It has been reported that netted muskmelons (Cucumis melo var. cantalupensis) treated with moist heat (steam or hot-water immersion) have reduced populations of vegetative surface organisms that may be responsible for spoilage, or that may be pathogenic to consumers. It is unknown, however, what affect a similar heat treatment may have on infesting bacterial endospores (which are dormant, nonreproductive structures that are resistant to environmental stress). Also, any heat treatment used must be effective without exceeding the treated melon's thermal damage threshold. In this study, natural microflora on muskmelon rind pieces treated from 75 to 95 °C for 3 minutes and whole fruit rinds inoculated with Bacillus atrophaeus spores and treated at 85 °C for 3 minutes were observed as a model system to explore the efficacy of moist heat in reducing surface populations of bacterial spores. There were significant reductions in populations of aerobic, nonspore-forming microbes, although the treatments had little to no effect on either the recoverable populations of inoculated B. atrophaeus spores or indigenous spore-forming bacteria. Recovery studies suggested a less than 2 log10 unit reduction of inoculated B. atrophaeus spores after a 3-minute, 85 °C moist heat treatment, and no heat injury symptoms developed on melons during storage for 2 weeks at 5 °C. Increasing treatment temperature from 75 to 95 °C resulted in no increase in efficacy in terms of recovery of indigenous vegetative bacteria. The results of this study suggest that aqueous heat treatment is not a suitable method for reducing populations of the resting structures of spore-forming bacteria from the surface of netted muskmelons.
Produce items such as netted muskmelons (Cucumis melo var. cantalupensis) and other melons have been associated with numerous outbreaks of food-borne illness (Centers for Disease Control and Prevention, 1991, 1993, 2002; Gayler et al., 1995; Mohle-Boetani, et al., 1999; Rise et al., 1990). Because the fruit of muskmelons and other melons are in contact with the soil during growth, they may have an increased chance of exposure to soil-resident microorganisms compared with other commodities. This exposure may increase the risk of contamination by various spore-forming bacteria that are plant or human pathogens, such as Bacillus cereus, a common cause of food-borne illness.
Research on the efficacy of sanitizer treatments in decontaminating honeydew (C. melo var. inodorus) or netted muskmelons has shown promising results for pathogens such as Escherichia coli O157:H7, various Salmonella species, and naturally occurring microflora (Annous et al., 2004; Park and Beuchat, 1999). However, when inoculated bacteria have been allowed to remain on the melon surface for more than a few days before treatment, substantially smaller reductions have been obtained (Ukuku et al., 2001a, b). Although chlorine treatment was able to inactivate microbes in solutions or on equipment, chlorination had only minor effects on cells from a five-strain Salmonella cocktail that was inoculated on the surfaces of tomato (Lycopersicon esculentum) fruit (Allen et al., 2005) or E. coli inoculated on iceberg lettuce (Lactuca sativa) leaves (Li et al., 2001).
It has been reported that fruit treated with moist heat (steam or hot-water immersion) have reduced populations of surface organisms that may be responsible for spoilage, or other organisms, such as Salmonella spp., which may be pathogenic to consumers. Fleischman et al. (2001) showed that hot-water immersion was a highly effective means of reducing E. coli O157:H7 from the surface of apples. A 6- to 7 log10 unit decrease was observed in pathogen populations after a 30-s dip at 80 or 95 °C. Pao and Davis (1999) reported that hot-water immersion of oranges (Citrus sinensis) at 80 °C for 1 min, or at 70 °C for 2 min, resulted in an approximate 5 log10 unit reduction of four different strains of E. coli on the orange peel. Ukuku et al. (2004) reported that whole muskmelons surface inoculated with a 3-, 6-, or 8 log10 unit cocktail of Salmonella strains Poona (RM2350), Stanley (HO558), Newport (H1275), Anatum (F4317), and Infantis (F4319), when treated with hot water at 70 or 97 °C for 60 s, resulted in a 2.0- or 3.4 log10 cfu/cm2 reduction of Salmonella, respectively. Any heat treatment used to reduce microbe populations on produce must be effective without exceeding the fruit's thermal damage threshold. In the studies mentioned, the peel of oranges insulated the fruit from heat and avoided damage to the juice, apples (Malus domestica) had poor heat conductivity and treatment did not increase their internal temperatures significantly, and muskmelons had no external evidence of damage.
In studies examining the efficacy of produce sanitizers against human pathogenic bacteria, E. coli O157:H7, Salmonella spp., and Listeria monocytogenes have been the most extensively studied organisms. However, only limited data are available on the effects of moist heat treatments or sanitizers on endospore-forming, Gram-positive bacteria such as members of the genus Bacillus (Bloomfield, 1999). The resting structures of Bacillus spp. are very resistant to high temperature, radiation, and various toxic chemicals (McDonnell and Russell, 1999; Nicholson, et al., 2000; Russell, 1990; Setlow, 2000). Spore germination is influenced by temperature, humidity, pH, and the presence of O2, CO2, and other nutrients. The prevalence and resilience of Bacillus spp. spores to sanitation steps makes them a good indicator organism for a biocide's efficacy against the resting structures of other potentially harmful spore-forming bacteria.
The goal of this study was to assess the efficacy of hot-water treatment, used for the surface decontamination of muskmelons, against the resting structures of spore-forming bacteria. Spores from the organism B. atrophaeus were used in this study as a model system for B. cereus, or other spore-forming human pathogens. The inactivation of indigenous spore-forming and nonspore-forming organisms was also examined.
Materials and Methods
Muskmelons were acquired from a local grocer the day of testing. The origins and the age of the melons were unknown for most studies. For testing water temperature effects on whole melons, muskmelons (imported from Honduras at the one-quarter to full-slip ripeness stage) were preordered for delivery on the day of testing directly from the grocer's distribution warehouse.
Melon temperature profiles.
Temperature probes were used to determine the depth of heat penetration into melons treated at 76 °C in the nonrecirculating water bath used for these studies (18 L, model 285; Precision Scientific Incorporated, Chicago), compared with what is reported in the literature. Flesh temperatures were recorded with thermocouple probes (SuperLogics 8018; SuperLogics, Waltham, MA) that were used at three sampling points: on the surface, between 2 to 3 cm below the surface (pericarp), and within the pericarp adjacent to the rind of the melon (2–4 mm below the surface). Each trial consisted of two probe readings at each depth. One melon was cut in half after being allowed to equilibrate to room temperature (25 °C) for each trial. One melon half was then placed in a nonrecirculating water bath. The waterline was limited to ≈1 cm below the cut surface of the fruit to ensure no water would flow into the cavity. The other melon half was held in air at room temperature as a control. Probes were inserted through the cut surface of the melon until they came to rest at about the required depth below the rind surface. In this manner, the probes did not puncture the submerged surface of the melon. Three such trials were performed.
Temperature effect on background microflora.
Paper templates with uniform, square holes (5 × 5 cm) cut in them were wrapped in aluminum foil, autoclaved, and aseptically stored until needed. Individual templates were placed on the surface of a test melon and a permanent marker was used to delineate the 25-cm2 test area. Enough melons were used to make 53 such outlines. Melon rinds were scored along the marked 25-cm2 outlines using a scalpel and then excised with a flamed knife to take as little pericarp tissue underlying the rind as possible (≈2–3 mm thick). Rind sections were cut in half diagonally to produce two equal triangular pieces, which were then organized into replicate sets of five (≈12.5 cm2 each, ≈62.5 cm2 total/set). Seven treatments were sequentially performed with water temperatures of 23 (room temperature water control), 75, 80, 85, 90, or 95 °C, plus an untreated control (23 °C air control). Treatment replicates consisted of sets of five triangular rind tissue pieces placed into a stomacher bag with 100 mL 0.1 N phosphate buffered saline solution (PBS; pH, 7) at 23 °C. There were three bags per treatment, for a total of 21 bags. Triplicate bags were placed in the water bath such that the buffer solution in the bags was completely submerged. A mercury thermometer was used to monitor the buffer temperature within the treatment bags. When the buffer reached the desired treatment temperature (after ≈45 s), a timer was started. The bags were removed from the water bath after 3 min and were briefly cooled in room-temperature water to quench any further heat-related bactericidal action.
Temperature effect on whole melons.
Melons averaging 1.74 kg (sd, 0.29 kg) were used in this experiment. Nine melons were individually heat treated for 3 min in a 25-L circulating water bath (model 960; PolyScience, Niles, IL) set to 85 °C and verified with a thermocouple temperature probe. Three melons were simultaneously treated in a larger (90-L) recirculating water bath set to 25 °C as a room-temperature control treatment, and an additional three melons were kept untreated as a negative treatment control (total of 15 melons). After treatment, all melons were placed in a walk-in cooler set to 5 °C for storage. A plastic sheet was loosely draped over the fruit to help maintain ≈95% relative humidity (RH) and a data logger was used to verify the RH. After 5 d, all trays were moved to a reach-in chamber set at 5 °C and 85% RH. Melons were observed at 24-h intervals for signs of breakdown, decay, or other visual clues of damage that could lead to loss of marketability. Melons were rated for marketability using a 10-point subjective quality scale based on the hedonic rating system of Kader et al. (1973). Melons rated as ≤3 (Table 1) were culled. All melons were rated for final quality, external and internal, after 14 d of storage. The date and identification of culled melons was recorded for later interpretation.
Subjective scale for the determination of melon quality after 14 d of storage.
Spore activation study.
Spores of B. atrophaeus [formerly B. subtilis ssp. niger (Fritze and Pukall, 2001)] were obtained from a commercial vendor as spore-impregnated sample disks, rated as 107 cfu/disk (catalogue no. 9372 sub-1-7; Raven Laboratories, Omaha, NE). Spores were first tested to determine whether they required a heat treatment for activation, as has been suggested (Leuschner and Lillford, 1999). A sample of spores taken from the stock was added to 10 mL buffer solution (0.1 N PBS; pH, 7) in a test tube. The spore suspension was heated to activation temperatures of 80 to 85 °C and held for 10 min, allowed to cool to the touch, then was serial diluted in sterile PBS and pour plated to tryptic soy agar (TSA; BD Diagnostics, Franklin Lakes, NJ). Spores were also plated to TSA agar without a separate heat activation step. Total dilution factor and the number of colonies observed after storage at 37 °C for 48 h were compared with the reported concentration of the stock solution. Each sample was plated using the pour or spread plate technique to determine differences in recovery.
Rind treatment, with and without spores, at 85 °C for 3 min.
Heat treatments consisted of water bath submersion followed by storage and recovery steps, as suggested by Annous et al. (2004). As previously described, enough melons were used to created 50 diagonally bisected 25-cm2 rind sections, which were placed into weigh boats and grouped into treatment. There were five replicates per treatment, with each replicate consisting of five, 12.5-cm2 rind pieces. All the pieces were placed in a randomized order on sterile paper towels with the outside (netted) surface facing up. One half of the replicate group was spot inoculated with 100 μL 107 cfu/mL B. atrophaeus spores each, and then allowed to air dry visibly (≈20 min). The other half of the replicate groups remained uninoculated. Each five-piece replicate was then placed into a labeled stomacher bag containing 100 mL 0.1 N PBS (pH, 7) that had previously been stabilized at a specified treatment temperature in a hot-water bath (85 °C). A total of four treatments were applied (heat with inoculation, heat without inoculation, unheated with inoculation, and unheated without inoculation), with five bags (replicates) each. After addition of the rind pieces, the buffer temperature of one bag from each treatment in the hot-water bath was monitored with a glass mercury thermometer. When the buffer attained the treatment temperature, a 3-min timer was started. After 3 min, all the bags from that treatment were removed from the water bath, cooled to room temperature by brief submersion in cool tap water, and immediately analyzed for surviving bacteria.
After heat treatment, bags were individually placed in a commercial stomacher (Mix 1; AES Laboratories, Combourg, France) for 2 min to macerate and homogenize the contents. After stomaching, a 1-mL sample was serially diluted three times in 9 mL 0.1 N PBS (pH, 7). A 1-mL sample from each diluted suspension was then pour plated on TSA and incubated at 37 °C. Plates were observed for growth at 24 and 48 h after pouring.
Individual melons in heat penetration studies, groups of three melons in temperature effects, and treatment bags containing melon rind sectionals were treated as replicates. Each repetition in spore recovery or bacterial count study was plated in three replicates and the results were pooled to determine average growth. Data sets were compiled and tabulated, and confidence intervals were calculated using Excel 2004 (ver. 11.3.7, for Mac) using the Analysis Toolpack add-in (Microsoft Corp., Redmond, WA).
Purchased stock of B. atrophaeus spores germinated as expected, confirming initial populations of 107 cfu/disk (data not shown). Results were similar with or without a heat treatment activation step, or when using either pour-plating or spread-plating recovery techniques (data not shown).
Melon surfaces did not reach the water bath temperature of 76 °C after 5 min of submersion; however, the melon surfaces reached ≈80% of the maximum temperate attained within 30 s of submersion (Fig. 1). Subsurface rind temperatures measured ≈3 mm below the surface increased more slowly, reaching a maximum of 50 °C after 5 min, or ≈70% of the maximum surface temperature. The temperature of the flesh 2.5 cm from the rind surface changed slightly, increasing only 4 °C after the 5-min treatment.
After 14 d of storage at 5 °C, whole melons treated at either 25 or 85 °C were rated for quality. On a 10-point hedonic scale (Table 1), all fruit were in the poor (2–3 points) to fair (4–5 points) range of marketability based on external visual appearance. One melon from the treated samples showing evidence of excessive bruising damage on the second day was culled. The average rating for melons treated at 85 °C for 3 min was 3.3 points (se, 0.3 point), melons treated in room-temperature water (25 °C) averaged 4.3 points (se, 0.8 point), and melons that were not treated averaged 3.3 points (se =1.2 points). When melons were culled, or after the 14-d observation period, the pericarp was rated using the same rating scale by cutting the melons in half and observing internal tissues. All melons were internally of significantly higher rating than they were externally, being in the good (6–7 points) to excellent (8–10 points) range. Pericarp of melons that were treated at 85 °C were rated as 6.4 points (se, 0.8), melons treated at 25 °C were rated as 7.7 points (se, 0.3), and melons that were not treated were rated as 8.3 points (se, 0.3). There were no significant differences between treatments at P = 0.05.
Recovery of indigenous bacteria from uninoculated melon rind sections was transformed from plate counts to colony-forming unit per square centimeter of melon rind. All counts from uninoculated rind sections, treated at five different temperatures, were significantly different from controls at P = 0.05 (Fig. 2). However, there were no significant differences in recovery between treatments from 75 to 95 °C, nor were there significant differences between untreated controls and pieces treated in room-temperature (≈23 °C) water.
Recovery of active B. atrophaeus spores from inoculated rind sections after treatment for 3 min at 85 °C was 1 log10 unit less than from inoculated rinds treated at room temperature (Fig. 3). Rinds that were not heat treated, with or without inoculated B. atrophaeus spores, had similar levels of bacterial recovery. Uninoculated melons that were heat treated had 3 log10 units fewer colonies (P < 0.05) than were recovered from nonheat-treated rinds (inoculated or not), and 2 log10 units fewer recovered cells (P < 0.05) than rinds that were heat treated and inoculated.
Samples of bacteria recovered from uninoculated, homogenized melon sections, both before and after heat treatments, were sent to an independent diagnostic microbiology laboratory for identification (ABC Research Corp., Gainesville, FL). The report of remaining organisms and approximate total concentrations of each are presented in Table 2. No spore formers were recovered before heat treatment, and only spore formers (all of various Bacillus spp.) were recovered after treatment.
Identity and approximate total titer of indigenous colonies recovered before and after heat treatments of uninoculated melons.
Discussion and conclusion
Heat treatments to obtain melon temperature profiles were performed using melon hemispheres. The hemispheres were used so probes could be placed through the exposed pericarp tissue to reach the desired depths relative to the fruit surface without puncturing the rind subsurface and allowing treatment water to infiltrate the melon. As heat treatments were applied over the 5-min trials, surface temperatures increased rapidly, whereas subsurface temperatures increased more slowly. Core temperatures were largely unaffected by hot-water exposure at the melon surface. These results were similar to observations reported by Annous et al. (2004), who stated that the rind seemed to “… [insulate] the flesh from extreme temperatures” when whole melons in steam chambers were exposed to temperatures similar to those used in this study. It was also noted in the same report that if the flesh (at 7 mm or deeper) reached ≈70 °C or more during steam treatment, rind tissue necrosis was observed. This was reported for melons that had been treated at 96 °C for 4 min. For these reasons, none of the melons in this study were exposed to temperatures in excess of 95 °C for 3 min. Also, the time/temperature combination of 85 °C for 3 min was selected for use in the inoculation/recovery trials, because it was less than the temperature that had been reported to cause melon damage, but was still efficacious in reducing vegetative bacterial cells.
The inability of the noncirculating water bath model to raise external rind temperatures successfully to the desired water temperature demonstrates the need for commercial hot-water treatment systems for bulky produce items, such as melons, to have sufficient heating capacity and transfer capability to overcome the thermal load of the product. Any commercial application should use a heating system sufficiently designed to reach target temperatures in the first few millimeters of tissue, subjacent to the fruit surface, in the least amount of time possible, disallowing heat to transfer deeper into the tissue. Excessive heating times, potentially allowing deeper penetration of heat, may lead to injury of fruit tissues.
Melons treated at 85 °C for 3 min to determine the effect of that time/temperature treatment on marketability suggest that melons are inherently resistant to damage incited by moist heat treatments, as has been reported previously (Annous et al., 2004; Ukuku et al., 2001b, 2004). Only one melon had significant damage that caused it to be culled before the end of the shelf life study. This melon, upon being cut open, had evidence that it was bruised before treatment, leading to its poor quality. After the 14-d storage period, when the remaining melons were cut open for observation, visual inspection suggested a minimum of “good”-quality pericarp. Such melons, although likely to be culled based on whole-melon rind observations, may be candidates for minimal processing or value-added sales, such as ready-to-eat packaging, and so forth.
For activation, B. atrophaeus spores only required the presence of nutrients and moisture, not a heat activation step. Spore requirement of a heat activation step may be necessary in a minimal media, such as when suspended in water (Leuschner and Lillford, 1999), but in a location with readily available nutrients (e.g., TSA plate or muskmelon surface), moisture and time appear to be the only other requirements for germination, survival, and replication.
In this study, 85 °C moist-heat treatments failed to produce greater than a 1 log10-unit reduction in recoverable spore-forming bacteria. This is less than has been reported for nonspore-forming bacteria subjected to similar treatments (Annous et al., 2004; Ukuku et al., 2001b). Heat treatments at lower temperatures (76 °C) have been suggested to be efficacious in controlling nonspore-forming bacteria (Annous et al., 2004). Although such treatment may not necessarily increase the germination rates of any bacterial spores that could be present, it would likely have little to no effect in reducing any such spore populations. This is not surprising, because a reduction of only 0.39 log10 unit of B. anthracis (for which B. atrophaeus is often used as a surrogate) spores was observed after heat treatment in various aqueous suspensions (deionized water, skim milk, and brain–heat infusion broth) at 78 °C for 90 min (Novak et al., 2004).
It should also be noted that we did not observe a reduction of populations of recoverable, indigenous background microflora below the minimum detection limit of 2 log10 cfu/mL after moist-heat treatments of up to 95 °C for 3 min (Fig. 2). The log reduction of background microflora may be as great as 3 log10 units (Fig. 3) or as little as 1 log10 unit (Fig. 2). In both cases, population recovery was near 3 log10 units, with the different reduction levels based entirely on the initial native background population. This is consistent with other observations in which hot-water treatments did not reduce recovery to less than 3 log10 units (Ukuku et al., 2004). Also, as the temperature of heat treatment increased, the level of recoverable background microflora was not significantly different (P < 0.05). This suggests that a certain portion of vegetative cells may be shielded from the effects of the moist-heat treatment, perhaps by becoming embedded in the porous netting tissues of the melon. This is consistent with previous observations when significant reductions of treated populations on smooth surfaces, such as processing surfaces (Li et al., 2001) and less porous fruit (Fleischman et al., 2001; Pao and Davis, 1999) were attained, but similar treatments on porous surfaces, such as the netting of a melon, had significantly lower efficacy rates (Ukuku et al., 2001b).
Although moist-heat treatment may reduce vegetative bacterial populations on the surface of treated melons, as observed in this study (Fig. 3) and as reported by others (Annous et al., 2004), it cannot be presumed to significantly reduce populations of those bacteria capable of producing resting structures resilient to heat, such as the spores of B. atrophaeus. It thus follows that other spore-forming bacteria, including other members of the genus Bacillus, such as B. cereus, a common cause of food-borne illness, would not be controlled by heat treatment.
Centers for Disease Control and Prevention1993Multistate outbreaks of Salmonella serotype Montevideo infections. Publication EPI-AID 93-79Centers for Disease Control and PreventionAtlanta, GA