 Hi, how do you pronounce your surname? Duxbury. OK, Duxbury, so it's fine. So the title of your talk is Chicken Gut Microbiome members limit the spread of antimicrobial resistance plasma in E. coli. Please. Thank you. Can you see? OK, great. Well, thank you for the invitation. It's a pleasure to speak to this meeting. So very interesting range of topics covered. So I'm currently a postdoc researcher at the University of Warwick and the group of Orpensøya working on microbial diversity and interactions and photosynthetic microbial communities. However, today I'm going to talk on my previous postdoc position that I completed in the laboratory of genetics at Vahnenin University in the Netherlands. And this was in the group of I in the business in collaboration with the Institute of Ecology, NIO, and veterinary epidemiologists based at Utrecht University. And this paper was published last year. So what I'm presenting on. So to give you some background. So as we've heard about this meeting, Plasmids mediated. So antimicrobial resistance genes are commonly transmitted on Plasmids. And this can occur both vertically and horizontally within bacterial populations, but then also at the level of the microbiome. So microbiomes, so for example, Cuban and animal microbiomes, dense concentrations where bacteria can transfer. And this can occur between commensal strains, but also between commensal and pathogenic strains leading to multi-resistant pathogenic bacteria. So our conjugation is a major contribution to the spread and persistence of these persistence genes. And there are two main determinants controlling the spread of these genes within a community. And these are firstly the fitness facts of the Plasmids. So if they impose a metabolic burden on the host cell, interactions between with other mobile genetic elements in the cell. And secondly, the conjugation rate is so the speed of the Plasmid transfer. And in particular, we can develop methods to quantify this as a rate rather than a frequency measure to take into account, given cell-to-cell contact opportunities, how fast will this Plasmid transfer under different conditions. So the context of my work is understanding the dynamics of Plasmids spread within the context of the microbiome. So from genomic studies, we can learn that looking longitudinally that during the evolution of bacterial lineages, they can gain multi-resistance, which is associated with the transfer of Plasmids from another coexistent lineage within the same population. And Plasmids can spread to not only just within strains, but within the whole community. And in some studies in the literature, this has been described as Plasmids permissiveness. So the diversity of different bacterial strains in taxes, which are Plasmids can spread. So that's one idea of understanding Plasmids spread within a community. But then we lack understanding of another aspect. So how biotic factors within a microbiome, so the role of microbial interactions, we know this plays an important role in microbial community dynamics. So for example, competitive or cooperative interactions between bacteria, how can these sorts of interactions between microbes and a microbiome influence the core process of Plasmid transfer and maintenance. So understand then, intra-community microbial interactions and the role of Plasmids spread. So this leads to the main research question I focused on, understanding how microbiome, so well, asking the question whether microbiome members affect Plasmids spread of a pro-core donor or second pair. And to study this, we look at the growth rates firstly. So the fitness effects, secondly, the conjugation rates. And then thirdly, what this can mean for the final within population Plasmid frequency. So the system we focus on to study Plasmid dynamics was the chicken gut microbiome. Because in the Netherlands, resistant genes have been described as particularly prevalent within the chicken gut and also in a range of other livestock animals. And what's interesting is that the extended spectrum beetle atomases, are these involved in antibiotic degradation and resistance to your full range of beetle atoms, including sephataxine. And a particular resistance gene carried on a large Plasmid. So this is called the BLAS CTXM1 resistance gene associated with an NCI1 Plasmid. It's particularly common in the chicken guts. So in chickens, particularly in boiler chickens used for meat production and also across a range of livestock animals. And also if you look, there was a study, a 10 year study looking at the location of these ESBL resistance genes and there was a strong bias towards them being carried on Plasmids rather than on the chromosome. So looking at this specific combination, we wanted to understand more about the prevalence of this specific gene carried on an NCI1 Plasmid within the microbiome context. So this is a large Plasmid, 100 kilobases. It has a narrow host range. So it exists in E. coli and salmonella species. And it has a low copy number. So you say typically one, possibly two copies per cell. And then we wanted to understand the role of the microbial interactions of the chicken gut on the spread of this Plasmid in a controlled system. So we set this up by looking at the sequel microbiome, well, of some microbes that could be cultured from the chicken gut microbiome. So we took our samples naturally from farms. We collected sequel samples. So this is from the end part of the large intestine. You're not familiar. Whether it's a high concentration and microbial diversity of bacteria, we know that's conjugation also occurs readily in this part of the gut. So we took samples from adult chickens and then brought them into the lab and preserved them by freezing them, mixing with glycerol, and then cultured them in a rich medium. So this medium is described in the literature. The endolepium medium, so it's a rich medium. It contains a high nitrogen source and has been described to somewhat mimic the chicken gut microbiome environment. So we grew up at replica cultures in this medium in the lab. And then we spun down the cultures to collect the cell pellets of whatever microbes had grown from these gut samples and then collected the medium and so the supinate until the spend medium from these cultures. So this would contain any secreted products from the microbes from the chicken gut. So we filtered this and then by replenishing all nutrients by mixing with fresh medium, we could then use this as a new type of medium to test the role of microbial secretions on our specific processes of plasma transfer. So this is a medium containing any same metabolites or other molecules secreted from bacteria. So we use this for growth and transfer assays. So this brings me on to the second part of how we set this up. So this was all done in a lab strain of the typical MG1655 background. But we moved to a natural plasma and so the combination I described, so the CTXM1 resistance exists in our NCI1 plasma. So we moved this by a conjugation from a natural chicken isolate strain into the MG1655 background. So that would be our donor strain on the far left here. And we isolated a spontaneous nanodixic acid resistant mutant of this MG1655 background. So it had a label called selection. So when we, yeah, so, well, I described that later for the conjugation assay it could be selected when plated on agar. So we could profile the growth of this strain and also the recipient strain we could use for conjugation. So also in the MG1655 background, this was a different label fashion. It had a prime clinical resistance marker in its chromosome via a gene insertion. And then the transcontrugant strain we wanted to profile the growth of this strain as well. This was formed in a prior conjugation assay between the natural ESBL E. coli strain. We transferred this into the recipient strain background so then we created a transcontrugant strain. So we have these three strains in advance that we could profile the growth of and we did this via automated optical density profile. And so measuring the OD at 600 nanometers in a plate reader. So in a microtized plate, we can set up multiple conditions and we can profile these three strains and replicates to quantify their growth rates with and without the plasma to compare between our three populations. And then secondly, we performed the conjugation assays. This was done in a similar way within the 96th world plate. We mixed a 50-50 culture of the donor recipient strain together and to incubate this at 37 degrees over a four-hour transfer period. So relatively short. And after this period, we could play to the start and end point to select for our three populations. So by the selective plating for nanodexic acid for the donor strain, the recipient strain plated on chlorophanicole containing agar and the transcontrugant could be selected with sephataxym and chlorophanicole. So firstly, looking at the results of what we found in the chicken gut samples. So going back to the first part of the study, we characterized by a 16S RNA sequencing, the V3 to V4 hypervariable region. So we did this for the culture sample. And so what's shown here is the gut microbes that could actually grow under the lab conditions we had as we were cultured under aerobic conditions. That's what we had access to and that's how we performed our conjugation assays. So we wanted to keep the same conditions, but we'd grown the microbes at 41 degrees and control the pH to try to somehow replicate the chicken gut environment. So we saw a strong enrichment for lactic acid bacteria in three genera and we had three replicates. So supernatants one to six with three replicates of age. So we saw consistency across replicates. And so two of the genera, so we saw a single genus in almost all of our samples apart from supernatants six where we saw a mix of enterococcus and pediococcus. So lactobacillus and pediococcus are both of them are lactobacillacy, which is the third most abundant family in the chicken seeker. Enterococcus found in low abundance and studies have shown that these types of bacteria to have probiotic effects and to reduce the growth of E. coli helps via toxin production. So is it interesting to see that we could culture these bacteria and now look at the role of them on the plasmid transfer and growth assays. So we then wanted to select a suitable concentration of the supernatant to use to grow our strains. So this is a pilot experiment. Firstly, we're going to look at the dose response effect of adenine, a certain concentration of supernatant into the medium and fully replenishing for all nutrients. So this is an endpoint measurement of growth. Of course, different concentrations of supernatant in a fully nutrient replenished medium. So what we saw that there was this dose response effect. So adenine more supernatant really inhibited the growth of our E. coli strains even if we fully replenish with nutrients. So it showed that there was something strongly inhibitory being produced by these gum microbes that was affecting the growth of E. coli. So then I selected this moderately inhibitory concentration so of 20% supernatant to perform for and to selects for the growth in conjugation assays because we still wanted to see some growth of our strains and not completely inhibit it. So we used 20% supernatant and then replenished with nutrients. Assuming that all nutrients had been maybe exhausted during growth in the spam medium. So then this, we firstly looked at the effective growth rates of spam media. So relative to control treatments we had two control treatment groups. So one is over replenishing all nutrients but then we had a slightly higher nutrient concentration in the medium we were mixing with supernatant which resulted in a slightly higher concentration of nutrients when it was mixed with the supernatant. So we compared to between these two control groups and the supernatant groups. And what we saw was that there was a significant reduction in growth rates of both for both the donor and the second strains. This was more noticeable for the donor with strain in the recipient strain but we saw similar magnitudes of growth rate reduction particularly in this supernatant at six that there was a mix of pediococcus and enterococcus. Overall, we didn't see a significant fitness cost of the plasma surprisingly in this energy 1655 background. So there was a very small, say 2% overall looking across all the supernatants and the control media. However, there are some slight differences we do see between control and supernatants. So then we looked at the natural E. coli strain background and we also saw a reduction in growth rates in at least while a couple of the supernatants we saw quite a lot of variability across data points. But we did see reductions at least in the pediococcus and the mixed supernatants that also reduce growth rates of natural E. coli strain. So it wasn't just reducing growth rates of this low-adapted strain, it was also found in a natural strain. And then we looked at conjugation rates. So the effect of the supernatant media on the rates of plasma transfer. And in particular we use this endpoint conjugation rate measure rather than a conjugation frequency. I won't go into a lot of detail here because Yana will talk about this tomorrow. But basically we took into account different growth rates of the three populations and the start and end point population densities of all three populations to control for cell contact opportunities across different conditions. So different growth rates and densities during across different treatments. So for example between control and supernatant treatments. And we saw quite a lot of variability across these data points and no significant effect of the supernatant media on conjugation rates. So then we wanted to input this data and in collaboration with modelists they developed this simple ecological model to use to look at our growth rate and conjugation rate measures and use these two parameters. Simple ecological model simulating an environment of a situation of a platelet invasion into an otherwise isogenic population of E. coli. So something that may be typical in the checking data with a continuous flow of nutrients. So a constant population size but a continuous turnover of cells. And we predicted that the within population password frequency. So within any parallel population would be dependent upon the growth rate difference between the donor recipient strain, the conjugation rate and the population density. So just the E. coli here. And for reference a cell density of 10 to the 8 cells per milliliter or less would be roughly equivalent to E. coli in the checking data. So we developed this simple model and looked at so defining the parameters faces for three equilibrium situations. So where the plasmids will be lost in the grey region where the plasmid will be fixed at an intermediate frequency in the dark region and where the plasmid will become completely fixed in the population in the white region. And what we saw was that due to subtle differences in the growth rate difference between the donor recipient population this growth rate difference was slightly increased compared to the plasmid loss compared to the plasmid treatment. So it's shown by the colour points compared to the black points. We saw a shift towards the plasmid loss compared to the plasmid fixation due to these small differences in growth rates. So this model would predict that over time and the microbial secretions would be suggested to lead to the plasmid loss in the population. To conclude what this has shown is that checking out microbiome levels specifically of the lactic acid bacteria can limit the spread of antimicrobial plasmids specifically in E. coli and this was due to reductions in growth rates and the small growth rate difference in that reduction on the donor strain compared to the recipient strain so via fitness differences rather than any significant reductions in conjugation rates we predict that this could lead to an increased chance of plasmid loss say within an in vivo environment in the gastrointestinal tract and the approach that I show here presents a new and a feasible approach to test to extend beyond standard in vitro conjugation assays that are often done in a rich laboratory medium often with well lab plasmids not for the natural system so this allows an approach to incorporate additional factors so for example looking at the world of this could be extended to look at the world of anaerobic conditions on plasmid transfer and additional environmental parameters so combining both experimental and theoretical approach to predict the effect of specific environmental factors on plasmid stability within population so I'd like to thank my supervisor in the laboratory of genetics and also collaborators at Utrecht University and at the Utrecht Institute of Ecology in the Netherlands and I'd like to thank the funders too and thank you for the opportunity to present today and look forward to your questions Well done, very interesting are there let me check this time properly if there are no there are no hands and no questions are there any questions the experimental setup was like awesome thank you so experimental so controlled but so real yeah it's kind of wow so does anybody have any questions for Sarah comments so do we pass on to the next speaker and thanks thanks