Facultative control of matrix production optimizes competitive fitness in Pseudomonas aeruginosa PA14 biofilm models

Abstract

As biofilms grow, resident cells inevitably face the challenge of resource limitation. In the opportunistic pathogen Pseudomonas aeruginosa PA14, electron acceptor availability affects matrix production and, as a result, biofilm morphogenesis. The secreted matrix polysaccharide Pel is required for pellicle formation and for colony wrinkling, two activities that promote access to O2. We examined the exploitability and evolvability of Pel production at the air-liquid interface (during pellicle formation) and on solid surfaces (during colony formation). Although Pel contributes to the developmental response to electron acceptor limitation in both biofilm formation regimes, we found variation in the exploitability of its production and necessity for competitive fitness between the two systems. The wild type showed a competitive advantage against a non-Pel-producing mutant in pellicles but no advantage in colonies. Adaptation to the pellicle environment selected for mutants with a competitive advantage against the wild type in pellicles but also caused a severe disadvantage in colonies, even in wrinkled colony centers. Evolution in the colony center produced divergent phenotypes, while adaptation to the colony edge produced mutants with clear competitive advantages against the wild type in this O2-replete niche. In general, the structurally heterogeneous colony environment promoted more diversification than the more homogeneous pellicle. These results suggest that the role of Pel in community structure formation in response to electron acceptor limitation is unique to specific biofilm models and that the facultative control of Pel production is required for PA14 to maintain optimum benefit in different types of communities.

FIG 1

Electron acceptor availability, Pel production, and biofilm architecture are linked in P. aeruginosa PA14. (A) Reduced availability of O₂ and/or phenazines induced a red and wrinkled colony morphology, while colonies grown with elevated availability of these electron acceptors were smoother and exhibited less Congo red (CR) staining. Δphz, phenazine-deficient mutant; phz⁺, phenazine overproducer. (B) A Δpel mutant formed smooth and pale colonies (with less CR bound), while pel⁺, a constitutive Pel producer, formed highly wrinkled, dark red colonies. Colonies shown are 5 days old and were grown on colony morphology assay medium. Colonies were grown at 21% O₂ unless otherwise stated. (C) Pellicles formed in static monocultures. Overnight precultures were diluted to an optical density (OD) at 500 nm of 0.5 and then diluted 1:100 in LB medium. These new cultures were grown statically (in 18-mm by 150-mm glass tubes) for 2 days at 37°C. The pellicle and planktonic fractions have been highlighted. (D) CR binding by pellicles grown as described for panel C. CR binding was calculated as the difference between absorbance for reference solution containing the starting concentration of CR and the absorbance of the sample. (E) CFU counts of both the pellicle and planktonic fractions of WT, Δphz, Δpel, and Δphz Δpel strains. Fewer CFU were found in the planktonic fractions of Pel producers (WT and Δpel strains) than non-Pel producers (Δpel and Δphz Δpel strains) (P = 1.6 × 10⁻⁴). More total CFU were observed in cultures of Pel producers (error bars denote ± standard errors of the means [SEM]; n = 3; 2-tailed heteroscedastic student t test).

FIG 2

Pel production is advantageous in competition with non-Pel producers but only under electron acceptor limitation. (Top) Final percentages of each strain in shaken (A) and static (pellicle) (B) cocultures. All cocultures were inoculated with an initial ratio of 1:1 (n = 3; error bars denote standard deviations [SD]; P values are from two-tailed unpaired equal-variance t test). (C) Imaging of coculture pellicles reveals variations in fluorescence intensity consistent with relative strain success, as indicated by the CFU counts for static cultures reported in panel B. Pellicles were imaged using a Zeiss AXIO Zoom.V16 microscope fitted with an ApoTome.2 module (excitation, 488 nm; emission, 509 nm).

FIG 3

Δpel strain shows a competitive disadvantage against the WT in pellicles formed on static cultures regardless of the initial coculture mixing ratio. (A) WT and Δpel strains were mixed at different initial ratios and used to start static cultures incubated under the same conditions as those used for results shown in Fig. 2. The initial percentage of the Δpel strain is plotted against its final percentage in the pellicle. Each data point represents an individual coculture, with the YFP tag on either the WT or the Δpel strain. (B) Pellicles reach a generally consistent final population size regardless of the initial mixing ratio of WT and Δpel strains. The initial percentage of the Δpel strain is plotted against the final total CFU (WT plus Δpel strains) in the pellicle.

FIG 4

Ability to produce Pel does not provide a competitive advantage against Δpel strains in mixed-colony biofilms. Samples were obtained from the colony center or edge (A) after 5 days of growth on colony morphology assay medium and plated for CFU (B). n = 3. Error bars represent SD. P values are from two-tailed unpaired equal-variance t test.

FIG 5

Experimental evolution in pellicles selects for strains that show an advantage in competitions for pellicle residence and in shaken liquid cultures. (A) Cartoon showing the experimental evolution regimes used to generate the ePel strain and the shaken-culture control. Representative scans of eShak, WT, and ePel colonies after 48 h of growth are shown. The ePel strain then was evaluated for its competitive fitness against the WT in pellicles of static cultures (B) and shaken cultures (C). n = 6 for static cultures, n = 3 for shaken cultures. Error bars represent SD. P values were calculated using unpaired, 2-tailed t tests.

FIG 6

ePel strain exhibits a competitive disadvantage in colony biofilms, both in the colony center (A) and at the colony edge (B). Colonies were grown for 7 days before sampling. n = 3. Error bars represent SD.

FIG 7

Experimental evolution in the O₂-limited (center) and O₂-replete (edge) regions of colony biofilms gives rise to mutants with diverse Pel production phenotypes. (A) O₂ profiles taken in the Pel-producing and non-Pel-producing regions (as indicated in the photo, inset) of a WT colony on day 6 (n = 5; error bars denote SD). O₂ was measured with a Clark-type microelectrode inserted at the top of the colony, taking readings every 5 μm as the electrode tip moved toward the agar medium. (B) Cartoon showing the experimental evolution regimes used to generate the eCol^whole, eCol^center, and eCol^edge lineages. (C) Representative scans of colony biofilms grown from the eCol^whole, eCol^center, and eCol^edge lineages. For each passaging regime, scans are shown for 4 replicates after the 1st, 2nd, and 3rd passages. Scans were taken after 3 days of growth. (D) Pel production, quantified using the CR binding assay, for colonies formed by WT, Δpel, and eCol^edge strains. n = 4, and error bars represent SD. P values were calculated using unpaired, 2-tailed t tests.

FIG 8

eCol^edge lineages exclude the ancestral strain from the colony edge. Colonies shown at the left were grown for 7 days on colony morphology assay medium. Sampling was performed as diagrammed in Fig. 4A, and the percentage of each strain represented by CFU is shown at the right. Experiments performed with the eCol^edge-1 and eCol^edge-2 lineages yielded similar results; results for eCol^edge-1 are shown. n = 3. Error bars represent SD. The scale bar is 1 cm.