Multifunctional Acidocin 4356 Combats Pseudomonas aeruginosa through Membrane Perturbation and Virulence Attenuation: Experimental Results Confirm Molecular Dynamics Simulation

Multidrug-resistant bacteria are a major threat to global health, and the Pseudomonas bacterium with the ability to form biofilms is considered one of the main causative agents of nosocomial infections. Traditional antibiotics have failed because of increased resistance. Thus, finding new biocompatible antibacterial drugs is essential. Antimicrobial peptides are produced by various organisms as a natural defense mechanism against pathogens, inspiring the possible design of the next generation of antibiotics. In this study, a new antimicrobial peptide was isolated from Lactobacillus acidophilus ATCC 4356, counteracting both biofilm and planktonic cells of Pseudomonas aeruginosa. A detailed investigation was then conducted concerning the functional mechanism of this peptide by using fluorescence techniques, electron microscopy, and in silico methods. The antibacterial and antibiofilm properties of this peptide may be important in the treatment of Pseudomonas infections.

ABSTRACT

A longstanding awareness in generating resistance to common antimicrobial therapies by Gram-negative bacteria has made them a major threat to global health. The application of antimicrobial peptides as a therapeutic agent would be a great opportunity to combat bacterial diseases. Here, we introduce a new antimicrobial peptide (∼8.3 kDa) from probiotic strain Lactobacillus acidophilus ATCC 4356, designated acidocin 4356 (ACD). This multifunctional peptide exerts its anti-infective ability against Pseudomonas aeruginosa through an inhibitory action on virulence factors, bacterial killing, and biofilm degradation. Reliable performance over tough physiological conditions and low hemolytic activity confirmed a new hope for the therapeutic setting. Antibacterial kinetic studies using flow cytometry technique showed that the ACD activity is related to the change in permeability of the membrane. The results obtained from molecular dynamic (MD) simulation were perfectly suited to the experimental data of ACD behavior. The structure-function relationship of this natural compound, along with the results of transmission electron microscopy analysis and MD simulation, confirmed the ability of the ACD aimed at enhancing bacterial membrane perturbation. The peptide was effective in the treatment of P. aeruginosa infection in mouse model. The results support the therapeutic potential of ACD for the treatment of Pseudomonas infections.

IMPORTANCE Multidrug-resistant bacteria are a major threat to global health, and the Pseudomonas bacterium with the ability to form biofilms is considered one of the main causative agents of nosocomial infections. Traditional antibiotics have failed because of increased resistance. Thus, finding new biocompatible antibacterial drugs is essential. Antimicrobial peptides are produced by various organisms as a natural defense mechanism against pathogens, inspiring the possible design of the next generation of antibiotics. In this study, a new antimicrobial peptide was isolated from Lactobacillus acidophilus ATCC 4356, counteracting both biofilm and planktonic cells of Pseudomonas aeruginosa. A detailed investigation was then conducted concerning the functional mechanism of this peptide by using fluorescence techniques, electron microscopy, and in silico methods. The antibacterial and antibiofilm properties of this peptide may be important in the treatment of Pseudomonas infections.

INTRODUCTION

Bacterial infections are one of the most common causes of mortality in clinical settings (1). The increased resistance to antibiotics is causing a serious threat to global health (2). A significant number of nosocomial infections are caused by Gram-negative bacteria due to a gradual decrease in available antibiotic sources to which they are susceptible (3). Hence, the development of new antibacterial agents targeting Gram-negative bacteria is of great importance (4). Natural compounds produced by a variety of probiotic bacteria are a reliable source for the discovery of new therapeutic drugs (5). Recent therapeutic progress in the treatments of bacterial infections has been focused on the reduction of harmful side effects on the human microbiome, minimizing their resistance to available antibiotics (6). Antimicrobial peptides (AMPs) have recently been introduced as potential alternatives to traditional antibiotics and are produced by all species of life forms as natural antibiotics (7–9). These AMPs are produced in multicellular organisms as a defense mechanism against microbial pathogens and are capable of modulating the immune system in mammals (7). The peptides produced by the bacteria are called bacteriocin (9).

AMPs are small peptides with a positive charge and significant amounts (>30%) of hydrophobic residues (9). AMPs interact with bacterial cell membrane through electrostatic interactions at the molecular level and cause physical and morphological damages (8). An alternative way of coping with the resistance of pathogenic bacteria is to use functionalized antivirulence agents to produce novel therapeutic agents (10). In contrast to conventional antibiotics, these compounds inhibit the activity or production of virulence factors without killing bacteria and have no direct severe or selective effects (11).

Pseudomonas aeruginosa is recognized as one of the most dangerous threats to human health; it is classified as one of the bacteria in the ESCAPE list of the Centers for Disease Control and Prevention (12). Owing to the inherent ability to increase resistance to antibiotics, biofilm formation, and the release of virulence factor arsenals, combating P. aeruginosa has been generally considered a long-established challenge (13).

P. aeruginosa produces several virulence factors that contribute to an orchestration framework, enabling it to colonize and adapt to the host (14). Virulence factors are bacterial products that cause diseases through the destruction of the host or the immune system (15). These virulence factors are controlled by a signaling system called quorum sensing (QS) (14). Since the pathogenesis of many bacteria is regulated by the QS system, the inhibition of this system can attenuate the virulence mechanisms of the pathogen, protecting the host against infections (16).

The biofilm formation by P. aeruginosa is associated with the development of chronic pulmonary diseases (17). The bacteria in biofilms are better protected against the human immune system and antibiotics so that they are up to a thousand times more resistant than their planktonic forms (18–20).

In an earlier study, Sarikhani et al. (31) reported the inhibitory effects of partially purified bacteriocin-like substances, 48 and 68 kDa in size, from Lactobacillus acidophilus ATCC 4356, against Bacillus subtilis. The focus of the present research is to detect a new potential antimicrobial peptide produced by this strain to inhibit other pathogens. We introduce a new prospective bacteriocin, designated acidocin 4356 (ACD), produced by the probiotic strain L. acidophilus ATCC 4356, a human isolate used as a dietary adjunct in dairy products. Our results show that this bacteriocin is a multifunctional bacteriocin against P. aeruginosa ATCC 27853 with antivirulence, antibacterial, and antibiofilm functions. Surprisingly, this peptide was defined as being able to cope with Pseudomonas-induced lung infection in a mouse model. We also report here a combined approach in which the fluorescent test data and the molecular dynamic (MD) simulation complement each other, defining the membrane-interacting mode of action of this bacteriocin against P. aeruginosa.

RESULTS

Purification and identification of bacteriocin.

As a preliminary step, the whole-cell-free culture supernatants (CFS) of L. acidophilus ATCC 4356 cells grown in MRS broth were concentrated and subjected to Tricine-SDS-PAGE (see Fig. S1a in the supplemental material [https://zenodo.org/record/3706572#.XmjKgKgzbIU]). Then, the bioassay-guided fractionation was carried out to identify the bioactive substances from crude proteins of concentrated CFS. The entire bacterial culture supernatants were precipitated with various ammonium sulfate (AS) concentrations to define the best percentage of AS for partial purification of the proteins with higher concentration. The saturation concentrations of ammonium sulfate of 50 and 80% were found to be ideal for precipitation and separation of the peptide (fraction I) and proteins >40 kDa (fraction II), respectively (Fig. S1b). Consequently, the activities of fractions I and II were examined and compared (Fig. S1c). According to the results, fraction I, at very low concentrations, was able to inhibit pyocyanin biosynthesis. However, fraction II was unable to inhibit pyocyanin production and only reduced the pyocyanin level when its concentration increased up to 50-fold compared to fraction I. Accordingly, the association between inhibitory activities with the peptide was confirmed.

To confirm the preliminary results, the purification of the active peptide was performed directly from CFS of L. acidophilus ATCC 4356 with 50% ammonium sulfate (see Fig. S1 in the supplemental material [https://zenodo.org/record/3706572#.XmjKgKgzbIU]), followed by C₁₈ (Fig. S2) and C₈ column high-pressure liquid chromatography (HPLC) analysis (Fig. 1a). To separate the peptide-related band from the other bands, the different freeze-dried fractions obtained from C₈ HPLC were characterized by SDS-PAGE analysis. It is worth mentioning that the fractions obtained in the separation with the C₁₈ and C₈ columns were used as semipurified and purified forms of ACD, respectively. Figure 1 shows typical chromatograms of a semipurified (Fig. 1a) and a pure (Fig. 1b) peptide. The SDS-PAGE image of the purified peptide is shown in Fig. 1c. The band of interest in the 5- to 10-kDa molecular weight range (Fig. 1c) was cut out, extracted from the acrylamide gel, and identified by liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis after tryptic digestion of in-gel proteins. The identification of the MS spectrum is summarized in Table 1 (for more details, see File S1 in the supplemental material [https://zenodo.org/record/3706572#.XmjKgKgzbIU]). The peptide structure prediction determined using the QUARK server is shown in Fig. 1d. Figure 1e shows that the computerized analysis of the sequence obtained is similar to a hypothetical protein of L. acidophilus ATCC 4356 with a domain of the bacteriocin-ΙΙc superfamily (21, 22). This newly identified bacteriocin, with a molecular weight of 8.3 kDa, was named acidocin 4356 (ACD). This bacteriocin contains a double glycine leader peptide that is cleaved through secretion outside the cell and has a mature form of 6.2 kDa. As shown in Table 1, three sequenced peptides that have full coverage with ACD do not consist of the leader peptide because the leader peptide is cleaved during secretion and could not be detected in the purified mature peptide fraction. Hence, the validation of the presence of the leader peptide in the bacteriocin was performed by DNA sequencing of the gene encoding ACD 4356 (see Fig. S3 in the supplemental material [https://zenodo.org/record/3706572#.XmjKgKgzbIU]).

FIG 1.

Purification and proteomic identification of acidocin 4356 (ACD) peptide. (a) Purification was performed using an Agilent 1260 infinity HPLC system. Acetonitrile–0.1% TFA in water was used as a mobile phase with a gradient of 5 to 95% and column C₈. The peptide-related peak is indicated by black dotted lines. (b) HPLC purity check. The collected peptide-related fraction using water-acetonitrile with a gradient of 5 to 95% was monitored at 220 and 280 nm. (c) Tricine-SDS-PAGE profile of ACD peptide. The desired peptide fraction was characterized by SDS-PAGE after HPLC, and the band of interest was then cut from the gel. The peptide-related band cut from the gel is indicated with a red dotted line (ladder [lane i] and peptide [lane ii]). (d) Three-dimensional model of ACD. The model was constructed using QUARK-based ab initio protein structure prediction and evaluated using DOPE score analysis. (e) Amino acid sequence of ACD determined using LC-MS/MS. The peptide of interest was extracted from the gel, digested with trypsin, and analyzed by LC-MS/MS. The bacteriocin-specific structural domain is indicated in the red box. The cleavage site of this bacteriocin is also indicated in red.

TABLE 1.

Peptide identification by LC-MS/MS

Parameter	Value
Locus tag (NCBI accession no.)	NH13_08920 (KHE29331.1)
Mol wt (kDa)	8.3
Description	Hypothetical protein
SEQUEST HT score	125.86
Sequence coverage (%)	43.47
Sequenced peptides	VAHCASQIGR, GSAACVSYLTR, GSAACVSYLTRHRHH

Anti-infective potential.

The spectrum of antibacterial activity was determined by the evaluation of the inhibitory effect of cell-free supernatant (CFS) of L. acidophilus ATCC 4356 culture on two groups of Gram-positive bacteria (Listeria monocytogenes ATCC 19115 and Staphylococcus aureus ATCC 25923) and Gram-negative bacteria (Escherichia coli ATCC 25922 and P. aeruginosa ATCC 27853). As can be seen in Fig. 2a, although CFS had antibacterial effects against all bacteria examined, it was able to inhibit the growth of P. aeruginosa by more than 80%. Therefore, further studies were conducted in order to characterize the antimicrobial and antivirulence features of ACD against P. aeruginosa. Both of these functions are parts of the process of discovering potential anti-infective treatments.

FIG 2.

Antimicrobial activity of CFS and ACD from L. acidophilus ATCC 4356. (a) Inhibitory effect of CFS on bacterial growth. Bacterial growth assays were performed by measuring the absorbance at 600 nm after 12 h of the same CFS treatment. As shown, CFS has an inhibitory effect on the growth of all four bacterial strains tested. However, there was substantially more impact against P. aeruginosa (*, P < 0.01 [Student t test]). (b) Antimicrobial activity of ACD against P. aeruginosa. The growth rate of bacteria after 12 h of treatment with different concentrations of ACD was evaluated using a fluorescence-based method (FDA, arbitrary units [AU]) and spectrometry (optical density [OD₆₀₀]). (c) Evaluation of antimicrobial activity of ACD under physiological conditions. (i) Comparison of inhibition of growth by ACD at its MIC₅₀ levels in the presence of human serum (HS), human plasma (HP), fetal bovine serum (FBS), high salt concentrations, and acidic conditions versus control. (ii) Growth inhibition of P. aeruginosa by ACD before and after treatment with eukaryotic and prokaryotic proteases. Each assay was performed with at least two independent experiments and three repetitions. *, P < 0.01 (one-way analysis of variance [ANOVA], significant difference from the control).

Antimicrobial properties.

The antimicrobial activity of ACD against P. aeruginosa was evaluated by measuring the optical density at 600 nm (OD₆₀₀) and cell viability with fluorescein diacetate (FDA) staining approaches. As shown in Fig. 2b, the growth inhibition is concentration dependent, and the growth rate of P. aeruginosa decreases with increasing ACD concentration.

From the comparison of the two methods, it can be assumed that the fluorescence-based method for cell viability assay presented more reproducible results, as reported earlier (20). Consequently, this measurement method was used for the rest of the analysis. The FDA fluorescence graph was plotted against various peptide concentrations, and linear regression analysis was used to determine the MIC (see Fig. S4a in the supplemental material [https://zenodo.org/record/3706572#.XmjKgKgzbIU]).

Antimicrobial activity of ACD under physiological conditions.

One of the most important obstacles in the use of AMPs is the impairment of their antimicrobial activities after exposure to bacterial and host proteases or intolerance to physiological conditions, leading to a limitation for their clinical use (23). As shown in Fig. 2c, it appears that the ACD retained its antibacterial efficacy under conditions similar to the physiological conditions (serum, plasma, high concentrations of salt, and acidic solutions). There was no significant reduction of ACD activity under these conditions compared to normal condition, except in the presence of 5% fetal bovine serum (FBS), in which the activity of the peptide decreased but its function was not completely inhibited. Comparison of P. aeruginosa growth inhibition by ACD at MIC₅₀, treated with different types of proteases, revealed resistance to proteases (Fig. 2c). The treatment of ACD with proteinase K and trypsin at concentrations at or below the MIC₅₀ did not adversely influence peptide function.

Evaluation of virulence factor production.

The results showed that the purified form of ACD could reduce the production of all four virulence factors of P. aeruginosa (pyocyanin, pyoverdine, elastase, and protease) (Fig. 3). These changes in virulence factor production occurred at ACD concentrations that did not affect bacterial cell growth (Fig. 3a). The reduction of all tested virulence factors was concentration dependent. However, significant differences were observed for the protease only upon treatment with a concentration of 20 μg ml⁻¹ (Fig. 3d). The most substantial effect was associated with the production of pyoverdine, since a significant difference was observed even at a concentration of 5 μg ml⁻¹ (Fig. 3b). The linear curve was drawn, separately, for the percent inhibition of all tested virulence factors versus the peptide concentration. The minimum virulence factor inhibition concentration (MVIC) index is defined as the average of changes in all four virulence factors (see Fig. S4b in the supplemental material [https://zenodo.org/record/3706572#.XmjKgKgzbIU]). Inhibition of pyocyanin (PCN) synthesis was not due to the interaction of the peptide with PCN, since the incubation of PCN-containing cell-free media of P. aeruginosa cultures with different concentrations of the peptide (0 to 20 μg ml⁻¹) did not significantly change the level of PCN detected (Fig. S5).

FIG 3.

Effect of ACD on the production of P. aeruginosa virulence factors. (a) Pyocyanin. The production of pyocyanin was measured at 695 nm after 24 h of treatment with concentrations of ACD that did not inhibit growth. The lack of blue-green color (right tube) indicates the inhibition of pyocyanin production at an ACD concentration of 20 μg ml⁻¹. (b) Pyoverdine. The production of pyoverdine was estimated by fluorescence spectroscopy (excitation, 400 nm; emission, 470 nm). Measurements are averages of three independent experiments. (c) Elastase. Elastase activity was measured using elastin-Congo red substrate at 495 nm. The color changes of the samples show a decrease in soluble hydrolysis products due to a decrease in elastase activity. (d) Protease. Proteolytic activity is estimated using a qualitative halo on skim-milk plates. Halo size indicates the proteolytic activity of the treated sample compared to the control. *, P < 0.01 (one-way ANOVA, significant difference from the control).

Effect of ACD against P. aeruginosa biofilm formation.

Although the ACD peptide had an inhibitory influence on planktonic cells, no significant effect of the peptide was detected for biofilm formation by P. aeruginosa over a period of 24 h (data not shown). This is different from the fact that the results obtained from confocal laser scanning microscopy (CLSM) and field emission scanning electron microscopy (FE-SEM) analysis showed consistently that the ACD has a significant effect against established biofilm (Fig. 4a and b). Based on the results of CLSM analysis of biofilms, the cells in the control sample (untreated) form a smooth and uniform layer, and all the live cells are stained with green fluorescence (Fig. 4b). However, the sample treated with peptide at its MIC level lost membrane integrity against propidium iodide (PI), and >90% of the P. aeruginosa cells in the biofilm were killed (Fig. 4c). The presence of dead cells in the internal layers of biofilms demonstrates the ability of ACD to penetrate into the depths of the biofilm (see Video S1 in the supplemental material [https://zenodo.org/record/3706572#.XmjKgKgzbIU]). The results of the FE-SEM analysis also showed that the biofilm structure is fairly preserved at this concentration (MIC) and that biofilm fragmentation only occurred in some areas (Fig. 4a). However, the degradation of biofilm just occurred when the concentration of peptide was increased up to 2× MIC (Fig. 4a). In addition, the results from crystal violet staining confirmed that the treatment of sample with peptide at its MIC level caused only lower than 25% reduction in the overall biofilm biomass (Fig. 4d). The changes in the numbers of live and dead cells determined by CLSM at various concentrations of peptide are shown in Fig. 4c. Based on the CLSM results, ACD caused significant killing (>90%) of bacteria in the biofilm at the MIC level, whereas eradication of large parts of the biofilm occurred as the concentration of peptide is increased up to 2× MIC (Fig. 4c and d). A decrease in biofilm biomass occurred, along with a reduction in the numbers of live and dead cells in the biofilm. Therefore, upon increasing the concentration of ACD up to 2× MIC, a decreased number of live cells were detected in a dose-dependent manner. However, the increased detachment of dead cells from the biofilm surface made it impossible to observe this (Fig. 4e).

FIG 4.

Effect of ACD peptide on P. aeruginosa biofilm. Images were taken from P. aeruginosa biofilms formed in TSB plus 0.2% glucose culture medium after 72 h without treatment and after an hour of treatment. (a) SEM images. (i) Control; treatment at MIC (ii) and at 2× MIC (iii). (b) Three-dimensional CLSM images from biofilms. (i) Control; (ii) treatment at its MIC level. The experiment was repeated three times; three regions were examined each time, and a representative image is shown for each of the conditions. The live cells are green, and dead cells are seen in red. A video of peptide effects on P. aeruginosa cells in various biofilm layers is accessible in the supplemental material (https://zenodo.org/record/3706572#.XmjKgKgzbIU). Images were taken with a 40× lens objective. (c) Subsets of dead and live cell populations in biofilms. The live/dead ratio was calculated using ImageJ software using six points. *, P < 0.01 (Student t test, significant difference from the control group). (d) Determination of detachment of P. aeruginosa biofilm after 1 h of treatment with various concentrations of peptide. Biofilm mass was measured by crystal violet staining. The results were derived from an average of three independent repetitions as ± the standard deviations. *, P < 0.01 (one-way ANOVA, significant difference from the untreated control). (e) Comparison of the detachment of P. aeruginosa biofilm with bacterial killing after ACD treatment. The pictures were taken from a mature biofilm after treatment with ACD. In the fluorescent images, green and red fluorescence shows live and dead cells, respectively (scale bar, 50 μm).

Selectivity evaluation of ACD.

The cytotoxicity of ACD toward human erythrocytes was assessed by measuring the hemolysis percentage of human red blood cells (hRBCs) and calculating the minimum hemolysis concentration (MHC). At the highest concentration tested, 400 μg/ml (>3× MIC), the hemolysis was less than 20% (see Fig. S6 in the supplemental material [https://zenodo.org/record/3706572#.XmjKgKgzbIU]). Subsequently, the therapeutic index (TI), a parameter showing the specificity of pathogens toward the mammalian cells, was calculated (Table 2).

TABLE 2.

Therapeutic indices of peptides for anti-infective properties^a

Peptide	MVIC50	MIC50	MHC50b	TIc	TId
Antivirulence activity	13.90		>400	57.55
Antimicrobial activity		79.43	>400		10.07

^aMVIC, minimum virulence factor inhibition concentration; MHC, minimum hemolysis concentration; TI, therapeutic index.

^b50% hemolysis was not observed.

^cTI = MHC₅₀ (μg ml⁻¹)/MVIC₅₀ (μg ml⁻¹). When 50% hemolytic activity was not observed at 400 μg ml⁻¹, a value of 800 μg ml⁻¹ was used to calculate the therapeutic index. Larger values indicate greater antimicrobial specificity.

^dTI = MHC₅₀ (μg ml⁻¹)/MIC₅₀ (μg ml⁻¹).

Preliminary mechanistic studies.

Our studies were conducted at three stages to gain insight into the interaction of ACD peptide with P. aeruginosa cells, as detailed below.

Flow cytometry analysis.

It is well known that cell membrane permeability to vital stains, such as propidium iodide (PI), is associated with complete population death (24, 25). The precision of FCM quantifications of bacterial cells and the correlation of PI-positive bacteria with cell death are shown in Fig. S7 in the supplemental material (https://zenodo.org/record/3706572#.XmjKgKgzbIU). The live and dead cell population was examined by flow cytometry (FCM) with FDA and PI dyes. The death rate of 100% was adjusted using isopropyl alcohol to kill bacteria (26). According to Fig. S7, in a live cell population, more than 97% of the stained cells show only the green color, indicating the presence of a continuous intact membrane that does not have the ability to penetrate the PI. In the dead cell population, 99% of the cells showed the ability to absorb PI. However, ∼14.99% of the cell population showed both colors. On the other hand, almost all of this population in single staining with PI showed red fluorescence. In fact, it seems that in the short term the remaining activity of the esterase enzyme of dead cells caused the creation of a green fluorescence (27), which gradually disappeared, and the cell population was pulled down to quadrant 4 of flow cytometry plot (Q4). Therefore, based on fluorescence of PI-treated cells with isopropanol, PI-positive cells were considered dead cells, and the kinetics of ACD-induced bacterial killing were studied based on PI absorption by using a flow cytometry technique.

The study of the killing kinetics of ACD indicated that treatment of P. aeruginosa with >4× the MIC of the peptide for 15, 30, and 60 min could destroy the cell membrane of more than 95% of the live cells after 30 min (Fig. 5a). It appears that ACD has a very quick action on the P. aeruginosa cell membrane, and the fast kinetics of cell death caused by ACD is similar to that of an antimicrobial peptide with fast membranolytic action reported by Lam et al. (24) (Fig. 5a). Surprisingly, the flow cytometry results obtained from treatments with different concentrations of peptide at 2× and 4× the MICs at 30 and 60 min showed no significant difference in PI permeability, indicating that the mechanism of bacterial inactivation is time dependent, and only the absorption of FDA fluorescence is altered during this time (Fig. 5b). The difference between the treatment times of 15 and 30 min, at 4× the MIC, confirmed an explicit dependence on the time. Our results are in agreement with a previous report indicating that the kinetics of the antimicrobial compound PepR, which induced the penetration of PI into the E. coli, were sigmoidal, and thenceforth the penetration curve entered the plateau phase (26). Altogether, the FCM analysis demonstrated that the ACD exerts its inhibitory action on P. aeruginosa growth through membrane disruption. Bacterial cell death after 60 min of treatment with 4× MIC was verified by measuring the CFU. At this concentration, the CFU/ml level was zero. Therefore, based on the minimum bactericidal concentration (MBC) as the lowest concentration of peptide in which the CFU/ml is zero, the MBC of ACD is ≤4× MIC.

FIG 5.

Mechanistic studies of action of ACD against P. aeruginosa. (a) Kinetics of ACD-induced membrane permeabilization. A flow cytometry correlogram (PI versus FDA) is shown at the specified time points after treatment with peptide at a concentration of 4× MIC. PI-positive and FDA-positive results are indicated in red and green, respectively. The population gating selection was determined based on live cells (control without treatment) and cells killed with isopropyl alcohol (dead cells). Two independent measurements were performed with two replications for each variable. (b) Comparison of the kinetics of antimicrobial activity of ACD at various concentrations by flow cytometry. The permeability of P. aeruginosa cell membrane was evaluated after treatment with different concentrations of ACD for 1 h in a linear fashion with time. Comparison of two concentrations at different times indicated that the antibacterial action of the ACD peptide is time dependent. On the dot plots, the x and y axes represent the fluorescent emissions of PI and FDA, respectively. (c) Evaluation of morphological changes of P. aeruginosa cells 1 h after treatment. TEM images of P. aeruginosa without treatment (i) and treated with ACD at its MIC (ii to iv) are shown. Untreated cells display a normal cell shape and undamaged membrane (i). Deformation of cell membrane and a loss of integrity (ii), detachment of the bacterial membrane from the bacterial cytoplasm (iii), and complete cell lysis (iv) occurred over time. (d) TEM micrographs of P. aeruginosa treated with ACD after 4 h. (i and ii) Cells without peptide treatment (control); (iii and iv) cells after treatment with an MIC level of ACD. (i) Cells have complete cell membrane and homogenous cytoplasm; (ii) cells are dividing without any damage; (iii) more damaged cells with signs of general lysis are observable; (iv) the membranes of the dividing cells are degraded, and leakage of the intracellular contents from the cytoplasm is visible. The regions of interest are indicated by red arrows.

Transmission electron microscopy.

Transmission electron microscopy (TEM) imaging was used to study the direct effect of ACD on cell morphology of P. aeruginosa. Figure 5c shows the images of cells after peptide treatment for 60 min. An untreated P. aeruginosa sample displayed an intact cell structure (Fig. 5c), while the cells exposed to the ACD revealed extensive structural alterations, in particular, in the bacterial cell wall. As shown in Fig. 5b, there is a considerable amount of aggregates in all of the treated samples, possibly as a result of the interaction of peptide with medium components (26). The intracellular integrity was impaired after treatment with ACD. As shown in Fig. 4, the bacterial membrane became detached from the bacterial cytoplasm, cell lysis was observed, and the cell ripping of P. aeruginosa was induced after ACD treatment for 4 h (Fig. 5d). In the treated sample, the membrane of the dividing cells is degraded and the leakage of the intracellular material is visible in the cytoplasm. These microscopic observations are in line with FCM data and confirmed that the inhibitory function of this peptide is related to cell wall integrity disruption.

Coarse-grained molecular dynamic simulations.

Based on a DOPE score analysis of peptide QUARK models, a reliable model of ACD was predicted for the rest of the analysis (see Fig. S8 in the supplemental material [https://zenodo.org/record/3706572#.XmjKgKgzbIU]). To gain insight into the location of ACD in bilayer and the membrane destruction of P. aeruginosa, coarse-grained molecular dynamic (CG-MD) simulations were carried out using a self-assembly of mixed peptide-lipid system. In MD simulations using CG models, two-layer, single-bilayer, parallel bilayer, and cylindrical bilayer structures are constructed (28). In addition, in the self-assembly of lipids, by increasing the lipid concentration, the two-layer parallel structure is converted to a cylinder form (29, 30). Accordingly, in our study, two-layer parallel and cylindrical structures were created. Among these models, a double-layer cylindrical membrane was selected due to the simultaneous binding of a single peptide and the accumulation of peptides in the membrane (Fig. 6a). The root mean square deviation (RMSD) diagram (Fig. 6a) shows that this structure was stable until the end of simulation time. Hydropathy analysis of ACD sequence by ProtScale tool shows that the 34 first amino acids and 6 amino acid endpoints are hydrophilic and have a role in attachment to the outer surface of the membrane. Amino acids 35 to 77 are rich in hydrophobic amino acids that play a role in penetrating the inner part of the membrane (Fig. 6b). A cross-sectional area of peptide-membrane interaction was used to analyze the density and distribution of phosphate groups. A different view of the interaction of a single peptide and peptide complex with membrane, at specified times of the simulation, and the corresponding electron density profile are shown in Fig. 7a and b. Thus, although the peptide in both single and complex forms has the ability to properly interact and integrate into the membrane, the membrane perturbation occurs only when the peptide is in its complex form (Fig. 7a and b).

FIG 6.

Equilibrated bilayer configuration containing ACD. (a) Representative structure of the end of simulation time and fluctuations of RMSD values during simulation time. In the inset image, the interactions of a single peptide and the peptide complex with P. aeruginosa membrane are shown. The RMSD value is seen to be constant after about 100 ns until the end of the simulation time period, and this indicates that this simulation was stable. (b) Hydropathy plot analysis of the ACD sequence using the ProtScale tool. Hydrophilic and hydrophobic regions in the secondary structure are indicated in red and green, respectively. The configuration of peptides in the membrane in complex form shows that their hydrophobic regions pass through the membrane and interact with the hydrophobic tails of the lipid in the inner part of bilayer.

FIG 7.

Self-assembly of a random distribution of ACD peptides, lipids, and water from CG-MD simulations. (a) Snapshots of bilayer formation during simulation. (i) Initial random distribution of molecules; (ii) time interval (0 to 200 ns); (iii and iv) final configuration. Simulations indicate that the final shape is remarkably cylindrical, and side views of the cylindrical bilayer are shown. Cross-sections of peptide-membrane interaction are also shown for a single peptide (iii) and the peptide complex (iv). (b) Density analysis of peptide, lipid, water during molecular dynamics simulation. Density profiles are presented from left to right corresponding to stages i to iv in panel a. All data are normalized. EDP, electron density profile. (c) Study of lipid bilayer perturbation by assessment of the distribution of phosphate groups within membrane. (i) Single form; (ii) complex form. PO₄ atoms, NH₃ atoms, and lipid acyl chains of lipids are indicated in yellow, blue, and cyan, respectively. For clarity, water is not depicted. The peptide is represented as a ribbon. The colors represent the hydrophobicity of the amino acid residues (green, hydrophobic; red, hydrophilic).

The distribution of phosphate groups within the membrane during the interactions proved that although the presence of a single peptide does not have much effect on membrane integrity, the peptide in its complex form caused disorganization of the phosphate distribution chart, representing membrane perturbation (Fig. 7c).

Anti-infective activity of ACD in a mouse model.

An in vivo study was conducted to determine whether the ACD has the capability of protection against P. aeruginosa infection (Fig. 8a). Images of lung sections from histopathological examination of mice are illustrated in Fig. 8b. Pulmonary histopathological scoring of each group was performed to grade based on hallmarks of lung pathology as shown in Fig. 8c. In addition, the details of hematoxylin and eosin (H&E) scoring of the stained sections are indicated in Fig. S9 in the supplemental material (https://zenodo.org/record/3706572#.XmjKgKgzbIU). As shown in Fig. 8b, in the control group, the ciliated columnar epithelium was intact, no macrophage infiltration was detected, and the airway was free from exudate. In the lungs of the infected group, increased infiltration of immune cells, peribronchial inflammation, and hemorrhage were observed. Moreover, the normal columnar epithelial cells were replaced by cuboidal epithelium cells. The lung sections obtained from peptide-treated mice (the treated I group) displayed less recruitment of macrophage and pneumocyte, epithelial hyperplasia, and structural degeneration than the control group. Although there were small focal areas accompanied by hemorrhage, the inflammation was significantly reduced. The lungs of the peptide-treated II group show multifocal hemorrhagic areas, interstitial thickening, and increased macrophage infiltration. However, these parameters showed a significant reduction compared to the infected group.

FIG 8.

In vivo analysis of ACD activity. (a) Schematic illustration of the design of the in vivo experiment. (i) Mice were anesthetized and then infected with intranasal instillation with 20 μl of PBS buffer containing 10⁷ CFU of P. aeruginosa/ml. One group was inoculated with a nonbacterial PBS buffer (control group). (ii) At 1 h after infection, intranasal administration of ACD was performed for two groups: the treated I group (12 mg kg⁻¹) and the treated II group (6 mg kg⁻¹). An untreated group was also included as infection group. (iii) At 24 h after treatment, the mice were euthanized, lung tissues were isolated, and slides of lung tissues were prepared to evaluate the histology of the specimens. (b) Comparison of mouse pulmonary tissue properties and histology. (i) Tissue sections were stained with H&E for comparison of the lung inflammation properties. (ii) PAS-stained lung tissue sections to measure goblet cells (arrowheads). Representative images are shown for n = 3 mice. (c) Histopathological tissue scoring of each group based on the characteristics of the lung pathology. (d) Goblet cell count. The data are based on two independent experiments, and at least two replicates were used for each variation. *, P < 0.01 (one-way ANOVA, significant difference from the control group).

The goblet cell population, an important criterion in lung pathology, was evaluated using periodic acid-Schiff (PAS) staining (Fig. 8b and d). The results showed that the number of goblet cells escalated as a result of a further increase in macrophage infiltration and hyperplastic changes in epithelial cells (Fig. 8d). Histology investigations also revealed a reduction of goblet cells in the lung tissues of the two treated groups compared to the infected group. Excitingly, no noticeable difference was recorded between the treated groups (I and II) and the control group, except in the H&E scoring of the treated II group.

DISCUSSION

The interest in bacteriocins produced by lactic acid bacteria (LAB) is due to a long history of their consumption. They are known as “generally recognized as safe” (GRAS), and the use of the semipurified form of bacteriocins has been recommended as user-friendly therapy, like that of the cost-effective semipurified form of nisin, which is available as Nisaplin and used for food or pharmaceutical purposes (9). The present work is a landmark study demonstrating the capacity of a new bacteriocin, acidocin 4356 (ACD), produced by the probiotic strain L. acidophilus ATCC 4356, as an anti-infective agent against P. aeruginosa in vitro and in vivo.

The inhibitory effect of the semipurified form of CFS obtained from L. acidophilus ATCC 4356 cultures on the growth and swarming of Proteus sp. and Bacillus subtilis BM19 has already been described (31, 32). Sarikhani et al. (31) reported the inhibitory effect of a partial purified form of bacteriocin-like substances with a molecular weight of >40 kDa from Lactobacillus acidophilus ATCC 4356 on Bacillus subtilis. However, considering the partial purification of the antibacterial molecules and the lack of study on peptides with a molecular weight of <10 kDa, ACD may also play a role in its antibacterial activity.

Also, Han et al. (33) reported the isolation of 3.1-kDa peptide from L. acidophilus ATCC 4356 with antibacterial activity against Gram-positive bacteria such as Staphylococcus epidermidis, Listeria innocua, and Staphylococcus intermedius. However, no antimicrobial effect was detected against Gram-negative pathogenic bacteria. The present study reports the isolation and identification of an ACD peptide of 8.3 kDa. It seems that L. acidophilus ATCC 4356 produces two bacteriocins of different sizes, showing at least 8 amino acids that are identical at the N-terminal region. However, ACD has 40 more amino acids than the 3.1-kDa bacteriocin and exhibited antibacterial activity against Gram-negative bacterium P. aeruginosa.

The antimicrobial activity of ACD (MIC₉₀ = 128.22 μg ml⁻¹) is comparable to that reported in earlier studies of peptide-based antimicrobials. For instance, in a previous study, the MICs of cationic antimicrobial peptides, such as melitin, defensin, and magainin II, against standard and clinical strains of P. aeruginosa were reported to be >128 mg ml⁻¹ (34). ACD was able to inhibit the major virulence factors (35, 36), including pyoverdine siderophore, pyocyanin toxin, secretory protease, and elastase enzyme, and therefore can be considered as an anti-virulence strategy to counteract P. aeruginosa. Since ACD plays its role as an antibacterial agent at higher concentrations and displays antivirulence at lower concentrations, it would seem to be a promising AMP for combination therapy to control the pathogenicity of Pseudomonas sp. Thus far, many studies have been conducted to investigate the inhibition of QS-controlled virulence factors for controlling the pathogenicity of P. aeruginosa using various synthetic molecules (37–40) or those structurally resembling natural compounds (41). The present study describes a bacterial natural peptide with such an ability. The containment of biofilm formation is a potential field approach to reduce the persistence of bacterial pathogens. The inability of antibacterial compounds to rapidly penetrate every part of biofilm is one of the causes of antimicrobial resistance in bacterial biofilms. The penetration limitation can be due to the attachment of antibacterial compounds to the biofilm matrix or to their possible inactivation by enzymes such as β-lactamase (42). Although the growth of bacteria was affected by ACD, the ACD had no inhibitory effect on biofilm formation. At first sight, it seems indeed reasonable to conclude that the ACD can kill the planktonic bacterial cells in the initial inoculum, preventing the formation of biofilms at early hours. However, this difference stops being noticeable with the prolonged time required to form biofilms by the proliferation of bacteria.

Here, an effort was made to provide a basis for the development of an efficient antimicrobial agent overcoming the penetration limitations in P. aeruginosa biofilm. Our results suggest a dual function for the bactericidal action of ACD against preformed biofilms of P. aeruginosa. At lower concentrations, the living cells in the biofilm often undergo cellular death, and the overall structure of the biofilm is maintained to some extent, whereas at higher concentrations biofilms are detached from the surface, and the general disruption of biofilm structure is observed. In a previous report, Chen et al. investigated the antibiofilm activity of thirteen host defense peptides (19). Although 12 of these peptides could kill bacteria in biofilm, only one of them could eliminate biofilm biomass. Our findings confirm the potential of ACD as a new strategy for coping with persistent infections caused by P. aeruginosa.

The term AMPs just refers to peptides that can kill bacteria under physiological conditions (9). The proper functioning of ACD under challenging physiological conditions confirms its potential as an antimicrobial peptide for the treatment of respiratory infections caused by P. aeruginosa. The potential of this peptide for resistance to proteolytic degradation is an advantage over other AMPs, such as LL-37, which hydrolyzed in the presence of trypsin and proteinase K (43, 44). The proteolytic degradation of trypsin is dependent on the concentration of the substrate. For example, it has been reported that trypsin inactivation of the antibacterial peptide LL-37 against P. aeruginosa occurred at low concentrations but not at higher concentrations (45).

An important criterion in the selection of drug candidates is the therapeutic index (TI), a quantitative relationship between pharmacology and toxicology (46). Since the main problem of amphipathic cationic peptides is their hemolysis activity, the TI is calculated based on the hemolytic activity of ACD (47). The low hemolytic activity and the high TI value of ACD are promising features for therapeutic options (46, 47).

The lack of basic knowledge concerning the mechanism of action of bacteriocins has confined the application of these peptides (48). Having information about the function, structural features, and interaction of peptides with bacteria is considered critical for discovering the new cellular targets and the development of antimicrobial drugs (49). AMPs play an antimicrobial role by interacting with cell membranes, followed by membrane disruption and degradation or translocation and interaction with plasma membranes. Changes in the membrane structure include thinning, pore formation, altered curvature, modified electrostatics, and localized perturbations (7).

Considering the low permeability of P. aeruginosa outer membrane compared to other Gram-negative bacteria, the results obtained from FCM analysis indicated that the ACD at concentrations of 2× and 4× MICs is able to change PI permeability, as a membranolytic compound (Fig. 6). Increased uptake of PI by the P. aeruginosa cells indicates a decrease in the number of live cells (>96%) after 30 min of treatment with ACD at a 4× MIC level. Consequently, the minimum bactericidal concentration (MBC) value, as measured based on killing levels of the peptide, was below 4× MIC, suggesting that the peptide is a bactericidal agent (50). Previous studies involving the antibacterial mechanisms of cationic peptides suggest that while membrane degradation occurs at very high concentrations of the drug, the main mechanism of cell destruction may be triggered through other intracellular targets (51). Hence, the effect of ACD at MIC level on the P. aeruginosa membrane was imaged using TEM. However, in contrast to our expectations, pore formation, as a frequent mechanism of action of AMPs, was not detected in the TEM images, as also found for the mechanism of action of the HD-6 antimicrobial peptide (52).

MD simulation of membrane-disrupting ACD was performed to investigate the dynamic behavior of the peptide-membrane system at the atomic level. The middle hydrophobic portion of this peptide enters the cell membrane. The presence of several positively charged and hydrophobic amino acid residues allows the peptides to adopt a three-dimensional amphiphilic structure, forming separate patches while reacting to the membrane by these positive charges and hydrophobic amino acids. The existence of the GXXXG motifs in the middle portion of the peptide is an interesting structural feature, which is involved in helix-helix interactions and the oligomerization of peptides in the membrane. This motif is common among transmembrane peptides due to its proper role for penetration into the membrane (53). Coarse-grain MD simulation provided information regarding the possible mechanisms involved in membrane perturbation by ACD. While single peptide can penetrate the membrane, the cumulative effects of the peptides in complex form caused the perturbation and membrane destruction. It has previously been reported that AMPs are attached to the membrane surface to reach the critical threshold and then self-organized to form a permeation pathway (54, 55). Thus far, the functional mechanisms of several antimicrobial compounds, such as LAH4 (56), Piscidin1 (57), PMAP-23 (58), and Plantaricin E (53), have been investigated using MD simulation. The results of MD studies confirmed a direct correlation between the antibacterial function of ACD 4356 and membrane destruction, as indicated by experimental data. Nevertheless, the exact function of this peptide may involve a multiple-step model that is not yet known to us. Respiratory tract infection is one of the most common pulmonary diseases, and P. aeruginosa, as the most common pulmonary pathogen, can induce fatal infections (59). Despite the high prevalence of these respiratory diseases and extensive research for the development of new therapies, there are few studies concerning the antimicrobial activity of AMPs in the animal model of respiratory tract infection with P. aeruginosa. For example, the antipseudomonal activity of AMPs such as pyocin SD2 (60), ESC (1, 2), and its diastereomer has been demonstrated in infected mouse lungs (59). Here, the changes that occurred in immune responses in a mouse model of respiratory infection after treatment with ACD using histopathological image analysis confirmed the significant potential of this peptide as an alternative therapeutic option for P. aeruginosa pulmonary infections.

All things considered, the anti-infective efficacy of ACD peptide against P. aeruginosa infections provides a reasonable ground for its development as a novel antibacterial drug candidate to cope with this pathogen. However, this natural peptide with proper bioavailability may still have a limitation for possible clinical use. The most important limitation of ACD is the low efficiency of its production and purification, wherein the development of a recombinant strain of L. acidophilus can overcome this problem. On the other hand, the large size of the peptide has made it difficult to synthesize the peptide chemically. It is assumed that nature inspires us because it proposes the most versatile and best form of AMPs. Accordingly, the de novo peptide design based on the expected pattern through a genetic algorithm approach may lead to the production of these kinds of small peptides with high activity (58). Further toxicity testing related to P. aeruginosa-induced lung infections is still needed to determine ACD's significance as a therapeutic agent and the exact mechanism of its antibacterial action.

MATERIALS AND METHODS

Purification and identification of peptide from L. acidophilus ATCC 4356 supernatant.

The bacteriocin was purified from the cell-free supernatant (CFS) of a grown culture through a multistep protocol, consisting of a combination of salt precipitation and chromatography steps, as described below. In the first step, to detect the anti-infective substances, the partial separation of the protein bands was performed using various ammonium sulfate concentrations. For precipitation with ammonium sulfate, it was slowly added to the cell-free supernatant solution (until the saturation reached 30%) and stirred for 2 h. The solution was then centrifuged, and ammonium sulfate was added to the supernatant until the optimum saturation was reached and then stirred for two more hours. The precipitate was collected by centrifugation and then dissolved in sodium citrate buffer (pH 5), followed by dialysis over a period of 16 h with at least three changes against this buffer. In the initial purification steps, a bioassay-guided fractionation technique was used, and all fractions were examined for their antibacterial and antipyocyanin properties against P. aeruginosa. Subsequently, peptide purification was carried out by ammonium sulfate precipitation of bacterial culture supernatants at 50% saturation, followed by HPLC analysis (using an Agilent 1260 Infinity HPLC system equipped with an ODS-3 C₁₈ column and a C_8-3 column, respectively). The resulting protein extract was separated by using an Agilent 1260 Infinity LC system equipped with an ODS-3 C₁₈ column (250 by 4.6 mm, 5 μm). The column was washed with a mobile phase of water-acetonitrile containing 0.1% trifluoroacetic acid (TFA) and a gradient from 5 to 95% acetonitrile at a flow rate of 1 ml/min. The loading volume of the sample was 100 μl. The fractions related to the same peak were pooled and, after condensation, used for Tricine-SDS-PAGE. The peptide-containing fractions were lyophilized after buffer exchange and used as a milky powder for evaluation of their antimicrobial activity. The inhibitory effects of the protein extract on both cell growth and PCN synthesis of P. aeruginosa were investigated.

The peptide-containing fraction was further purified by using a C₈ HPLC column (C_8-3 column, 250 by 4.6 mm, 5 μl) according to the same procedure described above, except that the mobile phase was water-acetonitrile containing acetic acid at pH 4.5. Once again, Tricine-SDS-PAGE analysis was performed, and the fraction containing the pure peptide was detected. The active peptide band was cut from the SDS-PAGE gel and identified by LC-MS/MS analysis. Peak purity analysis in HPLC with a C₈ column was performed under the same conditions as mentioned before. The purified fraction was used, after buffer exchange, for its inhibitory effect on PCN synthesis. Due to the low efficiency of HPLC purification with a C₈ column, further experiments were carried out using the semipurified peptide.

Peptide identification.

LC-MS/MS analyses were performed on a Thermo EASY nLC II LC system coupled to a Thermo LTQ Orbitrap Velos mass spectrometer equipped with a nanospray ion source. The silver-stained peptide gel samples were in-gel digested using trypsin for 16 h at 30°C. A volume of 2 μl of each sample containing about 100 ng of tryptic peptides was injected onto a column (10 cm × 100 μm) in-house packed with Michrom Magic C₁₈ stationary phase (5-μm particle diameters and 300-Å pore size). Peptides were eluted using a 35-min gradient at a flow rate of 400 nl/min with mobile phases A (96.9% water, 3% acetonitrile [ACN], and 0.1% TFA) and B (97% ACN, 2.9% water, and 0.1% TFA). A full MS spectrum (m/z 400 to 1,400) was acquired in the Orbitrap at a resolution of 60,000, and then the ten most abundant multiple charged ions were selected for MS/MS sequencing in a linear trap with the option of dynamic exclusion. Peptide fragmentation was performed using collision-induced dissociation at a normalized collision energy of 35% with an activation time of 10 ms. The MS data were processed using Thermo Proteome Discoverer software (v2.2) with the SEQUEST search engine. Database searches were against UniProt L. acidophilus proteome database (UniProt UP000030883) and Antimicrobial Peptide Database (http://aps.unmc.edu/AP/main.php). The enzyme for the database search was chosen as trypsin (full), and a maximum missed cleavage site was set at 2. Mass tolerances of the precursor ion and the fragment ion were set at 10 ppm and 0.6 Da, respectively. Dynamic modifications on methionine (oxidation, +15.9949 Da) and protein N terminus (acetyl, +42.0106) and static modifications on cysteine (carbamidomethyl, +57.0215 Da) were allowed. Only peptides and proteins with high confidence (false discovery rate, <1%) are reported.

Antimicrobial activity.

The bacteria L. monocytogenes ATCC 19115, S. aureus ATCC 25923, E. coli ATCC 25922, and P. aeruginosa ATCC 27853 were inoculated in Muller-Hinton broth until they reached an OD₆₀₀ of 0.01 (∼10⁶ CFU/ml). A dilution series of CFS and ACD, at the desired concentration, was inoculated into 50 μl of culture medium in a 96-well plate, followed by incubation with 150 μl of the bacterial solution for 12 h (61). The antibacterial activity was monitored by measuring the absorbance at 600 nm in a microplate reader (24). For P. aeruginosa, determination of the antimicrobial activity was carried out according to two different qualitative methods: the OD₆₀₀ and fluorescence methods. For this purpose, a stock of FDA (10 mg ml⁻¹) in acetone was diluted 1:100 in phosphate-buffered saline and used as a working solution. After incubation of the bacteria for 12 h, 4 μl of working solution was added to each well, followed by incubation at 30°C for 20 min. The fluorescence emission intensity was then measured using a fluorescence microplate reader with excitation and emission wavelengths of 485 and 520 nm, respectively. In each experiment, the control contained a bacterial solution without peptide.

Effect of ACD on virulence factors (pyocyanin, pyoverdine, protease and elastase).

P. aeruginosa cells in mid-log phase were collected by centrifugation and inoculated into the medium until the OD₆₀₀ reached 0.1. The purified peptide was then diluted in 50 μl of medium in the range of 0 to 20 μg/ml, mixed with bacterial suspension, and incubated for 24 h at 37°C. Cell-free culture fluids were used to measure PCN with Epoch microplate spectrophotometer in the range of 200 to 800 nm, and 690 nm was considered for calculation (35, 40).

The production of pyoverdine was also quantified by pyoverdine-specific fluorescence (excitation, 400 nm; emission, 470 nm) using a plate reader (35). Protease production was also measured by six-well plate agar (5 g liter⁻¹ LB, 4% [wt/vol] skim milk powder, 15 g liter⁻¹ agar). Next, 2 μl of bacterial culture was placed on the agar center, and the plates were incubated at 37°C for 20 h. Protease production was determined by the resulting halo with ImageJ software and calculated using the following formula (35, 62):

$$\text{protease production}\hspace{0.17em}=\hspace{0.17em}\frac{\text{diameter (halo)}-\text{diameter (colony)}}{\text{diameter (colony)}}$$

The analysis of elastase activity was conducted as previously described (with semimodification) (63). Bacterial culture was performed in 24-well plates at 37°C for 24 h in culture medium (25% tryptic soy broth [TSB], 5% peptone). After centrifugation, the supernatants were separated, and 500 μl of supernatant was mixed with 500 μl of buffer (200 mM Tris-HCl [pH 8.8]) containing 10 mg of elastin Congo red, followed by incubation for 4 h at 37°C. After centrifugation, the absorbance was read at 495 nm.

Resistance to proteolytic digestion and physiological conditions.

The antimicrobial activity under different conditions against P. aeruginosa was evaluated by a fluorescent microplate assay, as described above. Different concentrations of ACD were treated with trypsin and proteinase K in 50 mM Tris-HCl buffer (pH 7.4) at 37°C for 2 h. In addition, the peptide activity at MIC₅₀ was evaluated under different physiological conditions, including salt, human serum (HS), human plasma (HP), FBS, and acidic pH conditions (23).

Eradication of biofilm.

Inoculation of P. aeruginosa overnight cultures in TSB plus 0.2% glucose medium was carried out, where the initial OD was 0.1, and the plate incubated for 72 h at 37°C until a matured biofilm formed. After incubation, the culture medium was aspirated gently, and the biofilm was washed with PBS buffer to remove planktonic bacteria. The biofilm was treated with different concentrations of peptide and then incubated for 1 h at 37°C. The supernatant was then poured out, and the biofilm washed with PBS buffer (19).

Crystal violet staining.

Determination of biofilm biomass was performed by crystal violet staining (19).

FE-SEM analysis.

FE-SEM was used to observe biofilm morphology. Biofilms on coverslips were fixed in 3% glutaraldehyde for 3 h, followed by successive washes with increasing concentrations of alcohol to dehydrate the slides. Finally, the samples were prepared for imaging by using a Tescan MIRA3 XMU FE-SEM.

Confocal laser scanning microscopy.

Live/dead double staining of the P. aeruginosa biofilm was performed with FDA and PI. FDA staining was performed as described above. After the addition of the FDA and after 20 min, PI was added. PI staining was performed with a 1-mg/ml solution in water, so that the concentration of the PI in each sample was 5 μM. Subsequently, the solution was incubated at 30°C for 5 min, and the surfaces of the biofilms were imaged by CLSM. Images were also processed using NIS-Element viewer 4.5 imaging software. ImageJ software was used to measure the live/dead ratio (62).

Hemolysis assay.

Fresh human blood was diluted 1:10 with an isotonic solution (0.9% [wt/vol] NaCl). The blood cells were separated by centrifugation (at 1,000 × g for 10 min) and washed three times with saline solution, dissolved in the same volume of buffer. Next, various concentrations of ACD at a ratio of 1:1, in a final volume of 200 μl, were retained in close proximity to the blood cells at 37°C for 2 h. The multiwell plate was then centrifuged, and 100 μl of supernatant was transferred to another 96-well plate. The release of hemoglobin, based on the hRBC lysis, was examined at a wavelength of 540 nm with a microplate reader. Positive and negative controls for hemolysis, respectively, included hRBCs lysed with 1% Triton X-100, as fully lysed cells, and RBCs in saline solution (61). The percent hemolysis was calculated according to the following formula:

$$\%\text{hemolysis}=\left(\frac{A_{540}\left[\text{test sample}\right]-A_{540}\left[\text{negative control}\right]}{A_{540}\left[\text{positive control}\right]-A_{540}\left[\text{negative control}\right]}\right)\times 100$$

TI calculation.

The TI is a ratio of the drug concentration that causes adverse effects to the dosage that leads to the desired effect. In this study, the TI was calculated according to the following formula (46):

$$\text{TI}=\frac{\text{toxic dose in}50\%\text{of subjects}({\text{TD}}_{50})}{\text{efficacious dose in}50\text{\% of subjects}({\text{ED}}_{50})},$$

where the TD includes cell hemolysis and the ED is also considered based on both optimal peptide activities, including PCN and cell growth inhibition. Thus, two different TIs were reported based on these parameters. In cases where the highest concentration of the peptide did not affect 50% of the population, the test dose was doubled and, accordingly, the TI calculated based on this value was also doubled (47).

Kinetics of membrane permeability.

The degradation of the bacterial membrane was evaluated with live/dead double staining as described above using FCM (26). P. aeruginosa cells (∼10⁶ CFU/ml) were incubated with acidocin at 2× and 4× MICs at 37°C for 1 h. Samples were taken at 0-, 15-, 30-, and 60-min intervals and stained. The percentage of stained cells was calculated using FCM analysis (with a FACSCalibur flow cytometer) with a standard filter setup. FDA green fluorescence and PI red fluorescence were assessed, respectively, using FL1 and FL3 channels. Finally, data analysis was performed by using the Windows Multiple Document Interface computer program (FCS Express 6 flow cytometry software, Flows Software, version 2.4.1; FlowJo Software).

TEM analysis.

P. aeruginosa cells (∼10⁶ CFU/ml) were incubated with ACD at the MIC level at 37°C. Cells were separated after incubation at different time intervals using centrifugation (6,000 × g, 10 min), washed in PBS buffer, and fixed in 2.5% glutaraldehyde. Further steps regarding sample preparation were performed according to a previously published protocol (64).

Molecular modeling and molecular dynamic simulation.

A three-dimensional model of peptide was obtained by using the QUARK webserver (65), and the model was evaluated using DOPE score analysis (66). To observe the initial mechanism of peptide interaction with the lipid bilayer, the CG-MD self-assembly simulation (67) was applied to a mixture of P. aeruginosa membrane lipids (POPE, POPG, and CL; molar ratio, 60:21:11) (68) and ACD peptides using the GROMACS software package (v5.1) (69). This system was simulated for 2 μs. A MARTINI force field (54) was used for simulations. Simulation was performed in which spontaneous self-organization of lipids occurred in the presence of peptides. The ratio of peptides and lipid beads was considered about 0.03. The software packmol (70) was used to generate a starting configuration in a cubic box with a 25-nm edge in each dimension. Then, the system was filled with a coarse-grained model of water molecules that is considered 10% polarizable water. To neutralize the system, 160 Na⁺ ions were added to the box and according to the pH of the working buffer (pH ∼5), the charge of −2 was considered for Cl (chlorine) (71). The water geometry was constrained with the SETTLE algorithm (72). Pressure coupling was applied anisotropically, using a Parrinello-Rahman barostat with a reference value of 10⁵ Pa (73). A 50-ns equilibration CG-MD simulation was performed in an NPT ensemble at 300 K, followed by 2,000-ns production runs. Running the simulation for a sufficient time led to the occurrence of bilayer self-assembly and peptide optimization regarding both single and complex positions in the lipid bilayer and interaction with surrounding lipids. Hydropathy plots of peptides were determined by using the ExPASy ProtScale tool (74). Figures and plots were obtained using YASARA simulation software (75).

Mouse model of P. aeruginosa infection.

Twelve 2-week-old male C57BL/6 mice (weighing ∼20 g) were obtained from Pasteur Institute of Iran and assigned to four groups consisting of three mice each. Three groups were inoculated intranasally with 20 μl of bacterial culture including 10⁷ CFU ml⁻¹ after anesthesia with a ketamine-xylazine mixture. A group without infection was considered the control group. At 1 h after inoculation, two groups of intranasally infected mice were treated with ACD at concentrations of 6 and 12 mg kg⁻¹, and an untreated group was considered an infectious group. The mice were monitored over time to scrutinize their survival after infection. Mice that did not show evidence of animal distress were considered survived mice. At 24 h after starting a course of infection, the mice were euthanized, and their isolated lungs were fixed in 10% formalin. The fragments of the mouse lung samples were stained with H&E and PAS and then blindly evaluated and scored by a pathologist. Animal studies were conducted according the “Guide for the Care and Use of Laboratory Animals in NIGEB Institute.”

DNA sequencing of the gene encoding ACD.

Bacterial cells were collected in the logarithmic phase of growth, and RNA extraction was achieved using RNAX-Plus solution. After quantification of RNA quality by using electrophoresis and NanoDrop, the total RNA was treated with DNase to eliminate DNA contamination. Then, 1 μg of RNA was used to synthesize cDNA according to the factory production control test protocol (RevertAid Thermo Scientific). PCR of the acidocin gene was performed using primers designed based on whole-genome sequencing available in the NCBI database (GenBank accession no. JRUT01000019.1). The sequencing was achieved according to the Sanger method (Pishgam Biotech Co.).

Data availability.

All data collected and reported in this study, including SDS-PAGE gel, LC-MS/MS, FCM, and histology examinations, are available upon request from the corresponding authors. The mass spectrometry proteomics data have been deposited in the ProteomeXchange Consortium via the PRIDE (76) partner repository under the data set identifier PXD011556.