Antibacterial electrospun chitosan‐based nanofibers: A bacterial membrane perforator

Abstract This study investigates the antibacterial action of chitosan‐based nanofibers (CNFs) obtained by the electrospinning process on the permeability of bacterial membranes. The bactericidal efficiency of CNFs was first determined against Gram‐negative Escherichia coli and Salmonella Typhimurium, and Gram‐positive Staphylococcus aureus and Listeria innocua bacteria as a baseline. The results strongly suggest that CNFs interact with the negatively charged bacterial cell wall causing membrane rupture and inducing leakage of intracellular components among which are proteins and DNA. Results clearly indicate that the release of such components after contact with CNFs is an indication of membrane permeabilization and perforation, as pore formation was observed in transmission electron microscopy (TEM). This work suggests a plausible antibacterial mechanism of action of CNFs and also provides clear evidence in favor of chitosan as a bacterial membrane disruptor and perforator. As a result, CNFs can find promising applications as bioactive food packaging materials capable to extend shelf life of food products while inhibiting the spread of alteration flora and foodborne pathogens.

However, few have examined the antibacterial properties of CNFs. In a review article, Martínez-Camacho et al. (2011) point out that most reports on the antimicrobial activity of CNFs have used chitosan solutions instead. In most cases, the proposed mechanism for CNFs was indirectly related to the presence and release of protonated amino groups from CNFs mats, which were no longer nanofibers. The authors highlighted that further investigation would be useful in order to determine whether CNFs follow the same presumed mechanism, since it might be affected by the structural conformation these nanomaterials can adopt (Kong et al., 2008). The mechanism of action by which chitosan, in solution state, is able to inhibit or kill bacteria is a complex phenomenon that has not been fully explained either (Hammer et al., 2010;Kong, Chen, Xing, & Park, 2010;Raafat, Von Bargen, Haas, & Sahl, 2008). Moreover, no information is available regarding the mechanism underlying the antimicrobial activity of CNFs. To our knowledge, no study has reported the effect of CNFs on bacterial cell membrane integrity, nor their mode of action. A cytological study of the effect of CNFs on the bacterial membrane permeability is necessary to understand their exact mechanism of action and to avoid the outbreak of potential resistance phenomena.
In this study, we investigate the antibacterial mechanism of action of CNFs against four common alteration flora and foodborne pathogens, most frequently incriminated in food spoilage and food poisoning, respectively. All tests were performed under standardized and controlled experimental conditions to facilitate reproducibility and allow comparative studies. A plausible mode of action in which CNFs act as membrane permeability disruptor and even perforator is postulated. In this context, CNFs represent ideal biomaterials that can be used as suitable bactericidal barriers to prevent bacterial infections in several areas, including food packaging and biomedical applications. As part of active food packaging, CNFs can be applied to extend the shelf life of food products and prevent spoilage and foodborne diseases caused by Escherichia coli, Listeria, Staphylococcus, and Salmonella.

| Chemicals and polymers
Water-soluble chitosan (CS), a Venzym™ grade obtained via enzymatic treatment of chitin derived from shrimp shells was generously donated by Ovensa (Ontario, Canada). The water-solubility of this CS grade is due to the presence of a low amount of residual acetic acid (AcOH), as confirmed by the supplier. The corresponding degree of deacetylation (DDA) and number average molecular weight (M n ) are 95% and 50 kDa, respectively, with a narrow molecular weight distribution. Poly(ethylene oxide) (PEO), a cospinning agent for chitosan, with a molecular weight of 600 kDa, and acetic acid (AcOH, glacial, 99.7%) were purchased from Fisher Scientific (Saint-Laurent, QC, Canada). All materials were of analytical grade and used as received. transferred into a culture medium and finally incubated at 37°C for 24 hr in an orbital shaker (New Brunswick) to achieve an initial concentration of 10 9 colony forming unit per milliliter (CFU/ml).

| Culture media
Luria-Bertani (LB) broth and brain heart infusion (BHI) were used as growth media to start the bacterial cultures. To reach the required final concentration, cultures were diluted using phosphate buffer saline (PBS, pH 5.8, adjusted with 1 mol/L HCl). LB agar and BHI supplemented with agar (15 g/L) were used as solid media for counting the surviving bacteria.

| Preparation of chitosan and PEO stock solutions
Chitosan (CS) and PEO stock solutions (7% w/v and 3% w/v, respectively) were individually prepared by dissolving polymer powders in 50% (v/v) AcOH under overnight magnetic stirring. The CS/PEO blends were obtained by magnetic stirring of the two polymer solutions in a proportion of 80/20 (w/w) ratio for 4 hr agitation. The advantage of using aqueous acetic acid solutions is their nontoxic and ecofriendly character.

| Preparation of chitosan-based nanofibers via electrospinning
CS/PEO nanofibers were prepared according to Pakravan et al. (2011) using the electrospinning process. Electrospinning of the blend solution was performed using a horizontal homemade setup containing

| Scanning electron microscopy (SEM)
The morphology of the electrospun chitosan-based nanofibers (CNFs) was examined according to a slight modified method of Moayeri and Ajji (2015), using a field emission scanning electron microscope (FESEM JEOL JSM-7600TFE), operated at 1.5 kV. Samples were observed as collected on an aluminum foil after 2 hr electrospinning.
SEM results revealed that uniform and beadless fibers were obtained in the presence of the cospinning agent, PEO in this specific case.
The average fiber diameter was evaluated using Image-Pro Plus ® software. Approximately 600 nanofibers randomly chosen from three independent electrospun mats (200 fibers from each sample) were used for the quantification of fiber morphology parameters.

| Antibacterial efficiency of CNFs
The antibacterial activity of electrospun CNFs was evaluated in vitro following the American standard test method (ASTM E2149−13a, 2013). Commonly found bacteria, E. coli, S. aureus, L. innocua, and S. Typhimurium, in food contamination and skin infections were selected for this purpose. Samples of 1 cm 2 and 2.5 cm 2 swatches of CNFs were prepared in aseptic conditions. Bacterial suspensions (10 6 CFU/ml, 5 ml PBS, pH 5.8) were put in contact with CNFs. It is noteworthy that even though the CS grade used in this study was water-soluble, the resulting nanofibers were visually insoluble in aqueous media post-electrospinning due to solvent evaporation during processing. Negative controls of bacteria suspended in PBS without CNFs were also prepared. All tubes were placed at 37°C, optimal temperature for bacterial growth, for 4 hr incubation in an orbital shaker. Serial dilutions were performed and spread on agar plates incubated overnight at 37°C for further counting of survivors. All tests were conducted in triplicate. Finally, the antibacterial efficiency was expressed as a function of the reduction rate (R) of the total number of test bacteria. R was calculated according to Belalia, Grelier, Benaissa, and Coma (2008) using the following equation: where, A and B are the numbers of surviving bacteria in the controls and test samples, respectively.

| Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)
The release of intracellular proteins from CNF-treated bacteria was investigated by SDS-PAGE. In this section, E. coli (Gram-negative) and S. aureus (Gram-positive) were selected in order to appraise the effect of Gram-type on the strains' susceptibility/resistance to CNFs.
Overnight cultures of E. coli and S. aureus were resuspended in PBS (~10 8 CFU/ml) and incubated at 37°C in the presence of CNFs. After 0, 1, 2, 3, and 4 hr contact time, 5 ml aliquots were withdrawn and centrifuged at 3,000 g/10 min at 4°C. The supernatants were then mixed with trichloroacetic acid (TCA 10:1) and left for precipitation at 4°C overnight. After a series of wash, samples were resuspended in SDS-loading buffer and subjected to SDS-PAGE according to the method of Laemmli (1970). Positive controls (Ctrl + ) of extracted proteins from E. coli and S. aureus were also prepared by chemical lysis of both bacteria using a lysis solution containing 50 μl of chloroform and 25 μl of SDS (0.5% v/v). For more sensitivity, revelation was performed using silver nitrate staining of proteins.

| Agarose gel electrophoresis of released DNA
Because of its importance in fundamental research, its use in the industrial field and its involvement in the agri-food sector, the E. coli laboratory strain has been fully sequenced and its genome is currently 100% known. In the following section, E. coli (DH5α) bacterium was chosen to study the effect of CNFs on membrane permeability and subsequent DNA leakage.
F I G U R E 1 Schematic representation of the homemade electrospinning set up heat treatment (CtrlL + and CtrlH + , respectively) were also prepared. An additional step of pH adjustment (pH 7.0) with 1 mol/L NaOH in order to deprotonate the CNFs and break up CS-DNA interactions was necessary. A polymerase chain reaction (PCR) for the rrnB gene 16S RNA was performed in order to amplify the released DNA fragments from chitosan-treated cultures. Finally, DNA extracted sequences were loaded on a 2% (w/v) agarose gel and migrated for 20 min at 90 V.

| β-Galactosidase assay
In this section, E. coli DH5 hxt 55632-Lac Z + , a strain that overexpresses the gene encoding the β-galactosidase (β-gal) activity (without addition of lactose to the medium) was selected to assess the effect of CNFs on membrane permeabilization. To this end, the release of intracellular β-gal was evaluated by enzymatic titration according to Miller

| Transmission electron microscopy analysis of bacterial membrane integrity
Transmission electron microscopy (TEM) was performed to investigate the effect of CNFs on cell morphology and membrane integrity.
Sample preparation was performed following the guidelines of Tao (Figure 3).

| Antibacterial efficiency of CNFs
However, S. aureus and S. Typhimurium showed lower susceptibility to the action of CNFs. Nevertheless, a significant dose-dependent decrease of bacterial population (5 logs and 4 logs, respectively) was still observed (Figure 3). Furthermore, in order to increase the antisalmonella or anti-staphylococcal activity of CNFs, it is possible to combine chitosan with other antimicrobial agents such as ethylenediamine tetraacetic acid (EDTA, 0.2%) (Olaimat & Holley, 2015) and essential oils (Shahbazi & Shavisi, 2016), for a synergistic effect.

| Proteins leakage
The release of intracellular proteins is an indication of membrane deterioration. Figure 4 shows SDS-PAGE patterns of released cytoplasmic soluble proteins from chitosan-treated E. coli and S. aureus. In the case of E. coli, the protein content in the cell-free supernatant was similar to that of the positive control (Ctrl + ) that refers to bacterial

| DNA leakage
The release of bacterial genomic DNA in the supernatant was detected by PCR amplification of the rrnB gene (16S) for E. coli ( Figure 5) of the disruption of membrane permeability caused by CNFs ( Figure 5A and B). In contrast, no DNA was detected in the extracellular medium of untreated sample (Ctrl − , Figure 5D), which was synonymous with membrane integrity. The observed brightness at the loading spots of the treated samples was probably due to a deposition of small cationic chains of CS itself, which did not migrate towards the cathode. This can be also attributed to a deceleration of the electrophoretic mobility of genomic DNA caused by the chelation effect of chitosan, as suggested by Xing, Chen, Liu, Cha, and Park (2009b). Negatively charged phosphate groups present in nucleic acids, such as DNA and RNA, might be an intracellular target for CS and contribute to its interaction with bacterial cells. This conjecture was verified when CS was deprotonated (at neutral pH) in order to prevent CS-DNA complexation.
As a consequence, genomic DNA was detected both qualitatively and quantitatively. These results point out that the leakage of bacterial DNA would not occur without membrane perforation and strongly suggest a membranolytic effect in CNFs' mechanism of action. The concentrations of released DNA after exposure to CNFs, as measured using a NanoDrop spectrophotometer (ND-1000, Thermo Scientific), after PCR were 18.2, 19.5, 20.9, 60.2, and 172.3 ng/μl, after 0, 1, 2, 3, and 4 hr exposure times, respectively. Quantification of released DNA from CNF-treated E. coli clearly indicates that genomic DNA could be detected in the extracellular medium and its concentration was proportional to the contact time between E. coli and CNFs.

| Release of intracellular β-galactosidase enzyme
The release of cytoplasmic β-galactosidase (β-gal) was also an evidence of membrane permeabilization. Figure 6 shows the release of β-gal en- of CNFs to permeate bacterial membrane and coincide with the findings of Tao et al. (2011), who reported similar results for CS solutions.

| Transmission electron microscopy analysis of membrane permeabilization effect of CNFs
The effect of CNFs on membrane morphology and integrity was investigated by TEM (Figure 7). Untreated cells of E. coli (Gramnegative) and S. aureus (Gram-positive) were intact and did not show any membrane lesion or anomaly (Figure 7a and 6e). After exposure to CNFs, a remarkable alteration of membrane integrity was observed. TEM images of exposed cells to CNFs revealed that after 10 min contact, both E. coli and S. aureus strains showed membrane permeabilization by perforation (Figure 7b  E. coli and S. aureus was also observed and was proportional to contact time (Figure 7b, 7c and 7d). This might be due to (1)  ions, essential to bacterial growth. However, the common mechanism behind these different modes of action is undeniably due to the protonated functional groups of CS. The results clearly demonstrate that CNFs' bactericidal effect involves permeabilization of bacterial membrane with pore formation, contrary to what has been reported so far. However, no evidence of penetration of the membrane can be inferred, even though pore formation assuredly occurred. The next challenge should aim at clarifying the molecular mechanisms behind the bactericidal activity of CNFs and identifying the membrane elements and metabolic pathways involved in the internalization of chitosan into the bacterial cell wall. These further studies will not only be critical for the application of such materials in food packaging, but also for the prevention of outbreak of resistance phenomena toward chitosan.

| CONCLUSIONS
The results of this study show that the antibacterial activity of chitosan nanofibers (CNFs) can be attributed to membrane disruption and perforation. Consequently, this resulted in the leakage of intracellular components such as proteins and nucleotides. The bioavailability of NH 3 + functional groups on CNFs favored and maximized cell adhesion and attachment to the surface of the mats. The model established here, regarding CNFs' mode of action suggests that bacteria migrate to the surface of the nanofibers and not the reverse.
Since bacteria use adhesion and attachment surfaces to better grow and multiply, CNFs showed the ability to efficiently attract and trap bacteria through electrostatic interactions, on account of their large surface-to-mass ratio and high porosity. Our results also suggest that adsorption of CS to the bacterial surface is the first step in CNFs' mechanism of action, followed by membrane perforation, leakage of cytosolic compounds, and finally cell lysis and disintegration.
Nevertheless, it is not excluded that part of the antibacterial activity might be due to partial dissolution of the nanofibers, making chitosan available in solution. As promising practical application, CNFs can be used as part of active food packaging in order to extend the shelf life of food products along with preventing spoilage by bacteria such as

CONFLICT OF INTEREST
None declared.