Our lab work always complied with iGEM safety policies. All our work with bacteria was done in a BSL 2 lab in biosafety cabinets or next to the bunsen burner. Additionally we used a BSL 1 media kitchen. Safety was also ensured when working with flammable or corrosive chemicals. All team members who worked in the wet lab received security training before starting their lab work, covering topics such as lab access and behavior, proper technique when working with microbes, sterilization and the use of biosafety equipment, as well as chemical and fire safety. They also received specific training on how to handle the organisms and chemicals we used. Contaminated material was collected, autoclaved and disposed centrally by the institution.
E. Coli prolif.
→
Rhl-QS conc.
⊺
↓
Healthy (High) pH
←
Lactic Acid conc.
Figure 1: The negative feedback loop of LactoBack minimizes runaway bacterial proliferation.
We only worked with organisms from the white list, specifically a non-pathogenic strain of E. coli, as well as Lactobacillus crispatus and Lactobacillus gasseri (though the latter two were not genetically engineered) and took care to respect all of iGEM's safety policies. We used no gene drives, no human or animal experimentation, and used no novel resistance factors. When it comes to the safety of the hypothetical finished product, meaning LactoBack as a medicine, we would ensure containment and patient safety using a kill switch. Due to time constraints we could not add the kill switch in vitro. We also hoped to integrate our parts into the bacterial chromosome to further ensure containment, though the realization of that plan would require more time.
Without clinical trials it is difficult to say whether there is a risk of infection; however the self regulatory nature of LactoBack should keep that risk to a minimum, since the engineered organism produces lactic acid at a certain population threshold, reducing its own population (Figure 1). Additionally, the added parts introduce a fitness defect into the engineered E. coli by forcing it to produce lactic acid at the expense of other metabolic pathways. This makes uncontrolled growth even more unlikely.
To work with genetically modified organisms a safety form had to be submitted to the government. (ECOGEN safety form) The safety form was filled out and submitted via the account of Diana Albertos-Torres (technical lab manager). Diana Albertos-Torres and Prof. Dr. med. Dr. phil. Adrian Egli proofread and confirmed the validity of the safety form.
A useful kill switch in LactoBack must induce cell death in two situations: escape of bacteria outside the patient body, and antibiotic treatment in case of patient intolerance, side effects or uncontrolled infection. Additionally, such a kill switch must be evolutionarily stable, since the engineered bacteria are meant to persist in the vaginal microbiome.
Rottinghaus et al.[1.] have engineered a very stable CRISPR based 2-input chemical- and temperature-responsive kill switch that fulfills both of these aims, killing bacteria efficiently at temperatures below 32°C or upon anhydrotetracycline (aTc) exposure. However, we could not simply add a plasmid to LactoBack, since the stability of the 2-input switch depends upon knocking out key components of the bacterial SOS response. To be exact, multiple genes that drive DNA mutagenesis were knocked out of the bacterial genome, and a lethal mutation in the infA gene was introduced, making the bacteria dependent on the kill switch plasmid, which also had an intact copy of infA. Additionally the 2-input switch was characterized in E. coli Nissle 1917 (EcN), while LactoBack is currently based on E. coli K12 MG1655, and in a medical application should be based on a chassis that is naturally present in the vaginal microbiome (e.g. L. gasseri, though the dominant species varies geographically). A first step would be to add the LactoBack parts to an EcN chassis already containing the Rottinghaus et al. kill switch. The authors claim that, since their kill switch is based upon genetic parts not specific to E. coli, "similar kill switches can be engineered for a larger panel of probiotic microbes", making this kill switch a promising avenue if we plan on moving LactoBack to a different chassis.
A further promising possibility is the Cryodeath kill switch proposed by Stirling et al.[2.]. This only causes cell death at lower temperatures, and seems to be less sensitive than the 2-input kill switch mentioned above, as it only causes efficient killing at temperatures of 22°C and below. A useful property which would aid quick implementation is that Cryodeath has already been characterized in E. coli MG1655.