 now so we go to the next and last speaker this afternoon who is Willem Van Schijk and he will talk about mobile genetic elements in microbiomes. Are you there? Willem? He was. You're muted, Olem. I'm sorry, I was struggling a little bit with trying to share my screen but let's try this again. There you go. For some reason of course this always works when you try to do it and then when you are in a meeting it basically fails but there we go. That happens to me all the time so. Right, okay there we go. Can you see this? Great, we see your presentation. Okay, so thank you very much Fernando for that introduction and I feel properly intimidated having to follow two excellent talks by Jamie and Mike. My talk actually originally very similar title to Jamie and Mike's but even while they were talking I actually decided to change the title a little bit and it's now opportunities and challenges in detecting mobile genetic elements in microbiomes which I'm sure is a topic some of you hopefully most of you will find interesting. I'm going to talk mostly about one study which I will introduce in a minute but and I'm going to give a very short introduction on why I'm interested in antibiotic resistance in microbial ecosystems. Well we all know that we are being surrounded by bacteria, there are bacteria on our bodies, there are bacteria in the environment, there's soil in water and of course some of these bacteria will have antibiotic resistance genes and some of these antibiotic resistance genes will be present on classmates. We also know and this is of course the well-known one health concept that we frequently share as humans bacteria with animals and with the environment and so basically all of these reservoirs of different bacteria of different microbiomes are interlinked and there are certainly cases where bacteria that colonize humans basically cycle through this entire circle and up back in humans so we know and I find that fascinating that some of these generalist bacteria like E. coli can actually do that and and can spread everywhere and of course in addition to that and that's of course often being discussed over the last few days there's a transfer of antibiotic resistance placements from one bacterium to another. So what I'm going to do first is I'm going to focus on the environment and I'm not going to focus on the environment in Birmingham where I currently live and work but I'm going to take you to the country of Bangladesh in Asia. So we've done a study there a few years ago now on the spread of antibiotic resistance in rural and urban surface water in Bangladesh and there was a collaboration with Dr. Sirajul Islam of the International Centre for Diarrheal Disease Research in Bangladesh and so what we did is that we sampled different sites in Bangladesh so we sampled sites that are history sorry we sampled sites in rural Bangladesh and when I say rural Bangladesh you should remember that fish farming is incredibly important as an economic activity in Bangladesh and indeed fish is the main source of animal protein in the diet of the people of Bangladesh. So what we did is when we went to rural Bangladesh and we sampled water and sediment from the ponds in which fish were farmed and some of these ponds had a history of antibiotic use by interviewing the farmers but others did not so we sampled you can see that here in the in this middle picture with Ross McKinnon is one of my PhD students they sampled these sites but in addition to that we also sampled sites in in Dhaka which you can see in the right corner there Dhaka is one of the largest cities on the planet there's 21 million people living there and it's a city with large differences between rich and poor but most of the parts of Bangladesh as you probably all know are just about sea level and there's a lot of rivers and lakes and other water bodies even in a city the size of Dhaka so what we did is that we basically sampled these different lakes in in Dhaka as well and this is actually a site which is very close to the slum one of the slums in Bangladesh where human waste is basically ends up in this in the surface water without any treatment so we did we took two approaches we took just a sort of classical microbiology so we quantified fecal coliform sorry esbl producing coliforms in both our sediment and water samples and then we started to see a signal that these type of material were present more prevalent in the urban samples compared to the the rural samples we didn't have the opportunity to really follow up on this as we spent most of our time doing the metagenomic work so for metagenomics what we did is that we isolated the DNA from the surface water and the sediment and then we sequence shotgun sequence that's on an Illumina high-seq system and so this is the here on the left you can see the populations that we identified in those in those different sites in red the sediment in the blue circles are the water samples and you can see that they are two distinct populations because exactly what you expect when you do a study like this is that clearly sediment and water are different environments different ecology so you expect different bacteria there and and we were able to to show that and also when you look at this this stacked bar graph here you can see that the levels of cyanobacteria are relatively high in the are very high actually the water samples of course that's where photosynthesis is taking place so can that makes perfect sense and quite low in the sediment samples where we see much more proteo bacteria okay so one of the things that we did and if you start wondering when where are the plasmids in this story trust me they will come at some point one of the things that we did is that we then used source tracking of the bacteria that we found so we used a tool that's called feast which identifies eight typical species in an environmental ecosystem and for this study we basically asked it to identify whether there's human gut bacteria to quantify the the number of human gut bacteria that we found in our samples and you can see that the outcome of that study here on the right so in sediment we didn't see that many of these human gut bacteria but certainly in some of these urban sites in Dhaka and these were the sites that were very close in these slum areas we could see very high levels of human gut bacteria and that translated in very high levels of antibiotic resistance genes in these samples as well so here on the left you can see basically individual antibiotic resistance genes that we found and you can see that we find more antibiotic resistance genes in these urban samples compared to the rural samples but also when you look here on the right you can see that in a subset of these urban water samples we see very high levels of antibiotic resistance genes I think from a public health perspective this slide is probably the most important that I'm going to show you today because what we show here is on the x-axis is that that level of human gut bacteria in these environmental microbiomes and on the y-axis you can see the total abundance of antibiotic resistance genes and we saw an extremely strong and highly significant correlation between the abundance of these human gut bacteria and the levels of antibiotic resistance genes and I'll say that in a few more minutes as well but clearly this highlights that let's say if we can stop the spread of human waste into the environment we are making an important step in minimizing the spread of antibiotic resistance genes in the environment for those of you that want to use ESBL producing coliforms as a sort of surrogate for presence of antibiotic resistance genes in an environment the data looks far less clear cut so we saw a very weak correlation there between the abundance of ESBL producing coliforms and total antibiotic resistance loads so of course these are very important bacteria and it's interesting to quantify them and to further characterize them which we haven't done in this study but it is clear that to be using it as a proxy for total antibiotic resistance load in the environment is potentially challenging so then we of course wondered and that's also the theme of this meeting are these resistance genes associated with Plasmids so we used a tool that is called Plasflow which is quite widely used in the metatone fields these days which based on genomic signatures employs a neural network approach for to identify which fragments of a metagenome are chromosomal or Plasmids or Plasmid located sorry I should probably should have said that before because we first what we did is we identified which context in our metagenomic assembly had wrapped genes or replication genes and that wasn't I have to say that wasn't particularly successful so this is the total list of context that we found that were associated with with wrapped genes most of these are fragments of small Plasmids or small Plasmids in their entirety because we could circulate two of these these context and 10 out of 11 of these were originating from gram-negative bacteria that might be because that's a database bias and we just have more information about antibiotic resistance Plasmids in gram-negatives but it is more likely perhaps that particularly in those sediment samples where we saw high levels of proteal bacteria that this is real and these Plasmids are really there and there might not be that many other Plasmids in other filers that are that are present so one Plasmid that I just wanted to highlight is this so I'm now pointing out which of the two Plasmids could be circularized and one of these is that Plasmid that we call PWD1 and so it's interesting and I'm sorry my cancer is now working in front of the laptop it's interesting that this Plasmid is PWD1 is almost identical to RSF-1010 which has been discussed earlier today by Fernando de la Cruz over 81 percent of its sequence so it's a it's an incredibly well characterized Plasmid it's a broad host range Plasmid that was isolated in the early 1970s in Czechoslovakia from Pikminu and it seems to be globally distributed but what we think is that actually our Plasmid seems to be ancestral to RSF-1010 so you can see that here on the on the right basically it seems that there's in the insertion of two antibiotic resistance gene in RSF-1010 while PWD1 does not have that and we think that this type of Plasmid might be sort of the ancestral version and I know Rob Moran is in the audience as well he supposedly worked on this project but as far as I can remember we could not identify this sort of ancestral version of PWD1 sorry the ancestral version of RSF-1010 in any of the databases that we do tend at the time so but it is interesting that you can basically pick up these incredibly well studied Plasmids in the environmental samples literally on the other side of the planet from where they were initially isolated so then what we did is that we classified these context context in the method genomic assembly by Plasphode is two to identify which which context are Plasmid Plasmids and which ones are chromosome and again we saw that there's there's um higher numbers of these contact with antibiotic resistance genes in in these urban samples compared to these rural samples so clearly we do not only see more antibiotic resistance genes in urban samples where it's correlated with higher levels of human gut bacteria but we also see that they have an increased potential for horizontal gene transfer so on to the conclusions of this first part of my talk so we see this strong correlation between the abundance of human gut bacteria and antibiotic resistance genes we show that resistance genes in the environment are associated with Plasmids and highlighting their potential for horizontal gene transfer but perhaps the most important one in terms of public health is that when we want to invest in in sanitation maybe that sorry when we want to stop the spread of antibiotic resistance maybe we should consider investing in sanitation and sewage treatment as interventions particularly in low meal income countries to reduce the spread of antibiotic resistance because clearly that seems to be an important driver the lack of sanitation that is for the spread of antibiotic resistance in the environment in those countries right so i'm going to end just a few minutes to talk about the human gut microbiota which as you know is a very complex ecosystem there are lots of links with health and disease emerging but we also know that it is a reservoir of antibiotic resistance genes as i actually discussed in this presentation and opportunistic pathogens like E. coli and entrococcus just to name a few so what we're interested in is to identify what our bacterial hosts are of antibiotic resistance genes in microbiomes so one thing that we can do if we do shotgun sequencing is identify which resistance Plasmids are there but we don't know in which organisms they are replicating because essentially what we do is basically if you if you're lucky you get a nice circular assembly of a Plasmid but of course then you can't say whether it's an E. coli or another proteobacteria or in something else entirely so that's why we spent quite a bit of time on using proximity ligation techniques to link antibiotic resistance genes to their microbial hosts and to potentially to mobile genetic elements as well so this technique is is is sort of outlined here at the top so what you do is you cross-link DNA you digest the DNA you ligate the DNA then you remove the cross-links and then you sequence everything and so what you then are looking for is these these fragments that are highlights here the top one here where you sometimes have a cross-link a true cross-link where a piece of DNA for example Plasmid has been cross-linked to chromosome and we are hoping that they can identify the microbial hosts of Plasmids and antibiotic resistance genes in the gut microbiome. I won't have the time to really to discuss this all in a lot of detail but the biofrags here are not trivial and it's it's very important to remove all kinds of confounders particularly repeat elements which which can really screw your results in ways that you do not want them to be biased. So this is just one recent result that we have where we basically take a human stool sample and did this high C approach this cross-linking approach identify different antibiotic resistance genes here on the on the bottom and link them to different bacteria that are different material hosts so we one of the things that we do is we always use a spike in so we use in this case we use an SNA2 vector which has multiple antibiotic resistance genes SNA2 vectors are very rarely present in in the human gut so it's actually a really good control to use and in this case more than 90% of of these links that go from these resistance genes actually go to SNA2 vectors so there is still some noise in this system but we we are able to filter out most of it and this data does seem to suggest that most gut commensals that's sorry that some gut commensals at least like like pre-votella here or some bactroides or some some very understudied groups of gram-positive gut commensals can be reservoirs for antibiotic resistance genes. We're still very much working on then linking these with plasmids and other mobile genetic elements and I can say which yeah if I can speculate now a little bit or is that that will still remain challenging I think that most of the resistance genes were very doing a very good job in linking chromosomal resistance genes to their hosts but actually linking plasmid media plasmids located resistance genes is still quite challenging maybe it's just because a plasmid and a chromosome just doesn't uh they're not particularly closely closely chartering the cell so maybe that's William you froze or I froze no you didn't I hear yeah yeah he froze yeah he froze okay oh too much excitement for him yeah try to link what antibiotic resistance genes to plasmids in the gut microbiome is challenging yeah but video conferences are challenging as well