We first met the 2022 Groningen iGEM team, NanoBuddy, at the European iGEM meetup organized by iGEM Hamburg in July. During this weekend full of scientific and non-scientific events we got to know that NanoBuddy was also working on engineering a probiotic. In both our projects a GMO is ingested by a living organism, humans or chickens. Therefore, both teams thought of an extensive plan for the biocontainment of the engineered organism to avoid its environmental release. Specifically, during the poster presentation in Hamburg we realised that we were testing the same temperature-sensitive promoter, pTlpA (BBa_K2500003). This promoter is constitutively active at 37°C , the body temperature of humans and chickens, and higher temperatures, while it is repressed at 36°C or lower temperatures. In this way, when the organism leaves the human or chicken body, it will be killed, avoiding the release of the GMO in the environment.
After Hamburg we had a first meeting to discuss each other’s biosafety strategies and to see if we could create a strong partnership that would benefit both teams. We realised that, even though we shared the use of the temperature sensitive promoter pTlpA36, our constructs were quite different. In our system the pTlpA36 controls the expression of the antitoxin ccdA, while the toxin ccdB is under a constitutive promoter (BBa_K4244019). When the temperature drops, the promoter is inhibited and therefore blocks the expression of the antitoxin, leading to cell death. In contrast, NanoBuddy used pTlpA to control the expression of a tetR gene, which in turn inhibits the expression of a fatal CRISPR-Cas9 casette. The TetR in this system inverts the signal, so the chickens body temperature inactivates Cas9 and prevents cell death. Therefore, we decided to compare the two systems with the following research questions in mind.
What is the more efficient kill switch below 37 °C?Throughout the summer we regularly met online to discuss and troubleshoot the two biocontainment projects. Besides pTlpA, our team was testing another temperature sensitive promoter, pcspA, to compare them and to choose the one that suits our project best. When comparing the results, the cell death assay of pTlpA was not showing decreasing colony numbers with decreasing temperature, as expected. To try to solve this, the Groningen team sent us the pTRKH3 plasmid that contains a broad-range host ORI (pAMβ1 ori) [1]. We inserted our circuit with pTlpA containing the ccdA-ccdB in the pTRKH3 plasmid and tested it again in E. coli Nissle 1917 with a cell death assay. Even with their plasmid, the desired outcome was still not achieved.
Due to time constraint we did not manage to fully compare the two systems and we have not been able to answer our initial research question. This partnership allowed us to conclude that the other temperature sensitive promoter, pcspA, the results of which can be seen here, was the best choice for our project. Since our proof of concept should be modular and therefore also work in other organisms, it could be implemented in NanoBuddy instead of the pTlpA system.
Through a social media post we realised that our team and the Leuven team, SenSkill, had a shared goal: fighting colorectal cancer! In July, we had our first meeting in which we discussed shared interests and realised that both teams were working on a questionnaire directed to the general public. The aim was to understand the perception and the concerns regarding GMOs. Specifically, our goal was to understand how people perceive an engineered bacterium as a diagnostic. Meanwhile, SenSkill focused on colorectal cancer patients and their concerns, beside their perception on GMO. While discussing the answers to the questionnaire, one of the main findings was that people taking the SenSkill biosensor would like to know when to take it. This was in agreement with our stakeholders, as they had highlighted the importance to hand out the information regarding when to take the Colourectal self-test in the clearest way possible. Since we were constructing a model to study the behaviour of Colourectal in the human colon, we tried to answer this recurring question and used our model to design an administration strategy for our living diagnostic. Click here to read more about it.
During one of our meetings, we also realised that our projects had more in common! SenSkill was using a synthetic sensor to sense naringenin, while our team was working on a chimeric two-component system (TCS) to sense mucin. Our main challenge with the chimeric TCS was the fact that we experimentally observed leakiness of the genetic circuit as, even in absence of the input (mucin), our chimeric TCS showed signal (fluorescence). While troubleshooting our genetic circuit, we found out that SenSkill was developing a model for their synthetic receptor. Although the two systems were different, they had something in common: both chimeric sensors are activated by a specific input (mucin or naringenin) and a specific output is desired (control of gene expression). Hence, in August, when we obtained our experimental data about the GFP fluorescence response to different mucin concentration in a plate reader, we shared these with SenSkill to help them investigating the flexibility of their model. In return, they studied the quality of our dataset by exploring its signal/noise ratio. By validating the quality of our data, they helped us to design the data collection strategy for one of our modelling projects, click here to find out more.
Unfortunately, even considering the similarities of the two chimeric sensors, our data was not useful to fit their model, so we could not use it to investigate our leakiness problem. SenSkill’s main goal was to understand if their sensor behaves in a behaves in a dose-dependent way. However, the TCS does not display this behaviour and the data could not be applied. Although we could not solve the leakiness in our system, we think that having continuous exchange of ideas and meeting to troubleshoot problems that both teams were encountering has been really helpful!
