In-home or food service antimicrobial treatment options for fresh produce are limited. Hot water treatments for whole (unpeeled) produce have been proposed, but data to support this practice for onions are not available. Separate cocktails of rifampin-resistant Escherichia coli O157:H7, Listeria monocytogenes, and Salmonella were cultured on agar and suspended in sterile water. The outer papery skin at the equator or root or stem ends of the whole yellow onions was spot inoculated at 6 log CFU per onion. After drying for 30 min and, in some cases, storage at 4°C for 6 days, onions were immersed in water at ca. 100°C for 5 s or 85°C for 10 to 180 s. No significant difference (P > 0.05) in the mean decline of Salmonella was found on onions that were exposed to hot water after drying the inoculum for 30 min or after storage at 4°C for 6 days. Exposure of whole onions at 100°C for 5 s reduced E. coli O157:H7 and L. monocytogenes populations by >5 log CFU per onion at all inoculum sites and Salmonella populations by >5 log CFU per onion at the stem end and equator but not consistently at the root end. Mean root-end reductions of ≥5 log CFU per onion of E. coli O157:H7, L. monocytogenes, and Salmonella were achieved consistently when the root end was fully immersed in 85°C hot water for 45 or 60 s except in a small number of cases (4 of 57; 7%) when the root end was oriented upward and above the water line during treatment. When onions were held at 85°C for 180 s with the root end above the water line in an uncovered water bath, no significant declines in Salmonella populations were observed; significant mean declines in Salmonella were achieved (mean, 5 log CFU per onion; range, 3.49 to 6.25 log CFU per onion) when the water bath was covered. Short exposure to hot water can significantly reduce pathogens on the surface of whole onions. Reductions are more consistent when the root end is submerged and when the water bath is covered.
Short hot-water exposure significantly reduces pathogen populations on whole onions.
Salmonella reductions were greater at the stem and equator than at the root end.
Pathogen reduction was greater for fully submerged root ends in covered water baths.
Onions (Allium cepa) can be purchased as fresh cut or frozen, can be prepared in-house from whole bulbs (by a restaurant or consumer), or can be prepared at a food service distribution kitchen, which then ships the refrigerated products to various customers in the supply chain. Onions are used in cooked dishes and are consumed raw as an ingredient or garnish. Although reported foodborne illness outbreaks associated with onions are not common, uncooked diced onions prepared and used as a garnish for hamburgers at a fast-food chain restaurant were linked to an outbreak of Escherichia coli O157:H7 gastroenteritis in Canada in 2009 (19). In 2020, a large multistate (United States) and multiprovince (Canada) Salmonella Newport outbreak was linked to consumption of whole red onions (5, 33). In 2016, a multistate outbreak of listeriosis was linked to frozen vegetables, including frozen chopped onions (2, 4). Isolation of Salmonella (30) or Listeria monocytogenes (3, 29) from product and environmental samples has led to recalls of fresh-cut onions alone or in vegetable mixtures, and recovery of presumptive L. monocytogenes and Salmonella isolates resulted in a recall of frozen sautéed diced onions (and other vegetables) (31).
No published data are available on prevalence or levels of foodborne pathogens that might be associated with bulb onions. The outer surface of the onion bulb would most likely be contaminated in the field, during harvest, or during postharvest handling (33). The survival of foodborne pathogenic bacteria or surrogate organisms introduced to growing onion plants has been evaluated both in greenhouses (7) and in the field (11, 18, 35). Onions are typically dried or cured during or after harvest, allowing the neck and the skin of the bulb to dry and increasing onion shelf life by restricting ingress of plant pathogens and thus reducing decay during storage (25). Curing leads to a decrease in microbial populations, including surrogate organisms applied via drip irrigation (7, 18).
After harvest, bulb onions are often stored at 0°C but they are commonly shipped and retailed at ambient temperatures (26). When E. coli O157:H7, L. monocytogenes, or Salmonella were inoculated onto the outer skin of whole onions, >4-log reductions were observed after 2 weeks of storage at 23°C and >3-log reductions were observed after 8 weeks of storage at 4°C (16, 17).
Onions are not commonly treated in the kitchen or food service environment prior to peeling and preparation. Although the onion skin is usually removed prior to consumption, contaminants on the skin can transfer to the flesh during preparation, as occurs with other types of produce from which the skin or peel is removed prior to consumption (12, 34). Populations of E. coli O157:H7 and Salmonella increased significantly in chopped onions held at ambient temperature but not during 6 days of storage at 4°C (17). Populations of L. monocytogenes increased significantly in chopped onions during storage at ambient temperature or at 10°C; however, populations did not change significantly over 28 days of storage at 4°C (16).
Washing under running water while rubbing and then drying with a clean towel is often recommended (e.g., fightbac.org) to consumers or in food service settings before preparing whole produce for consumption. For some produce types, especially those with smooth surfaces, significant reductions in microbial levels can be achieved with these methods (13, 15, 22, 23, 32). However, data on the impact of washing whole onions on surface microbial reduction are not currently available.
Hot water treatments are used during postharvest processing of some types of fruits or vegetables to inhibit insect damage, reduce plant pathogens, and extend produce shelf life (21, 28). Hot water dips can decrease postharvest incidence of disease and decay on produce (8). Hot water treatments also have been evaluated in the home, at retail, and in commercial settings as a means to reduce levels of foodborne pathogens on the surfaces of various fruits and vegetables: apples (9, 10), cantaloupe (1), celery and cucumbers (10), oranges (20), tomatoes (24), and peas, spinach, broccoli, potatoes, and carrots (6). In the present study we evaluated the efficacy of hot water treatments for reducing foodborne pathogens inoculated onto whole onions.
MATERIALS AND METHODS
Whole yellow onions (five to eight onions in 1.36-kg mesh bags or individual onions in bulk bins) were purchased from local retail stores (Davis, CA) and stored for up to 1 day under ambient conditions (ca. 23°C; ca. 45% relative humidity [range, 30 to 60%]) before use. The onions weighed 387 ± 52 g (range, 290 to 487 g).
The rifampin-resistant strains used in this study were the same as those previously used with whole and fresh-cut onions (16, 17): E. coli O157:H7 strains CDC658, EC4042, F4546, H1730, and Odwalla strain 223; L. monocytogenes strains LIS 0077, LIS0087, LIS0110, LIS0133, LIS0234, and LIS0235; and Salmonella enterica serovars Agona, Enteritidis PT 30 (ATCC BAA-1045), Gaminara (F2712), Michigan, and Montevideo (G4639). All cultures were stored at −80°C in tryptic soy broth (TSB) supplemented with 15% glycerol.
Individual frozen stock cultures were streaked onto tryptic soy agar (TSA) supplemented with 75 μg/mL rifampin (TSAR; Gold Biotechnology, St. Louis, MO) and incubated overnight at 37°C. Isolated colonies were then transferred into 10 mL of TSB and incubated overnight at 37°C; this step was repeated twice. The second overnight liquid culture was plated (1 mL) onto large (150 by 15 mm) TSAR plates and incubated at 37°C for 24 h. To suspend cultures, 9 mL of ultrapure water (Milli-Q Advantage A10, MilliporeSigma, Burlington, MA) was added to each plate, and the cell lawn slurry on the plate surface was scraped with an L-shaped spreader. Separate cocktails of each pathogen type were prepared by combining equal amounts (5 mL) of each strain. Serial dilutions were made in ultrapure water to achieve target populations. Unless otherwise specified, all culture media were from Difco (BD, Franklin Lakes, NJ).
Onion preparation and inoculation
Circles (3.3-cm diameter, 9-cm2 area) were drawn onto the outermost papery surface of individual whole yellow onions at the equator (middle) or at the root or stem end to define areas to be inoculated. Prepared E. coli O157:H7, L. monocytogenes, or Salmonella inoculum was spot inoculated (10 1-μL spots) randomly within each circle to give an initial target level of 6 log CFU per onion, then the onions were held in a biosafety cabinet for 30 min to allow the inoculum to dry. Inoculation of the root end included root tendrils and skin within the marked area, and inoculation of the stem end included skin at the base of the stem and within the marked area. Hot water treatments were performed immediately after drying. For onions inoculated on the root end with Salmonella, hot water treatments also were performed after storage at 4°C for 6 days.
Hot water treatment
Whole inoculated onions were placed individually in a sterilized metal strainer and then lowered into a water bath (High-Temp Bath 160A, Thermo Fisher Scientific, Waltham, MA) containing 3 L of ultrapure water. No attempt was made to either orient the onion or facilitate mixing. Orientation was noted for each onion at the time of treatment. In some cases, the water bath was covered with an aluminum baking tray (41 by 41 cm; Wilton, Darien, IL).
Each onion was exposed to water heated to 100 ± 1°C for 5 s or 85 ± 1°C for 10 to 180 s, timed from release of the onion into the hot water. Water temperature was monitored using a handheld data logger thermometer (HH378 series thermocouple type K, Omega, Norwalk, CT). After the allotted time, each onion was recovered from the water bath with sterilized metal tongs and placed in a 1,630-mL Whirl-Pak bag (Nasco, Modesto, CA) containing 200 mL of cold TSB (ca. 4°C).
For some experiments, whole onions were inoculated on the root end and then individually placed, root end up, within a 500-mL flask platform clamp (four-pronged holder attachment for a shaking incubator). The base of the apparatus used to hold the onions was weighed down with metal washers and placed at the bottom of the water bath. The volume of water in the bath was increased to 3.75 L to ensure that onions were at least 65% submerged in the water (based on height of each onion), which simulated the degree of floating that occurred when onions were placed freely into the water.
Bacterial recovery and enumeration
Sample bags containing the whole single onions in cold TSB were shaken by hand for 30 s and then rubbed for 30 s. When necessary, appropriate dilutions of the TSB from the prepared samples were made in 0.1% peptone. All samples were spiral plated in duplicate (Autoplate 4000, Advanced Instruments, Norwood, MA). Uninoculated control samples were plated onto TSA to determine levels of the native mesophilic microbiota, and inoculated samples were plated onto nonselective TSA supplemented with rifampin (75 μg/mL) and cycloheximide (50 μg/mL) and selective media CHROMagar O157 (CHROMO157), CHROMagar Listeria (CHROML), or CHROMagar Salmonella (CHROMS) (DRG International, Springfield, NJ), all of which were supplemented with 75 μg/mL rifampin. For samples treated with hot water, 50 mL of each sample was filter plated onto selective media using 0.45-μm-pore-size filters (Thermo Fisher Scientific). All agar media were incubated at 37°C for 24 h (TSA, TSAR, CHROMO157, and CHROMS) or 48 h (CHROML) prior to counting colonies.
When colony counts were expected to fall below the limit of detection by plating (LOD; 0.60 log CFU per onion) for the whole onion samples, 250 mL of TSB supplemented with 50 μg/mL rifampin was added to each sample, and samples (including the onion) were incubated at 37°C for 24 h. Presence or absence of the inoculum was detected by plating 50 μL of enriched sample onto selective media and incubating at 37°C for 24 or 48 h (for L. monocytogenes only).
Impact of treatment on onion flesh
Preliminary tests were conducted to evaluate the impact of hot water exposure on the visual appearance and texture of the onion flesh. After each hot water exposure time, a single uninoculated onion was removed from the water bath and cut into quarters. One to three outer layers of each onion were removed, evaluated visually for transparency, and twisted to determine malleability. The transparency and flexibility of the layer(s) were compared with those of the control and scored as either the same or different.
Three onions for each inoculation location (equator, root, and stem) were evaluated at each time point. Each experiment was replicated at least twice to give a total of six to nine analytical units at each time point, from which the mean and standard deviation (SD) were determined. Two uninoculated control samples were also evaluated for each experiment (n = 2 to 10). Values of 0 were assigned to whole onion samples when plate counts were below the LOD (4 CFU per onion [0.60 log CFU per onion]) and enrichment cultures were negative. A value of 3 CFU per onion (0.48 log CFU per onion), that is, 1 CFU below the LOD, was assigned to samples that were below the LOD after plating but enrichment cultures were positive. This approach was used so the data could be plotted on graphs and mean values could be calculated. The Each Pair Student's t test performed with JMP Pro 14 software (SAS Institute, Cary, NC) was used to compare survival of Salmonella on the root end of whole onions exposed to 85°C water at various times immediately after drying or after storage at 4°C for 6 days. Differences in the mean values were considered significant at P < 0.05.
RESULTS AND DISCUSSION
Hot water treatments: onion orientation
When the whole onions were immersed in water they floated, leaving the top ca. 2.5 cm of onion above the water line. The onions oriented most often with the root down and under the water and the stem end up and above the water. However, in a small number of cases, the onions oriented in the opposite direction with the root facing up and out of the water. Across all experiments, for 53 (93%) of the 57 onions the orientation was root down and stem up, whereas for 4 onions (7%) the root end was up and above the surface of the water.
Hot water treatments: background microbiota
Different lots of onions were used for each experiment. Initial populations of native mesophilic microbiota on whole yellow onions recovered on TSA were 5.48 to 7.31 log CFU per onion (mean ± SD, 6.38 ± 0.61 log CFU per onion) (Fig. 1), 6.18 to 7.95 log CFU per onion (6.88 ± 0.94 log CFU per onion) (Fig. 2), and 5.14 to 6.87 log CFU per onion (6.07 ± 0.87 log CFU per onion) (Fig. 3).
Insignificant mean reductions (P > 0.05; mean, 0.56 log CFU per onion) in populations of native microbiota were observed when whole onions were exposed to 100°C water for 5 s (Fig. 1). At 85°C, mean declines in native microbiota of 0.44 to 1.47 log CFU per onion were observed across 10 to 180 s of exposure (Figs. 2 and 3). Mean differences compared with time zero were insignificant (P > 0.05) at shorter exposure times (10, 15, and 20 s) (Fig. 2). At longer times, mean differences compared with time zero were either insignificant (45 and 60 s) (Fig. 3) or were significant (P < 0.05) at 30 s (1.27 log CFU per onion), 45 s (1.30 log CFU per onion), and 180 s (1.47 log CFU per onion) (Fig. 2). The differences among these experiments might be due to the use of separate lots of onions, each of which likely had unique initial microbial diversity.
Hot water treatments: inoculated onions
After inoculation and drying, mean initial pathogen levels recovered from the equator, root end, and stem end of inoculated whole yellow onions were 5.39 ± 0.62, 5.80 ± 0.19, and 6.00 ± 0.26 log CFU per onion, respectively, for E. coli O157:H7; 5.74 ± 0.50, 5.96 ± 0.25, and 6.18 ± 0.33 log CFU per onion, respectively, for L. monocytogenes; and 5.58 ± 0.17, 5.86 ± 0.28, and 5.82 ± 0.34 log CFU per onion, respectively, for Salmonella (Fig. 1). Populations of E. coli O157:H7, L. monocytogenes, and Salmonella inoculated at the equator or stem end declined by >5 log CFU per onion after exposure to 100 ± 1°C water for 5 s. When inoculated at the root end and exposed to the same time and temperature, >5-log reductions were observed for E. coli O157:H7 and L. monocytogenes but not consistently for Salmonella. After exposure to 100°C water for 5 s, Salmonella populations ranged from 2.82 log CFU per onion (one onion) to below the LOD by plating (0.60 log CFU per onion) but positive by enrichment culture (three onions; >5-log reduction) or to below the LOD by plating and negative by enrichment culture (two onions; >5-log reduction) (Fig. 1). All onions exposed to 100°C, including those positive by plating or enrichment for Salmonella, had oriented in the water bath with the root down.
To evaluate the impact of time between inoculation and thermal treatment, onions inoculated with Salmonella were held for 30 min under ambient conditions only or for 30 min under ambient conditions and 6 days at 4°C prior to treatment (stored onions). For the inoculated onions that were not stored, preliminary experiments evaluated the survival of Salmonella after exposure to water at 85°C for 5 s. Mean populations declined from 6.02 ± 0.23 to 2.96 ± 1.03 log CFU per onion after 5 s (ca. 3-log reduction); longer exposure times were evaluated in subsequent experiments.
Mean population levels of Salmonella were 5.97 ± 0.26 log CFU per onion at 30 min after inoculation and 5.59 ± 0.26 log CFU per onion after 6 days of storage at 4°C (Fig. 2). In all but one case, after 60 s of exposure to 85°C water, Salmonella was below the LOD by plating (0.60 log CFU per onion; >5-log reduction) (Fig. 2). Three of six samples that were not stored and one of five stored samples were positive by enrichment culture. For a single stored sample, the root was oriented up during the 60-s treatment; Salmonella on this sample was 2.95 log CFU per onion. After 180 s of exposure, no samples were positive by enrichment culture; none of the 12 onions tested had oriented in the water bath with the root up. No significant difference in the average Salmonella reductions were found irrespective of onion storage (P > 0.05) at any of the time points. In another study, Salmonella reductions were also significantly impacted by storage time when Salmonella was inoculated onto cantaloupe surfaces and stored at 5°C for either 3 or 5 days before exposure to 70 or 95°C water (28). Because declines in population of pathogenic bacteria were consistently lower on the root end than at other sites and storage did not influence survival, all other experiments were performed with onions that were inoculated on the root end and were dried for 30 min without further storage.
Onions inoculated separately at the root end with E. coli O157:H7, L. monocytogenes, or Salmonella and held for 30 min at ambient temperature were exposed to 85°C water for 45 and 60 s. Mean initial pathogen levels recovered from the onions were 5.89 ± 0.32, 5.51 ± 0.43, or 6.10 ± 0.17 log CFU per onion, respectively (Fig. 3). Mean reductions of ≥5.0 log CFU per onion were observed at both 45 and 60 s for all samples except for two onions that were inoculated with E. coli O157:H7 and had oriented with the root up (45-s treatment). Levels of E. coli O157:H7 recovered from these two samples were 2.06 and 2.79 log CFU per onion (3.83- and 3.10-log reductions, respectively) (Fig. 3). For E. coli O157:H7, L. monocytogenes, and Salmonella, one, two, and four of nine samples at 45 s, and three, two, and one of nine samples at 60 s, respectively, were positive by enrichment culture. One sample inoculated with Salmonella oriented with the root up; however, the root end remained mostly submerged in the water for the full 60-s exposure time, and this sample was negative by enrichment culture.
Similar results have been achieved with other types of produce. When apples were surface inoculated with E. coli O157:H7, >5-log reductions were observed after exposure to 80°C water for ≥15 s. However, when apples were inoculated by submersion, only 2-log reductions of E. coli O157:H7 were obtained, even after 60 s of exposure to 80 and 95°C water, or 4.2-log reductions when apples were dip inoculated with L. monocytogenes and exposed to boiling water for 25 s (10). When celery was inoculated by submersion, 1.2-log reductions were obtained after exposure to boiling water for 25 s (10). The researchers hypothesized that these limited reductions were due to internalization of a portion of the inoculum.
For rind-inoculated cantaloupe, 3.6- to >5-log reductions of E. coli and Salmonella were observed after 60 to 180 s of exposure to hot water (76 to 97°C) (1, 27, 28) or steam (99°C) (14). In whole oranges, ca. 5-log reductions of E. coli inoculated at the stem scar, overall surface, and blossom end were observed after 120 s of exposure to 80°C water (20). Similar reductions of L. monocytogenes were achieved on whole cucumbers after exposure to boiling water for 25 s (10). Differences among these study results can be attributed to differences in target pathogens or strains, water volume, overall structural topography, size, and temperature of the produce item.
Hot water treatments: inoculated onions, root end up
In a preliminary trial, smaller onions did not fit in the flask apparatus and freely floated root down. Larger onions were constrained by the apparatus and remained oriented with the root up but were positioned higher above the water line than were free-floating onions. Onions with a diameter of >6.0 cm (6.3 to 8.5 cm; mean ± SD, 7.38 ± 0.47 cm) and a height of ≤9.0 cm (6.0 to 9.0 cm; 7.18 ± 0.56 cm) were used in these experiments because they remained oriented with the root up and above the water line at a height similar to that of free-floating onions.
Mean initial levels of Salmonella inoculated onto the root of whole yellow onions were 5.86 ± 0.21 log CFU per onion (Fig. 4). No significant mean decline (P > 0.05) in the population of Salmonella was observed during 180 s of exposure to 85°C water. Mean levels of Salmonella recovered from the onions were 5.66 ± 0.47 and 5.39 ± 1.36 log CFU per onion after 90 and 180 s of exposure to 85°C, respectively. Significant declines of ≥2 log CFU per onion were observed in single onions after 45, 60, and 180 s of exposure.
In a separate experiment, the impact of covering the water bath was evaluated. Onions oriented with the root up were held for 180 s in a water bath that was left uncovered or was covered. Mean levels of native microbiota did not change significantly when the water bath was left uncovered but decreased significantly (P < 0.05) by 0.78 log CFU per onion when the water bath was covered (Fig. 5). The mean initial level of Salmonella inoculated onto the root of whole yellow onions was 6.25 ± 0.22 log CFU per onion (Fig. 5). No significant change (P > 0.05) in the mean Salmonella population was observed at 180 s (6.49 ± 0.14 log CFU per onion) (Fig. 5) when the water bath remained uncovered during heating. The Salmonella population declined significantly (P < 0.05) to 1.37 ± 0.99 log CFU per onion after 180 s (mean 4.88-log reduction) when the water bath was covered during heating. Population declines from the mean initial level for individual onions ranged from 3.49 to 5.55 log CFU per onion.
Impact of hot water treatments on onion flesh
When exposed to 100°C water for 5 s, the first layer of onion flesh under the skin was visibly indistinguishable from that of the control onions. After exposure for 60 and 240 s, the outer layer and outer three layers of flesh (but not those below), respectively, were visibly distinct (more translucent) and could be easily bent without breaking. At 85°C for 10 to 60 s, the onion flesh exposed to water was indistinguishable from that of the control samples. After 180 s, only the outer layer of onion flesh was visibly distinct (more translucent) and could be easily bent without breaking.
For produce that is intended to be consumed raw, hot water treatments may be a useful risk-reduction strategy when the outer peel is not consumed and potential surface damage can be tolerated. The root and stem end of onions and the outer skin are typically removed and discarded, which should significantly reduce microbial populations. However, cross-contamination from hands and cutlery or processing equipment during onion preparation would remain a concern. The potential for cross-contamination from contaminated onion skin to the edible flesh during preparation was not evaluated in this study. However, in previous studies hot water treatments significantly reduced E. coli levels in juice extracted from E. coli–inoculated oranges that had been exposed to hot water (20) and reduced the transfer of Salmonella onto fresh-cut pieces of cantaloupe from contaminated rind (28).
In the present study, minimal visible damage to onion flesh was observed during hot water treatments of whole onions that resulted in significant reductions of E. coli O157:H7, L. monocytogenes, and Salmonella. A hot water treatment for whole onions might be useful in a home setting, especially with individuals at higher risk (10), or in a food service setting when large numbers of onions are being prepared for use in a final raw form. When scaled, the hot water treatment might be a useful process control in a fresh-cut or frozen product operation (6). Although these treatments can be effective, time, temperature, and pathogen location influence their effectiveness, and standard operating procedures that include a means for complete and continuous submersion of the onions or covering of the water bath is recommended.
This research was supported by the U.S. Food and Drug Administration (FDA) of the U.S. Department of Health and Human Services (HHS). The contents of this article reflect the opinions of the authors and do not necessarily represent the official views of nor an endorsement by the FDA, HHS, or the U.S. Government. Special thanks are given to Sylvia Yada for her editorial assistance.