INTRODUCTION
More than 95% of human
Salmonella infections are due to the ingestion of contaminated foods such as meat, dairy products, and vegetables (
1). Nuts, which include tree nuts and peanuts are low-moisture foods with a low water activity (a
w) and are not usually favorable for bacterial growth. Despite this, several outbreaks of
Salmonella infection associated with peanuts and peanut butter have been reported since the mid-1990s. In low-a
w foods such as peanut butter, ingesting as few as 10 to 100
Salmonella colony forming units (CFU) is enough to cause infections (
2). One of the biggest challenges with processing and sanitation is that
Salmonella exhibits a markedly increased heat resistance in low-a
w food products, such as nuts (
3).
Effective cleaning and sanitation are essential for preventing initial and cross-contamination of bacterial pathogens such as
Salmonella (
4). Traditional cleaning and sanitation utilize water-based cleaning (or wet cleaning), followed by a sanitizing step (
5). Wet processing environments can encourage microbial growth, and splashing during wet cleaning can contribute to the aerosolization and the potential spread of pathogens (
6). Wet cleaning of peanut butter and other nut pastes is especially challenging due to the high dust levels, high-fat content of peanuts, and the immiscible nature of nut dust and butter with water (
7–9). Therefore, nonaqueous cleaning and sanitizing (also known as dry cleaning and dry sanitizing) procedures are used (
10). However, traditional dry-cleaning methods have been reported to be inefficient for the processing of peanut butter and other low-moisture or low-a
w food products (
11). Disinfection of low-moisture food products is often based on the use of isopropanol or ethanol that evaporates after use (
4). The major disadvantage of these compounds, however, is that they are flammable and require equipment to cool before disinfection (
12,
13). This can lead to significant processing downtimes, causing economic losses. In this study, we evaluate the use of oils such as peanut oil as vehicles for delivering organic acids against bacterial contamination. The advantage of using oils for delivering antimicrobial compounds is the opportunity of using high temperatures, thus eliminating processing downtimes (and possibly, economic losses) associated with dry sanitization methods. Organic acids have been used effectively against enteric pathogens by other researchers (
14). The effectiveness of organic acids against bacteria is related to the undissociated proportion of the acid, which in turn, is dependent upon the pKa of the acid and the pH of aqueous systems (
15,
16). The undissociated form is not charged and is thought to pass through lipid bilayers and enter the bacterial cytoplasm (
17). The undissociated form of the acid then dissociates into protons and anions in the cytoplasm, dropping the intracellular pH. Under normal conditions, ATP is generated in cells by the entry of H
+ ions into the cytoplasm along a pH gradient. However, continuous entry of acid molecules inside bacterial cells leads to the accumulation of protons, leading to a disruption of intracellular pH homeostasis and proton motive force (
18). We hypothesize that this will contribute to the antimicrobial effects of the acidified oils.
The use of oils can be an effective delivery approach to supply antimicrobials in environments where wet cleaning cannot be used. Food industries dealing with low-moisture foods often maintain their processing units at low a
w, which allows contaminating microbes to exist in a desiccated state (
19). In addition,
Salmonella has been reported to demonstrate higher heat resistance in high-fat and low-a
w environments such as in almonds, peanuts, and ground nut pastes (such as peanut butter) (
20,
21). Hence, in this study, we have tested the efficacy of acidified oils against desiccated
S. Enteritidis cells. The objective of this research is to provide a proof-of-concept that formulated oil-based sanitizing agents have the potential for combatting
Salmonella enterica, which is a common foodborne pathogen in low-moisture food processing environments.
DISCUSSION
In this study, we have evaluated the antimicrobial properties of acidified oils against desiccated
S. Enteritidis. We observed that organic acids dissolved in oils had significant antimicrobial properties. We quantified the MLR of 500 mM a variety of fatty acids with chain lengths between C
1 to C
10 against desiccated
S. Enteritidis (
Fig. 1). Formic acid was the most effective organic acid and is a common additive in diets of poultry and other livestock to limit the presence of
S. enterica in poultry feed (
25). However, formic acid is known to have toxic effects on humans and its use is mostly limited to livestock feed (
26). Although formic acid and acetic acid were highly effective, propionic acid, which is another very short-chain organic acid, was not as effective against
S. Enteritidis. Since our experiments revealed that among the food-grade acids tested, acetic acid was the most effective in causing a significant reduction in
S. Enteritidis MLR, we conducted an in-depth analysis of the antimicrobial effects of acetic acid in oil, along with the mechanisms of its antimicrobial action.
Higher levels of MLR were observed with the increase in the concentration of acetic acid and treatment duration (
Fig. 2 and
3). We observed a synergistic effect of heat and acetic acid in oil on MLR of desiccated cells whereby the efficacy of the acid was significantly enhanced at high temperatures (
Fig. 1 and
Table 1). This is an important observation since it can be used to formulate effective sanitizing treatments against desiccated
S. enterica which are highly resistant to heat, especially in low-moisture food products (
27). In addition, acetic acid in oil was found to be effective against different
S. enterica strains, when treated individually as well as in a cocktail (
Fig. 8), providing more evidence for the wide range of efficacy of acidified oils against
S. enterica. For the other acids tested in this study, the relationship between the acids and heat were found to be additive (
Fig. 1 and
Table 1). Exposure to heat below 60°C is reported to be associated with minor increases in bacterial membrane permeability in
S. enterica (
28). This is a possible factor contributing to the higher MLR of acidified oils at high temperatures. The effect of heat in enhancing the antimicrobial properties of organic acids against other foodborne pathogens such as
E. coli has been reported in a previous study (
29).
Our lab has previously used the bright-field microscope of the oCelloScope to study the effects of antimicrobial compounds on bacterial cells (
23). The advantage of using this method is that the antimicrobial treatment is performed on cells desiccated in microtiter plate well, and then neutralized, without washing steps that may contribute to cell loss, such as when staining cells on a microscope slide. Using this method, we observed that cells treated with higher concentrations of acetic acid had lower BCA values. In other words, with the increase in concentration of acetic acid in oil, there was a reduction in the number of detectable cells after treatment. The BCA values are dependent on the pixel density recorded by the bright field microscope of the oCelloScope. Since light microscopy cannot differentiate between live and dead cells, a nonlytic antimicrobial treatment would be expected to have similar BCA values before and after treatment (
23). The measurement of reduction in BCA values after treatment was likely due to cellular lysis. In our study, after treatment with acidified oils, MLR values calculated from the BCA values of the oCelloScope and plate count methods were the same, indicating that lysis likely occurred during treatment. The effect of organic acids in inhibiting growth of Gram-negative pathogens has been reported by other researchers (
30–33). Proposed mechanisms include disruption and destabilization of outer membranes, osmotic imbalance, buildup of reactive oxygen species (ROS), and inhibition of ATP synthesis (
34–37). When desiccated cells containing damaged macromolecules and ROS are rehydrated, the damaged macromolecules and ROS gain mobility inside cells, leading to lethal damage to the cells (
38). A combination of the above-mentioned factors is likely responsible for the cellular lysis in
S. Enteritidis. Acetic acid stress is reported to cause apoptosis and necrosis in
Zygosaccharomyces bailii and
Saccharomyces cerevisiae (
39,
40) and programmed cell death in
Chlamydomonas reinhardtii (
41). Acetic acid in oil was effective against
S. Enteritidis desiccated on stainless steel surfaces (used for majority of the experiments in this study), as well as on cells desiccated in polystyrene wells in 96-well plates (
Fig. 4). Since stainless steel and plastic materials are commonly used in food processing industries (
42), this highlights the effectiveness of acidified oils as antimicrobial compounds against
Salmonella contamination on different surfaces.
Confocal microscopy helped in the further elucidation of some of the mechanisms of action of acetic acid in oil on recovered
S. Enteritidis cells. The results (
Fig. 5 and
6) indicate that an increase in the concentration of acetic acid, as well as an elevation in the treatment temperature, led to increased cell membrane permeability and reduction in cellular respiration. The integrity of cellular membranes is closely linked to the cellular respiration of aerobically respiring bacteria (
43). Gram-negative pathogens such as
E. coli and
S. enterica contain membrane-bound enzymes that help in the generation of electrons that are coupled to the electron transport chain (ETC) of the respiratory pathway (
44). Thus, disruption of cell membranes by acidified oils coupled with heat has a negative effect on cellular respiration, possibly by disrupting the ETC, which leads to the reduction in cellular respiration. Also, disruption of the ETC and cellular respiration has been reported to lead to the production of ROS, which in turn damages cellular components such as membrane lipids and proteins (
45). Thus, disruption in cellular respiration can promote the further degradation of bacterial cell membranes. Evidence for this phenomenon was observed in
Table 2, where an increase in the percentage of cells with membrane damage was associated with a similar increase in the percentage of cells with reduced cellular respiration. Disruption of bacterial cell membranes and changes in bacterial respiration after exposure to organic acids has been reported previously (
30–33,
46,
47). The absence of any significant increase in the number of cells with membrane damage or reduced respiration after exposure to heated unacidified oils indicates that oils alone do not significantly affect the integrity of desiccated
S. Enteritidis cells. This is consistent with findings by other researchers who have concluded that heated oils alone were not sufficient to eliminate
S. enterica contamination (
4).
TEM revealed the presence of structural changes in recovered
S. Enteritidis cells after treatment with acetic acid in peanut oil (
Fig. 7). The TEM images revealed cells with damaged membranes, irregular membranes, membrane ruffling, and changes in cellular periplasm of the bacterial cells. Similar changes in
S. enterica cell structure have been reported to be brought about by bactericidal compounds and treatment with other acids such as malic acid (
48). The formation of protein aggregates in
E. coli cells after exposure to environmental stresses such as heat and exposure to antibiotics has been reported (
49). We observed the presence of granularity in the cytoplasm, or regions of high and low electron density in the bacterial cytoplasm of cells treated with acidified oils at both temperatures, and after treatment with hot oil alone at 45°C. It is possible that the granularity in our system is indicative of aggregates of cytoplasmic contents as a stress response against heat and acidified oil treatment in
S. Enteritidis, as seen in other systems (
49).
Cells desiccated at lower %ERH values were found to be more resistant, while at higher %ERH values, cells became more sensitive to acidified oil treatment (
Fig. 9). We hypothesize that the hydrophobic environment created by the oil, facilitates the entry of acid molecules inside bacterial cells, possibly through an affinity of the acid to the water remaining within the desiccated cellular cytoplasm. Thus, the cellular water concentration may be important for the sensitivity to the acidified oil treatment. Alternatively, it is possible that the increased resistance to antimicrobials after exposure to low %ERH is due to an adaptive stress response in cells, as reported by researchers in other pathogens (
50,
51). Exposure to reduced %ERH is known to cause a differential change in the expression of several genes in
S. enterica which confers resistance against future exposure to environmental stresses (adaptive stress response) (
52,
53).
Edible oils such as palm oil and sesame oil have been used for the delivery of compounds in the pharmaceutical and cosmetic industry (
54,
55), and peanut oil emulsions have been reported for the delivery of insulin in the pharmaceutical industry (
56). In the present study, peanut oil has been used as the primary vehicle for the delivery of organic acids. However, we tested the efficacy of corn oil, mineral oil and MCT oil (see Fig. S2) for the delivery of acetic acid as an antimicrobial compound. The results indicate that the oils used in this study do not possess significant intrinsic antimicrobial properties and primarily serve as a carrier of the organic acids. The high oxidative and thermal stability of MCT oils (
57,
58) make them suitable alternative carriers for delivery of antimicrobial compounds at high temperatures. To our knowledge, use of oils as carriers for organic acids is a novel method of delivery of antimicrobial compounds against foodborne pathogens. Acidified oils could be used as an effective means of sanitation in low-moisture food processing facilities where use of water-based sanitization can be challenging.
Conclusion.
The results obtained in this study indicate that peanut oil acidified with acetic acid is effective against S. enterica contamination. This study demonstrates for the first time, the use of acidified oils for the destruction of a common foodborne pathogen, S. Enteritidis, on food contact surfaces. The results described here indicate that acidified oils can reduce reliance on the use of water for cleaning and sanitation. Use of oil-based antimicrobials is relevant, especially for the sanitization in industrial settings associated with processing of low-aw foods. The antimicrobial formulations described here have the potential to serve as cost-effective robust replacements to the dry sanitation methods currently in use and can improve the safety of low-aw foods.
MATERIALS AND METHODS
Bacterial isolates and growth media.
The bacterial isolates used in the study are S. enterica subsp. enterica serovar Enteritidis (ATCC BAA-1045, phage type 30), S. enterica subsp. enterica serovar Michigan (ATCC BAA-709), S. enterica subsp. enterica serovar Montevideo (ATCC BAA-710), and S. enterica subsp. enterica serovar Gaminara (ATCC BAA-711). The bacterial strains were maintained as frozen stocks in a −80°C freezer in Trypticase soy broth (TSB; B.D. Diagnostic Systems) supplemented with 15% glycerol. The bacterial cultures were revived in tryptic soy agar (TSA; B.D. Diagnostic Systems) and used for inoculum preparation. Frozen cultures of Salmonella were revived by streaking on TSA and incubated at 37°C for 18 to 20 h to create a bacterial lawn.
Chemicals and antimicrobial compounds.
Organic acids formic acid (Fisher Chemical, A1199P-1), acetic acid (glacial; Fisher Chemical, A38), propionic acid (Sigma-Aldrich, 402907), butyric acid (Sigma-Aldrich, B103500), valeric acid (Sigma-Aldrich, 240370), hexanoic acid (Sigma-Aldrich, 21530), heptanoic acid (Sigma-Aldrich, 21530), caprylic acid (Sigma-Aldrich, W279900), nonanoic acid (Sigma-Aldrich, N29902), and decanoic acid (Sigma-Aldrich, C1875) were used as antimicrobial compounds in this study. Stock solutions (500 mM) of the acids were prepared in sterile peanut oil and further diluted in sterile peanut oil to reach concentrations of 250, 100, 50, and 25 mM. Acidified oils of the above-mentioned concentrations were also prepared in sterilized corn oil, mineral oil, and medium-chain triglyceride (MCT) oil. The oils were purchased from the local supermarket and were sterilized by autoclaving for 30 min. Subsequently, 100-μL portions of the oils were plated on TSA to confirm their sterility. To desiccate S. Enteritidis cells at 33, 75, and 90% equilibrium relative humidity (ERH), saturated solutions of MgCl2 (ACS Organics, 7786-30-3), NaCl (Fisher Chemical, S271-1), and KCl (ACS Organics, LC187901), respectively, were used.
Evaluation of the susceptibility of S. Enteritidis to acidified oils on stainless steel slides.
Inoculum preparation was done using previously reported methods (
59) with variations. Stainless steel slides were used for inoculation of
S. Enteritidis and the other serovars of
S. enterica since it is one of the most common materials used in processing surfaces in food processing facilities (
60). For evaluation of the effect of acidified oils against desiccated
S. Enteritidis cells using stainless steel, stainless-steel slides (type 304, no. 2B, 5 × 1.2 cm) were soaked in acetone overnight to remove grease (
61) and then washed with water and autoclaved. The inocula were prepared from 20-h lawns on TSA, and cells were suspended in filter-sterilized deionized water to an optical density at 600 nm (OD
600) of 0.9. An inoculum of 20 μL was added to achieve an initial cell level of 10
6 to 10
7 CFU/slide. For the cocktail study, the strains were individually streaked on TSA and, after 20 h, the OD
600 values for the individual strains were adjusted to 0.9 in sterile-distilled water and then combined, prior to adding a 20-μL drop on a degreased sterile stainless-steel slide. The slides were incubated inside desiccators containing saturated solutions of MgCl
2, NaCl, and KCl at 20°C (33, 75, and 90% ERH, respectively). After 20 h, the dried bacterial inoculum films on the stainless-steel slides were used for subsequent experiments.
Antimicrobial treatment of the desiccated cells on stainless steel slides was performed as follows. To each slide, 50 μL of acidified oil was added over the desiccated inoculum film, followed by incubation at room temperature (20°C) or in a 45°C air incubator for 30 min. For treatment at 45°C, the oils were preheated at 45°C prior to treatment. After treatment, the slides were transferred into sterile tubes containing 10 mL of TSB buffered with 0.25 mM HEPES (Sigma-Aldrich) at pH 7.2. Sterile glass beads were added to aid in the removal of surviving bacteria from the surfaces of the stainless-steel slides. The tubes were vortexed for 5 min, and serial dilutions of the bacterial cells were plated on TSA for enumeration of the total number of surviving cells. After 20 h of incubation at 37°C, the colonies were counted, and the CFU/slide were calculated. Nontreatment control samples were stainless-steel slides containing desiccated bacterial cells that were not treated with acidified oils. The MLR was calculated by plating on TSA (traditional MLR quantification assay) using the following formula: MLR = log CFUnontreatment – log CFUtreatment.
Numbers of injured survivors were evaluated during select experiments using xylose lysine deoxycholate (XLD) agar (Remel, R459902) (
62,
63). It was assumed that calculated survivors on TSA contained both injured and healthy cells and that only healthy cells would grow on XLD agar. The levels of injured cells were calculated using the following formula: injured cells = log CFU
TSA – log CFU
XLD. The MLR quantification and injury assays were run in three technical and three biological replicates for each concentration of the acidified oils.
Evaluation of the susceptibility of S. Enteritidis to acidified oils using light microscopy.
For analysis of MLR of S. Enteritidis using an automated bright field microscope (oCelloScope; BioSense Solutions, Farum, Denmark), an inoculum was prepared from overnight lawns of S. Enteritidis that were suspended in filter-sterilized deionized water to obtain an inoculum with an OD600 of 0.1. The inoculum (20 μL) was added inside the wells of a 96-well microtiter plate (Thermo Fischer, model 266120) to obtain an initial cell level of 104 to 105 CFU/well. The plates were incubated inside desiccators maintained at 75% ERH by adding a saturated solution of NaCl and stored at 20°C for 20 h.
Oil formulations (50 μL) for each acid concentration were added to wells containing desiccated cells in a 96-well plate (in triplicates for each concentration), followed by incubation for 30 min. After treatment, 200 μL of TSB buffered with 0.25 mM HEPES (pH 7.2) was added to the wells to neutralize the acidified oils. The plates were centrifuged for 10 min at 2,000 rpm until the oils rise to the top of the wells. The 96-well plates were then placed inside the oCelloScope to analyze bacterial growth according to previously described methods (
23). The background corrected absorption (BCA) algorithm and BCA normalized algorithm on the software UniExplorer (version 10.1) were selected for analysis. BCA values from treated and untreated samples were used from the zero-time point to calculate MLR. Using data from the oCelloScope, MLR was calculated with the following formula: BCA
nontreatment – BCA
treatment. Nontreatment control samples were wells in a 96-well plate containing desiccated bacterial cells that were not treated with acidified oils. All experiments were run in three technical and three biological replicates for each concentration of the antimicrobial compounds.
Analysis of cell membrane damage and changes in cellular respiration of S. Enteritidis after treatment with acetic acid in peanut oil.
S. Enteritidis cells were desiccated overnight on stainless steel slides and held at 75% ERH, as described above. After 20 h, the slides were treated with different concentrations of acetic acid dissolved in peanut oil for 30 min. Six slides containing the desiccated inoculum of S. Enteritidis were treated with a specific concentration of acetic acid in peanut oil to increase the number of cells that would be used for further analysis using confocal microscopy. After treatment, the acidified oils were neutralized by adding the slides to sterile tubes containing 10 mL of TSB buffered with 0.25 mM HEPES (pH 7.2). The tubes were vortexed to ensure the transfer of the bacterial cells from the surfaces of the slides to the TSB. The slides were removed from the tubes aseptically, and the tubes were centrifuged at 6,000 rpm for 15 min. After centrifugation, the medium was discarded, and the bacterial pellets were suspended in 5 mL of fresh buffered TSB and centrifuged at 6,000 rpm for 10 min. The medium was discarded again, and the pellets were dissolved in 1 mL of buffered TSB. The dissolved pellets from the different tubes containing S. Enteritidis cells that were treated with the acidified oils (six tubes for each concentration of acetic acid) were combined into one fresh tube. The resulting tubes, each containing 6 mL of bacterial suspensions recovered after treatment with a specific concentration of acetic acid were further centrifuged at 6,000 rpm for 10 min. The centrifugations were repeated three more times with the pellets being dissolved in 1 mL, 500 μL, and 250 μL of sterile-distilled water. The bacterial cell suspensions in 250 μL were used for staining to analyze bacterial cell membrane damage and changes in cellular respiration after treatment with acetic acid in oil.
The cell suspensions were treated with Live/Dead BacLight (Thermo Fisher Live/Dead bacterial viability kit, L7012) and held in the dark for 15 min for analysis of bacterial membrane damage using confocal microscopy. The kit contains a mixture of Syto 9 (membrane permeable green, fluorescent stain) and propidium iodide (membrane impermeable red fluorescent stain). Using this staining kit, cells with damaged membranes were stained red with propidium iodide, and all cells, irrespective of their membrane integrity, were stained with the green-fluorescent stain Syto 9 (
64). For analysis of changes in cellular respiration, the bacterial suspensions were treated with BacLight Redox Sensor stain (Thermo Fisher CTC vitality kit, B34956) according to instructions provided by the company. The kit consists of CTC (which forms a fluorescent red precipitate, formazan, in actively respiring cells) and Syto 24, which is a green nucleic acid counterstain. The cells were initially stained with CTC and held at 37°C for 30 min, followed by staining with Syto 24 for 15 min. Actively respiring cells take up CTC, which is reduced by the bacterial electron transport chain to a red florescent formazan precipitate, and all cells were stained with a membrane-permeable green florescent stain Syto 24 (
65–67). After staining for analysis of cellular damage and cellular respiration changes, 50-μL portions of the stained cell suspensions were transferred to poly-
d-lysine-coated glass-bottom microwell dishes (Mat Tek, P35GC) and air dried in a fume hood for 2 h. The cells were then imaged using a A1R-SIMe confocal laser scanning microscope. For quantification of cell membrane damage and changes in cellular respiration at 20 and 45°C, each treatment consisted of three biological replicates and, each time, the images were recorded from 20 fields of view. The percentages of cells with membrane damage and reduced cellular respiration were quantified for each of the images using the quantification software of NIS elements (Nikon Instruments, Inc.). Data from 60 images for every sample were analyzed to quantify the percentages of cellular damage and reduction in cellular respiration.
Analysis of changes in cellular structure of S. Enteritidis cells after treatment with acetic acid in peanut oil.
Desiccated inoculum of S. Enteritidis was prepared on stainless-steel slides and treated with acetic acid in peanut oil as described above. Surviving cells were recovered in 300 mL of buffered TSB from a total of 30 stainless steel slides and centrifuged to obtain a cell pellet. The bacterial pellets were fixed overnight using 1.6% paraformaldehyde and 2% glutaraldehyde in 0.1 M sodium cacodylate buffer. Next day, the pellets were rinsed twice in 0.1 M sodium cacodylate buffer for 15 min. The pellets were stained with 2% osmium tetroxide in 0.2 M sodium cacodylate buffer (1:1) for 2 h and then rinsed three times with 0.1 M sodium cacodylate buffer for 15 min each time. The pellets were dehydrated with 10, 20, 30, 50, 70, 80, and 95% ethanol for 20 min each. This was followed by dehydration in 100% ethanol for 20 min three times. The pellets were then incubated with propylene oxide and resin (1:1) for 20 min three times, followed by incubation in propylene oxide and resin overnight. The next morning, the resin was changed with fresh resin three times and then left overnight in resin. The following morning, the pellets were placed in a 60°C oven for 48 h. The resin pellets were then sectioned using ultramicrotomy to produce sections of 70-nm thickness and then imaged by using a Philips CM10 transmission electron microscope.
Data analysis.
All the data are expressed as means ± the standard deviations. GraphPad Prism was used to generate all the graphs in this project. Statistical significance was calculated with GraphPad Prism using two-way ANOVA. A P value of ≤0.05 was considered statistically significant.