Episode 1 Transcript: Trusting a Gut Instinct

Introduction (00:03): Great science and engineering often begin with a singular hypothesis, but how does a lone spark of innovation become popular science? From Caltech, this is The Lonely Idea.

Rich Wolf, host (00:16): Welcome to The Lonely Idea. I'm your host Rich Wolf. Today, I'm speaking with Sarkis Mazmanian. We'll explore how he has been studying the relationship between the gut and the brain to make discoveries about the microbiome that bring us closer to treatments for autism and Parkinson's disease. Sarkis, welcome to the program.

Sarkis Mazmanian, guest (00:34): Hi Rich, how are you?

Wolf (00:35): Doing great, thanks. Thank you for taking the time today.

Mazmanian (00:37): My pleasure.

Wolf (00:38): Sarkis, before we delve into the gut brain axis and the microbiome and all of the stuff that hopefully we will uncover today, one of the things that I know people would be interested in is how did you, as a young guy growing up in California, get interested in microbiology and bacteria?

Mazmanian (00:54): Well, it was a circuitous route. It didn't happen early on in my life. It happened in college and as I explored different areas of biology, I really gravitated towards microbiology for a very simple reason. I found microbes to be incredibly versatile and efficient machines that they were doing so many things in such a small cell, shaping their environment in ways that I thought were profound. And I fell in love with the biology and I fell in love with the potential, the impact that they would have on their environment. And immediately I started thinking about humans as their environment and I thought to myself, we're just one more environment, one more surrounding that they could shape.

Wolf (01:38): That's the thing that's so strange. When I, when I was a student 25 years ago, dating myself, microbiology to the biologist was like not an interesting field. It was a very interesting field for environmental scientists, you know, the notion was could we develop bacteria that would chew up oil and oil spills and you know, if you want it to do something groundbreaking, you were going to do genomics, you're going to do molecular biology. Yet you saw something in microbes that others didn't see and you saw it very early. What was it?

Mazmanian (02:09): It was when I transitioned from studying microbial pathogenesis to beneficial organisms, what we now call the microbiome. And so my formal training was in infectious disease because again, at the time, you know, there was a large effort in trying to understand pathogens so that we can understand and develop ways to inhibit their ability to cause disease,

Wolf (02:30): To develop antibiotics, for instance.

Mazmanian (02:33): Exactly. And so, and this was at a time, and we're talking late nineties, early two thousands, when drug companies had abandoned many of their antibiotic programs, because people figured the problem was solved and we now know that it's not. And even at the time, microbiologist knew that that bacteria would find ways to overcome, you know, antibiotics and, and develop resistance, which we know that they do. But as I fell in love with microbiology and was thinking about the next phase of my career, I thought, well, I can continue to do microbial pathogenesis, continue to study infectious disease, or I can take a different route. And that's more of my personality, I think, is to, you know, go down the untraveled path if you will. And I can remember the moment when I decided I wanted to study the microbiome. Again, this was at a time before people even called it the microbiome.

Wolf (03:25): And where were you at the time when this happened?

Mazmanian (03:27): I was sitting at the base of the Hancock tower in Chicago reading a two page article about all these bacteria that live in our intestines that help us digest our food and make nutrients and minerals that we use, and nobody was studying them or very, very few people were studying these organisms. And there are a hundred times as many bacteria that live in your body than bacteria that will cause disease. Yet everyone was studying the disease causing organisms and not the bacteria that we were living with for the entirety of our lives. And I thought as a microbiologist, what an opportunity to go into this new area.

Wolf (04:00): Incredible. Could you frame that for people again, just so people know the numbers, the number of bacteria that aren't us, that are in our body, the number of different species that are in our gut. What are those numbers?

Mazmanian (04:10): So we have about a hundred trillion bacteria in or on our bodies. And so if you were to compare that to number of human cells, there are more bacteria in and on us than there are human cells in our bodies. And so we are more bacteria than we are human on a cellular level. And so our interactions with microbes are quite profound. And the numbers of species that inhabit the human race approaches about 10,000, at least that's our current knowledge. Less than 50 species of bacteria cause disease.

Wolf (04:42): And of the 10,000, how many would be living inside our GI system versus on our skin or elsewhere?

Mazmanian (04:47): Yeah, so we have approximately two to 500 different species of bacteria in our intestines, roughly the same number in the oral cavity. A few dozen different species on our skin.

Wolf (04:59): So, as we start to think about how do we design experiments around this, so you now you fall in love with the notion of how we coexist with this bacteria and these bacteria are doing good things for us, but you don't just stop there. You have to come up, you have to start to think of like, what experiments do I want to do? So what was your next step?

Mazmanian (05:18): I thought exactly about how am I going to study organisms that already live in an animal, microorganisms that already live in an animal. For infectious disease research you just add an infectious bacteria to a mouse or you know, people get, get infected. And so you're introducing an organism from the outside, so that allows you then to study. But how do you remove a particular organism that's already there? So we took the opposite approach. We said, why don't we raise animals, in particular mice, that have no bacteria so that we can selectively control what organisms we colonize those animals with.

Wolf (05:55): Incredibly hard to do.

Mazmanian (05:56): Yeah, and then study the effects of those organisms.

Wolf (06:00): When you said you wanted to do this, did people at the time say, this is crazy, this makes no sense.

Mazmanian (06:06): At that stage, not to my face. That came later in my career. I had gotten my own fellowships. I was able to support part of the research and I found an environment, a laboratory that was able to provide me with the infrastructure to test these ideas. And so, initially, it was me taking a chance on being able to set up the system to be able to study what I wanted to study. And at the time it was asking the question, how does specific bacteria improve the function of the immune system? So that was the specific hypothesis. And of course we all grew up thinking bacteria were, you know, set out to cause disease, that there were these horrible little creatures, and all they want to do is make us sick. We now know that that's not the case. How do you study that in the context of an animal system? Well, we went to what is called gnotobiology-what I just described. It's setting up these germ-free colonies and adding those organisms back. And at the time, again, there was, there were not enough people knowing what I was doing to comment. The comments about being crazy came when I started applying for funding for this research.

Wolf (07:20): When you were being reviewed by peers?

Mazmanian (07:22): That's right. And I remember, and I can probably dig up the letter, my first fellowship application was denied because, the comment was, this is nothing more than a hunch.

Wolf (07:35): I love that. We could change the name of the show from the lonely idea to chasing a hunch and we could probably just spend an hour exploring the difference between a hunch and a guess. There's another aspect of this that I want to quantify for people as well, which is these bacteria, you can't just go get a can of yogurt or old milk and get these bacteria. These bacteria are anaerobic. Not only is it hard to sterilize the environment for these mice, but then you have to try to manipulate them in a way that's very difficult. Could you say a little bit more about that?

Mazmanian (08:02): What we're trying to do is identify rational experimentally supported organisms, meaning organisms that come from the human gut, have co-evolved with people, and are more likely to have an effect on our biology. And yes, they're harder to grow. Yes, they're harder to study. But we believe these are the organisms that are going to have the most impact on human health.

Wolf (08:22): Is there any evidence inside ourselves of how bacteria have evolved with us and have become part of our normal biologic process?

Mazmanian (08:32): There's evidence of co-evolution based on the fact that these microbes will produce molecules that bind to our receptors, which are proteins or other molecules on the surface of the cells that then change their activity. And these are very specific interactions, like a lock and a key. And it's hard to believe that this would've just happened randomly in nature, that a microorganism would produce a molecule that fits perfectly into the receptor on a human cell. So I think that's very strong evidence for co-evolution. But there's also genomic co-evolution, which there's very strong evidence for, [and] is that many microbes have actually picked up genes, picked up fragments of DNA from their hosts. And if those hosts are people, then in humans, they've picked up DNA and genes that are more humanlike than other bacterial genes and incorporated those genes into their genomes.

Wolf (09:31): Incredible.

Mazmanian (09:32): And so again, there's only a select number of examples of this so far.

Wolf (09:36): Now let's go back to your clean mouse. How did we get from the clean mouse where we're introducing this complicated set of experiments of introducing this bacteria into this mouse to understanding how we coexist with bacteria and affect our immune system?

Mazmanian (09:52): So what we wanted to understand was what does the immune system of an animal look like in the absence of bacteria and how does that change when specific organisms are introduced? In fact, our first piece of evidence that the immune system doesn't fully develop in the absence of microbes, was the fact that we showed that subsets of T cells were either completely nonexistent in the germ-free animal or reduced in number. And so when we added back particular organisms and later identified the microbial molecules in most organisms that then led to the development and function of those T-cell subsets.

Mazmanian (10:30): And so this led us to this hypothesis that we're still, that we're still working on, that our human genome doesn't encode all the information we need for our health. It's that we also rely on information from our second genome, which is our microbiome that works in concert with our own DNA and provide signals that when they come together, they give us our full repertoire, our full arsenal of immune cells.

Wolf (10:57): It's incredible when you think about that. If we were born void of all of this bacteria, we wouldn't have the same immune cells. We wouldn't have the same number of T cells.

Mazmanian (11:05): I can only speculate, obviously, but you know it's important to remember that we were never sterile. We were never germ-free. We [were] the first metazoans.

Wolf (11:17): We started as germs.

Mazmanian (11:18): That's right. We evolved in a world covered with microbes. Bacteria inhabited the earth 2 billion years before the first eukaryotic cell, which again was a product of a microbial symbiosis. And so the bugs were there the whole time. And as we evolved, one can again potentially project or speculate that organisms, microorganisms found it in their own best interest, if they're going to live in a host, to have that host have a very strong immune system to help them fight off infectious agents.

Wolf (11:54): So let's come back to our germ free mouse. We now have identified that when we change the gut bacteria in the mouse, we changed the immune system of the mouse. What was the next set of experiments that you wanted to do?

Mazmanian (12:08): We wanted to understand how. We started to develop disease models in the laboratory or utilize disease models in laboratory and ask, if we had this organism that's augmenting immune function, can we then understand how that process would improve the ability of the immune system to fight off infection? Or would prevent that immune system from causing an autoimmune reaction or an inflammatory reaction. And so over the years we were able to identify many of the molecular and cellular interactions between various bacteria and the host. And we do this in the context of disease models.

Wolf (12:48): What was the first disease mouse that you wanted to look at or the first model that you created?

Mazmanian (12:52): We initially studied inflammatory bowel disease or colitis in experimental systems. So inflammatory bowel disease is inflammation of the intestines. It affects about 1.4 million people in the United States.

Wolf (13:04): The combination of Crohn's disease and ulcerative colitis.

Mazmanian (13:07): But we discovered in, again, a series of studies, that there are bacteria that induce an anti-inflammatory response in our immune system. The bacteria turn on a different subset of T cells. These anti-inflammatory T cells called regulatory T cells, which then dampen that immune response. They do this in a very natural way. The bacteria that we've discovered, at least in mouse models so far, what they show is they balance the immune response, meaning that they don't immunosuppress or immunocompromise the animal. What they do is they suppress the inflammatory response to a level that no longer cause disease, but it's still able to respond to a microbial infection.

Wolf (13:48): And do they do this by varying the different types of T cells?

How do they actually do it?

Mazmanian (13:49): They augment the development and the function of this anti-inflammatory regulatory T cells.

Wolf (13:57): There must be something on the bacteria that senses when there's a problem. So are you interrogating these bacteria to say, how did the bacteria know that something bad is happening?

Mazmanian (14:05): We have a significant amount of evidence that suggests that these bacteria turn on the ability to produce the molecules that leads to that anti-inflammatory response only in the context of inflammation. But we don't know what that environment is. We don't know what those signals are.

Wolf (14:25): So your next thought must be, could I create a therapy that mimics what these bacteria do? How do you do that?

Mazmanian (14:32): We actually believe in the power of the bacteria itself. These organisms make molecules in the right place, in the right amounts, at the right time. And so our belief is that bacteria, or living organisms, would be those therapeutics, would be the avenue to try to understand how one can treat an immunologic or neurologic, a metabolic disorder.

Wolf (14:58): So another way of saying that is you could go in and, in a very sterile environment, synthesize these anaerobic bacteria and find a way to introduce them into my gut. That's gotta be complicated, creating an anaerobic environment and then somehow being able to put that into a pill that I can take. Is that possible? Can we do that?

Mazmanian (15:17): It is possible. It does require cutting-edge manufacturing. The probiotics that are being mass produced are not anaerobic and they're quite easy to grow cause they can grow in ambient air. But the infrastructure that people use to grow aerobic organisms, the probiotics, can be modified to grow anaerobic organisms. And so there's still quite a few challenges to mass production, but the ability to grow organisms in the absence of oxygen is something we do in the laboratory all the time. It just now has to be transferred to a more commercial scale.

Wolf (15:52): In a manufacturing setting. Do you believe that that's a reasonable approach to drug development? And do you have evidence for that as well?

Mazmanian (15:59): I do. And we do have evidence for that because we've been studying this particular organism called

Bacteroides fragilis that makes a molecule called polysaccharide A, or PSA, that has this anti-inflammatory activity that I've been telling you about. And so PSA itself is impossible, to our knowledge, impossible to synthesize in a laboratory. So we produce it by purifying it away from the bacteria that naturally make it. But ultimately what we're, what our goal is, is to put a purified microbial molecule into a pill. So it's very similar from a delivery standpoint to traditional drugs because you're taking a purified molecule. The difference is that molecule didn't come from a laboratory, it came from nature, and nature meaning an organism that already lives in the healthy human gut.

Wolf (16:51): You're actually using the bacteria as the engine and the manufacturing for making the molecule that you want to use.

Mazmanian (16:57): Correct.

Wolf (16:57): We're now several years back, this is maybe five or six years ago or longer. And we're past the reviewer that called this a hunch. Have people now said this is no longer a hunch, we're all on board, or were you still finding some skepticism in the approach that you were taking?

Mazmanian (17:12): I believe immunologists have embraced the power of the microbiome and how important the microbiome is, not just to their own research but to the field of immunology. And of course we had some small part in that. But at the end of the day, it was quite gratifying, or it is quite gratifying, to look back and to think that the people who may have been skeptical a decade or more ago have now incorporated that line of research into their own laboratories.

Wolf (17:41): But that wasn't enough for you. You're very interested in the central nervous system and neurobiology. So maybe tell our listeners a little bit more about that.

Mazmanian (17:49): In past few years, we have transitioned our individual research program away from immunology. Not completely. We will always study some aspect of immunology, but more towards trying to understand the connections between the gut and the brain. There was some evidence at the time, even when we started at work, that bacteria were affecting other parts of the body beyond just the gut. And Rich, I'm sure I've told you this story before. It came into conversation with Paul Patterson, about seven or eight years ago. Paul was a neuroscientist here at Caltech. He sadly passed away in 2014. And he and I were having lunch at our faculty club, at the Athenaeum, and I was probably going on and on about our work on colitis and how we have bacteria that improve gut function, and Paul at the time had just developed a mouse model to study autism.

Mazmanian (18:41): The model was based on an environmental risk factor, and he had shown that by introducing this environmental risk factor into mice, that the offspring of those animals had many of the features of autism. Paul was a man of few words, so I'm sure as I was talking and talking, he interrupted me with the following sentence that changed the trajectory of my research. He said, "I think some kids with autism have GI issues." So both he and I did some digging and again, you know, many years ago there were a few papers that suggested that there were abdominal issues in the autism population. So we took another hunch. We looked at some of the colons of the animals in this mouse model of autism that he had created. And on our first experiment we showed an elevated immune response in an autism model and a phenotype called leaky gut, or permeability of the intestinal barrier, which is a surrogate to show that gastrointestinal function had somehow been compromised.

Wolf (19:49): Now we hear leaky gut all the time in the news. It's a very real phenomenon and you've started to identify what's actually happening and that potentially these autistic children have a preponderance of it. They have much more pathology here than the neurotypical population.

Mazmanian (20:06): We know, we believe we know what's causing, what can cause leaky gut in mice, and it may be similar to what's happening in people. All animals have a single layer of cells that separate their gut, their intestines, from their bodies. In other words, the barrier between us and our microbiome is a single layer of cells called epithelial cells. And these epithelial cells will form structures called tight junctions. So think of this a sa zipper. And so these cells come together, they form a zipper, they form a nice barrier, and they have various functions. Most importantly in my mind are to allow good molecules in and keep the harmful molecules out. On the mouse, the proteins on the epithelial cells that form that zipper are compromised. They don't function as well as they would in a person who had an intact barrier. And so molecules that weren't meant to get across the intestine do leak out of the gut into the circulation and potentially can cause some health issues.

Wolf (21:13): In these autistic mouse models. So you create mice that mimic the symptoms of an autistic patient, child, and you can also obviously create mice that are neurotypical. And in these, in these autistic mice, you're able to actually observe what you're describing in terms of the lining of their gut.

Mazmanian (21:32): Yes, models both Paul have created, as well as other laboratories around the world have generated, based on various different genetic or environmental triggers for autism, in some of those models it's been shown that the animals have this leaky gut phenotype.

Wolf (21:49): So let, let's pause there for a second because the first lonely idea is "I'm interested in studying the bacteria that coexists with us and I think they could have important function." A hunch. The second lonely idea was, " Can I create these sterile models, these germ-free models, that will enable me to do the experiments that I want to do?" The third lonely idea is, "I think potentially the immune system is regulated in part by this." You've proven all three of those to the point where you've won over a large swath of the immunology community and now you're embarking on what is truly a third rail, which is, "Are neuroatypical patients, autistic patients in this case, potentially not all of them, but some subset of them, potentially experiencing those symptoms as a result of something wrong with the bacteria in their gut?" Huge. A huge leap.

Mazmanian (22:51): We don't know what's going on in humans relating to our discoveries and animals, at least not completely. We have just a very, very partial view, extrapolating our human, our mouse data to humans. So what we know is this: is that in the animal models where we've shown leaky gut, we've actually tracked what molecules are leaking out of the gut into the circulation. And in one example, we've shown that a particular molecule, which doesn't leave the gut of a healthy animal, leaves the gut of a mouse with features of autism and actually winds up in the brains of those animals.

Wolf (23:33): Wait a minute, hang on. I have to stop there. That's incredible. So you have normally healthy mice, you've identified that there's a molecule that stays in their gut, never goes anywhere. Somehow they managed to keep that molecule in the gut.

Mazmanian: That's right.

Wolf: You have autistic mice. Somehow the molecule leaves the gut, probably through this leaky gut mechanism, and ends up in the brain of these mice.

Mazmanian: That's correct.

Wolf: And what is that molecule?

Mazmanian (23:57): It's called 4-ethylphenyl sulfate. It's a derivative of tyrasine. Tyrasine is an amino acid that's found in our diet. So we all ingest tyrasine when you eat anything with protein in it and our gut bacteria then transform tyrasine, that amino acid, into 4-ethylphenyl sulfate. And that's the molecule that not only does it get across the gut, get across the gut as well as getting into the brain, but it's sufficient to induce many different types of neurological or behavioral deficits in the animal. So far simply correlation, but what this may lead us to is a diagnostic, maybe a companion diagnostic to the drug. Is that ability to identify a patient population and treat them in a personalized way.

Wolf (24:43): I mean that would be a remarkable thing to treat the symptoms associated with autism. I apologize for the naive way I'm asking this question, but how does it get out of the gut and then travel and end up in my brain? We have a blood brain barrier. We have lots of things to cause that to not happen. Now does that happen?

Mazmanian (24:59): We don't know exactly exactly, but this is an important topic because just like the , our blood brain barrier is there for a reason. It is to keep the brain safe from toxic molecules. Right? And this is clearly a toxic molecule.

Wolf (25:12): So now you have to, now I have to start saying to myself, what other neurological diseases potentially have the same mechanism where there's bacteria in our gut that are affecting the concentration of any whatever bad molecule is out there. Now I happen to know, having read some of your research that you've already identified one associated with Parkinson's disease. Let's make the leap from autism to Parkinson's. Tell us a bit about your Parkinson's research because the paper there was just mind blowing when I read it.

Mazmanian (25:41): So similar to autism, in fact even more well studied and well documented, are the GI symptoms in Parkinson's. And so I often refer to James Parkinson's essay on shaking palsy published in 1817 where he wrote about six of his patients and their tremors. He noted that in those individuals that had shaking palsy, that four out of the six who were described in the case report had GI symptoms as well. So he documented that. And so 202 years ago we knew that there were GI symptoms in Parkinson's disease. And for one of them, he gave them a laxative that led to improvement of their motor symptoms.

Wolf (26:33): Incredible. 200 years ago?

Mazmanian (26:34): So this is not a new concept. Now in the last three or four years, you know, both our research and research of others, there's been an increased emphasis on the gastrointestinal symptoms, on constipation. So the fact that the GI symptoms appear before the motor symptoms and the motor symptoms are driven, by lesions or pathology in the brain suggests that there may be pathology in the gut prior to pathology in the brain.

Wolf (27:05): As you start thinking about other diseases that potentially are regulated or affected by the insult caused by the bacteria, how do you, where do you go next? What are the diseases that we should go after?

Mazmanian (27:18): It's a good question because we need to be forward looking and think about other indications that one may use the microbiome to develop therapies for. It's an inexact science. Where we've been successful in autism and Parkinson's, both were because individuals with those disorders have GI issues. But that's not a prerequisite. You don't necessarily have to have a GI issue for your microbiome to impact your neurological status.

Wolf (27:46): You may not have that symptom, but you could still have the problem with the bacteria.

Mazmanian (27:49): Correct. The evidence for the microbiome impact to the immune system is much, much stronger than that of its ability to affect the nervous system just because it's been studied longer and maybe those interactions are more robust. And, several neurological disorders are believed to have an immune component. There's an emerging body of evidence suggesting that the immune system is integral to the pathophysiology of Parkinson's. It's the same for Alzheimer's. In fact, the evidence, I believe is stronger that the immune system induces or immune activation is required for the cognitive defects in Alzheimer's disease. Potentially the immune system is how you connect the dots between the microbiome and the brain--is that by altering the immune status of the individual, then that might lead to some neurological outcome that leads to the death of neurons that may lead to a neurodegenerative disorder such as Parkinson's or Alzheimer's. But again, at this point we're speculating.

Wolf (28:46): And as you said, the best validation is so many other scientists around around the world now who are starting to research these same things. And a year or so ago, you along with others formed a company that is trying to develop therapies in the field of Parkinson's and autism.

Mazmanian (29:04): Yeah. So the company is called Axial Biotherapeutics. And I would say not only would a clinical success validate our research, but I would argue that clinical success is required to validate the microbiome field. But of course mice are not humans. And there's a long history of both biology and drugs that have been discovered in mice, validated in mice, that never translated to people. We're studying biology. And I think there's, there's beauty and elegance in discovery for the sake of discovery. But if one can then leverage that to help people, I think, then that effort was, was even more worthwhile. And so there was an opportunity to found a company that would essentially take the learnings from our laboratory and apply them to people.

Wolf (29:58): The thing that's so exciting about it. And one of the reasons I wanted to sit down with you today is that it's such a new tool. And let's face it, there were people that were skeptics, there were probably skeptics all over the country and you told a great story about how you were hired here. I wonder if you would share that because I think it's a great way to punctuate the whole concept of the lonely idea and what you were able to do here and why you've decided to stay here.

Mazmanian (30:23): So now you're taking me back to 2005 when I was on the job market interviewing for faculty positions. And I was obviously very interested in identifying an institution that would share my vision for the research program that I wanted to build. And so I took my show on the road, visited maybe half a dozen or eight institutions for formal interviews and was offered a handful of jobs. And Caltech was the only place that I sincerely believe shared my vision. I believe the other offers, I'm only, again speculating, I believe the other offers were based on my CV and not based on a scientific alignment. And I believe this to be true based on the enthusiasm that I felt when I visited Caltech, on the types of questions that were asked of me and on comments from members of the search committee that told me "You want to cure disease with bacteria? That's so crazy, it might actually work." No one else, nowhere else had that same reach for the stars mentality.

Wolf (31:41): Well, with that, I'd like to thank you for your time today, Sarkis. And on behalf of our listeners, thank you for the work you're doing and on behalf of the patients, hopefully we'll have a great outcome. It's been great spending time with you today.

Mazmanian (31:52): Thank you, Rich.

Wolf (31:54): The Lonely Idea is produced at Caltech. Learn more about Caltech innovators and their research on our website at www.caltech.edu or connect with Caltech on Facebook, Twitter, Instagram, and YouTube.