Please now via live we’ll just wait for a few minutes so that all some audience in join like for two minutes or something then we can continue. Okay okay we’ll start things Hi everyone, Welcome all to the fifth chapter by battalion webinar series. Today with us is a remarkable and distinguished scientists Dr. bareback Khomeini, Assistant Professor of Boston College. Before we start, let me take the liberty to speak a few words about us. Dr. pebax is trained as an electrical engineer. He attended Shattuck University of Technology events, where he received his bachelor’s in electrical engineering, and a master’s in microwave and optical communications. He then moved to Georgia Institute of Technology Atlanta, where he received an MS in physics and a PhD in electrical engineering in 2007. his doctoral training at the Georgia Institute of Technology involves designing and implementing compact on chip for Tony de multiplexes. He continued his work at Georgia Tech as a postdoctoral fellow to build spectrometers for biosensing applications into the bio biological systems. He joined optishot lab at Fred Hutchinson Cancer Research Center, Seattle, Washington in 2009, where he studied the stability and spatial organization of microbial communities in synthetic communities. He joined the Department of Biology at Boston College in 2016. As an assistant professor, Dr. bareback and his team combine theory and experiments to study microbial communities that affect our health and environment. This current lab project includes examining the stability of microbial communities, characterizing and controlling microbes in human nasal cavity and repurposing bacteria and their insights for efficient removal of food contaminating substance. Sir, it’s a pleasure for us that you took the time to be here with us today. Also for the audience, if you have any doubts or questions regarding the presentation, please post your questions in the comment box. And I will be moderating it at the end of the presentation. So I welcome you Dr. bareback to the fifth chapter of vibe Italian webinar series. Over to you Dr. Beck. Thank you very much for the introduction. It’s my pleasure to talk about some of the research that’s going on in Purdue Owl. So today, I’m going to do a more of a survey of different types of projects that are happening in the lab, and all of them come within this umbrella of harnessing microbial potential. Before starting, I would like to thank people in black, who actually did most of the work is our postdoctoral researchers and graduate students as well as several undergraduate researchers who contributed to this work. When we are talking about harnessing microbial potentials, the idea is that there are these microbes around us that are very diverse with many different metabolic capabilities. And they are present in several different environments. And that makes it likely that if we are looking for solutions to some of the current challenges, we can potentially find the solutions already within the microbial world. In our lab here in particular, focusing on two directions. One is microbiota based therapy, and by that I mean the use of microbes to prevent or treat diseases. And the second direction is microbial bio remediation, which means the use of microbes to remove waste and toxins. They should give you this disclaimer that there is a lot of breadth in this presentation, but not a whole lot of depth. So if anything is unclear, please feel free to ask questions, and I’m happy to clarify things. So we’ll start with microbiota based therapy. And the idea comes from this observation that microbes that are associated with us offers several benefits. And that includes modulation of our immune system so that it can work does its job properly, as well as protecting us from pathogens, there are additional benefits, for example, they help us in digestion or provide some of the vitamins as well. And because of that disruption of these microbiota can lead to diseases and examples of that are things that happened with eczema or colitis. But if we accept that as a fact, then we can also potentially intervene and manipulate the microbial community to prevent or treat some of these diseases. So that’s the advantage of manipulating microbiota. But there’s an additional advantage that it reduces our reliance on antibiotics, because if we are using other microbes to treat infections, we don’t need to use antibiotics. And that’s helpful because currently we are facing the rise of antibiotic resistance, which can be a major challenge in in the coming years. But if we want to do microbiota based therapy, what are the challenges it sounds like a, like an easy task. The main challenge is that microbes are complex ecosystems. There are a multitude of species present, there are several unknown interactions, either among microbes, or between microbes and their hosts that are not characterized and not well understood. And the environment is complex. It’s temporally and spatially variant and heterogeneous. So it’s really hard to know what is actually happening in that environment, if you want to control. Another challenge is that the rules for intervention are not completely understood. So we, there are some intervention strategies that are currently being used, for example, probiotics are being used as a way of bringing back the microbial community to help or there is fecal matter transplant that in some cases of infection seems to be very beneficial. But in all these cases, we don’t fully understand what is happening. And because of that, there are many cases that probiotics work don’t work, or several cases of fecal matter transplant has unintended consequences that we didn’t plan and could potentially be really harmful. So there are even cases of this after fecal matter transplants. And we think the reason for that is that these are only partially successful because we don’t fully recognize and appreciate the complexity of the situation. And with those challenges, what was the things that we can do? Today, the approach that we’re taking starts with formulating predictive models to capture the properties of interested in microbiota. And the idea is that if you have a predictive predictive model, then we can start from there and navigate the system towards the state that we are more interested in. The second step that we have taken is to find a system that’s not too complex, and not too simple, as a way of allowing us to do mechanistic studies the microbiota. And goal of this step would be to ensure that the models that we’re using are realistic, and also to test and validate the hypothesis that we come up with. And then the third step is to search for general principles that can guide intervention strategies. So rather than focusing on specific cases and specific situations, we are looking for principles that could inform and inspire how we should intervene and modify microbial communities. So let’s get started. The modeling point for modeling community is that traditional approach has been to use less cultural technology. And these are models that have a long history of being used in the context of ecological communities. They are relatively simple in the sense that regardless of the mechanism, they abstract all the interactions into a fitness effect between different species. Let me explain what I mean by that. So let’s assume that you have two species. These types of models just tell you that depending on how many of the species one, you have present how much a species to benefit from them. Or depending on how much of the species two are present how much a species one, for example, gets inhibited. And the history comes from the interaction between links and here and prey predator injection, this is that. So these models have a long history, and they have been successful in capturing some of the dynamics that happened, because logically, so in this particular case, there was data from 1840s to 1930s, that shows oscillations between populations of Harrison. And this simple alternative model was capable of capturing these oscillations. So people got really excited, it was a really simple model, but it was capable of giving realistic predictions about how the populations would fluctuate in nature. There are several advantages. One is that there is no need to know interaction mechanisms, when you’re forming those alternative models. They are easy to estimate parameters, all you need to do is to measure population densities, which is a fairly easy thing to do, especially these days using sequencing, you can easily measure the population composition, there is some empirical support for it. So there are cases that they have worked well. And they are easy to extend to multi species committees. So all you need to do is to take pictures of a species, calculator interactions, and do that for all different periods within the community, and then put all of those models together and form a network of interactions that describes the community as a whole. So this all sounds good. But this type of modeling applies to microbial communities. So that traditionalism actual ecology of non larger species. So to answer that question, we started from a simple thought experiment, we said that if the has a ground truth, and we know exactly what happens within the system, and if we formulate a pairwise model of loss as an approximation for that model, and we simulate the dynamics using both models and compare predictions to the class, do they agree with each other or not? And what are the conditions that they agree and what are the conditions that they do? So this line of question tells us tells us whether this type of model could in general work or not. And instead of using their realistic or illogical system, we are using a mechanistic model as a reference, because in this case, we know exactly what is happening within the community. And we are not bound by experimental errors or measurement restrictions. So we did that. For our mechanistic models, we formulated a mediator explicit model. And by that, what I mean is that instead of only looking at the interactions between species, we also include the chemicals that mediate those interactions. So this can cause could be metabolites or toxins or antibiotics. And in our model, we assume for simple types of interactions. The fact that they show us an arrow up here tells us that the species is removing chemical see from the environment. So this could be for example, consumption of the metabolites. This type of arrow shows that species S is producing a chemical in the environment. So this could be something that gets secreted, or even potentially something in the environment that gets converted into C. And then in return, chemical C can either stimulate the growth of this, or it can inhibit the growth. So these are fairly simple types of links. But they capture the majority of cases that we can think of that species are injecting through chemicals or metabolites in their environment. And then we take these links, put them all into a single model with a certain species, a set of chemicals, and each of these species produce or consume a subset of chemicals. And each of the chemicals either positively or negatively affect the substance which and that’s our ground truth as our reference model that could represent a realistic microbial community. And then we asked a few simple questions. First question was can the last overthrow type model represent different types of microbial interactions and for that, we looked at two simple types of interactions one that we called reuseable in the sense that the chemical is reusable in this interaction species one is producing a chemical species two is getting affected by it, but the species two which is the recipient does not change the concentration of chemical plants. So it’s not consuming it or removing it from the environment. In contrast, you can have consumable interaction that one species is producing a chemical and the other species is consuming it at the same time getting infected. So, an example of reusable attraction could be made even species one changing the pH in the environment, and that change in pH can affect species two. And in the second case, we can have a species one producing the metabolites. For example, an organic acid and a species two is consuming that organic acid and may be benefiting from it. And what we saw was that if you formulate the models, pairwise model for these cases, you end up with two different types of instance, equations. And these two types are in are not interchangeable in the sense that if you are simulating a reusable type of interaction, you can’t use the equation for that consumer type of interactions and vice versa. And what that tells us is that the simple look of ultratech model which is shown here cannot capture different types of microbial interactions even something as simple as metabolites in the environment it gets consumed is a properly approximated and does not. And if we look back, this is almost trivial. In this case, if species one has produced a chemical, it gets distributed among recipients, whereas in the other cases, it affects all the recipients without getting distributed among them. So you would expect that the number of recipients does not affect the injection in this case, and it affects the injection in this case. So we naturally would expect that these two would not be consistent and that the list of attack model is not adequate for capturing other types of interaction. The other question we asked was what happens if we go to more than two species and we put together simple committees to see the actual situation in this case, we have three species that are interacting through a single mediator fish’s body is producing it and getting inhibited by for example, its acidic acid, we can have this type of interaction, the species two consumes c one and benefits from it and a species three is just being inhibited by. So if we do pairwise interactions, we can come up with this type of network connectivity. For example, in the absence of a species one, there is no chemical one species two and species three are not interacting with each other. But when we simulate the dynamics using the original models shown with solid lines here, versus the attract the local will approximate which is shown by dotted lines here, we see that the dynamics are very different. And I would like to emphasize that in this case, each pairs of interactions work pretty fine. So if we measure the deviation between the maturity model and the mechanistic model, all the pair’s s, one s, two s, three s, one, S, two s, three are pretty good approximations. But when we put the three of them together, they are no longer a good approximation. And the reason for that is the presence of two, for example, is affecting the interaction between this one ministry, because it’s removing the mediator from the environment, and weakens the junction between this one city. And that’s an effect that’s not included in the local search act. In the last couple teletype models, we’re assuming that all the attractions are pairwise and independent of each other. So another failure of two cultures. So what should we do? We know that many interactions among microbes are happening through this chemical environment. So should we just abandon our model and move on with our lives? Or maybe there is more to it. So when we look at the literature, there are several examples that show that show that the look of alternative models appear to work fine. There are cases with community assembly, or some synthetic communities of gut microbiota show that show that local sensor models work just fine. But there are other examples that show that they are not enough. There are examples with again, analysis of gut microbiota data that shows of course, and people have also done some controlled experiments. And they see that doesn’t always work. So we change our approach and ask this question is what are the conditions under which these simple was Google search our problems work? So answering this question with a very, very beneficial because it tells us where we can use the simple model and We cannot use it and we should look for maybe more complex models that capture the actual dynamics. And to answer this question, our solution is to test our ideas in a controlled biological, relevant setting. And see in a situation that we can control the microbial community and grow them in a lab environment, and we can reliably measure them when and why lotto check would or would not work in that situation. So the system that we picked for this study is the Nizam microbiota. And our main motivation for it is that neither microbiota has it has an important impact on our respiratory health. It’s been observed that neither mcwethy can prevent condensation of opportunistic pathogens. And an example of that is simple caucus aureus, which is one of the members of the community. In general, around 30% of the population has several caucus areas in their nose, and even around 2% carry the antibiotic resistant version of it in their nose. And for most people, this is harmless, they are present in that environment, but they don’t cause an infection. But if the host host immunity gets weakened, or the patient is dealing with other illnesses, then stop Misaki serious can become pathogenic, it can cause bacteremia, or blood infection, and infection in several other organs. And that’s an important infection type because it has been shown that cause around 120,000, invasive infections, most of them in the hospital setting and more than 20,000 deaths in the US. It’s a global issue. But unfortunately, there are not reliable numbers to Article. interesting for us, it’s been observed that other microbes that are present in the nasal cavity environment, can control the conversation of staff. So that 30% that has the staff persistently in their nose versus 70% that they don’t have it. The thought is that in those 70%, the presence of other microbes has prevented the colonization of suburbia. And that brings this idea that can be restructured the community, maybe even using the members that are already present in that environment to reduce the load of staph. aureus, and reduce the risk of people getting infected with their own microbes. As an additional bonus, neither microbiota has a relatively low diversity. If it is this makes it a good system to study in the lab. So if you sample the human population, around 20 species, explained the majority of the race more than 90% of the reads have come from sequencing and 19 out of those 20 are easily culturable in the lab, which means we can reliably capture most of the things have happened in the case of macrobiotic. Even more if you look at individual persons, you see that in each community, somewhere between three to eight species are done. So each of these microbial communities are not really that complex, they have only a handful of a species press, which makes it a simple model to work with. So we have taken isolates from the nasal microbiota. And the isolates that we took was based on the data that’s available in terms of what species are dominant in that environment. So these graphs show the general that explain the majority of abundance in the sequencing data to see according to bacteria, to the bacterium and staphylococcus are the major, major general in that environment. And if you look at the species or Supra species level, you see some of these species that show up as the very prevalent within the Magneto microbiota. So because of that, we collaborated with the lemon lab now at Baylor College of Medicine. And they were kind enough to give us some of these species that are clinical isolates, and they have come from healthy individuals, but they also represent some of these species that are highly represented within the metro bill within the nasal microbiome. So it took those species and we studied them in the lab. By studying them in the lab, I mean the grow them and look at their gross properties. But we also looked at their interactions with each other. And to calculate these interactions, we use a simple type of experiment. Because because it’s so freely spent media or cssm in which what you do is you grow individual species in monocultures. You sense of use them to remove most of the cells and then filter them out to have self respect media, which of the species and then you expose other species that still freezes, but media and ask how well do they grow in this new environment. And that will tell us something about the metabolite mediated interaction between different species if a species one has maybe removed one type of sugar from the environment, then you see how species two response to the definition of a chicken or if species one has produced something, some byproducts of metabolism as part of their regular activities, for example, some organic acid then you can see how species to respond to the presence of that organic acid in the environment. So what we do is that we systematically do this for all days of a species and measure growth rate and carrying capacity of each of these species in the cell free spent media of other species. And then, important thing that we observed just based on the equations, if that if we do this, the ratio of growth rate your carrying capacity in spent media turns out to be equal to the growth rate, the ratio process gain capacity and in fresh media, it looks also tech model is true. And that gives us a simple test to ask whether rocoto SAP model is a good approximation or not. Because we know that if it is true, when we do these experiments and calculate these ratios, they should stay the same. So just to reinstated, it looks normal tourtech model holds the ratio growth at carrying capacity in selfies and video are super naked, that is independent of the species that are used for this, for getting this cssm. And Sandra graduates, Lindsey lab did all these experiments of looking at pairs of a species. And she saw that in many cases, we we observed this trend to be to be true. So if you look at r squared values, they are pretty high. There is one exception, because there is an outlier in this case. But overall, it seems that in many of these situations, when we look at the interactions between nasal bacteria, the ratio growth rate, the carrying capacity seems to be a constant regardless of what supernatant we’re using. So we started from this point, that last verse, maybe the resource availability explained the strength. And we did the experiments in low nutrient environments first. So we do a typical growth, media for the suspicious to quite a bit. So searching from 100%, we went to 5%, all the way down down to point three 2%. And these low concentrations of the media, we saw that the trends. So what this means is that if you take the media from one species, and you just diluted, both the growth rate and carrying capacity will is still the same way. So this suggests that maybe resource availability could be a good explanation for why we see the same trends when we’re looking at the attractions is simply one species removing some of the resources from the environment and the other species has to grow in the lower nutrient environment. So it just follows this trend. Importantly, when we did the same experiment that high region consideration so this is going from 10% th vi all the way to 100% th by the linear trend did not hold anymore. So whatever trends we are seeing applies to low nutrient environments, but not too high nutrient environments. Which is one more step and if we change individual metabolites or mediators in the in the environment, how how does that affect the growth of different species. So here I’m showing you a subset of that data with some of the compounds that we tested and if you have the species and as you can see, the expensive individual species to the mediators is not necessarily the same between different species. That maybe if you try acetic acid, Staph aureus gets affected by Excuse me, and it doesn’t grow as well. But growing the bacterium to Virgo circle does not get affected by pH much. Or by acidic acid. My sleep state, in contrast, is affected by Ph. But maybe it’s Aquarius doesn’t get affected by Ph. And overall, we saw that the linear trend did not necessarily hold for individual compounds. So if you manipulate one of the compounds in the environment, that’s the guarantee that local church and model would still hope. So we saw that the responses were compounds specific. So it was different between let’s say, acetic acid and micro Meissen, which is an antibiotic. And it was also species specific that not all species responded the same way to different compounds in the environment. So from these observations, or model is that in a low nutrient environment, a combination of effects that could come from removal or addition of resources in the environment, or removal or addition of some of these additional metabolites in the environment, the effect of all of those combined on average average creates a new habitat in which the other species can grow. And that effect falls under linear trend. So from this observation, we went back to literature and asked whether this observation has had support in what people have already explained. And we found this 2015 paper from Lipson, and what he was doing was that he was trying to explain inconsistent relationships that have been observed in terms of ratios are the carrying capacity in previous data. So people have done different experiments. And they have seen that in some cases, they would see positive correlation between race and carrying capacity. And in some cases, they would see negative correlation. So what he did was that he tried to cap capture different situations and came up with this theoretical model that has two different zones. So in low nutrient environments, they call it maintenance energy, so that he thinks that in this case, there is a positive correlation. But if you go to high nutrient environment, there’s a negative correlation. And I’m oversimplifying things. But in effect, what he’s saying is that in low nutrient environments, because the nutrients are so low, any nutrients will be used for producing new biomass. And that contributes equally to both rate and carrying capacity. And because of that, you see a positive correlation between processing carrying capacity real in the donor chain. In contrast, in nutrient rich environments, the soul can pick different strategies. So the soul can grow really fast, but be a little bit more wasteful. Or it can grow more slowly, but take advantage of all the resources and the environment more effectively. And because of that, there is a trade off between how fat cells grow, and how much salt they produce or because of that, you see this negative correlation between growth in carrying capacity. And what we have seen in our experiments is consistent with with this view, and importantly, in this in the range that we are interested in, based on the amount of nutrients that are measured in the base of the environment, we know that we are in the Low Energy Department. And because of that, we think that local water type model could work fine for these types of communities. Moving on from that point to please search for general principles, and one of the principles you’re interested in is this idea of coexistence. So when we talk about some of the diseases, it’s come from dysbiosis, that some of the members that are usually present in our microbiota is a cool way that we will face all the consequences. So we asked ourselves, what keeps them together to begin with? What allows them to coexist with each other? And that’s the main question that we were trying to answer what allows it I’ve been checking macros to correct this, even though on their own, they have a different Growth Properties and needs. And the approach that we took was that we use the model of injections through metabolized mediators. This is the same mechanistic model that they introduced earlier. and restart from an initial pool of Microsoft are randomly interacting with each other, and we simulate those committees over rounds of growth and dilution until we end up with an injury. community that has maybe fewer species, but also species are stable together. And repeat this process 1000s and 1000s of times. And the idea is that if we look at the properties of initial pools of microbes and properties of the final community that gets enriched, then we can get some ideas about what influences coexistence. So the main trends that we observe come from this capability to do many different simulations of the order of 10s of 1000s of simulations and look into this large data set to come up with trends. This is something that you can’t easily do experimentally, just because of the amount of effort that it takes. And from these experiments, the things that we’ve learned, I’m showing one of the examples of the graphs, but there are a few things that they want to emphasize. One is one of their vision was that. So here, we are simulating an initial pool of 30 species and 10 mediators, and simulating how the richness of what’s the frequency of observing communities of different races in the end. And first, let’s compare the one the graphs on the left to the graphs on the right. So these are the same colors to present the same day exactly same community with one difference. In this case, we have the majority of injections to be facilitation or positive influence of species on each other. In this case, we assume that most of the interactions are negative. And you can easily see the difference. So in this case, it’s very unlikely to come up with community that has high richness. Whereas in this case, you can find some cities with high richness relatively easily. So with relatively high frequency, let’s say 10% of so you can find a committee or priestess four, which means there are four different species coexisting with each other. Another aspect that they wanted to emphasize is that, within these examples, you’re also changing the amount of the ratio of average consumption to average production of different metabolites. And you can see that, as we go to higher and higher ratios, the richness in the communities increases. So for example, in this case, as we are increasing so that there is more consumption and less production, we end up with higher court systems. And our interpretation is that first, interspecies facilitation allows improve coexistence. And we see that clearly by contrasting the left hand side with the right hand side. And we also see that consumption of beneficial metabolites regulates the gland balance of the species, because one species, if it’s benefiting from something, and it’s consuming it, it’s also removing that benefits. And that acts as a balancing force to keep the species to get the further question that we try to answer with the idea of colonization resistance. And that means the ability of a species microbiota to prevent colonization of invaders or pathogens. And this is very important for our health. We use probiotics or microbiota transplant, for example, fecal matter transplant. In this context, there is resistance from the microbiota to accept new species. We are in some of these cases, we want to introduce a new species, or in other cases, we want to prevent the introduction of new species, for example, in pathogen and we ask how interactions among species contribute to the success or failure of invention? And to answer that question, we came up with this conceptual essay to to categorize different types of outcomes that you can get. So we are starting from a resident community at the beginning, we’re introducing any major into that community. And we’re asking whether the invader is maintained in environments where it drops out, and whether the residents committee remains intact, or it loses some of its members. And based on these two axes to define the outcome, you can have augmentation, for example, if the new invader becomes part of the community, and everything else stays there, or you can, for example, have resistance if invader goes away, and the initial committee maintains its diversity. You can also have displacement which one of the one or more of the initial species gets replaced by the invader. Or we can have disruption that the invader knocks out some of the residents community but it’s not capable of staying in that environment either. So what we did was that we simulated again, many different examples and quantified the fraction of these outcomes how often for example, So we see resistance versus assumption versus augmentation versus displacement. So I’m showing the frequency of these outcomes over some parameter of interest that is that. In this case, you’re looking at the perfect shoe size, which is the number of invaders introduced into the company. And importantly, what we see is that for a wide range of professional sizes, there is no real change in the outcomes. And that’s something that we have to consider when we are considering intervention strategies. For example, if you’re introducing a probiotic, you want to have more fragmentation, but just adding more of it up to really large numbers like 20% of so which is huge considering how much Microsoft can practically introduce. So we’ve up to that point, there is no real benefit from adding more and more of the probiotics. And only and only at really high concentrations, you see an effect. But even in that case, it’s not contributing to augmentation. Instead, it’s displacing or disrupting the community and losing resistance and then Introducing the new member. As another example, we, again, similar had many cases, but remove different types of interactions. And what we observed was that the mutual beneficial interaction between the invader, and the resident community is the main factor that determines the outcome of augmentation. So if you want to recommendation, instead of trying to add more and more numbers, you have to focus on interactions and create a situation that both the probiotic that you’re trying to introduce, and resident microbe benefit from that interact. So to summarize everything so far. In this context of microbiota based therapy we have done, we have used in vitro nasal communities, along with responding mathematical models, as a tractable system to formulate and test hypothesis. And hopefully, that would help us design future intervention strategies. Because these are systems that we can easily control. We found that simple with alternative models might be adequate, but only if the environment is nutrient poor. So it’s relatively low in nutrients. And it’s complex, that there are multiple injection sites that are happening that environment. And we also search for general principles. And we came up with some ideas that that could help us decide what to do to change coexistence or causation. So now I switch gears and talk what’s the other project in the lab that I wanted to share it with you. And this project is about bi remediation. And the context is that we want to remove some of these contaminations so that the defective foods that we consume. The types of contamination that you focus on are mycotoxins. So these are fungal, secondary metabolites that contaminate a lot of different resources and are really harmful for consumption. They are fairly widespread, they contaminate around 25% of both crops, and they cause a lot of damage to the body from inducing new mutations to cancer to tissue damage, for example, liver damage, to immune depression. And it costs somewhere around $3 billion to farmers in the US to get rid of these compounds, or sometimes because they have to throw out the stock that is contaminated. I want to emphasize that this is not just the sort of adjusted conceptual concern. There are strict regulations by FDA that controls the quality of food. And once in a while there are less that some stock is not safe for consumption, and it has to be removed from from circulation. Here in the US, it mostly affects farmers and food producers. Because if they’re stuck, it’s contaminated. They can’t sell it anymore, they can collect it back from the market and destroyed. in low income countries, these regulations are not enforced as much. And that means some of these toxins go through the food chain and people consume them. And as a result is around four times higher rate of liver cancer in developing countries, and liver cancer itself is the second leading cause of death. So it costs communities even more because of the health consequences. And even in Kenya not too long ago, there were cases that the amount of contamination was so high that it’s even led to direct death because of consumption of contaminated core. There are approaches to remove the contamination from the environment, there are physical and chemical approaches, but they are relatively expensive, they are not reliable. And they might also affect food quality. So food producers are not keen on using these approaches. There has been suggested that biological detoxification could be a promising alternative. So use of biological organisms or their enzymes to remove these toxins from our environment. And the main challenge is that it seems that we have been so far in nature are not good enough for the applications. So they are not fast enough in removing the toxin or they might have side effects or they are not safe to introduce into the food. And we think that the true potentials of microbes is not still known because we haven’t really sample dinner. And because of that, this is the area that we thought we could potentially contribute. So the main idea is that to take something that the nature is already offering us and build upon it to improve the detoxication performance. So take a simple example of detoxification by an extra solar enzyme. So there is an enzyme in some organism that’s potentially regulated that produces the enzyme, that enzyme potentially gets secreted, and that could also be regulated, and then outside the silly Texas toxin and makes it less bucks. So if you think about this overall picture of how desertification is taking place, you can think about how to improve it as well. So one approach is to just find better host organisms, things that have better enzymes or produce them more or secrete and more. And we can do that by screening for species and mutants to find better detoxifiers. Another approach is to engineer better detoxifies, so you could take your host organism already has some of these capabilities, and genetically engineered them, for example, overproduced enzyme, or overexpress the secretion machinery to to make the overall process more effective. But this requires us to know what exactly is happening within the cell. And the third approach is to take the enzyme. So rather than focusing on the host itself, take the enzyme and modify it to improve its performance against the target microtasks. So we have done a little bit of work and in each of these directions for finding better detoxifiers. a graduate student in the lab, Natalie has looked at screen several strains from our environment, and asked how good they were for detoxifying aflatoxin, which is one of the major categories of mycotoxins. And interestingly, she found that she could, she was able to find many different ISIS isolates that showed detoxification. And she did this on glucose and starch for isolation and also tested their performance and glucose and starch as the main carbon source in the environment. And the major thing she found was that change that she isolated on glucose, and then exposed to starch when she was capturing the performance showed the best performance. And we think the reason for that is isolation is on glucose. Let’s just check this. So it allows us to get more of the species in our environment in our screen, but when we are testing them on a starch, the presence of a source in them in their environment, triggers the cells to produce more of the enzymes for complex carbons, and some of those enzymes degrade the mycotoxin of interest. And then another important aspect is that, based on our observations to detoxification of aflatoxin is not a rare event, we can easily find many species that show really high performance for detoxifying aflatoxin using a simple string. And that’s something that wasn’t appreciated before in the literature, that they had fun examples of different species, but the prevalence of aflatoxin detoxification wasn’t known and appreciated. You’re also using droplet based screening as an alternative to find good detoxifiers. So the idea is that instead of putting individual cells in large scale wells, we can put cells inside droplets and watch how they perform. And in our case, because the toxin is fluorescent, we have an easy essay to watch the fluorescence and decide whether the cell is doing a good job or not. And you can see that the droplets that have solecism also has a drop in fluorescence between They are effective in removing the toxins from the environment. But these types of screens allow us to go at a much higher throughput, we can potentially even screen 1000s of doses per seconds. And this allows us to screen several, many mutants and find out what types of genes might be involved in the process. The main limitation is that we need fluorescence for assay, but for many, for several appetizers, that’s not the restriction. We also looked at the enzymatic capability of microbes. So we asked, can we figure out what what the mechanisms are. And in this essay, the surface associate substrates of rhodococcus species which are known to have a production detoxification capability. And unlike our expectation, we saw that we observed the enzymes completely depleting the Aqua toxin, they slow down and eventually stopped. And we use a simple model of assuming that there is decay in the enzyme itself. And that simple model showed shown by top left here seems to capture the actual dynamics pretty well. So we think what’s happening in this environment is that it’s called the enzyme constantly moving the toxin, it loses its activity over time. So we’ve compared different species of the rhodococcus, our previous warrants and our droplist. And we saw that there was a difference in their enzymatic activity at the beginning, and also a difference in their in the lifetime of their activity. And you might expect it or not, but there’s also seems to be a trade off that the one that has lower activity has higher lifetime. So we did this for a whole bunch of cases, for both glucose and starch environments and for prison warrants. And I’ve traveled this species. And what he saw was this consistent trend between them, and it seems to be an intrinsic trade off between them, in the case of the enzyme is proportional to the initial concentration of the enzyme environments or initial activity of the enzyme environment. And this trend doesn’t seem to be strain dependent or condition dependent, all of them sort of more or less while at the same time. And that tells us that there is if we include this restriction in our model, we see that there is a there is a limit on how much of the toxin is the the species can effectively remove from the environment. And the toxin concentration cannot be too high. Because even if it’s even if we use too much of the enzyme, the lives of after they’ve dropped, so it ends up not doing as good of a time. And there’s ongoing for that you’re still trying to figure out why the straight up exists, and what is the cellular machinery that you also looked at the enzyme itself. So we collaborated with a lab in Europe that they do larger scale moleculer modeling using quantum mechanics. So we simulated one of the examples of enzymes laccase from dermatosis versicolor, which has shown to have the capability to degrade our production. But when we tested this in the lab, we saw that two different versions of our production efforts in v one v two showed very different performances. And this is something that existing models can’t explain why this, this difference exists. So we got a little bit different to the quantum mechanical modeling and tried to figure out what might be the reason for that, and what we learn from this process. But that was at first that this exhibition itself takes place by lactone drink opening, which is this drink over here following an oxidation event. So that is what it’s doing is that it’s oxidizing the molecule and that makes it unstable and ends up breaking down the molecule and making that toxin. The other thing we learned was that the strength of this oxidative capability is not the rate limiting factor. So it’s not like the lack is that better at oxidizing the substrate will do a better job. Instead, it seems that the mismatch between the apple toxin and dematic pocket is the rate limiting factor. And we also observed that there is a difference between ASP one and SG two SP one D essentially uses electrons from one end of the molecule, whereas the fg two, there are several places on the molecule that can lose an electron. And we think that this gives this FC two more flexibilities so that even if it doesn’t completely fit within the intermatic Pathak pocket, it might still show, lose the electron and get oxidized. And then we also observed that there might be some nonspecific binding of SB one to lock it away from the active site that might be responsible for the difference that we have. But this is still an ongoing work, and we’re trying to figure things out. So to summarize, this last part, there are existing mycotoxin detoxifiers, that shows some potential, but they are typically not enough for the current needs in the food and seed industry. And there are opportunities to improve them, you can screen for better species or screen for mutants and find the next genetic routes of detoxification, we can potentially engineer this change to be better at detoxification, or we can take their enzymes I think what what changes in those enzymes could make them better detoxifying? And that’s the conclusion. I’ll leave it here for you to read. And I’m happy to answer any questions that you might have. grade it was indeed an interesting and wonderful presentation. That you’re that all the audience was watching, this would be benefited from your job. So we have got a few questions here. first session is some I shall graduate. And then how about studying the interactions of the multi species interactions using the loadsa? Was Ramadhan? Right? That’s a really good question. It’s an ongoing research in the lab that we are trying to see how far we can expand it. So far, it seems in don’t change environments, things still work pretty well. But it’s something that we have to confirm. And we are doing those experiments using it. Some Wireshark a vision? how effective is fecal transplant therapies under these theories? So the issue of fecal transplant is a little bit of a black box. So usually the screening that goes into it, it’s only to screen whether there are pathogens present to impact. pathogens are present, they won’t use that. But other than that there’s not a whole lot of information was what’s actually in that environment, sometimes or that metabolites in that environment or making an impact. Some thoughts are that there are some species make an impact. And probably the reality is that it’s a combination of both. But we don’t know about that system to to apply any of the theories yet. So there is a gap between what we know and what we can design. And right now they’re the things that you know about optical metal transparencies. So below that we can effectively design anything. Next, we have access to some light shafts. Out of curiosity, how was your switch from engineering to biology? This is like, is it Sorry? What was your motivation? And what were the challenges? So that I guess to completely be honest, initially, when I made the transition, the thought was that I need to learn more about biology. So I went to a lab to do that. And then I was thinking, maybe naively, that I will learn something about biology and then go back to technology, which was my comfort zone. But when I started working on this synthetic systems and simple microbiota is really interesting. And I sort of got hooked up hooked in there and then never really left that. The main challenges is sort of not having the background that standard biology students have had during their normal training. So it’s a little bit challenging to go into a new field and try to find your footing in that new field and that that’s an ongoing challenge for me, there’s so things that I wish I knew earlier that I don’t and, but it’s part of sort of learning experience and I’ve been enjoying that process. Again, some lightstep very much Feel free to help in modeling complex microbial interactions? Well, the short answer is yes, definitely. And I think genomics has already taught us a lot about what’s happening in these communities. Now, the transition is to locate metabolomics, and proteomics on lot more complex, but also a lot more informative. So if we can track, for example, how metabolites in in our microbiota are affecting different species and different environments that could potentially open up new intervention strategies and new treatments as audience may have a question, since the huge percentage of microbiome resides in gut, how far have we come in using microbiome based therapies in treating in this sinus disorders or autoimmune disorders like phones. So there have been steps important steps taken. But it’s not nearly enough to to propose it as a treatment, especially for things related to autoimmune diseases. There are issues that are that remain unsolved. There are cases like C. difficile infection that have a pretty good success rate. But presumably, that’s because the the normal gut microbiota gets completely switched during that process during the infection, it completely switches out. And the thought is that once it’s completely other horror generated, it might be easier to introduce new species into that environment. And that might be sort of the reasons that FMT works fine for many of the C. difficile patients. But if the community is maybe a little bit disrupted, but not completely destroyed or other word unity, and it’s harder to manipulate it as well. Okay, I think that’s given to you. It was sent indeed interesting and wonderful presentation. Thank you once again, Dr. babett, for such an informative presentation. And Dr. De Becque after the live streaming, we’ll meet back in the broadcasting studio. So I kindly request you to stay after the live streaming. So thank you, audience for joining us today. Also, for the audience who didn’t subscribe to our YouTube channel. Please subscribe to know about the upcoming events, please go to our website link given in the description box and register so that you will be notified. See all soon in the next door. And so we are going to end the livestream now. Thank you. Bye, everyone.
