Dr. Rob Knight, co-founder of American Gut, Discusses the microbiota and microbiota testing. This was a fascinating discussion about the current state of microbiota research. We cover what the newest breakthroughs are and what place your American Gut microbiota results have related to gut interventions. There are a few gems here you won’t want to miss.
If you have questions about microbiota testing, click here.
Dr. Rob Knight Bio….. 0:58
Microbiota Defined….. 1:57
Heart Disease and the Microbiota….. 5:29
Obesity and the Microbiota….. 7:40
Fecal Transplants….. 14:35
Microbiota Assays and Disease Risk Prediction….. 20:52
Small Intestine Assays….. 27:52
Episode Wrap-up….. 30:50
Right click on link and ‘Save As’
Dr. Rob Knight, co-founder of American Gut, Discusses the Microbiota and Microbiota testing
Dr. Michael Ruscio: Hey, everyone. Welcome to Dr. Ruscio Radio. This is Dr. Ruscio. I am here with Dr. Rob Knight who is really a leader in microbiota research. And I’m super excited for this conversation. So, Rob, thanks for being on the show today.
Rob Knight: Sure. Thanks so much for having me.
Dr. Rob Knight Bio
DrMR: Can you tell people just briefly about your background and how you’re active in this field?
RK: Oh, yeah, sure. So currently, I’m a professor of pediatrics and computer science and engineering at UC San Diego where I’ve been for a year and a half. Before that, I was at the University of Colorado at Boulder for about 15 years as a post doc and as a faculty member in the chemistry and biochemistry and computer science.
So my formal training is in biochemistry and evolutionary biology. Essentially analysis of DNA sequences and molecules sent me on a path that has led to developing some tools that let us read out the microbiome much faster and much more efficiently that it has been possible before.
DrMR: Which is such an amazing field right now that’s just booming in terms of the importance of the microbiota. And I’m sure most people listening have heard of the microbiota.
DrMR: But just briefly, the microbiota, to put it very simply, is just the world of bacteria that live on or in us—on our skin, in our digestive system, in our urinary tracts, and so forth.
And we’ve been learning an immense about the impact and the co-relationship that we have with these bacteria and these bacteria have with us. And Rob’s work has really helped bring us way forward in our understanding and our capabilities with understanding how this world of bacteria impacts us and how we have this important delicate balance, or this delicate relationship.
So, Rob, I think people get the basic concept that bacteria are important. And they can either help you or hurt you. But what might be some new things that are exciting that you’re seeing, someone being so involved in this field, that either a practitioner or a lay person might want to be aware of?
RK: Well, one thing that’s exciting is that it’s not just the bacteria. So there are fungi. There are other single-celled organisms including microbial eukaryotes and including the archaea. And there are viruses that are arguably part of the microbiota as well. And so one thing that’s very interesting is looking at the links between those different compartments, those different kinds of organisms and how they make up the complex ecosystems in our bodies.
Another thing that’s really interesting is long range connections. So what starts off in the gut doesn’t stay in the gut. And we’re finding out that the microbes in our gut have effects on our liver, on our joints—so for example, rheumatoid arthritis—and even on our brains. So it’s fascinating seeing how you can have these long range effects even from a microbe that’s primarily active in the gut through chemicals it releases in the bloodstream or through signaling to the vagus nerve and other mechanisms.
And another thing that’s just been fascinating is linking microbes to all kinds of diseases that we had no idea that they were involved in. So you probably expected that your gut microbes were involved in things like irritable bowel syndrome and inflammatory bowel disease. But you might be surprised to learn that they also contribute to things like atherosclerosis and even, at least in mouse models, to things like autism and depression.
DrMR: And it is really amazing the—both of those components, certainly the far-reaching effects of the gut. And I think that’s something that many practitioners that have helped patients or their clients improve their gut health have often observed, “Boy, when we help someone’s IBS, their skin clears up. And their brain fog goes away.”
And it’s just remarkable how there’s such an impact on the gut to the rest of the body. And it’s neat to see the science kind of catching up with that to help validate what I think many of us have seen clinically for awhile.
RK: Yeah, absolutely. And especially with IBS and with celiac, there’s been such a long-known comorbidity with major depression. And being able to explain some of that in terms of particular bacterial and metabolic interactions I think is very exciting.
DrMR: Absolutely, yeah. It’s such a fascinating field. Now, you mentioned other diseases. And there’s a number I’d like to kind of touch on to get your perspective and your thoughts.
Heart Disease and the Microbiota
DrMR: Heart disease I think is really interesting. And some of the papers I’ve seen have suggested that heart disease may predominantly have an autoimmune component. And a lot of the bacteria seem to affect the immune system, especially if we look at early life microbiota colonization and how that impacts things like allergy, eczema, atopic dermatitis.
I think those are the best studied where we know that anything that negatively affects the gut bacteria—using the term bacteria loosely because, as you said earlier, it’s more than bacteria. There’s fungus and viruses and protozoa and what have you.
But things that negatively perturb the microbiota early in life seem to really increase the risk for some of these autoimmune or pseudo autoimmune conditions like allergy, eczema, asthma, atopic dermatitis later in life.
Do you think heart disease might fall into that immune category? Or is it different? What are your thoughts on that?
RK: Yeah. Well, heart disease definitely has an inflammatory component, although most of the research, like from Stan Hazen’s lab at the Cleveland Clinic, linking gut bacteria to heart disease has been focused on a chemical called TMA or trimethylamine which gets metabolized in your liver to TMAO or trimethylamine oxide. And in mouse models, it’s very clear that that’s what the pathway is contributing to atherosclerosis.
In human populations, it’s been a bit less clear. So there have been so promising studies from Stan’s lab suggesting that the same principles can apply in human populations. But one thing that we’re fighting is that for many disease, even when microbes are involved in that disease in different populations, it might be different microbes in different places. And we’re still starting to understand that.
Say, for example, for Crohn’s disease, the signatures are very consistent, one population to another. But for obesity, although we can often tell whether you’re lean or obese with very high accuracy based on your microbes, it seems to be different microbes that matter in different parts of the world.
DrMR: And I definitely want to circle back to the Crohn’s and inflammatory bowel disease piece because I believe you had published a pretty noteworthy paper on that. And I definitely wanted to get some commentary on that in a moment.
Obesity and the Microbiota
DrMR: But you also mentioned obesity which I find really interesting. And one of the things that I have observed as I have been going through the literature on this is we have to be really, really careful with extrapolating too much, of course, from animal data because it seemed like the early animal data pretty strongly supported the altered Firmicutes/Bacteroidetes ratio in obesity. But that didn’t seem to really be borne out consistently in the human studies. What are your thoughts on that?
RK: Yeah, that’s correct. And so the rodent data continues to support very strongly phylum level changes and usually the same phyla being involved with maybe some contribution from Proteobacteria as well which are also pro-inflammatory.
In different human populations, we see different things basically. So the two things that are fairly constant are that you can discriminate lean from obese with high accuracy, frequently as high as 90% based on the gut bacteria.
And you might be thinking, “Well, what about the other 10%?” But if you try to do that based on human genes, if you use every SNP associated with obesity by genome-wide association studies, you only get to about 60% classifier accuracy. So 60% ability to tell lean from obese based on human genes versus 90% based on bacterial genes. So that’s not so bad really.
RK: But it does seem to be different bacteria in different populations. Now, the other thing that’s fascinating is you can actually confer obesity on germ-free mice by transplanting bacteria—well, transplanting bacteria I say specifically because you can isolate the strains that do it. But basically by transferring bacteria from individual people who differ in their obesity stages into mice.
So what’s exciting about that is that it shows the role for causality. But it’s a lot more complicated than just the phylum level changes that initially seemed promising. Now, it is important to remember that in the lay-it-out 2006 nature paper where they were looking explicitly at human weight loss, some of the patterns they saw there resembled what we had seen in the mouse data. But it seems like that may have been more of a coincidence about that particular population or possibly due to the difference in energy balance rather than the difference in obesity per se.
DrMR: And it’s an interesting question or string to pull at which is cause/effect. And I’ve seen differing things published in different papers. There is some data that I think certainly suggests a causal relationship. But there’s definitely other data that suggests the alterations in the phylum maybe a consequence of the obesity state itself.
And one of the papers that I found to be interesting that kind of supports that loosely was the fecal microbiota transplant therapy in type II diabetics from lean donors to these diabetic patients where there wasn’t really any appreciable change in weight loss.
So I guess regarding cause/effect and maybe treatments for this, where do you think we stand right now?
RK: Right. Well, so one thing that was interesting about that paper is they did see a restoration of insulin sensitivity on fecal transplant. So one issue is always what aspects of phenotype are you looking at. And in the early obesity papers, we were specifically not looking at type II diabetics as opposed to people who were obese without other metabolic complications. So again, that comes back to different populations likely have different things going on.
But where we are as far as microbial treatment for obesity—there are things that are pretty successful in rodent models. But as you know, the biomedical research enterprise is a lot better at treating rodents than it is at treating humans.
RK: The translation of how, there’s still a lot of work to go on that. What I can say is that a number of different interventions have worked pretty well in rodents as proof of principle. And then the question is always, what’s going to work in humans? And will it be based on the same principles?
But I think a good analogy is leptin where in mice and in rats you can control weight very precisely using leptin injections. In humans, that generally doesn’t work unless you have one of a handful of specific mutations. And so, although leptin is a very important part of the story of obesity in humans, you can’t treat obesity in most people just be injecting leptin.
And I think it’s going to be the same sort of thing in the microbiome where the stage where we’re at at the moment is finding all the parts of the complex metabolic circuit that we now know is distributed between people and microbes. And so rodents are very useful for finding out what the parts of the circuit are. They’re not quite so useful for figuring out what the switch is that you can use to affect the state in humans.
DrMR: Yeah, I think that’s incredibly well said. And I think it’s just worth reiterating. And I thought you said it really well. We’re much better at treating rodents at treating humans. And I think we still have a lot to learn.
And the reason why I just want to echo that—and I think the audience probably knows what I’m going to say—is I see a fair number of people in the clinic who—if I may be a little bit frank—have been kind of duped by a kind of pseudo weight loss miracle that’s been wrapped in a very interesting microbiota kind of sales pitch. And it’s unfortunate.
And the thing that I really try to protect the audience from is when something is in vogue and something is popular, you have great scientists like yourself contributing immensely. And then you also have people who are taking advantage of that science and misconstruing it for their own monetary gain.
And so I just want to echo that I think we have a lot to learn here. And we should definitely continue to follow this and support this in any way that we can. But just be a little bit cautious with things that you read on the internet where maybe someone’s talking about an animal model that promised weight loss. And now they want you to buy a test and a special kind of prebiotic or what-have-you with big promises of weight loss. Not to say that that can’t be helpful, but just be cautious with that because I’ve certainly seen a fair number of people trying to take advantage of some of this.
RK: Oh yeah, sure. Yeah, so your listeners should definitely be aware that rodents have completely different microbiomes from humans. And just because something works in one species doesn’t mean it works in another species. And yeah, it’s really important to check that what you’re about to take has been clinically validated with placebo control trials in humans rather than just relying on pre-clinical data like animal data.
RK: But what’s amazing about fecal transplant—for C. diff., it’s absolutely life transforming. So recurrent C. diff. infection, the fecal transplants have more than a 90% success rate in clinical trials in humans. And you have many people walking around alive who would be dead had they not gotten the fecal transplant by this point. That’s the experience of thousands of patients.
But that does not necessarily mean that it’s going to work for everything else that the microbiome is involved in. And while it’s amazing that we can literally cure this disease using human fecal material, when not going to cure anyone with bullshit. So it’s really important to figure out what’s real in these studies and what’s an exaggeration or extrapolating things too far.
DrMR: Completely agreed. And FMT as a treatment for resistant C. diff. infection is FDA approved.
DrMR: I’m thinking the next condition that will be FDA approved will probably be IBD, just looking at some of the studies and the promise that the studies are showing and also looking at what the other treatment options for end-stage IBD are—pretty powerful drugs or even surgical intervention. So it’s not the best frontline therapies we have to offer these people if they’re failing out of some of the milder drugs.
So what are your thoughts on IBD? And was it you or your group that wrote the paper that showed a pretty consistent signature in the microbiota that correlated with IBD?
RK: Yeah, the [inaudible 16:07] Paper. So we were involved in that, as were many other labs in consortia. But yeah, we did a fair amount of the data analysis for that paper and worked with Dirk Gevers. He’s now leading the J&J Human Microbiome Innovation Center to develop the dysbiosis index.
One thing that’s really cool about that is in additional work, we’ve been able to extend that kind of concept to other populations. And that’s one of the things that does seem to translate across species. So it’s pretty exciting.
As far as fecal transplant for IBD goes—for both Crohn’s disease and for ulcer colitis, the trials have been a lot more mixed than the trials for recurrent C. diff. And one thing that’s interesting that for recurrent C. diff. it seems that basically anyone’s stool sample will do. But for inflammatory bowel disease, there’s probably a lot more detail about how you have to match up donors to recipients.
So one thing the American Gastroenterological Association is doing right at the moment is setting up a nationwide registry for fecal transplant so that we can capture specimens from donors and from recipients and also capture long term outcomes and really build up that kind of information that you need about whose stool works for therapy for which kinds of patients rather than assuming it’s just going to work most of the time like it does with C. diff.
DrMR: Agreed. It’s miraculous the results that we see for C. diff. And we’ve had a couple different doctors that offer FMT discuss how, for C. diff., almost anything will work. One doctor even joked that you could probably even use dog poop, and it would probably work for C. diff. But for IBD, it’s much more mixed.
RK: Right. So one thing that’s important about dog poop, which I hope was a sarcastic suggestion—but there are some microbial species that seem to be beneficial to some species and harmful in others. B. fragilis which is beneficial in humans can be an outright pathogen in mice. And the same sort of things is almost certainly true for dogs where they have a much carnivorous lifestyle in the wild. And so there are a number of microbes that are adapted to that sort of gut ecology that are not so good if they get into our gut.
DrMR: Sure. Sure. And he was saying that kind of tongue in cheek. But I think the comment he was making was you could use almost any donor as long as they were screened, and you’d get a positive response where, with some of the IBD trials, it seems like potentially the donor/recipient matching is important or certain donor characteristics are really important.
RK: Yeah, exactly. Of course, my dog loves to give himself cross-species fecal transplants all the time. And he’s okay. So I think there’s no limits to what you can achieve. [Laughs]
RK: So I think there are limits to what you can achieve.
DrMR: Absolutely. Now, in your paper, or the paper you were a part of, it seemed like we could maybe use microbiota assay results as a predictive tool for future IBD. And I’m pretty conservative and cautious in the statements that I make. But it seemed like this is one of the first times we had really seen clear evidence that we could use a microbiota analysis to correlate or predict future risk for a disease. Am I accurate in saying that?
RK: Yes, that’s accurate. And since then, we did a couple of papers with [inaudible 19:53] at [inaudible 19:54] Chinese Academy of Sciences in Qingdao, looking at gingivitis and caries where there are microbial indicators of future disease, not just current disease, also worked very effectively. So in the case of caries, you can tell which kid is going to get a cavity several months before there are any clinical signs.
And this general concept of being able to not just read out current state from the microbiota or from the microbiome but being able to predict future state is really exciting.
One of the things that’s in progress at the moment is we’re working with Dr. Bill Sanborn who is the chair of gastroenterology here at UC San Diego and an expert on clinical trials in IBD to try to predict future flares with ongoing monitoring. And then also to try to stratify patients in the best drug response.
DrMR: That’s really, really interesting.
Microbiota Assays and Disease Risk Prediction
DrMR: Now, if someone is listening to this and they wanted to try to have a microbiota assay and know if that predicted risk, it seemed like the—I don’t want to call it an algorithm—but it seemed like the pattern that you guys were determining was pretty advanced. And I’m not sure many clinicians would, A, have the access to the type of testing that you were using and, B, would have the know-all to be able to interpret that. Would you say we’re there yet for this to be done in clinical practice? Or is some of this more research based only?
RK: It’s all research based only. And in particular, you should not try to diagnose yourself with IBD with something that’s not an FDA approved test. So where we are at the moment is the type of sequencing we used in that paper. You can do that through American Gut which is a citizen science project that I run. And there are also a number of commercial companies that offer things that are like it with different degrees of accuracy.
But the problem is in interpreting the data. And in a research context where we have a well-defined cohort, we can develop these indicators. We can do cross validation between different cohorts and that kind of thing. If you’re getting a single sample processed and you’re trying to incorporate a single sample into that existing body of knowledge, that’s a lot harder. And that’s not something that can be offered in a clinical context yet, although that’s exactly what the field is trying to work towards.
DrMR: Gotcha. So for people who run an American Gut—and I encourage people to run American Gut or some of the similar testing. But I also encourage them to be cautious in how much you can take away from that in terms of clinically and if that means you should make any clinical interventions from that. So it sounds like we’re still a little bit early for someone to be able to take that and then make any kind of predictive remarks for their future IBD risk. Is that accurate to say?
RK: Yeah, that’s accurate. And that’s through, not just of American Gut, but of any of the testing services out there that are looking at the whole microbiota or whole microbiome. So what we assume everyone else is trying to do right now is to figure out how well can you incorporate a single, new sample from a member of the public into the existing data resources.
RK: And then you have to take the next logical step to make it easy enough that your physician can explain it to you in the 15 minutes or whatever you have together when you’re in the clinic. So we’re trying to make it that easy.
And we have a program with the AGA that we ran this year to run a bunch of gastroenterologists own samples through American Gut and then compare them to their patients. And then we’ll be doing that again in an expanded way next year.
There was a lot of demand for it where, essentially, we’re trying to do that kind of physician education piece and give you an idea of what you can and can’t get out of those kinds of reports and other kinds that are higher tech that are just about to become more widely available than they have been.
But at this point, it’s still very much a research tool rather than something that you can expect to use in the clinic.
And as you know, this is really a problem for clinicians at the moment because you have a lot of people who are coming in with their reports and saying, “Okay. I got this readout of 1000 species in my gut.” Or better, or perhaps worse yet, a readout of a million genes that were in my gut. So, “Hey, doc, can you tell me what’s wrong with me?” And about all you can do at the moment is refer them to your colleagues in psychiatry for being crazy enough to think that that’s going to help.
But the goal is to make it not crazy and to make it so that you can actually get something that’s useful and actionable and where you can put people on a scale from high to low risk or predict the date that you’re going to have the next recurrence just like we can figure out—say, on a forensics project, we can figure out how long ago someone died based on the microbes in their decaying corpse or figure out how long ago someone touched a particular pen or glass or other object based on the microbial succession.
So our goal is to do that kind of thing for the future as opposed to reading out the past. And it’s likely to be technically feasible in the near future. But it’s not something that you can just do easily today, unfortunately, although we hope that within in the next few years, that’ll be something that’s routine and unremarkable.
DrMR: I hope so also. And I’m really happy to hear you say that because I have been watching this literature for quite awhile. And I’ve not seen anything that has been clinically actionable yet. And my recommendation to people has always been continue to support these different projects because we need to continue to gather data to get to a point where we can have clinical interpretation and then clinical actionables. But we’re not there yet.
But I also have witnessed other people on the internet really making very clinically prescriptive recommendations based upon some of these tests. And I said to myself, “Either I am dumb and just not able to pull any of that out of the literature, or some of these people,” as you had put it earlier, “might need to go see a psychiatrist because they’re maybe overstepping.”
So I think that’s a really important point. Thank you for saying it. But also, I just want to reiterate that, everyone listening, I would still encourage you to participate in these projects because we need this research to get to that clinically actionable point. And it sounds like, Rob, that we’re not too far off from that. So that is definitely very exciting.
RK: Yeah. Absolutely. This is exactly the right time to participate too because we’re probably right at the boundary between being able to use your sample only for something that’s useful for people sometime in the future and being able to use your sample for something that’s going to be useful and cross validated on a time scale that’s relevant to you.
DrMR: Yeah. No, this is really fantastic stuff. And I’m looking forward to when these things are able to be done in a clinical setting and really helping people be more prescriptive with getting their guts tuned up, so to speak.
RK: Yeah, exactly. That’s exactly why I moved from a university that wasn’t directly attached to a med school into a clinical department. I think the time between now and when we can actually use this to directly benefit patients, not just sometime in the future but today, is getting very close.
And it’s really exciting to be able to work with multi-disciplinary teams including clinicians to figure out how we can bridge that gap from something that lets us answer fun research puzzles versus something that helps patients directly.
DrMR: Absolutely agreed. And now, Rob, I know you have to go any minute here.
DrMR: And if you have to go now, let me know. Do you have time for one more question? Or are you at your time?
RK: One more question, I guess.
DrMR: Okay. All right. I will limit you to this. But I know people are going to be wondering.
Small Intestine Assays
DrMR: We don’t get a good assay of the small intestine with the current tests.
DrMR: How much relevance do think that has? Do you think we need to get that data? Do you think it’s not that important? What are your thoughts on that?
RK: Well, sometimes, it’ll be important, and sometimes, it won’t. One thing we’re doing at the moment with Dr. Laura Pace who’s a very talented clinician here working primarily with Pieter Dorrestein is doing biopsies all the way along the length of the gut of healthy controls versus IBD patients so we can test directly what’s most useful to read out of the biomarker.
And one thing to remember is that for ileo Crohn’s disease, for example, even though the disease is in the ileum, you get very distinct signatures in the stool from people with ICD versus controls and versus CCD.
So I think probably what’s going to happen is sometimes you’re going to get that biopsy directly. Sometimes, you’ll be able to read it out indirectly, like hydrogen breath test for SIBO, for example. And sometimes, you’re just going to need to do the biopsy and get the small intestinal data.
And a lot of what we try to do in my lab is figure out when can you get away with less invasive readout of a more invasive sample you’d like to do.
RK: Say, for example, on a cystic fibrosis project, we’ve been dicing out into little cubes lungs removed from transplant patients so we can ask, if we had a full view of the cystic fibrosis lung, what would we be able to see? And then how well can we approximate that if we can just do a BAL or have the patient cough up some sputum or that kind of thing to avoid having to remove the lung every time?
DrMR: Sure, yeah. No, I agree. Yeah. No, I think that’s great. And I think it’s a very important question to ask which is, where can we get away with the least invasive testing? And then what specific population or subset of a population might we need to be more invasive? That’s actually a really good point.
RK: One thing that’s amazing about that is sometimes you might not need to go to the gut at all, even for something that’s a GI disease. So for example, for let’s see, I think it was type II diabetes and for cirrhosis and for rheumatoid arthritis, oral microbiome biomarkers are looking pretty good as a readout.
So if you want to diagnose as opposed to understand the underlying biology, there might be a lot of cases where you don’t even need to go to the gut at all. You might be able to get away with a tongue swab or spitting in a little tube, which obviously people tend to like a lot more than having to collect some of their stool.
DrMR: People don’t like stool tests. I can tell you that—
RK: Absolutely true.
DrMR: From experience in the clinic.
RK: Yeah, I can absolutely see why. I’ve been collecting my own about every day for the last eight years. And I totally get why you would want to minimize the number of times you had to do that.
DrMR: Well, Rob, thank you so much for taking the time. This was a really great call. I think this will help a lot of people out there who are kind of wondering how to interpret all this new data and what to do with it. So thank you again.
Is there anywhere people can track you down if they wanted to follow you or read more from your or what have you?
RK: Yes. So my lab has a website. So if you search for “Knight lab” on Google, I think we come up fifth. And then the American Gut Project is also very easily discoverable. But yeah, thanks again for your interest in this.
And I’ll definitely be in touch again when we’re getting to the point that we can get clinical utility for the IBD samples. I think that’s going to be soon. But you know how it is with research.
RK: It always takes longer than you think even when you take into account that it’s going to take longer than you think.
DrMR: Sure. Well, please let me know when. It’d be great to have you back on to expand on that when we get there.
RK: Oh, great! Well, thanks so much. Maybe next year around DDW.
DrMR: Sure. Sounds good. Well, thanks again, Rob.
RK: Cool. Thank you. Have a great evening.
DrMR: You too. Take care.
If you have questions about microbiota testing, click here.