Conditional privatization of a public siderophore enables Pseudomonas aeruginosa to resist cheater invasion

Abstract

Understanding the mechanisms that promote cooperative behaviors of bacteria in their hosts is of great significance to clinical therapies. Environmental stress is generally believed to increase competition and reduce cooperation in bacteria. Here, we show that bacterial cooperation can in fact be maintained because of environmental stress. We show that Pseudomonas aeruginosa regulates the secretion of iron-scavenging siderophores in the presence of different environmental stresses, reserving this public good for private use in protection against reactive oxygen species when under stress. We term this strategy "conditional privatization". Using a combination of experimental evolution and theoretical modeling, we demonstrate that in the presence of environmental stress the conditional privatization strategy is resistant to invasion by non-producing cheaters. These findings show how the regulation of public goods secretion under stress affects the evolutionary stability of cooperation in a pathogenic population, which may assist in the rational development of novel therapies.

Fig. 1

Environmental stresses trigger the accumulation of a bacterial siderophores in the periplasms of P. aeruginosa. Representative bright-field (gray) + confocal images show that a illuminations of violet-laser (405 nm, 6.00 $\mathrm{mW}{\mathrm{cm}}^{-2}$) trigger the accumulation of a bacterial siderophores (pyoverdine, PVDI) in wild-type P. aeruginosa within 4 min and b $\Delta \mathrm{pvdA}$ cells did not produce PVDI despite illuminations (6.00 $\mathrm{mW}{\mathrm{cm}}^{-2}$), where the PVDI is indicated by blue colors. c Distributions of PVDI fluorescence intensities along the vermillion line across the bacterium shown in a for the time t = 0, 60, 120, 240 s, from bottom to top. d Photon-stress dependence of PVDI fluorescence intensity in single cells, where cells were exposed to violet-laser for 2 min. The dots and the line present data arising from single bacteria and from the average of multiple bacteria, respectively. e, f Normalized histograms of PVDI (blue bar) fluorescent intensities or ROS-specific dye H₂DCFDA fluorescent intensities (bluish green bar) arising single cell in the presence of e different photon-stresses (0.75, 3.00, 4.50, 6.00, and 9.00 $\mathrm{mW}{\mathrm{cm}}^{-2}$) or in addition of f different amount of tobramycin (0.0, 1.0, 2.0, 3.0, and 4.0 µg mL⁻¹), where cells were exposed to violet-laser for 2 or 8 min or tobramycin for 1 or 7 h, respectively. Scale bar for all images are 2 µm

Fig. 2

Accumulation of pyoverdine in bacteria allows them to survive in the presence of environmental stresses. a Representative bright-field (gray) + multi-color confocal images show that the accumulation or the exogenous addition of PVDI (5.0 $\mu \mathrm{M}$) enables wide-type P. aeruginosa or $\Delta \mathrm{pvdA}$ mutant to survive in the presence of a stronger photon-stress $\left(3.00\mathrm{mW}{\mathrm{cm}}^{-2}\right)$, where blue, bluish green or vermillion colors represents PVDI, H₂DCFDA or propidium iodide (PI) fluorescent intensity respectively, $\Delta \mathrm{pvdA}$ or $\Delta \mathrm{pvdA}\Delta \mathrm{fpvA}$ represents a mutant that deficient in production of PVDI or deficient in both production of PVDI and uptake of exogenous PVDI, respectively. +exo PVDI represents the exogenous addition of 5.0 $\mu \mathrm{M}$ PVDI. b Photon-stress dependence of PVDI (blue), H₂DCFDA (bluish green) or PI fluorescence intensity (vermillion) in single cells, where the dots present data from single bacteria. $\Delta \mathrm{pvdA}$ and $\Delta \mathrm{pvdA}\Delta \mathrm{fpvA}$ + exo PVDI groups show significantly higher percentage than that of the wide type group, both for H₂DCFDA channel and PI channel (one-way RM ANOVA versus wide type data, p < 10⁻⁶ for $\Delta \mathrm{pvdA}$ and $\Delta \mathrm{pvdA}\Delta \mathrm{fpvA}$ + exo PVDI). $\Delta \mathrm{pvdA}$ + exo PVDI group shows no significantly difference with wide type group, both for H₂DCFDA channel (p = 0.8054) and PI channel (p = 0.2557). The black dashed line represents an intensity threshold to distinguish whether the fluorescence intensities arising from single cells is significantly greater than the average. Single cells were exposed to light stimulations for 5 hours. Scale bar for all images are 4 $\mu \mathrm{m}$

Fig. 3

P. aeruginosa tunes down the efflux of PVDI in the presence of environmental stresses. a Representative regular + fluorescent photos show that the amount of PVDI in the supernatant of bacterial cultures negatively relates to the concentration of tobramycin, where all the bacterial cultures were harvested at an identical bacterial density (O.D.₆₀₀ = 0.3). b Decline of PVDI fluorescence intensity in the presence of different photon-stresses indicates that bacteria tune down the efflux of PVDI, where dots and lines represent the average intensities and the fitting lines using exponential fitting, respectively, with the magnitude of the photon-stress of 0.1, 0.15, 0.3, 0.6, 0.66, 0.72, 0.82, 1.2, 1.8, and 3.0$\mathrm{mW}{\mathrm{cm}}^{-2}$, from bottom to top. c-e Photon-stress dependence of c $\left(\gamma +\mu \right)$, d μ, and e $\gamma ∕\mu$, where $\gamma$ or μ represents the PVDI efflux rate or bacterial growth rate, respectively. The error bars in c–d are the standard deviation of experimental values

Fig. 4

Conditional privatization stabilizes bacterial cooperation in the presence of environmental stresses. a–c Direct evolution of the producer (cooperator, wild-type) and non-producer strain (cheater, $\Delta \mathrm{pvdA}$ mutant) of PVDI in the presence of different amounts of FeCl₃ (a 1 × 10⁻⁵ $\mathrm{\mu M}$), (b 0.005 $\mathrm{\mu M}$), and (c 0.05 $\mathrm{\mu M}$) and different amounts of tobramycin (0.0 to 2.0 $\mathrm{\mu g}{\mathrm{mL}}^{-1}$), where symbols and lines represent the experimental data and the theoretical calculations, respectively. The final concentration of tobramycin in each panel is 0.0, 0.125, 0.25 0.5, 1.0, and 2.0 $\mathrm{\mu g}{\mathrm{mL}}^{-1}$, from bottom to top, respectively. d Producer fraction as a function of the ferric abundance and the normalized ROS stress, where the normalized ROS stress = [Tobramycin]/IC₅₀, the symbol [] represents the concentration, IC₅₀ = 2.0 $\mathrm{\mu g}{\mathrm{mL}}^{-1}$. Colors represent the fraction of producers in the end of direct evolution, red or blue color indicates the cooperator or cheater dominated, respectively. Symbols represent the evolutionary outcomes arising from experiments in the presence different amount of ferric ions or tobramycin; i.e., 10⁻⁵ $\mathrm{\mu M}$ (a), $0.005\mathrm{\mu M}$ (b), 0.05 $\mathrm{\mu M}$ (c) FeCl₃, 0.125 (II), 0.25 (III), 0.5 (IV), 1.0 (V), 2.0 (VI) $\mathrm{\mu g}{\mathrm{mL}}^{-1}$ tobramycin. e Schematic showing the principle response curves of the efflux of the public goods in the presence of various environmental stresses, where the curve with cyan, red or magenta color represents the conditional privatization, loyal cooperation or selfishness respectively. The error bars in a–c are the standard deviation of experimental values

Fig. 5

Evolutionary outcomes dominated by different strategies. a Theoretical calculations show $\gamma ∕\mu$ dependence of producer fractions as a function of the abundance of ferric ions and the normalized environmental stress, where $\gamma ∕\mu$ ranges from selfishness ($\gamma ∕\mu$ = 0.1) to loyal cooperation ($\gamma ∕\mu$ = 50). b Schematic showing that conditional privatization stabilizes bacterial cooperation in the presence of environmental stresses