Human Practices

Impact

At William & Mary, we understand the importance of strengthening research through collaboration with experts and stakeholders in the field. Our project, SaPh:IRES, has many moving components from bioinformatic software development, complex engineering, and mathematical modeling, to design of intricate hardware. To combine these into a cohesive project, we needed extensive feedback from experts and stakeholders who would benefit from our advances.

This year for IHP, we conducted interviews with experts and stakeholders. Most notably, we maintained continuous conversations with several contacts, which shaped our project's bioinformatic pipeline, engineering strategies for satellite phage, and therapeutic applications. Every piece of feedback impacted the project’s direction. On the right, we diagram how IHP influenced various aspects of our work.

Sample Image

Experts in the Field

→ Dr. Fa-arun - University of Bristol Bioengineering

Summary

Impact on Project

Meeting with Dr. Fa-arun twice, at the beginning and end stages of our project, provided a unique opportunity to speak with the scientist responsible for generating the research that forms the basis of our project. These conversations were extremely valuable in developing our engineering approaches for the engineered P2/P4 system as well as host-range expansion, with numerous specific suggestions to improve the engineering aspects of our project. Additionally, and extremely valuable for the broader context of our project and integrated human practices, Dr. Fa-arun’s input on how her engineered P4 cosmid and transducing units could be used in novel ways in real-world environments informed how we employed our engineered cosmid systems in the model colon and soil microcosms. Moreover, Dr. Fa-arun very graciously took the time out of her busy schedule to meet with us twice, continuing to give feedback and inform the direction of our project.

Background and Relevance

Dr. Jidapha Fa-arun is a Postdoctoral research associate at the University of Bristol who designed and engineered the P4 cosmid system that we are using for our project. Throughout the 2024 iGEM season, we had the opportunity to meet with Dr. Fa-arun two times which had a significant impact on the development of our project due to the massive influence of her research on our project.

First Meeting (Beginning of Project)

P4 Cosmid

During our first meeting with Dr. Fa-arun, we had the opportunity to ask questions about her engineered system to further understand the P4 cosmid and the way it works. After explaining our engineering plans to her, she gave us some suggestions: she stated that the size of the cosmid needs to stay around 11.5kb due to the possibility of it having a head-full packaging mechanism, meaning the titer would decrease drastically if that size was not maintained. Our original design for the RFP construct did not align with that, given that it would only be 7kb, creating a large possibility for a titer that would not be employable in real-world environments. We found this to be true, and the titer for our engineered RFP cosmid was exponentially less than the titer of the CRISPR-based cosmids.

Host Expansion

Secondly, at this initial meeting, we spoke to Dr. Fa-arun about her experience making chimeric tail fibers, since host expansion via engineered tail fibers in Mycobacteria is one of the three engineering applications of our project. Dr. Fa-arun attested that this process consisted largely of trial and error since there is currently no way to systematically predict tail fiber binding in silico. Based on this information, we engineered several chimeric tail fibers to attempt infection with, to see if at least one would be compatible. In addition to this, we have developed a novel AI-based software to predict tail fiber binding and generate chimeric tail fibers that would subsequently expand phage host range.

Second Meeting (End of Project)

P4 Cosmid

During our second meeting with Dr. Fa-arun, we talked to her about the issues we encountered as we applied her system in our simulated environments. One of these was the lack of significant results in our model colon. She suggested that if the transducing units are degrading in the colon, a solution would be to add proteinase inhibitors as a supplement in the event that the proteinase in the colon is degrading the transducing units. We discussed the difficulties in titer with the RFP, and she said that when engineering P4 constructs, the size of the construct must be an integer divisor of the size of P2, which is 33kb because of the packaging mechanism of P4. Therefore, it is not necessary to maintain the size of 11.5 kb, but to increase the titer we could just increase the size to the nearest integer divisor of P2.

Future Directions

To close out our first meeting with Dr. Fa-arun, we discussed potential future directions of our system, and where she could see it being employed outside of the lab. Dr. Fa-arun stated that phage therapy will likely not be a widespread form of treatment in humans in the near future and she does not see her system being used in therapeutics due to difficult policy and cost setbacks. However, Dr. Fa-arun stated that an application of these transducing units in an animal gut could be as a preventative treatment for pathogenic E. coli, since most of these pathogens are foodborne.

While we did not have the time to incorporate the feedback from our second meeting into our engineering design, we are planning to pursue the changes Dr. Fa-arun suggested making in future research.

→ Los Alamos National Laboratory (LANL)

Summary

Impact on Project

Speaking with Dr. Sheila Van Cuyk and Dr. Kumar was exceedingly useful in informing our approach for physical discovery of satellites, particularly regarding the location of samples and methods of isolation. This conversation was also critical for broadening our search for relevant literature, resulting in more robust literature reviews.

Background and Relevance

Dr. Sheila Van Cuyk is a microbiologist at Los Alamos National Lab with an expertise in environmental engineering and environmental threat detection, and Dr. Anand Kumar is a microbiologist at Los Alamos National Lab with experience in bacteriophage therapy in mammalian gut microbiota and using phage satellites.

Dr. Kumar is experienced in advancing phage therapy as a promising alternative to traditional antibiotic treatments. His work involves utilizing a bacteriophage cocktail, consisting of both identified and novel phages, to effectively target antibiotic-resistant bacterial pathogens like methicillin-resistant Staphylococcus aureus (MRSA) and Pseudomonas aeruginosa. This innovative approach is evaluated through a novel microfluidic-based method, which, when combined with flow cytometry, allows for precise assessment of the bactericidal effects on host bacterial cells. In biofilm models, the phage therapy cocktails have consistently demonstrated potent bacterial killing without the emergence of phage resistance, even after multiple generations. Furthermore, the cocktail has shown considerable potential for disinfecting surfaces contaminated with bacterial pathogens. Metagenomic analysis revealed that more than 50% of the phages in the cocktail are novel, adding to the therapeutic potential of the treatment. Dr. Kumar's team is currently focused on identifying key host-phage receptor binding sites to further refine and enhance the therapy.

Screening for Satellites

During our meeting, Dr. Kumar suggested that the best place to find relevant satellite phages is in sewage and wastewater because if the host is present, the medium will most likely have compatible phages and satellites. We have taken this into account in our project as we have collected soil and water samples from areas near areas with sewage and wastewater to use in plating in our search for novel satellites.

Another important consideration that Dr. Kumar brought up: how do we separate phages from satellites? We have been using sequencing which has several drawbacks: cost, time, and is limited to satellite phages which have already been confirmed and characterized. Therefore, current bioinformatic approaches may not be sufficient. Dr. Kumar recommended separating phages versus satellites by size using a column and further looking into how to differentiate satellite phages from phages in the lab. We are also currently limited by our screening mechanics to screening for DNA phages only. In light of this feedback, we have focused on developing methods of identifying phage satellites both bioinformatically and in the lab. Notably, we have developed the software SaPhARI which identifies novel satellite phages bioinformatically.

Literature Reviews

Another recommendation by Dr. Kumar was that a lot of phage identification and experimentation had already been done in the 1900s by Russian and French scientists. Many of these papers have not been translated to English but could greatly assist in our research and ensure we do not unnecessarily repeat research which has already been done. We have looked into these literature sources, specifically the Russian papers, and one of our team members who speaks Russian herself has been able to conduct a literature review so that we can build from the research done by these scientists.

→ Jonathan Frey - William & Mary Makerspace Director

Summary

Impact on Project

Receiving ongoing feedback from Jonathan Frey throughout the iGEM season to construct the model colon and soil microcosms has been an incredibly useful connection and provided us with invaluable feedback that impacted our project. Jonathan Frey provided suggestions at every stage from designing to building to testing the model colon and soil microcosms: he recommended we create designs at multiple cost levels and create materials lists for our designs, he had the idea to build our colon model in an aquarium tank and use aquarium heaters and wavemakers to heat and mix the system, and at his suggestion we designed a tank top to house electrical components and made it detachable to allow for easy troubleshooting, among other key feedback and input. He also provided us with tools and resources from the Makerspace, which we used virtually everywhere to build our models. We have been incredibly grateful to have his expertise and the Makerspace resources available to us throughout this project.

Background and Relevance

Jonathan Frey is the Makerspace Director at William & Mary. He has experience working as an engineer and scientist on a variety of projects, from wireless sensors and measurement devices for personal fitness to precision laser microfabrication equipment. His engineering expertise was vital to the design and build processes of the model colon and soil microcosms, as he provided invaluable guidance and Makerspace resources which both informed our design and accelerated our build progress.

Model Colon

Design and Construction

Our hardware team had the privilege of working closely with Jonathan Frey to develop our physical model of the human colon. Jonathan aided us early into the summer, helping us develop a feasible plan for the design and timeline for the build of the model colon. He suggested that we make designs for the model at different cost levels, which allowed us to explore different options for the chamber we stored the model in, quality of materials like the peristaltic pumps, and additional functionalities. He also recommended we keep an organized list of materials throughout the design and build process and, once we started building, establish milestones along the way to keep us on track and organized. He continued to serve as a resource to us throughout the summer and after the construction of the model, and helped work through malfunctions and troubleshoot complications with the model’s operation, like weighing down the feeding bags we used as chambers in our model to prevent them from floating up to the top of the tank.

Material Selection

Jonathan also significantly informed our material selection. Once we had settled on using enteral delivery bags as our chambers representing the components of the gut instead of commonly used double-jacketed glass vessels, Jonathan helped us obtain the bags, along with silicone tubing compatible with the bags, cost-effective peristaltic pumps that would work for our project, and tubing connectors. During the first stages of the design process, we explored a number of options for housing the model at his suggestion. We first considered repurposing spare 3D printers from the Makerspace, using the printers’ heating elements to keep the system at temperature, before looking at building our own heated chamber using the tools and resources available to us at the Makerspace and machine shop. We finally settled on using a spare aquarium tank, in which Jonathan suggested using aquarium and terrarium heaters to efficiently keep the system heated, and aquarium wavemakers to replicate the mixing mechanism of the gut. To house the motors and electrical components, he provided us with a sheet of PVC plastic, and taught us how to use a CNC router to cut the sheet for our design, which we used as the tank’s top. When assembling the top, he helped us select and obtain PVC primer and cement to secure elements to the newly-cut PVC sheet, as well as bulkhead through-barbs to make tubing connections through the top.

Throughout the build, we utilized the functional 3D printers in the Makerspace to print pump stands, caps for effluent containers, and other components that were integrated into the build. We had originally planned on using epoxy to seal the top to the aquarium, but Jonathan advised us to keep the top detachable from the tank and use weatherstripping to insulate the system. This proved to be a crucial design decision that gave us access to the bags and elements inside the tank for troubleshooting, which had to be performed throughout its operation. To control the system, Jonathan provided us with an Arduino Mega and three AdaFruit Motor Shield v2’s, as well as electrical wire, electrical tape, and the right power supplies for our components. Finally, Jonathan also provided input on a number of additional functionalities, including an automated pH control system to keep each chamber of the colon at its own pH and a nitrogen gas purge system to keep the entire model under anaerobic conditions, both of which did not make it into the final design and operation of the model for our project, but are components that can be added to our model to increase its applicability to the human gut for future experiments.

Soil Microcosms

During the summer, Jonathan also provided resources and design insight into the construction of the soil microcosms and the rainfall mechanism. Tools and resources from the Makerspace were used to modify the microcosm’s plastic containers and other microcosm components, and build the rainfall mechanism. Once again, Jonathan provided feedback at every step of the process, and while the soil microcosm construction was not as complex as that of the model colon, his continuous assistance was invaluable.

To learn even more, click here to visit our hardware page!

→ Dr. Scharf - Virginia Tech

Impact on Project

Speaking with Dr. Birgit Scharf provided us with meaningful feedback on our project and numerous suggestions for the use of phage satellites, specifically how to increase the titer of our engineered P4 transducing units and how to more effectively image phage satellites. Most notably, this conversation offered us an alternative method to chimeric tail fiber engineering for phage host range expansion that we had not previously considered.

Background and Relevance

Dr. Birgit Scharf is a professor of biological sciences at Virginia Tech. Her lab is devoted to understanding the molecular mechanisms of bacterial motility and chemotaxis. Dr. Scharf has also conducted significant research on multiple families of lytic phage as well as phage host range expansion.

Flagelotropic Phage

In her research with phage, Dr. Scharf partially focused on lytic phages, which primarily bind to their hosts flagella – “Flagellotropic Phages”. She shared her experience in working with this understudied phage group and made a few recommendations for how to get around roadblocks in our projects. To help increase the titer of our P4 transducing agents and phagelets, Dr. Scharf recommended a number of protocols that her lab has performed which she thinks would work well for our purposes. Chief among these was the use of a cesium gradient centrifugation and Centricon ultrafiltration; we are currently looking into these procedures.

Host Range & In Situ Phage Training

Dr. Scharf’s lab has also worked on expanding the host range of phage, although she is taking a much different approach than we initially thought to. She recommended “Phage Training” (aka. Applemans protocol), an alternative method for phage host range expansion using in vitro evolution. The advantages of this approach are that it requires no manual design of chimeric receptor-binding proteins and instead works through homologous recombination between phages of the same family inside a shared host.

Finally, Dr. Scharf also pointed out an interesting additional application that we could consider for our model colon: this “Phage Training” could be done inside of our simulated natural environment to adapt the host range of phage to the bacteria present in it. This approach has the potential to inform our knowledge of in vivo evolution of satellite systems in a way typical lab techniques may not afford.

Phage Characterization and Imaging

When we brought up the challenge of attempting to image our novel “phagelets” to further characterize them, Dr. Scharf had invaluable advice. One problem we have faced is reaching the high titer needed for imaging these satellite-like elements; our titer was several orders of magnitude below what is generally accepted as sufficient for analytical visualization techniques. Dr. Scharf recommended using a cesium gradient to concentrate phage based on density, and shared her experience with the technique to isolate flagellotropic phages in her own lab. She also recommended that we use TEM due to the high success rate with imaging phage, specifically using TEM in association with cesium gradient isolation.

→ William & Mary Biology and Math Departments

Summary

Impact on Project

Presenting multiple iterations of our mathematical model and bioinformatic software at two William & Mary Biomath Seminars gave us integral feedback from professors and students in the Biology and Math departments. This allowed us to make decisions on which modeling approaches to take and to refine the presentation of our models and their resulting graphs for optimal clarity.

Background and Relevance

As part of the William & Mary Biomath program that hosts a seminar once a week, our team gave two presentations about the mathematical modeling and bioinformatics components of SaPh:IRES. We sought feedback from professors, graduate students, and undergraduate students with knowledge in a variety of areas of mathematical modeling about how we can improve our approaches to modeling satellites in SynBio.

Reworking the Model(s)

At our first presentation last spring, we outlined the mathematical basis and applications of genome-scale metabolic models (GEMs), showed off an early iteration of our satellite phage behavior PDE model, and discussed our plan for what would eventually become our satellite discovery and characterization pipeline (SaPhARI). The feedback we received at this meeting was largely centered on the GEM and whether it would fit in with our project goals and the rest of our modeling. After considering this feedback, we decided not to continue with the design of a GEM for our 2024 project, but are exploring this form of modeling for future research.

At our second presentation this fall semester, we presented the fully developed and functional SaPhARI pipeline along with results, however, the main focus was on our multiple PDE models of satellite dynamics in soil and colon systems. One of the major issues that was addressed at this seminar was related to the clarity of the output graphs. To resolve this, we added clearer axis labeling to the graphs as well as changed graphs for both the soil and colon models to be 1-dimensional so that they can be more easily compared. Additionally, we received feedback that the initial presentation of our parameters and variables was confusing. Taking this into account, we separated the parameters, state variables, and constant functions for each model and included units to convey the information more clearly.

To learn even more, click here to visit our modeling page!

→ Noblis - Dr. Negrón (Bioinformatics Scientist) and Dr. Abramson (Senior Scientist)

Impact on Project

Meeting with Noblis researchers has been an integral piece of our 2024 project; specifically the feedback we have received on our bioinformatic software has been extraordinarily valuable in addressing technical questions and refining the software aspects of our project. Speaking with Dr. Negrón and Dr. Abramson has allowed us to further refine our bioinformatic pipeline for efficiency, optimize our database framework, and explore new avenues for evaluating satellite families.

Background and Relevance

Noblis has been an invaluable partner in supporting our project, offering expert guidance across several key areas. In particular, Dr. Daniel Negrón, a Bioinformatician at Noblis, and Dr. Bradley Abramson, a Senior Scientist at Noblis, provided critical insights into the bioinformatics portion of our project across two meetings.

First Meeting with Dr. Daniel Negrón (August)

Metagenomic Assembler Selection

We sought advice on whether the metagenomic assembler we were using, MEGAHIT, was the best approach for our project, given the diversity in metagenomic assembly software available. Dr. Negrón emphasized that there is no universally "best" assembler and that the suitability of the tool depends on its ability to generate reliable results aligned with the specific requirements of our project. After explaining how MEGAHIT met both of our needs for reduced computational time and low memory usage, he emphasized that if MEGAHIT meets the needs of our project, then we should continue to utilize the software.

Database Optimization

Our second topic of discussion concerned how to filter the NCBI Non-Redundant (NR) database, as utilizing the whole database was too computationally expensive for us to annotate large genomic samples efficiently while running SaPhARI . We searched for an efficient way to filter the NCBI NR database to include the protein sequences from only prokaryotic, archaea, and viral species. While we could find pre-filtered databases available on the NCBI FTP server for the NCBI Nucleotide database, such options did not exist for the NCBI NR database. The only form available for download was the entirety of the NR database.

Dr. Negrón offered two suggestions for optimizing database efficiency, the first being to construct a database utilizing a list of only our desired taxids, which would involve downloading and creating a multifasta file. His second suggestion was using the –taxids option within the BLAST command, which only searches the query across organisms limited to the given taxids. Dr. Negrón issued a warning of caution for the former option, due to the need to update the database frequently in the event NCBI switches the taxid for any species. This may cause a broader issue the future BLAST+ searches, thus we would need software that would be able to reassign the old taxid numbers to the new, as these outdated labels could compromise the results of the search.

We took an approach based on Dr. Negrón’s broader warnings on filtering the NCBI non-redundant (NR) database, as we sought to find pre-filtered databases publicly available online to avoid the potential issues associated with filtering the database ourselves. We ultimately decided on using the RefSeq Bacterial, Viral, and Archea non-redundant files from the NCBI FTP server, which we then compiled into a protein DIAMOND database. His advice helped us evaluate the computational trade-offs and influenced our final decisions on database management.

Pipeline Efficiency

Dr. Negrón recommended exploring SnakeMake for streamlining our workflow, a comparable software to NextFlow, the workflow management software of SaPhARI. As with metagenomic assemblers, his guidance reinforced the importance of selecting a tool that best supports the project’s needs and our team’s familiarity with the software. While SnakeMake seemed very promising in his description, we felt our needs were being suitably met through the use of NextFlow for this project. However, we are inclined to look into using the software for future projects, due to the increased simplicity of SnakeMake. For example, Snakemake operates on Python, a more universal language than the Java-based language, Groovy, utilized for NextFlow.

Furthermore, Dr. Negrón introduced us to NextStrain, a tool for tracking viral variants, which holds potential for future phases of our research. NextStrain and NextClade are software that predict the clade and similarity of viruses given a query sequence and a reference database of similar viral sequences. This tool could be extremely useful in identifying satellite families and visualizing their differences through NextStrain’s graphic user interface, especially once we have a comprehensive database of each phage satellite family.

Overall, Dr. Negrón reviewed our bioinformatics pipeline and found it clear and well-structured. His input provided sage professional guidance to our strategies and played a significant role in refining our overall approach, enabling us to make informed decisions that balance computational efficiency with the project’s objectives.

Second Meeting with Dr. Bradley Abrahamson and Dr. Daniel Negrón (September)

Further guidance on NextStrain

Following our meeting with Dr. Daniel Negrón in August, a follow-up meeting to discuss new bioinformatics-related questions was arranged to include both Dr. Abramson and Dr. Negrón of Noblis. As we discussed our initial reservations about using NextStrain, mostly due to concerns that our satellite phage database would not meet the sufficient amount of reference material to accurately model each phage satellite family. However, Dr. Negrón highlighted that not many reference genomes are required to utilize the software and encouraged us to experiment with the software locally on our computers.

Hypothetical Proteins

Following the construction of SaPhARI, we have sought additional metrics to contrast satellite phage families from each other. This metric could lie in the vast number of hypothetical proteins found in phage satellites, which are proteins of uncharacterized function. If we could assign each hypothetical protein a unique identifier code based on genomic sequence, then perhaps we could compare these unique identifier codes to compare genetic or proteomic similarities within satellite families. While assigning function to hypothetical proteins is an empirically difficult task, we could compare the similarities in their nucleotide or amino acid sequences to add an additional layer of similarity.

A tool potentially useful for this in the future is Conserved-Domain Search BLAST, which looks for functional or structural units within amino acid sequences that can provide insights into a protein’s structure or function. Hypothetical proteins could be mapped into broad protein families, which could then be compared among satellites to add an additional parameter when classifying a suspected satellite phage into a specific family. We look forward to exploring and potentially implementing this software tool in the future.

Satellite Family Clustering

After such excellent advice on how to leverage hypothetical proteins to strengthen our satellite identification software, we continued the conversation on how to strengthen our current method of differentiating satellite families. Both professionals suggested utilizing the Jaccard Index, a method of determining the similarity between two sets. Our current code takes protein sets from multiple files and finds combinations that maximize the number of files sharing a minimum threshold of proteins, akin to how the Jaccard index measures similarity between sets. By focusing on the intersection of protein sets, our code identifies the optimal set that maximizes coverage across files based on the ratio of shared proteins to the total unique proteins, similar to calculating the Jaccard index.

In addition to the Jaccard Index suggestion, Dr. Negrón suggested utilizing the Python SciKit Learn machine learning algorithms to build models based off of distinct satellite families. He recommended looking into the Bag of Words algorithm, which would model which protein names are present within a given satellite family. SciKit Learn package CountVectorizer does exactly this, where we could gather annotated protein names into single strings for each file. Then, by applying CountVectorizer, we could create a count matrix representing the frequency of these protein names across files, enabling further analysis of similarities between protein families.

We were extremely grateful for another opportunity to discuss with Dr. Negrón and Dr. Abramson how to better improve our software to better classify satellite families. Both scientists gave great suggestions on novel software to potentially implement into improving our SaPhARI software for future use.

→ Interactions with Scientists from Noblis, Inc.

Impact on Project

Presenting at the Noblis headquarters, and subsequently meeting with Noblis researchers was vital in informing our project on two levels: technical and human impact. Noblis professionals gave us insights into the following areas: policy aspects surrounding future applications of our research, safety concerns and their potential solutions, presentation skills, the importance of public outreach, and considerations for our in situ model systems.

Background and Relevance

Noblis is a non-profit organization centered around science and technology, employing experts in a variety of fields to collaborate on projects for government clients such as the health, space, and defense civil sectors. In the spirit of iGEM, Noblis has a large focus on tailoring research solutions to real-world problems, from portable DNA sequencing in the field to optimizing aircraft flight paths and much more.

Beginning last spring, we have had the opportunity to meet with professionals at Noblis on multiple occasions to discuss and further develop our project. These ongoing conversations with Noblis professionals have been massively valuable for our project and their feedback will keep influencing our continuing research.

Presentation and Discussion

This spring, we had the opportunity to travel up to Reston, VA to give a presentation at the Noblis headquarters. Following a presentation from the 2023 William & Mary iGEM team, we delivered a half-hour presentation on phage satellites, emphasizing and exploring their potential in the field of synthetic biology. We discussed potential future directions for our project, including the use of bioinformatic tools, mathematical models, and novel assays and circuits. Our goal was to gather expert opinions on the potential challenges and benefits of our project; we received invaluable guidance and feedback on choosing real-world applications to pursue for our project within the scope of relevant contracting projects Noblis has expertise working with. Overall, this opportunity to present in front of professionals at Noblis gave us a chance to improve our presentation skills early on in the iGEM season.

During our presentation, we were asked questions by Noblis researchers present both in-person and remotely. These questions centered mainly around policy and safety concerns regarding deploying phage satellites in real-world environments, both immensely important considerations for any iGEM project. Not only did Noblis professionals pose these questions, but they also offered solutions and feedback specifically regarding the potential future directions we presented to them. Additionally, Noblis professionals gave us valuable feedback on the importance of public outreach and how to integrate the foundational advances at the core of our project with solving global problems.

Since this presentation was still very early on in the iGEM season, we were able to fully consider this feedback and incorporate it into our project by choosing different applications of phage satellites to explore that, in future research, would be easier to work with on a policy level. Furthermore, we have taken into account the safety concerns that were brought up, as the main system we are working with is non-replicative without manual induction.

Once we concluded our presentation, Noblis offered us a tour of their laboratories in their headquarters building, which featured cutting-edge technologies catered towards their clients. As we toured the building, Noblis employees were kind enough to answer any questions about any projects or equipment we noticed, giving great insight into their design and engineering cycle.

After the tour concluded, a panel of Noblis employees held a forum to answer our questions about the biotechnology industry, highlighted their career paths, and offered general advice for young professionals. We were very grateful to receive such attentive advice from these professionals and sought to incorporate these principles back into our project.

Further Discussion

After our presentation in the spring, we were able to meet with professionals from Noblis on a Zoom call. In addition to this meeting, we had two meetings with Noblis Life Sciences Group Senior Manager, Dr. Katherine Jennings, in person to discuss updates to our project and future directions.

Technical Feedback

During our initial meeting over Zoom, the focus was on the fieldability of our engineered systems in real environments. Noblis noted that in testing in real-world environments such as our model colon, we needed to consider factors such as how sugar sources affect the balance of microbial communities, guaranteeing adherence of cells to the matrix, and flow rate in creating a representative microbiota for testing constructs. They emphasized that these steps in testing engineering solutions before deployment require teams to consider as many factors as possible due to the huge impact a small variable can have on results, which is especially relevant to the mathematical models informed by this data.

Taking this into consideration, we decided to design our mathematical models specific to the model colon itself rather than a real human gut to take into account inherent limiting factors that keep the system from being completely biologically accurate.

Human Impact Feedback

During our two meetings with Dr. Jennings, the focus was more on the human impact aspect of our project than the technical side. At the first of these two meetings, Dr. Jennings gave us valuable feedback on a timeline to meet the deadlines of our project as well as on our presentation of the project as a whole. We applied this feedback immediately and by the second meeting with Dr. Jennings, our project progressed significantly regarding data collection. Additionally, we improved upon the presentation of each aspect of our project to better convey the simultaneous broadness and cohesiveness of our research. We hope to continue meeting with Dr. Jennings and other researchers at Noblis to keep incorporating their invaluable feedback into our continuing research efforts.

Stakeholders

→ Seed Health

Summary

Impact on Project

Meeting with Seed Health played an important role in informing our understanding of policy surrounding development of products such as probiotics and in expanding our understanding of biofilms. Their input also gave us important information regarding the utilization of a probiotic designed by Seed Health in the inoculation of our model colon in order to make our model more realistic.

Background and Relevance

Dr. Shanthi Parkar is a Research and Development Senior Scientist for Seed Health, a biotechnology company. We met with Dr. Parkar to discuss biofilm formation regarding our model colon, and the testing of the Seed Probiotic DS-01, a Seed Health product, which we used to inoculate our model colon.

Seed Probiotic DS-01

During our meeting, Dr. Parkar referred us to a research paper from the Simulator of Human Intestinal Microbial Ecosystem (SHIME) testing by Dr. Peter Bron. The product testing for the DS-01 probiotic demonstrated full dissolution of the capsule in the small intestine, implying delivery of the released probiotic bacteria to the colon. The SHIME used for this testing differed from the one we constructed for our project as it was focused on the stomach and small intestine rather than the colon. Seed Health built upon this “in situ” testing with human trials. The strains used in Seed DS-01 were a combination of well-vetted strains and new strains that Seed Health studied and proved to be beneficial.

Dr. Parkar informed us that Seed probiotic DS-01, like most probiotics, will simulate a mixture of bacteria but would not simulate complete gut microbiota or form a community. Seed DS-01 has bifidobacteria and lactobacilli, which represent about <10% and <1% of the average human adult gut respectively. Therefore, probiotics will only represent about 10% at most of the gut microbiota of a healthy person. Therefore, our colon model is a mixed culture model rather than a true gut microbiota model. Considering this information, we have inoculated our model colon with the DS-01 probiotic for its second run to test its effectiveness in our “in situ” testing model.

Regarding the future of phage probiotics, Dr. Parkar believes the future of phage probiotics is largely dependent on consumer opinion. Much like the early introduction of consuming bacteria as a probiotic, phage probiotics must be proven to be safe for consumption. She also believes they could be more easily accepted in clinical settings.

Biofilms

Dr. Parkar has previously worked with biofilms, specifically dental biofilms. We expressed an interest in a future application of satellite phages to treat medical biofilms, for example, the biofilms that form on silicone catheters. Dr. Parkar suggested measurement in a microplate with discs of the materials we are interested in and washing with our biological sample to form the biofilm. Although we have not had time to include this biofilm application in our current project, we are planning to utilize this valuable feedback and Seed’s suggestions for future research and incorporate this into our continuing research on this project.

→ Dr. Otchy - Colorectal Surgeon

Summary

Impact on Project

Speaking to Dr. Otchy provided us with a unique insight from a medical professional on both the potential and the challenges associated with phage therapy generally and the possible use of satellite phage in the human gut. In addition to educating our team on the policy and ethical concerns, this conversation led us to incorporate specific technical improvements into our project, specifically pre-growing the probiotic before inoculation into our model colon experimental design.

Background and Relevance

Dr. Daniel Otchy is a retired colorectal surgeon affiliated with INOVA Fairfax Hospital in Fairfax, VA with vast experience in this field. As a medical professional for multiple decades, Dr. Otchy has experience dealing with antibiotic resistance and has insight into public and professional opinion regarding alternatives to antibiotics including phage therapy.

Antibiotic Resistance and Phage Therapy

Dr. Otchy noted the increase in antibiotic-resistant infections in his time as a surgeon, with more and more patients suffering from multidrug-resistant bacteria. While he was not initially familiar with phage therapy, after further discussion, he believed that it could be an innovative way to solve the issue of antibiotic resistance, albeit there are many roadblocks ahead. Mainly, his concern was based on a lack of current literature and negative-leaning public opinion.

When asked to predict the medical and public opinion on phage therapy as an option for drug-resistant bacteria, Dr. Otchy initially stated that the general public would likely be skeptical of this treatment due to the generally negative connotation of the word “virus”. However, he then assessed that if this therapy was supported by ample amounts of literature, and both animal and human trials, then the medical community would most likely support it and patients would thereby follow.

Model Colon

In regards to our human colon model, Dr. Otchy stressed the importance of chemical content of the gut chambers as the main purpose of the digestive system is to chemically break down the feed that first enters the gut. He agreed that having separate chambers with different chemical compositions, pH levels, and bacterial loads was a good way to go about this. The major feedback he provided us was regarding the probiotics in that they should be grown before inoculation into the gut to verify their livelihood due to the high variance of probiotic supplements that can be found online; we have since taken this advice into account in the testing of our model colon.