Progress 01/01/23 to 12/31/23
Outputs
Target Audience:
Nothing Reported
Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?
Nothing Reported
How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?Objective 1 "To conduct interaction assays betweenE. coliO157:H7 strains, and phages known to be active againstE. coliO157:H7 to determine interactions of selecting phages from different activity groups to use in phage cocktails". This work is nearly complete. We are waiting on phage receptor data to select phages in the cocktail that are using different models of infection. Once we have that we will carry out final phage cocktail testing on bacteria attached to epithelial cells. Objective 2 "To build a machine learning classifier to allow phage cocktails to be developed based only on the sequence of anE. coliO157:H7 isolate without repeating the interaction assays" Machine learning classifiers have already been built by the Roslin group for phage prediction from E. coli/phage interaction datasets. We will carry out further O157 interactions with the additional set of 30 O157:H7 strains from the USA and use this data with our existing machine learning models for future prediction of phage that will be active on O157 isolates. This should be achievable within the final year of the project. Objective 3 "To validate the tailored phage cocktail through a cattle challenge study by administering the cocktail rectally (at the known site of colonization) to cattle inoculated with a STEC O157:H7 strain known to persistently colonize cattle" is scheduled to be completed during the third year of the project so there is nothing to report for this objective. We expect to complete this objective this year.
Impacts
What was accomplished under these goals?
Beef feedlots are a known persistent reservoir for Shiga toxin-producing Escherichia coli (STEC) O157:H7. Cattle entering the abattoir with low STEC O157:H7 levels result in lower STEC O157:H7 prevalence in finished product; therefore, we consider that pre-harvest interventions provide the most effective approach to reducing STEC O157:H7. Bacteriophages are viruses that infect bacteria and have been used in the past for treating bacterial infections. We propose to reduce or eliminate STEC O157:H7 shedding in cattle using bacteriophage cocktails. A key aspect of this work is testing bacteriophage activity under 'in-vivo' like conditions in the laboratory to increase the likelihood of the intervention working in cattle. The results from this research project would be of interest to anyone active in any aspect of the beef continuum from farm to fork. Others benefiting from this research would be scientists who study bacteriophages along with computer scientists who use machine learning. The specific USMARC E. coli O157:H7 feedlot strains were received by the Roslin Institute (Scotland) in Spring 2023 after significant courier and paperwork issues during the previous year. On receipt of the strains the team in Scotland genetically modified strain 22-2-9-1 to be shiga (stx) toxin negative so further work can be carried out in the Category 2 laboratory (BSL 2). This stx- strain was named ZAP2615 and was confirmed to be toxin negative with the RIDASCREEN® Verotoxin test kit (Clinical Diagnostics) and has recently been sent for whole genome sequencing (sequence data pending). In addition, around 30 other USMARC E. coli O157:H7 feedlot strains were sent to the Roslin Institute and these will be used to expand our database of phage-E. coli interactions towards predictive phage cocktail development. Objective 1 "To conduct interaction assays betweenE. coliO157:H7 strains, and phages known to be active againstE. coliO157:H7 to determine interactions of selecting phages from different activity groups to use in phage cocktails". We have a collection of over 300 E. coli phages that are active across a broad range of E. coli strains and include 20 phage isolated from the environment and cattle feces from the midwestern USA feedlots. Previously, an initial screen of phage activity with a set of E. coli O157:H7 strains demonstrated that phage active against E. coli O157:H7 strains cultured in rich broth conditions could easily be isolated. However, for some O157 strains, it became apparent that although the phage retained activity in liquid LB medium, many of the phage showed considerably less activity when the strains were cultured in other media such as MEM, a common tissue culture medium showing the importance of understanding the interaction dynamics of your strain of interest. These results highlighted the importance of generating interaction data in the most realistic conditions in the laboratory as a proxy for in vivo conditions and the importance of initially selecting and enriching phage under conditions in which they need to be effective. The USMARC E. coli O157:H7 strain (ZAP2615) was tested for phage interactions in different media conditions. An initial screen was performed using an agar spot test as a quick screen across around 200 phage. The 18 phage showing the highest activity from this screen were then tested in 4 different liquid media (LB, MEM, M9 and M9+bovine mucus) with phage activity assessed using an adapted Omnilog Biolog assay. The key to a successful phage intervention for cattle colonized with E. coli O157:H7, or to limit E. coli O157:H7 colonization in the first place, will be to utilize phage that are active at the terminal rectum of cattle where the bacteria colonize and multiply. This includes activity on the bacteria that are intimately attached to epithelial cells. E. coli O157:H7 employs a type-3 secretion system (T3SS) to drive host cell cytoskeleton re-arrangements to promote intimate tight binding to host cells. While screening of phage in liquid media (such as MEM which promotes expression of the T3SS in O157 strains) can narrow down phage selection, a key part of the project is to screen phage sets on E. coli O157:H7 that are tightly attached to epithelial cells. The group in Roslin has optimized their 24-well plate assay to measure phage activity on bacteria interacting with cattle epithelial cells (EBLs) and has carried out initial screening of phage activity on ZAP2615 attached to cells. ZAP2615 is relatively slow at forming attaching and effacing lesions on EBL cells (around 6 hours compared to other O157:H7 strains which can form lesions in 2-3 hours). We tested 9 single phage and some phage combinations in the EBL assay and did both imaging and bacterial counts. The imaging confirmed the formation of attaching and effacing lesions and some evidence of phage activity. This was quantified by counting surviving bacteria after 2 hours of phage treatment. It should be noted some of the surviving bacteria may be internalized in the EBL cells and further experiments will be carried out in the New Year to determine this proportion. For another strain we work with, cell invasion accounts for about 100 bacteria recovered in the assay. Several of the phage (Well, Tape, Mac1A and B) resulted in around a 2 log reduction of bacterial counts over the 2 hour time period. These phage also showed good activity in the different liquid media so we expect these phage to form the basis of a phage cocktail going forward. We have demonstrated that we have phage in the collection that are active against ZAP2615 under a variety of in vitro conditions. We have generated phage resistant mutants against this set of active phage to identify the bacterial receptors that these phage are utilizing. This will be important knowledge to design phage cocktails for further testing. We expect these sequencing results back early in 2024. Objective 3 "To validate the tailored phage cocktail through a cattle challenge study by administering the cocktail rectally (at the known site of colonization) to cattle inoculated with an STEC O157:H7 strain known to persistently colonize cattle".With another couple of months of in vitro testing of phage cocktails with phage we have identified as the most active against ZAP2615, we should be in a strong position to supply phage cocktails to Dr Bono at USDA-MARC to evaluate on ex vivo samples from cattle colonized with strain 22-2-9-1 ahead of the main challenge study in the coming year. In summary, overall great progress has been made after the strains were received by the Roslin group. Progress has been made in expanding the phage collection, optimizing the assays for bacteriophage susceptibility testing on tissue culture cells with more O157 strains, and developing machine learning models for other E. coli /phage data sets. Preliminary results with the strain being used for the phage trail (ZAP2615) are promising and we are moving forward with running the interaction assays in liquid media, bovine mucus, and on EBL cells to develop the bacteriophage cocktail for the challenge study.
Publications
|
Progress 01/01/22 to 12/31/22
Outputs
Target Audience:Nothing to report Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?
Nothing Reported
How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?The USMARC feedlot strains have finally arrived in Scotland. We should be able to accomplish the goal of bacteriophage susceptibility testing of these strains under different growth conditions within 6 months. If the susceptibility testing takes longer than 6 months, we may need to extend the grant in order to complete objective 3. The information from the susceptibility testing will allow us to generate phage cocktail formulations that should be active on the USMARC feedlot strain 'in vivo'. There isdiscussion of a preliminary cattle colonization experiment to check how the feedlot strain behaves following rectal and oral inoculation of cattle with inoculums grown in two different media. It should be possible to gain some information on the efficacy of the cocktails using ex vivo samples from colonized cattle. The agreement with The Roslin Institute was changed to allocate the majority of their consumables dollars to salary to allow completion of the screening work and cocktail selection.
Impacts
What was accomplished under these goals?
Beef feedlots are a known persistent reservoir for Shiga toxin-producing Escherichia coli (STEC) O157:H7. Cattle entering the abattoir with low STEC O157:H7 levels result in lower STEC O157:H7 prevalence in finished product; therefore, we consider that pre-harvest interventions provide the most effective approach to reducing STEC O157:H7 infection in humans. Bacteriophages are viruses that infect bacteria and have been used in the past for treating bacterial infections. We propose to reduce or eliminate STEC O157:H7 shedding in cattle using a specific bacteriophage cocktail targeting E. coli O157 in the bovine host and immediate local environment. The key to this work will be testing bacteriophage activity under 'in-vivo' like conditions in the laboratory to increase the likelihood of the intervention working in cattle. The results from this research project would be of interest to anyone who is active in any aspect of the beef continuum from farm to fork. Others benefiting from this research would be scientists who study bacteriophages along with computer scientists who use machine learning. Continued delays in shipping the specific U.S. Meat Animal Research Center (USMARC) E. coli O157:H7 feedlot strains to the Roslin Institute (Scotland) have persisted this year and unfortunately meant delays to the specific interaction assays for the feedlot strains that we had planned to have carried out by this stage of the project. A suitable courier has now been secured but internal issues within this company has meant the strains still haven't been shipped to Scotland. We are hopeful this will happen early in the new year and the team in Scotland are prepared to carry out these assays as soon as they receive the strains. However, we have continued to expand work on other parts of the objectives. Objective 1 "To conduct interaction assays betweenE. coliO157:H7 strains, and phages known to be active againstE. coliO157:H7 to determine interactions of selecting phages from different activity groups to use in phage cocktails". We now have a collection of over 250 E. coli phages (around 100 new phages isolated this year) that are active across a broad range of E. coli strains. During the first year of the project, we did an initial screening of phage activity with a set of E. coli O157:H7 strains and it showed we could easily find phage active against E. coli O157:H7 strains cultured in rich broth conditions. The next stage of the work was to test these phage interactions in other growth media as the first part of collecting interaction data for prediction models. However, for some O157 strains, it became apparent that although the phage retained activity in liquid LB medium, many of the phage showed considerably less activity when the strains were cultured in other media such as MEM, a common tissue culture medium. Further testing this year with more O157 strains has shown for some strains, activity was retained across the different media showing the importance of understanding the interaction dynamics of your strain of interest and using as diverse a strain collection as possible to build up data sets for the machine learning component of the project. These results also highlighted the importance of generating interaction data in the most realistic conditions in the laboratory as a proxy for in vivo conditions, and the importance of initially selecting and enriching phage under conditions in which they need to be effective. The key to a successful phage intervention for cattle colonized with E. coli O157:H7, or to limit E. coli O157:H7 colonization in the first place, will be to utilize some phage that are active at the terminal rectum of cattle where the bacteria colonize and multiply. This includes activity on the bacteria that are intimately attached to epithelial cells. E. coli O157:H7 employs a type-3 secretion system (T3SS) to drive host cell cytoskeleton re-arrangements to promote intimate tight binding to host cells. While screening of phage in liquid media (such as MEM which promotes expression of the T3SS in O157 strains) can narrow down phage selection, a key part of the project is to screen phage sets on E. coli O157:H7 that are tightly attached to epithelial cells. During this last year the group in Roslin has optimized their 24-well plate assay to measure phage activity on bacteria interacting with cattle epithelial cells (EBLs) and used it to screen phage activity on different O157 strains attached to cells. Initial work was carried out with a single O157 strain for which we struggled to find active phage however further screening with other O157 strains attached to EBL cells with an increased phage set has successfully identified many more phage that are active on intimately attached E. coli O157. As soon as we receive the USMARC E. coli O157:H7 feedlot strain in Roslin we will use this phage set to identify phage which are active on this strain when it is attached to EBL cells. While it is encouraging that we have phage in the collection against representative E. coli O157:H7 attached to epithelial cells, it is also important to use phage that are specific to the USMARC E. coli O157:H7 feedlot strain for this project. Phages specific to this strain were isolated from 9 of 12 feedlots in the midwestern United States. These phages have been shipped to the lab at the Roslin Institute and will be included in the phage set for activity testing. We may need to isolate more phage specific to the USMARC E. coli O157 strain using media representing the 'host-conditions' during the initial phage isolation if the phage activity screening with the current phage sets are not successful. Objective 2 "To build a machine learning classifier to allow phage cocktails to be developed based only on the sequence of anE. coliO157:H7 isolate without repeating the interaction assays" For this we will generate host-bacteriophage susceptibility data to develop the training data set for the machine learning classifier. We reported last year we had carried out a pilot study using genomic data available online for 427 E. coli O157:H7 isolates, and phage activity of a phage called TP13 based on isolate phage type. The prediction of this phage activity was 97% accurate using the machine learning models. Although we have not been able to build the training set for this O157 project without the specific strains in Scotland the Roslin group has continued to generate this data for other E. coli strain sets, in particular a set of 300 urinary tract E. coli isolates against 31 phage. These models are in the final stages of validation and once completed it will be straightforward once we have our training set, to put the O157 data into the models that we have already generated. Objective 3 "To validate the tailored phage cocktail through a cattle challenge study by administering the cocktail rectally (at the known site of colonization) to cattle inoculated with a STEC O157:H7 strain known to persistently colonize cattle"is scheduled to be completed during the third year of the project so there is nothing to report for this objective. In summary, overall progress continues to be slow due to the continued delays with the strains being shipped to Roslin. Progress has been made expanding the phage collection, optimizing the assays for bacteriophage susceptibility testing on tissue culture cells with more O157 strains and development of machine learning models for other E. coli /phage data sets. As soon as the USMARC feedlot strains can be shipped, we are ready to move forward with running the interaction assays in liquid media, bovine mucus and on EBL cells to develop the bacteriophage cocktail for the challenge study.
Publications
|
Progress 01/01/21 to 12/31/21
Outputs
Target Audience:
Nothing Reported
Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?
Nothing Reported
How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?We have met all the goals and objectives for the previous year, and we should be able to accomplish them this year also. We look to finish the bacteriophage susceptibility testing of E. coli O157:H7 strains in the different liquid media. This work is being accomplished at the Roslin Institute by a post-doc and lab technician. This will provide the data for the training data set for the machine learning classifier for predicating a strains susceptibility to bacteriophages. We have started preliminary work on developing the classifier and we are currently advertising for a post-doc that will be located at USMARC to work on this objective. We are set to sequence the bacteriophages at USMARC that were isolated from cattle feedlots in the United States once they are purified. We will continue screening phage using the 24-well plate assay to measure phage activity on bacteria interacting with cattle epithelial cells. This work is being completed at the Roslin Institute by a post-doc and lab technician. We hope to be using phage cocktails by the end of the year with the epithelial cell assay to determine those that will move forward to screening on the bovine rectal organoid assay. We are planning fora scientist to travel to Scotland to be trained on using epithelial cell and bovine rectal organoid methods for determining bacterial susceptibility to bacteriophages and help with machine learning approaches. This will allow these assays to be performed in the lab at USMARC. However, this will be determined by travel restricted due to the COVID-19 pandemic. While we won't know until more susceptibility assay are completed, we may need to isolate more phage specific to the USMARC E. coli O157 strain using media representing the 'host-like conditions' during the initial phage isolation. As described in the progress report, we are concerned that phages that were isolated using LB might not be as effective in 'host-like conditions' which reduces the number of bacteriophages that will make through this part of the screening process. The lab at USMARC will collect bovine feces from additional cattle feedlots and isolate the bacteriophage using media and/or additives that represents a "host-like condition" if additional bacteriophage are need for screening.
Impacts
What was accomplished under these goals?
Beef feedlots are a known persistent reservoir for Shiga toxin-producing Escherichia coli (STEC) O157:H7. Cattle entering the abattoir with low STEC O157:H7 levels result in lower STEC O157:H7 prevalence in finished product; therefore, we believe that pre-harvest interventions provide the most effective approach to reducing STEC O157:H7. Bacteriophages are viruses that infect bacteria and have been used in the past for treating bacterial infections. We propose to reduce or eliminate STEC O157:H7 shedding in cattle using an animal specific bacteriophage cocktail. The results from this research project would be of interest to anyone who is active in any aspects of the beef continuum from farm to fork.Others benefiting from this research would be scientists who study bacteriophages along with computer scientist who use machine learning. Delays in shipping the specific US Meat Animal Research Center (USMARC)E. coliO157:H7 feedlot strains to the Roslin Institute (Scotland) have unfortunately meant delays to the specific interaction assays for this strain that we had planned to have carried out by this stage of the project. The transportation issue with the feedlot strains is due to finding a suitable courier, obtaining the proper permits and delays with legal teams signing off the Material Transfer Agreement for the strains However, we have continued to work on other parts of the objectives.For objective 1 "To conduct interaction assays betweenE. coliO157:H7 strains, and phages known to be active againstE. coliO157:H7 to determine interactions of selecting phages from different activity groups to use in phage cocktails",we have a collection of over 150E. coliphages that are active across a broad range ofE. colistrains that we have startedworking through initial screening interactions with a set ofE. coliO157:H7 strains.The aim of screening a selection of phage from our library andE. coliO157:H7 strain is to compare their activity under different bacterial growth conditions. From an initial screening on solid agar, it is evident that around 40-50 percent of this phage set had lytic activity on at least four of ourE. coliO157:H7 strains, in other words it is easy to find phage active againstE. coliO157:H7 strains cultured in rich broth conditions. The next stage of the work was to test these phage interactions in other growth media as the first part of collecting interaction data for prediction models. However, it became apparent that although the phage retained activity in liquid LB medium, many of the phage showed considerably less activity when the strains were cultured in other media such as DMEM, a common tissue culture medium. It is not surprising that the phage in our collection show the highest activity in LB, as this was the medium used for the initial enrichment of phage from samples (you get what you select for). These results highlight the importance of generating interaction data in the most realistic conditions in the laboratory as a proxy forin vivoconditions, and the importance of initially selecting and enriching phage under conditions in which they need to be effective. While it is encouraging that we have phage in the collection against representativeE. coliO157:H7, it is also important to use phage that are specific to the USMARCE. coliO157:H7 feedlot strain for this project. Phages specific to this strain were isolated from 9 of 12 feedlots in the midwestern United States.These phages are currently being purified and characterized before being shipped to the lab at the Roslin Institute for further characterization using bovine mucus and in the presence of eukaryotic tissue culture cells to help mimic a more realistic host like environment. It is essential to have phage that are active in 'host-conditions', hence the inclusion of bovine mucus and tissue culture cells at the enrichment stage.We are concerned that phages that were isolated using LB might not be as effective in 'host-conditions' so we may need to isolate more phage specific to the USMARC E. coli O157 strain using media representing the 'host-conditions' during the initial phage isolation media. The key to a successful phage intervention for cattle colonized withE. coliO157:H7, or to limitE. coliO157:H7 colonization in the first place, will be to utilize phage that are active at the terminal rectum of cattle where the bacteria colonize and multiply.This includes activity on the bacteria that are intimately attached to epithelial cells.E. coliO157:H7 employs a type-3 secretion system to drive host cell cytoskeleton re-arrangements to promote intimate tight binding to host cells. While screening of phage in liquid media (such as DMEM) can narrow down phage selection, a key part of the project is to screen phage sets onE. coliO157:H7 that are tightly attached to epithelial cells. We have developed a 24-well plate assay to measure phage activity on bacteria interacting with cattle epithelial cells (EBLs). Fifty phage from our collection known to be active againstE. coliO157:H7 strain ZAP198 in LB were tested with the bacteria attached to EBL cells and only a few phage showed any evidence of activity. This indicates the challenge of finding phage active under these more realistic growth conditions. We appreciate the results could also relate to an inhibitory impact of the tissue culture medium on phage activity and this is also being investigated.We have also trialed phage activity in diluted bovine mucus isolated from the terminal rectum of cattle, but we are now waiting for the USMARCE. coliO157:H7 strains before we continue the phage selection work under these more specific conditions. For objective 2 "To build a machine learning classifier to allow phage cocktails to be developed based only on the sequence of anE. coliO157:H7 isolate without repeating the interaction assays",we are generating host-bacteriophage susceptibility data to develop the training data set for the machine learning classifier.As a pilot study, we carried out using genomic data available onlinefor 427E. coliO157:H7 isolates, and phage activity of a phage called TP13 based on isolate phage type. Prediction of this phage activity was 97% accurate using the machine learning models. Once we have our training set, it will be straightforward to put the data into the models that we have already generated. Objective 3 "To validate the tailored phage cocktail through a cattle challenge study by administering the cocktail rectally (at the known site of colonization) to cattle inoculated with a STEC O157:H7 strain known to persistently colonize cattle" is scheduled to be completed during the third year of the project so there is nothing to report for this objective. In summary, a frustratingly slow start but a lot of progress has been made developing the assays for bacteriophage susceptibility testing in different media and with gathering phage interaction data to be used as the training data set for predicting a bacterium's susceptibility to infection with a bacteriophage. As soon as the USMARC feedlot strains can be shipped, we are ready to move forward with running the interaction assays in liquid media, bovine mucus and on EBL cells to develop the bacteriophage cocktail for the challenge study.
Publications
|
|