Copper selects for siderophore-mediated virulence in Pseudomonas aeruginosa

Abstract

Background: Iron is essential for almost all bacterial pathogens and consequently it is actively withheld by their hosts. However, the production of extracellular siderophores enables iron sequestration by pathogens, increasing their virulence. Another function of siderophores is extracellular detoxification of non-ferrous metals. Here, we experimentally link the detoxification and virulence roles of siderophores by testing whether the opportunistic pathogen Pseudomonas aeruginosa displays greater virulence after exposure to copper. To do this, we incubated P. aeruginosa under different environmentally relevant copper regimes for either two or twelve days. Subsequent growth in a copper-free environment removed phenotypic effects, before we quantified pyoverdine production (the primary siderophore produced by P. aeruginosa), and virulence using the Galleria mellonella infection model.

Results: Copper selected for increased pyoverdine production, which was positively correlated with virulence. This effect increased with time, such that populations incubated with high copper for twelve days were the most virulent. Replication of the experiment with a non-pyoverdine producing strain of P. aeruginosa demonstrated that pyoverdine production was largely responsible for the change in virulence.

Conclusions: We here show a direct link between metal stress and bacterial virulence, highlighting another dimension of the detrimental effects of metal pollution on human health.

Keywords: Coincidental selection; Evolution of virulence; Metal detoxification; Opportunistic pathogen; Pyoverdine.

Fig. 1

Per capita pyoverdine production (log₁₀-transformed standardised fluorescence units per OD₆₀₀) by Pseudomonas aeruginosa populations after growth in different concentrations of copper for (A) two days + one day in the absence of copper, or (B) twelve days + one day in the absence of copper. Circles show individual replicates (n = 6), with colour indicating total pyoverdine production (relative fluorescence units) and size showing optical density at 600 nm. Asterisks indicate significant differences between groups (*** = 0.001, ** = 0.01, * = 0.05, NS = non-significant), with the left value comparing the control to the low copper treatment, the middle value comparing the control to the high copper treatment and the right value comparing the low and high copper treatments

Fig. 2

The proportion of Galleria mellonella dead 18 h after being injected with Pseudomonas aeruginosa grown in the presence or absence of copper (A) two days + one day in the absence of copper, or (B) twelve days + one day in the absence of copper. 20 G. mellonella were injected per replicate (circles show individual replicates; n = 6 per unique treatment combination). Asterisks indicate significant differences between groups (*** = 0.001, ** = 0.01, * = 0.05, NS = non-significant), with the left value comparing the control to the low copper treatment, the middle value comparing the control to the high copper treatment and the right value comparing the low and high copper treatments

Fig. 3

The density (log₁₀ CFU mL⁻¹) of Pseudomonas aeruginosa populations incubated with copper for (A) two days + one day in the absence of copper, or (B) twelve days + one day in the absence of copper. Asterisks indicate significant differences between groups (*** = 0.001, ** = 0.01, * = 0.05, NS = non-significant), with the left value comparing the control to the low copper treatment, the middle value comparing the control to the high copper treatment and the right value comparing the low and high copper treatments

Fig. 4

The relationship between per capita pyoverdine production (log₁₀-transformed relative fluorescence units per OD₆₀₀) and virulence of P. aeruginosa populations grown with copper for (A) two days + one day in the absence of copper, or (B) twelve days + one day in the absence of copper. Virulence was quantified using the Galleria mellonella infection model and expressed as the proportion of G. mellonella dead (out of 20) 18 h after injection. Individual replicates are represented by circles (n = six per treatment); red points containing a △ represent the control (no copper) treatments, blue points containing a + the low copper treatment, and black points containing a ✕ the high copper treatment. The line shows the best model fit, and the shaded area shows the 95% confidence interval

Fig. 5

The proportion of Galleria mellonella dead 18 h after being injected with an isogenic non-pyoverdine producing strain of P. aeruginosagrown in the presence or absence of copper for (A) two days (+ one day in the absence of copper) or (B) twelve days (+ one day in the absence of copper). 20 G. mellonella were injected per replicate. Circles show individual replicates (n = 6), with size showing optical density at 600 nm and colour indicating total pyoverdine production (standardised fluorescence units) to demonstrate the mutant did indeed not produce any pyoverdine. Asterisks indicate significant differences between groups (*** = 0.001, ** = 0.01, * = 0.05, NS = non-significant), with the left value comparing the control to the low copper treatment, the middle value comparing the control to the high copper treatment and the right value comparing the low and high copper treatments

Fig. 6

Schematic of the experimental design used to test whether copper selects for siderophore-mediated virulence. Microcosms (n = 6 per treatment) containing KB medium at a concentration of either 0.0, 0.1 or 1.0 g/L of copper sulphate (CuSO₄) were inoculated with Pseudomonas aeruginosa (either the pyoverdine producing or non-pyoverdine producing strain), incubated at 28 °C and transferred every two days into fresh media. On days two and twelve, cultures were transferred into copper free medium for 24 h before being homogenised, their per capita pyoverdine production quantified and frozen in glycerol at a final concentration of 25%. Virulence and density assays were performed using defrosted samples