 First of all, I'd like to thank Lita and the organizers for inviting me here to talk to you. And we're going to talk about the microbiome and infectious and non-infectious colitis in 20 minutes and cover both fields. So hang on. But what I'm going to do is I want to summarize the field, and I'm going to summarize the field using some examples from the literature that I've chosen. And not because they are the only examples, and I know looking around the room I'm going to be talking about some people's work from people who are in the room, and I will apologize ahead of time. If I misquote your work, you can come up to me during lunch and set me straight. But to me, these are examples that helped me raise some of the gaps and challenges, because that's what we want to address. We understand this idea that we have this balance with our microbes, and when this balance is lost, something can go wrong. And in the gut, the manifestation of something going wrong is often the development of inflammation. If you look at a medical textbook, if you think about colitis or enteritis, they'll divide it into infectious and non-infectious. You go through infectious, they'll come down a whole long laundry list of the various organisms that might cause an infectious colitis. And then there are these non-infectious, which inflammatory bowel disease is one that we've already heard about, and I will talk about a little bit here. But let me start at the end. Let me bring up some of the gaps, needs, and challenges that we'll talk about. And this is kind of a wordy slide. I've bolded some things. You will see this slide several times during the talk. And we've already heard about some of these gaps, needs, and challenges. We have to move away from these surveys. Surveys are great in the beginning, because we get signals. We get an idea that there's biology there. But we have to move from those associations and those signals of potential something going on to actually finding out what the functional consequences are of any changes in the microbial community. Because if we're thinking about functional consequences, that's moving us closer to mechanism away from hypothesis generating and actually having testable hypotheses. We've already heard from Gary and others that we need to actually have model microbial communities. We can't always work in the human. Or if we work in the human, we'd like to be able to back it up with work in a model system where perhaps we can do more hypothesis testing, go back, observe in the humans and move back and forth in that translational scheme. We've already heard about the ability to handle and analyze large multi-omic data sets as an important need and a gap and a challenge for us to be able to do that. And then finally, at least as a physician, which I am, the idea is that somewhere in the back of my mind, the reason that my mom wasn't able to come visit me in my office where my nurse was able to set her in a chair, and instead I'm in the lab is that I want to get someday two novel therapeutics that will be broadly distributed. And how do we get there? And we've heard about using microbes as therapies. In some cases, there's something that we need to think about. One of the challenges is we actually have to get this microbe. We have to be able to propagate it. And even though we have a lot of great molecular techniques, the first way we studied a lot of microbes, i.e. cultivation is actually important if we're going to try to get certain types of therapies. So let's start talking about IBD. And we've heard enough about IBD, this formerly idiopathic condition that's basically characterized by uncontrolled gut inflammation. And there's been a lot of evidence that there's a role for the intestinal microbiota in the pathogenesis of IBD. And we've already heard this. There's been a lot of association between so-called dysbiosis. You take a patient with IBD. You take a matched control. You do some sort of analysis of microbiome. It could be 16S clone libraries, DGGE, pyrocytesine, pyrocytesine metagenomics, metaproteomics. They're different somehow. And I'm going to talk about two studies. And these are somewhat older studies. And when I think about my grad students and postdocs, I guess these are older studies because they were in elementary school, middle school, and high school for some of them when these came out. I remember these like, oh yeah, they just came out. But you talk to your students and they go, oh wow, those are your old papers. So we'll do Dan Frank's paper first. And he was doing a study with clone libraries. And he was kind of looking at patients with inflammatory bowel disease versus controls. And they sort of fell into two different groups. Now, interestingly enough, the controls with no colitis kind of all grouped into one group. But mixed in with that group were also patients with colitis. However, the other group that sort of came out were patients that were basically with the exception of this one individual was only patients who had colitis. And when they did looking to see who was there, characterizing it clearly, the ones in this IBD subset over here, the communities were quite different than those in the control subset. And I could probably spend the next 50 minutes. I won't. I could spend the next 50 minutes going over many, many different studies like this that show patients with IBD. Many of them are different from controls. Now, there are other ways to look at the communities. We actually haven't talked about actually, again, sort of an older thing. Tyler Jackson used to say, it's amazing what you can do in science when you look. And those are the people who go to the microscopes. And here is comparing now patients with inflammatory bowel disease versus those without colitis or those with an intermediate phenotype of colitis. And they're staining the bacteria to actually look spatially. When we actually get a biopsy, when we get a fecal specimen and we mash it up in our bead beater and we get the DNA out, we destroy any sort of spatial heterogeneity and all the information of where the bugs actually are. And so in this study from Alexander Swenzinski and his colleagues, in no colitis, you can actually get this idea that there is diversity. They stain for ones that will stain the B. fragilist group, the u-bacterium rectally group, or if you didn't stain with either of those, you'll stain with sort of a broad range bacterial probe. And you have quite a mix in no colitis. And interestingly enough, you have this sort of void zone where the mucus layer is. Contrast that with patients with Crohn's disease, active colitis. First of all, you sort of shift the community again. Once again, there's sort of a dysbiosis. The most predominant probe is this B. fragilist probe. And these organisms are forming what they termed as a biofilm, right on the surface of the epithelium. And in sort of the intermediate case, a self-limited colitis, there is some of loss of bacteria encroachment into this mucosal area. But perhaps you could have some suggestions that the community more resembles the control community. So we have these great associative studies. So now one of the first needs, one of the major needs is how, and a challenge is how do we move from these associations between disease states and something about the microbiota, say, if be it community structure or whatever, to some sort of functional consequence of these changes. Couple of recent papers came out. I'm just picking one. There's many. Here's one where the title sort of says it all. They want to connect dysbiosis, bile acid, dysmetabolism. There's a function. Gut inflammation, there's a host response in inflammatory bowel diseases, okay? So in order to go over this paper, we have to go over one of my favorite topics a little bit and that's biometabolism. So most people realize that bile is important in the digestion of lipids. And these are produced by the liver and when they're produced by humans, they're usually conjugated to glycine or torene and then excreted into the bile duct, sometimes hanging out in the gallbladder and eventually making it into the gut where they can basically solubilize, act as detergents, solubilize lipids and allowing them to be absorbed. But that's not the end of the story. So these are the so-called primary bile acids, the ones that are produced by humans. The microbiota has a lot of activity, including bile salt, hydrolyse activity and 7D hydroxylase activity that changes these primary bile acids into so-called secondary bile acids. They remove the conjugations, they remove hydroxyl groups so they create a whole different set of so-called secondary bile acids. Now, both the host and the microbiota also can do things like add or remove sulfate groups, actually the host does sulfation in some ways to quote unquote detoxify and also allow them to move out of the gut and not cause any damage. Bile acids themselves are quite active compounds, they can actually directly damage since they are detergent like, they can actually damage the epithelium. So the host has ways to try to detoxify them after they've been used for lipid metabolism. So if that's normal bile acid metabolism and in this paper was published in Gut and Gut, they have a nice box that basically you don't read anything, my students just read the gray box. What are the new findings? It's even shorter than the abstract because they bulletize it. They found that there's this metabolism of bile acids. So they took the patients with IBD and now they're looking for a functional difference. Well, gee, the bile acids are different and we'll go over what those are a little bit. And somehow this is associated, well not somehow, this is associated with changes in the microbial communities, all right? And one of these changes that they find in the bile acids besides changing the ratio of primary and secondary, they also saw increases in sulfated bile acids and they begin to speculate that this may actually change with this inflammatory loop that we see over and over in IBD. And so what do they find? Well, they found that because you have less enzymatic activity, so you're not having as much deconjugation, you're not having seven dehydroxylation, you actually have an increase in the primary bile acids and a corresponding decrease in the secondary. And they also observed this changed in the sulfated bile acids. There was an increase in the sulfate bile acids. And they brought up the fact that this actually may contribute to the inflammatory loop. They didn't talk so much about the sulfated, but there's a lot known about how secondary bile acids can actually change gut inflammatory state. And this is through signaling through a G-coupled protein receptor known as TGR5. Interestingly enough, at the same time this paper came out, there was another paper came out that caught my eye and basically they were looking at TGR5 signaling in patients with Crohn's disease. So what's interesting about this TGR5 G-coupled protein bile acid receptor is that normally when it's engaged, it actually decreases the inflammatory response. If you take macrophages that express TGR5, you treat them with an agonist, either the primary bile acids themselves or sorry, the secondary bile acids themselves or an agonist, you'll actually have decreases in TNF-alpha expression. They also took, in this particular case, after doing a bunch of in vitro studies, they took laminopropyl mononuclear cells from patients with inflammatory bowel disease. And what did they find? They actually saw that there were increases in TGR5 expression. And they saw that actually that they saw increases in TNF-alpha expression when they treated them with an agonist. Now I'm saying it wrong. I hope these authors and I actually said it backwards that the TGR5 agonist will actually increase the expression of TNF-alpha. But basically, to cut a long story short, they're showing how a change in the microbiota can cause a change in the function of that microbiota. That is changing bile acid metabolism and switching that bile acid metabolism such that it interacts with the host through a known receptor and actually increase inflammatory tones. So people are trying to do those kinds of links. Again, this is not so much the take-home message of exactly what they found, but the idea that you can link changes in the structure and function of the microbial community, have it influence the host, have it influence the entire system, which in this particular case is the gut. Gene Chang said, I'm gonna tell you more about our particular study and he lied. You will go see Marius's poster and he will tell you, so I'm passing the buck even further. He'll tell you more about our pouchitis study, but I wanna pause on this for just a second to address one more challenge. And this is sort of related to Gary's discussion and even though we're both at the University of Michigan, we weren't kind of in cahoots in bringing this up. But our particular study involved people from five different institutions. And if you look at these institutions, we have a marine biology lab, two different medical centers, a land grant university and a national lab, people with expertise in soil microbiology, ocean microbiology, infectious disease, gastroenterology, computation, technology. In the current way that we have to put together grants and get them reviewed, how do you keep and form such interdisciplinary teams to address these types of problems? These kinds of teams that we think are essential for addressing these interdisciplinary and I don't have an answer for that, but that's something to think about as a challenge. So let's in the last kind of third of the talk, let's jump a little bit to my favorite topic now, which is infectious colitis. And Ralph Frater has already been brought up in terms of the Frater and nutrient niche hypothesis. And Ralph Frater, when he was a postdoc, was already speculating in the 50s that the normal microbiota might be a factor in influencing human resistance to enteric disease. And he's talking in infectious diseases. And so let me give you an example that I always give on how this can come up. So you can have people who have a lung disease, they're one of Gary's subjects, they have chronic obstructive pulmonary disease, they have an exacerbation of their COPD, get treated with broad-spectrum antibiotics, do well from the pulmonary standpoint, but then develop severe abdominal pain, diarrhea, and hypertension. And what have they developed? They've developed an infection. They've been infected with clostridium difficile. But what does this have to do with the microbiota? That's coax postulates, you have the pathogen, it's clostridium difficile. Well, normally, with a normal microbiota, clostridium difficile, if you get exposed to it, it doesn't cause any disease. You have to alter the microbiota, for example, by giving antibiotics. Somehow you lose colonization resistance against clostridium difficile, which you generally encounter in the form of spores. These spores then germinate, bile acids are actually a good germinate, getting to my favorite topic again. The vegetative cells begin to grow, they can begin to compete against this altered microbiota. They'll produce the toxin, that's what causes disease. So we know that this happens in people. 25 years ago, when I was a med student, he says, yeah, there's something about the flora, the flora's all messed up by the antibiotics, that's what comes in. But we couldn't study it, because at that time, we didn't really have great model microbial communities with which to study these processes. And we also didn't have the technology to look at the microbiota. I was told when I was a med student, I was an MD-PhD student, don't try to study it, that's impossible to study these complex microbial communities. So just kind of go ahead and study your favorite pathogen, put it in cell culture, watch how it interacts with an individual cell, and you'll get your PhD. Well, now we do have model microbial communities in microbial systems. And remember that for 90 plus percent of the HMP, remember H stands for human. And so the four genome centers, the major studies were all on humans, and there wasn't much work that was done on model systems. And so actually, when I started working on CDIF, I thought I wanted to work in humans, but I also wanted a model system. And generally our favorite model systems are rodents. Koch's postulates were fulfilled with hamsters, but mice for a lot of reasons are good to work with. Initially it was found that only germ-free mice could be made ill with CDIF-facil. If you gave them antibiotics, clindamycin was the classic, they wouldn't get any disease. Karen Kelly in his group published one of the more recent mouse models. We've also modified the mouse models to the point that we can use a single antibiotic, give it to mice, and then challenge them with clostridium difficile, both spores and vegetative cells. And we will get disease. And the disease is relatively rapid. We can judge this by looking at the weights of the mice. Basically after treating them with antibiotics and then infecting them with CDIF-facil, animals that are gonna get sick within two to four days will drop 80% of their body weight, they'll become morbid and they'll stop eating and they'll have to be sacrificed. And instead of having a normal looking gut like this, they'll have a lot of edema, they'll have a lot of epithelial destruction. If you look closer, you see the characteristic pseudomembrane that you see in people. This is fibrin, this is blood, this is an inflammatory exudate that you see there. And so this model's pretty good at capitulating a lot about the human disease. So we now need to be able to look at the microbiota. We've published a few papers on that. But if you have a model, can you use that model for therapeutics? And as I said, when you have therapeutics, you have to go after particular members of the microbiota, and then understand how to grow these out, how to propagate them, and then how to administer them to patients. I said we did a number of surveys in this particular mouse model, and I'll summarize it by saying one of our results that we got early on that one group of organisms, the lactosporaceae, members of the Fermiqutes, you heard about some of these, were associated with protection or minimum disease in this model. You can adjust the model, so some animals get less severe disease, and also you can find certain antibiotics that can change the community without having disease. And basically, the simple-minded conclusion is lactosporaceae, good. You can also get a lot of E. coli. We've heard about the blooms of E. coli already. Gary just talked about them. And if you have a lot of E. coli in your mouse and you have a lot of E. coli in your gut, that's associated with susceptibility disease and development of severe disease. So it's great, we can do these 16S surveys. But how do you move from a 68S survey to actually a cultivated organism? Tom Schmidt, my friend and collaborator, now joining us at the University of Michigan, along with John Bresnick at MSU, published ways that you can actually use molecular probes to guide complex cultivation efforts. And I'll point you to this particular paper that's almost 10 years old now. So again, back when my grad students were on swings. But they can use 16S probes as ways to guide the cultivation efforts and say, okay, I'm going the right way. I'm varying the culture conditions such that I'm beginning to get what I want. So we did basically that. We designed probes so that we can look at lagnosporaceae and, well, E. coli was relatively easy to grow in selective media. And we got a handful of these from the mice and several E. coli's from our mice, mono-associated germ-free mice and tested on a one-on-one against C. difficile what happens? Well, what happened? If you take a germ-free mouse, and as I said, if you take a germ-free mouse, give it C. difficile, drop its weight and die within two days. Now, if you pre-colonize them with E. coli, and this E. coli which is about 10 to the 10th per gram in the gut, it doesn't protect them at all. In fact, they die even faster probably because they get endotoxymic once the seed of starts growing and damaging the gut. But this one particular strain that we had tried, this lagnosporaceae strain, some of the animals lost weight, but on the whole at the end of two days, and we've extended it much further than that, they'll survive. And when we looked at what was going on, this one organism, this lagnosporaceae strain, could decrease the levels to which C. diff would grow in the gut and it produces toxin, the toxin levels go down. If by stochastic reasons, you know, you were one of the unlucky mice that still had relatively high, but somewhat lower colonization levels in toxins, you'd actually die. And those were the three mice out of the 14 in this experiment that died. So we're gonna hear about fecal transplantation. We are gonna hear about some of these defined communities that can be used both in mice and people to try to treat it, and I won't try to steal anybody's thunder, but we are getting to the point that we can use cultivation to guide therapies, okay? I'm gonna point you to one more poster. This has already went yesterday, but she'll be back at the poster on Friday. You don't always have to even have a host. Rob Britton and his group has been using bioreactors to look at complex communities outside of the host, and they're doing great work at looking to see how antibiotics can change this and how these communities can interact with C. diff. We've already heard about multi-omic data sets and the challenges there. We've already heard from Janet about her work, and my final job is to send you to yet one more poster from my postdoc, Casey Terria, who will be presenting right after lunch today. We're basically, we're looking to see making that leap. All right, we have these organisms. What do they actually do? What do they do to the metabolism in terms of carbohydrates and bile acids? So I just want us to think about these gaps, needs, and challenges. Yes, this is specific to infectious and non-infectious colitis, but I think a lot of the things that are coming up in terms of model systems, getting away from association to function, multi-omics data sets, all of that. I think that applies to all of our work when we're looking at complex communities. Finally, I'd like to acknowledge that, this is fun to stand up here, but I'm up here speaking for a lot of people. There were six of us last night who were all the PIs of that multi-disciplinary group that I told you about. We all managed to gather in Bethesda, and we were trying to figure out how do we keep something like this going? The places that we all gotten funded before were quite diverse. We've taught each other so much about each other's science, and I think back, this is now four years down the road from the HMP when we first gathered to try to put together our project. We communicate so much better with each other. Our people who work with us understand so much more about fields that they never thought they would understand about, and I hope we can try to keep what we've started with the HMP moving forward. And of course I'd like to thank the NIH who has found it interesting, not just within NIH, NHGRI, can never keep the alphabet soup together, but many of the other institutes and centers that have been interested in microbiome research. And thanks for your attention, and you guys can get to lunch even a tiny bit early. Thank you. A couple of questions? It's okay? There was very little data for my group, so there were no questions, I'm sure. Okay. Okay, so now we're moving on to lunch and the poster session, and we come back here at 1.45, so we'll see you then.