Antimicrobial activity of Epsilon-Poly-l-lysine against phytopathogenic bacteria

Antimicrobial peptides (AMPs) are components of immune defense in many organisms, including plants. They combat pathogens due to their antiviral, antifungal and antibacterial properties, and are considered potential therapeutic agents. An example of AMP is Epsilon-Poly-l-lysine (EPL), a polypeptide formed by ~ 25 lysine residues with known antimicrobial activity against several human microbial pathogens. EPL presents some advantages such as good water solubility, thermal stability, biodegradability, and low toxicity, being a candidate for the control of phytopathogens. Our aim was to evaluate the antimicrobial activity of EPL against four phytobacterial species spanning different classes within the Gram-negative phylum Proteobacteria: Agrobacterium tumefaciens (syn. Rhizobium radiobacter), Ralstonia solanacearum, Xanthomonas citri subsp. citri (X. citri), and Xanthomonas euvesicatoria. The minimum inhibitory concentration (MIC) of the peptide ranged from 80 μg/ml for X. citri to 600 μg/ml for R. solanacearum and X. euvesicatoria. Two hours of MIC exposure led to pathogen death due to cell lysis and was enough for pathogen clearance. The protective and curative effects of EPL were demonstrated on tomato plants inoculated with X. euvesicatoria. Plants showed less disease severity when sprayed with EPL solution, making it a promising natural product for the control of plant diseases caused by diverse Proteobacteria.

www.nature.com/scientificreports/ of the lipopolysaccharide layer in cell membranes, compromising their physical integrity, and causing the extravasation of cellular content 18 . Epsilon-Poly-l-lysine (EPL) is a linear homopolypeptide, generally composed by 25 to 35 identical l-lysine residues with a unique structure characterized by the peptidic bond of lysine monomers to gamma-amino functional groups and carboxyl groups 19 . The homopolymer is biodegradable, water soluble, non-toxic and highly stable at high temperatures. EPL has wide antimicrobial activity, dependent on the molecular weight of the peptide 20 . Its application against several animal pathogens is well-documented [21][22][23][24][25] and it has been used as a food preservative since 1980 and considered safe for human consumption 26 . However, there is a lack of studies that demonstrate its efficacy against phytopathogenic bacteria and the possible control of plant diseases by topical application. The aim of this work was to investigate the in vitro antimicrobial potential of EPL against the phytopathogens R. solanacearum, X. euvesicatoria, X. citri and A. tumefaciens and to verify the in vivo action in the control of bacterial spot on tomato plants.

Results
epL minimum inhibitory concentration and bacterial clearance dynamics. The antimicrobial effect of EPL was first evaluated through in vitro spotting assays. The minimum inhibitory concentration (MIC) was determined based on the lowest concentration of the peptide solution that was able to inhibit any bacterial growth (Fig. 1). EPL efficiently inhibited the growth of all four tested bacterial species. The MIC of EPL varied between 80 μg/ml to inhibit X. citri growth, 400 μg/ml for A. tumefaciens and 600 μg/ml for R. solanacearum and X. euvesicatoria. In order to assess a putative bactericidal effect, further investigation was conducted. Incubating the cells with the determined MICs of EPL confirmed the antibacterial activity of the peptide in the first 30 min, when the quantity of colony forming units (CFUs) in the treatment group was already statistically different from the water-treated control (Fig. 2). The peptide efficiently inhibited the growth of colonies after 30 min of EPL treatment with a reduction in bacterial growth close to 100%. After two hours, the reduction remained, confirming the bactericidal activity of EPL (Fig. 2).
The effect of EPL on cell surface integrity and on cell viability of phytopathogenic bacteria. After confirming the growth inhibition effect of EPL on all four bacterial species, we further investigated its effect on cell integrity to better understand if there was a bactericidal effect or simply growth suppression (bacteriostatic effect). To assess this, we used spectroscopy and microscopy methods. First, we measured the EPL effect on the fluorescent signal coming from cells treated with SYTO 9, a dye that binds DNA (Fig. 3). In all four bacterial species the fluorescence emission was lower in cells incubated with MIC levels of EPL compared to controls treated with water. This reduction indicates EPL is effective disrupting the cells instead of simply halting their multiplication. Next, an additional dye was used, propidium iodide, which only enters cells with damaged membranes. The combination of both dyes allows for a direct visual inspection of EPL effects, as shown in Fig. 4. Figure 1. Inhibition of bacterial growth by spotting assay. Each bacterium was grown on LB agar plates and different concentrations of the peptide (70-1,000 μg/ml) were tested in different spots to determine the minimum inhibitory concentration of EPL able to inhibit bacterial growth. Water was used as negative control. The plates were placed at 28 °C for two days and evaluated. Three biological replicates were done. www.nature.com/scientificreports/ Fluorescence microscopy images showed that cells emitted red fluorescence due to interaction of propidium iodide with DNA. In the control group without EPL, the number of cells that emitted green fluorescence from intact membranes was considerably higher (Fig. 4). This result suggested that EPL-treated cells had their cell membrane damaged by the peptide.
To assess membrane integrity and bacterial morphology in finer detail, the bacterial cells were visualized under scanning electron microscopy (SEM) after 1 h of treatment with EPL at the MIC (Fig. 5). The SEM images revealed the rupture of the cytoplasmic membrane and the extravasation of cellular content due to the membrane integrity disruption effect of EPL. Compared with treated cells, the surface of untreated cells was bright and smooth. Taken together these results provide further support that EPL is actively damaging the bacterial cells and not just simply arresting their proliferation. epL protects tomato plants against bacterial spot disease. After attesting the in vitro antimicrobial activity of EPL, we performed in vivo assays in tomato plants infected with X. euvesicatoria to establish the potential of this peptide to control bacterial diseases by topical application.
Twenty days post-inoculation of the pathogen the initial symptoms such as spots and irregularly shaped watery patches on the leaves started appearing in all 12 control plants treated with water only. Three out of twelve plants that were treated preventively with EPL (at MIC level) developed initial symptoms. Interestingly, plants that were treated preventively with twofold MIC did not show any symptoms. At thirty days post-inoculation the disease was more severe in the control group (yellowish and bottom leaves death), and some plants were classified as level 5 according to the disease scale proposed by Mello et al. 27 . Plants that were treated preventively with EPL (onefold MIC) developed some disease symptoms, however no symptom was observed on plants that were treated preventively with twofold MIC, demonstrating the long-lasting effect of preventive EPL treatment (Fig. 6a).
We also tested the potential curative effect of EPL after X. euvesicatoria infection. Plant recovery, however, was not as robust as prevention with EPL. When EPL was applied two days after X. euvesicatoria infection all the plants showed initial symptoms of bacterial spot (onefold MIC and twofold MIC), however, by the end of the experiment, no plants were classified as level 5 according to the disease scale, meaning that EPL treatment can slow down the disease progress or alternatively reduce its severity (Fig. 6b).

Discussion
Antimicrobial peptides are widespread in nature, among competing microbial communities and also among plant and animal hosts as a means to fight off infections. They hold great potential in agriculture, since phytopathogenic bacteria are still problematic to control due to lack of effective bactericides and the emergence of resistance.
Here we evaluated the potential use of EPL to control some of the most devastating bacterial plant pathogens. Its effectiveness against bacterial pathogens in mammalian and safety for human consumption has been previously confirmed, making it a promising candidate for agricultural applications. A pioneering study revealed EPL to be nontoxic in a dosage level of 10,000 μg/ml and nonmutagenic in usage-relevant concentrations in a two-generation reproduction study using rats 28 . This value is 125-fold higher than what we determined here to be necessary to control X. citri, 25-fold higher than necessary to control A. tumefaciens, and around 17-fold higher than the MIC to control R. solanacearum and X. euvesicatoria, again reinforcing its safety at effective levels for bacterial clearance. Zhao et al. studied the effect of EPL and Epsilon-caprolactone (CPL) copolymer-based nanoparticles on mammalian cells viability using human breast carcinoma and human umbilical vein endothelial cells as models, and reported no apparent cytotoxic effects for both cell types even at the concentration of 1,000 μg/ml 29 , but studies on acute toxicity are also required before recommendation for commercial application. Some countries including Japan, South Korea and Canada have already regulated EPL use as a natural preservative for the food industry 19,30 . In 2004, the US Food and Drug Administration has given EPL GRAS status (generally recognized as safe), and approved its use as an antimicrobial agent in cooked or sushi rice at levels up to 50 mg/kg of rice 31 .  www.nature.com/scientificreports/ We demonstrated that the in vitro antibacterial activity of the peptide was able to inhibit growth of all four tested phytobacteria. These results are correlated with loss of cellular viability verified by SEM and fluorescence assays and encouraged further investigation on its therapeutic potential in vivo. Previous studies have shown great variability of microbial sensitivity to EPL, with Gram-negative bacteria consistently showing higher sensitivity compared to Gram-positive. EPL was tested against the Gram-negative Escherichia coli and Gram-positive Listeria innocua, for which the MIC was defined as 74 µg/ml for E. coli and 750 µg/ml for L. innocua 32 . To understand these differences, it is necessary to correlate the mechanism of action of the peptide, the membrane and cell wall composition of the pathogens, and the phospholipid stoichiometry of the different microbial targets 33,34 . The kill curves obtained in this study confirmed the effective and fast bactericidal effect of EPL. Our results are in accordance with other studies testing AMPs. For example, the mortality ratio of E. coli O157:H7 treated with different concentrations of EPL ranged from 5 to 50 µg/ml, and after 15 h was higher than 95% 25 .
Bacterial cell death is defined as the incapacity of the cell to grow as a viable colony in bacteriological media. However, there are different ways to evaluate cell viability without culturing cells. One of them is based on fluorescent probes that can be detected through fluorescence microscopy and spectroscopy 35 . Fluorescence detection is a fast and reliable technique that is valid to quantify bacteria of different genera 36 . A known effect of EPL on bacterial cells is the damage of the cell wall, compromising cellular integrity. Cell viability can be monitored using fluorescent dyes that differ in spectral characteristics and in the ability to penetrate bacterial cells. This allows reliable quantitative distinction between bacteria with intact or damaged citoplasmic membrane, consequently differentiating living and dead cells 35 . The results obtained through fluorescence spectroscopy and SEM are in accordance with the initial results that showed reduced colony formation following incubation with EPL. The predominance of red-fluorescent cells confirms that the growth inhibition is due to the bactericidal effect of the peptide and not only a bacteriostatic effect. In order to study the effects of EPL on S. aureus cell membrane, Tan et al. showed that after EPL treatment there was a remarkable increase in fluorescence intensity measured by propidium iodide assay indicating an increase in cell membrane permeability 37 . When S. aureus cells were treated with 250 μg/ml of EPL, the cells appeared collapsed, lysed and with non-integral cell morphology 37 . Similar results were obtained after treating E. coli cells with 50 μg/ml of EPL for 4 h. They observed that the external membrane and cytoplasm were damaged, surrounded by cell debris and with wrinkled appearance 25 . In another study the interaction between S. aureus and B. subtilis cells with nanoparticles composed by EPL and CPL were evaluated 29 . They observed the effect of EPL treatment on cell structure by SEM and similar to our findings revealed cells broken in appearance, with rupture of cell wall and membrane, lysis of cellular content, and extravasation of cytoplasm. All these observations confirm that EPL can adsorb on the surface of the microbe membrane resulting in physical damage to the cell. The previous reports showed that EPL can affect the cell membrane permeability and compromise their viability 25,37,38 . The EPL mechanism of action assures that the microorganism does not easily develop resistance.
Despite its in vitro efficacy, it is important to understand how EPL performs in vivo under more realistic conditions encountered in the field. In this study, tomato plants that were sprayed with EPL before bacterial infection were protected against bacterial spot disease. Some spots were observed in some leaves, but the disease did not fully develop. When a higher concentration was applied (twofold MIC), no symptom was observed. Therefore, our data shows the higher effectiveness of EPL when it is applied as a preventive method at higher dosages such as twofold MIC. However, we also observed symptom reduction when EPL was applied after bacterial infection, even at MIC level.
Here we concluded that EPL, even at low concentrations, has significant in vitro antimicrobial activity against diverse phytobacteria, attesting to its broad range of activity towards microbial cells. It is known that EPL is biodegradable, non-toxic, resistant to thermal degradation, and possess antimicrobial activity against a wide spectrum of microorganisms. Our results confirm that EPL is a promising alternative to control phytobacteriosis prophylactically. In light of these findings, more investigations will determine the optimal method for the application of this peptide in agricultural contexts, and the effect of this peptide in the pathosystem and phytosphere microbiome in general.

Minimum inhibitory concentration assay.
To obtain the minimum inhibitory concentration (MIC) of EPL necessary to prevent bacterial growth, a spotting assay 39,40 was performed. Briefly, bacteria were cultured in LB broth at 28 °C, 200 rpm, for 12 h. Ten microliters of these bacterial suspensions were diluted in LB to 100 μl and spread onto LB agar plates. Ten microliters of EPL solutions at different concentrations (70-1,000 μg/ml) were spotted on the agar plates that had previously received the bacterial suspensions. Plates were then incubated at 28 °C until formation of clearance zones where growth inhibition could be clearly seen. MIC values were registered as the least concentration of EPL that would inhibit bacterial growth. Distilled water was used as negative control spotted in the center of each plate. All assays were performed in biological triplicates. www.nature.com/scientificreports/ medium (~ 10 6 CFU/ml). EPL was added to each bacterial suspension at the previously determined MICs and these were incubated for two hours at 28 °C and 200 rpm. Aliquots were taken at 0, 30, 60, 90, and 120-min intervals, serially diluted with LB broth and plated. Plates were kept for two days at 28 °C and the number of CFUs was used to determine the efficiency of EPL in clearing the pathogen. Three biological replicas were performed.

Kill-curves.
fluorescence spectroscopy and microscopy. Bacterial viability after EPL treatment was assessed by fluorescence emitted by EPL-treated compared to non-treated bacterial cells. Initially log-phase bacterial cultures were adjusted to OD 600nm 0.1, and centrifuged at 10,000×g for 15 min. Supernatant was removed and pellets suspended in EPL solutions at MIC or distilled water control. Cells were incubated at 28 °C for one hour when 1 µl of SYTO 9 from the Live/Dead BacLight bacterial viability kit (Life Technologies) was added to 1 ml of the bacterial suspensions. The samples were then incubated in the dark for 15 min and fluorescence measured using flat bottom black plates and fluorimeter (PerkinElmer). Samples were excited at 470 nm and emission spectrum (490-700 nm) recorded. Three biological replicas were performed, and two technical replica readings of each sample were taken. Controls included water-only treatment (non EPL-treated) and EPL + fluorophores (without bacterial cells). Fluorescence microscopy was used to visualize viable and non-viable cells stained with 3 µL of a 1:1 mixture of SYTO 9 and propidium iodide from the Live/Dead BacLight kit components. Cells were prepared as described above and imaged on an Evos FL fluorescence microscope (ThermoFisher), in which viable cells fluoresce in green and non-viable cells in red.
Scanning electron microscopy. Morphological  in vivo epL protection assay. Tomato plants (Solanum lycopersicum L.) cv. Santa Cruz Kada were grown in the greenhouse at 25 ± 2 °C to the V 2 developmental stage 41 . Xanthomonas euvesicatoria EH 2009-130 was cultured in LB medium and adjusted to 10 7 CFU/ml with NaCl 0.85% (w/v). The bacterial suspension was sprayed on the tomato leaves until dripping. Plants were kept in a humid chamber made with a transparent plastic bag cover for 24 h before and after inoculation to facilitate infection. To test protective and curative effect of EPL to bacterial infection, onefold and twofold MIC diluted in water plus 1% Pentra-Bark surfactant were sprayed on entire tomato plant surfaces (twelve plants for each treatment). Water was used as negative control. Applications were done two days before inoculation with X. euvesicatoria (prophylactic) or two days after inoculation (curative). Progress of symptom development was recorded as visual appearance of leaf spots until 30 days postinoculation, and scores were given from 1 (1% of leaf area affected) to 5 (50% or more of leaf area affected), according to the diagrammatic scale proposed by Mello et al. 27 .
Statistical analysis. The necessary assumptions required for the analysis of variance (ANOVA) were verified. Normality of errors and variance of homogeneity were evaluated with Shapiro-Wilk and Levene tests. Killcurve assays were analyzed in a split-plot arrangement placing the EPL treatment and the incubation time in the main and sub plots, respectively. Complex variances were applied when significant interactions were observed. Averages of peptide treatment and incubation periods were compared by Tukey test and polynomial regression, respectively. All analyses were done considering significance of 0.05.