 depending on the size of the chamber. And this is an example picture from one of my experiment where the chamber width is roughly 60 microns. So, to study what happens when an amino acid oxidative arises in a wild type population, I basically start with two monocultures where I grow them in M9 supplemented with a carbon in this case glucose with methionine which is the amino acid that the oxidative cannot produce. And then I harvest them at mid exponential phase because that is when the wild type cells tend to leak most amount of amino acid. And I wash these cells thoroughly to get rid of the amino acids that are there in the extracellular environment. And then I mix them in a specific ratio and then I load them in the microfluidic device and then I start imaging them. And just as a control after harvesting them what we do is we also grow the monocultures of the wild type and the oxytrofe and the oxytrofe basically cannot grow. Whereas the wild type cells do grow just in M9 supplemented with carbon. So, this is an example video from one of my experiment to quickly remind you again the pink cell is the oxytrofe cell here and the green cell is the wild type cell. And this is just half a chamber not the full chamber. And so now I will quickly talk about how we go from transforming these or from these images to more biological information. So, I quickly want to acknowledge Francisca who is a data scientist and she is developed the pipeline to look at single cell growth rates. So, this is a video from one of my experiment of a full microfluidic chamber and this is basically different snippet. So, images taken at different time points put together to make it into a movie and what we can do is we can segment these images where the segmentation algorithm identifies individual cell and then we can track individual cells over different frames and what we can do is we can measure single cell growth rates. And then here let us say each point represents a particular frame and we can see how big a cell is growing over a given let us say period of time. And now I will discuss the main questions that I am interested in testing and some of the findings that we see for now. So, the first question of course was to see if the oxytrophs grow faster than the wild type cells and the reason why I expect to see this. So, the pink cells are the oxytrophs and the green cells are the wild type cells and the reason why I expect to see this is because I was reading a lot about this black queen hypothesis and stuff where gene loss or genome streamlining can be potentially beneficial for a cell and can have a growth advantage. So, I wanted to see because the oxytroph lacks a big gene. Yeah. So, then if the pink is going to grow it is going to be because the blue guys excrete methionine and you expect that that is what is going to happen. Because I would not have thought necessarily that that is what is going to happen. Okay. So, I said that in the beginning that these wild type cells passively leak amino acids and the only way that the pink guys would grow is because they are able to take up these amino acids leaked by the blue cells and that is how they grow. And however, what we end up seeing is that the oxytrophs tend to grow roughly three times slower than the wild type cells and I will talk a bit more about why I think this is actually happening in a later slide. So, just quickly tell you. So, I consider one microfluidic chamber as a unit of replication and I have looked at about 1900 cells across experiments. And the next question that I was interested in testing was to see how the growth rate of an oxytroph cell is influenced by fraction of oxytrophs in a local neighborhood. So, for example, here you have an oxytroph which is the pink cell that emerges and this oxytroph grows and divides and it forms a cluster eventually because it is able to take up the amino acid that is leaked by the wild type cell that is in the neighborhood. And now let us say I am interested in looking at this focal oxytroph cell. I am interested in understanding how the growth of this oxytroph cell is dependent on fraction of other pink cells or other oxytrophs in that local neighborhood and local neighborhood is here roughly 4 to 5 cell lengths from that cell. And here since the oxytroph is the taker and the wild type cell is the giver, one would expect that with increase in fraction of oxytroph neighbors in that local neighborhood you would see that the growth rate of oxytroph decreases and this is exactly what we observe. So, for example, here I show that this is the local micro environment that I am looking at and that is the focal oxytroph cell. So, what I do here is that I look at the growth rate of the oxytroph cell and if the local micro environment has less than 33 percent oxytrophs I put it into the first bin. And if the local micro environment has somewhere between 33 to 66 percent of oxytrophs it goes to the second bin. And if it has more than 66 percent of oxytrophs it goes to the third bin. And essentially what you see is that with increase in fraction of oxytrophs in that neighborhood you see that the growth rate of oxytroph decreases. And here on the left hand side you can see the cartoon of that, for example. And this is just I am just demonstrating this for one cell on the left, but essentially this has been done for all cells across chambers and this trend seems to prevail. And so essentially here what I think is happening is that these oxytrophs are like black holes where they are taking up the amino acid that is being leaked by the wild type. And the more oxytrophs you have the least beneficial it is for the focal oxytroph that is there let us say in the center somewhere. And the next question that I had was to look at the wild type cells and to understand how the growth rate of a wild type cell is influenced by fraction of oxytrophs in that local micro environment. So even here I fix the local neighborhood radius and I look at a wild type cell, the green cell and I ask the same question. So how many, how does the growth rate of the green cell depend on fraction of pink cells in that local neighborhood? And essentially even here since what I would expect is that with increase in fraction of pink cells which is the oxytrophs I would expect that the growth rate of the wild type cells would decrease. I am sorry, why? Why in this case? Yeah, if you just give me, yeah, I will get to that. So even here what we see essentially is that when you look at the wild type cell and you look at a local neighborhood I do the same set of, same type of analysis here. So if the neighborhood has less than 33 percent of oxytrophs the cell is binned into the first category. If it has oxytrophs between 33 to 66 percent goes to the second bin and if it has more than 66 percent of oxytrophs a lot of oxytrophs in that local neighborhood then it goes to the third bin. So even here you see that the growth rate tends to decrease with increase in fraction of oxytrophs in the local neighborhood. So these oxytrophs also tend to slow down the growth of wild type cells in the local neighborhood. So what is happening? So what I think is happening is that these wild type cells are inevitably leak amino acids as part of their metabolism and these wild type cells and oxytrophs then try to take up the leaked amino acid but the oxytrophs tend to be more efficient either because they have more transporters or because of the concentration gradient that they are facing and they grow and divide but they are not growing as fast as the wild type cells. So that means they are getting compensated but they are not compensated at the right level that they are able to overtake the wild type cells and grow. 0.3 to 0.6 and 0.6 to 1. So are you controlling for the total number of cells in this neighborhood because one could imagine that the denser your neighborhood is the slower you grow and so but that might also if you are a denser neighborhood then it is also more likely there is more than so many percent pink ones. So did you control for this? Yeah. So I also make sure that the local density is the same. So as I was showing these movies before these microfluidic chambers are always full. So cells are always packed. It is not like you know there is one particular place where there are fewer cells and there is another region where there are more cells. It is always uniformly distributed more or less. Does that answer your question? Yeah there is some variation in terms of orientation of cells but then they are all packed. There is no empty space. Sorry so since the chambers are all packed like you are sure that the nutrients that you are flowing in the glucose you are flowing in the channel is getting all the way to the end of the chamber right? Yeah so we do control experiments where for example we look at growth rates at different regions of the chamber. So on the top of the chamber and bottom of the chamber at different locations and you see that there is no new ingredient basically. Can I ask you another quick thing? Do you know how much methionine would they need to grow as fast as the wild type? Yeah this is something that I am testing right now where I provide methionine. So I just do batch culture experiments where I provide them with Casamino acid methionine and see how much I need to provide so that they grow as fast as the wild type. So to go back so they are being compensated but they are not fully compensated that they are able to grow let us say as fast as the wild type or even faster than the wild type. So my understanding is that these oxotrophs are a triple disadvantage. So they are slow first of all to begin with because they cannot produce methionine. Second of all they are even slow when they are surrounded by more oxotrophs because these other oxotrophs are taking up or scavenging on these amino acids and they also slow down their suppliers who are in the immediate neighborhood. So basically this is why I think they are growing super slow. So next what I want to do is that I want to test the generality of this observation with amino acids that have different costs of production. So methionine is a relatively expensive amino acid to produce for the cell. So I would also like to work with other amino acids that are less expensive to produce like isoleucine or tryptophan and I want to see if I see a similar result and with that I will quickly summarize what I just told you about. So I see that in my system methionine oxotrophs tend to grow roughly three times slower than the wild type cells and oxotrophs tend to grow slower when they are surrounded by other oxotrophs and it is beneficial to be surrounded by a wild type than an oxotroph and wild type cells that are close to oxotrophs tend to grow slower than wild type cells that are far from oxotrophs essentially. And that brings me to the end and thank you so much for your attention and I am happy to take questions. More questions. Thank you that was very interesting. The only thing that irks me a little bit is this fact that the wild type is releasing amino acids into the environment. Like do you have any insight on why they would do that because I mean my idea would be that if I have something growing in glucose minimal medium I take glucose I make the amino acid that I need to make biomass and that is it. Why would I ever want to actually release some of these amino acids that I need into the environment? So do we have any insight into why this is happening? I think this is something that people are still trying to understand and do not fully understand why this is happening because it is a costly metabolite and there is some evidence that it is not actively secreted like it is not going through. But also that these are fairly big and they cannot just pass through the membrane. So they are definitely going through the transporter but it is a leaky thing. So I think that is what is going on there. Yes. We have done this. I have not measured the concentration but what I do is I take the supernatant and I grow the oxytrophs on that for example and the oxytrophs tend to grow and they do not grow when you just grow them in M9 supplemented with glucose. So that means they excrete methionine and then they need to re-import it in order to grow their full growth, right? I am not sure I understand because this is your explanation why the wild type is also slow in the presence of oxytroph, right? So they excrete it and then try to re-import it immediately. Yeah. So it is a leaky thing. It is also something that I am still trying to understand is that it is a leaky function. So it inevitably leaks no matter what the cell does. So it just goes out. So then the cell immediately tries to collect it back. This is my hypothesis and then either because of concentration gradient or because of more transporters the oxytroph does a better job than the wild type that is there. A simple question. Methionine is essential for the blue ones, I believe. The blue one is metabolically independent. So it does not need methionine. It can produce its own methionine. The blue ones are the wild type cells, yes. I see everything backward now. All right, I have a line but I will go on. A lot of people, we have time. So do you think that the oxytrophs are just evolving randomly or they are actually, they are being selected. Is there a benefit to losing? Yeah, exactly. That is where I started and that was the idea to actually look at selection at single cell level to see if this is beneficial and from what I have seen, it turns out that being an oxytroph is probably not so beneficial. So I have a question. Maybe you said that and I just missed it about the methionine re-import. Can you kind of elucidate about in which system this whole or what the re-import transport system of methionine looks like in the wild type or the oxytroph or I mean preferentially both, I guess. Yeah, I am still trying to understand that so I will tell you when I understand it better. And maybe as a follow-up question do you, can you see if you understand this better or maybe an experimental approach to validate your hypothesis about the re-import of methionine? Yeah, I think one way would be to see if there is any way to quantify the number of transporters that transport methionine back in both cases and see if one has more than the other or maybe add fluorescent tags and see if that helps, yeah. I think that my question is almost irrelevant right now because there have been so many answers but I just want to put my five cents here is that if it's a passive leakage by diffusion only then by locally decreasing the concentration of methionine they will increase the flux out of the cells that will make the blue cells to produce, they will force them to produce more methionine and that will slow down their growth. Of course the only way to check it is to actually do the metabolomic study, local metabolomic study where you will measure the concentration but that's not feasible at all. Yeah, exactly. Yes. So I just want to say that let's say if I was a reviewer of your paper I would ask for some controls. Yeah. To really like add methionine to the system and show that it goes away or look for growth rates as a function of where you are in the channel because it's sort of difficult for me to really accept that you're making really important gradients in methionine locally but not in any of the other nutrients. Right? That seems, I mean, a priori a bit hard to swallow to me. I think everywhere where there are more guys between you and where the nutrients come in the levels are going to go down. It's very hard to stop this. Yeah, so just to answer that I have done a few of those controls where I grow them independently so when I just grow the oxytrope for example the oxytrope does not grow at all in the chamber and you see cell death eventually and you start imaging them because there's no methionine and also when I grow wild type cells I see that they're able to grow even without methionine and when I add methionine they're able to grow and also when I let's say supplement methionine in the medium I see that the oxytrope is also able to grow. So I've done multiple controls and I'm still doing more experiments of course. Well, no, it's not a question. It's more a comment in general for all the people that has been doubting on the secretion of different nutrients. I think it's something very common in communities given the heterogeneity in gene expression that can arise in a clonal population. I would think it's actually very common to see the exchange of nutrients maybe not all of them are secretion methionine but very different and that's why it could affect that the fact that there are a lot of oxytrophs taking up the methionine could also not affect or force these cells to produce more but it in general affects the local neighborhood by depleting that part of the neighborhood of that nutrient. I don't think it's that hard to believe. This is very nice to see this happening. Methionine seems a very key component that is expensive. Do you see density differences when you have like sort of when you mix them? It seems like you get about 10% maybe oxytrope. When you increase that number, do you see eventually a shift in global population like in the wild type? Does that also eventually how much oxytrope do I need to sort of have an effect? Yeah, so I haven't done that experiment yet and the idea was to recreate this scenario where let's say you have a wild type population and one or two mutants pop up and the idea was to see if these mutants spread and if they spread what happens in different local micro environments. So I haven't tried that yet and that would be one of the controls that I would like to do eventually. So yeah, just a curiosity. So your wild type cells passively leak amino acids. Was just wondering if that's very common across different strains in E. coli. Yeah, it's fairly common and it has been documented a lot with synthetic strains and also evolved strains that E. coli wild type tends to leak. Amino acids and it's sufficient that it can also let's say support an oxytrope for example to grow and this has already been published by other groups. Thanks. One more question. If there are normal questions, let's thank the speakers.