Animal & Veterinary
Inc A/C Plasmids in Salmonella enterica by Rebecca Lindsey, Ph.D.
DR. LINDSEY: Okay. First of all I would like to thank the organizers for giving me this opportunity to discuss my research and I would like to thank the audience for listening.
Okay, so in my research I have used NARMS animal isolates for my research and these NARMS isolates are an amazing resource for scientific studies because of the sheer number and variation in these samples.
So basically there are two groups of NARMS isolates. There are the diagnostic or clinical isolates and then there is nondiagnostic isolates. I wanted to use the diagnostic clinical isolates because these came from ill animals and frequently the isolate was the source of the illness. It is important to note that these animals do not enter the food chain.
We also have a group of nondiagnostic isolates and these isolates come from healthy animals. They can come from either on farms through the NAHMS project or from slaughter isolates. That usually comes through the FSIS. That can be in the form of carcass rinses, carcass swabs, ground product, and these do enter the food chain.
So to start, I looked at 2005 diagnostic isolates because I started this project in August of 2007. I started with ampicillin and tetracycline-resistant isolates.
I kind of see the world in papers so this talk is basically divided into three papers. So for the first group of studies I did replicon typing, PFGE analysis, for the second group of studies I did DNA microarray analysis, and my last group of papers is going to be on 454 sequencing of these Inc A/C plasmids.
I did this work sequentially and it made sense to me to do it this way and hopefully it will make sense to you all as well.
So in general, in our isolates with the multidrug resistant Salmonella, we saw beta-lactam resistance and sometimes this included third generation cephalosporins, tetracycline and chloramphenicol resistance, as well as streptomycin, kanamycin, and gentamicin resistance. We even had some isolates with trimethoprim/sulfamethoxazole resistance.
So the multidrug resistance, the genetics of it is usually found either in integrons or on plasmids. So these integrons -– sorry technical difficulties. Okay, there we go. These integrons are often associated with “pentaresistant” phenotype and S. typhimurium DT-104. These are in Salmonella genomic island 1.
What I was specifically interested in is the plasmid mediated multidrug resistance. These often we see five or more antimicrobial resistances on these plasmids. They are usually associated with beta-lactamase CMY2, and they are very large plasmids, 150 kb.
So in 2005 in our NARMS samples we had 1,584 diagnostic isolates and that was a little bit too many for me to start off with. So I narrowed it down. I looked for Amp and Tet resistant isolates because Amp and Tet are carried on the Inc A/C plasmids and that is what I wanted to study with the Inc A/C plasmids specifically.
Then I narrowed it down further. I wanted isolates that had different resistances, serotypes from different regions and hosts, and I ended up with 437 isolates to analyze further. These ended up being from 55 different serotypes from 28 states and 17 different animal sources.
In these isolates, 47 were ampicillin resistant, 166 were tetracycline resistant, and 224 were both Amp and Tet resistant.
So one of the first things I did was multiplex PCR looking at 18 out of 26 plasmid replicon types. These are commonly found in Enterobacteriaceae. Dr. Carattoli, her lab came up with the basic PCRs and then it was Drs. Nguyen (sic)and Johnson took it a little bit further and made this three multiplex PCR reactions. Carattoli had about five PCR reactions, multiplex reactions.
This is a great assay because we are able to characterize the plasmids a little bit further. In my isolates, I had a high number of Inc A/C plasmids followed by I1 and then HI1, which was good because that is what I am studying is the Inc A/C plasmids.
Okay, so this is going to be hard to read but it has been published so it is probably easier to read in the paper. This is a figure from one of my papers and the dark boxes show presence and the white boxes are absence. This is a PFGE dendogram down along the side. I mean this is antibacterial resistance profiles. This is my replicon typing profile. Down the side is the sample number and serotype.
Just from this big graph what you can see is that this is Inc A/C and where Inc A/C is present we also have a lot of multidrug resistance.
There as well. On this side a well. You can see that these Inc A/C also group with serotypes where as I1, the next most prevalent one, is more random throughout. As I mentioned, that has been published.
The second set of studies on these isolates, I wanted to look closer at the genes on the Inc A/C plasmids to learn more about them. So I reduced my sample size again. I used a –- I looked for different serotypes, different animal sources, from different states and regions, different resistance phenotypes, and the ones that had the greatest distance on the PFGE dendogram. So I took 59 of my original isolates for the next step and that was to put them on a microarray.
So first we had to design and print the microarray. I worked with Jonathan Frye and he has 775 antimicrobial resistance probes that represent genes on a microarray. I added to that and added 494 multidrug resistant plasmid genes. These are mostly from Inc A/C plasmids, but there is one HI1 plasmid.
So when we made this microarray, we started with pYR1 and we took all the genes present in that Inc A/C plasmid. Then we went on to the next plasmid which is from Yersinia pestis and any gene that was not already represented in pYR1 we added it to the array. So these are 494 unique plasmid genes.
Again, I apologize for this big figure. I am in the process of writing this up and hopefully it will be published soon and you can read the paper and it will be easier to read.
But again, in general, I can show you some things on this graph. First of all, in blue means it is present, the gene or probe is present, in read shows absence. Down here, the gene names on the far left. Across the top are the isolates and down the side are the, I guess, the class, like hypothetical protein or ampicillin resistant, things like that.
So this whole first side and part of this side is all from pYR1. These genes represent -- go around the plasmid in order. So you can see that we have a lot of big chunks of genes that are present in the Inc A/C backbone.
What is interesting about this data is that we see a number of isolates, one whole lineage that has a big chunk missing down here. These genes end up being mostly hypothetical genes but also type IV conjugation genes. We also have a great degree of variability and these relate mostly to mercury resistance genes as well as some transposon-related genes. This is summarized on the next slide.
So as I mentioned, we do see high variability and also the mer, aadA, and floR, as well as tet and sul genes. I saw almost exclusively beta-lactamase CMY-2 genes, but I did have one isolate that had a different beta-lactamase present. Then we saw variability with the related transposon genes which is expected.
So as I said, I am working right now on writing that data up and at the same time I am working on my third part of this project which is high-throughput sequencing of some Inc A/C plasmids.
So the first thing I did to pick my next set of samples is we made a tree based on the PAR microarray using PAUP. I again wanted to select the most distantly related isolates to get the greatest variability. I picked 31 out of the 59. In addition, I am also sequencing 12 multidrug resistant E. coli isolates that have Inc A/C plasmids and these have been put on the microarray as well and gone through replicon typing.
So we did a lot of this high-throughput sequencing work ourselves. We did the DNA extractions in Athens and then we did the first part of the library prep kit in Athens as well. I am collaborating with people at Clay Center. Dr. Jim Bono and Dr. Tim Smith, so I went to Clay Center and learned how to make the plasmid libraries, the sequencing libraries and brought that technology back to our lab. Linda Genzilinger is doing the work now.
They are doing the 454 sequencing in Clay Center as I mentioned. They are sending –- they are doing a BLAST search there and then they are sending the information and all the sequence back to me. That is what I will be doing for the next year or so, analyzing all this sequence.
So I can give you some preliminary information from the BLASTS that were done. So in this graph we basically have the isolate Salmonella or E. coli down the side. Here is the first BLAST, the second BLAST, and the subsequent BLAST. We did find Inc A/C in most all of the plasmids. This coverage right here is just an indication of how much coverage we have of each plasmid and we wanted at least 20X coverage.
These blue and green ones are mostly done now and we had some troublesome isolates that we had to transform into a different background because some of these we had a lot of other plasmids present or phage present which was giving us problems with sequencing the Inc A/C plasmids.
So to just kind of summarize what I had on the previous table, we do see a lot of other plasmids besides Inc A/C. My first job will be to analyze the Inc A/C plasmids but there is a lot of other information in there that we can analyze and write good stories and publish as well.
We see other Salmonella enterica plasmids, as well as Shigella. We see some R Inc I2 type of plasmids, F2 type plasmids. I see a lot of large APEC plasmids so if anybody here works on APEC plasmids and is interested in collaborating let me know because I have more sequence than I have time. I am interested in all of it though.
We see some more other virulence in F plasmids, Xi1 type of plasmids. We see a number of isolates have a lot of phage associated with them. We see different types of beta-lactamase genes as well. We have found a bunch of ColE1 plasmids which was not surprising since I am working on that with somebody else and I wanted to sequence those as well.
We have pRA1 also. And then we have some things I really did not expect to see. We have some Pasteurella plasmids as well as Rhodopseudomonas plasmids, so it is going to be really interesting to analyze this data further to kind of learn what is present and maybe try to figure out how they are getting there. We know that they are mobile so we know how they are getting there, but it is just surprising to see them that far from their original host.
Okay, as I mentioned, I have a lot of work to do finishing the sequencing. Right now we have just finished the transformations and we are doing the last set of libraries. Then I am going to assemble the plasmids and first study the Inc A/C backbones. Then we can do some SNP analysis and look closer at the plasmid lineages, evolution and epidemiology. Maybe learn more about how often they are transferred, how they are transferred, and if there are any patterns there.
Future studies, there are a lot of future studies we could do, but a few examples are it would be very interesting to look at the non-diagnostic slaughter isolates to see where the Inc A/C plasmids as well as other plasmids are present and how many there are.
We could also look at different bacteria isolated from the same animal sample. We already have done preliminary studies on microarrays with these isolates.
We can always refine the microarray. There are always new genes that are being defined that we could add to the microarray and then there are some old genes and probes that have changed over time.
I would also love to add more plasmid genes so it would be easy to differentiate all these different types of plasmids.
Then finally, everything seems to be going to high throughput sequencing and so we could screen isolates using the cheaper microarray and then sequence the ones that are the most interesting.
So I first of all want to thank everyone in the BEAR Unit. I really enjoy working there. I enjoy working with the scientists. They have been a lot of help and I have been able to collaborate quite a bit. Dr. Paula Cray and Dr. Jonathan Frye and Rick Meinersmann as well.
Linda, as I mentioned, is doing a lot of the library prep work, actually all of it once I showed her how to do it. She is doing all the prep work. And a number of other people who have helped me over the last three years.
Dr. Jim Bono and Tim Smith in Clay Center have been a lot of fun to work with. Then, of course, everybody in Athens involved with NARMS.
Anybody who knows about NARMS knows that there are a lot of people. It takes a lot of people to process these samples and so probably everybody in the room from Athens has been involved in this process and I would like to thank them, as well as anybody else associated with NARMS at the FDA, CDC, and USDA.
Thank you for your attention and I will be happy to answer questions in a few minutes or at the break.
DR. ZHAO: Thank you Rebecca. We have about five minutes for questions. If anybody has questions for the section speaker, please stand up.