Abstract
The aim of this study was to examine the synergistic bactericidal effects of medium-chain fatty acids (MCFAs; caprylic, capric, and lauric acid) and organic acids (OAs; acetic, lactic, malic, and citric acid) against Escherichia coli O157:H7 and to identify their underlying mechanism(s) of action. E. coli O157:H7 was treated with MCFAs, OAs, or different combinations of MCFAs and OAs. Membrane damage and cell morphology were examined by flow cytometry and transmission electron microscopy, respectively. Combined treatment resulted in an additional log-unit reduction compared with the sum of the reductions obtained after individual treatment. For example, caprylic acid (1.0 mM, or 0.016%) and citric acid (1.0 mM, or 0.012%) alone showed negligible bactericidal effects (0.30- and 0.06-log-unit reductions, respectively); however, a marked synergistic effect (>7.15-log-unit reduction) was observed when the two were combined. Although flow cytometry and microscopic analyses of bacteria treated with individual MCFAs and OAs showed evidence of membrane disruption, the bacteria were still able to form colonies; thus, the cell damage was recoverable. In contrast, cells exposed to combined treatments showed clear membrane disintegration and/or cell death (irreversible damage). The mechanism underlying the antimicrobial effects of combined treatment with MCFAs or OAs may involve disruption of the bacterial membrane, which then facilitates the entry of other antimicrobial compounds into the cytoplasm. The main advantage of combined treatment with very low concentrations of natural antimicrobial compounds is that it is very cost-effective. Thus, this approach may be an alternative to more conventional antimicrobial treatments, such as those currently used in public health, medical centers, and the food industry.
INTRODUCTION
The Shiga toxin-producing organism Escherichia coli O157:H7 causes serious illnesses in humans, which are characterized by a number of clinical manifestations, including anemia, stomach cramps, bloody diarrhea, and life-threatening infectious disease (1, 2). The most severe condition caused by E. coli O157:H7 is hemolytic-uremic syndrome (HUS), which is characterized by acute renal failure and damage to the brain and other organs (3, 4). Most patients with HUS recover within a few weeks, but some (10 to 15%) suffer permanent damage or die (5). E. coli O157:H7 is highly virulent and can cause disease at an extremely low infectious dose (<10 to 100 CFU) (6). The bacterium is responsible for large disease outbreaks in many parts of the world, including the United States, Canada, the United Kingdom, and Japan (6–10). The U.S. Centers for Disease Control and Prevention (CDC) reported that E. coli O157 infections (and their associated complications) cause 73,000 illnesses/year in the United States alone, resulting in at least 2,000 hospitalizations and 60 deaths; the estimated annual cost of illness caused by E. coli O157 is $405 million (11).
Microbiological safety is a major concern for both consumers and industry, and a wide range of chemical and physical treatments has been developed to eliminate pathogens. The words “chemical compound” and “additive” provoke a negative reaction from many consumers, causing them to avoid products labeled with such words (12, 13). Instead, consumers prefer to use natural compounds rather than chemicals. Therefore, researchers have begun to examine the antibacterial effects of natural compounds with a view to using them to replace synthetic antimicrobials. In addition, an increase in the number of antibiotic-resistant bacteria means that we must identify novel natural drugs that can act as viable alternatives to the current batch of antimicrobials.
Medium-chain fatty acids (MCFAs) are a family of saturated medium-chain fatty acids that includes caprylic acid (CLA; C8:0), capric acid (CRA; C10:0), and lauric acid (LRA; C12:0) (14). They are naturally present in foods, such as coconut oil, palm kernels, cow's milk, and human breast milk (14–16), and are often used as nutritional supplements (17). MCFAs and their corresponding monoglycerides (monocaprylin, monocaprin, and monolaurin) also have antimicrobial properties and show broad-spectrum activity against a number of bacterial species (18–20). Organic acids (OAs) are natural compounds that are present in various foodstuffs and are produced by some microorganisms (21). They are widely used by the food industry as antimicrobials and preservatives to improve the microbiological safety of products and to prevent the deterioration of food (22).
Future research in the field of antimicrobials is expected to focus on the combined effects of antimicrobial compounds; this is because the use of two or more antimicrobials in combination is predicted to have a synergistic effect (23). Although MCFAs and OAs have been used to inhibit and eliminate pathogens (18–20, 22), no studies have examined their bactericidal activity when used in combination. Therefore, the objective of the present study was to examine the bactericidal effects of three MCFAs (CLA, CRA, and LRA) and four OAs (acetic acid [AA], lactic acid [LA], malic acid [MA], and citric acid [CTA]) against E. coli O157:H7. These compounds were tested both individually and in different combinations, and the integrity of the bacterial membrane and changes in cell morphology were examined by flow cytometry and electron microscopy, respectively, to identify the underlying mechanism(s) of action.
MATERIALS AND METHODS
Bacterial strains.
Three strains of E. coli O157:H7 (ATCC 35150, ATCC 43889, and ATCC 43890) were obtained from the Food Microbiology Culture Collection at Korea University (Seoul, South Korea). Each strain was stored at −20°C in tryptic soy broth (TSB; Difco, Becton, Dickinson, Sparks, MD) containing 20% glycerol. Bacteria were subcultured on a monthly basis.
Preparation of cell suspensions.
The laboratory stock cultures (100 μl) of E. coli O157:H7 were expanded by growing the bacteria overnight at 37°C in screw-cap tubes containing 10 ml TSB. Equal quantities of culture containing each bacterial strain were mixed in a plastic 50-ml centrifuge tube (Becton, Dickinson, Franklin Lakes, NJ), and the bacteria were harvested by centrifugation (Centra-CL2; International Equipment Company, Needham Heights, MA) for 15 min at 3,000 × g. The supernatant was removed, and the pellet was washed twice in 0.85% sterile saline. After centrifugation, the final cell pellet was resuspended in 0.85% sterile saline.
Bactericidal effect assays. (i) Preparation of stock solutions.
Stock solutions of each MCFA (CLA, CRA, and LRA) and each OA (AA, LA, MA, and CTA) were prepared by dissolving them in 98% ethanol or sterile distilled water, respectively. Each stock solution was vortexed at high speed for 1 min to create an emulsion, which was clear at ambient temperature. All stocks were used within 1 week of preparation. Unless otherwise stated, all the reagents used in the present study were purchased from Sigma Chemical Co. (St. Louis, MO).
(ii) Treatment with individual MCFAs or OAs.
An aliquot (0.1 ml) of each stock solution (12.5, 25, 50, 100, or 150 mM) of MCFA (CLA, CRA, or LRA) was added to 9.8 ml of sterile 0.85% saline in a sterile tube to yield a final concentration of 0.125, 0.25, 0.5, 1.0, or 1.5 mM. The tubes were then preheated at 37°C in a shaking water bath (Vision Scientific Co. Ltd., Daejeon, South Korea). An aliquot (0.1 ml) of E. coli O157:H7 was added to 9.9 ml of each MCFA solution to yield an initial inoculation level of approximately 107 CFU/ml. For the OAs, 0.1 ml of bacterial suspension was added to 9.9 ml of a preheated solution containing a single OA (final OA concentration, 0.125, 1, 50, 100, 150, or 200 mM). The tubes containing the MCFA or OA plus the bacteria were then incubated at 37°C in a shaking water bath (100 rpm) for 5 min. Bacteria were also added to distilled water and a 1% ethanol solution as solvent controls.
(iii) Combined treatment with MCFAs and OAs.
The individual MCFAs or OAs that did not show bactericidal activity in the above-described experiments were combined to examine whether they showed any synergistic effects. Each of the three MCFAs (at a concentration of 0.125 or 1.0 mM) was combined with each of the four OAs (at a concentration of 0.125 or 1.0 mM) to yield 24 different treatment combinations in all. Briefly, a 0.1-ml aliquot of each stock solution (12.5 or 100 mM) of MCFA or OA was added to 9.7 ml of saline in a sterile tube (9.7 ml of saline, 0.1 ml of MCFA, and 0.1 ml of OA; total, 9.9 ml). After preheating, 0.1 ml of the E. coli O157:H7 suspension was added to 9.9 ml of the solutions containing MCFAs and OAs. Thus, the final concentration of MCFA and OA individually in each sample was 0.125 mM or 1.0 mM. E. coli O157:H7 was incubated at 37°C with shaking at 100 rpm for 5 min. To examine whether pH had any influence on the bactericidal effects, E. coli O157:H7 was treated with different concentrations of CLA plus CTA (0, 0.5, 0.75, 1.0, and 1.25 mM CLA combined with 0.5, 0.75, 1.0, 1.25, and 1.5 mM CTA; a total of 25 different treatment combinations) according to the method described above.
Microbiological analysis.
Bacterial samples (1 ml) that had been treated with the antimicrobial agents were diluted 10-fold in 0.85% sterile saline (9 ml), and 0.1 ml was spread plated onto duplicate tryptic soy agar (TSA) plates. For the lower detection limit, 0.2 ml of undiluted sample was spread plated onto five plates (a total of 1 ml was spread plated; the detection limit in this study was 1 CFU/ml). The plates were then incubated at 37°C for 24 h to allow the surviving E. coli O157:H7 cells to form colonies. The number of colonies on the TSA plates was then counted. Each experiment was repeated six times.
pH measurement.
The pH of the solutions containing either untreated or treated bacteria was measured at room temperature using a combination electrode (SevenEasy pH meter S20; Mettler-Toledo, Greifensee, Switzerland).
Flow cytometry.
Stock solutions of SYTO9 (Molecular Probes, Invitrogen, Eugene, OR) and propidium iodide (PI; Molecular Probes) were prepared by dissolving them in dimethyl sulfoxide (Sigma Chemical Co.) or sterile distilled water, respectively. The stock solutions were then stored in the dark at −20°C and at 4°C, respectively. Samples of bacteria that were untreated, treated with MCFAs alone (1.0 mM CLA, 0.125 mM CRA, or 1.0 mM LRA), treated with OAs alone (0.125 or 1.0 mM AA, LA, MA, or CTA), or treated with different combinations of MCFA and an OA (1.0 mM CLA plus 1.0 mM AA, LA, MA, or CTA; 0.125 mM CRA plus 0.125 mM AA, LA, MA, or CTA; or 1.0 mM LRA plus 1.0 mM AA, LA, MA, or CTA) were diluted to approximately 106 to 107 CFU/ml in sterile phosphate-buffered saline (PBS; pH 7.4). The concentrations of the MCFAs and OAs in these solutions were based on the results of the plating experiments described above. One milliliter of each diluent was transferred to a 1.5-ml Eppendorf tube and centrifuged (13,000 × g) at 4°C for 3 min (VS-550; Vision Scientific Co.). The resulting cell pellet was washed twice in sterile PBS and resuspended in 1 ml of sterile PBS. The bacteria were then double stained with SYTO9 (5 μM the working solution) and PI (30 μM) and incubated at 37°C for 15 min in a dark room. After staining, the bacteria were washed twice with sterile PBS and placed on ice in the dark until use.
Flow cytometric analysis was performed using a FACSCalibur flow cytometer (Becton, Dickinson, Franklin Lakes, NJ) equipped with a 15-mW argon ion laser with an excitation wavelength of 488 nm. The green fluorescence emitted by SYTO9 was detected at 530 ± 15 nm (fluorescence 1 channel [FL1]), and the red fluorescence emitted by PI was detected at ∼670 nm (FL3 channel). The signals were detected in the direction of forward light scatter using a photodiode. The side scatter, FL1, and FL3 signals were collected by a fluorescence collection lens and split by dichroic mirrors. Data acquisition was set to 50,000 events at a flow rate of 12 μl/min. Data were analyzed using BD CellQuest Pro software (Becton, Dickinson, San Jose, CA), and the results were depicted as density plots (FL1 versus FL3). All parameters were detected as logarithmic signals.
TEM.
E. coli O157:H7 was untreated or treated with 1.0 mM CLA, 1.0 mM CTA, or 1.0 mM CLA plus 1.0 mM CTA and then collected by centrifugation (13,000 × g) at 4°C for 3 min. The pellets were washed twice with sterile PBS, and the bacteria were fixed overnight in a mixture of cold 2.5% glutaraldehyde in 0.1 M phosphate buffer (pH 7.2) and 2% paraformaldehyde in 0.1 M phosphate or cacodylate buffer (pH 7.2). The bacteria were then postfixed for 90 min in 2% osmium tetroxide in 0.1 M phosphate or cacodylate buffer at room temperature. The samples were washed briefly with deuterated H2O2 and then subjected to two rounds of dehydration in a graded series of ethanol solutions (50, 60, 70, 80, 90, 95, and 100% ethanol). The samples were then treated with a mixture of propylene oxide and Epon epoxy resin (Embed 812, nadic methyl anhydride, poly Bed 812, dodecenylsuccinic anhydride, and dimethylaminomethyl phenol; Electron Microscopy Polysciences). Finally, the samples were embedded in epoxy resin. The epoxy resin-embedded samples were loaded into capsules and polymerized by heating to 38°C for 12 h and then to 60°C for 48 h. Sections (1 nm thick) were cut for light microscopy and stained with 1% toluidine blue on a hot plate at 80°C for 45 s. Ultrathin sections were cut using an RMC MT-XL ultramicrotome and collected on a copper grid. Areas appropriate for thin sectioning were cut into 65-nm sections and stained with saturated 4% uranyl acetate and 4% lead citrate before examination under a transmission electron microscope (TEM; JEM-1400; JEOL, Tokyo, Japan) at 80 kV.
Statistical analysis.
The average CFU counts on duplicate plates from six repeated experiments were converted to the log10 number of CFU/ml for examination by analysis of variance (ANOVA). The log bacterial population and the pH of the treated cell solutions were calculated using the general linear model (GLA) within the SAS package (version 9.13; SAS Institute Inc., Cary, NC). When ANOVA indicated a significant result (P < 0.05), the mean values were separated using Tukey's multiple-range test.
RESULTS
Bactericidal effects of individual MCFAs (CLA, CRA, or LRA).
Control treatments (sterile distilled water or 1% ethanol) showed no bactericidal effects in any of the experiments (data not shown), whereas all of the MCFAs showed bactericidal activity over a wide range of concentrations (0.125 to 1.5 mM) (Table 1). CRA showed the highest level of activity, followed by CLA and LRA. Treatment with 0.125 mM CRA resulted in a 0.49-log-unit reduction in the E. coli O157:H7 population (to 7.05 CFU/ml); however, this was not significantly different from the initial population (7.55 log CFU/ml) (P > 0.05). Treatment with CRA at 0.25, 0.5, 1.0, and 1.5 mM resulted in 2.75-, 5.68-, 7.29-, and 7.45-log-unit reductions, respectively. Treatment with CLA at a concentration of <1.0 mM did not have any significant effect on the bacterial population (P > 0.05). However, treatment with 1.5 mM CLA resulted in a 7.32-log-unit reduction. LRA had little antibacterial effect at any of the concentrations tested (<0.28-log-unit reduction; P > 0.05). The pH of the bacterial suspensions treated with LRA was higher (5.22 to 5.25) than that of the suspensions treated with CLA (3.82 to 4.58) or CRA (4.25 to 4.53). The pH of the suspensions treated with CLA or CRA tended to decrease in a concentration-dependent manner (P < 0.05); however, the LRA-treated suspensions showed no significant differences in pH (P > 0.05).
Table 1.
Bactericidal effects of three MCFAs at different concentrations and pH values of the bacterial suspensionsa
| MCFA | Molarity (mM [% concn]) | pHb | Microbial cell count (log no. of CFU/ml)b |
|---|---|---|---|
| Initial population | 6.53 ± 0.02a | 7.55 ± 0.05a | |
| CLA | 0.125 (0.002) | 4.58 ± 0.04c | 7.52 ± 0.09a |
| 0.25 (0.004) | 4.45 ± 0.03d | 7.48 ± 0.10a | |
| 0.5 (0.008) | 4.14 ± 0.04f | 7.41 ± 0.17a | |
| 1.0 (0.016) | 3.98 ± 0.01g | 7.25 ± 0.18a | |
| 1.5 (0.024) | 3.82 ± 0.03h | 0.23 ± 0.36d | |
| CRA | 0.125 (0.0024) | 4.53 ± 0.03c | 7.05 ± 0.31a |
| 0.25 (0.0048) | 4.43 ± 0.02d | 4.79 ± 0.54a | |
| 0.5 (0.0096) | 4.24 ± 0.02e | 1.87 ± 0.41c | |
| 1.0 (0.019) | 4.24 ± 0.04e | 0.26 ± 0.37d | |
| 1.5 (0.029) | 4.25 ± 0.04e | 0.10 ± 0.25d | |
| LRA | 0.125 (0.0028) | 5.25 ± 0.03b | 7.43 ± 0.16a |
| 0.25 (0.0057) | 5.24 ± 0.02b | 7.44 ± 0.13a | |
| 0.5 (0.011) | 5.22 ± 0.03b | 7.40 ± 0.15a | |
| 1.0 (0.022) | 5.22 ± 0.03b | 7.39 ± 0.15a | |
| 1.5 (0.034) | 5.23 ± 0.01b | 7.27 ± 0.21a |
Data are for 6 suspensions. The detection limit was 1 CFU/ml.
Mean ± standard deviation values denoted by different superscript letters are significantly different from the values for the initial population (P < 0.05).
Bactericidal effects of individual OAs (AA, LA, MA, or CTA).
The antimicrobial effects of AA, LA, MA, and CTA were examined at concentrations of 0.125, 1.0, 50, 100, 150, and 200 mM. As shown in Table 2, none of the OAs had any significant bactericidal effect at a concentration of 0.125 or 1.0 mM (P > 0.05). LA and MA showed the greatest bactericidal effect, followed by CTA and AA; treatment with 150 mM LA, MA, CTA, or AA resulted in a 7.45-, 7.07-, 5.55-, or 0.33-log-unit reductions in the bacterial population, respectively. E. coli O157:H7 was completely eradicated upon treatment with 200 mM LA or MA. The pH of the CTA-treated bacterial suspension was the lowest, followed by that of the MA-, LA-, and AA-treated suspension; the pH values of the AA-, LA-, MA-, and CTA-treated suspensions were 2.60 to 4.48, 2.03 to 4.27, 1.93 to 4.13, and 1.76 to 4.06, respectively.
Table 2.
Bactericidal effects of four OAs at different concentrations and pH values of the bacterial suspensionsa
| OA | Molarity (mM [% concn]) | pHb | Microbial cell count (log no. of CFU/ml)b |
|---|---|---|---|
| Initial population | 6.54 ± 0.02a | 7.53 ± 0.06a | |
| AA | 0.125 (0.00072) | 4.48 ± 0.02b | 7.50 ± 0.12a |
| 1.0 (0.0057) | 3.93 ± 0.01f | 7.33 ± 0.17abc | |
| 50 (0.29) | 2.92 ± 0.01j | 7.31 ± 0.23abc | |
| 100 (0.57) | 2.77 ± 0.01k | 7.23 ± 0.12abc | |
| 150 (0.86) | 2.68 ± 0.04l | 7.21 ± 0.09abc | |
| 200 (1.14) | 2.60 ± 0.02m | 6.70 ± 0.21bcd | |
| LA | 0.125 (0.00093) | 4.27 ± 0.05c | 7.40 ± 0.17ab |
| 1.0 (0.0075) | 3.46 ± 0.01g | 7.22 ± 0.20abc | |
| 50 (0.37) | 2.34 ± 0.03n | 5.84 ± 0.67e | |
| 100 (0.75) | 2.19 ± 0.04o | 1.47 ± 0.65hg | |
| 150 (1.12) | 2.07 ± 0.02p | 0.08 ± 0.19i | |
| 200 (1.49) | 2.03 ± 0.01pq | NDi | |
| MA | 0.125 (0.0010) | 4.13 ± 0.02d | 7.40 ± 0.06ab |
| 1.0 (0.0083) | 3.28 ± 0.05h | 7.25 ± 0.21abc | |
| 50 (0.42) | 2.20 ± 0.04o | 6.34 ± 0.25de | |
| 100 (0.83) | 2.08 ± 0.02p | 1.30 ± 0.69hg | |
| 150 (1.25) | 1.97 ± 0.02qr | 0.46 ± 0.57i | |
| 200 (1.67) | 1.93 ± 0.02r | NDi | |
| CTA | 0.125 (0.0014) | 4.06 ± 0.01e | 7.50 ± 0.13a |
| 1.0 (0.012) | 3.20 ± 0.04i | 7.47 ± 0.12a | |
| 50 (0.58) | 2.07 ± 0.04p | 6.67 ± 0.30cd | |
| 100 (1.15) | 1.94 ± 0.02r | 4.23 ± 0.52f | |
| 150 (1.73) | 1.81 ± 0.01s | 1.98 ± 0.30g | |
| 200 (2.31) | 1.76 ± 0.03s | 1.20 ± 0.42h |
Data are for 6 suspensions. The detection limit was 1 CFU/ml. ND, not detected.
Mean ± standard deviation values denoted by different superscript letters are significantly different from the values for the initial population (P < 0.05).
Bactericidal effects of different combinations of MCFAs and OAs.
The concentration of each MCFA and OA used in the combined-treatment experiments was chosen according to the results of the individual-treatment experiments. Individual treatment with 1.0 mM CLA or 0.125 mM CRA did not show any bactericidal effect at any of the concentrations tested. Therefore, to ascertain whether the combined use of MCFAs and OAs had a synergistic effect, we treated bacteria with 0.125 mM CRA, CLA, or LRA plus 0.125 mM AA, LA, MA, or CA or with 1.0 mM CRA, CLA, or LRA plus 1.0 mM AA, LA, MA, or CA.
Figure 1 shows the bactericidal effects of MCFAs (0.125 mM) and OAs (0.125 mM), used either individually or in combination, against E. coli O157:H7 incubated at 37°C for 5 min. The bactericidal effects of CLA or LRA in combination with each of the four OAs (0.125 mM plus 0.125 mM) resulted in a negligible reduction in the bacterial population; indeed, the reductions were so small that it was not possible to ascertain whether there were any synergistic effects. However, when bacteria were exposed to CRA plus OAs (0.125 mM CRA plus 0.125 mM each OA), we observed a 0.61-log-unit reduction for CRA plus AA, a 2.30-log-unit reduction for CRA plus LA, a 2.89-log-unit reduction for CRA plus MA, and a 6.04-log-unit reduction for CRA plus CTA. The pH of the bacterial suspensions ranged from 3.95 to 4.38 for CLA plus OA, 3.97 to 4.39 for CRA plus OA, and 3.96 to 4.34 for LRA plus OA.
Fig 1.
Bactericidal effects of MCFAs alone (□), OAs alone (
), or combined treatment (■). Bacteria were treated with each of the MCFAs (CLA, CRA, and LRA; 0.125 mM), each of the OAs (AA, LA, MA, and CTA; 0.125 mM), or different combinations of MCFAs and OAs (0.125 mM plus 0.125 mM). *, the data in parentheses denote the pH of the suspension. The bars represent the standard deviation. Mean values denoted by the same letter (a to e) were not significantly different at P < 0.05.
Figure 2 shows the results of a similar experiment performed using MCFAs and OAs at a concentration of 1.0 mM. When tested individually, CRA caused a 7.29-log-unit reduction in the bacterial population; therefore, combined treatment with 1.0 mM CRA plus any of the four OAs (at 1.0 mM) resulted in the complete eradication of bacteria. The combined use of 1.0 mM CLA or 1.0 mM LRA plus 1.0 mM each OA also showed bactericidal effects that were significantly greater than those observed upon treatment with the individual MCFAs or OAs (P < 0.05). Whereas treatment with individual MCFAs or OAs showed negligible bactericidal effects, the combination of CLA plus AA (both at 1.0 mM) resulted in a marked reduction in the bacterial population (1.32-log-unit reduction), as did CLA plus LA (5.48-log-unit reduction), CLA plus MA (7.33-log-unit reduction), and CLA plus CTA (>7.51-log-unit reduction; no detectable bacteria). LRA also showed synergistic antimicrobial activity: treatment with LRA plus AA, LRA plus LA, LRA plus MA, or LRA plus CTA resulted in reductions of 2.53, 5.21, 5.50, and 6.27 log CFU/ml, respectively. Each MCFA showed the greatest bactericidal activity when used in combination with CTA, followed by MA, LA, and AA. The pH values of the bacterial suspensions treated with CLA plus OAs, CRA plus OAs, or LRA plus OAs were 3.14 to 3.81, 3.18 to 3.86, and 3.17 to 3.87, respectively.
Fig 2.
Bactericidal effects of MCFAs alone (□), OAs alone (), or combined treatment (■). Bacteria were treated with each of the MCFAs (CLA, CRA, and LRA; 1.0 mM), each of the OAs (AA, LA, MA, and CTA; 1.0 mM), or different combinations of MCFAs and OAs (1.0 mM plus 1.0 mM) (n = 6). *, the data in parentheses denote the pH of the suspension. The bars represent the standard deviation. Mean values denoted by the same letter (a to f) were not significantly different at P < 0.05.
Flow cytometric analysis.
E. coli O157:H7 was treated with MCFAs and OAs at concentrations that showed a synergistic effect in the previous experiments. Bacteria were treated with individual MCFAs (1.0 mM CLA, 0.125 mM CRA, or 1.0 mM LRA), individual OAs (AA, LA, MA, or CTA at 0.125 or 1.0 mM), or a combination of the two (1.0 mM CLA plus each OA, 0.125 mM CRA plus each OA, or 1.0 mM LRA plus each OA) and analyzed by flow cytometry. Viable cells showed a strong green fluorescence signal (FL1; R0), whereas permeabilized cells or dead cells showed a strong red fluorescence signal (FL3; R1).
The data derived from E. coli O157:H7 treated with individual MCFAs or OAs and then double stained with SYTO9-PI are shown in Fig. 3. The dot plots show that, compared with untreated cells (Fig. 3A), the signal emitted by cells treated with MCFAs shifted from R0 to R1; i.e., the samples contained more damaged cells (Fig. 3B to D). The damaged cells within the untreated E. coli O157:H7 population represented only 5.35% of the total population within R1; however, this increased to 55.16%, 24.74%, and 24.29% after treatment with CLA (1.0 mM), CRA (0.125 mM), and LRA (1.0 mM), respectively. A similar pattern was observed after treatment with individual OAs (Fig. 3E to L), regardless of the type of OA or its concentration, although the percentages of dead/dying cells were smaller than those observed after treatment with MCFAs (range, 10.70% for 0.125 mM LA to 15.84% for 1.0 mM MA).
Fig 3.
Flow cytometry plots showing E. coli O157:H7 treated with individual MCFAs or with individual OAs for 5 min at 37°C and then double stained with SYTO9-PI. (A) Untreated cells; (B) CLA (1.0 mM); (C) CRA (0.125 mM); (D) LRA (1.0 mM); (E to H) AA, LA, MA, and CTA (0.125 mM each), respectively; (I to L) AA, LA, MA, and CTA (1.0 mM each), respectively. The black line divides live cells in gate 1 (R0) and damaged or dead cells in gate 2 (R1). The percentage of cells in R1 is also shown.
Figure 4 shows dot plots derived from E. coli O157:H7 bacteria treated with different combinations of MCFAs and OAs. Combined treatment resulted in a marked shift in the signal from R0 to R1. For all treatment combinations, >90% of the E. coli O157:H7 bacteria clustered in R1. The percentage of bacteria in the R1 region was lower after treatment with 1.0 mM CLA plus 1.0 mM AA (91.86%), 0.125 mM CRA plus 0.125 mM AA (95.50%), and 1.0 mM LRA plus 1.0 mM AA (93.86%) than after combined treatment with CLA, CRA, or LRA plus LA, MA, or CTA (all >98.03%).
Fig 4.
Flow cytometry plots showing E. coli O157:H7 treated with different combinations of MCFAs and OAs at 37°C for 5 min and then double stained with SYTO9-PI. (A to D) CLA (1.0 mM) plus AA, LA, MA, or CTA (1.0 mM each), respectively; (E to H) CRA (0.125 mM) plus AA, LA, MA, or CTA (0.125 mM each), respectively; (I to L) LRA (1.0 mM) plus AA, LA, MA, or CTA (1.0 mM each), respectively. The black line divides live cells in gate 1 (R0) and damaged or dead cells in gate 2 (R1). The percentage of cells in R1 is also shown.
TEM analysis.
The combination of CLA plus CTA showed the most marked bactericidal effect. Accordingly, bacteria were treated with CLA alone (1.0 mM), CTA alone (1.0 mM), or CLA plus CTA (both at 1.0 mM) and observed by TEM. Untreated E. coli O157:H7 bacteria (Fig. 5A) showed a normal morphology and had an intact cell membrane. E. coli O157:H7 treated with CTA alone also showed a normal morphology and an intact membrane (Fig. 5C). The cells treated with CLA did sustain some damage to the cell membrane, although the morphology was relatively normal (Fig. 5B). In contrast, bacteria treated with CLA plus CTA showed a highly deformed structure. Some showed a ruptured cell membrane; the membrane was shrunken and separated from the cell wall, and the intracellular components were released (Fig. 5D).
Fig 5.
Transmission electron microscopy images of E. coli O157:H7 treated at 37°C for 5 min with CLA, CTA, or a combination of both. (A) Untreated cells; (B) CLA (1.0 mM); (C) CTA (1.0 mM); (D) CLA (1.0 mM) plus CTA (1.0 mM).
Effects of pH on bactericidal effects mediated by CLA plus CTA.
To determine the effects of pH on the bactericidal effects of MCFAs and OAs, we examined the reductions in the bacterial population in the solutions containing different concentrations of CLA plus CTA (Table 3). The pH of the combined solutions decreased significantly as the CTA concentration increased (P < 0.05); however, the CLA concentration had no effect on pH (P > 0.05). To compare the effects of pH on the bactericidal effects, the different treatment combinations outlined in Table 3 were categorized according to the CTA concentration; thus, each combination with the same pH falls into the same group. Table 3 shows that the pH values of the combined solutions containing CTA at 0.5, 0.75, 1.0, 1.25, and 1.5 mM were 3.36 to 3.38, 3.26 to 3.28, 3.18 to 3.20, 2.96 to 2.98, and 2.86 to 2.87, respectively. The bactericidal activity of the combined treatments did not always correlate with the pH of the solution. Within the same pH group (at the same CTA concentration), different concentrations of CLA showed different bactericidal effects: from no reduction in the microbial population (i.e., 0.5 mM CLA plus CTA) to complete elimination of all bacteria (i.e., 1.25 mM CLA plus CTA). Although the pH of the solution containing 0.5 mM CLA plus 1.5 mM CTA (pH 2.87) was lower than that of the solution containing 1.25 mM CLA plus 0.5 mM CTA (pH 3.36), the bactericidal activity of the latter (complete elimination, >7.47-log-unit reduction) was higher than that of the former (no reduction).
Table 3.
Effect of pH on bactericidal effects of different concentrations of CLA combined with different concentrations of CTAa
| Molarity of CTA (mM) | Molarity of CLA (mM) | pHb | Microbial cell count (log no. of CFU/ml)b |
|---|---|---|---|
| Initial population | 6.54 ± 0.02a | 7.47 ± 0.15a | |
| 0.5 | 0 | 3.38 ± 0.02b | 7.26 ± 0.03a |
| 0.5 | 3.38 ± 0.01b | 7.27 ± 0.02a | |
| 0.75 | 3.37 ± 0.01b | 6.53 ± 0.34ab | |
| 1.0 | 3.36 ± 0.01b | 3.63 ± 0.69d | |
| 1.25 | 3.36 ± 0.02b | NDf | |
| 0.75 | 0 | 3.28 ± 0.01c | 7.26 ± 0.03a |
| 0.5 | 3.27 ± 0.01c | 7.19 ± 0.16a | |
| 0.75 | 3.27 ± 0.02c | 5.78 ± 0.14bc | |
| 1.0 | 3.27 ± 0.01c | 0.77 ± 0.68f | |
| 1.25 | 3.26 ± 0.01c | NDf | |
| 1.0 | 0 | 3.20 ± 0.03d | 7.25 ± 0.02a |
| 0.5 | 3.20 ± 0.02d | 7.13 ± 0.04a | |
| 0.75 | 3.19 ± 0.02d | 4.96 ± 0.24c | |
| 1.0 | 3.18 ± 0.01d | NDf | |
| 1.25 | 3.18 ± 0.02d | NDf | |
| 1.25 | 0 | 2.98 ± 0.01e | 7.21 ± 0.06a |
| 0.5 | 2.97 ± 0.02e | 7.09 ± 0.04a | |
| 0.75 | 2.97 ± 0.01e | 2.62 ± 0.72e | |
| 1.0 | 2.96 ± 0.01e | NDf | |
| 1.25 | 2.96 ± 0.02e | NDf | |
| 1.5 | 0 | 2.87 ± 0.02f | 7.19 ± 0.08a |
| 0.5 | 2.87 ± 0.02f | 7.03 ± 0.20a | |
| 0.75 | 2.87 ± 0.02f | 1.92 ± 0.57e | |
| 1.0 | 2.86 ± 0.03f | NDf | |
| 1.25 | 2.86 ± 0.01f | NDf |
Data are for 6 suspensions. The detection limit was 1 CFU/ml. ND, not detected.
Mean ± standard deviation values denoted by different superscript letters are significantly different from the values for the initial population (P < 0.05).
DISCUSSION
Combined treatment with MCFAs and OAs showed a marked synergistic bactericidal effect against E. coli O157:H7; indeed, the effects of combined treatment were much greater than the sum of the individual effects. For example, the reductions achieved upon individual treatment with 1.0 mM CLA or 1.0 mM CTA were 0.30 and 0.06 log unit, respectively; however, the combination of CLA plus CTA (both at 1.0 mM) completely eradicated all bacteria in the sample (>7.51-log-unit reduction). Thus, combination treatment achieved an additional >7.15-log-unit reduction in the bacterial population compared with the sum of the log unit reductions achieved by individual treatment.
The mechanism(s) underlying the bactericidal activity of MCFAs is not fully understood; however, many studies suggest that they act as nonionic surfactants, which become incorporated into the bacterial cell membrane (18, 19, 24). Studies also show that MCFAs diffuse through the bacterial cell membrane and create transient or permanent pores, resulting in altered membrane permeability and cell death (18, 19, 24). The mechanisms underlying the bactericidal activity mediated by weak OAs have been widely investigated. Undissociated OAs penetrate the cell membrane and enter the cytoplasm, where they dissociate into charged anions and protons, thereby altering the hydrogen ion equilibrium inside the cell and raising the pH (22, 25). The presence of charged anions and protons within the cytoplasm can upset intracellular pH homeostasis, inhibit essential metabolic reactions, and cause the accumulation of toxic anions (22, 25).
We suggest that the increased antimicrobial activity observed after combined treatment with MCFAs and OAs may be due to changes in cell permeability and the loss of cell membrane integrity. We propose the following mechanism for these synergistic bactericidal effects: (i) OAs or MCFAs (mainly MCFAs) disrupt the integrity of the cell by damaging the cell membrane, (ii) membrane disruption accelerates the entry of other antimicrobial compounds (MCFAs, OAs, or hydrogen ions) into the cell, and (iii) the rapid and direct influx of other antimicrobial compounds results in increased bactericidal activity. In particular, hydrogen ions must penetrate the cell and enter the cytoplasm to have any antimicrobial effect. It may be that, after treatment with individual MCFAs or OAs alone, hydrogen ions are not able to enter the cells due to their polarity. On the other hand, the membrane damage caused by combined treatment with MCFAs and OAs may allow hydrogen ions to enter the cell; therefore, in such cases, the bactericidal activity is much greater than that observed after treatment with individual MCFAs or OAs.
Of the individual OAs, MA and LA showed the highest levels of bactericidal activity, followed by CTA and AA (Table 1). However, when used in combination with MCFAs, the order was CTA > MA > LA > AA (Fig. 1 and 2). This reflects the pKas of these OAs: the pKa values of CTA, MA, LA, and AA are 3.13, 3.40, 3.86, and 4.75, respectively (14, 26). OAs with lower pKas showed greater bactericidal activity when combined with MCFAs. CTA has the lowest pKa; this means that it produces more hydrogen ions when exposed to an aqueous environment (the higher concentration of hydrogen ions was also reflected in the pH of the treated suspensions; Tables 1 and 2). CTA molecules were most likely present in solution in a dissociated form; therefore, they were not able to enter the cell, resulting in weak bactericidal activity. However, when CTA was combined with MCFA, the resulting damage to the cell membrane allowed hydrogen ions to pass into the cell, causing a marked bactericidal effect.
Flow cytometry was then used to examine the ability of these compounds to damage the cell membrane. Although the conventional plating method used for the earlier experiments showed that treatment with individual MCFAs or OAs did not show any appreciable bactericidal effect (Fig. 1 and 2), flow cytometry showed that they did cause some damage to the cell membrane (Fig. 3). These results suggest that, although cells treated with individual MCFAs or OAs still retained the capacity to form colonies on agar plates, they had sustained some damage to the cell membrane (the cells were able to repair this damage when cultured on nutrient-rich medium). However, when cells were treated with a combination of MCFAs and OAs, a significant number were killed (as determined by plate counting; Fig. 1 and 2) and showed evidence of membrane damage (as determined by flow cytometry; Fig. 4). This suggests that, in this case, the damage was irreversible.
TEM analysis allowed us to visualize changes to both the bacterial cell membrane and intracellular components after treatment with individual agents and after combined treatment (Fig. 5). The results were consistent with those from flow cytometry. Although flow cytometry showed that there was an increase in membrane permeability after treatment with CTA alone, no significant morphological changes were observed on the TEM images. Treatment with CLA alone did damage the membrane; however, the cells were still viable (the cells retained their shape and formed colonies on agar plates). In contrast, bacteria treated with CLA plus CTA showed marked morphological changes, including a deformed structure and a completely disrupted membrane that allowed the release of cellular material. Taken together, the results outlined in the present study support the hypothesis that combined treatment with MCFAs and OAs results in a synergistic bactericidal effect.
MCFAs and OAs are weak acids; therefore, the pH of the combined solutions differed according to the concentrations of these components. The pH of the bacterial suspensions was primarily determined by the CTA concentration rather than by the CLA concentration; this is because CTA has a lower pKa than CLA. Treatments performed within same pH range showed different E. coli O157:H7 killing activity; indeed, combined treatments at a lower pH showed less of a bactericidal effect than treatments at a higher pH (Table 3). These results indicate that it is the concentration of the antimicrobial compound rather than the pH of the suspension that has the greatest effect on bactericidal activity; in particular, the concentration of CLA appeared to have a more marked effect on bactericidal activity than the concentration of CTA. However, within combined solutions that contained the same concentration of CLA, the pH also had an effect on bactericidal activity (Fig. 1 and 2 and Table 3).
The use of natural antimicrobials to kill pathogens has received renewed attention because of public concern about the use of synthetic chemicals. MCFAs and OAs are natural compounds found in many foods (14–16, 21). The CLA and OAs used in the present study have all been awarded generally recognized as safe (GRAS) status by the U.S. Food and Drug Administration (CFR §184.1025 for CLA, §184.1005 for AA, §184.1061 for LA, §184.1069 for MA, and §184.1386 for CTA) (27). In addition, MCFAs and OAs are generally inexpensive, and it is possible that strong bactericidal effects can be achieved by combined treatment with very low concentrations of each agent. For example, very small amounts of CLA (0.016%) and CTA (0.012%) were required to eradicate bacteria when the two were combined. The obvious strengths and advantages of the method proposed herein are as follows: (i) all the antimicrobial compounds used are natural, (ii) their use should be acceptable to the consumer because OAs are widely used for industrial and food applications, (iii) they are relatively inexpensive, and (iv) the loss of nutrient content in products containing these molecules will be minimal because the concentrations required are very low.
Previous studies report that it is more difficult to inactivate Gram-negative bacteria using medium-chain triglycerides and their corresponding monoglycerides than it is to inactivate Gram-positive bacteria; this is due to differences in the structure and permeability of the outer membrane (28, 29). The bactericidal activity of the combined treatments examined in the present study was closely related to the level of cell membrane damage; thus, synergistic bactericidal effects may differ according to the bacterial species being treated. We also found that combined treatment with CLA plus CTA was effective against Cronobacter sakazakii, Salmonella enterica serovar Typhimurium, and Listeria monocytogenes (data not shown). Further studies are required to better understand the spectrum of antibacterial activity shown by these combined treatments.
In conclusion, the results of the present study clearly show that the combined use of MCFAs and OAs at low concentrations has a synergistic bactericidal effect, indicating that they may be promising alternatives to the conventional antimicrobial drugs used to control pathogens. In addition, we used flow cytometry and TEM to provide clear evidence for the mechanism underlying these antibacterial effects. These findings highlight the potential use of such combined treatments to improve microbiological safety in public health, medical centers, and the food industry. Indeed, these compounds could be used as the active ingredients in surface disinfectants, as sterilization agents for household appliances and medical equipment, as prophylactic agents coated onto contact surfaces, and as food additives/preservatives. The next logical steps are to optimize the combinations and concentrations of these molecules required to achieve the strongest bactericidal effect and to examine their effects on clinical or food specimens.
ACKNOWLEDGMENTS
This study was supported by a Korea University grant. We also thank the Institute of Biomedical Science and Food Safety, South Korea University Food Safety Hall, for providing the equipment and facilities.
Footnotes
Published ahead of print 16 August 2013
REFERENCES
- 1. Armstrong GL, Hollingsworth J, Morris JG., Jr 1996. Emerging foodborne pathogens: Escherichia coli O157:H7 as a model of entry of a new pathogen into the food supply of the developed world. Epidemiol. Rev. 18:29–51 [DOI] [PubMed] [Google Scholar]
- 2. Faith N, Shere J, Brosch R, Arnold K, Ansay S, Lee M, Luchansky J, Kaspar C. 1996. Prevalence and clonal nature of Escherichia coli O157:H7 on dairy farms in Wisconsin. Appl. Environ. Microbiol. 62:1519–1525 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Lee GY, Jang HI, Hwang IG, Rhee MS. 2009. Prevalence and classification of pathogenic Escherichia coli isolated from fresh beef, poultry, and pork in Korea. Int. J. Food Microbiol. 134:196–200 [DOI] [PubMed] [Google Scholar]
- 4. Rangel JM, Sparling PH, Crowe C, Griffin PM, Swerdlow DL. 2005. Epidemiology of Escherichia coli O157:H7 outbreaks, United States, 1982-2002. Emerg. Infect. Dis. 11:603–609 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Dundas S, Todd WTA, Stewart AI, Murdoch PS, Chaudhuri A, Hutchinson SJ. 2001. The central Scotland Escherichia coli O157:H7 outbreak: risk factors for the hemolytic uremic syndrome and death among hospitalized patients. Clin. Infect. Dis. 33:923–931 [DOI] [PubMed] [Google Scholar]
- 6. Kaper JB, Nataro JP, Mobley HLT. 2004. Pathogenic Escherichia coli. Nat. Rev. Microbiol. 2:123–140 [DOI] [PubMed] [Google Scholar]
- 7. Doyle MP, Schoeni JL. 1987. Isolation of Escherichia coli O157:H7 from retail fresh meats and poultry. Appl. Environ. Microbiol. 53:2394–2396 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Mead PS, Griffin PM. 1998. Escherichia coli O157:H7. Lancet 352:1207–1212 [DOI] [PubMed] [Google Scholar]
- 9. Rhee MS, Lee SY, Dougherty RH, Kang DH. 2003. Antimicrobial effects of mustard flour and acetic acid against Escherichia coli O157:H7, Listeria monocytogenes, and Salmonella enterica serovar Typhimurium. Appl. Environ. Microbiol. 69:2959–2963 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Tsuboi I, Ida H, Yoshikawa E, Hiyoshi S, Yamaji E, Nakayama I, Nonomiya T, Shigenobu F, Shimizu M, O'Hara K, Sawai T, Mizuoka K. 1998. Antibiotic susceptibility of enterohemorrhagic Escherichia coli O157:H7 isolated from an outbreak in Japan in 1996. Antimicrob. Agents Chemother. 42:431–432 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Frenzen PD, Drake A, Angulo FJ. 2005. Economic cost of illness due to Escherichia coli O157 infections in the United States. J. Food Prot. 68:2623–2630 [DOI] [PubMed] [Google Scholar]
- 12. Bae YY, Lee HJ, Kim SA, Rhee MS. 2009. Inactivation of Alicyclobacillus acidoterrestris spores in apple juice by supercritical carbon dioxide. Int. J. Food Microbiol. 136:95–100 [DOI] [PubMed] [Google Scholar]
- 13. Kim SA, Kim OY, Rhee MS. 2010. Direct application of supercritical carbon dioxide for the reduction of Cronobacter spp. (Enterobacter sakazakii) in end products of dehydrated powdered infant formula. J. Dairy Sci. 93:1854–1860 [DOI] [PubMed] [Google Scholar]
- 14. Zentek J, Buchheit-Renko S, Ferrara F, Vahjen W, Van Kessel A, Pieper R. 2011. Nutritional and physiological role of medium-chain triglycerides and medium-chain fatty acids in piglets. Anim. Health Res. Rev. 12:83–93 [DOI] [PubMed] [Google Scholar]
- 15. Jensen RG. 2002. The composition of bovine milk lipids: January 1995 to December 2000. J. Dairy Sci. 85:295–350 [DOI] [PubMed] [Google Scholar]
- 16. Sprong RC, Hulstein MFE, Van der Meer R. 2001. Bactericidal activities of milk lipids. Antimicrob. Agents Chemother. 45:1298–1301 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Korea Food and Drug Administration 2004. Korea food additives code. Korea Food and Drug Administration, Chungbuk, South Korea: http://fa.kfda.go.kr/foodadditivescode.html Accessed 1 July 2013 [Google Scholar]
- 18. Bergsson G, Arnfinnsson J, Steingrímsson Ó, Thormar H. 2001. Killing of Gram positive cocci by fatty acids and monoglycerides. APMIS 109:670–678 [DOI] [PubMed] [Google Scholar]
- 19. Desbois AP, Smith VJ. 2010. Antibacterial free fatty acids: activities, mechanisms of action and biotechnological potential. Appl. Microbiol. Biotechnol. 85:1629–1642 [DOI] [PubMed] [Google Scholar]
- 20. Wang LL, Johnson EA. 1992. Inhibition of Listeria monocytogenes by fatty acids and monoglycerides. Appl. Environ. Microbiol. 58:624–629 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Piper P, Calderon CO, Hatzixanthis K, Mollapour M. 2001. Weak acid adaptation: the stress response that confers yeasts with resistance to organic acid food preservatives. Microbiology 147:2635–2642 [DOI] [PubMed] [Google Scholar]
- 22. Brul S, Coote P. 1999. Preservative agents in foods: mode of action and microbial resistance mechanisms. Int. J. Food Microbiol. 50:1–17 [DOI] [PubMed] [Google Scholar]
- 23. Davidson PM, Branen AL. 2005. Food antimicrobials—an introduction, p 1–10 In Davidson PM, Sofos JN, Branen AL. (ed), Antimicrobials in food, 3rd ed. CRC Press, Boca Raton, FL [Google Scholar]
- 24. Altieri C, Bevilacqua A, Cardillo D, Sinigaglia M. 2009. Effectiveness of fatty acids and their monoglycerides against gram negative pathogens. Int. J. Food Sci. Technol. 44:359–366 [Google Scholar]
- 25. Doores S. 2005. Organic acids, p 91–141 In Davidson PM, Sofos JN, Branen AL. (ed), Antimicrobials in food, 3rd ed. CRC Press, Boca Raton, FL [Google Scholar]
- 26. Hsiao CP, Siebert KJ. 1999. Modeling the inhibitory effects of organic acids on bacteria. Int. J. Food Microbiol. 47:189–201 [DOI] [PubMed] [Google Scholar]
- 27.U.S. Food and Drug Administration 2013. Database of Select Committee on GRAS Substances. U.S. Food and Drug Administration, Rockville, MD: http://www.accessdata.fda.gov/scripts/fcn/fcnNavigation.cfm?rpt=scogsListing Accessed 1 July 2013 [Google Scholar]
- 28. Nair M, Joy J, Vasudevan P, Hinckley L, Hoagland T, Venkitanarayanan K. 2005. Antibacterial effect of caprylic acid and monocaprylin on major bacterial mastitis pathogens. J. Dairy Sci. 88:3488–3495 [DOI] [PubMed] [Google Scholar]
- 29. Nobmann P, Smith A, Dunne J, Henehan G, Bourke P. 2009. The antimicrobial efficacy and structure activity relationship of novel carbohydrate fatty acid derivatives against Listeria spp. and food spoilage microorganisms. Int. J. Food Microbiol. 128:440–445 [DOI] [PubMed] [Google Scholar]