2,3-Dihydroxybenzoic Acid-Containing Nanofiber Wound Dressings Inhibit Biofilm Formation by Pseudomonas aeruginosa

Abstract

Pseudomonas aeruginosa forms biofilms in wounds, which often leads to chronic infections that are difficult to treat with antibiotics. Free iron enhances biofilm formation, delays wound healing, and may even be responsible for persistent inflammation, increased connective tissue destruction, and lipid peroxidation. Exposure of P. aeruginosa Xen 5 to the iron chelator 2,3-dihydroxybenzoic acid (DHBA), electrospun into a nanofiber blend of poly(d,l-lactide) (PDLLA) and poly(ethylene oxide) (PEO), referred to as DF, for 8 h decreased biofilm formation by approximately 75%. This was shown by a drastic decline in cell numbers, from 7.1 log₁₀ CFU/ml to 4.8 log₁₀ CFU/ml when biofilms were exposed to DF in the presence of 2.0 mM FeCl₃ 6H₂O. A similar decline in cell numbers was recorded in the presence of 3.0 mM FeCl₃ 6H₂O and DF. The cells were more mobile in the presence of DHBA, supporting the observation of less biofilm formation at lower iron concentrations. DHBA at MIC levels (1.5 mg/ml) inhibited the growth of strain Xen 5 for at least 24 h. Our findings indicate that DHBA electrospun into nanofibers inhibits cell growth for at least 4 h, which is equivalent to the time required for all DHBA to diffuse from DF. This is the first indication that DF can be developed into a wound dressing to treat topical infections caused by P. aeruginosa.

INTRODUCTION

Pseudomonas aeruginosa is a versatile opportunistic pathogen commonly associated with a number of nosocomial infections of the respiratory tract, urinary tract, and wounds (1, 2, 3). The antibiotic-induced killing of P. aeruginosa is extremely poor in chronic wounds such as leg ulcers or burn wounds (4). The species protects itself from antibiotics by forming biofilms (5–7) and by doing so develops resistance to antibiotics (6, 8, 9).

Cystic fibrosis is one of the best-studied examples of chronic infections caused by biofilms of P. aeruginosa (10). Zhao et al. (11) have shown that biofilm formation by P. aeruginosa is stimulated by Psl exopolysaccharide on the cell surface. However, Singh et al. (12) and Banin et al. (13) attributed biofilm formation by Pseudomonas spp. to higher levels of free iron in the environment. Free iron may retard wound healing and may even be responsible for persistent inflammation, increased connective tissue destruction, and lipid peroxidation (14). These pose unique challenges in the development of antimicrobial drugs effective against P. aeruginosa.

Lactoferrin, an iron chelator in mammals, inhibited biofilm formation (12, 13, 15). Similar results were obtained with synthetic iron chelators such as 2,2′-dipyridyl (2DP), diethylenetriaminepentaacetic acid (DTPA), EDTA, deferoxamine mesylate (DM), and ethylenediamine-N,N′-diacetic acid (EDDA) (16). As far as we could determine, the effect of the iron chelator 2,3-dihydroxybenzoic acid (DHBA) on biofilm formation by Pseudomonas spp. has not been reported.

DHBA has been used orally as a noncytotoxic iron-chelating agent in patients suffering from β-thalassemia (17, 18). DHBA is produced during aspirin metabolism in humans and is present in many plants (19, 20). DHBA has higher binding constants for iron than transferrin and lactoferrin (21). Incorporation of DHBA in nanofiber wound dressings could thus control bacterial growth and biofilm formation in chronic wounds, especially in diabetic patients. In this study, we report on the effect that DHBA, electrospun into nanofibers, has on biofilms formed by P. aeruginosa.

MATERIALS AND METHODS

Bacterial strain and growth conditions.

P. aeruginosa Xen 5, a bioluminescent derivative of a clinical strain (ATCC 19660) isolated from human septicemia, was obtained from Caliper Life Sciences (Hopkinton, MA). Luria-Bertani (LB) broth, supplemented with 60 μg/ml tetracycline, was used to maintain the plasmid containing the lux operon.

Electrospinning of nanofibers.

Nanofibers were prepared by spinning DHBA (50 mg/ml) into a blend of poly(d,l-lactide) (PDLLA) and poly(ethylene oxide) (PEO) (24%; 50:50), as described by Heunis et al. (22). N,N-dimethylformamide (DMF) was used as the solvent for PEO:PDLLA. The relative humidity was controlled at 50 to 55% by keeping the temperature between 28 and 30°C. Control nanofibers were produced using the same method, but without DHBA. Nanofibers containing DHBA (DF) and control nanofibers (CF) were cut into sections of 0.5 cm². Nanofibers were coated with gold to increase conductivity and images were recorded by using a Leo 1430VP scanning electron microscope (Zeiss). The diameters of the nanofibers were determined using ImageJ Software (version 1.46; Scion Corporation).

Release of DHBA from nanofibers.

Nanofiber samples of 0.5 cm² (DF) were placed in 0.5 ml sterile Milli-Q water and incubated at 37°C. At specific time intervals, the nanofibers were transferred to 0.5 ml sterile Milli-Q water. Release of DHBA in the water was determined by using the universal chrome azurol S (CAS) assay (23). Five-hundred microliters of the DHBA suspension was mixed with 500 μl CAS reagent. Absorbance readings were recorded at 630 nm (A_s), with Milli-Q water as the blank. Milli-Q water with CAS reagent served as the reference (A_r). The percentage of chelator units was defined as [(A_r − A_s)/A_r] × 100. Chelator units recorded for nanofibers were compared against a standard curve compiled with known concentrations of DHBA.

Biofilm formation and changes in cell numbers of P. aeruginosa Xen 5 when exposed to nanofibers containing DHBA.

P. aeruginosa Xen 5 was grown overnight at 37°C in Luria-Bertani (LB) broth supplemented with 60 μg/ml tetracycline. The culture was diluted to 8.5 ± 0.02 log₁₀ CFU/ml in M63 biofilm assay medium [3 g KH₂PO₄, 7 g K₂HPO₄ and 2 g (NH₄)₂ SO₄ in 1 liter of water], as described by O'Toole (24). The medium was supplemented with tetracycline (60 μg/ml), magnesium sulfate (1 mM), and arginine (0.4%, wt/vol) to stimulate biofilm formation. Wells in a 48-well polystyrene flat-bottom multidish (BioLite MultiDishes, Thermo Scientific, NY) were each filled with 200 μl of the cell suspension, thus ∼7.8 log₁₀ CFU per well. Nine wells each received 0.5 cm² of DF, nine wells 0.5 cm² CF, and the rest no nanofibers. The multidish was incubated at 37°C without shaking.

The total numbers of P. aeruginosa Xen 5 cells, dead and viable, in a biofilm were determined by staining the wells with crystal violet, as described by O'Toole (24). In brief, 250 μl of 0.1% (vol/vol) crystal violet was added to the biofilms in the wells and incubated at 25°C for 10 to 15 min. After incubation, the wells were carefully rinsed three to four times with sterile distilled water, blotted dry by placing the multidish upside down on paper towels, and then left for 4 to 5 h to air dry. Acetic acid (250 μl; 30%, vol/vol) was added to each well and left at 25°C for 10 to 15 min to solubilize the crystal violet. The density of the crystal violet in solution was determined by recording optical density at 595 nm (OD₅₉₅) readings in a microplate reader (model 680; Bio-Rad). Acetic acid (30%, vol/vol) served as the blank.

The number of viable cells of P. aeruginosa Xen 5 in the biofilm was determined by removing planktonic cells and nanofibers from the wells, followed by gentle rinsing of the wells with three volumes of sterile distilled water. Special care was taken not to disturb the biofilms. Two-hundred microliters of phosphate-buffered saline (PBS) (pH 7.3) was added to each well and gently agitated with a glass rod. The cell suspension was then transferred to a sterile tube, vortexed for 1 min, and serially diluted with sterile PBS. One-hundred microliters of each dilution was spread plated onto LB agar and the plates incubated at 37°C for 24 h.

Effect of iron on biofilm development.

An overnight-grown culture of P. aeruginosa Xen 5 was diluted to 8.5 ± 0.02 log₁₀ CFU/ml in M63 biofilm assay medium supplemented with tetracycline, magnesium sulfate, and arginine, as described elsewhere. The cell suspension was divided into three equal parts, to which 1.0 mM, 2.0 mM, and 3.0 mM FeCl₃ 6H₂O were added, respectively. Wells of a 48-well polystyrene flat-bottom multidish were each filled with 200 μl of the cell suspension. Nine wells each received 0.5 cm² of DF, nine wells 0.5 cm² CF, and the rest no nanofibers. The multidish was incubated at 37°C without shaking. At specific time intervals, the cell suspensions and nanofibers were removed from the wells and biofilm formation was recorded as described above. Viable cell numbers were determined by plating onto LB agar, as described above.

In another experiment, the effects of DHBA, iron, and a combination of DHBA plus iron on the twitching mobility of strain Xen 5 were studied as described by Singh et al. (12), but with minor modifications. Overnight-grown cells of P. aeruginosa Xen 5 were stab inoculated into M63 agar (1%, wt/vol, agar), supplemented with DHBA (2.0 mM in 50%, vol/vol, DMF), FeCl₃ 6H₂O (1.0 mM in Milli-Q water), and a combination of 2.0 mM DHBA plus 1.0 mM FeCl₃ 6H₂O, respectively. Tetracycline (60 μg/ml) was added to the growth medium. Plates were incubated at 37°C for 3 to 4 days and examined for the presence of growth between the interface of the plate and the surface of the agar. The distance the cells migrated from the stab was measured in mm. The agar was carefully removed, the cells that adhered to the surface of the plates were stained with crystal violet (0.1%, vol/vol), and the migration patterns of the cells were studied under a light microscope.

Antimicrobial activity of DHBA.

The antimicrobial activity of DF compared to CF was evaluated by using a slide culture model. In brief, 100 μl M63 molten agar (1%, wt/vol, agar), containing 60 μg/ml tetracycline, was dispensed in a circle (15-mm diameter) on a sterile microscope glass slide. The agar was allowed to solidify and 10 μl of a P. aeruginosa Xen 5 cell suspension (8.9 ± 0.06 log₁₀ CFU/ml) was deposited in the center of the solidified medium. Samples of DF and CF (0.5 cm²) were placed on the surface of the agar containing the cells. The glass slides were placed in petri dishes and incubated at 37°C. Drying out of the slides was prevented by sealing the plates with Parafilm. At specific time intervals, the slides were observed for bioluminescence using the in vivo imaging system (IVIS 100) of Caliper Life Sciences. The number of photons per second per cm² per steradian (p s⁻¹ cm⁻² sr⁻¹) emitted from each reactive oxygen intermediate (ROI) was calculated using the Living Image software (version 3.0) of Caliper Life Sciences. The agar was scraped from the glass slides, suspended in 1 ml sterile PBS, vortexed for 1 min, and serially diluted in sterile PBS. One-hundred microliters of each dilution was spread plated onto LB agar and the plates were incubated at 37°C. Colonies were counted after 24 h of incubation.

In another experiment, the effect of DHBA on the growth of P. aeruginosa Xen 5 was tested in the presence of iron. An overnight-grown culture of P. aeruginosa Xen 5 was inoculated into 100 ml iron-deficient LB broth to obtain 6.4 ± 0.02 log₁₀ CFU/ml. The LB both was deferrated by the addition of 0.5% (wt/vol) 8-hydroxychinolin (8HQ). After 30 min of stirring at 25°C, the Fe-8HQ complex was extracted by adding 20% (vol/vol) chloroform to the broth, in three consecutive steps, and then autoclaved. The cell suspension was divided into three equal volumes, to which DHBA (116 μg/ml in 50%, vol/vol, DMF), FeCl₃ 6H₂O (1.0 mM in Milli-Q water), and a combination of DHBA (116 μg/ml) and FeCl₃ 6H₂O (1.0 mM) were added, respectively. The control culture contained no FeCl₃ 6H₂O, but received 0.1 ml 50% (vol/vol) DMF. Incubation was at 37°C for 8 h. Samples were taken after every 2 h, serially diluted, and plated onto LB agar, and then colonies were counted after 24 h of incubation at 37°C.

Determination of MIC.

DHBA was dissolved in 50% (vol/vol) DMF and added to LB broth to final concentrations ranging from 0.05 mg/ml to 5.0 mg/ml. Each of the test tubes was inoculated with P. aeruginosa Xen 5 to a final OD₅₉₅ of 0.01 and incubated at 37°C for 24 h. The MIC was defined as the lowest concentration of DHBA producing complete inhibition of visible growth.

In another experiment, P. aeruginosa Xen 5 was inoculated into LB broth, incubated at 37°C to an OD₅₉₅ of 0.01, and then treated by adding DHBA at the MIC level. Changes in optical density was recorded at 595 nm over 24 h. Cells grown in the presence of 50% (vol/vol) DMF served as controls.

Statistical analysis.

Statistical analysis was performed by one-way analysis of variance (ANOVA) and with the t test by using GraphPad Prism (version 6.03 [Trial] for Windows; GraphPad Software, Inc., CA). Differences were considered statistically significant at P values of <0.05.

RESULTS

Electrospun nanofibers.

Electrospinning of DHBA into a blend of PDLLA and PEO (24%; 50:50) produced nanofibers of 400 to 450 nm in diameter. Nanofibers spun without DHBA had a diameter of 450 to 500 nm (Fig. 1).

FIG 1.

Schematic representation of the electrospinning of DHBA into a blend of PDLLA and PEO to produce nanofibers of 400 to 500 nm in diameter.

Release of DHBA from nanofibers.

Almost all DHBA was released from DF within 2 h (18.8% CAS decolorization units), followed by a slow, linear release over the following 2 h to no detectable levels of DHBA (Fig. 2). Deduced from a standard curve with known concentrations of DHBA (not shown), 18.8% CAS corresponded to 116 ± 0.03 μg DHBA per 0.5 cm² nanofiber.

FIG 2.

Release of DHBA from DF. Data points show average values from three independent experiments (mean ± standard deviation).

Biofilm formation.

Optical density (OD₅₉₅) readings recorded for biofilms that formed after 8 h in the presence of DF were 1.2 and significantly (P < 0.001) lower than OD readings of 1.7 recorded for biofilms that formed in the presence of CF and 1.8 in the absence of nanofibers (Fig. 3a). The number of viable cells in an 8-h-old biofilm and in the presence of DF were 6.4 log₁₀ CFU/ml and significantly lower (P < 0.001) than the cell numbers of 7.4 log₁₀ CFU/ml in biofilms exposed to CF and 7.5 log₁₀ CFU/ml in biofilms not exposed to nanofibers (Fig. 3b).

FIG 3.

Biofilm formation of P. aeruginosa Xen 5, expressed as optical density (OD₅₉₅) of crystal violet diffused from stained cells (a, c, e, and g) and viable cell numbers, log₁₀ CFU/ml (b, d, f, and h) when cells were grown in the presence of nanofibers containing DHBA (DF) or nanofibers without DHBA (CF) and in the absence of nanofibers (control). The effects of different iron concentrations on biofilm formation are shown in panels c to h. Data points are the averages of three independent experiments (mean ± standard deviation). *, P < 0.001.

Effect of iron on biofilm formation.

After 8 h of incubation, the OD₅₉₅ readings recorded for crystal violet-stained cells from biofilms that were not exposed to nanofibers (control) and biofilms that were exposed to nanofibers, but without DHBA incorporated (CF), were similar at each of the iron concentrations (Fig. 3a, c, e, and g). Similar results were obtained when viable cell numbers were recorded (Fig. 3b, d, f, and h). However, OD₅₉₅ readings recorded for crystal violet-stained cells from biofilms that were exposed to nanofibers containing DHBA (DF) decreased from 1.17 in the absence of FeCl₃ 6H₂O (Fig. 3a) to 0.89 (Fig. 3c), 0.48 (Fig. 3e), and 0.46 (Fig. 3g) in the presence of 1.0, 2.0, and 3.0 mM FeCl₃ 6H₂O, respectively. A slight decrease in cell numbers (from 7.5 log₁₀ CFU/ml to 6.4 log₁₀ CFU/ml) was recorded when biofilms that formed in the presence of 1.0 mM FeCl₃ 6H₂O were exposed to DF (Fig. 3d). A drastic decline in cell numbers, from 7.1 log₁₀ CFU/ml to 4.8 log₁₀ CFU/ml (Fig. 3f), was recorded when biofilms that formed in the presence of 2.0 mM were exposed to DF. A similar decline in cell numbers, from 6.8 log₁₀ CFU/ml to 4.7 log₁₀ CFU/ml (Fig. 3h), was recorded when biofilms that formed in the presence of 3.0 mM FeCl₃ 6H₂O were exposed to DF.

Twitching mobility studies.

Cells migrated up to 10 mm from the point of inoculation when incubated in the presence of 2.0 mM DHBA and 2.0 mM DHBA plus 1.0 mM FeCl₃ 6H₂O, but not when incubated in the presence of 1.0 mM FeCl₃ 6H₂O (Fig. 4A). Microscopic examination of stained cells clearly showed that cells in the presence of DHBA were more motile than cells in the presence of iron (Fig. 4B).

FIG 4.

(A) Mobility of P. aeruginosa Xen 5 recorded in the interface between the plate and the agar surface after 3 to 4 days of incubation at 37°C. The top three images (Aa through Ac) were taken after the agar was removed from the plates, and cells that adhered to the plates were stained with crystal violet. (Aa) Growth in the presence of 2.0 mM DHBA; (Ab) growth in the presence of 2.0 mM DHBA plus 1.0 mM FeCl₃ 6H₂O; (Ac) growth in the presence of 1.0 mM FeCl₃ 6H₂O. Data points presented are the average values of three independent experiments (mean ± standard deviation). (B) Microscopic images of P. aeruginosa Xen 5 in the presence of DHBA (a) and iron (b), as observed at ×10 magnification. *, P < 0.01. Arrows indicate direction of cell mobility.

Antimicrobial activity of DHBA.

After 8 h of incubation at 37°C, cell numbers in the presence of CF increased from 8.7 to 9.2 log₁₀ CFU/ml and bioluminescence from 1.50E + 07 to 2.97E + 07 p s⁻¹ cm⁻² sr⁻¹ (Fig. 5a). However, cell numbers in the presence of DF remained between 6.5 log₁₀ CFU/ml and 6.7 log₁₀ CFU/ml and bioluminescent readings at approximately 3.0E + 07 p s⁻¹ cm⁻² sr⁻¹ throughout the 8-h experiment (Fig. 5a).

FIG 5.

(a) Effects of nanofibers containing DHBA (DF) and nanofibers without DHBA (CF) on the growth of P. aeruginosa Xen 5. Cell numbers (CFU/ml) are shown in bar graphs and bioluminescence in line graphs. (b) Growth in the presence of sub-MICs of DHBA (116 μg/ml) and DHBA (116 μg/ml) in the presence of 1 mM iron, and in the absence of DHBA, with and without 1 mM iron. (c) Growth of strain Xen 5 in the presence of 1.5 mg/ml DHBA (MIC levels) and in the absence of DHBA. Data points are the average values of three independent experiments (mean ± standard deviation). *, P < 0.001.

The lowest cell numbers (6.3 log₁₀ CFU/ml) were recorded in the absence of iron and in deferrated LB broth supplemented with 116 μg/ml DHBA after 8 h of incubation (Fig. 5b). In the presence of 1.0 mM FeCl₃ 6H₂O, cell numbers increased to 7.0 log₁₀ CFU/ml after 2 h but declined to 6.5 CFU/ml during the following 6 h (Fig. 5b). Cells exposed to the same iron concentration (1.0 mM FeCl₃ 6H₂O), but also to 116 μg/ml DHBA, remained at ∼6.5 log₁₀ CFU/ml for the first 4 h and then increased to 7.3 log₁₀ CFU/ml over the next 4 h (Fig. 5b).

The MIC of DHBA was 1.5 mg/ml (not shown). At this concentration, growth of P. aeruginosa Xen 5 was suppressed for 24 h, whereas the optical density of cells in the control (without DHBA) increased to 1.3 after 24 h (Fig. 5c).

DISCUSSION

The decrease in diameter of nanofibers electrospun with DHBA compared to nanofibers without DHBA may be ascribed to the change in viscosity and conductivity of the solution. Similar findings were reported by Heunis et al. (22) when nisin was incorporated into PDLLA:PEO nanofibers. Nanofibers with a smaller diameter have a larger surface-to-volume ratio and release incorporated compounds much more quickly (25, 26). Nanofibers with these properties are thus ideal to incorporate into dressings used for treatment of topical infections.

Almost all DHBA diffused from the nanofibers within 2 h (Fig. 2), rendering 24% PDLLA:PEO (50:50) the ideal carrier for the chelating agent. The release of DHBA from nanofibers chelated free iron, and by doing so repressed the growth of cells in the biofilm, as shown with decreases in optical density readings and a decline in viable cell numbers (Fig. 3). The drastic decline of cell numbers in iron-enriched (2.0 mM and 3.0 mM FeCl₃ 6H₂O) biofilms (Fig. 3f and Fig. 3h, respectively) is not surprising. Higher concentrations of iron are toxic to the cells (27). Based on these results, the 116 μg/ml DHBA incorporated into the nanofibers did not chelate all the iron.

The increase in mobility of strain Xen 5 in the presence of DHBA (Fig. 4) further supports our supports our finding that the cells are less likely to form biofilms in the presence of low concentrations (1.0 mM) of free iron. These results confirm the findings of Singh and coworkers (12), i.e., a decrease in mobility of Pseudomonas cells (and an increase in biofilm formation) in the presence of freely available Fe²⁺.

Results obtained with the in vitro slide model (Fig. 5a) corresponded to the release of DHBA from nanofibers, i.e., DHBA released from nanofibers depleted the freely available iron levels from the environment and slowed down the growth of strain Xen 5. The decline in bioluminescence recorded when cells were exposed to DF suggested that the metabolic activity of the cells decreased, which also led to a decrease in viable cell numbers. Lowering of iron levels to <10⁻¹⁸ M in wounds assists the chelating function of lactoferrin and transferrin (28). These proteins bind iron as a natural defense mechanism against bacterial infections and aid phagocytic cells and tissue fluids in resisting bacterial invasion. This is extremely important, as freely available iron can severely damage or destroy the mechanism of natural resistance and lead to rapid bacterial or fungal growth in tissue fluids.

The decrease in cell numbers recorded in deferrated LB broth supplemented with DHBA to cell numbers equivalent to those obtained in the absence of iron (Fig. 5b) indicated that the cells needed iron to sustain growth. This was confirmed by recording slightly better growth in the presence of 1.0 mM FeCl₃ 6H₂O, without DHBA (Fig. 5b). The increase in cell numbers (from 6.5 log₁₀ CFU/ml to 7.3 log₁₀ CFU/ml) after the first 4 h of incubation (Fig. 5b) is difficult to explain. It is possible that at this point in growth the remaining free iron left in the medium after chelation is at a concentration supporting cell division and hence a rapid increase in cell numbers.

DHBA at MIC levels (1.5 mg/ml) inhibited the growth of P. aeruginosa for at least 24 h (Fig. 5c). At this concentration, DHBA chelated iron to levels lower than required to support growth. It is also important to note that the DHBA concentrations used in the experiments with results shown in Fig. 3, 4, and 5a and b were performed with 116 μg/ml DHBA, far below MIC levels. Reduction of biofilm formation thus was not due to a direct antimicrobial effect of DHBA but rather was due to a decrease in free iron in the medium. Nanofibers containing DHBA (DF) exhibit two modes of action, i.e., (i) a decrease in biofilm formation due to the chelation of iron and (ii) an antiseptic effect that increases in the presence of iron, a mechanism that merits further investigation.

DHBA is less cytotoxic than other chelators, such as 1,2,3,4,6-penta-O-galloyl-β-d-glucopyranose (29) and 2DP (30). The chelators DTPA, EDTA, DM, and EDDA are also regarded as nontoxic (16). Dressings prepared from nanofibers containing DHBA may thus be an alternative treatment of topical infections caused by P. aeruginosa.