Safety and responsibility

Fault tree: How to choose the best strategy for our treatment ?

In the consequence tree, we consider here the Iron Coli as if it already produced the siderophores, and thus have iron under a chelated form. We imagine then, what Iron Coli may become after their action in the jejunum.

(1) Free iron needs to be chelated by siderophores to be internalized by bacteria. Thus, a higher concentration could favor any bacterial growth. However, the pathogenicity of harmful strains is characterized by the expression of a higher amount of siderophores, thus positively selecting them among others.

(2) If Iron Coli dies after it released its siderophores without internalizaing them after chelation, both pathogenic and non-pathogenic strains have access to the chelated iron. This should not favor one or another.

(3) In case Iron Coli survives, the expression of siderophores in the presence of iron is a suicidal behaviour if the ion comes to be in excessive amount. We hypothesize that the bacteria may lose the plasmids for its own survival because the selective pressure is lost. However, if the overexpression of siderophores appears as beneficial for bacterial growth, the strain has an advantage compared to the rest of the microbiome.

(4) As expected, mutations and genetic information sharing among bacteria are the main concerns for the patients and the general public. We can see through this fault tree that this scenario is the less likely to happen.

How to improve the safety of our system ?

To minimize the risks for our iron chelating bacteria, it is very clear that the best case scenario is the death of our bacteria.With the help of our fault tree, we decided to use a temporary strategy call "flush strategy". However, zero risk does not exist, so we need to give more thought to new ways of improving our system. We propose there some possible improvement for our treatment.

Physical containment

To deliver Iron Coli, we designed a capsule which disolve in the upper intestines. Our bacteria would be trap in the viscous matrice (composed hydroxypropylmethylcellulose) and be evacuated in the bowels. With this physical containment, we limitate the colonization of the upper intestine by Iron Coli and the interaction with other species of the microbiom.

Use of safer chassis

For our experiments, we used common E.coli strain as MG1655, BL21 or TOP10 for pratical reasons. However, for medical implementation, safer strain would be recquired. Here are some chassis, commonly used as probiotic that we could choose:

• E.coli Nissle 1917 strain

  E.coli Nissle 1917 (EcN) is a commensal strain used as a probiotic for treatment of various diseases from diarrhoea to inflammatory bowel diseases¹. As we worked on E.coli, our system would work without any modification. Nevertheless, even if this strain has no Virulence factor, EcN is also an excellent colonizer of the human gut², which does not match up with our flush strategy (even if we use our physical containment). However, we could use an engineered EcN that could not form bio-film and thus not colonize the intestine.

• Lactic acid bacteria
  Lactococcus and Lactobacillus are also frequently used as probiotic³. Nevertheless, those bacteria are also good colonizer of the human gut. And as we worked with genes of E.coli, our systems are not adapted for them. However, it had also recently be demonstrated that some strain, as Lactobacillus acidophilus, could be used as a DNA delivery factor in mouse ( the mechanism of plasmid transfer from bacteria to eukaryote cell remains unknown)⁴. Thus, we could imagine an alternative strategy for our treatment: an engineered Lactobacillus that would transfer a plasmid with correct version of the gene HFE to human enterocytes !

Destruction system implementation

We do not want our bacteria to colonize human gut, or to spread in the environment. Thus we will need to add a destruction system in our Iron Coli. Here are some exemple of systems we could use:

• Kill-switch
  

  Kill-switch are one of the most know destruction system and are frequently used in the iGEM competition (see the database created by Paris Bettencourt 2012 team). As we designed a capsule to deliver our Iron Coli in a viscous matrice (with HPMC), we could imagine to incorporate a molecule that will repress the kill-switch. Once the viscous matrice is no longer on the intestines, and so our repressor molecule, the kill-switch would be activated, and then kill the bacteria that remain in the guts. A suitable kill-switch would be for exemple the Biosafety-System araCtive system developed by Biefield 2013 team, which is repressed by L-rhamnose.

• Unnatural amino acids
  

  It is possible to introduce unnatural amino-acids (UAA) with orthogonal system. As for the kill-switch, the orthogonal system would be regulated by the presence of a molecule in the viscous matrice. Then we could use in 2 different ways. First, we can add UAA in our enterobactins enzymes genes. When the molecule is no longer in the intestine, the orthogonal system would no longer be produced and so our different enzymes. Then we can also use UAA as a kill-switch by mutation of an essential gene, exemple Tyrosyl tRNA synthetase, that would be expressed only on the presence of our molecule. Without the recquired molecule, our Iron Coli will die. For more detail, see P.A.S.E 2 System designed by BGU Israel 2013 team.

Chromosomal integration

Horizontal transfer, intra or inter specific, is very common with bacteria. To limit this problem, we could integrate our systems directly into bacterial chromosom of our Iron Coli, thus stopping at least plasmid transfer with the microbiom.

Enterobactin engineering

Even if we made a safe Iron Coli using previous system, a major issue remains: enterobactins produced by our bacteria are released in the lumen of the duodenum and can be catch by both non-pathogenic and pathogenic bacteria. To settle this problem, we could engineer our enterobactin, with slightly modifications, to make them exportable by our Iron Coli but non-importable for any bacteria. Thus, enterobactins would chelate iron and be eliminated in the bowels without inducing modification of the microbiom.