Inhibition and Dispersal of Pseudomonas aeruginosa Biofilms by Combination Treatment with Escapin Intermediate Products and Hydrogen Peroxide

Abstract

Escapin is an l-amino acid oxidase that acts on lysine to produce hydrogen peroxide (H₂O₂), ammonia, and equilibrium mixtures of several organic acids collectively called escapin intermediate products (EIP). Previous work showed that the combination of synthetic EIP and H₂O₂ functions synergistically as an antimicrobial toward diverse planktonic bacteria. We initiated the present study to investigate how the combination of EIP and H₂O₂ affected bacterial biofilms, using Pseudomonas aeruginosa as a model. Specifically, we examined concentrations of EIP and H₂O₂ that inhibited biofilm formation or fostered disruption of established biofilms. High-throughput assays of biofilm formation using microtiter plates and crystal violet staining showed a significant effect from pairing EIP and H₂O₂, resulting in inhibition of biofilm formation relative to biofilm formation in untreated controls or with EIP or H₂O₂ alone. Similarly, flow cell analysis and confocal laser scanning microscopy revealed that the EIP and H₂O₂ combination reduced the biomass of established biofilms relative to that of the controls. Area layer analysis of biofilms posttreatment indicated that disruption of biomass occurs down to the substratum. Only nanomolar to micromolar concentrations of EIP and H₂O₂ were required to impact biofilm formation or disruption, and these concentrations are significantly lower than those causing bactericidal effects on planktonic bacteria. Micromolar concentrations of EIP and H₂O₂ combined enhanced P. aeruginosa swimming motility compared to the effect of either EIP or H₂O₂ alone. Collectively, our results suggest that the combination of EIP and H₂O₂ may affect biofilms by interfering with bacterial attachment and destabilizing the biofilm matrix.

INTRODUCTION

In their natural environments, microorganisms most frequently exist as biofilms, or communities of microorganisms attached to surfaces and encased in a self-produced extracellular matrix (1). The properties of this matrix afford these microorganisms protection from environmental challenges, including nutritional starvation and chemical treatments such as antibiotics. Biofilms have a well-documented impact in both industrial and clinical settings. In microbial infections, the protective and recalcitrant nature of the biofilm state leads to problems with treatment and clearance. Biofilms on medical devices such as catheters or implants can result in chronic infections that are resistant to therapeutic drugs (2, 3). Nosocomial infections, often associated with biofilm formation on medical devices or at wound sites, contribute to higher morbidity and mortality rates as well as to increased health care costs (3, 4). Industries such as wastewater treatment, food, and agriculture are heavily impacted by the adverse effects of biofilms as well (5, 6). Consequently, the search for effective antibiofilm strategies is an ongoing quest that looks to both natural and synthetic agents that are capable of preventing, disrupting, or eradicating biofilms, while reducing selective pressures that contribute to resistance.

An effective antimicrobial agent against planktonic microbes has been found in the ink of the marine gastropod mollusc Aplysia californica (sea hare) (7). The ink is the product of two simultaneously released glandular secretions; upon attack by predators, the sea hare releases both products into the mantle cavity where they are mixed before being ejected from the animal (8, 9). One of the bioactive ingredients in the secretion is escapin, an l-amino acid oxidase (10). Escapin and its major natural substrate, l-lysine, are secreted at nearly 2 mg/ml and 150 mM, respectively (11 – 13). A series of escapin-catalyzed and nonenzymatic chemical reactions yields an equilibrium mixture of a diverse set of molecules referred to as escapin intermediate products (EIP), which can be synthesized and are effective as naturally produced products (Fig. 1) (14, 15). Hydrogen peroxide (H₂O₂) and ammonium are also produced. The equilibrium among the components of EIP is dependent on pH, with the cyclic form, Δ¹-piperidine-2-carboxylic acid (compound 3), dominating at any pH.

FIG 1.

Summary of the chemistry of the reaction of escapin with l-lysine, including the effects of pH on the relative composition of the molecular species in the equilibrium mixture (reprinted from Ko et al. [14]). Compounds are identified by number as follows: compound 1, l-lysine; compound 2, α-keto-ε-aminocaproic acid; compound 3, Δ¹-piperidine-2-carboxylic acid; compound 4, Δ²-piperidine-2-carboxylic acid; compound 5, γ-aminovaleric acid; compound 6, γ-valerolactam; compound 7, 6-amino-2-hydroxy-hex-2-enoic acid; compound 8, 6-amino-2,2-dihydroxy-hexanoic acid; compound 9, 2-hydroxy-piperidine-2-carboxylic acid.

The combination of EIP and H₂O₂, annotated as EIP + H₂O₂ (and with various concentrations) throughout this work, is bactericidal against a wide range of planktonic microbes, including Gram-negative and Gram-positive bacteria, Saccharomyces cerevisiae, and fungi (10, 14). At low-millimolar concentrations, EIP + H₂O₂ produces rapid, powerful, and long-lasting bactericidal activity against planktonic cells, probably through condensation of DNA (14, 16). EIP + H₂O₂ is an especially effective agent against planktonic cultures of P. aeruginosa (14). Given the bactericidal effects of EIP + H₂O₂ against planktonic bacteria, and in particular Pseudomonas aeruginosa, we focused on the effectiveness of EIP + H₂O₂ against bacterial biofilms. P. aeruginosa is a well-known opportunistic pathogen whose biofilms cause chronic infections, morbidity, and mortality (17 – 19). Taking into account the effectiveness of EIP + H₂O₂ against this bacterium (14) and its clinical relevance as a formidable pathogen, the objective of this study was to determine the effectiveness of EIP + H₂O₂ in preventing the formation of P. aeruginosa biofilms and disrupting existing biofilms of this pathogen.

MATERIALS AND METHODS

Culture preparation.

Pseudomonas aeruginosa PAO1 was grown in Pseudomonas basal mineral (PBM) medium, supplemented with glucose (80 mM final concentration) (PBM-glucose medium) (20) at 37°C with shaking at 200 rpm for 16 to 18 h. Frozen stocks (10% glycerol at −80°C) were thawed, and 35 μl was added to 30 ml of PBM-glucose medium in a 50-ml flask. Overnight cultures were diluted with fresh PBM-glucose medium to obtain initial inoculum optical densities at 600 nm (OD₆₀₀s) of 0.01 or 0.10 for biofilm formation and dispersal assays, respectively.

Chemicals.

Escapin intermediate products (EIP) were synthesized as described in Kamio et al. (15) based on Lu and Lewin (21) using a nonenzymatic synthesis starting with pipecolinic acid ethyl ester. Δ¹-Piperidine-2-carboxylic acid (compound 3) is the major product, and Δ²-piperidine-2-carboxylic acid (compound 4) is the minor product of this synthesis although, in solution, compounds 3 and 4 form an equilibrium mixture with other compounds, as shown in Fig. 1. The preparation of EIP used in each treatment is derived from the synthetic preparation of Δ¹-piperidine-2-carboxylic acid and used as the initial molecule to generate the EIP equilibrium mixture. This synthesis allows for the independently controlled presentation of these two major components of escapin's products, EIP and H₂O₂. Freeze-dried EIP was stored at −80°C and dissolved in sterile deionized (DI) water as a 1 M stock and diluted at the time of experiment. Hydrogen peroxide (H₂O₂; 30%) was purchased from Fisher Scientific (catalog number H325-100). For experiments, treatment concentrations of EIP and H₂O₂ were prepared in Pseudomonas basal mineral (PBM) medium without glucose in order to prevent further growth during treatment periods.

A Live/Dead BacLight bacterial viability kit (L-7012) was purchased from Life Technologies (CA, USA). This kit includes two different nucleic acid stains: SYTO 9 and propidium iodide (PI). SYTO 9 is a green fluorescent dye that labels bacteria with either intact or damaged membranes. PI is a red fluorescent dye that can penetrate only bacteria with damaged membranes and that reduces SYTO 9 fluorescence when both dyes are present.

Assay of biofilm formation.

Biofilms were cultured in 96-well polystyrene microtiter plates (22 – 24) using PBM-glucose medium as the growth medium. Briefly, biofilms were grown for 5 h (attachment phase) in the presence of EIP + H₂O₂, EIP alone, H₂O₂ alone, and PBM-glucose medium alone as the control. After incubation and treatment periods, biofilms were rinsed with sterile, deionized (DI) water by consecutively submerging microtiter plates in three separate tubs for 10 s each, after which extra water was shaken out. The biofilms were stained with 125 μl of 0.02% crystal violet (125 μl of 0.3% crystal violet [Becton Dickinson, NJ] in 2 ml of DI water) for 15 min at room temperature with shaking at 200 rpm. After the staining step, 96-well plates were rinsed with DI water three times. Plates were allowed to air dry for a minimum of 1 h or overnight. For quantification, bound crystal violet was dissolved with 125 μl of 95% ethanol for 30 min at room temperature with shaking at 200 rpm. Absorbance of dissolved crystal violet was measured by spectrophotometer at 570 nm using 95% ethanol as the blank. Treatment conditions were normalized to those of the untreated controls, and biofilm inhibition, as determined by biomass accumulation, is expressed as biomass (percentage of the control value [% control]).

Assay of biofilm dispersal.

Biofilms were grown in flow cells as described previously (23, 25, 26) using PBM-glucose medium as the growth medium. Frozen stocks of strain PAO1 were inoculated (35 μl added to 30 ml of medium) into PBM-glucose medium in a 50-ml flask and incubated at 37°C with shaking at 200 rpm for 16 to 18 h. The overnight culture was diluted with fresh PBM-glucose medium to obtain an OD₆₀₀ of 0.10. Biofilms were grown in flow cells for 20 h in recirculation mode and were rinsed for 20 min in PBM medium without glucose (rinse buffer). Biofilms were then treated with EIP + H₂O₂, EIP alone, H₂O₂ alone, or PBM medium without glucose as a control for 30 min, followed by a 10-min rinse with PBM medium without glucose prior to staining. One milliliter of a staining solution from a Live/Dead BacLight bacterial viability kit (3.34 μM SYTO 9 and 20 μM propidium iodide) was used to stain the biofilm for 15 min, followed by a 5-min rinse with PBM medium without glucose. Microscopic imaging of each flow cell was performed using a Zeiss LSM 510 confocal laser scanning microscope (CLSM; Carl Zeiss, Thornwood, NY). A minimum of 10 image stacks with a 1-μm z-step were taken for each channel of the flow cell using a 40× oil immersion lens. The excitation/emission wavelengths of 480/500 nm and 490/635 nm were used for SYTO 9 and propidium iodide, respectively, using argon and helium-neon (HeNe) lasers. Quantitative analysis of image stacks was performed using the statistical package COMSTAT (27). Biovolume is quantified as biomass volume divided by substratum area (as the number of cubic micrometers/number of square micrometers). It provides an estimate of the biomass in the biofilm, and thus it is generally referred to as biomass (27). Area layer is the fraction of the area occupied by biomass (as a percentage) in each image of a stack (i.e., distance from the substratum [in micrometers]) (27). Image stacks are 1-μm slice images that are stacked by the CLSM program to generate a three-dimensional image of the biofilm. Area layer analysis determines what fraction of each 1-μm slice is occupied by biomass from the substratum to the apex of the biofilm.

Motility assays.

Motility assays were performed as described previously (28, 29) with some modification. Medium used for the assay was Difco Luria-Bertani broth (Miller's LB; 10 g/liter tryptone, 5 g/liter yeast extract, 10 g/liter sodium chloride) (catalog number DF0446-07-5; Fisher Scientific) containing 0.3% (wt/vol) Bacto agar (DF0140-15-4; Fisher Scientific) for swimming, 0.5% (wt/vol) Bacto agar for swarming, and 1% (wt/vol) for twitching plates. For initial screens of the effects of EIP on motility, 10 μl of either PBM medium without glucose (control) or various concentrations of EIP (50, 100, 200, 400, 800, and 1,000 μM) (treatment) was spotted at the center of each corresponding motility plate and allowed to dry for approximately 5 to 10 min prior to the inoculation of bacteria. In subsequent swimming motility experiments, EIP (50 μM or 100 μM), H₂O₂ (3 μM or 6 μM), or EIP + H₂O₂ (3 μM H₂O₂ + 50 μM EIP or 6 μM H₂O₂ + 100 μM EIP) was applied to each plate in the same manner. To account for the effect of the single 10-μl application of EIP and H₂O₂, in comparison to that of the constant flow presented in flow cell treatments, increased concentrations of each compound were tested, also taking into account diffusion of the compound through the agar matrix as well as the possibility of chemical instability. Swimming and swarming plates were gently inoculated at the agar surface at the center of each plate with bacteria from an overnight culture streaked on LB agar plates (1.5%, wt/vol) using sterile toothpicks. Twitching plates were inoculated by stabbing the toothpick through the agar at the center of each plate, making sure that contact was made with the bottom surface of the plate. Plates were sealed with Parafilm to prevent dehydration and incubated at 37°C for 24 h. The diameter (measured in millimeters) of the motility zone was measured at intervals of 2, 4, and 24 h and used to determine the area (in square millimeters) of the zone. Values for the treatments were normalized to the mean values for each replicate of the untreated controls, and motility zones are expressed as area (percentage of the control value).

Statistical analysis.

Prevention of biofilm formation was analyzed using two-way analysis of variance (ANOVA) (α = 0.05). Analysis of biofilm dispersal and undamaged/damaged cell ratios was performed using one-way ANOVA (α = 0.05). A repeated-measures ANOVA (α = 0.05) was used in analyzing area layer data. Motility experiments were analyzed using an independent-samples t test (α = 0.05) and independent-samples Kruskal-Wallis test (α = 0.05).

RESULTS

The EIP + H₂O₂ combination inhibits P. aeruginosa biofilm formation at micromolar concentrations.

Preliminary data collected to determine effective concentrations of EIP and H₂O₂ for biofilm prevention studies indicated that micromolar concentrations of both EIP and H₂O₂ were of particular interest. To examine these conditions further, P. aeruginosa was grown in microtiter plates for 5 h, simulating the attachment phase of the biofilm life cycle, in the presence of various concentrations of H₂O₂ alone, EIP alone, or EIP + H₂O₂ (Fig. 2). H₂O₂ alone resulted in reduced biofilm formation, particularly at the concentrations of 48 μM and 96 μM, which resulted in approximately 44% and 30% reductions in biomass, respectively, relative to levels in the untreated controls. EIP alone, at either 3 μM or 30 μM, resulted in 25% and 17% less biofilm formation, respectively, than that in untreated controls. EIP + H₂O₂, at H₂O₂ concentrations of ≥24 μM, resulted in up to 47% less biofilm formation than that in untreated controls. The greatest effect on biofilm formation was observed when EIP was paired with 96 μM H₂O₂, resulting in more than 65% less biofilm formation than that in untreated controls. Two-way ANOVA indicated a significant treatment effect (H₂O₂, 3 μM EIP + H₂O₂, and 30 μM EIP + H₂O₂) and a significant concentration effect (H₂O₂ at 0 to 96 μM) but a nonsignificant treatment-concentration interaction. Post hoc analysis of the treatment effect indicated that combination treatment of EIP (3 μM or 30 μM) plus H₂O₂ resulted in significantly less biofilm formation than single treatments. Post hoc analysis of the H₂O₂ concentration effect showed that higher concentrations resulted in significantly less biofilm formation than lower concentrations. Thus, while EIP and H₂O₂ alone resulted in only 20 to 30% less biofilm formation than that in the untreated controls, EIP (3 μM or 30 μM) plus H₂O₂ (96 μM) resulted in nearly 70% less biofilm formation than that in the control. The effects of EIP and H₂O₂ were assessed after 12 h to determine if inhibition of biofilm formation was maintained (data not shown). However, the inhibitory effects diminished over this period, most likely due to a reduction in the chemical stability of the EIP.

FIG 2.

Effects of EIP and H₂O₂ on P. aeruginosa biofilm formation. P. aeruginosa biofilms were grown for 5 h in the presence of the following: H₂O₂ alone (black bars), 3 μM EIP alone or in combination with H₂O₂ (gray bars), or 30 μM EIP alone or in combination with H₂O₂ (white bars). The negative control (untreated) was PBM-glucose medium. Prevention of biofilm formation was determined by 96-well microtiter plate crystal violet assay. The values for each treatment including the control (PBM-glucose medium) are means ± standard errors of the means for three replicates for each experimental condition. The total number of measurements for each treatment ranged from 23 to 48. Two-way ANOVA indicated a significant effect for the treatment factor [F_{(2, 473)} = 18.57; P < 0.05]; post hoc tests show that the effects of treatment with H₂O₂ alone are significantly different from those of H₂O₂ + 3 μM EIP and of H₂O₂ + 30 μM EIP (P < 0.05). Additionally, a significant effect was determined for the H₂O₂ concentration factor [F_{(6, 473)} = 11.43; P < 0.05]; post hoc tests show that the values are equal for 0 μM, 3 μM, and 6 μM (a), >12 μM and 24 μM (b), and >48 μM and 96 μM (c). The interaction between the treatment factor and the H₂O₂ concentration factor was not significant [F_{(12, 473)} = 0.91; P > 0.05].

EIP and H₂O₂ work synergistically to disperse P. aeruginosa biofilms.

To examine the dispersal effects of EIP + H₂O₂ on established biofilms, a range of concentrations of H₂O₂ plus one concentration of EIP (50 μM) was tested using biofilms cultivated in flow cells for 20 h. Preliminary experiments (data not shown) indicated various EIP concentrations (above and below our treatment condition) that resulted in biofilm disruption; a concentration of 50 μM resulted in more pronounced disruption when paired with H₂O₂ and thus was selected as the treatment concentration. Representative CLSM images of 20-h-old biofilms treated with 3 μM H₂O₂ alone, 3 μM H₂O₂ + 50 μM EIP, and 50 μM EIP alone show the disruptive effects of the combined treatment versus H₂O₂ or EIP alone (Fig. 3A). The combined treatment resulted in greater biomass clearance (black, indicating no cells) and less stained biomass (yellow) than that in the control and with other treatments (Fig. 3A). One-way ANOVA showed that combination treatments, including the combinations of 50 μM EIP plus either 0.03 μM H₂O₂ or 3 μM H₂O₂, but not the respective single treatments, significantly reduced biofilm biomass by 42% and 37%, respectively, relative to control levels (Fig. 3B). The effects of treatments with 30 μM and 300 μM H₂O₂ alone were not significantly different from those of their corresponding combined treatments with 50 μM EIP, suggesting a small window of concentration ranges in which synergistic effects take place.

FIG 3.

Effects of EIP on P. aeruginosa and biofilm cell viability and biomass. (A) Representative confocal microscopy images of 20-h P. aeruginosa biofilms following treatment with 3 μM H₂O₂ alone, 50 μM EIP alone, 3 μM H₂O₂ + 50 μM EIP, and the control (PBM medium without glucose). Shown is cell viability labeling using Live/Dead BacLight nucleic acid stain where green labeling represents live and undamaged cells, red labeling represents cells that are dead or have damaged membranes, yellow represents areas where green and red labeling are colocalized in the biofilm, and black labeling represents area without cells. The bottom panel shows representative three-dimensional projections of the representative confocal images. Scale bar, 50 μm. (B) Effects of EIP + H₂O₂ against P. aeruginosa biofilm (i.e., biofilm disruption). Flow cell-cultivated P. aeruginosa biofilms (20-h) were analyzed posttreatment by CLSM. The image analysis software package COMSTAT was used for biomass determination, and values for all treatments were normalized to the levels of the untreated controls. Open diamond, untreated control; open square, 50 μM EIP alone; open circle, H₂O₂ alone; filled circle, EIP + H₂O₂. Values are means ± standard errors of the means for three replicates for each experimental condition. A range of 5 to 10 image stacks was taken for each biofilm; the total number of measurements for each treatment ranged from 4 to 172. ANOVA showed that the seven treatments significantly differ in their effects on biofilm biomass [F_{(12, 472)} = 8.21; P < 0.05], and post hoc tests show that the effects of EIP + H₂O₂ but not of EIP or H₂O₂ are significantly different from those of the control (P < 0.01). Asterisks indicate that the values for EIP + H₂O₂ at concentrations of 0.03 μM or greater and of H₂O₂ alone at 30 μM and 300 μM are significantly lower than those of the untreated control and of EIP + H₂O₂ at concentrations of ≤0.003 μM.

Treatment with EIP or EIP + H₂O₂ disperses but does not increase membrane damage within P. aeruginosa biofilms.

The ability of EIP + H₂O₂ to cause membrane damage and impact viability of biofilm cells was assessed by measuring the ratio of green- to red-stained cells in biofilm images collected by CLSM. One-way ANOVA indicated that treatment with 3 μM H₂O₂ + 50 μM EIP or with 50 μM EIP alone significantly reduced biofilm biomass compared to levels with untreated controls and treatments with 3 μM H₂O₂ alone (Fig. 4A). Measurements of green- and red-stained biomass from these treatments were used to determine a ratio of undamaged to damaged cells to determine the impact of treatment on membrane integrity (Fig. 4B). An undamaged/damaged ratio of less than 1 is indicative of a greater presence of damaged cells. The undamaged/damaged ratios for all treatments, including with the untreated control, were all greater than 2, suggesting that the effects of the treatments did not result in increased membrane damage to the biofilm cells. The undamaged/damaged ratios for 3 μM H₂O₂ + 50 μM EIP and for 3 μM H₂O₂ were not significantly different from those of the untreated control, and the undamaged/damaged ratio for 50 μM EIP was significantly greater than the ratios of all other treatments. Taken together with our other results, these experiments support the idea that EIP and H₂O₂ are dispersive but not through a mechanism of membrane damage.

FIG 4.

Effects of EIP and H₂O₂ on P. aeruginosa biofilm disruption. (A) Flow cell-cultivated biofilms were analyzed posttreatment by CLSM. The image analysis software package COMSTAT was used for biomass determination, and values for all treatments were normalized to the levels of the untreated controls. Values are means ± standard errors of the means for three replicates. Ten image stacks were taken for each biofilm; the total number of measurements for each treatment ranged from 30 to 109. ANOVA showed that two treatments significantly differ in their effects on biofilm biomass [F_{(3, 205)} = 10.24; P < 0.05]; post hoc tests show that the effects of 3 μM H₂O₂ + 50 μM EIP and of 50 μM EIP but not of 3 μM H₂O₂ are significantly different from the effect of the control (P < 0.05). Asterisks indicate that the values for 3 μM H₂O₂ + 50 μM EIP and for 50 μM EIP are significantly lower than the values for the untreated controls and for 3 μM H₂O₂. (B) Undamaged/damaged cell ratios were derived by dividing green-stained biomass measurements by red-stained biomass measurements. Treatments evaluated were 3 μM H₂O₂, 50 μM EIP, and 3 μM H₂O₂ + 50 μM EIP and the untreated controls. ANOVA showed that the undamaged/damaged cell ratios significantly differ across the treatments [F_{(3, 233)} = 2,951.10, P < 0.05]; post hoc tests show that the undamaged/damaged cell ratio for 50 μM EIP was significantly different from all the ratios for other treatments (P < 0.05). (C) Area layer was determined by COMSTAT analysis and is a measurement of the fraction of the area occupied by biomass (percent) in each image of a stack (i.e., distance from the substratum [in micrometers]). The differences in the mean area layer of biofilms in each treatment group relative to values of biofilms of the untreated controls were used to determine how the biofilm structure (from substratum to apex) was affected by our treatments. A repeated-measures ANOVA showed a significant effect on area layer by treatment condition; post hoc tests showed that the effects of treatments of 3 μM H₂O₂ + 50 μM EIP and of 50 μM EIP were significantly different from those of treatment with 3 μM H₂O₂ alone.

EIP and EIP + H₂O₂ disrupt the biofilm structure from substratum to apex.

Comparison of biofilm biomass after treatment with H₂O₂, EIP, and EIP + H₂O₂ revealed that the combination treatment of these compounds resulted in significant dispersal of P. aeruginosa biofilms relative to that of untreated controls (Fig. 4A). To determine if EIP + H₂O₂ affects the biomass distribution within the biofilm, the area layer function of COMSTAT was used to analyze CLSM-derived image stacks. Area layer measures the fraction of the area occupied by biomass (percent) in each image of a stack as a function of distance from the substratum. By calculating the differences in the mean area layers of biofilms in each treatment group relative to those of biofilms of untreated controls, we determined how the biomass distribution within the biofilm was affected by the treatments (Fig. 4C). Treatment with 3 μM H₂O₂ + 50 μM EIP or with 50 μM EIP led to a significant decrease in biomass from substratum to apex relative to that of the untreated controls. On the other hand, 3 μM H₂O₂ alone caused biomass accumulation near the substratum relative to the level in the untreated controls.

EIP and EIP + H₂O₂ enhance P. aeruginosa swimming motility.

To determine if biofilm dispersal, in the absence of increased membrane damage, was mediated through a motility mechanism, a series of agar plate-based motility assays was performed to test the effects of EIP on swimming, swarming, and twitching motility. Initially, a concentration of 1 mM EIP was tested in order to account for the diffusion of the compounds through the agar matrix as well as compound stability over the duration of the assay. These preliminary experiments showed that exposure to 1 mM EIP did not enhance or inhibit swarming or twitching motility; however, swimming motility was significantly enhanced relative to that in untreated controls (data not shown). Subsequently, a series of EIP concentrations below 1 mM was tested (50, 100, 200, 400, and 800 μM) to identify the range of effective treatment concentrations. Motility was monitored over a period of 2 and 4 h to determine the duration of any effects. Exposure to 50 μM EIP did not result in a significant increase in swimming motility relative to that in untreated controls (Fig. 5). However, only a 2-fold increase in concentration (100 μM EIP) was required to significantly enhance swimming motility relative to that of untreated controls. This effect was observed over the course of 4 h (Fig. 5). To determine the combined effects of EIP + H₂O₂ on motility, concentrations used in biofilm dispersal assays (3 μM H₂O₂ and 50 μM EIP) were tested either alone or in combination, but they did not result in any enhancement in swimming motility over the course of 4 h (data not shown). However, by increasing the concentration of H₂O₂ 2-fold (6 μM), its combined effect with 100 μM EIP enhanced P. aeruginosa swimming motility significantly (∼80% after 2 h; ∼40% after 4 h) compared to that with each treatment alone and in the untreated controls (Fig. 6).

FIG 5.

Effects of EIP on motility at 2 and 4 h. (A) Effects of 50, 100, 200, 400, and 800 μM EIP on P. aeruginosa swimming motility after 2 h at 37°C. Swimming motility for each treatment was quantified by measuring the diameter (in millimeters) of each motility zone and calculating the area (in square millimeters) of each zone after incubation. Treatments are indicated as follows: black bar, untreated control (PBM medium without glucose); checkered bar, 50 μM EIP; dark gray bar, 100 μM EIP; light gray bar, 200 μM EIP; white bar, 400 μM EIP; striped bar, 800 μM EIP. Values are means ± standard errors of the means for three replicates for each experimental condition. Values for the treatments were normalized to the levels of the untreated controls after the mean of the values for each of the control replicates was determined. An independent-samples Kruskal-Wallis test indicated a significant effect of treatment on swimming motility [χ²(5) = 40.118; P < 0.05] at 2 h. The asterisk indicates that the mean rank values for treatments of ≥100 μM EIP were significantly different from the mean rank values for the untreated control and for treatment with 50 μM EIP (P < 0.05). (B) The experiment is the same as that described for panel A, except that the incubation time is 4 h instead of 2 h. A Kruskal-Wallis test indicated a significant effect of treatment on swimming motility [χ² (5) = 40.399; P < 0.05]. The asterisk indicates that the mean rank values for treatments of ≥100 μM EIP were significantly different from the mean rank values for the untreated control and for treatment with 50 μM EIP (P < 0.05).

FIG 6.

Effects of EIP + H₂O₂ on motility at 2 and 4 h. (A) Effects of 6 μM H₂O₂ + 100 μM EIP on P. aeruginosa swimming motility at 2 h at 37°C. Swimming motility for each treatment was quantified by measuring the diameter (in millimeters) of each motility zone and calculating the area (in square millimeters) of each zone after incubation. Treatments are indicated as follows: black bar, untreated control (PBM medium without glucose); dark gray bar, 100 μM EIP; light gray bar, 6 μM H₂O₂ + 100 μM EIP; white bar, 6 μM H₂O₂. Values are means ± standard errors of the means for two replicates for each experimental condition. Values for treatments were normalized to levels in the untreated controls after the mean of the values for each of the control replicates was determined. An independent-samples Kruskal-Wallis test indicated a significant effect of treatment on swimming motility [χ² (3) = 30.251; P < 0.05] at 2 h. The asterisk indicates that the mean rank value for the treatment with 6 μM H₂O₂ + 100 μM EIP was significantly different from the mean rank values for each compound alone and for the untreated control (P < 0.05). (B) The experiment is the same as that described for panel A, except that the incubation time is 4 h instead of 2 h. A Kruskal-Wallis test indicated a significant effect of treatment on swimming motility [χ² (3) = 14.530; P < 0.05] at 4 h. The asterisk indicates that the mean rank value for the treatment with 6 μM H₂O₂ + 100 μM EIP was significantly different from the mean rank values for each compound alone and for the untreated control (P < 0.05).

DISCUSSION

Our results show that EIP + H₂O₂ acts in combination against P. aeruginosa biofilms at micromolar concentrations in two ways: prevention of biofilm formation and disruption of established biofilms. Preventing biofilm formation is an important antibiofilm strategy, and it encompasses the use of compounds that modulate gene expression linked to virulence factors, cell-to-surface adhesion, and interference with exopolysaccharide production (30). However, in many cases, the specific mechanisms of agents that prevent biofilm formation have yet to be elucidated. The extent of biofilm inhibition caused by EIP + H₂O₂ is similar to the biofilm-inhibiting effect caused by the Bacillus subtilis S8-18-derived α-amylase, a type of hydrolase that prevented biofilm formation in P. aeruginosa and other pathogens (31). As is the case with α-amylases, EIP + H₂O₂ could play a direct role in inhibiting biofilm formation by interference with bacterial adhesion, which is a critical step in initial biofilm formation and has been shown to occur within the first several hours in P. aeruginosa (32). There is an ecological interpretation for the biofilm prevention activity of EIP + H₂O₂: escapin, the l-amino acid oxidase from which EIP is derived, is a paralog of aplysianin A, an l-amino acid oxidase used by the sea hare A. californica to prevent microbial biofouling of its egg capsules (12, 16, 33).

A notable finding of this work is the ability of H₂O₂ to inhibit P. aeruginosa biofilm formation at micromolar concentrations. This is of particular interest due to the fact that millimolar concentrations are commonly used to trigger sublethal effects of oxidative stress in P. aeruginosa (34). P. aeruginosa is adapted to detect and overcome oxidative stress, particularly at these low-millimolar concentrations (35, 36). Low-millimolar concentrations of H₂O₂ have been shown to actually enhance biofilm formation, most likely through a quorum-sensing mechanism (37). Transcriptomic analyses have shown that exposure to H₂O₂ results in an increase in mRNA levels of genes necessary to deal with oxidative stress as well as virulence factors (34). These adaptive capabilities are not unique to P. aeruginosa. Salmonella enterica serovar Typhimurium becomes resistant to H₂O₂ treatments as high as 10 mM after exposure to sublethal concentrations of H₂O₂ (60 μM) (38). Similar observations have also been reported in Escherichia coli (39). However, here we have identified concentrations of H₂O₂ that, when paired with EIP, inhibit biofilm formation at levels far below those commonly tested against P. aeruginosa.

Oxidizing agents such as H₂O₂ have well-documented antimicrobial effects through DNA damage and oxygen radical toxicity (40, 41). The antimicrobial effects are often more pronounced in planktonic cells as they are genotypically and phenotypically different from their biofilm counterparts and are generally more susceptible to treatments (36, 42). In fact, this same pattern of susceptibility was observed in our antimicrobial treatment in that EIP + H₂O₂ was more effective against planktonic cultures of P. aeruginosa (14, 43). In addition to inhibiting biofilm formation, the combination of EIP and H₂O₂ is effective against established P. aeruginosa biofilms at micromolar concentrations, which are at or below concentrations often used in published treatment assessments. For example, Stewart et al. (36) showed that a steady treatment of 50 mM H₂O₂ for 1 h had little effect on wild-type P. aeruginosa biofilms, a result linked to the combined effects of reduced penetration of the compound through the biofilm matrix and the protective role of catalase production in the biofilm. Similarly, Bjarnsholt et al. (44) treated established P. aeruginosa biofilms with 100 mM H₂O₂ and demonstrated a decrease in susceptibility, most likely due to a quorum-sensing mechanism. Although microbial biofilms are generally less susceptible to the effects of H₂O₂, specifically at concentrations in the low-millimolar range, our results suggest a treatment strategy in which H₂O₂ is effective at micromolar concentrations.

Disrupting established biofilms is a critical antibiofilm strategy in applied contexts. Several factors promote detachment of P. aeruginosa, including enzymatic disruption of the surrounding extracellular polymeric substance (EPS) matrix, oxygen radical-dependent killing of bacteria (45), prophage-mediated bacterial death that enhances dispersal of cells from the biofilm (18), or the release of amyloid fibers linking cells in the biofilm together, a process regulated by d-amino acids (46). Area layer analysis indicated that introduction of EIP, either alone or in combination with H₂O₂, significantly affected biofilm structure down to the substratum. The fact that treatment with H₂O₂ alone appeared to have no significant structural effect on the biofilm was not completely unexpected. In fact, H₂O₂-mediated cell lysis has been shown to contribute to extracellular DNA (eDNA) release in P. aeruginosa biofilms (47). This eDNA release, coupled with poor penetration of the H₂O₂ through the biofilm matrix or its inactivation by catalases, could account for the largely unchanged biofilm structure, particularly at the substratum. The introduction of EIP, on the other hand, either alone or in combination with H₂O₂, significantly affects biofilm structure, specifically down to the substratum. EIP may not be susceptible to the same inactivation mechanisms seen with H₂O₂, which would allow it to penetrate and disrupt the biofilm matrix more effectively. Since previous work with EIP in planktonic cultures suggested DNA condensation as a mechanism underlying its bactericidal properties (16), we initially hypothesized that EIP may be affecting the structural stability of the biofilm matrix by targeting the eDNA. This is of particular importance because eDNA is an important structural component to P. aeruginosa biofilms and has been viewed as a viable target for biofilm disruption using enzymes such as DNase (48). However, the possibility of EIP initiating biofilm dispersal through a motility-dependent mechanism was also considered. Bacterial motility such as swimming, swarming (flagellum-mediated), and twitching (type IV pilus-mediated) is essential in P. aeruginosa biofilm formation and maturation (49). In fact, the enhancement of P. aeruginosa motility, such as swimming, and its subsequent role in biofilm dispersal by compounds such as nitric oxide (50) and ginseng extracts (28) are well documented. Interestingly, we determined that exposure to 1 mM EIP significantly enhances swimming motility in P. aeruginosa compared to that in untreated controls. Follow-up experiments showed that concentrations as low as 100 μM EIP resulted in enhanced swimming motility relative to that in untreated controls, while pairing of 6 μM H₂O₂ + 100 μM EIP enhanced swimming motility to a greater degree than that of either compound alone. This effect was observed over a period of 4 h and provides further support to the effectiveness of these compounds. Altogether, the coupling of EIP with hydroxyl radicals generated by H₂O₂, which have also been shown to trigger DNA damage (40, 51), results in a significant dispersal effect to established P. aeruginosa biofilms.

There is additional significance in that the presence of endogenous H₂O₂ in the biofilm environment has been documented. Liu et al. (52) measured H₂O₂ concentrations in the range of 0.7 to 1.6 mM in Streptococcus gordonii biofilms and suggested that H₂O₂ concentrations can vary by species composition. Likewise, many oral streptococci produce H₂O₂ as a means of competitive advantage (53). The production of oxygen radicals, including H₂O₂ by polymorphonuclear leukocytes (PMNs) as a means of eradicating microbial infections, is yet another potential source of endogenous H₂O₂ that could be encountered within a biofilm environment (44). Thus, introduction of EIP alone could potentially enhance the inherent disruptive effects of H₂O₂ in these environments.

EIP + H₂O₂ is a potentially valuable therapeutic for antivirulence strategies because it negatively impacts biofilm development and promotes dispersal at sublethal concentrations. Antivirulence strategies are currently being pursued to overcome widespread microbial multidrug resistance (54). In general, these strategies aim to control microbial pathogenesis by targeting virulence mechanisms (e.g., cell adhesion, quorum sensing, biofilm formation, and toxin production) while minimizing the selective pressure on the microorganisms that often leads to resistance (54, 55). Further investigation into the potential application of these compounds in combination with existing treatment strategies is both warranted and the focus of future work.