Antimicrobial Characterization of Erythorbyl Laurate for Practical Applications in Food and Cosmetics

Department of Agricultural Biotechnology, Seoul National University, Seoul 08826, Republic of Korea Department of Food Science and Biotechnology, Wonkwang University, Iksan 54538, Republic of Korea Center for Food and Bioconvergence, Seoul National University, Seoul 08826, Republic of Korea Research Institute of Agriculture and Life Sciences, Seoul National University, Seoul 08826, Republic of Korea


Introduction
Lipid oxidation of food and cosmetics has been considered as a major hazard for consumer health [1]. Moreover, microbial contamination influences the physical and chemical properties of food and cosmetics [2][3][4]. erefore, preservatives have been employed to control the lipid oxidation and the microbial contamination in food and cosmetic industry [5]. Previously, erythorbyl laurate (EL) was suggested as a novel multifunctional emulsifier with antioxidative and antibacterial activity to control the lipid oxidation and microbial contamination with a single compound [6].
A previous study investigated EL had antibacterial activity against Gram-positive food-borne pathogens including Staphylococcus aureus, Bacillus cereus, and Listeria monocytogenes.
e minimum inhibitory concentrations (MICs) and minimum bactericidal concentrations (MBCs) of EL against these Gram-positive bacteria were determined, and the antibacterial mechanism was thought to be disruption of the bacterial cell membrane [7]. In addition, it was investigated EL exhibited more effective antioxidative activity than that of erythorbic acid in oil-in-water (O/W) emulsion system since its antioxidative moiety is located on the interface of droplets where lipid oxidation mostly occurs [8]. However, no study has examined the antibacterial properties of EL in an O/W emulsion, which is thought to be the best system for application of EL. Food and cosmetics are made at various pH values, which affect the antimicrobial activity of preservatives, caused by structural changes [9]. In addition, structure of EL might be changed once it begins to act as an antioxidant. It was reported that ranalexin, an antimicrobial peptide, retained its antimicrobial activity after being oxidized, while catechins with antioxidative activity exhibited antimicrobial effects when functioning as an antioxidant [10,11].
Antimicrobial spectrum of an antimicrobial agent is one of the crucial considerations for practical applications since microbial contamination of food and cosmetics can be caused by bacteria, yeasts, and molds [12]. Moreover, increasing concerns about bacterial resistance to antibiotics have prompted the requirement to search for alternative antibacterial agents for clinical application, as well as use in food and cosmetics. Since there are no reports of bacteria developing resistance to fatty acids, EL may not induce bacterial resistance [13,14].
Fundamental investigation on antibacterial properties of EL with regard to antibacterial activity and mode of action was carried out in the previous study [7], but no study has been performed to bring detailed knowledge of antimicrobial properties of EL for the practical applications. erefore, the aim of this study is comprehensive evaluation of antimicrobial properties of EL for the practical applications in food and cosmetics. e antimicrobial spectrum of EL against bacteria, yeasts, and molds was investigated, and resistance study was performed to figure out bacterial resistance against EL. Furthermore, the effects of structural changes in EL due to protonation by pH variation and oxidation on its antibacterial activity were evaluated. Finally, the antibacterial activity of EL in an O/W emulsion system was investigated to assess its applicability to emulsion-based foods and cosmetics.

Production of Erythorbyl
Laurate. EL was produced in gas-solid-liquid multiphase system [15]. e esterification between erythorbic acid and lauric acid was catalyzed by the immobilized lipase (Novozym 435) for 95 h. e total volume was 450 mL, and the erythorbic acid (0.97 mol) to lauric acid (1.93 mol) molar ratio was 1 : 2. e enzyme amount was 840 PLU/mL (0.12 g/mL), and nitrogen gas was bubbled at a flow rate of 6.0 L/min. e lauric acid was preincubated at 60°C for 20 min on the reactor to be melted.
Quantitative analysis of the EL was conducted using high-performance liquid chromatography (HPLC; LC-2002; JASCO, Tokyo, Japan) with a silica-based column (5.0 μm, 4.6 × 150 mm; Luna C18; Phenomenex, Torrance, CA, USA) and an ultraviolet (UV) detector (UV-2075; JASCO). e mobile phase was acetonitrile/water/acetic acid (90 : 5:5, v/v/v) at a 1.0 mL/min flow rate. Peaks in the HPLC chromatograms were confirmed according to the retention time of an EL standard. e EL concentration was determined using the standard curve of EL, generated by integrating the peak area at 265 nm using Borwin (ver. 1.21; JASCO). S. pyogenes ATCC 19615 was cultured in tryptic soy agar (TSA) at 37°C for 24 h in an atmosphere of 5.0% CO 2 . C. perfringens ATCC 13124 and P. acnes ATCC 6919 were cultured in TSA at 37°C in an anaerobic atmosphere for 1 and 5 days, respectively. All molds were cultured in potato dextrose agar (PDA) at 25°C for 5 days, and the yeast was cultured in yeast malt agar (YMA) at 25°C for 2 days.

Disk Diffusion Assay.
e antimicrobial susceptibility to EL was assessed using disk diffusion assays with reference to the Clinical and Laboratory Standards Institute (CLSI) guidelines. Inocula of Gram-positive bacteria and yeast were made by direct colony suspension in tryptic soy broth (TSB) and yeast mannitol broth (YMB), respectively, equivalent to 0.5 McFarland suspension, and molds were inoculated (5 × 10 4 spores/mL) in potato dextrose broth (PDB). e bacterial and fungal suspensions were inoculated on the surfaces of agar plates. Gram-positive bacteria, molds, and yeast were inoculated on the surfaces of TSA, PDA, and YMA, respectively. EL, a negative control (TSB, PDB, and YMB with 2.0% dimethyl sulfoxide), and antibiotics (10 ppm ampicillin and 250 ppm amphotericin B) were prepared. Paper disks (8.0 mm thickness) were laid on the surface of the agar, and 40.0 μL samples were dispensed on the disk; the plate was then dried for 5 min. e plates inoculated with S. pyogenes ATCC 19615 and C. perfringens ATCC 13124 were incubated at 37°C, anaerobically for 1 day, and that with P. acnes ATCC 6919 was incubated for 5 days. e plates inoculated with molds were incubated at 25°C for 2 days and T. mentagrophytes ATCC 18748 was incubated for 7 days; the yeast was incubated at 25°C for 2 days, and the clear zones were then analyzed [16][17][18].

Structural Changes in EL.
e degree of protonation of EL according to the pH was predicted using Chemicalize. org. EL was dissolved in 2.0% dimethyl sulfoxide in 100 mM universal buffer at pH 5.0, 6.0, or 7.0 to protonate the EL.
At neutral and alkaline pH, ascorbic acid is highly unstable due to its much faster conversion to dehydroascorbic acid. Dehydroascorbic acid also degrades more rapidly at alkaline pH (7.0-8.0) than at acid pH (3.0-5.0) [19]. Similarly, EL might also be oxidized and degraded at alkaline pH. erefore, EL was added to 5.0% dimethyl sulfoxide in 50 mM Tris-HCl buffer at pH 8.0 for oxidation. e degree of oxidation was presented as the mean of decrement in the reduced form of EL measured by HPLC in triplicate.

Evaluation of Antimicrobial Activity.
To evaluate the antimicrobial activity of EL according to the degree of protonation, a 100 μL inoculum of S. aureus ATCC 12692, L. monocytogenes ATCC 19115, or B. cereus ATCC 10876, equivalent to 0.5 McFarland suspension, was added to 100 μL universal buffer containing EL. e colony-forming units (CFUs) in a 100 μL aliquot were enumerated by preparing serial dilutions in 100 mM universal buffer at pH 5.0, 6.0, and 7.0; plating was performed in triplicate for each dilution with TSA plates, for 1 h after incubation at 37°C for 1 h. en, each plate was incubated at 37°C for 18 h. e antimicrobial activity of EL according to the degree of oxidation was evaluated using the MIC test after storage for 0, 6, 15, 24, and 48 h at 40°C. e reduced form of EL was quantified by HPLC in triplicate. For the MIC test, Tris-HCl buffer containing EL was added to 1.25% dimethyl sulfoxide in TSB. S. aureus ATCC 12692 was tested to determine the MIC of the oxidized form of EL, and the bacterial inoculum was diluted in 2.00% dimethyl sulfoxide in TSB as 0.5 McFarland standard. e MIC was determined using the broth microdilution assay. Serial dilutions of each desired concentration of oxidized EL were prepared in sterile TSB to a final volume of 100 μL in 96-well microplates. en, each well was inoculated with 100 μL of the test organisms in TSB. e MIC was obtained as the lowest concentration at which the test compound inhibited bacterial growth after incubation for 12 h at 37°C [20,21].

Multipassage Resistance Selection
Study. An inoculum was prepared as 0.5 McFarland suspension after incubating S. aureus ATCC 12692 in TSB at 37°C for 18 h. Serial passages were performed every 18 h in TSB (2.0% dimethyl sulfoxide). e strain was treated with a 2-fold dilution series of EL in 200 μL broth in 96-well plates. For each subsequent passage, a 2 μL aliquot was taken from the wells with concentrations below the MIC that matched the turbidity of the growth control and was used to inoculate the dilution series for the next passage. Every passage was performed until 20 consecutive passages were completed [22].

Preparation of the Emulsion and
Inoculum. An O/W emulsion containing EL was formulated using sunflower oil, Tween-20, EL, and sterilized water. e concentrations of sunflower oil (5.0%, w/w) and Tween-20 (0.5%, w/w) were fixed for all emulsion formulations. e emulsion was prepared by mixing the oil and Tween-20 with or without EL (5.6 mM) followed by the addition of sterilized water. e emulsion was then sonicated with Sonomasher (ULH-700S; Ulso High-Tech Co., Cheongwon, Korea) at 4°C for 20 min at 210 W using pulses of 10 s on followed by 30 s off [23].
Bacteria related to food-borne and infectious skin diseases were selected. e three Gram-positive bacteria tested were S. aureus ATCC 12692, L. monocytogenes ATCC 19115, and B. cereus ATCC 10876; the two Gram-negative bacteria tested were E. coli ATCC 35150 and P. aeruginosa ATCC 10145. ese bacteria were incubated in TSB at 37°C for 18 h. en, cultures of the bacteria were diluted in 10 mM phosphate-buffered saline (PBS) at pH 7.4 as a 0.5 McFarland standard.

Bactericidal Assay.
e MBC was determined as the standard criterion for evaluating the antimicrobial effect with some modifications, to determine the effective concentration of EL in the O/W emulsion [24]. Serial dilutions of each desired concentration of emulsion containing EL were prepared in an emulsion without EL to a final volume of 180 μL in 96-well microplates. en, each well was inoculated with 20 μL of the inoculum. After incubation at 37°C for 12 h, the MBC of the emulsion containing EL against each bacterium was determined as the lowest concentration producing a 99.9% reduction of the viable bacteria count in the subcultured wells. e serially diluted subcultured well contents were spread on TSA plates, and the colonies were counted after incubation at 37°C for 24 h.

Time-Killing Assay.
e time-killing assay was performed to study the concentration and time-dependent killing effect [25]. For the O/W emulsion containing EL, a 200 μL inoculum was added to 1,800 μL of O/W emulsion containing EL. After inoculation, the solutions were incubated at 37°C with shaking (220 rpm). After incubation for 0, 1, 2, 3, or 4 h, a 100 μL aliquot was removed and serially diluted. e diluted samples were inoculated on TSA plates and incubated for 24 h at 37°C. e number of survivors (CFU/mL) was determined by counting the colonies, and time-killing curves were constructed by plotting the log CFU/mL versus time. e limit of detection in the assay was 10 CFU/mL (1 log CFU/mL). e experiments were conducted in triplicate.

Statistical Analysis.
e statistical analysis was performed using SAS software (SAS Institute, Cary, NC, USA). Experiments were conducted in triplicate. Mean separations were evaluated using Duncan's multiple-range test. A p value <0.05 was taken to indicate a significant difference.

Antimicrobial Spectrum of EL.
e three bacteria, the three molds, and the one yeast, causing food-borne diseases and infectious skin diseases, were selected for the evaluation of the antimicrobial spectrum of EL (Table S1 in Supplementary data) [26][27][28][29][30][31].
In the disk diffusion assay, EL showed antimicrobial activity against the three Gram-positive bacteria, one yeast, and three molds. P. acnes, S. pyogenes, C. perfringens, A. nidulans, R. oryzae, and T. mentagrophytes were susceptible to 3 mM EL and C. albicans to 6 mM EL (Table 1). ese results demonstrated that EL can be applied to food and cosmetics to control the molds, yeast, and Gram-positive bacteria that lead to food-borne diseases and infectious skin diseases.

Effect of pH.
ree Gram-positive bacteria were used to examine whether the degree of protonation of EL altered its antimicrobial activity (Table 2). When S. aureus ATCC 12692 was treated with 0.5 mM EL, no significant reduction in viable cells was seen at pH 7.0. By contrast, 4.11 ± 0.28 and more than 6.89 ± 0.20 log CFU reductions in viable cells were observed at pH 6.0 and 5.0, respectively. A similar result was obtained for L. monocytogenes ATCC 19115. ere was no significant reduction in viable cells at pH 7.0, but reductions of 7.02 ± 0.18 and 7.04 ± 0.20 log CFU were seen at pH 6.0 and 5.0, respectively. For B. cereus ATCC 10876, there was no significant difference in log CFU reduction by pH variation. e results might be caused by surviving spores of B. cereus, spore-forming bacteria, germinated [32].
At acidic pH, the chemical structure of EL changes with the protonated form of EL increasing. e percentage of the protonated form of EL at pH 5.0, 6.0, and 7.0 was 22.03, 2.75, and 0.28%, respectively, because EL has a strongly acidic pK a of 4.5 according to the prediction ( Figure S1 in Supplementary Data).
Protonation of the head group of EL with decreasing pH should reduce the electrostatic repulsion between EL and the outer membrane of bacteria.
Other antimicrobial peptides with antibacterial mechanisms involving damage to the bacterial cell membrane also have greater antibacterial activity at acidic pH. Under acidic conditions, the anionicity of the peptides was decreased by the protonation of aspartic acid and glutamic acid residues of which pK a values are 3.65 and 4.25, respectively. e reduced anionicity of the peptide enhances its ability to interact with negatively charged phospholipids on the cytoplasmic membranes of Gram-positive bacteria, increasing its antibacterial activity [9]. erefore, the reduced anionicity of the head group of EL caused by protonation at acidic pH enhanced its antibacterial activity against Grampositive bacteria.
It was reported that pH values of human skin and most of the foods are all in the pH 4.2-6.5 [33]. erefore, it was demonstrated EL is adequate to be applied in food and cosmetics, since the antimicrobial activity of EL can be enhanced in acidic conditions than neutral conditions.

Effect of Oxidation of EL.
e MIC of EL against S. aureus ATCC 12692 did not increase significantly with oxidation ( Figure 1). During oxidation, 58.99% (mol/mol) of EL was oxidized, whereas its MIC did not increase significantly.
Ascorbic acid is unstable in alkaline solution, which leads to the formation of dehydro-1-ascorbic acid on exposure to active oxygen species or UV light [34]. During ascorbic acid oxidation, dehydroascorbic acid is formed by the removal of two hydrogen atoms and two electrons from ascorbic acid [35]. In addition, the lactone ring of dehydroascorbic acid induces its irreversible conversion to 2,3diketo-1-gluconic acid by degradation in alkaline solution [36]. erefore, in this study, the form of the EL head group was predicted to change during the oxidation of EL ( Figure S2 in Supplementary Data).
Although the structure of EL changed with oxidation, degree of oxidation did not affect the antibacterial activity of EL. e reason why EL retained the antibacterial activity after being oxidized could be explained in terms of hydrophilicity. It was reported hydrophilic/hydrophobic balance between the head and tail group of fatty acid esters is one of the crucial factors determining their antibacterial activity [37]. In addition, it is necessary hydrophilicity of the head group of lauric acid esters is above a certain level to possess antibacterial activity.
Hence, hydrophilicity of EL and oxidized form of EL were assessed by calculation of octanol-water partition coefficient (log P) with the atom/fragment contribution method [38]. As a result, the log p values of EL and oxidized form of EL were calculated as − 0.686 and − 1.753, respectively. e values indicated that the hydrophilicity of EL was not decreased after being oxidized, and the hydrophilicity of the oxidized form of EL is above the certain level to possess antibacterial activity. erefore, EL can retain the antibacterial activity after it functions as an antioxidant in food and cosmetics.

Antimicrobial Activity of EL in O/W Emulsion.
e MBCs of EL in the O/W emulsions against the Grampositive bacteria L. monocytogenes and S. aureus were 2 mM and less than 0.50 mM, respectively. In comparison, the MBCs of EL in the O/W emulsions against the Gramnegative bacteria E. coli and P. aeruginosa were both 2.00 mM ( Figure S3 in Supplementary Data). e previous study on the antibacterial activity of EL against bacteria found that Gram-negative bacteria were not susceptible to EL in the aqueous phase [7]. However, the result showed that EL in the O/W emulsion had antimicrobial activity against both Gram-positive and Gram-negative bacteria. Even though not fully understood, the difference in the antibacterial spectrum of EL between in the O/W emulsions and the aqueous phase could be related to negatively charged characteristics of EL as described in Section 3.2.1. e outer membrane of Gram-negative bacteria consists of lipopolysaccharide (LPS) with negatively charged phosphate groups [39]. In the aqueous phase, the electrostatic repulsion between EL and outer membrane (especially LPS) of the Gramnegative bacteria would harshly prevent EL interacting with the outer membrane.  (Figure 2). After treatment for 4 h, the numbers of P. aeruginosa, E. coli, S. aureus, and L. monocytogenes were reduced by 5.29 ± 0.24, 6.01 ± 0.18, 5.95 ± 0.13, and 6.24 ± 0.30 log CFU/mL, respectively.

Multipassage Resistance Selection Study.
Resistance of microorganisms to EL can be selected for multipassage resistance selection, which enables strains to acquire greater resistance to antibiotics [40]. e initial MICs of EL and    Figure 3). erefore, it was impossible to isolate bacteria that acquired resistance to EL. As a β-lactam antibiotic, ampicillin is an irreversible inhibitor of the transpeptidase needed to make the bacterial cell wall. Generally, S. aureus acquires resistance to ampicillin by producing penicillinase and triggering the resistance pathway [41]. In comparison, evolution of bacteria acquiring resistance to free fatty acid was less problematic than conventional antibiotics [42]. Lauric acid and esters of fatty acids such as monocaprin kill microorganisms by disrupting bacterial cell membranes; hence, the emergence of resistance is unlikely [43]. erefore, it was expected EL did not develop resistance since the antibacterial mechanism of EL is based on the damages to the bacterial cell membrane physically.

Conclusions
EL showed antimicrobial activity against Gram-positive bacteria, yeasts, and molds that cause food-borne diseases and infectious skin diseases. e protonation of EL in acidic conditions increased its antibacterial activity, while the oxidation of EL did not alter its antibacterial activity. EL exhibited bactericidal effects against both Gram-positive and Gram-negative bacteria in O/W emulsion system. A multipassage resistance study implied EL might not develop the drug resistance of bacteria. is study demonstrates EL is a promising additive to control microbial contamination in food and cosmetics, especially emulsionbased products.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest. Table S1: criteria to select microorganisms for the susceptibility tests based on potential diseases related to food and cosmetics. Figure S1: predicted chemical structures of erythorbyl laurate in various pH values. Figure S2: predicted chemical structures of erythorbyl laurate according to the oxidation. Figure