Scientist Stories: Emily Balskus, Using Chemistry to Decipher the Human Microbiome

Good evening, and welcome, everybody, to this evening's Science for All Seasons talk. These talks are intended to give you a chance to explore hot topics in genomics and biology with some of the leading experts from the Broad Institute of MIT and Harvard. My name is Tom Ulrich. I'm the Associate Director for Science Communications here at the Broad Institute, and it's my pleasure to welcome all of you here tonight. Our speaker for this evening is Emily Balskus. Emily is an Institute member and part of our Chemical Biology and Therapeutic Science program, as well as a professor of chemistry and chemical biology at Harvard University. Emily is going to talk to us tonight about the gut microbiome, that collection of bacteria and more that live in our intestines, each and every one of us, and about her lab's efforts to understand the chemistry of those microbes and how they metabolize the drugs and nutrients that we all then take in. If you have any questions for Emily, please ask them using the Zoom Q&A function, and we'll do our best to discuss them during the Q&A portion of the evening. And if you happen to be tweeting tonight, please use the hashtag #BroadSFAS, Broad SFAS. And Emily, over to you.

Thanks, Tom. It's a real pleasure to be able to share my research. This evening. And I'd like to start by inviting everyone to imagine a chemist at work. So I'm guessing that just about everyone is picturing something like this, and, and I did too at the start of my career as a chemist when I was first learning how to build molecules in a lab. But my career and life changed when I realized that some of the most innovative chemists chemists have never worn a lab coat. So these chemists I'm referring to are, of course, microbes, individual creatures like bacteria, archaea, viruses, and fungi. Microbes are by far the most numerous organisms on the planet, and they live together in complex communities called microbiomes. Now microbes are truly astonishing chemists, and we all benefit from their talents. Their chemistry is truly all around us. For example, we can find microbes in the soil where they actually make complex bioactive molecules, including many important medicines like the antibiotic vancomycin. Microbes are also found in the ocean. Here they perform very challenging chemical processes like photosynthesis on a scale that affects Earth's climate. Climate and geochemistry. But we are also a habitat for microbes. In fact, one of the densest microbial communities on Earth lives within each one of us, the trillions of microbes that inhabit the human gut. So each one of us is colonized by a unique set of microbes. And we know that these organisms play critical roles in keeping us healthy. For example, they help us to digest our food. And we know that they help to train our immune system. So if you follow science in the news, I'm sure you've heard of the human microbiome, especially the organisms that live in the gut. It's a topic of much excitement right now, both in the popular science press and certainly among researchers. But it turns out that this is not really a new area of science. So we've actually known about these organisms and been interested in them for quite a long time, essentially since the birth of microbiology. The first observation of the human microbiome was by Antonie van Leeuwenhoek in the late 1600s, shortly after he had invented the microscope. He actually wrote the statement here on this slide describing organisms he observed in dental plaque that he had obtained from a man who'd never cleaned his teeth. You can see immediately upon realizing that there were microbes associated with with humans, scientists found this very intriguing. But if we've known about our close relationship with microbes for this long, why are we suddenly so interested in them right now? Well, it turns out that recent advances in DNA sequencing technology have changed how we can study microbes and microbiomes. So now we can start with either a single strain of microbe or a community of organisms directly from a habitat like the human gut or soil. We can extract DNA directly from that organism or sample, and we can sequence it to learn about all of the genes that are present in that microbe or in the collection of genomes associated with a microbial community. This type of collection of genomes directly obtained from a microbial habitat is called a metagenome. So by analyzing the genes encoded by microbial DNA, you can start to learn many interesting things about a microbe or a microbial community without even having to study these organisms in a lab. And so in the case of especially the human microbiome, one of the things we've been able to do with DNA sequencing is actually associate changes in the composition of the human microbiome, which microbes are present in a community, with different human diseases. So over the last several decades, we've learned that there can be very strong correlations between the different organisms present in a microbiome and various human diseases. But it turns out that even though we have obtained all of this very interesting correlative information, we still don't have a great understanding of exactly how the gut microbiome affects host biology. Do these organisms directly cause these diseases, or are these connections just correlations? This lack of mechanistic knowledge currently limits our ability to treat and potentially prevent many of the diseases that are linked to these organisms. But I just told you that DNA sequencing allowed us to identify microbes in the microbiome. Why isn't this enough information to give us an understanding of how the microbiome works? Well, it turns out that, that, you know, many of the microbes in the human gut are not well studied. Many have never been grown in a lab. But even for our most well-studied microbes, Knowing who they are, you know, being able to identify them doesn't necessarily tell us about what they can do. So one of my favorite examples is the microbe E. coli. So let's compare E. coli K12, which is the strain of E. coli that scientists like to study in the lab, it's very domesticated, with EHEC. This is E. coli O157:H7. 157:H7, the pathogenic strain of E. coli that's responsible for outbreaks of food poisoning. Now, these two strains of E. coli, even though they're very closely related, are very different from a functional potential. They'll behave very differently. For example, if they enter the human body. And this functional difference is reflected in the whole genomes of these organisms. So it turns out only about half of the genes in these two E. coli genomes are shared. But if you look instead just at the bacterial gene that we use as a marker of identity, the gene that we use to, to tell which microbes are present in a microbial community, we see that this gene is essentially identical between these two organisms. So when we find E. coli in a microbiome sequencing data set, that doesn't necessarily tell us much. About what that E. coli strain can do. So knowing who a microbe is doesn't tell us much about its function. So to really understand the microbiome, we need to do more than just identify which, which microbes are present in a microbiome. We need to focus instead on what they're doing in the human body. We need to understand their chemistry. So it turns out that microbes are doing an enormous amount of chemistry. In the human gut. And this is what my lab finds so fascinating about these organisms. Gut microbes can make and modify molecules in ways that human cells can't. They expand the chemistry of, of the human body. They can make their own specialized molecules. They can modify ingested compounds like dietary components and drugs in ways that our cells can't. They can even change the chemical structures of metabolites made by our body. It turns out that just like the composition of the microbiome varies with health and disease, metabolite production does as well. So the chemistry of the microbiome can also be very strongly linked to, to health and disease states. So to understand the microbiome, we need to really understand this chemistry and especially how it happens. So these chemical reactions are carried out by proteins called enzymes. These are protein-based catalysts that microbes and all living organisms use to accomplish chemistry. And my lab really believes that a key to unlocking the secrets of the microbiome lies in actually discovering and characterizing these enzymes. So for example, if we can discover a microbial enzyme that performs a particular type of reaction and the gene that encodes that enzyme, we can actually look at DNA sequences from human microbiomes and make more accurate predictions about what kinds of chemistry the gut microbiome can do, maybe correlate the abundance of those genes with, with disease. Enzymes also provide us with a really potentially exciting way of manipulating the gut microbiome, so as chemists, we can potentially synthesize small molecules that block the activity of enzymes. Indeed, this is what many of the drugs that we take do: block the activity of enzymes, prevent them from working. And if we can inhibit the activities of these microbial enzymes, we now can have powerful tools to study the role of microbial chemistry in microbiomes and potentially even You could imagine developing drugs that actually target the microbiome and may help to treat or prevent disease. Over the last almost 10 years now, my lab has investigated many different types of gut microbial chemistry. We've studied how microbes consume nutrients derived from food in the human host, We've studied how microbes can reduce the effectiveness of drugs used to treat heart failure and Parkinson's disease. And finally, we've also studied how microbes make unique complex molecules called natural products that can actually cause changes to our own cells and may lead to cancer. And so I thought today I would tell you a brief little vignette from each of these research themes that really emphasizes the different things that we've learned about the chemistry of the microbiome. So first, in our investigations of gut microbial nutrient consumption, we've learned that gut microbial chemistry is often distinct from our own. And one great example of this is the metabolism of the essential nutrient choline. So this is the work of two talented former graduate students in my labs, Miranda and Ben, and a former postdoc, Spencer. So this story begins with the essential nutrient choline, which is found in foods like eggs, meat, and soy. Choline plays many important roles in the human body. It's really important for cell membrane, um, uh, structure and liver function. So it's really important that we consume choline in the diet. Uh, but we aren't the only ones who can consume choline. Gut bacteria can also metabolize choline, and it turns out they do it in a completely different way than our own cells. They actually convert choline to a molecule called trimethylamine, which is the substance that causes rotting fish to smell. Now, trimethylamine production by the gut microbiome turns out to have been strongly linked to a variety of different human diseases, including heart disease, liver disease, and diabetes. But before my lab got interested in this pathway, we didn't know how microbes actually did this transformation. So in starting this project, my lab actually figured out the gut bacterial enzyme that's responsible for converting choline to trimethylamine. This enzyme turns out to be a member of a large group of enzymes called the glycyl radical enzymes. So our enzyme, CUT-C, is a member of a larger enzyme family. Now this slide is meant to give you a sense of the, the diversity of the different chemistry this enzyme family can do. So these are very special enzymes. You've probably heard of radicals and how they can do damage to cells. Radicals are species that contain unpaired electrons. They're very reactive. It turns out that these glycyl radical enzymes have evolved the ability to harness and control the powerful reactivity of radicals to do useful chemistry. Now, humans don't have glycyl radical enzymes, but it turns out that they're among the most abundant enzymes in the human gut microbiome. So many different types doing many different reactions. So again, our enzyme, CUT-C, was a new member of this family that metabolized choline. So we really wanted to understand it in detail to understand exactly how it differed from the rest of the family, what made it unique. So we were able to obtain the enzyme in purified form, study its chemistry and structure in detail. So I wanted to share a crystal structure of the enzyme that we obtained with Kathy Drennan's lab at MIT. So this crystal structure shows choline in blue bound in the active site of the enzyme, which is shown in green. When we analyzed this structure and did experiments based on the structure, what we found was that there were certain amino acids in this enzyme that were found in all other glycyl radical enzymes. They were shared among the family, and they're very important for doing the radical-based chemistry. We also found some amino acids here highlighted in red that are totally unique to our enzyme CUT-C, and we think play a really important role in recognizing choline as a substrate and promoting the reaction with this particular molecule. Now, this turns out to be really useful information. We can use our deep understanding of the chemistry of this enzyme and the specific roles of these parts of the protein scaffold to really help us identify this enzyme in microbiome sequencing data. So this is what we next set out to do. Given that choline metabolism is so strongly connected to human disease, can we actually find genes encoding this CUT-C enzyme in patient microbiomes? Is there perhaps an increased abundance of this enzyme in individuals who have heart disease or liver disease. This turned out to be a little more challenging than we expected, as it turns out. Because the glycyl radical enzymes are such an abundant and diverse family within the human microbiome, telling different members of this enzyme family apart from one another when we looked at DNA sequencing data turned out to be really challenging. So this made it difficult to accurately identify choline-metabolizing CUT-C enzymes in the microbiomes of patients and investigate their connection to disease. So to tell you how we solved this problem, I'd like to take a little detour and actually introduce you to my family. So this is me and my 3 siblings. And it turns out that human families are actually a lot like enzyme families. So just like enzyme families, we are related to one another phylogenetically. And just like enzymes, we also have very different functions or jobs, including my little sister who's a grad student. She's kind of like an enzyme of unknown function. We don't really know exactly what, what her activity is going to be yet. So now for a moment, imagine that I asked you to predict our jobs from these photographs. If you were to consider just the information that might indicate we were related to one another, for example, our faces, this actually might be really challenging. Instead, what you'd like to do is, is to actually look at some of the tools we do to use to do our jobs. Features of our photographs that more directly relate to our functions in the world. And it turns out that this is what we needed to do to better detect enzymes in our microbiome sequencing data. Instead of just looking for elements of the sequence that told us that the enzymes were related to one another, we really needed to focus on the unique features of the sequences that were more directly related to enzyme function. So we developed a method to do this bioinformatically, and we were then able to look at gut microbiome sequencing data, apply our very specific knowledge of the chemistry of this enzyme, and really improve the accuracy with which we could identify choline-cleaving glycerol radical enzymes, differentiate them from the many other members of this enzyme family. And so when we were able to then actually do this analysis with patient microbiome sequencing data, we were able to show that patients with liver disease actually have higher levels of CUT-C in their gut microbiomes than healthy individuals. So this is just a start, but it potentially supports the proposal that gut bacterial choline metabolism may be involved in liver disease. Okay, so switching gears now to our studies of gut microbial drug interactions. In this field of our work, we've really come to appreciate the variability of gut microbial chemistry. So as I'm sure we're all very familiar with, with the variable responses of different individuals who are receiving the same drug. This is a big problem for patients, and the factors underlying this variability still remain a big mystery. But my group really believes that part of this may actually involve the gut microbiota. So, we know, for example, in many cases, gut bacteria can actually consume drugs that we ingest orally. Before they can actually enter the bloodstream. So this can mean that the structure of the drug is transformed, changing its activity. This can mean that less of the drug actually gets to where it needs to go in your body. So we've been studying a number of different drugs, but today I wanted to tell you a little bit more about our efforts to understand gut bacterial metabolism of the Parkinson's disease medication L-DOPA. And this is the work of a former graduate student, Vayu. Now, it turns out that's been known for a very long time that there's a great deal of variability in how patients respond to L-DOPA, both in terms of how much drug is required for efficacy, but also differences in side effects that result from taking the drug. Now, we got very intrigued by the idea that perhaps microbial metabolism might be involved in this effect. So, it actually had been known for many years that the microbiome as a whole could transform the structure of L-dopa. So, the community as a whole was known to perform a reaction called decarboxylation in the gut. Now, this reaction is actually very problematic for Parkinson's patients because it actually prevents L-DOPA from entering the brain, which is where it needs to act. Metabolism of L-DOPA outside of the brain is also associated with unwanted side effects. So, when we started this project, the first goal we had was to identify the organism and enzyme that were responsible for this reaction, and we were able to do that. We were able to link this chemistry to an organism called Enterococcus faecalis, and a specific enzyme in its genome called tyrosine decarboxylase. So once we'd identified the enzyme, similar to our work with choline metabolism, we wanted to actually try to detect this gene in the microbiomes of Parkinson's patients. When we did this, we learned that there was actually quite a variable distribution of the tyrosine decarboxylase enzyme and E. faecalis in the gut microbiomes of Parkinson's disease patients. So only a subset of patients had this organism and had a microbiome that could metabolize L-dopa. So this variability is really striking given the variability in patient responses and side effects that are already known. So we're wondering whether in at least a subset of Parkinson's patients, microbial metabolism of the drug might be really limiting the amount of L-DOPA that can enter the brain. But now, of course, the question is, can we actually use this knowledge to try to benefit patients? So, one thing we've begun to do is actually think about how we might prevent bacterial metabolism of L-DOPA using small molecule inhibitors. So, using our understanding of the chemistry of the E. faecalis tyrosine decarboxylase enzyme, we've been able to identify a small molecule that interferes with this activity. This molecule is called alpha-fluoromethyltyrosine, or AFMT for short. You can kind of tell from the chemical structure it's mimicking L-DOPA in some ways. Basically reacts with the enzyme and destroys its activity. So, when we gave this molecule AFMT to mice along with L-DOPA, it actually boosted levels of L-DOPA in the blood when it would, in these experiments suggesting that, you know, if we can develop a compound like this, we maybe, we might be able to improve L-DOPA therapy for some Parkinson's patients through drugging the microbiome and preventing this type of metabolism. All right. So, finally, In our efforts to investigate how gut bacteria make unique disease-associated molecules, we've really come to appreciate just how complex the chemistry of the microbiome can be. So this product that I'm going to end with was the work of a large team of lab members, including former grad students Carolyn and Lee, and former postdocs Matt, Yaniv, and Paul. So to me, one of the most intriguing potential links between the gut microbiome and disease is the link to colorectal cancer. So not, not only are our gut microbes in very close physical proximity to these types of tumors, but also colorectal cancer rates have been dramatically increasing over the last several decades in younger individuals. So this rate is actually too fast to be linked to any change in human genetics. So it must be some sort of environmental factor. And there's a lot of reasons to believe that it might be something to do with the gut microbiome. For example, there are many gut microbes that have been found on tumor tissue and been correlated with the presence of cancer. So my lab got really intrigued a few years ago when a set of bacterial genes was correlated with colorectal cancer. So, so these genes, which are called the PKS gene cluster shown here, are grouped together in the genome. And when we analyze these genes, we notice that they were predicted to encode something known as an enzymatic assembly line. So just like we use assembly lines to build cars and other complex machines, assembly line enzymes take very small molecules and assemble them into more complex molecules step by step. So when these gut bacterial enzymes were discovered, it was proposed that they worked together to produce a complex toxic molecule called colibactin. And because these genes were associated with cancer, there was a proposal that maybe colibactin might be playing a role in cancer initiation or progression. But it was really hard to test this hypothesis because still no one has been able to isolate and actually characterize this molecule. So we actually don't know anything about what the chemical structure of colibactin might look like. So when we started this project, we decided to kind of take a really unconventional approach. We decided to go ahead and start studying colibactin anyway, even though its chemical structure was unknown. What we thought we might be able to do was to learn about colibactin's structure and activity by actually studying these assembly line enzymes that build the molecule. Perhaps if we studied these enzymes, we may be able to get clues about what colibactin looked like and potentially how it might be causing toxicity in cancer. So eventually, through, through many years of effort, we learned enough about colibactin and these enzymes to be able to predict one possible final structure of the colibactin toxin. Now, there are a few important things to note about this, this structure. First, it's obviously very large and complex. Second, it contains two cyclopropane rings. So these are three-membered carbon-based rings They are very small and strained, and they are often chemically reactive. So once we realized that there were cyclopropane rings in colibactin, we wondered whether these structural features might actually be allowing colibactin to react with DNA, causing DNA damage and ultimately toxicity. So to test this hypothesis, what we did was to actually expose human cells to colibactin-producing bacteria and then try to use various chemical techniques to look for pieces of DNA that had undergone reaction with colibactin. And it turns out we were actually able to find such a product. Shown here on the slide, and this really confirmed that colibactin was, was capable of reacting with and damaging DNA. Now interestingly, we were able to detect this same product when we exposed mice to gut bacteria that make colibactin, so this suggests that the product of this reaction between colibactin and DNA might potentially be a biomarker for exposure to the colibactin toxin. And this finding could actually be really useful when we think about studying colibactin's connection to cancer in humans and also for diagnostic purposes. So this could potentially allow us to monitor the activity of colibactin-producing bacteria in patients it may potentially predict future risk for developing colorectal cancer. So more broadly, our studies of colibactin have revealed that our own gut microbes have the potential to damage the human genome, compromising its integrity if this damage isn't repaired correctly. So I think this just, this further reinforces the idea that some gut microbial chemistry may have the potential to contribute to disease. All right, so where is this work taking us next? Well, one of the big future goals we have for this research is to explore the development of drugs that actually target the chemistry of gut microbes. So these would be different from drugs that we take that manipulate our our own cells. They would also be different than existing antibiotics, which, which kill both pathogens as well as many gut bacteria. So our idea that would be, you know, rather than killing these organisms, we would just like to prevent them from performing specific chemistry, for example, inhibiting their ability to metabolize choline. This might provide new ways to treat disease that might be complementary to host-targeted approaches and also might be able to manipulate the microbiome without harming the many beneficial organisms that are present in this community. So my dream is that someday, in addition to taking drugs that manipulate processes in your cells, you'll also take medications that have been specifically designed to treat your gut microbes. But I want to also emphasize that understanding gut microbes and their chemistry could lead to more than just new medicines. There could be diagnostics that are microbiome-based that could help us to diagnose and monitor disease. We might be able to design personalized diets that promote beneficial gut microbial chemistry. We can even imagine structuring the environments we live in to enable beneficial interactions with microbial diversity. This is the future I think that could emerge when we start to think of ourselves not just as individuals but as ecosystems. All right, so I'd now like to wrap up by just reemphasizing what I think the key take-home message of this talk is. There's a really large need to expand our understanding of the chemistry of the microbial world. Right now, when we look at microbiome sequencing data and we try to generate hypotheses about what these communities might be doing, if we use only the chemistry that we, we understand, well, it's a lot like looking for lost keys under a single streetlight. We really only have a small amount of knowledge about, about what these organisms can do. To truly understand these communities, We need new ways of studying microbes and microbial communities that illuminate and discover new microbial chemistry. And my lab is very excited to continue our efforts and continue to work closely with biologists to pursue this goal. So I'd like to end by thanking the lab for their hard work and many contributions to to all of the science that I've talked about. This is us this summer over Zoom. Thank you all for listening and for the opportunity to share our science, and I'd be happy to answer any questions.

Thank you so much, Emily. And we do have questions that are coming in, but before I start diving into those, I wanted to ask one really quickly. How did you get interested in studying the chemistry of the gut microbiome?

Yeah, that's a great question. So, as you might have alluded, or might have, you know, kind of been thought from the beginning of the talk, I am actually a trained, originally was a trained organic chemist. So, I myself was trained to synthesize molecules in the lab. As I started to really have a great appreciation for how hard it was for synthetic chemists to make molecules in the lab, I got interested in how nature made molecules. And then, you know, was drawn to microbes because they make some of the most beautiful natural products and do these beautiful complex molecules. And that had been so hard for me to make in the lab as a chemist, I wanted to know how they did it. And I think that the timing was pretty great. And when I started, my postdoctoral research after graduate school, that was right when the Human Microbiome Project was beginning to sequence the microbiomes of healthy individuals. And in the context of just following that work, I think I started to pretty quickly realize that there was a lot of chemistry that we just didn't understand. Most genes, 85% of the genes in the healthy human gut microbiome can't be linked to known activities. So it seems like a really great opportunity for someone who is interested in microbial chemistry.

Actually, that's a great answer leading into one of the questions that came in, which is whether or not anyone has ever cataloged all of the enzymes— I'm assuming the question means both bacterial or microbial and host— that are present in the human gut.

So there have been a number of studies that have attempted to assemble gene catalogs So through sequencing the microbiomes of many, many people, trying to identify all of the genes that are present in those collective habitats, and you can ask some really interesting questions like, are there genes that are shared, genes, enzymes that are shared by everyone? What fraction of the genes are, unevenly distributed what fraction of the genes are found in microbes that we have cultivated. So all of those studies have kind of reinforced something that we knew even from sequencing the genomes of organisms like E. coli. There's a huge fraction of genes in microbes in any organism that have unknown functions. So in the case of E. coli, I think it's something like 30 to 40% of the genome consists of genes of unknown function. So that's true of other microbes, the numbers are even higher. So yeah, so we do have a sense of the scope of the genetic potential associated with microbiome. As we sequence more and more people, that catalog of genes just gets bigger, but the problem of not understanding what those genes do also just gets bigger and bigger.

Great, thank you very much. And then another question that's come in, and you've touched on this a little bit, I think relates to diversity in some ways. So how similar is your microbiome to my microbiome to other people's microbiomes? And do you ever find cases where you have people that have a completely unique gut microbiome?

So actually, in some ways, everyone's microbiome is completely unique. So there are no two people who have the exact same set of genes, the exact same set of species in their gut microbiome. That was one of the really kind of surprising initial findings from the Human Microbiome Project. So most of the comparisons of microbiomes happen using genes that are identifying microbes. So they're comparing the composition of the gut microbiome. And if you just look for those genes I mentioned that are the markers of bacterial identity, and you look to catalog the species that are present in different individuals' microbiomes, that's where you see this really enormous variability. One thing that's kind of interesting is that if you look at more genes than just the markers of identity, so if you start to look for, look at the fraction of genes that we can kind of tentatively assign a function, or say that they're enzymes or they're involved in DNA synthesis, protein synthesis, if you compare our microbiomes using that metric, by looking at all of the genes we can assign a function, then we look more similar, which is kind of neat. This was done with the Human Microbiome Project data across healthy individuals. From a functional perspective, healthy individuals looked more similar. So one hypothesis to explain that observation is maybe that the healthy human microbiome consists of a core set of functions or activities that the microbiome supplies. And maybe different organisms can supply those functions in different people. But, you know, so different people's communities, even though they look different based on species composition, maybe functionally they're doing some similar things. So I think that's a compelling hypothesis, although I will also point out that certainly for some activities we have studied, there's definitely variability between people. So I think it kind of depends on what type of chemistry, what type of activity you're kind of thinking of.

Okay, great. One of our viewers had another question about what you were saying about colorectal cancer, noting that you mentioned that colorectal cancer is on the rise. Do you find that the bacteria that produces Colibacterium is also on the rise?

So this is a very interesting question. So is, is the abundance of colibacterium-producing bacteria growing? So there are some studies that suggest that yes, colibacterium-producing bacteria are more frequently found in human gut microbiome. So, so colibacterium is made by a number of different bacteria, including certain strains of E. coli. And the strains of E. coli that make colibactin are from a very specific phylogenetic group. And even before colibactin had been discovered, it had been noted in some previous studies that that phylogenetic group was appearing more and more frequently in individuals from certain Western nations, US and Canada, sort of within a 30-year timeframe that kind of nicely corresponds to the timeframe in which we've been seeing this increase in colorectal cancer incidence. It's also really well established that if you move, you actually immigrate from certain places of the world to the US or Canada, your risk for colorectal cancer changes within a generation, really speaking to some very dramatic environmental factor that's at play.

Wow, that's incredible. So one other question that's coming from one of our viewers, given that we are in the middle of a coronavirus epidemic, have you seen or have your colleagues seen any impacts of the coronavirus on the gut microbiome? So— Or is it too early to tell?

Yeah, I mean, so I know that there are studies examining that. I think many people are interested in whether the gut microbiome, and particularly its ability to prime and train the immune system, whether that might affect our response to coronavirus. And certainly in the context of developing coronavirus vaccines, vaccines, gut microbiome is going to be very interesting because, again, kind of the gut microbiome's ability to interact with the immune system is associated with vaccine efficacy in other types of studies of other types of vaccines. So, yes, I think there are some opportunities to study the microbiome in the context of problems that are very relevant right now.

And sticking with the theme of viruses for a second, this is the question that I had come to mind, was you've talked a lot about the gut, the bacterial portion of the gut microbiome, but there are viruses and fungi in there as well. Do they have any chemical impact on us?

Yeah, yeah. So that's a great question. And I think that the fungal and viral components of the microbiome have not been as well studied as the bacterial components, and that's not great. So I think that's starting to change, but certainly there's a lot of evidence to suggest that those additional components of the microbiome are very important. One great example that relates to fungi is actually inflammatory bowel disease. So if we look at the host genetics associated with Crohn's disease in particular, we see a lot of associations between genes that are part of our innate immune system, and specifically the part of our innate immune system that recognizes fungal molecules. So parts of the fungal cell wall. And that suggests that something that could be going awry in Crohn's disease is some sort of impairment in terms of how our immune system recognizes fungi in particular. But we don't know that much about how fungi change in the context of Crohn's disease, or even, you know, kind of, we don't know that much about their functions within the microbiome and their chemistry. So that's a great opportunity for further investigation. And with viruses, I'll just maybe throw out there that, you know, certainly there are, if you sequence all of the DNA associated with the gut microbiome, you can find a lot of viral sequences, not just viruses that might interact with human cells, but viruses that attack bacteria. So they're known as bacteriophage, and they can play a large role in shaping microbial communities, but their role in the human gut microbiome is not well understood yet. That's another really exciting area for further study. We don't know that much about how our body reacts to phage in terms of how our immune system reacts to these viruses that are present in us but are actually targeting bacteria.

One of our viewers is interested in the microbiome and antibiotics. And whether long-term antibiotic treatment has any kind of impact on the microbiome.

Yeah, so, so that is a great question. So there have been lots of studies focused on antibiotic use and the microbiome. So, so I think, you know, there's definitely changes in the microbiome that can occur upon taking antibiotics. One of the things that we know can happen is that you can impair the microbiome's ability to exclude other pathogenic microbes. So one of the most famous examples is Clostridium difficile, a really nasty anaerobic human pathogen that people tend to actually get C. difficile infections when they've been especially hospitalized and are taking antibiotics. Because by killing members of our healthy gut microbiome, we actually kind of open up a niche, open up space in the microbiome for the pathogen to occupy and start to grow and flourish. Studies have looked at your sort of microbiome recovery after treatment with antibiotics, and so there is some, some good news, I think, to be had there in that, you know, certainly there's been plenty of studies that do show the microbiome can, you know, can recover after taking antibiotics. But plenty of animal studies also show that especially with repeated antibiotics and especially early life exposure to antibiotics, there could be some evidence of increased disease risk later on. So I think that there are many reasons to think about being very careful with how we administer and manage antibiotics, and collateral damage to the microbiome is certainly one thing to consider.

All right. Are there any other ways that the microbiome and and us interact besides enzymes? Are there other kinds of biomolecules that signal back and forth or have any kind of activity?

Yeah, so that's another great question. So essentially every type of molecule associated with a microorganism is built by an enzyme. So I think that is important to remember. But in terms of molecules, maybe beyond small molecules, the sort of macromolecular structural features of bacteria and other microbes are really important in terms of how they interact with the host. So these types of molecules, things like parts of the bacterial cell wall, the bacterial flagellum, the sugars that coat the surfaces of bacterial cells and membranes. These are the types of molecules that our immune system recognizes and reacts to. And so those are much larger than the molecules that I showed today, but they are very important in terms of interactions. Small proteins and peptides are also things that we're learning that microbes can make and release and are ways that they might signal to host cells. So there's a wide diversity of different mechanisms by which these organisms can potentially be signaling or interacting with us.

And as we're talking about all these different kinds of molecules, What amounts are we talking about? Like, what are the concentrations that are involved here? I mean, I assume it's not gallons of enzyme that are getting released into our gut on any given day. So what are we talking about in terms of amounts?

Yeah, so it varies from molecule to molecule. So some molecules are bioactive in very small, very low concentrations, so they don't need Colibactin is an example of that. You don't need to make a lot of colibactin to sort of see a biological effect. Other microbially produced small molecules accumulate and circulate at concentrations that are as high or higher than some orally administered drugs. So that does kind of give you a sense of the concentration range we can be kind of thinking about. But yeah, I would say there's a huge variation in terms of concentrations of bacterial metabolites.

As you were talking about potentially coming up with ways to counteract the effects of microbial enzymes on the gut so that you could ingest, you know, have better, you know, influence of drugs, or, you know, be able to take in more of the nutrients and drugs that you need. Do you think— this is a question from a viewer, um—

do you—

you know, we all know the microbiome and bacteria in particular are pretty adaptable, um, to any kind of external influence. Do you think that the microbiome would adapt to whatever treatment you would come up with?

Yeah, so that's a really good question. If we're developing these, um, inhibitors that, that aren't intended to kill the bacteria but are intended to change how they're growing, is there a likelihood that we'll develop resistance? So I think that depends on, you know, really how essential the process we're trying to manipulate is to the organism. One thing that maybe gives us some hope is that, you know, a lot of the organisms that live in the gut environment are metabolically very versatile. So they will be able to metabolize hundreds of different substrates to support their growth. And so, you know, one might hope that we could, you know, block one pathway that might be involved in supporting microbial growth, but then, you know, as long as there is another another niche that's open, and as long as there's another source of nutrients that's present, that, you know, these organisms would just shift their metabolism. Now, we're starting to test this with some of the molecules that we've developed already, and we're— you know, it's very early days, but we are having a hard time raising resistance. To these compounds. So hopefully, yeah, hopefully we will be able to— again, if we're not targeting processes that are essential under all conditions, we may avoid that selective pressure to evolve. But we'll see.

Okay. We had a question come in that I think I think may have to do with the educational pathway or the research pathway one might take to do the kinds of work that you do. What aspects of organic chemistry do you think are most useful to use and to know if you're trying to decipher the chemistry of the human gut microbiome?

That's a really interesting— that's a really interesting question. So I actually teach the second semester of organic chemistry at Harvard. It's called The Organic Chemistry of Enzymes. And I think that there's a kind of a core set of organic chemical reactions that are really important for central metabolism. And I think that, you know, really kind of understanding those I think is very, very important. And I think there's obviously also more exotic chemistry happening in the microbiome, so colibactin is a great example of a really complex molecule. But I think it kind of depends on what types of metabolites and activities you're interested in. I mean, I think At the end of the day, what is probably most important is the biological effects of these molecules. And a lot of very simple molecules can have very profound biological effects. So yeah, I think organic chemistry is, I see it as as being a critical part of understanding biology. I don't know if everybody would agree, but I think that certainly for us, having an understanding of reactivity and a knowledge of organic chemistry, it allows us to make predictions about what enzymes might do. It also allows us to start from a new transformation that may not have an enzyme assigned to it. Generate hypotheses for what type of enzymatic chemistry might be possible. So I think we've found it to be very, very useful in many different contexts.

So in the couple of minutes that we have left, I wanted to throw one more question to you. Where do you see our understanding of these interactions between the micro— our microbiomes and us going in the next 5 or 10 years?

Wow, yeah, that's a, that's a really fun question to think about. You know, I think that we're going— our knowledge is only going to increase. I think that there's— back when I started our lab almost 10 years ago, the microbiome field was not thinking that much about chemistry. They were very much at the stage where the sequencing aspect of this was still very new, and there was a huge effort to just sequence as much as possible. And now I think that the microbiome research community has kind of realized that there's a limitation to how much mechanistic information you can really get out of that sequencing data. So it's been really fun for me to see that a lot of people in the microbiome space are now starting to get very interested in metabolites and in the chemistry microbiome can do. So for example, there have been in the last year two large-scale screens of gut microbes for drug metabolism that revealed hundreds of potential new metabolic events that we didn't know were happening. So I think we'll see more of that. But what I hope we'll get to in the next 5 to 10 years is really kind of fully integrating this chemical knowledge with some of the big data, and especially with clinical studies. So to really design clinical efforts that will incorporate efforts to also characterize chemistry and metabolism. And by that time, we may also have a better understanding of this concept of drugging the microbiome. I think microbiome-based therapeutics, whether they're small molecules, organisms themselves, food for microbes, prebiotics, these things are all kind of progressing towards really being applied clinically, and hopefully within that time we'll really know what types of microbiome-based therapeutics are working, and hopefully we will also have more information about how they're working. And that mechanistic information I think will then be really useful in kind of deciding how do we move forward and develop more therapeutics that could target these organisms.

Well, thank you so much, Emily, for your wonderful talk, for answering all these questions from our viewers. And to the viewers, thank you to all of you for joining us this evening. We hope you enjoyed tonight's talk. If you like what you heard, then consider watching some of the other talks we in the Science for All Seasons series. You can find all of them at broadinstitute.org/sfas. A recording of tonight's talk will also be posted to our YouTube channel at youtube.com/user/broadinstitute. To see this recording and to find out about other livestreamed events and recorded videos that we've made, consider subscribing, and please tell your friends and family to subscribe as well. Thank you so much again for joining us this evening, and have a great night.