Current BiologyEvolutionary BiologyPeer-reviewed journal article
Cheating on cheaters dramatically affects social interactions in Pseudomonas aeruginosa
Özhan ÖzkayaiD, Roberto BalbontíniD, Isabel GordoiD, Karina B. XavieriD
Abstract
Bacterial cooperation can be disrupted by non-producers, which can profit from public goods without paying their production cost. A cheater can increase in frequency, exhausting the public good and causing a population collapse. Here we investigate how interactions among two cheaters for distinct social traits influence the short and long-term dynamics of polymorphic populations. Using as a model Pseudomonas aeruginosa and its extensively studied social traits, production of the siderophore pyoverdine and the quorum sensing regulated elastase, we analyzed the social dynamics of polymorphic populations under conditions where the two traits are required for optimal growth. We show that cheaters for either trait compete with both the wild type and each other, and that mutants for pyoverdine production can prevent a drastic population collapse caused by quorum sensing cheaters. A simple mathematical model suggests that the observed social dynamics are determined by the ratio of the costs of each social trait, such that the mutant which avoids paying the highest cost dominates the population. Finally, we demonstrate how quorum sensing regulation can avoid the full loss of cooperation.
- organism: bacteria (other)
- research focus: host-pathogen & immune response
- approach: experimental assay
- approach: simulation
- approach: statistical model
In-depth summary
Claude Sonnet 4.6 · xhigh effort · 8 Nov 2026 · from the full text of version 2 (23 Jan 2018) · CC BY 4.0
When pyoverdine and elastase are both needed, a pvdS mutant outcompetes the lasR cheater in P. aeruginosa and averts population collapse; a cost-ratio model with quorum sensing explains this.
Cheater invasion and population collapse have been seen in laboratory studies of a single social trait. This paper asks what happens when two traits are required at once, where a cheater for one trait can act as a cooperator for the other.
Key results
ShownWhen both elastase and pyoverdine are required, the pvdS mutant stops the lasR mutant from cheating, and pvdS itself cheats on both the wild type and lasR.
In iron-depleted casein medium, lasR increased in frequency against the wild type alone (9:1) but not in the 8:1:1 triple co-culture. pvdS had relative fitness above 1 against the wild type and against lasR. No cheating was seen where neither trait was required. · Cheating capacity of lasR mutant depends on…
ShownWhen lasR expanded, it caused a drastic population collapse in three of the four scenarios: WT+lasR with elastase only (5 of 6 replicates), WT+lasR+pvdS with elastase only (all 6; pvdS did not expand), and WT+lasR with both traits required.
With elastase only, lasR reached up to 90% and collapse followed by day 12 in 5 of 6 WT+lasR replicates, and in all 6 triple co-cultures. With both traits required, WT+lasR collapsed by day 18 as lasR neared 80%. · Invasion of lasR mutant leads to a drastic…
ShownWhen both traits are required and pvdS is present, pvdS takes over without a drastic collapse and lasR stays rare. A small share of pyoverdine producers can sustain the whole culture.
At day 18, lasR was about 3% and pvdS averaged 96%, with no collapse (6 replicates). Mixtures with 3-4% WT cells had yields similar to WT monocultures. Starting at 75-85% pvdS, a collapse was observed. · pvdS prevents the drastic population collapse…
ShownChanging the environment mid-propagation reverses the outcome, which points to social interactions rather than de novo mutations as the driver of the dynamics.
Switching WT+lasR from casein to casamino acids after day 6 removed the lasR advantage and prevented collapse. Adding iron to the two-trait medium after day 6 reverted pvdS expansion and led to collapse at day 18 (N=6 each). · Alterations in carbon or iron source…
SuggestedA three-strain public-goods model predicts that the cheater avoiding the costlier trait dominates. Adding quorum-sensing regulation explains why lasR does not reach fixation, and AHL addition supports this.
The model needs a cost ratio c1/c2 and predicts cheater coexistence only when c1 = c2. In WT+lasR propagations starting at 80-90% lasR, lasR failed to fix; with 5 μM 3OC12-HSL it expanded to fixation. Model parameters were chosen to resemble Figures 1-2. · A mathematical model of a 3-way public goods…; Figure legends
Methods, data and reproducibility
Short-term relative-fitness competitions and 18-day serial propagations (48 h transfers) of P. aeruginosa wild type, lasR and pvdS mutants in media that make elastase, pyoverdine, both or neither essential. A simple mathematical public-goods model, with and without a Hill-function quorum-sensing term, was simulated alongside.
- P. aeruginosa PA01 wild type plus lasR::GmR and ΔpvdS::GmR knockout mutants, in mono-, double and triple co-cultures · Pseudomonas aeruginosa · n = at least 6 biological replicates per panel and per propagation condition
- Methods & tools
- GraphPad Prism 6, Mann-Whitney test, Kruskal-Wallis test with Dunn's correction, colony-count plating (CFU/ml) with fluorescence and morphology scoring, PCR genotype check, 3OC12-HSL autoinducer, public-goods mathematical model (Models 1 and 2)
- Reproducibility gaps
- No code or data deposit is stated. Model equations and the parameter table are in the Supplemental Information. The pvdS lasR double mutants from longer runs are cited as data not shown.
Limitations (3)
- The basic model predicts full fixation of the expanding mutant and cannot explain the lack of lasR fixation; it needed an added quorum-sensing term. It ignores spatial structure, diffusion and privatization. · stated by the authors
- In one of six WT+lasR replicates under elastase-only conditions there was no collapse, possibly from a non-social beneficial mutation in the wild type. Double pvdS lasR mutants arose in longer propagations and their effects were not examined. · stated by the authors
- Only one strain background (PA01) and well-mixed liquid media were tested. The model cost parameters were chosen to resemble measured fitness ratios, not fitted and validated on held-out data. · noted in this summary
Read this if you study microbial cooperation and cheating and want an experimental and modelling case for how multiple social traits and quorum sensing shape population collapse.
- Citations
- 64
- FWCI
- 2.93
- Field percentile
- top 7%
Reproducibility ledger
Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.
Sign in to file a reproducibility report.
Related papers
- Privatisation rescues function following loss of cooperation
S. B. Andersen, M. Ghoul, … A. S. Griffin·bioRxiv·May 22, 2018
- Ecology drives the evolution of diverse siderophore-production strategies in the opportunistic human pathogen Pseudomonas aeruginosa
A. R. T. Figueiredo, A. Wagner, R. Kuemmerli·bioRxiv·Jan 27, 2021
- Modification of social cheating facilitates the stabilization and chronic infection of polymorphic Pseudomonas aeruginosa population
- Cheats can boost the success of a cooperative invader
Luke Lear, Angus G. J. Buckling, Elze Hesse·bioRxiv·Apr 30, 2025
- Biophysical Constraints Dictate the Stability of Social Traits in Pseudomonas aeruginosa Aggregates
Kathleen O'Connor, Nicholas Bradley Larson, … Stephen P. Diggle·bioRxiv·Mar 19, 2026
- In vivo microscopy reveals the impact of Pseudomonas aeruginosa social interactions on host colonization
C. Rezzoagli, E. Granato, R. Kuemmerli·bioRxiv·Mar 20, 2019