 Okay, thanks everyone for coming and welcome. Rachel, it's a pleasure to have you here at OIST as a theoretical sciences visitor. Let me just very briefly introduce Rachel, which is a fellow in biology at the University of Oxford, starting her lab and Queens University in Belfast early next year. She's an expert on ecology evolution of microorganisms or evolutionary microbiology, especially microorganisms associated with some other organisms, plants, animals, bacteria such as pseudomonas, rhizobia, many other organisms, and the ways how they compete or live together, such as chrysporacas, antibiotic resistance and so on. She's a rising star, I would say, in the field of evolutionary microbiology and forward to that. Also, thank you very much everyone for coming to my talk, and thank you for the lovely introduction. So I'm going to be talking to you today about how does antibiotic resistance evolve in pathogenic bacteria and what can we do about it. It works a second, okay. So my talk is going to be split into three parts today. So starting off, I'm going to give quite a general introduction to antibiotic resistance, what are antibiotics, what's their history, and how can we learn and understand resistance. I'm going to give an example from my research. So I'm going to talk about, well, so how can we understand how resistance evolves in a patient with a bacterial infection. So I'm going to wrap things up by talking about some perspective across the field and thinking about kind of what are some leading ideas on how can we combat antibiotic resistance moving forwards. So starting off, if we think about antibiotics, they're super useful for bacterial infections and bacterial infections have been a major killer throughout time in history. I'm going to highlight a couple of examples. So if we look back in the Middle Ages, we've got the bubonic plague, or the Black Death, which spread throughout Europe, and all the rest of the world and it's estimated to have killed around a third of the total world population during its spread in the 1300s. And we now know that the bubonic plague is caused by this bacteria eucinia pestis. Moving forwards, we've got tuberculosis, which has been a big problem throughout history, and antibiotics are still today really useful for treating tuberculosis which is still present. And this is showing an image here, so this is London in the 1930s, where the standard treatment for tuberculosis prior to antibiotics was fresh air. And tuberculosis again has a bacterial agent, so we now know it's caused by this bacterium called myco bacterium tuberculosis. So we've seen bacterial infections as a major cause of disease and death throughout history. And if we look at some major milestones in public health over the last century, one that really stands out to me as this first use of the antibiotic penicillin. But it's in kind of collaboration with a lot of other changes that have happened. So this is looking at the crude annual death rate from data in the United States, and it's highlighting some kind of key points over the last 100 years. So we've got the introduction of fluorine disinfectants into public water supplies. And we've got, so linking back to the first slide, the last recorded human to human transmission of plague in the US. Then over here in the 1940s we've got the first use of the antibiotic penicillin, which was really widely used after this. And then we've got some introduction of nationwide vaccine programs after that. So there's this famous story about the discovery of the antibiotic penicillin by Alexander Fleming in 1928. And the story behind this is it was actually an accidental discovery. What happened is Alexander Fleming was growing some pathogenic bacteria, Staphylococcus aureus on these agar plates. So this image just shows a plate filled with nutrients the bacteria like to grow on so it's how we can culture bacteria in the lab. And he was growing Staphylococcus on these plates and he went away for holiday left these plates out and came back and noticed something really interesting. So he noticed while he'd been away, his plates have got contaminated with mold. But what was really interesting about this if we look at this image here, so we've got the mold in this top corner here which we now know is the fungus called penicillium. We've got these little white dots are the colonies of the bacteria so the colonies of Staphylococcus growing. What he noticed was this zone of inhibition around the penicillium colony. So he saw where the mold was, there weren't any bacteria growing. And we can repeat this in a modern experiment where you can perhaps it a little bit more clearly. So this big white splodge is our fungal colony. The bacteria is streaked on the top of this plate and this is this white smear, and you can really clearly see the zone of inhibition where the bacteria are no longer growing in the proximity of this fungi. So this is really interesting he noticed okay this mold this fungi is producing some kind of substance that is killing back pathogenic bacteria, which is of course very important. And so moving forward, a lot of people started working on this, and then we see in 1942, the purification and then the effective use of this penicillin antibiotic in injection. I think this is kind of an interesting statistic in parallel. So in 1941 the worldwide stock of penicillin is estimated to have been probably around a few milligrams. So we move forward to 1945 this is absolutely skyrocketed. By 1945 just four years later it was estimated that around 4,000 kilograms of penicillin were being produced per month. And a lot of this was driven by use in World War Two. So this is a lot of production. And then this kind of really caught on and we've moved forward obviously using a lot of antibiotics. And what I think is really interesting is so the Nobel Prize in 1945 was awarded to Fleming Fleming and two others chain and flurry for their work with the antibiotic penicillin. In the Nobel Prize lecture that Fleming gave he predicted that antibiotic resistance could become a problem. So he said basically in the lab, it's not very difficult to make bugs resistant to palicillin. We've already noticed that this can occasionally happen in people. So depending on how this drug is used moving forward this could become a really big problem. And I think this is really a really interesting microbiological observation as well when we think about. So we know about evolution but actually kind of the Nobel Prize for the structure of DNA, the kind of molecule of evolution wasn't actually till later than 1945. So it is a really interesting observation. And this is of course as everyone in here probably knows this is exactly what we've seen. In this penicillin there's been a bunch of different classes of antibiotics that have been discovered. So we've got different classes and their data discovery along here. Then up in this panel is date of antibiotic resistance identified and effectively what you can see is where antibiotics go resistance will follow. We're currently in what is referred to as a bit of a discovery void with antibiotics where we haven't really identified very many new classes but we've been making modifications to the classes. We already have where there's already widespread and kind of easily evolveable resistance. And so if you look at what how is this actually happening so how do antibiotics work and how does resistance evolve. So it's broadly work by killing or inhibiting the growth of bacteria, and they can achieve this by a number of different mechanisms but generally the idea is they target something that's pretty essential for cell function. So for example, we've got the beta-lactam penicillin is an example of a beta-lactam antibiotic that inhibits cell wall synthesis. We've got antibiotics that target protein synthesis, nucleic acid synthesis, and the cell membrane. And then kind of in parallel there's a variety of different mechanisms by which resistance can evolve. I've highlighted a couple on this slide here. So for example, something we can quite often see is the up regulation of efflux pumps in bacteria. So efflux pumps are basically things that sit on the bacterial cell membrane that can pump out antibiotics and other things. And a common mechanism resistance we can observe is the up regulation of these so basically antibiotic comes into the cell but then it's quickly pumped out. So bacteria can produce enzymes that inactivate the antibiotics themselves. They can also evolve resistance via target alterations. So basically changing something about the sequence and structure of the target of the antibiotic. So the antibiotics can no longer kind of bind as effectively. The final one I've highlighted is decreased uptake. So a bunch of these antibiotics, they need to of course travel inside the cell. They do this by forens that already exist on the bacterial cell membrane and the bacteria can mutate these, stop these working so the cells no longer able to take up as much antibiotics. So there's a couple of mechanisms of action and mechanisms of resistance. But how does this actually work from an evolution perspective. So here I've just highlighted kind of two major mechanisms that we can think of. So we can think of we see the NOVO mutation. So this is a key mechanism of resistance evolution. This is basically spontaneous mutations. So just kind of random mutations that arise as an artifact of replication that just by chance provide enhanced resistance to an antibiotic. So this means is that if this mutation just by chance is present in a population of bacteria. If this mutation is treated with antibiotics, it's going to be really strong selective pressure for those individuals that carry this mutation. And so all the kind of susceptible cells will be killed off. And all the cells that have this resistant mutant mutation will be able to replicate and thrive. And this is kind of a method as well we can think of as mother cell to daughter cell transmission. On the other hand, an interesting one is horizontal gene transfer. So by this I mean basically existing resistance genes can be transferred between bacteria in a population and this can happen between neighboring cells. So for example, an antibiotic resistance gene might be carried up carried on a plasmid, and this can move by a conjugation to another bacterial cell. So these those are kind of two key mechanisms you can think of. And just as Fleming predicted back in the 1940s antibiotic resistance is indeed on the rise. And there's been a number of attempts, a lot of interest in trying to quantify the scope of this problem. I think a really nice paper came out that did this last year where they looked at global death data in 2019 and tried to categorize deaths that were either directly attributable to antibiotic resistance or associated with resistance. And from this they saw that already at least a million deaths a year are directly attributable to antibiotic resistance. And this is calculated on data from 2019 and a further four million associated with resistance. And at current rates this number is only really predicted to rise so again there's been some modeling and some estimates to efforts to estimate kind of what is this looking like in the future. And if we look back actually at this first image to the left, we can see so this is deaths associated with different infectious syndromes. And if we look further to the left here so kind of responsible for the largest budget burden of deaths. We're seeing so LIR means lower respiratory infections, so infections in the lungs. So we're seeing kind of causing a really significant burden of death is respiratory infections in the lungs. So things like acute short term infections like pneumonia as an example. So I think this makes trying to understand so well how does antibiotic resistance evolve during the spiritual infections a good question. And this is something we're really interested in during my postdoc and the research story I'm going to present today is a project from my postdoc which I did in the lab of Craig McLean in Oxford. In this room involved as part of a large scale clinical trial called compactor, which basically stands for combating antibiotic resistance in Europe. And this is kind of a large scale collaboration and involve hospitals, kind of all over Europe and lots of different countries and lots of different locations. And the overarching kind of aim of this project was to understand learn more about antibiotic resistance and we were working on a very very small part of that. And we were interested specifically in how is antibiotic resistance evolving in a specific pathogen. And the pathogen we were working on is an opportunistic bacterial pathogen called pseudomonas originosa. And this for a couple of reasons so pseudomonas is a major cause of hospital acquired infections. So it's an opportunistic pathogen which means if you're quite well and you encounter it, you'll probably be fine you won't really notice. But it's a really big problem for people who are already sick or immunocompromised which means it's a really big problem in hospitals and causing secondary infections in hospitals. So the World Health Organization critical priority list of pathogens that show very worrying levels of antibiotic resistance already and we're in need of new treatments for so high levels of resistance. And individuals that do get infections the infections themselves are associated with significant mortality. So for pseudomonas pneumonia infections, it's estimated around kind of 30% mortality out of people who acquire those. And I'm just showing some pictures of pseudomonas on a plate here. So you can see it's a very just looking at it growing it's a very diverse pathogen it can produce lots of interesting colored pit pigments that help it with virulence. And the type of infections we were specifically interested in is these acute respiratory infections so by this I mean short term infections are rather than kind of chronic long term infections an example of this being ventilator associated pneumonia. And what we're doing is so patients were coming into the ICU intensive care unit they were on ventilators. And they were enrolled in this study and basically their lungs were sampled every couple of days for a period up to 30 days because these are short term kind of intense infections. And then we were looking at the pseudomonas population in their lungs. And we were kind of looking at a collection of six to 12 pseudomonas isolates per patient per time point, which differs from the kind of standard thing done with this sort of clinical microbiology where the standard thing is you look at one isolate per patient per time point coming out of some of these lungs or another sort of infection. So really interested in the population biology of resistance evolution. And the story I'm going to tell you about is what happened when we looked across this whole patient cohort and so I'm going to break down what we did here. So we ended up with a total of 35 patients from ICUs across Europe. And we were looking what's going on in their lungs using endotracheal aspirate. So this is when patients are on mechanical ventilation, a kind of extra tube can go down there and get out this aspirate so it's really representative of what's going on in the lower respiratory tract. And we were looking at, as I said, so six to 12 pseudomonas isolates per patient per time point. And this is what I mean by this. So it's a quick exclamation of what I mean by isolates in microbiology. So basically what happens is the sample from the lungs is plated out onto an agar plate. And a single one of these circles or colonies is an isolate. So that's picked and that said, okay, that is one pseudomonas isolate. So this is an illustrative example where you can see if you've got a mixed sample of lots of different things in there, you want to get this solution down to a concentration where single cells are sticking across the surface. And so a colony the idea is it's growing up from a single cell population so you can kind of think of it as like one similar very similar genetic entity. So in this picture here, this is a mixed culture that's been plated out, where the different bacteria, the different species that have been placed out, make different colors on these plates. And so you can see when they're in these colonies, each colony is a different single color, rather than a smudge or a mix of colors. So that with a total of 441 isolates in total. And what we're doing with these are a bunch of different things but probably we're interested in so how resistant are they to antibiotics, and which antibiotics. So we're doing things called minimum inhibitory concentration assays, where we basically have a plate that has lots of different tiny tiny wells in, and you create an increasing gradient of antibiotics, you put your cells in all these wells. And then you see what concentration of antibiotics that they stop growing. So this is a way that you can characterize antibiotic resistance from a phenotypic perspective. Then we were also sequencing their genomes and looking at what's going on there. So we're doing whole genome sequencing of each isolate, and I'm going to highlight two things from this that we are particularly interested in. So we wanted to know, kind of what strains or sequence types of pseudomonas do we have in this population. So this was basically characterizing strain identity, and this is done by something called multi locus sequence typing, so that looks at the sequence variations in a set of seven housekeeping genes. And this is the way you can characterize strains within a species. And so these strains will be really quite different to each other because they've got variations in these seven quite core genes. And so then you call these like sequence type, then a number ST and a number and this is a method to use kind of across microbiology to characterize strains across bacterial species using kind of this core system. So when I say sequence type or strain, this is what I'm referring to I might use them interchangeably. So basically strain within a species, these things are really similar. And then of course we were also interested in so what's going on during the course of an infection. So analyzing mutations, so for example looking for any known antibiotic resistance mutations arising in this population. And we're also looking at gene gain or loss during infection so looking at, as I mentioned at start horizontal gene transfer of these acquired resistance genes. So what did we expect going into this. So this is kind of outlined the methods of what we were doing, but what were expectations about what we might see. So the kind of classical view and standing knowledge, particularly for these type of short term spiritual infections. So the two infections are clonal. So by this I mean the assumption is that infection will generally be started by a kind of single clonal identity entity that gets into the lungs and then replicates and proliferates over the course of the infection. And this is due to kind of extreme bottlenecks that happen during transmission it's quite difficult to get into the lungs. So our assumption is that each patient would have a single sequence type or strain of pseudomonas. And kind of as a result of that, we're expecting and what you often see is the antibody resistance emerges by a selection but did only for mutation. So the kind of. So single entity starts an infection growth growth growth replicates mutations randomly occur in this population and just by chance, some of these mutations can provide enhanced resistance to an antibiotic. And this is known by this little lightning over here. So when a patient is treated with antibiotics, so really strong selective pressure for those bacteria that just by chance have those resistance mutations. And so then you're ending up with okay, much increased antibiotic resistance by these bacteria that are now in a patient's lungs, and this could be kind of an antibiotic resistance infection. So these were assumptions going in and kind of your textbook view about how these source of infections happen. And indeed we've seen some good examples of this in our single patient case studies so far so when you dig kind of deep into the dynamics of a single patient by doing lots and lots of sequencing and lots and lots of sampling to get trying to get a really high resolution. So for example, we're able to see in a particular patient. They had one sequence type of pseudomonas. And the interesting things we saw were mutations to some kind of key antibiotic resistance genes. And we saw it again so again you've got another patient we sampled them over the course of the month. And we're seeing resistance is driven by mutations in these antibiotic resistance genes again, and this is something you'll see a lot of examples of it across the field. But so kind of that's the method that's what we expected what did we see. So if we look across this whole patient cohort. So what I'm showing you here is, I guess, quite a complicated bar charts so we've got along here we've got patients. And these bars show the number of pseudomonas isolates that we analyzed and the color of the bar tells you what sequence type or strain of pseudomonas that is. But all you really need to think about is the actually this graph kind of splits into two. So, on the left hand side you can see, all these patients have a single colored bar, I a single sequence type of strain of pseudomonas in their lungs. So this is what we're calling single strain populations, which is kind of a classical view, and indeed to the majority about two thirds of our patients were colonized by a single strain of pseudomonas. But quite surprisingly we saw actually there are a third patients which weren't so third patients we saw were colonized by multiple strains of pseudomonas at the same time. So we've caused called these are mixed strain populations and you can see in their lungs. They have two sometimes three different strains of pseudomonas at the same time. So this is pretty interesting. And so this allowed us to ask the question so what is the impact of within patients pseudomonas diversity on antibiotic resistance evolution. And to look at this we looked at a subset of patients. So we looked at a subset of 13 patients where kind of two criteria, they've been sampled on multiple days, so we could compare between time points, and they've been treated with antibiotics that we predict to be active against pseudomonas. And then we looked at the pseudomonas isolates that came out of their lungs and basically characterised them as multi drug resistant MDR or non multi drug resistant so resistant or non resistant isolates is the basic characterisation. And quite conveniently this population of 13 patients nicely split split into six patients who had these mixed strain infections and seven patients who had these single strain infections. And our hypothesis going into this. And so these mixed strain populations, purely as a kind of aspect of having multiple strains in there at the same time, are going to have an increased source of standing genetic variation compared to these single strains that just have a single strain. And we know from evolutionary biology and through theory and evolutionary biology that obviously standing genetic variation is really important for evolution. The hypothesis is that these mixed strain populations by having this increased source of standing genetic variation resistance may be able to emerge more rapidly in these population than the single strain counterparts. And so what we did to look at that is we could characterise each individual as so we've got single strain in blue, mixed strain in orange, and we could look at the initial prevalence of resistant isolates versus the change in prevalence of resistance isolates. And something you can see from this is this kind of line here, which quite, yeah kind of as you'd expect so when initial resistance is low we're seeing larger changes in resistance, larger increases. And then when you correct for this effect of initial resistance, and this is kind of our key finding here if we look comparatively at how does resistance change in single strain, compared to mixed strain populations. We indeed kind of saw agreeing with our hypothesis that mixed strain populations were indeed associated with larger increases in resistance. And we're seeing kind of 20% larger increases in resistance. And this is really our key finding here so mixed strain populations are indeed associated with the accelerated emergence of antibiotic resistance in patients, and this is significant. So our next question of course is so after we've observed this difference and this is a fact. But what are then the drive what the drivers of resistance in these mixed strain populations what is behind these larger changes in antibiotic resistance. And I'm just going to take you through an example of how we analyze this. So what I'm showing you here is a clinical timeline summary of an individual patient. So on the y axis here we've got percentage in MDR resistant isolates. And on the x axis we've got time between sampling so you can see this patient was sampled here and here. The percentage of resistant isolates in their lungs goes from around 60% to 100%. This colored bar here represents the different classes of antibiotics that they were treated with. So you can see this patient was only in for about two weeks and they were treated with a lot of antibiotics is a lot of different antibiotics that are going to be in front of them while they're in hospital. So if we look at these inset pies, this tells us what's going on with the pseudomonas. So this inner ring tells you which strain or sequence type of pseudomonas was there. And the outer ring shows you the percentage of these isolates that had this multi drug resistant or non multi drug resistance classification. So what we can see is we can see at our first time point here this patient has a mixture of this light blue ST235 in their lungs which is 100% multi drug resistant and this purple ST2211 which is 0% resistant. This is a mixture of loads of antibiotics and indeed at this last time point we can see now 100% of the population that we're sequencing is this multi drug resistant body ST235 and indeed 100% of the isolates and now multi drug resistant. And so this is showing an example of where these changes in resistance are being driven by changes in strain composition. So we can kind of partition these differences into differences that were driven indeed as this example shows changes in strain composition or changes due to within strain acquisition. So this is a classical view of de novo resistance mute mutations so when is resistance being driven by changes in who's there versus at the within strain level ie a strain acquiring new resistance mutations over the course of the infection. And when we did this. So what's the contribution to the change in multi drug resistance prevalence, we can see within strain changes, this is like our de novo resistance mutation here at zero or in one case 25% or changes in strain composition like the example I just showed you. What's really quite strongly is that resistance is emerging in these mixed strain populations due to selection for pre existing resistant strains. And so I'm going to wrap that story up there and just summarize kind of some of the take homes from that. So we saw in the study that mixed strain infections of pseudomonas was surprisingly common in ICU patients who had these acute respiratory infections. And importantly this within strain within host strain diversity is something that impacts antibiotic resistance evolution. And what we see is that resistance emerges much more rapidly in patients colonized by mixed strain populations, rather than single strain populations. And this is due to selection for pre existing resistance strains. So I think a kind of cool idea that you could think of staring off this. And so I think it raises the question. For example, as an application is so well could you measure within patient pseudomonas population diversity and use this to help you predict the likelihood of treatment failure. So high diversity populations. Okay resistance might emerge more rapidly. And this is something that's already done and very well developed with cancer. So with chemotherapy, you can predict the likelihood of treatment failure based on the kind of diversity in the cancer cells that you have. And this is an example of where you can go from basic to application science. And we still have a lot of questions about this project and stemming off this for example so I think a big one is well how are people getting mixed strain infections. So what is the location of these different strains in the lungs. We take one sample of endotracheal aspirate. How representative and what does this tell you about where the strains are given the lungs have a very large surface area. And so we don't actually know kind of the proximity of these different strains to each other. Okay well when you've got lots of different strains in there how are they interacting with each other and what's the significance of this. Okay, so I'm going to finally wrap things up. We're thinking back again a bit more big picture so trying to think about okay some perspective of the field. So we've kind of talked you through antibiotics, antibiotic resistance on the rise, an example of how we can look at this. But what are some ideas about how we can combat antibiotic resistance. So this is kind of the area that I'm kind of most linked to is this question of well, if we understand how antibiotic resistance evolves during infections and how patients are getting antibiotic resistant infections can we stop it. And I think kind of cool example of this is a paper we published last year. So big question is so when someone has an antibiotic resistant infection, where is it come from, especially in these very severe and fast up acting lung infections. And what we're able to see is an example of where a patient's lungs were getting repeatedly colonized by the translocation of Pseudomonas from their gut. They're being treated with an antibiotic for an unrelated thing they're being treated with an antibiotic for urinary tract infection, and Pseudomonas was just chilling in their gut. It evolved really, really quite severely increased antibiotic resistance while just chilling in their gut where I didn't think it was causing too many problems, but then translated up to their lungs where it causes an infection. So the idea kind of an example of what you can use from that is you can say okay well if we reduce the intestinal colonization of Pseudomonas, could this be an effective way to reduce lung infections in critically ill patients. I think another example of this kind of thinking kind of linking understanding to okay what can we do. So this is an example of identifying a drug so actually already an already approved drug that in this paper they were able to what they kind of called it an evolution slowing drug. So this drug they saw. These anti-coli encounters, the antibiotic Cip prophylaxis in a stress induced mutagenesis pathway is upregulated and they identified this drug inhibits this specific one specific stress induced mutagenesis pathway. And the idea there being that well if you get fewer new mutations. There's a lot of new mutations that provide enhanced antibiotic resistance. There's lots of work going on and looking at kind of well what effects mutation rates, what effects horizontal gene transfer in bacteria and if you can identify things that can inhibit kind of upregulated mutation rates or conjugation between bacteria is this one way that you could reduce the kind of emergence or spread of resistance in populations that's kind of an example of the thinking behind there. A very cool new therapy, phage therapy, so useful for me to know put your hand up if you've already heard of phage therapy. Okay, so maybe 5050. So bacteria phage, so just like we have viruses that can infect us, so COVID being a really good example, bacterial cells have viruses that can infect them too. These viruses are called bacteria phage, abbreviated to phage, and kind of just as what happens with us. So these phage or bacterial viruses can infect the cell get inside replicate replicate and burst. So the idea here is well could we use these like natural predators of bacteria to target bacterial infections and could this be a really good idea for example when we've got chronic antibiotic resistance infections where the antibiotics are no longer working. And there's a lot of work going on in this space and trying to develop this. And this I'm just highlighting here so this is a paper where they've looked across a bunch of well 12 cases at the moment it's really at the kind of customized level so at the moment it's working with people who have long term infections and developing what's called a customized phage cocktail so phage are understood to be quite specific. So they'll kind of get a sample of the bacteria and test or even evolve phage that can effectively effectively infect the bacterial cells burst them open. And then this was just kind of a review of these trial cases. And they said yes to actually yielded favorable clinical or microbiological outcomes in two thirds of the cases here. So that's pretty good. And then the other idea is so and this is something I'm also interested in is microbiome manipulation. So effectively using bugs to fight other bugs. So put your hand up if you've heard of microbiome manipulation. What about people microbiome transfer. So the idea here again said bugs bacteria can do lots of different nasty things to each other so actually our penicillin was isolated from a fungi but we have lots of antibiotics are actually isolated or modifications of bacterial products that are used like in their natural setting as toxins to kill competitors. So the idea here is that we can manipulate the microbiome with strains that are particularly effective at killing or inhibiting our pathogen of choice. And the really famous example of this is fecal microbiota transplantation. So this is very famously been used to treat chronic rostridium difficile infections. So these are basically chronic bacteria infections in the gut that we, yeah, they get to a point they get really resistant antibiotics are no longer working. And what's done is so basically healthy stool samples. So kind of a sample of somebody's healthy microbiome and transplant that into the intestine of a patient with one of these costridium difficile infections. So we can restore the healthy gut microbiome and indeed this has been successful in getting rid of the costridium in there. And so again just showing an example of paper where they've done that and have a pretty cool name. So again showing success in these proof of principle studies. And then kind of last but not least, I think something important to be aware of is so it's, yeah, there's lots of cool ideas, but a really good idea is well antibiotic stewardship and infection control. So the kind of aim of antibiotic stewardship is just to overall reduce the consumption of unnecessary antibiotics. So we see a really strong correlation between the uses of usage of antibiotics and the emergence of resistance so anything you can do to reduce the use of antibiotics, particularly if necessarily is good. And by this I mean so kind of well controlling the source of infection. So just like good hygiene and sanitation processes, lots of hand washing and hospitals trying to make sure there's too much not too much like into a hospital spread. Yeah, trying to use the appropriate antibiotics so using cultural results when available. And yeah kind of what only prescribing antibiotics when they're needed so not prescribing antibiotics for viral infections but using them for bacterial infections. And I'm going to kind of wrap that up with what I think is a pretty promising study and a bit of good news and links to antibiotic stewardship and infection control. So this paper came out a couple of months ago, and it's a Spanish national wide study, where they've looked at antibiotic susceptibility and resistance in pseudomonas so the bug I've just spoken about on a nationwide scale. And what we can see here in 2017. So basically the darker the color that higher the percentage of extremely drug resistant pseudomonas isolates that have come from that region. So we can see in 2017 it's not looking great there's some kind of extremely drug resistance this is when they're resistant to almost all classes of antibiotics we have available. It's I'd say pretty widespread. And then across Europe there were lots of kind of national action plans for antibiotics including I think a really good one in Spain, where they implemented a lot of these ideas of improving infection control and antibiotic stewardship. And this hasn't been like mechanistically linked, but I think this is kind of nice results, regardless, so actually when they looked at the picture in 2022 five years later, they saw a kind of generalized and widespread decrease in resistance to antibiotics, particularly these kind of extremely drug resistant isolates that are really concerning. So I'm going to wrap that up with a bit of I think a nice story there. And finally, I've done a bit of a deep dive into antibiotic resistance evolution, but I've got a bunch of other interests that are kind of moving forward and thinking about this in my new position and also while I'm here. I've been kind of understanding the interface between antibiotic resistance and the microbiome so for example how do interactions in the microbiome shape the emergence of resistance and vice versa how does a pathogen, having enhanced levels of antibiotic resistance shape that it's interactions and ability to invade a microbiome. And I'm also very interested in so CRISPR-Cas systems, kind of famous for their use in genetic engineering, but I originally identified as bacterial defense systems. I'm very interested in how they work as bacterial defense systems and using comparative genomics to understand more about that. So please get in touch if you're interested in any of these topics that I've presented and would like to chat about them more. And then finally I'll wrap that up by saying what a big thank you to everybody I work with, particularly in Oxford who contributed to this study and of course collaborators across Europe. The names highlighted there, but especially Craig McClain and Julio Diaz Caballero and Natalia Capell, acknowledgements to my funders and also a big thank you to the TSVP for inviting me on this program. Yes, thank you very much for listening and I'm happy to take questions. Thank you questions please use the microphone. Thanks for the very clear presentation. That was great. I had a bunch of questions I'll try to limit them. When you were doing the species analysis you were using seven poor housekeeping genes is that right. So, a couple questions about that one is how did you choose those seven because I think usually there's more housekeeping genes is that right. So it's a system called hub MLS T. So it's, it's, yeah pretty is very well developed as a method in microbial genomics. So they weren't chosen by me but this is quite an old and established pipeline for identifying and classifying bacterial species. Yeah, so I'm not sure exactly how they were chosen, but it's used as kind of the common genetic approach that's used. And then, if you had like, I'm just trying to figure out how you define like a different species so if you had like one of those seven that was a different amino acid sequence then it would be considered a separate species is that right. I mean that's a great question isn't it how do I how do we define species in bacteria. There's a lot of debates on it. So they went through a couple of different pipelines. So first of all, so the stuff that came out of the patient's lungs would play to doubt on what we call pseudomonas selective a girl. So it contains basically compounds that are supposed to be selective for pseudomonas originals specifically. The isolates were then randomly picked off that and they went through moldy toff. So moldy toff was used to characterize. Okay, are these pseudomonas originosa. And then we got sent the isolates and we extracted their DNA and we sequenced them. For the malady top. Are you looking at the secondary metabolite weights or like the weight of the isolates. But then we hold genome sequenced them. Yeah. And we could put them through, yeah, this M LST pipelines that said yes, pseudomonas and these are the sequence types. But we also had the whole genome so we could plot them out on a flat phylogeny and say yes okay this looks like pseudomonas originosa. Yeah. And then for your, like, when you had the patients and you had like the number of isolates per sample, like obviously a patient sample. Yeah, obviously that was non uniform. And is that because they're from different hospitals or like different clinics each with their own procedures or. Yeah, so it's from we wanted six to 12 per patient per time point. Sometimes they couldn't get 12. So we had patients where we're only sent three. So that's some of the variation, but we tried to sequence between six and 12 per patient so it's kind of stemming from like, yeah I guess the samples we got sent from the hospitals that was a variation. Questions here. Hello. That was a great talk. Very understandable even for a non biologist like me. I was wondering if you could go back to the evolution slowing slide. Yeah. I was just curious if there were any downsides to this drug. Yeah, so in this study, I think, for my understanding what they've done is they've tried to go okay well long approval processes for new drugs let's see if there's anything that's already out there that has a side effect that is beneficial. So at the moment they've just tested it on E coli and with one antibiotic resistance I guess I mean there's like this general stress response is pretty useful from the bacteria's perspective in that like this is a useful process for adaptation like generating mutations and is that thing so there's definitely downsides for the E coli. So we tested, at least not in this study. This is all done in the lab. So we don't know whether there's other effects if you put it into an animal or human system. But from like an evolutionary microbiology perspective. It's a pretty neat story in the study system. Yeah, cool. Thanks. Any other questions. I have a question about strain level diversity if so if in one third of the patient diversity is already there at the beginning. What would be the ideal therapy in that case if it's sequencing or the other way you see that the patients have multiple strains of pseudomonas. For me, like, the evolutionary solution would be like, as an evolutionary biologist would be to hit the bacteria with as many antibiotics as possible right from the beginning. Yeah, that's thing one by one. So what's now done in hospitals. Is that what is my interpretation. Yeah, so I think that's an interesting question I think there's a couple of things that I think I feel like the ideal or the ideal thing is it so why do these patients get mixed strain infections and can we target this is this kind of just an artifact of being in a high infection rate setting. Are there some strains that kind of promote or enhance the colonization with other strains. And I also think an interesting idea to agree as you want to hit them with loads of antibiotics at once, or kind of a cool idea could be so that example I showed there was quite a common pairing we saw so we saw this like multi drug resistant ST235 with the comparatively quite wimpy other strain. And actually if you compete these strains and the absence of antibiotics, you see the, the soup the more resistant one is comparatively much less fit. It's out compete competed by this wimpy ST. So I think that makes you think well, could you think about kind of how the dynamics of how you're using drugs. And could it be okay, you wait for it. I'm not sure whether this would be morally okay to do people, but you wait for kind of the wimpy strain to do better. And then you could hit with antibiotics but it's all the wimpy one there, and then the wimpy one will crash, crash, but ideally, kind of make it so the wimpy one out competes, the really resistant one. And could you use something like that, I don't know. And a related question is, did you see any cases of plasma and acquisition horizontal gene transfer in resistance acquisition this way. We were expecting to say we looked for that we looked at during the course of infections are we identifying any horizontal gene transfer. And we didn't non resistance friends were just simple out competed. Yeah, they were getting out competed. I think an interesting point here is as well the time scales were quite often quite short. I think quite often in people when we look at horizontal gene transfer. We're looking at longer time scales and maybe populations in the microbiome that coexisting for a long time, whereas the time scales here in some cases just a couple of days. And you'd see, yeah one of the strains would get completely wiped out just from a wet lab perspective, I think it's going to be harder to see the horizontal gene transfer. And if it's a plasmid, it's highly dependent on the DNA purification and sequencing protocols. So that's something that's easy to miss if you have like low coverage or you're doing like some kind of non DNA plasma purification method. So we did like a hybrid sequencing approach so we used a luminous short read and nano for long read. So we're able to complete the genome for at least one well for at least one isolate for each ST and use that to help us inform what was going on with the mobile genetic elements. So we were able to identify pseudomonas isn't a super plasmid species, maybe about 5% as an estimate have plasmids, but we were able to find plasmids and kind of look at what's going on using this long read sequencing as well. But I'm sure yeah I think I'm sure we missed some things and I think also methods for looking for horizontal gene transfer as well can be a bit tricky. Is there one more question. Hi. Are there estimations of the overall mutation rates of the of the bacteria that are resistant or not based on the whole genome sequence. Um, there's yeah there's general estimates of mutation rates for pseudomonas we didn't do specific and estimates of mutation rates for these strains. Do you know if the resistant ones are overall mutating more often. So we looked at. Yeah we looked at the accumulation of mutations both in resistance genes and then across the whole genome. We looked at the mixed strain and also the single strain and also by resistance phenotype and we didn't see any differences. Okay, thank you let's thank Rachel again for. Thank you guys for a few more months. So if you're interested, please find her and talk to her projects. Thanks.