Integrated Human Practices

Ideation Processes:

As we began our ideation process, our team decided to focus on sensing and detection schemes using synthetic biology and immediately reached out to get advice on the scope of our project. We discussed with Ph.D. candidate Ethan Jones, who is part of Pamela Silver's lab, with a speciality in synthetic and quantitative biology. He focused on the suggestion that for such projects involving detection, it would be crucial to focus on modularity and rapidly growing organisms, especially in emergency situations where there is a high turnover rate. Additionally, we met with Ph.D. candidate Meg Dillingham McCullough, who has expertise in detecting nucleic acid analytes. She provided us with the advice that the most effective way to enter the detection space would be to make a contribution towards improving an existing sensing mechanism. From these suggestions, our team decided to focus on platforms that would improve such sensing mechanisms.

Through literature review, we decided to focus on a foundational advance in multiplexed cell-free sensing systems that could be utilized affordably in low-resource detection scenarios as well as in wet lab settings to advance scientific research. However, upon initial discussions with Avery Normandin and Piyush Nanda, we realized that the current method of producing cell-free systems involves an extremely costly and time-consuming process of purifying and extracting cell-free contents, which would be unfeasible for low resource settings. This led us to our project on devising a new manufacturing scheme for generating cell free systems, where we use the benefit of biology to first isolate bacteria with the appropriate biosensor within protocell droplets, after which we can trigger a genetically engineered an inducible lysis switch to release the biosensor into the protocell for sensing.

Throughout our ideation process, we brainstormed through different inducers for the inducible lysis system, including temperature-based, light-based, small-molecule-induced, and quorum sensing based. Through discussions with Piyush Nanda and Avery Normandin, we decided that temperature-induction and light-induction would not be as reliable with temperature gradients and uneven distribution of light within a solution of protocells, especially in low resource settings since temperature and light conditions would vary around the world. For proof of concept, we focused on experimentally implementing the lysis cassette through small molecule induction by arabinose.

Education Outreach for Human Practices:

As our cell-free droplet system is a foundational advance, one of the many applications is that our sensing mechanism can be used in educational settings to expose students to synthetic biology concepts in a multiplexed manner using the protocells. Our team presented and received feedback from high school students that participated in the BioSTAR program, and more information can be found on our education page.

Project Feedback from Stakeholders:

To receive feedback on the constructs we had made in lab, we met with experts in the field of synthetic biology and diagnostics, with the focus on the following questions:

• What do stakeholders ideally require out of systems that perform multiplexed sensing, especially in the context of low resource settings? What changes to our protocell system can we incorporate in order to achieve these requirements?
• With our protocell sensing system being a foundational, where else do you see such a system being applied?
• What are the limits of our protocell technology?

We met with Dr. Jeffrey Way, a lecturer in the Department of Systems Biology and Laboratory of Systems of Pharmacology at Harvard Medical School. Dr. Way has expertise in the field of implementing biological technologies in practice: he previously served as a Senior Staff Scientist at Harvard's Wyss Institute, co-founded 64-x, and is the president of General Biologics. During our discussion, he mentioned that the easiest implementation of the Eco.DROP technology would involve storing the E. coli cells at room temperature in an agar stab before suspending the bacteria in the protocells and activating their lysis. Among the inducible lysis options, Dr. Way commented that lysis triggered by quorum sensing would have the potential issue of the cells lysing during storage; therefore, it would be more beneficial to have an additional external control to induce lysis, such as the arabinose inducible lysis plasmid that we created. Another suggestion he provided was to genomically integrate the lysis gene, especially with the long term use of our protocell system, since plasmids could easily be lost during multiple replications of the cells. Another application that we discussed was that the protocells could potentially be aerosolized and then sprayed for environmental monitoring.

Additionally, we met with Dr. Justin Rolando and Dr. Arek Melkonian M.D. Dr. Ronaldo is a senior postdoctoral fellow in Professor David Walt's laboratory, which has a focus on Advanced Diagnostics at Brigham and Women's Hospital, Harvard Medical School, and the Wyss Institute. Dr. Melkonian is a Clinical Fellow in the Pathology Department at Brigham and Women's Hospital. During our discussion, we received the feedback regarding the implementation of the protocell protocol: because our protocell generation involves vortexing, there is the risk of the bacteria concentration not being high enough due to variability in protocell size and the dispersion of bacteria. Therefore, they suggested using a step emulsifier that could maintain an even size of the droplets. We discussed about the option to place our inducible lysis system under quorum sensing, which they highly recommend in order to amend the concentration problem (since the cells would grow to a high enough concentration within the protocells before they lyse). For point-of-care detection in low resource regions, they mentioned that we would have to take into account the energy considerations of powering a vortex machine as well as the microscope. The best implementations they suggested would be having a completely autonomous disease detection scheme that would involve someone putting the sample inside and is able to get the result from the push of a button. One step to achieve this would be to implement hardware along with the protocell diagnostics that involves using cellphone-based microscopes. They stressed the importance of the different aspects for fully implementable diagnostics: while we are trying to build the tools for a disease detection process, there also needs to be another practitioner who is able to interpret these results and provide best practices for next steps for the patient.

From our discussions, we had the takeaway that in order to make the protocells more effective for applications in point-of-care diagnostics, we would need to have the inducible lysis cassette controlled by two factors:

• External control based on an inducer such as arabinose. Based on our discussion with Dr. Way, this would ensure that when storing the bacteria at room temperature before forming the protocells, we would not have lysis induced during that storage process.
• Lysis also under quorum sensing. Based on our discussion with Dr. Rolando and Dr. Melkonian, this would allow the bacteria to grow at a high enough concentration within the protocells before lysing in order to maintain a high concentration of the biosensor within the protocells.

Our discussions led us to focus on the quorum sensing aspect of inducing the lysis of the bacteria within the protocells, as this would make our system more autonomous. Taking this feedback, we devised a mathematical model of the quorum sensing lysis option, which presented us with options on how to best design this type of lysis cassette. More information on this can be found on our modeling page.

Returning back to our initial objective of a detection scheme, future directions to implement our foundational advance as a diagnostic device would include creating the hardware infrastructure for the device, considering power efficient vortexing and an on-built microscope. Such hardware tools could be coupled with computational software to include algorithms that can characterize the multiplexed diagnostic results.