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

A new study published in Stress Biology has now provided one of the most detailed long-term pictures of how this adaptation unfolds. Researchers compared Bgt populations collected in the Northwestern Himalayan state of Himachal Pradesh, India, across two separate windows: 1994 to 1998 and 2015 to 2019. In total, 285 single-colony isolates were analyzed, comprising 215 historical isolates (155 conidial and 60 ascosporic) and 70 contemporary isolates (45 conidial and 25 ascosporic). Sampling spanned the major wheat-growing agro-climatic zones of the region, including the districts of Lahaul and Spiti, Kangra, Bilaspur, Una, Chamba, Shimla, and Hamirpur. Ascosporic isolates were derived from chasmothecia collected in the dry temperate zone, where the fungus completes its sexual cycle, while conidial isolates were established from actively sporulating field colonies.

The laboratory methodology was meticulous. Conidia from distinct colonies were transferred with a sterilized camel-hairbrush onto primary leaves of fourteen-day-old seedlings of the susceptible cultivars Agra Local or Lehmi, and each isolate was maintained in spore-proof isolation chambers with inoculum renewed every fifteen to twenty days. For ascosporic isolates, individual cleistothecia were excised and placed on moist blotter paper, and the ascospores they discharged were captured over young susceptible seedlings, with fungal colonies appearing about seven days after discharge. Virulence phenotyping employed international powdery mildew differential lines, including near-isogenic lines in the Chancellor background, sourced from the Indian Institute of Wheat and Barley Research, CIMMYT, and Punjab Agricultural University. Infection types were recorded ten days after inoculation on a modified zero-to-four scale, with reactions of zero to two scored as avirulent and three to four as virulent.

The historical population proved remarkably varied. Pathogenicity analysis of the 215 isolates from 1994 to 1998 on nine differential lines delineated 51 distinct pathotypes among conidial isolates and 15 among ascosporic ones. The broadest-spectrum conidial pathotype, P28 from Kukumseri in Lahaul and Spiti, overcame seven of the nine tested resistance genes, while several other pathotypes were virulent on six or five. At the opposite extreme, pathotype P20 from Joginder Nagar was avirulent on all tested genes. Among ascosporic isolates, pathotypes P1a and P12a, both from Kukumseri, each overcame an average of six resistance genes. By the contemporary period, the picture had changed substantially: evaluation of 70 isolates on 20 differential lines resolved 48 pathotypes, with the most virulent conidial pathotypes, Pt17 and Pt25, defeating 15 of the 20 resistance genes, and the ascosporic pathotype Pt44 overcoming 14.

To enable direct temporal comparison, the team analyzed both populations on a common set of nine differential lines, and the results revealed a paradoxical but statistically robust evolutionary trend. The mean virulence complexity per isolate increased significantly, from 3.92 in the historical population to 4.61 in the contemporary one, indicating that individual isolates had gained pathogenic capability. Yet overall population diversity declined: Simpson’s index fell from 0.888 to 0.864, normalized Shannon’s index from 3.085 to 2.414, and Kosman’s index from 0.420 to 0.283, with a concurrent drop in the Gleason index. Hill number analysis confirmed the pattern, with non-overlapping confidence intervals at all diversity orders showing a significant decrease in the contemporary population. Together, these findings point to a selective sweep in which a few highly virulent, genetically similar lineages have risen to dominance, likely driven by the widespread cultivation of varieties with a narrow genetic base.

Multivariate statistics reinforced the conclusion that the pathogen population had fundamentally restructured. Roger’s, Kosman’s, and mean character difference distances all indicated significant genetic differentiation between the two temporal populations, and permutational multivariate analysis of variance validated the shift in virulence structure with an R-squared of 0.082 and a p-value below 0.001. A beta-dispersion test showed a significant reduction in multivariate variance within the 2015 to 2019 population, and principal coordinates analysis visually separated pathotypes by decade, with the first two axes explaining 24.94 and 21.23 percent of virulence variation respectively. Historical pathotypes were widely dispersed, reflecting higher diversity, whereas contemporary pathotypes formed a tight cluster, indicative of a more genetically homogeneous population.

The study also tracked how virulence genes recombine into new combinations over time. The most striking change involved the Pm3b-Pm8 pair, which showed a co-occurrence of minus 25.8 percent in the historical population and was completely absent in the contemporary one, a full dissociation. Other significant dissociations included Pm2-Pm3c and Pm3b-Pm3c, each at minus 19.7 percent, Pm3b-Pm4a at minus 18.2 percent, and Pm2-Pm5 at minus 16.7 percent. Conversely, strong new positive associations emerged, with the largest increases in co-occurrence for Pm5a-Pm8 at plus 48.2 percent, Pm3c-Pm8 at plus 47.2 percent, and Pm3c-Pm5a at plus 45.2 percent, alongside substantial gains for Pm1a-Pm3c and Pm1a-Pm5a. The authors suggest that dissociated gene pairs may represent promising candidates for pyramiding, since the pathogen may struggle to recombine those virulences simultaneously, while new virulence clusters could signal an adaptive response capable of threatening pyramided resistances.

The temporal analysis of resistance gene efficacy delivered both warnings and encouragement. Fisher’s exact tests identified significant virulence frequency changes for Pm1a, Pm2, Pm3b, Pm3c, Pm5a, and Pm8, with virulence against Pm1a escalating dramatically from 6 to 63 percent over the two decades. Linear regression confirmed significant negative efficacy slopes for Pm1a, Pm3c, Pm5a, Pm6, and Pm8, marking them as broken or eroding. The defeat of Pm8 is directly tied to its historical deployment: the gene was inadvertently introgressed into popular Indian wheat varieties through linkage with leaf rust resistance gene Lr26 and yellow rust gene Yr9, and varieties carrying the Pm8/Lr26/Yr9 complex, such as PBW-343, HS-240, HPW-284, and HD-2967, were cultivated on a massive scale across the North Western Plain Zone and North Hill Zone. Virulence against Pm8 rose from under 10 percent in the early 1990s to more than 50 percent within a decade and now exceeds 80 percent. Similar deployment-driven pressure is evident for Pm1a and Pm6 in cultivars such as HS-542, DBW-179, WH-1181, HPBW 01, and DDK-1051.

In stark contrast, Pm2 and Pm4a demonstrated complete and durable effectiveness throughout the entire study period, maintaining low virulence pressure and stable efficacy for more than twenty years. Pm3b also showed consistently low virulence pressure. The durability of Pm2 and Pm4a against Himalayan Bgt populations mirrors findings from Egypt and Hungary, and the combination of the two genes has proven highly robust, suggesting synergistic effects that enhance durability. Earlier surveys from Himachal Pradesh and Punjab likewise recorded very low virulence on these genes, indicating that their effectiveness has persisted across more than three decades of pathogen evolution. The study additionally noted that Pm1c, Pm3b, the Pm2 plus Mld combination, and the four-gene pyramid Pm1 plus 2 plus 9 plus 12 remained effective during the 2015 to 2019 period, marking them as potential assets for breeding programs.

The broader lesson is a familiar one in plant pathology, but rarely documented with such temporal depth: monogenic resistance deployed at scale exerts intense directional selection, and the pathogen eventually answers. The researchers advocate a strategic shift toward pyramiding validated durable genes such as Pm2 and Pm4a with other effective sources, guided by continuous, region-specific virulence monitoring. They also emphasize that India currently lacks a systematic, focused breeding program for Bgt resistance, even though susceptible commercial varieties and vulnerable advanced breeding lines remain widely cultivated. As the pathogen’s sexual recombination in the dry temperate zone and prolific asexual mutation continue to generate novel virulence combinations, the study stands as both a warning about the fragility of single-gene defenses and a practical roadmap for building wheat varieties that can withstand the next two decades of evolutionary pressure.

Subject of Research: Long-term virulence evolution of the wheat powdery mildew fungus Blumeria graminis f. sp. tritici and the durability of Pm resistance genes in the Northwestern Himalayas

Article Title: Deciphering a pathogen’s evolution: a two-decade longitudinal study reveals virulence shifts and identifies durable Pm genes against Himalayan Blumeria graminis f. sp. tritici populations

Article References: Mehta, A., Kaur, H., Basandrai, D., Basandrai, A. K., Ahanger, S. A., Sharma, B. K., Sharma, R., Sharma, S., & Basu, U. (2026). Deciphering a pathogen’s evolution: a two-decade longitudinal study reveals virulence shifts and identifies durable Pm genes against Himalayan Blumeria graminis f. sp. tritici populations. Stress Biology, 6(1), Article 37. https://doi.org/10.1007/s44154-026-00299-0

Image Credits: AI Generated

DOI: 10.1007/s44154-026-00299-0

Keywords: wheat powdery mildew, Blumeria graminis f. sp. tritici, virulence evolution, Pm resistance genes, durable resistance, gene pyramiding, Himachal Pradesh, pathotype diversity, selective sweep, plant pathology, wheat breeding, virulence surveillance