Fate of Salmonella Typhimurium and Listeria monocytogenes on Whole Papaya during Storage and Antimicrobial Efficiency of Aqueous Chlorine Dioxide Generated with HCl, Malic Acid or Lactic Acid on Whole Papaya

Papaya-associated foodborne illness outbreaks have been frequently reported worldwide. The goal of this study was to evaluate the behavior of Salmonella Typhimurium and Listeria monocytogenes on whole papaya during storage and sanitizing process. Fresh green papayas were inoculated with approximately 7 log CFU of S. Typhimurium and L. monocytogenes and stored at 21 or 7 °C for 14 days. Bacteria counts were determined on day 0, 1, 7, 10 and 14. Fresh green papayas inoculated with approximately 8 log CFU of the bacteria were treated for 5 min with 2.5, 5 and 10 ppm aqueous chlorine dioxide (ClO2). The ClO2 solutions were generated by mixing sodium chlorite with an acid, which was HCl, lactic acid or malic acid. The detection limit of the enumeration method was 2.40 log CFU per papaya. At the end of storage period, S. Typhimurium and L. monocytogenes grew by 1.88 and 1.24 log CFU on papayas at 21 °C, respectively. Both bacteria maintained their initial population at inoculation on papayas stored at 7 °C. Higher concentrations of ClO2 reduced more bacteria on papaya. 10 ppm ClO2, regardless the acid used to generate the solutions, inactivated S. Typhimurium to undetectable level on papaya. 10 ppm ClO2 generated with HCl, lactic acid and malic acid reduced L. monocytogenes by 4.40, 6.54 and 8.04 log CFU on papaya, respectively. Overall, ClO2 generated with malic acid showed significantly higher bacterial reduction than ClO2 generated with HCl or lactic acid. These results indicate there is a risk of survival and growth for S. Typhimurium and L. monocytogenes on papaya at commercial storage conditions. Aqueous ClO2 generated with malic acid shows effectiveness in inactivating the pathogenic bacteria on papaya.


Introduction
Papaya (Carica papaya) is one of the major tropical agricultural commodities amongst banana, mango, avocado and pineapple [1]. Annual global papaya production has increased by approximately 90% since 2000 and reached 13.7 million metric tons in 2019 [2]. The top three papaya-producing countries are India, Brazil and Mexico, among which 99% of Mexican papayas are exported to the United States [2]. However, along with the increased papaya demand and production worldwide, foodborne illness outbreaks linked to papaya have also been emerging in recent years [3,4]. In particular, outbreaks associated with whole fresh papaya have been frequently reported in the U.S. from 2011 to 2019, which affected the papaya industry in both US and Mexico [4,5]. Papaya grows best in tropic environments at 21-33 • C where the survival and growth of pathogenic bacteria are favored [6]. Microbial contamination of papaya might happen at any step of the production chain where the fruits are in contact with water, soil, harvest equipment and human handling [7]. Salmonella Litchfield was detected on whole papayas associated with an outbreak lactic acid, were more stable and more lethal to Bacillus cereus spores than ClO 2 formed using HCl. Our previous study has also shown that aqueous ClO 2 generated by mixing NaClO 2 with organic acids, including citric acid, lactic acid and malic acid, had higher antimicrobial efficacy against common foodborne pathogenic bacteria on Romaine lettuce than ClO 2 generated with inorganic acids [38]. For example, 5 min treatments with 5 ppm ClO 2 generated with lactic acid, citric acid and malic acid reduced S. Typhimurium on Romaine lettuce by 0.92, 1.39 and 1.37 log CFU/g, respectively, whereas lettuce treated with ClO 2 generated with HCl and sodium bisulfate reduced S. Typhimurium by 0.71 and 1.14 log CFU/g, respectively [38].
In numerous studies investigating the survival of foodborne pathogenic bacteria on fresh produce or decontamination of fresh produce using sanitizers, procedures used to recover and quantify bacteria cells from fresh produce vary. The ununiformed procedures make it difficult to compare and accurately interpret results of different studies [39]. For example, pummeling using a stomacher resulted in higher bacteria recovery than pulsifying, sonication and shaking by hand from iceberg lettuce, perilla leaves, cucumber and green pepper, while a lower level of bacteria was recovered from cherry tomato due to its acidity [40]. Sample preparation method, bacteria type and produce type may affect the efficiency of bacteria recovery and hence further affect the accuracy of a microbiological method. So far, there has been no recommendation of sample preparation methods specifically for whole papaya.
This study aimed to optimize homogenization parameters and enumeration methods for recovering S. Typhimurium and L. monocytogenes from papaya surface. It also sought to evaluate the behaviors of these pathogenic bacteria on whole papaya during storage and sanitizing process. Obtaining information in this regard would assist the papaya industry in selecting optimal sanitizer type, usage concentration and treatment time for papaya washing and sanitizing.

Bacterial Strains and Cell Cultures
Salmonella Typhimurium (ATCC 14028) and Listeria monocytogenes (F2365) were obtained from Food Microbiology Lab at the University of Hawaii at Manoa and stored in trypticase soy broth (TSB; Becton Dickinson, Franklin Lakes, NJ, USA) containing 50% glycerol at −80 • C. Working cultures were prepared by transferring 50 µL of stock culture into 5 mL of sterile TSB and incubating at 37 • C for 24 h. Working cultures were transferred twice in TSB before each experiment.

Preparation of Papayas and Inocula
Fresh papayas (Carica papaya L.cv. Rainbow Solo) were purchased on the day of experimentations on separate occasions from local grocery stores in Honolulu, USA. Non-injured whole papayas at mature green/color break stage were selected according to the maturity chart [41]. Papayas were rinsed with tap water and dried on a lab bench at room temperature for 1 h. Then an area of 2.5 × 2.5 cm 2 on the middle part of the fruit surface was marked with a thin-line non-toxic marker (Sharpie, Oak Brook, IL, USA). The marked whole papayas were placed on sterile Petri dishes in a biosafety hood before experimenting. S. Typhimurium and L. monocytogenes cultures were diluted with 0.1% peptone water (Becton Dickinson, Franklin Lakes, NJ, USA) to desired concentrations. 100 µL of the inoculum was spot inoculated on the marked area and the papayas were dried under a biosafety hood. For Sections 2.3 and 2.4, approximately 10 7 log CFU of S. Typhimurium or L. monocytogenes inocula were used, and the papayas were dried for 1 h to initiate the attachment before every experiment [42]. For Section 2.5, approximately 10 8 log CFU of the inocula were used, and the papayas were dried for two hours to ensure attachment and initiate colonization before being washed with sanitizer solutions [42].

Recovery Method
Optimization of homogenization parameters is essential for accurate assessment of bacterial behavior on fruit surfaces. The goal of this experiment was to maximize the number of bacteria cells recovered from the papaya surface. After inoculation and drying as described above, the skin of the inoculated area was excised with a sterile knife and placed in a sterile stomacher bag. Bacterial cells were collected by homogenizing the skin under different conditions described as follows. Tested homogenization buffers included phosphate buffered saline (PBS, pH 7.4), 0.1% peptone water (PEPT), PBS + 0.2% Tween 80 (PBS + T) and 0.1% peptone water + 0.2% Tween 80 (PEPT + T). 25 mL of each buffer was separately added into the stomacher bag containing the excised skin and homogenized at 150 or 250 rpm for 1 or 5 min using a stomacher (Seward Stomacher ® , Model 400 Circulator, West Sussex, UK). After homogenization, the homogenate was serially diluted with 0.1% peptone water and plated on selective agar or using the agar overlay method. The agar overlay method was to plate the serially diluted homogenate on Plate Count Agar (PCA, Becton Dickinson, Franklin Lakes, NJ, USA) and incubating the plate at 37 • C for 1 h to ensure the recovery of injured cells, followed by pouring warm selective agar at 55 • C over the PCA [43]. The agar plates were incubated at 37 • C for 24 h and then analyzed for bacterial counts. The selective agar for S. Typhimurium and L. monocytogenes were xylose lysine deoxycholate agar (XLD, Becton Dickinson, Franklin Lakes, NJ, USA) and modified oxford agar (MOX, Becton Dickinson, Franklin Lakes, NJ, USA), respectively. Bacterial colonies were counted and populations were expressed as log CFU/papaya. The detection limit was 2.40 log CFU/papaya.

PH of Papaya Skin Homogenate as Affected by Homogenization Parameters
Papayas were prepared as described in Section 2.2 except that they were not inoculated with pathogenic bacteria. The skin of the marked area was cut and homogenized with buffer in a stomacher bag under the conditions described above. Papaya skin was also homogenized with water as a control. pH of the homogenate was measured using a pH meter (Model pH 6+, Oakton Instruments, Vernon Hills, IL, USA).

Behavior of Pathogenic Bacteria on Whole Papayas Stored at Different Temperatures
After harvesting and packing, papayas are usually stored at 7-13 • C before being distributed to grocery stores [44]. At grocery stores and customers' homes, papayas are usually placed at room temperature (21-25 • C). Hence, we selected 21 and 7 • C to simulate the two papaya storage scenarios. Inoculated whole papayas were individually placed in large sterile beakers and stored at 21 and 7 • C for 14 days. One papaya was randomly sampled, with the skin of the inoculated area being sterilely excised and collected for bacteria count on storage days 0, 7, 10 and 14. The papaya that was inoculated and dried for 1 h on the day of inoculation was considered as the sample on day 0. To determine bacterial population on papaya, the excised skin was homogenized using the optimized method from Section 2.3, which was homogenizing in PBS + T buffer at 250 rpm for 1 min for both S. Typhimurium and L. monocytogenes. Subsequently, the homogenates were serially diluted with 0.1% peptone water and plated using the agar overlay method described above. After incubation, bacterial colonies were counted and populations were expressed as log CFU/papaya. Aqueous ClO 2 solutions were made on-site using a previous method [38]. Briefly, ClO 2 stock solutions were prepared by mixing 10 mL of 4.0% NaClO 2 (Fisher Scientific, Waltham, MA, USA) with 10 mL of 1 M HCl (Fisher Scientific, Waltham, MA, USA), lactic acid (VWR Chemicals, Radnor, PA, USA) or malic acid (Fisher Scientific) in aluminum foil-covered bottles. After reacting for 1 min, 100 mL of distilled water was added into the bottles.
The final mixture was set at 21 • C for 20 min before being placed in a refrigerator at 4 • C. We previously investigated the generation kinetics and the stability of ClO 2 [38]. As organic acids release hydrogen ions slowly, it took one week to achieve equilibrium. During the 14-day-experimentation, the ClO 2 concentration increased till up to day seven and then remained stable for those generated with organic acids. For ClO 2 generated with HCl, the reaction was quick and the concentration remained stable for up to eight days and eventually decreased. Therefore, the stock solutions were all stored for seven days to allow the completion of the reaction in malic acid-and lactic acid-produced ClO 2 solutions and ensure no loss of the effectiveness of HCl-produced ClO 2 solutions. On the day of experimentation, the concentration of ClO 2 in each stock solution was measured using Chlordioxid-Test kit (EMD Millipore Corp., Burlington, MA, USA). The stock solutions were diluted with distilled water to 2.5, 5 and 10 ppm to treat papayas. The pH of each diluted solution was determined.

Washing Papayas with Aqueous ClO 2 and Individual Acid Solutions
To wash artificially contaminated papayas, each papaya was inoculated with S. Typhimurium or L. monocytogenes as described in Section 2.2 and then submerged into a sterile container containing 1 L of ClO 2 made with HCl, lactic acid or malic acid at concentrations of 2.5, 5 and 10 ppm. The submerged papayas were mildly stirred at a rate of 150 rpm for 5 min [45]. Subsequently, the washed fruits were dried under a biosafety hood for 15 min. After drying, the marked surface was sterilely cut and homogenized in 25 mL of PBS + T buffer at 250 rpm for 1 min. The homogenate was serially diluted and plated by the agar overlay method with XLD and MOX agar for the selection of S. Typhimurium and L. monocytogenes, respectively. Bacterial populations were expressed as log CFU/papaya, and the detection limit was 2.40 log CFU/papaya. Washing with distilled water and 200 ppm bleach (sodium hypochlorite (NaClO), pH 6.5) diluted from Clorox ® (6.0% NaClO, The Clorox Company, Oakland, CA, USA) served as the control treatments.
Acid solutions were prepared by adjusting 1 L of distilled water individually with 1 M HCl, 1 M lactic acid or 1 M malic acid to the pH of 10 ppm ClO 2 made with the corresponding acid. Papayas inoculated with S. Typhimurium or L. monocytogenes were washed with the acid solutions, and the remaining bacteria were collected and enumerated following the procedures described above.

ClO 2 Residue on Papaya Surface after Washing
Papayas were washed with tap water and dried on a lab bench for 1 h. Subsequently, the papayas were washed with 1 L of ClO 2 made with HCl, lactic acid or malic acid at concentrations of 5, 10 and 20 ppm. After drying for 15 min, the papayas were placed in 1-gallon Ziploc bags containing 100 mL distilled water. The papayas surfaces were hand massaged and rinsed thoroughly for 2 min, followed by filtering the rinse water into a flask [46]. 10 mL of the filtrate was collected and measured for ClO 2 concentration using Chlordioxid-Test kit. The detection limit was 0.02 mg/L in the undiluted filtrate. The ClO 2 concentration was converted into mg/kg papaya.

Statistical Analysis
All experiments were conducted in three independent replicates. Bacterial cultures were separately grown following the same procedure for each replicate. ClO 2 solutions were prepared freshly for each replicate. Data were reported as mean ± standard deviation (SD). Analysis of variance and Tukey's multiple comparison test were performed using SSPS software (IBM ® SPSS ® Statistics 24.0 for Windows, IBM Corp., Armonk, NY, USA). A significance level of 0.05 was used to determine the differences between the means of treatment groups.

Recovery of S. Typhimurium and L. monocytogenes Cells from Whole Papaya Surface as Affected by Homogenization Parameters and Enumeration Methods
Statistical analysis revealed no interactions among homogenization parameters, and only buffer significantly affected the bacterial count (p < 0.05). For S. Typhimurium (Table 1), papayas homogenized in buffers with the non-ionic surfactant Tween 80 resulted in significantly higher bacteria counts than those homogenized in peptone water alone. Tween 80 interrupts the hydrophobic interactions between bacteria cells and papaya surface and promotes the detachment of cells [47]. Papayas homogenized in the combination of PBS and Tween 80 (PBS + T) had the highest S. Typhimurium counts; an average of 5.36 log CFU was recovered from the initial inoculum of approximately 7 log CFU. Among all treatments, homogenization at 150 rpm for 5 min using XLD plating resulted in the highest recovery of 5.64 log CFU from papaya surface. For L. monocytogenes (Table 2), homogenization in PBS + T collected significantly more cells than in PBS alone (p < 0.05). Homogenization time, speed or plating method did not play a significant role in the collection. Homogenization at 150 rpm for 5 min by the agar overlay method resulted in the highest count of 5.09 log CFU. However, homogenization at 250 rpm for 1 min also resulted in relatively high L. monocytogenes counts. Homogenization at 250 rpm for 1 min was chosen for collecting S. Typhimurium and L. monocytogenes from papaya surface to maintain the time efficiency and consistency of the experiment. Even though the agar overlay method did not result in significantly higher bacteria counts than using selective agar alone, incubating on nonselective media before adding selective media would help recover bacteria cells injured by sanitizers [43]. It is an essential step to avoid over-estimation of the antimicrobial efficiency of sanitizers. Therefore, homogenizing the inoculated papaya piece in PBS + T at 250 rpm for 1 min was chosen, and the homogenate was decided to be plated by overlaying selective agar on PCA. * "PBS", "PEPT", "PBS + T" and "PEPT + T" stand for phosphate buffered saline, 0.1% peptone water, PBS with 0.2% Tween 80 and 0.1% peptone water with 0.2% Tween 80, respectively. Enumeration methods "XLD" and "PCA + XLD" stand for xylose lysine deoxycholate agar and plate count agar overlaid with XLD, respectively. Numbers are mean ± standard deviation (n = 3). No significant interactions were found between the factors. Means in the same column with different superscripts are significantly different (p < 0.05). * "PBS", "PEPT", "PBS + T" and "PEPT + T" stand for phosphate buffered saline, 0.1% peptone water, PBS with 0.2% Tween 80 and 0.1% peptone water with 0.2% Tween 80, respectively. Enumeration methods "MOX" and "PCA + MOX" stand for modified Oxford agar and plate count agar overlaid with MOX, respectively. Numbers are mean ± standard deviation (n = 3). No significant interactions were found between the factors. Means in the same column with different superscripts are significantly different (p < 0.05).
pH values of the above-mentioned homogenates were measured with uninoculated samples to compare buffering capacity between homogenization buffers. Even with careful excision, papaya flesh attached to the skin could acidify the homogenate. Papaya flesh has a pH of 4.87-5.7 [16,18]. This pH range does not inhibit the growth of S. Typhimurium or L. monocytogenes; however, it could influence the recovery of cells injured by desiccation [43]. Tian et al. incubated sublethally injured E. coli O157:H7 cells in nutrient broth at pH 4.0, 5.0, 6.0, 7.2 and 8.0. They found that the cells showed no significant recovery at pH 4.0 and 8.0 whereas the cells recovered by 0.48, 0.49 and 0.72 log CFU/mL in pH 5.0, 6.0 and 7.2, respectively, indicating that pH even at relatively high levels (5.0 and 6.0) did affect the recovery of sublethally injured cells [48]. Shown in Table 3, homogenizing papaya skin in different buffers resulted in significant differences in homogenate acidity in a descent order of PBS, PBS + T, PEPT, water and PEPT + T (p < 0.05). The initial pH value of each buffer was measured with PBS, PBS + T and water being neutral whereas PEPT and PEPT + T being slightly acidic (pH = 6.5-6.7). PBS is known for its high buffering capacity, whereas water and peptone water have little buffering capacity. When mixed with the papaya juice, the pH of water and peptone water decreased to 5.89-6.26. The pH of the homogenate may affect the state of cells, and this is consistent with the higher cell counts observed in PBS + T. Peptone water is often used in studies involving fresh produce [20,23,49]. Researchers should carefully select homogenization buffers since peptone water alone may lead to experimental errors in studies with acidic produce. 6.05 ± 0.22 6.03 ± 0.17 6.10 ± 0.14 5.82 ± 0.08 6.00 ± 0.17 c * "PBS", "PEPT", "PBS + T" and "PEPT + T" stand for phosphate buffered saline, 0.1% peptone water, PBS with 0.2% Tween 80 and 0.1% peptone water with 0.2% Tween 80, respectively. Numbers are mean ± standard deviation (n = 3). No significant interactions were found between the factors. Means in the same column with different superscripts are significantly different (p < 0.05).

Behavior of Pathogenic Bacteria on Whole Papayas Stored at Different Temperatures
With about 7 log CFU of initial inocula, 5.46 and 4.67 log CFU S. Typhimurium and L. monocytogenes were detected on papaya surfaces on day 0, respectively ( Figure 1). Bacteria response to environmental stress differently. Salmonella showed higher desiccation tolerance than L. monocytogenes in powdered infant formula and desiccated shredded coconut [50,51]. S. Typhimurium had an interesting survival and growth pattern. At 21 • C, the population increased gradually to 7.34 log CFU on day 14. At 7 • C, S. Typhimurium level decreased to 4.10 log CFU on day 7 and then increased to 6.18 log CFU at the end of the storage period ( Figure 1A). Intrinsic factors of fruit, including surface roughness, surface hydrophobicity, nutrient and moisture availability and background flora, may affect the behavior of foodborne pathogenic bacteria on the fruit [8]. At ambient temperature, S. enterica level remained stable on whole mangos stored at 20-22 • C for nine days [52]. Salmonella was reduced by about 5 and 2 log CFU at high (~7 log) and low (~4 log) inoculation levels, respectively, on whole kiwifruits stored at room temperatures for 10 days [14]. On whole cucumbers stored at 23 • C, Salmonella level significantly increased by 1.7 log CFU within the first day of inoculation and remained stable for four days [53]. Looking at the fruit type alone, at commercial cold storage conditions (7-12 • C), S. Typhimurium level did not significantly change on whole papaya or mango at the end of the storage period [54]. However, Salmonella tended to decrease over time on other fruits, such as passionfruit, strawberry, cucumber and peppers [53][54][55][56]. Different from other tropical fruits, sugar accumulates on papaya surfaces as ripening progresses, which provides more nutrients for the attached microorganisms. Naturally present yeast may also aid the growth of S. Typhimurium by their saccharolytic interactions with the compounds permeated through papaya skin [57].
L. monocytogenes showed a major increase from 4.67 to 5.60 log CFU during the 1st day of storage at 21 • C, and then gradually grew to 5.91 log CFU in the following 13 days. At 7 • C, L. monocytogenes level remained stable on papayas for 14 days ( Figure 1B). The behavior of L. monocytogenes on fruits varies. L. monocytogenes grew on whole cucumbers stored at 4 • C and grew on fresh Gala apples stored at 5 • C and 25 • C [53,57]. However, on Granny Smith apples, 1.5 log CFU and 0.5-1.2 log CFU reductions were observed at 25 and 22 • C, respectively, in two studies [13,57]. The reductions of L. monocytogenes on whole cantaloupe and mango were also reported [12,49]. Aside from the intrinsic differences of the fruits, initial inoculation levels and the carrying capacity of the fruit may contribute to the varied behavior of L. monocytogenes [8,18]. Approximately three-fold more L. monocytogenes died on whole kiwi fruits inoculated with 7 log CFU than those inoculated with 4 log CFU at room temperature over 10 days [14]. In the case of organic Granny Smith apples, L. monocytogenes decreased by 1.8 and 0.7 log CFU at inoculation levels of 6.3 and 3.0 log CFU, respectively, at 22 • C over two weeks [13]. Papayas could have a higher carrying capacity than the above-mentioned fruits, leading to the growth of L. monocytogenes on papayas even at a relatively high inoculation level. Regardless, L. monocytogenes is known for its ability to adapt to cold temperatures through mechanisms of alternating membrane fatty acid composition, synthesizing cold shock proteins and cold acclimation proteins and activating energy providing pathways such as glycolysis [58].
S. Typhimurium and L. monocytogenes showed abilities to survive and grow on papaya, and hence effective sanitation methods are essential for papaya production. Error bars are standard deviations (n = 3). Different lower-case letters horizontally indicate significant differences between the means of different time points at each temperature (p < 0.05). Different upper-case letters vertically indicate significant differences (p < 0.05) between the means of different temperatures at the same time point. "a*" means p values were marginal, which were 0.058 and 0.059 on day 10 and day 14, respectively, compared with day 0.

Inactivation of S. Typhimurium and L. monocytogenes on Whole Papayas Using
Aqueous ClO 2 Figure 2A shows S. Typhimurium reduction by water, aqueous ClO 2 , and bleach on whole papayas. 10 ppm of ClO 2 was significantly more effective than 2.5 and 5 ppm (p < 0.05). 10 ppm of ClO 2 reduced S. Typhimurium from the initial inoculation level of 7.5 log CFU to an undetectable level. 200 ppm of bleach achieved the same result. Malic acid-produced ClO 2 reduced S. Typhimurium by 6.23 and 6.90 log CFU at 2.5 and 5 ppm, respectively. HCl-and lactic acid-produced ClO 2 reduced S. Typhimurium by 4.20 and 5.05 log CFU, and 3.89 and 4.67 log CFU at 2.5 and 5 ppm, respectively. Overall, ClO 2 solutions generated with malic acid inactivated significantly higher numbers of S. Typhimurium than the solutions generated with HCl or lactic acid (p < 0.05). 1.74-2.01 and 0.86-1.97 log CFU/cm 2 Salmonella was inactivated in 100 ppm free chlorine and 80 ppm peracetic acid with scrubbing by sponges/microfiber, respectively [35]. Comparing with these results, the microbial reduction on papayas achieved by immersing in ClO 2 for 5 min seems more effective. Error bars are standard deviations (n = 3). Bars labeled with different letters indicate significant differences between the means of treatments (p < 0.05). Lines labeled with "*" indicate significant differences between ClO 2 groups made with different acids ("*", p < 0.05; "**", p < 0.01).
Water treatment only removed 2.56 log CFU of S. Typhimurium from papaya surface, whereas 4.47 log CFU of L. monocytogenes was removed by water ( Figure 2). This may be partially due to that S. Typhimurium attached stronger to papaya surfaces than L. monocytogenes. In a study conducted by Collignon and Korsten [42], S. Typhimurium adhered to peach immediately after contact, whereas L. monocytogenes required 60 s for the adhesion. Higher numbers of S. Typhimurium cells were observed in one hour than L. monocytogenes on peach.
ClO 2 produced with HCl did not show higher effectiveness in reducing L. monocytogenes than water ( Figure 2B). ClO 2 produced using lactic acid had increased bacterial reductions than HCl-produced ClO 2 at 5 and 10 ppm but with large variations. Malic acid-produced ClO 2 showed the highest L. monocytogenes reduction among all ClO 2 treatments. However, there was no significant difference between the three tested concentrations. The group treated with ClO 2 made with malic acid showed statistically higher bacterial reduction than the group treated with ClO 2 made with HCl (p < 0.05). 2.5, 5 and 10 ppm of malic acid-generated ClO 2 reduced L. monocytogenes by 7.20, 6.63 and 8.04 log CFU, respectively. These reductions were higher than the L. monocytogenes reductions on apples, lettuce, strawberries and cantaloupe treated with 5 ppm ClO 2 made with phosphoric acid (~5.6 log CFU) [59]. L. monocytogenes-contaminated papayas treated with 200 ppm bleach also showed a relatively large variation with an average reduction of 5.5 log CFU, which was lower than all samples treated with malic acid-generated ClO 2 . However, the concentration of bleach was much higher than that of ClO 2 , indicating the high antimicrobial efficiency of ClO 2 . This result agrees with the higher reduction of L. monocytogenes on blueberries treated with 10 ppm ClO 2 (1.7 log CFU/g) than those treated with 200 ppm chlorine (1.0 log CFU/g) for 5 min [23].
ClO 2 generated with malic acid inactivated significantly more S. Typhimurium and L. monocytogenes than ClO 2 generated with HCl. This result is consistent with our previous observation of the high antimicrobial efficiency of ClO 2 generated with organic acids. In particular, malic acid-generated ClO 2 had higher efficacy in killing S. Typhimurium and L. monocytogenes than HCl-, sodium bisulfate-, citric acid-and lactic acid-generated ClO 2 [38]. This conclusion was drawn from experiments conducted on bacteria cell suspensions and Romaine lettuce. We hypothesized that synergistic effects between ClO 2 and the excessive organic acids in the ClO 2 solutions may contribute to the high antimicrobial efficiency of organic acid-generated ClO 2 . We treated contaminated papayas with individual acid solutions to confirm this hypothesis. Since the pH of ClO 2 decreased with the increase of its concentration (data not shown), pH values corresponding to 10 ppm ClO 2 were selected for the decontamination experiments with acids alone. This means the pH of HCl, lactic acid and malic acid solutions were adjusted to 3.15, 3.42 and 3.36, respectively. S. Typhimurium on papayas treated with the acids was reduced by 2.45-3.01 log CFU, which was not significantly different from the samples treated with water (Table 4, p > 0.05). Similarly, L. monocytogenes on papayas treated with the acids was reduced by 3.58-3.91 log CFU and was not significantly different from the samples treated with water (p > 0.05). Hence these results confirmed the high antimicrobial effect of ClO 2 solutions made with malic acid and lactic acid was contributed little by the excessive organic acids, but rather a synergistic effect between ClO 2 and organic acids. The combination treatment of 2.0% lactic acid and 10 ppm ClO 2 resulted in higher reductions of S. Typhimurium and L. monocytogenes on blueberries than the treatments by each sanitizer alone [60]. On papaya, ClO 2 produced with lactic acid interestingly had similar killing effects to ClO 2 produced with HCl, yet ClO 2 produced with malic acid still performed better than that with HCl. In many studies, lactic acid was either better or as good as malic acid in the inactivation of pathogens when used alone as the sanitizers [61,62]. The synergistic effect somehow altered the antimicrobial efficiency of lactic acid and malic acid. Another factor may contribute to the altered antimicrobial efficacy of the organic-acid-generated ClO 2 compared with HCl-generated ClO 2 is the intermediate compounds produced in the ClO 2 solutions. ClO 2 solution is a mixture of pure ClO 2 and oxidative chlorine compounds such as ClO 2− , ClO 3− , free chlorine (Cl 2 ), hypochlorous acid (HOCl) and hypochlorite ion (OCl − ) [32].  Additionally, CFR Sec. 173.300 specifies that ClO 2 can be used in fresh produce wash with a rinse procedure, and ClO 2 residue in the wash water of the applied fresh produce should not exceed 3 ppm [25]. EPA also specifies that ClO 2 is allowed to rinse fruits and vegetables at a concentration of 5 ppm [63]. Some literature also suggests that the residue on the washed produce should not exceed 3 ppm [64,65]. In this study, the ClO 2 residue on papayas after being treated with 5, 10 and 20 ppm ClO 2 solutions ranged from 8.0 × 10 −5 to 6.2 × 10 −3 mg/kg, which were far below 3 ppm (Table 5). These numbers were also far below the EPA regulation of 0.8 mg/L ClO 2 residue in public drinking water [27]. Tomatoes and strawberries treated with 0.5 ppm gaseous ClO 2 for 10 min had 0.09 and 0.37 mg/kg ClO 2 residue [29]. ClO 2 residue on produce treated with gaseous ClO 2 was much higher than ClO 2 residue on papayas treated with aqueous ClO 2 , providing insights into safety concerns in the application of ClO 2 . However, future studies of ClO 2− reside on food matrix treated with ClO 2 should be carried out as ClO 2− and ClO 3− are harmful disinfection by-products (DPBs) that can cause anemia and thyroid dysfunction in animals [26].

Conclusions
To provide potential solutions to the emerging issue of foodborne illness outbreaks associated with whole papayas, this study investigated the survival of S. Typhimurium and L. monocytogenes on whole papaya during storage at 21 and 7 • C and determined the efficiency of aqueous ClO 2 in inactivating the two pathogenic bacteria on whole papaya. Temperature played a significant role in the survival and growth of the bacteria on the fruit. S. Typhimurium grew by 1.88 log CFU on whole papaya in 14 days at 21 • C and remained at the initial inoculation level at 7 • C. L. monocytogenes grew by 0.93 log CFU on papaya during the 1st day of storage at 21 • C; the level remained stable thereafter at 21 • C and at 7 • C. The acid used to produce aqueous ClO 2 influenced the killing efficiency of ClO 2 against these pathogenic bacteria. ClO 2 solutions generated with malic acid reduced significantly higher levels of S. Typhimurium and L. monocytogenes than the solution generated with HCl. Methodology wise, we optimized the methods for recovering pathogenic bacteria cells from papaya surface, which was a crucial step evaluating bacterial behavior on fresh produce. This study also provided information on ClO 2 residue on the washed papayas. These results give insights on risk assessment and management of microbiological safety issues associated with whole papaya. Further studies including the intermediate compound compositions in various ClO 2 solutions and the residue of DPBs on ClO 2 treated food matrix are suggested to better understand the antimicrobial mechanisms and safety concerns regarding using aqueous ClO 2 .