Engineered probiotic Escherichia coli can eliminate and prevent Pseudomonas aeruginosa gut infection in animal models Original paper

What was studied?

Engineered Escherichia coli Nissle 1917 Pseudomonas aeruginosa control was tested as a live targeted therapy for gut infection. The authors rebuilt the earlier “sense-and-kill” circuit in a clinically used EcN background. The circuit sensed the P. aeruginosa quorum signal 3OC12-HSL. It then triggered lysis and release of pyocin S5, an antipseudomonal bacteriocin. It also released dispersin B to break biofilm matrix. The strain was made D-alanine auxotrophic to keep the plasmid without antibiotics and to improve safety. The study asked if this single modified probiotic could clear established gut infection and could prevent later colonization.

Who was studied?

Experiments used two in vivo infection models. First was infected Caenorhabditis elegans. Worms received GFP-labelled P. aeruginosa in the gut and then the different EcN variants. Survival and gut fluorescence were tracked. Second was a streptomycin-treated mouse model. Mice received 10¹⁰ CFU P. aeruginosa by oral gavage and developed stable gut colonization of 10⁴–10⁵ CFU/g. They were then treated with wild-type EcN, lysis-only EcN, EcN making only pyocin, EcN making only dispersin B, or the full engineered strain EcN SED (sensor, pyocin S5, E7 lysis, dispersin B). Prophylaxis was also tested by precolonizing mice with EcN before pathogen challenge.

Most important findings

The engineered EcN SED detected P. aeruginosa at far lower HSL levels than the earlier circuit and expressed killing factors for a longer period. In vitro, it killed P. aeruginosa even when EcN numbers were lower, and it dispersed mature biofilms by ~80%. In C. elegan,s all EcN strains improved survival versus infection alone. But EcN SED and EcN SE (without dispersin B) gave the most extended survival and the highest gut clearance. In mice with established gut infections, only EcN SED produced a significant decline in P. aeruginosa faecal counts, reaching approximately 77% clearance after 6 days.

Key implications

This work demonstrates that a probiotic can be engineered to target a specific gut pathogen with high specificity and execute two coordinated actions. It senses, then it kills, then it breaks biofilm. It also demonstrates that containment can be established through auxotrophy and self-lysing. For clinicians, this suggests a potential future tool for high-risk settings where gut P. aeruginosa serves as a silent reservoir, such as oncology, intensive care, and very-low-birth-weight infants. It could reduce the need for broad antibiotics that damage commensals. It also illustrates how microbiome therapeutics should be logged, categorized by chassis (EcN), sensor (AHL lasR), payloads (pyocin S5, dispersin B), and containment (Δalr ΔdadX), allowing databases to link them to Pseudomonas-dominant dysbiosis states.