How a Deadly Bacterium Outsmarts Amoeba Predators in Soil and Water

In the damp soils and water films where the opportunistic human pathogen Pseudomonas aeruginosa makes its home, the bacterium is not the top of the food chain. It is prey. Wandering phagocytic amoebae such as Dictyostelium discoideum, professional hunters that can swallow several bacteria per minute, patrol these environments constantly. Yet P. aeruginosa survives, thrives, and occasionally goes on to cause devastating infections in cystic fibrosis patients and immunocompromised people. A new study published in the journal iScience by Imen Ayadi, Pierre Cosson, Alexandre Persat and colleagues at institutions in Switzerland reveals the molecular armor that lets this bacterium shrug off its amoeboid predators, and in doing so offers a striking illustration of how human virulence can be an accidental byproduct of life in the wild.

The research team set out to dissect exactly how P. aeruginosa escapes two distinct killing strategies deployed by D. discoideum. In previous work, the same group had shown that the amoeba can destroy the bacterium in two ways: by ingesting cells and killing them inside phagosomes, and by secreting lytic effectors that destroy bacteria left outside the cell. The intracellular route depends on a phagosomal protein called Kil2, while the extracellular route relies on a secreted lytic protein called AplA and on NoxA, the sole superoxide-producing NADPH oxidase of the vegetative amoeba. Because P. aeruginosa largely evades ingestion in the first place, the extracellular mechanism is critical for the amoeba’s success, making the bacterium’s defenses against both routes a compelling evolutionary puzzle.

Using live-cell time-lapse microscopy with GFP-expressing bacteria, the researchers quantified phagocytosis by measuring the percentage of bacterial cells actually ingested after a contact with an amoeba. Wild-type P. aeruginosa PAO1 proved remarkably slippery: only about 17 percent of contacts ended in ingestion. The team then systematically tested mutants lacking individual virulence factors, including the two lipopolysaccharide antigens of the bacterial surface, the common polysaccharide antigen (CPA) and the O-side chain specific antigen (OSA), as well as mutants defective in the Las and Rhl quorum-sensing systems. Only one change mattered for phagocytosis. Bacteria lacking OSA-type LPS were swallowed far more efficiently, while mutants lacking CPA, Las, or Rhl remained as resistant as the wild type. An extended panel of mutants, including defects in the type 3 and type 6 secretion systems, showed no significant increase in uptake.

The finding held across serotypes. The researchers repeated the experiment with two additional clinical laboratory strains, PAK and PA14, which carry entirely different OSA O-antigen chemistries belonging to the O6 and O19 serotypes respectively. All three wild-type strains were poorly ingested, and in every case, disrupting OSA synthesis made the bacteria more vulnerable to amoeboid grazing. This suggests that resistance to phagocytosis by D. discoideum is a common trait of P. aeruginosa isolates, and that the protective function of OSA LPS transcends the specific sugar-repeat structure of any one serotype.

Intriguingly, the same LPS shield does nothing to protect the bacterium from human immune cells. When the team incubated wild-type and OSA-deficient bacteria with differentiated THP1 macrophages, both were ingested with equal efficiency. This divergence between amoeboid and mammalian phagocytosis hints at a fundamental difference in how the two types of predator recognize their prey. D. discoideum uses phagocytic receptors that bind a broad range of hydrophilic surfaces with low specificity, and the OSA layer of P. aeruginosa simply does not fit these receptors. Human phagocytes, by contrast, rely on opsonins such as complement and antibodies, and on receptors like TLR4 that recognize the lipid A portion of LPS, so the O-antigen cloak is largely irrelevant to them.

To confirm that the bacterium’s invisibility reflects a failure of recognition rather than an active anti-phagocytic signal, the researchers staged a molecular decoy experiment. They mixed GFP-expressing P. aeruginosa with mCherry-expressing Klebsiella pneumoniae, a bacterium the amoeba eats readily. When an amoeba encountered the two bacteria side by side, it consistently swallowed the Klebsiella and ignored the Pseudomonas, even after prolonged contact. If P. aeruginosa were emitting an inhibitory signal, the neighboring Klebsiella should have been spared as well. Even more telling, coating P. aeruginosa with anti-LPS antibodies dramatically increased its ingestion, providing the amoeba’s receptors with a protein surface they could grip. An irrelevant antibody that does not bind the bacterium had no effect. The conclusion is elegant: P. aeruginosa is not rejected, it is simply not seen.

Phagocytosis, however, is only half the battle. The amoeba’s second weapon, extracellular lysis, targets bacteria that escape ingestion. When wild-type P. aeruginosa touched an amoeba and moved away, only about 10 percent lysed. But mutants lacking the Rhl quorum-sensing system were lysed more often, and mutants lacking Las, or both Las and Rhl, were lysed far more efficiently. Since the double mutant fared no worse than the lasR mutant alone, LasR emerges as the main modulator of resistance to extracellular destruction. Notably, LPS mutants were just as resistant as wild type to this route of killing, showing that the two defenses are mechanistically independent: LPS blocks entry, quorum sensing blocks lysis.

What does Las quorum sensing actually buy the bacterium? The team probed this by testing sensitivity to the individual killing mechanisms of the amoeba. Because NoxA produces superoxide, they challenged bacteria with hydrogen peroxide. LasR mutants, but not Rhl mutants, were dramatically more sensitive to peroxide killing, indicating that Las induction confers resistance to reactive oxygen species. The researchers then collected the proteins secreted by D. discoideum and exposed bacteria to this cocktail. Again, only lasR-deficient strains were efficiently lysed, showing that Las induction also protects against the amoeba’s secreted bacteriolytic effectors. A proteomic analysis of the secretome identified 205 proteins, of which 177 were bona fide D. discoideum secreted proteins bearing signal sequences, including at least 21 putative lytic proteins and at least 63 degradative enzymes such as lipases, glycosidases, nucleases and proteases. This catalog, which overlaps substantially with previously identified bacteriolytic fractions, provides a roadmap for future work on how amoebae kill bacteria.

The mechanism behind Las-mediated resistance may lie in the bacterial membrane itself. Using the hydrophobic fluorescent probe NPN, which fluoresces strongly only when it inserts into bacterial membranes, the team measured how accessible the outer membrane was to external agents. Wild-type membranes admitted little probe, Rhl mutants admitted more, and Las mutants admitted the most. Induction of the Las quorum-sensing system therefore appears to tighten the bacterial envelope, limiting the penetration of free radicals and lytic proteins alike. Interestingly, when the researchers tested individual genes known to be induced by Las, none alone accounted for the resistance phenotype, suggesting that either an untested critical gene or the synergistic action of multiple Las-regulated factors is responsible. Genes encoding catalase and superoxide dismutase, which are quorum-sensing regulated in P. aeruginosa, are obvious candidates for future study.

The evolutionary implications are perhaps the most provocative part of the story. In the environment, P. aeruginosa faces relentless selection from amoeboid grazers, and this study shows that two of its most celebrated virulence-associated traits, OSA-type LPS and quorum sensing, are precisely the traits that shield it from predation. Yet in chronically infected patients, the bacterium often does the opposite: strains frequently evolve from smooth LPS with extended O-antigens to rough LPS lacking OSA side chains, and quorum-sensing-negative variants commonly arise and replace the infecting population. The explanation proposed by the authors is that these defenses pay off only in the wild. In the human body, where immune recognition differs and biofilm-protected bacteria face different pressures, the OSA cloak and the quorum-sensing program lose their advantage and can even become liabilities. This is a textbook case of coincidental evolution, in which virulence toward humans is a byproduct of adaptation to a completely different ecological arena. The authors caution that their work examined three laboratory strains and one amoeba strain under laboratory conditions, and that broader surveys of environmental and clinical isolates will be needed to confirm how general these selective pressures are. Still, the message is clear: to understand why P. aeruginosa is such a formidable pathogen, we may need to watch it where it learned its tricks, in the mud, hunted by amoebae.

Subject of Research: Molecular mechanisms by which Pseudomonas aeruginosa LPS and quorum sensing protect against predation by Dictyostelium discoideum amoebae

Article Title: LPS and quorum sensing protect Pseudomonas aeruginosa bacteria from phagocytosis and destruction by Dictyostelium discoideum amoebae

Article References: Ayadi, I., Pain, E., Oddy, J., Lamrabet, O., Vijay, H., Koehler, T., Distler, T., Persat, A., Munoz-Ruiz, R., & Cosson, P. (2026). LPS and quorum sensing protect Pseudomonas aeruginosa bacteria from phagocytosis and destruction by Dictyostelium discoideum amoebae. iScience, 29(11), Article 117912. https://doi.org/10.1016/j.isci.2026.117912

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117912

Keywords: Pseudomonas aeruginosa, Dictyostelium discoideum, lipopolysaccharide, quorum sensing, phagocytosis, amoebae, virulence evolution, coincidental evolution, host-pathogen interaction, reactive oxygen species, O-antigen, environmental microbiology