Pathogenic S. Enteritidis ATCC 13076, S. Stanley H0558, and L. monocytogenes Scott A 724, and non-pathogenic L. innocua 33090, L. innocua 33091, and L. innocua 51742 were obtained from USDA, ARS, ERRC culture collections. The non-pathogenic strains of Listeria are often used as surrogates for L. monocytogenes in studies because they behave similarly and have similar characteristics as L. monocytogenes.

Stock cultures were maintained at -80°C. Cultures of each individual strain were propagated on tryptic soy agar (TSA, Difco Laboratories, Detroit, MI, USA.) at 37°C and maintained at 4°C until use. Prior to the inoculum preparation, individual Salmonella and Listeria cells were grown in 10 mL tryptic soy broth (TSB: Difco Laboratories) for 18–20 h. After centrifugation at 1,800 ×g for 10 min, the cultures were suspended in 10 mL phosphate buffered saline (pH 7.2). The two Salmonella strains, and four Listeria cultures were combined to form two cocktails of cultures.

Sophorolipids were synthesized from C. bombicola ATCC 22214 using palmitic, stearic, and oleic acids as co-feedstocks, followed by extraction and purification (Ashby et al., 2011; Solaiman et al., 2015). Briefly, 10 L of Candida growth medium (CGM; 10% w/v glucose, 1% w/v yeast extract, and 0.1% w/v urea) was prepared in a 12-L capacity vessel of a bench-top bioreactor (Bioflo 3000 Batch/Continuous Bioreactor, New Brunswick, NJ, USA). Following sterilization by autoclaving and then cooling to 26°C, palmitic, stearic, and oleic acids (200 mL, technical grade, Sigma–Aldrich, St. Louis, MO, USA) were added to the medium to a final concentration of 2% (v/v). A stock inoculum culture (previously prepared and stored in -80°C freezer) was thawed and added to the 10-L CGM (containing 2% palmitic, stearic, or oleic acids) medium to initiate the fermentation. The bioreactor was set at the following parameters: 26°C, an impeller speed of 700 rpm, 2 L/min aeration, and no pH control. On day 2 of the fermentation, 7.5% (w/v) of granular glucose and 2% (v/v) of palmitic, stearic, or oleic acids were added to the culture. On day 5, 1% (v/v) of palmitic, stearic, or oleic acids was again added. On day 7, the entire culture (cells and broth) was collected and lyophilized in ∼2-L volumes (for ease of handling). The dried residues were transferred to Erlenmeyer flasks and extracted in excess ethyl acetate with shaking at room temperature for 2 days. The extract was filtered through Whatman No. 2 filter paper. The solids on the filter were additionally rinsed with two volumes of ethyl acetate to recover any residual SLs. The ethyl acetate fractions containing the SLs were combined and concentrated using a rotary-evaporator. The concentrate was slowly added to excess hexane to obtain the pure SLs in crystallized form. Each of the parental SL mixtures was analyzed by LC/MS according to the procedure previously outlined by Nuñez et al. (2004) and in all cases were found to be composed of greater than 95% lactones.

Free-acid SLs were produced by base-catalyzed ring-opening chemical reactions of the lactone form. A six molar solution of potassium hydroxide (KOH) was prepared in deionized water and 10 g of each lactonic SL (three total) was added to the KOH solution. This mixture was stirred overnight at room temperature, heated to 40°C and stirred for an additional 3 h. The mixture was then acidified to pH 3–4 using phosphoric acid and poured into a crystallization dish where a steady stream of nitrogen was used to help remove the water. Desalting was accomplished by stirring the dried SL mixture with excess 2:1 CHCl₃:ethanol and placing the mixture in a stoppered flask to inhibit CHCl₃ evaporation. After a few hours celite was added to the mixture (helps in filtration, otherwise the fine salts clog the filtration frit) and the free-acid SLs were recovered by vacuum filtration through Whatman #2 filter paper. The solids were washed an additional two times with 2:1 CHCl₃: ethanol, and the free-acid SLs were recovered and dried under vacuum until constant weight. In total, six SLs were obtained with three in the free-acid form and three in the lactonic form (Figure 1).

Purified palmitic, stearic, and oleic acid SLs in both free-acid and lactonic forms and TDS (Kanematsu U.S.A. Inc., Somerset, NJ, USA) at concentrations of 0.1, 0.5, and 1.0% (w/v) were tested in suspensions of Salmonella and Listeria cocktails. Stock solutions (5%, w/v) of palmitic, stearic, and oleic acid SLs and TDS were first dissolved in 100% ethanol. The stock solutions were then diluted in sterile water or ethanol to final concentrations of 0.1, 0.5, and 1.0% and various levels of ethanol (1–20%). Due to low solubility, 0.1, 0.5, and 1.0% of SLs and TDS could only be dissolved in minimum levels of 1, 5, and 10% ethanol, respectively.

For the antimicrobial test, 10 μL aliquots of bacterial inoculum were added into 990 μL of surfactant solutions giving a cell density of ∼7.2 log CFU/mL. The mixtures were then vortexed for 10 s. After 1 min and 1 h at 22°C, the suspensions were immediately serially diluted with neutralizing buffer (Difco Laboratories), and aliquots (100 μL or 1 mL) were spread plated on TSA. Plates were incubated at 37°C for 24 h and the number of colonies of bacteria was recorded. In a separate experiment, the effect of other treatment times (1 min, 1, 2, and 4 h) was also studied. Detection limit was 1 log CFU/mL. Propagation of bacteria and antimicrobial tests were carried out in a BSL-2 laboratory.

For the preparation of SEM samples, 100 μL aliquots of Listeria spp. inoculum were mixed with 900 μL of water, 20% ethanol, 1% TDS, or 1% stearic acid SL. The concentrations of TDS and stearic acid SL were chosen based on results in previous tests and the solubility of the compounds. Aliquots (20 μL) of the mixtures were placed onto acetone-cleaned 12 mm round cover glass slips (Thermo Scientific, Portsmouth, NH, USA) and allowed to adhere for 30 min. Samples were then covered with 50 μL of 2.5% glutaraldehyde [Electron Microscopy Sciences (EMS), Hatfield PA, USA] and were allowed to fix for 30 min. Samples were then rinsed twice for 30 min each with 2–3 mL 0.1 M imidazole buffer (EMS), and dehydrated in 2 mL ethanol solutions (50, 80, 90, and 100% sequentially). After dried using ethanol, the samples were placed in a Critical Point Drying Apparatus (Denton Vacuum, Inc., Cherry Hill, NJ, USA) using liquid carbon dioxide (Welco Co, Allentown, PA, USA) for approximately 20 min. The samples were mounted onto stubs and sputter gold coated for 1 min (EMS 150R ES), followed by viewing with a FEI Quanta 200 F Scanning Electron Microscope, (Hillsboro, OR, USA) at an accelerating voltage of 10 KV in high vacuum mode.

For TEM examination, Listeria cells were suspended in a 2.5% glutaraldehyde solution (EMS) after treatment with water, 20% ethanol, 1% TDS, or 1% stearic acid SL. The samples were then centrifuged to pellet the cells, and the cells were re-suspended in 10 μL of warm 1% agarose (FMC Bioproducts, Rockland, ME, USA). The pellets were then washed with 0.1 M imidazole solution for 30 min, and fixed with 100 μL 1% osmium tetroxide solution (EMS) in a fume hood for 1 h. The cells were re-suspended and the samples were allowed to stand for 1 h, and dehydrated using graded ethanol solutions of 50, 80, 90, and 100%. Ethanol was then replaced with propylene oxide twice for 5 min. EMbed-812 (EMS) was mixed and used at a 50–100 % solution with propylene oxide starting with 500 μL 50 % mixture. The EMbed 812 plastic was cured in a vacuum oven (VWR, Wayne, PA, USA) at 60°C and 25 inch Hg overnight. Thin sections at approximately 70 nm were cut using a Reichert Ultracut S (Leica, Wien, Austria) with a Diatome Ultra 45° diamond knife (Fort Washington, PA, USA). Sections were collected on a copper 200 mesh grid (EMS) and stained with 1% solution of uranyl acetate (EMS) for approximately 1 min, followed by rinses with water and counter stain with lead citrate for 1 min. Thin sections were observed using a Philips Transmission Electron Microscope CM 12 (Philips, Netherlands) with an accelerating voltage of 80 KV, and imaged with a DVC detector and processed with AMT software (Danvers, MA, USA).

Treatments were repeated three times for each experiment. Data were statistically analyzed using general linear model (GLM) of SAS 8.1 (SAS Institute Inc., Cary, NC, USA). Differences between the means were separated using the Bonferroni–Dunn test (P < 0.05).

Ethanol at 20% did not significantly (P > 0.05) reduce the population of Salmonella spp. within 1 h of contact at 22°C (Figure 2). Within 1 min treatment, none of the free-acid SLs in the presence of 20% ethanol significantly reduced the population of Salmonella spp. in any of the three concentrations (0.1, 0.5, and 1.0%) tested. After 1 h, however, all free-acid SLs at all three concentrations significantly lowered Salmonella populations. There was no significant (P > 0.05) difference among the three concentrations of free-acid stearic and oleic SLs in their effectiveness in reducing the populations of Salmonella while free-acid palmitic acid SL reduced Salmonella populations in a concentration dependent manner.

In general, lactonic SLs were more effective than free-acid SLs in reducing the Salmonella populations (Figure 2). For the 1 min treatment, 1% lactonic stearic SL reduced populations of Salmonella by 3.1 logs while other treatments at any concentration did not significantly reduce the populations. After 1 h treatment, all lactonic SLs at the three concentrations tested significantly (P < 0.05) reduced Salmonella populations. Compared with 20% ethanol, more than 5 log reductions of Salmonella were achieved by all three lactonic SLs at all concentrations after 1 h except that palmitic acid SL at 0.1% reduced the Salmonella population only by 4.1 log CFU/mL.

Thiamine dilauryl sulfate at 0.1% in the presence of 20% ethanol reduced Salmonella populations from 7.3 log CFU/mL to non-detectable level after 1 min of treatments (data not shown).

In the presence of 10% ethanol, none of the free-acid SLs at 1% significantly reduced Salmonella population after 1 h incubation (Figure 3). In the presence of 20% ethanol, significant reductions of Salmonella were achieved regardless of the fatty acid type. The results indicate that relatively high concentrations of ethanol were needed for SLs to significantly reduce Salmonella populations.

Within 1 min treatment, none of the free-acid SLs significantly affected the populations of Listeria spp. (Figure 4). However, after 1 h treatment time, the populations of Listeria were significantly reduced by all three types of free-acid SLs at all three concentrations tested. The populations of Listeria were reduced to the non-detectable level (1 log CFU/mL) by most of the treatments. Compared with their effects on Salmonella (Figure 2), the free-acid SLs were more effective against Listeria after 1 h of treatment. Ethanol itself at 20% did not significantly reduce the populations of Listeria spp. within 1 h.

Lactonic SLs were much more effective than free-acid SLs (Figure 5). After 1 min treatment, free-acid SLs in the presence of 20% ethanol only reduced Listeria population by 0.2–0.8 log CFU/mL while lactonic SLs reduced the populations by more than 5 logs. Initial Listeria population before treatments was 7.2 log CFU/mL.

After 1 min treatment, all three lactonic SLs at 0.1% in the presence of 1% ethanol reduced Listeria population by about 0.4–0.5 log CFU/mL (Figure 6). Lactonic SLs at 0.5% in 5% ethanol and 1.0% in 10% ethanol reduced populations of Listeria by 2.0–2.2 and 2.8–3.4 log CFU/mL, respectively. Increasing contact time (treatment time) from 1 min to 2 h reduced Listeria populations. At 0.5 and 1.0%, lactonic SLs reduced Listeria populations to non-detectable levels after 1 or 2 h of treatment. Stearic acid SL was more effective against Listeria than palmitic or oleic acid SLs as evidenced at 0.1% after 1 and 2 h treatments.

Thiamine dilauryl sulfate at 0.1% in the presence of 20% ethanol reduced Listeria population to a non-detectable level after 1 min of treatment (data not shown). In the presence of 1% ethanol, 0.1% TDS reduced Listeria populations by 3.6 log CFU/mL. Our results suggested that TDS was more effective than the SLs against Salmonella and Listeria spp., similar to our earlier results on E. coli O157:H7 (Zhang et al., 2016).

The SEM images showed that non-treated Listeria cells had typical plump rod shapes with smooth surfaces and various lengths (Figure 7). A few cells treated with 20% ethanol exhibited some shrinkages and depressions. Most lactonic oleic acid SL-treated cells were highly distorted, and had corrugated surfaces with folds, shrinkages, lumps, and protuberances. TDS-treated cells had lumps and protuberances on the surface. Unlike cells treated with SL, TDS-treated cells did not show any folds or shrinkages. The distorted and corrugated surfaces of cells treated with SLs may be results of the cell lysis and leakage of cellular contents while the lumps and protuberances on the surface of TDS-treated cells may be indicative of cellular leakages (Dengle-Pulate et al., 2014).

Transmission electron microscopy images showed the morphology of non-treated and treated Listeria cells (Figure 8). The cell walls and membranes of non-treated Listeria cells were mostly intact, showing well-preserved, dense cell contents. Cells treated with 20% ethanol had a similar morphology as the non-treated cells with occasional lysed cells. Oleic acid SL-treated cells had less defined cell membranes, and often with broken walls and membranes, and increased heterogeneity in electron density in the cytoplasm. It appears that some TDS-treated cells underwent lysis, resulting in the release of their cellular contents into the surrounding environment. It was common to find electron-dense particles or precipitates around damaged bacterial cells that were electron translucent in comparison to undamaged cells (Jung et al., 2008). Our earlier study (Zhang et al., 2016) on E. coli O157:H7 showed that SL and TDS-treated cells had lumps and protuberances on their surface. Similar SL-induced morphological changes (lumps and protuberances) of bacteria cells were observed by other researchers (Dengle-Pulate et al., 2014). TEM images of E. coli O157:H7 also revealed that the outer membrane of E. coli cells treated with SLs and TDS were separated from the plasma membrane (Zhang et al., 2016). Dengle-Pulate et al. (2014) observed, using SEM, that SLs synthesized using lauryl alcohol acted on the integrity of E. coli and S. aureus cell membranes, and the treated cells were disrupted with the outpouring of their cytoplasmic contents, which leads to cell lysis.

Our results demonstrated that Gram-positive Listeria spp. were more sensitive to SLs and TDS than Gram-negative Salmonella spp. The results are in agreement with several earlier reports on other bacteria (Kim et al., 2002; Shah et al., 2007; Dengle-Pulate et al., 2014). Dengle-Pulate et al. (2014) found that SLs synthesized using lauryl alcohol showed complete inhibition against E. coli at 30 μg/mL at a contact time of 2 h whereas for S. aureus, complete inhibition was observed at 10 μg/mL. Kim et al. (2002) reported that SLs showed antimicrobial activity against Gram-positive bacteria: B. subtilis, S. xylosus, S. mutans, and Propionibacterium acne, but no effect on E. coli. Reetika et al. (2012), however, found that linolenic SL mixture (containing 80% lactonic form) was more potent against Gram-negative bacteria than Gram-positive ones, with minimum inhibitory concentrations (MICs) of 20 and 10 μg/mL for B. subtilis and E. coli, respectively. Pulate et al. (2013) showed that SLs, produced using glucose as a hydrophilic source and lauryl alcohol C12–14 as a hydrophobic source by C. bombicola, had potent antimicrobial activity against both E. coli and S. aureus. The difference in antimicrobial activity of TDS and SLs against Gram-positive and Gram-negative microorganisms may be attributed to the structure of their respective cell walls. The cell wall of Gram-negative microorganisms is comprised of a thin peptidoglycan layer and two layers of membranes which control the efflux of substance across the membranes. In contrast, the cell wall of Gram-positive microorganisms is composed of one layer of membrane (cytoplasmic membrane). The cytoplasmic membrane is surrounded by a thick peptidoglycan layer. The peptidoglycan of the cell is sufficiently porous to allow the passing of compounds with molecular mass less than 50 kDa (Lang and Wagner, 1993). The molecular mass of SLs used in our study are typically 500–800 Da and maintain amphiphilic properties, which can allow SLs to enter the peptidoglycan layer and interact with the cytoplasmic membrane (Dengle-Pulate et al., 2014). Our results showed that TDS and SLs caused morphological changes and damage to the membranes of Listeria spp. which might be the cause of cell death.

Palmitic, stearic, and oleic acids were used as fermentation co-substances to synthesize three fatty acid SLs with different lengths (C18 vs. C16) and saturation (saturated and unsaturated). Palmitic acid is the most common fatty acid found in animals, plants, and microorganisms while oleic acid and stearic acid also occur naturally in various animal and vegetable fats and oils. It seems that there is no difference among the three fatty acid SLs in their antimicrobial activities with exception of stearic acid SL which was more effective against Listeria spp. than oleic and palmitic acid SLs at 0.1% after 1 and 2 h of treatments (Figure 6). An earlier study (Shah et al., 2007) found that SLs obtained from different sugar-containing media differed in their activities against several microorganisms including B. subtilis. R. erythropolis, S. epidermidis, and Moraxella spp. SLs from arabinose-containing medium were more effective against three of the four Gram-positive bacteria tested and against Moraxella sp. than SLs from glucose-containing medium. However, SLs from arabinose showed no inhibition of the growth of E. coli while SLs from lactose-containing medium were the most effective compound against B. subtilis.

Sophorolipids and TDS are generally insoluble in water, but soluble in organic solvents such as ethanol. Therefore, to incorporate SLs and TDS into water, the compounds were first dissolved in ethanol before dilution. Our results showed that ethanol itself at concentrations up to 20% had no significant effects on populations of Listeria or Salmonella spp. within 1 h of treatment. Our results further demonstrated that 20% of ethanol was needed for the antimicrobial effects against Salmonella spp. of free SLs at the concentrations tested. De Rienzo et al. (2016) reported that growth of Gram-negative Cupriavidus necator and Gram-positive B. subtilis were inhibited by undefined SLs only at high concentrations (5%). The presence of caprylic acid (0.8% v/v) enhanced the inhibition of biofilm formation of P. aeruginosa, E. coli, and B. subtilis. Kim et al. (2002) reported that SLs had no effect on E. coli. Sleiman et al. (2009) also reported that SLs did not show any significant antibacterial activity in vitro when tested at clinically relevant concentrations. The primary target of ethanol is the cell membrane (Ingram, 1990), and ethanol is known to inactivate bacteria by disrupting cell membrane, and serving as a membrane perturbant (Fried and Novick, 1973). Ethanol may enhance the penetration of SLs and TDS through lipid bilayers of bacteria, especially Gram-negative bacteria which had two membranes (cytoplasmic and outer membranes). Our results suggested that ethanol enhanced antimicrobial activity of SLs against Salmonella and Listeria, and there was a synergistic interaction between SLs and ethanol.

Our results show that SLs, particularly in lactonic form, were effective in inactivating Salmonella and Listeria spp. in bacterial suspension. It is envisioned that the compounds may be used in different ways to minimize the risk of pathogen contamination in various foods. For example, SLs may be incorporated into ready-to-eat meat or cheese products to reduce the populations of L. monocytogenes, a pathogen often associated with the products. Furthermore, the compounds may be used as a wash to minimize pathogen cross-contamination for the handling and processing of fresh and fresh-cut produce. With their strong surfactant property, the antimicrobial SLs could be valuable in the formulation of washing and cleaning solutions. In addition, the compounds may be used as a coating on various foods. However, since SLs and TDS require the presence of 10–20% ethanol for effective antimicrobial activity against Salmonella and Listeria, their commercial applications on foods may be limited. One viable possibility may be for disinfecting food contact surface. Furthermore, the compounds may be combined with other antimicrobials to achieve synergistic effects against human pathogens. The actual effectiveness and applications of these compounds in food need further investigation.

In summary, we synthesized six types of SLs from C. bombicola by using palmitic, stearic, and oleic acids in growth media, and evaluated their antimicrobial activity against two Salmonella and Listeria spp. Our results showed that lactonic SLs were more effective antimicrobials than free SLs, and Listeria spp. were more sensitive to SLs and TDS than Salmonella spp. TDS was a more effective antimicrobial than the SLs. In addition, ethanol enhanced antimicrobial activities of TDS and SLs. Overall, our results demonstrate that SLs and TDS are effective antimicrobials and may be used to inactivate foodborne pathogens.