Pseudomonas aeruginosa Mobbing-Like Behavior against Acanthamoeba castellanii Bacterivore and Its Rapid Control by Quorum Sensing and Environmental Cues

Abstract

Mobbing, group attack of prey on predator, is a behavior seen in many animal species in which prey animals use numbers and coordination to counter individually superior predators. We studied attack behavior of Pseudomonas aeruginosa toward the bacterivore Acanthamoeba castellanii. This behavior consists of directed motility toward and specific adhesion to the predator cells, enacted in seconds and responding to both prey and predator population densities. Attack coordination relies on remote sensing of the predator and the use of the Pseudomonas quinolone signal (PQS), a P. aeruginosa species-specific quorum sensing molecule. Mutants unable to produce the PQS show unspecific adhesion and reduced survival, and a corresponding increase in predator population occurs as a result of predation. The addition of an external PQS restored some predator-specific adherence within seconds, suggesting a novel response mechanism to this quorum sensing (QS) signal. Fast behavioral response of P. aeruginosa to PQS is also supported by the rate of signal accumulation in the culture, reaching relevant concentrations within minutes, enabling bacteria response to self population density in these short timescales. These results portray a well-regulated group attack of the bacteria against their predator, reacting within seconds to environmental cues and species-specific signaling, which is analogous in many ways to animal mobbing behavior. IMPORTANCE Pseudomonas aeruginosa was shown previously to attack amoebae and other predators by adhering to them and injecting them with virulent substances. In this work, we show that an active, coordinated group behavior is enacted by the bacteria to utilize these molecular components, responding to both predator and bacterial population density. In addition to their ecological significance, immediate behavioral changes observed in response to PQS suggest the existence of a fast QS signal cascade, which is different from canonical QS that relies on slow-to-respond gene regulation. Similar regulatory circuits may drive other bacterial adaptations and pathogenicity mechanisms and may have important clinical implications.

Keywords: PQS; adhesion; biofilm; predation avoidance; quorum sensing.

FIG 1

Time-lapse microscopy of adhesion of Pseudomonas aeruginosa to Acanthamoeba castellanii. (a to f) Time-lapse microscopy in green fluorescence protein (GFP) fluorescence wavelengths of bacterial adhesion to amoeba. (g) Bright-field image at 10-minute time point showing the position of amoeba cells within the field. (h) Superimposed image of 10-minute GFP fluorescence and bright-field to demonstrate the colocalization.

FIG 2

Sensation and attraction of Pseudomonas aeruginosa to amoebae. (a) Taxis measurement; the Fluoroblok system consists of a 24-well plate and a 3-μm pore size fluorescence blocking filter insert dividing each well into a top and a bottom chamber. Amoebae are added to the bottom chamber of a well and allowed to settle. GFP-tagged bacteria are added to the top chamber above the filter. Reading fluorescence from the bottom allows measurement of bacterial migration kinetics. (b) Migration kinetics of GFP-tagged PAO1 in the presence or absence of amoebae; n = 9 per treatment, circles stand for measurement times, flanking curves represent 1 SD.

FIG 3

Bacterial adhesion kinetics at different amoeba population densities. (a) Conceptualization of adhesion kinetic measurement. In the absence of dye (left), light in GFP excitation and emission wavelengths travels freely through the microtiter well. The bottom fluorescence measurement under these conditions detects both attached and unattached bacteria. When the dye is added (right) light can penetrate only a few microns into the well, eliminating signal from bacteria in the bulk liquid, enabling a specific measurement of adhering bacteria (36). (b) Bacterial adhesion kinetics in different amoeba population densities; n = 7 per treatment, dots represent measurement, flanking curves represent ±1 SD. (c) Average adhesion rates at first 5 minutes. Different letters stand for significant differences at a P value of <0.05 according to an analysis of variance (ANOVA) test. (d) Adhesion rate per amoebae at first 5 minutes. *, Note that results for 0,4,8,16,32 and 64 amoeba per well (Panel c) were not significantly different (Tukey’s HSD post-hoc test).

FIG 4

Effect of nutrient masking on Pseudomonas aeruginosa adhesion to amoebae. All wells contain LB Lennox medium, with or without 1,000 amoeba per well. n = 21 per treatment, error lines stand for ±1 SD.

FIG 5

Effect of PQS on Pseudomonas aeruginosa adhesion to amoebae. (a) Adhesion kinetics of ΔpqsA in the presence of amoebae in different PQS concentrations. (b) WT and ΔpqsA adhesion at 60 minutes in the absence or presence of amoebae (1,000 per well), and with or without addition of 160 nM PQS. (c) Bacterial adhesion at 60 minutes in different PQS concentrations in the presence (dark gray) and the absence (light gray) of amoebae. In all cases, n = 7 per treatment and error bars stand for ±1 SD. Different letters denote statistically significant differences.

FIG 6

PQS signal accumulation in WT and ΔpqsA cultures. PAO1 WT and ΔpqsA overnight cultures were washed once in M9 buffer and resuspended and incubated in buffer for a period of 5 minutes. PQS was quantified in spent media and in buffer against a PQS standard using a bio-reporter (Pseudomonas putida expressing β-galactosidase under control pqsR promoter). n = 6 per treatment; error bars represent ±1 SD; all groups are statistically different from each other, P < 0.05.

FIG 7

Population density kinetics of Pseudomonas aeruginosa WT (black) and ΔpqsA (gray) in the presence (closed symbols) and absence (open symbols) of amoebae; n = 14 per treatment, symbols signify measurement times, curves stand for ±1 SD. Different letters stand for statistically different groups at the last time point (P < 0.05).

FIG 8

Effect of prey type on amoeba survival or growth. Amoebae were cocultured in microscopy counting chambers with various bacterial prey types. Bars show percent change in amoeba population density relative to t₀; n = 5 per treatment, error bars stand for ± 1 SD. Different letters stand for significant differences at a P value of <0.05 according to an ANOVA test. (a) Pseudomonas aeruginosa WT and ΔpqsA; t₀ amoeba population = 35 ± 5 (avg ± SD) per counting field; starting bacterial OD₆₀₀ of 0.5 (1 cm) in all cases. (b) Escherichia coli DH5α; OD is calculated as OD₆₀₀ (1 cm) after dilution into the counting chamber; t₀ amoeba population = 35 ± 5 (avg. ± SD). (c) Pseudomonas aeruginosa WT, ΔpqsA, and ΔflgF—36-h time point; t₀ amoeba population of 151 ± 19; starting bacterial OD₆₀₀ = 0.5 (1 cm).