. Does EEN provide something good for patients with IBD that's not abundant in the regular diet or does the consumption of EEN exclude something that's bad for patients with IBD in the regular diet? 

Well, there is a notion that our diets are now fundamentally different over the past century than they were. And the notion that ultra-processed foods could be associated with, in this particular paper, increased mortality about individuals in France. It's a growing concern that food is one of the many environmental factors that could be different that could be playing a role in various types of disease states that we associate with industrialization. All right? 

There is evidence now in animal models that these food additives-- artificial sweeteners, dietary emulsifiers, can change the mouse microbiota that alters functionality and leads to an inflammatory response and metabolic syndrome in mice. Dietary emulsifiers, just so that you know, are things like carboxymethylcellulose, polysorbate 80, guar gum, soy lecithin. These names are actually familiar to you, because if you look at any processed food product that has a wrapper on it or an expiration date and you look at the labeling, it will contain dietary emulsifiers in it. 

So again, I'm not saying artificial sweeteners and dietary emulsifiers cause IBD. What I am saying is that, again, our food supply is different now. And it could be an environmental factor driving the increasing incidence of these different types of diseases we see with industrialization. The problem is that this is all in a mouse model. We don't know the relevance to human physiology. 

So last night, I showed the group the fact that we actually have a grant where we are trying to prove cause and effect relationships in humans. So we take healthy human subjects, bring them into the PENN CTRC, and feed them either a placebo or diet that contains high doses of carboxymethylcellulose. 

I won't have time to actually tell you about that study here today, but what I will tell you is that we don't think we're going to cause obesity and diabetes in an 11-day inpatient eating study. But we are collecting a lot of biospecimens to determine can we see biomarkers that Andrew actually saw in mice, either in stool or in plasma, and reproduce these in human physiology. If we're able to reproduce those biomarkers, now we have intermediate biomarkers that would allow us to do much larger outpatient cohort studies on risk exposure with dietary emulsifiers. 

Alternatively, it's possible that we're not able to reproduce anything that Andrew actually found in mice. That's actually also very impactful, because this suggests that this is a very interesting mouse physiology study with less relevance to human physiology. 

Either way, I think the results will be very interesting. We just completed the study. Samples have been sent all over the world. And we're waiting on the analysis. So we should probably have more information in a couple of months. 

All right. The notion that diet not only changes the structure of the microbiota but leads to the production of small molecules that then can have an effect on us as the host. This is a whole big story, but I just want to show you as proof of concept this one particular slide. 

So this is dietary amino acids that can be converted by the microbiota into a myriad of small molecules-- aromatic compounds that can do all these things. They're antibiotics, neurotransmitters. They can activate nuclear hormone receptors, G protein-coupled receptors. 

I'm old enough that there was an entire field called orphan receptor biology. They're called orphan receptors-- orphan nuclear hormone receptors, for example-- because we didn't know what the endogenous ligands for these receptors actually were. And so, for example, FXR-- Farnesoid X receptor-- the endogenous ligand for that are actually bile acids. 

Well, the notion is that some of the endogenous ligands for these receptors may not be post derived. It could actually be from the gut microbiota. 

Now, thinking about amino acids-- for example, phenylalanine, tyrosine, tryptophan-- these are aromatic amino acids. Chemists like Michael Fischbach have been working out the biosynthetic pathways by which these-- by which microbes actually convert these dietary compounds, amino acids, into various types of small molecules that we cannot produce-- for example, indolepropionic acid here. 

So these small molecules, as I mentioned, have a deep effect on human physiology. So just focusing on tryptophan, particular clustering organism, needs the production of indolepropionic acid, a product that we don't make. An indolepropionic acid is a ligand that activates the nuclear hormone receptor called PXR. 

PXR activation fortifies epithelial barrier function. And maybe that's important in inflammatory bowel disease because PXR is a genetic locus associated with development of inflammatory bowel disease, OK? It's early days right now, but what I have to say is that the metabolites that are induced by the microbiota has a rich source of bioactive molecules that we can use for discovery. 

So as I mentioned last night, when I first started at Penn, I actually went down to an oceanographic institute down in Florida, and they were taking deep sea submersibles to the bottoms of the world's oceans, and collecting sea sponges, and extracting bioactive molecules. 

These bioactive molecules-- organisms spend a lot of energy making these bioactive molecules, because they are bioactive. They have important consequences in biology. 

Well, the notion is that the microbiota is a rich source of a lot of these molecules, most of which we don't actually understand, have never been characterized. When we do mass spectrometry in stool, we can annotate by libraries maybe about 1,000 or 1,500 small molecules, but we see 50,000 spectral features in stool. It's a very rich source for small molecule discovery, and I think we're going to be seeing a lot more of that in the future. 

So let me talk a little bit about FMT, because this is actually relevant from a clinical standpoint. You know that. C. difficile is a big, big problem increasing in incidence. And you know that FMT is actually highly effective in the treatment of refractory Clostridioides difficile infection. 

From a personalized medicine standpoint, there are several reasons by which the composition of the microbiota is important in protecting or leading to predisposition for a C. difficile infection. This is one, and it involves bile acids. 

So primary bile acids in your small intestine is actually a germinate for C. difficile. But the conversion from primary to secondary bile acids by Clostridio species leads to secondary bile acids like deoxycholic acid. And secondary bile acids actually kill C. difficile, right? 

So if you take an antibiotic that you kill off these organisms, you no longer convert as well primary to secondary bile acid. So now you germinate the spores for C. difficile in your small intestine if you don't kill them off in the colon. And now you're at higher risk for developing C. difficile. 

So it's a notion of personalized medicine. Maybe you stratify for the risk for C. difficile infection by looking at these organisms that can make a conversion between primary and secondary bile acid. Or you should develop antibiotics that sort of preserve these organisms. Or maybe you could reintroduce these organisms as a protective way to prevent the recurrence of C. difficile. 

Now, there's enormous scientific value in looking at observations in FMT, because, again, this is human biology. And it's proof of concept that maybe you need to use a very robust and resilient microbial community to change that this dysbiotic microbiota for the treatment of disease. 

At the end of the day, this is human data, which is fundamentally important, and more of what we actually need in the field. And with that notion that it's so effective in the treatment of C. diff, there are a lot of clinical studies looking at many different types of diseases. This is a little bit old, but there are now over 250 clinical studies ongoing in the United States of many different types of diseases. 

Now, you know, I would love to believe that it could cure all these different types of diseases. I'm skeptical about a lot of these things, because C. difficile is an infectious disease process. A lot of these are sort of chronic diseases, but we will learn a lot if these are done in a well-controlled setting. 

But with the notion that it's so effective, there is a lot of interest in the lay public about fecal transplantation. This is just a YouTube video of a person that basically teaches you how to do your own FMT, right? This is the notion that the lay public will do bad things to themselves with basically a misunderstanding of the field. 

So I feel very strongly that as physicians and scientists in the field, it is our responsibility to try to educate people about the potential value of things like FMT, but potentially the risk, because people will do these unfortunate things to themselves. So that's the notion that just because you can do it, you shouldn't always do it. 

And there is actually a need for regulation in this space. And I use as an example the fact that even though we can quantify short-term infectious risk of FTM and it looks largely safe at this point, we don't know long-term consequences. Recall that we actually infected an entire generation of individuals with hepatitis C and HIV by blood transfusion before we actually recognized them as pathogens. 

And, you know, the hemophilia community always felt betrayed by the individuals that were giving blood transfusions even though there was no way that you would know that you're transmitting HIV. You have to keep in mind that the gut microbiota is incredibly complex with many different types of microbes, most of which we don't understand. 

When we do sequencing and we look at viruses, prokaryotic viruses in particular, we can only annotate about 10% 20% of those viruses. 80% of those are not in databases, have never been previously described. This is a dynamic living consortium that can change in ways that we currently can't predict. 

And there is certainly abundant animal model. I've already shown you a little bit. This suggests the microbiota could be important in disease pathogenesis. And it is for this reason that it's a very dynamic time right now, because we have clinical practice, but then there is possibly a need to regulate this space. 

But as you know, as you're all physicians, being a physician is all about being practical. If you have an intervention that will help a patient that has a disease that's suffering from something that is seemingly safe that has hypothetical risk, as long as you have informed consent, it is actually very reasonable to do that. And for that reason, FMT in recurrent Clostridium difficile infection is allowed in the United States without an IND, but for any other reason, you need to have IND. 

I won't get into this right now, but this ended up in The New York Times. This is a very difficult time when you think about regulation and the interaction with product development in the field, which is what the FDA is really there for. So it's a very dynamic field right now in terms of regulation and clinical practice. People are very passionate about this topic, because it directly affects their patients. I don't have time to get into that right now, but you'll be seeing a lot more of this. 

So let me just show you this study-- the impact of FMT. So this is by Peer Bork. He was looking about two or three months after FMT in five separate individuals. And here are all the different bacterial taxa. In all the bacterial taxa that are from the donor that are now ingrafted into the recipient, after three months are orange. 

So what he's essentially showing is that 30% to 50% of the individual's microbiota as recipient is now the donor's. So now if you're going to ingraft a donor's microbiota essentially at high levels into a recipient. 

Now remember, you've grown up with your own microbiota. It remains relatively stable throughout your life. But now you have to wholesale switch somebody else's microbiota into yours. Now, maybe there is no danger to this at all, but these are the types of things that give us pause. Like, we need to be careful. 

And my whole notion is that, if we didn't do our very best to try to safeguard our patients that are getting FMT. And part of this is that the AGA, for example, has a national FMT. Registry, try to collect as much information about specimens as possible. If we didn't do that, we would always be criticized. 10 years from now, if there were an outbreak, an infectious disease, or immune [INAUDIBLE] disease, that we didn't really spend the time thinking about using our best science to try to understand what we're actually doing to our patients during a period of time where there is really no regulation in this space, right? 

I hope 10 years from now, people would say, Gary you're were just a worry wart. There was nothing to be concerned about. That would be the best outcome, but we would be always criticized if we didn't try to save our patients. So thinking about the scientific evolution of the field, or you agree with FMT, it's a notion of full spectrum process, people products, and ultimately to find microbial consortia. Specific organisms grown in a laboratory assembled more as a drug. 

So for cluster to difficile infection, FMT works. We're hopeful that these other modalities will work for other types of diseases. I give you a little bit of evidence, maybe that's a little early signal an inflammatory bowel disease. But it is really the focus of technology development in various companies and in academia. Think about the best biology, the best science, to figure out how you can put together the right organisms to treat and or prevent disease. This Is like the next generation of probiotics. Not the ones that we're currently using, the next generation of probiotics. And the notion is that as you more better-- if you better define what you're actually putting into individuals in a more standardized way that it's actually going to be safe. 

So for initiation, we're beginning with stool. But hopefully, sustainability in the next couple of years we'll be moving on to additional products. And the final thing I'm just going to show in two or three slides is drug metabolism. So there's a lot of interest in the gut microbiome and liver disease because you know, through portal circulation, that everything you eat and microbiota products go to the liver first. 74% of blood actually is supplied by circulation direct to the liver. And, so, it's a lot of interest in many different types of liver diseases associated with the microbiota. We're actually just starting a liver disease program and a microbiota space with this notion in mind. 

And there are associations with the composition of the microbiota that predict the presence of liver cirrhosis. There are even studies here looking at NAFLD and NASH, predictive models based on the structure of the microbiota, as well as the metabolites, but also drug metabolism. So xenobiotics drugs can actually change and shape the physiology of the microbiome. You may know that Metformin, treatment for type 2 diabetes, has a very distinct effect on the microbiota. In fact, there was a paper in Nature that showed that type 2 diabetes had a very specific signature in the microbiome, while a subsequent paper that was published shortly thereafter showed that, no, it wasn't diabetes. It was the use of Metformin that was actually responsible for that. 

But it's not only that Xenobiotics can change this microbiota, but the microbiota can actually change functionality of drugs. So this is work done at the University of Pennsylvania, that the microbiota actually has beta-glucaronidases that will cleave off glucoronides. So you actually, by a phase II metabolism in the liver, you tag xenobiotics with sugars like glucaronide, and that tags that xenobiotic for expulsion, either in bio or through the kidneys. But if you cleave off that glucaronide, that drug continues to circulate. 

So bacteria actually have glucaronidases. There are six different types of glucaronidases that are now being targeted, because if you cleave off that glucaronide, you could can have drug toxicity, because that drug from a cut pharmacokinetics standpoint exists at higher levels now. And there's an association, at least in animal models, with NSAIDS recirculating causing more ulcerations because of the gut microbiota. And I think what I want to show you here is acetaminophen metabolism. So acetaminophen could neither be sulfated for metabolism or can be glucuronidated. 

So tyrosine, again, and one of those aromatic amino acids, can be converted by the microbiota to a product called p-Cresol. And p-Cresol is also sulfated, so in those individuals that have a certain type of microbiota that make a lot of p-Cresol, this will swap out the sulfation mechanism in the liver, and then you switch your acetaminophen metabolism from sulfation into glucuronidation. 

So their [INAUDIBLE] microbiota can have a big impact, potentially, on the way that you actually metabolize drugs. So this is my last slide. You know, it's like a sort of a vision that I've sort of promoted with national associations and I've promoted this at Penn, at both university of Pennsylvania and Children's Hospital. It's sort of the paradigm for integrative multidisciplinary research, thinking about the intersection between basic science and clinical research. Because everything that I just talked to you about in the last 40 minutes or so is really multidimensional, multidisciplinary team research. Trying to integrate basic science research with clinical research. 

And I think that as a department of medicine and as a clinical department in GI, this is one of our greatest strengths. So in the basic science world, we do mechanistic studies and animal model systems and we're trying to phenocopy what we actually see in a clinical world, that we actually describe by epidemiology by clinical phenotyping. 

But the way that we do science, now, is actually very different than it was two decades ago. Because we have these technologies-- genomics, metabolomics, [INAUDIBLE] immune profiling, transcriptomics, combined with computational algorithms, that allow us to think much more systems approach to understand the complexity by which mammalian systems actually respond to various types of interventions. 

You can do this in animal models, but I say that you can actually do this in humans if you collect the clinical metadata, you collect human biospecimens. And when you integrate these two things together, you think about doing human subject research. And there are enormous opportunities for precision medicine, new therapeutics, in patients suffering from disease. 

What do you need to do this type of work when you integrate these two, where you're trying to understand mechanism here, and you're trying to determine human relevance here? Well, unique clinical patient population. You need to phenotype these integrations, these patients, and maybe use your EMR as a database for this. You collect human biospecimens, you have these [INAUDIBLE] technologies, and you have computational statistical analysis. And I firmly believe that this is an opportunity in the future. So emblematic of that is my acknowledgment slide. 

So I'm one of a very large team of researchers. Multidisciplinary team of PhDs, MB PhDs, and MDs in different areas in the clinical domain, and the analytic domain. We all have to work together as a team, because nobody could do this type of research that I've been showing you over the past 40 minutes as an individual. It actually takes a team to actually do this. So with this, again, I'm deeply honored thank you very much for inviting me to give this lecture. I'm happy to answer any questions. 

Gary, thanks. My predecessors had the great wisdom to offer an incredible token of our appreciation. Everyone who comes to give the rest of this lecture has a permanent memory of Thomas Jefferson's university, which stands permanently and print prominently in our GI library. And we will be mailing you a copy of this, Gary. And you will hopefully prominently displayed in your office in Pennsylvania. Thanks very much. 

Thank you. 

We have time for a few more questions. 

So you had mentioned the mouse model, showing that giving lean mice the [INAUDIBLE] mice [INAUDIBLE] has shown the opposite. [INAUDIBLE] 

You know, it's actually very interesting. I'm not seeing a lot of evidence for that, to be honest with you. You know, germ free mice tend to be more lean. Part of that is just short chain fatty acids because they're a fermentative organism. So we, actually, have an interest in our own laboratory about the reversibility of these phenotypes. And some of the phenotypes are not actually reversible. 

So one of the limitations, and I'm going to be quite honest with you about these transfer experiments, germ-free mice are physiologically and developmentally very different from colonized mice, right? So some of this, I do wonder whether or not there is a developmental side of this that it's unique to germ free mice. Most of these good states, they have control. So the feces from somebody that has a different phenotype, as I showed you, they don't develop that phenotype. All right. 

But you're absolutely right. The reversibility of this and the developmental aspects, it's something that we continue to argue about. What's the right animal models to be doing this? And the argument against germ free mice is they are developmentally, fundamentally different from colonized mice. 

[INAUDIBLE] 

Again, that's exactly my point that, right now, there's a practical need for these. These are patients that are suffering. Some die, some go to surgery, right? And if you had a intervention that-- it's a fairly sizable body of literature, right now, that the short term risks are relatively modest. Most of these risks that I talk about are hypothetical. Some based on animal models, so it's a practicality that you have an intervention that you know right now will work. But we do hope that the field moves on beyond just using stool. You know, we do screen our patients a lot looking for blood borne transmission of disease and things like that, that could be transmitted by mammalian components, that we're actually transmitting in stool, too. 

So we do the best that we can, but it's the unknown that keeps us-- that we should be constantly surveying and being careful about in the future. Rick? 

Sure. [INAUDIBLE] think an antibiotic, we tend to think about dose-duration forgiveness organism. Have we characterized the general effects of classes of antibiotics? 

Yeah. No, we have not done that in a systematic way. And I think that's an excellent study that should be done. There are certain antibiotics that are deeply disruptive to the microbiota. Like vancomycin is deeply disruptive to the microbiota. And there are others that are not. We actually published a study where we use, I think, it was Metronidazole, Ciprofloxacin, Trimethoprim-sulfamethoxazole. Treated patients for like five days, and we sequenced, and we saw very little change. 

So it's actually a very important thing. Now, there's a whole field called the [INAUDIBLE]. It's an antibiotic resistance. And there are investigators-- Gautam Dantas at WashU is a big leader in the field on this. Antibiotic resistance is huge, so it brings their point that we should actually understand more about the impact of specific antibiotics on our microbiota. But it's not been studied as systematically as we probably should have. 

[INAUDIBLE] what's absolutely necessary. So years ago, [INAUDIBLE], why don't we go to octologists? So I'm perfectly healthy now. We'd need [INAUDIBLE], but the collection is so much simpler than collecting [INAUDIBLE]. It's my hope [INAUDIBLE] that one donor could [INAUDIBLE]. And so most people don't need four FMTs to recover after [INAUDIBLE]. But I got some poop in my freezer. I'm telling you. 

[LAUGHTER] 

Yeah. 

It seems crazy, but there's no risk of bacterial exposure against your own microbiome. 

So the group at Memorial Sloan Kettering is actually doing this. I talked to them years ago when they were trying to get an IAD to actually do this. And their interest is actually in Graft Versus Host Disease, where they have a notion that [INAUDIBLE] transplantation, after devastation, the use of antibiotics, and pre-conditioning, could actually reduce Graft Versus Host Disease. 

So the thing that they were actually doing was [INAUDIBLE] FMT. So before you get preconditioned for a BMT, you donate a stool sample. You get the stool sample back, right? And so no, I think it's a very interesting notion how you do that, you know, in the large populations and beyond because not everybody develops C. diff, right? So is that cost effective or not, right? But people are doing those types of things. 

Thank you again. This is fantastic. [INAUDIBLE] Can you comment on, especially, the Amish dust when you think more about the location of the microbiome? And the [INAUDIBLE] gets respiratory organisms that are in the wrong places that can [INAUDIBLE] as a whole complex set of questions. Can you comment on that? 

Yeah. You know, I mean, I think we're just scratching the surface. You know, one of the reasons that people have a lot of interest in microbiome spaces, for the very first time we have technologies to actually describe our environment with some greater level of precision, right? But what you're talking about is, for example, the expose of the-- collection of all types of things that we're exposed to. And so we're just sticking with bacteria. What's the entrance into the body? Is it through aerosolization in the lung, is it through your skin, is it through the ways that you eat, right? 

All of these things could have an impact. I only focused on the gut because, well, I'm a gastroenterologist. Most of the products, actually, focused more on the gut microbiota. But, there are people that are actually very interested in atopic disease and the skin microbiota. So different portals of entry could have different physiologic functions. And you know again, it's just very early days, right? 

[INAUDIBLE] 

You know, computational biologists, a lot of people in the microbiome field, have been thinking about that for a long time. And there is not-- it's not like a blood test where you can say, you know, your two standard deviations away from the mean. Right now, there is not a notion for what is the prototypical healthy microbiota because it could depend on where you're actually living. There are general concepts. The general concepts are the richer the microbiota, the better the microbiota. It's like a lawn with a lot of weeds in it, right? If you don't have a really rich lawn, you can get pathogens that actually grow in it. So most diseases are associated with a decreased richness, a decrease in the different types of microbes in the environment. 

In general, proteobacteria, like enterobacteriaceae, E. coli, Klebsiella, those types of organisms tend to be more associated with disease. But there's really no standard definition of what a healthy microbiota actually is at this point. I think it's really sort of a challenging way to come up with a standard notion of that. All right. Thank you very much.