International Journal of Food Microbiology 202 (2015) 66–72

Contents lists available at ScienceDirect

International Journal of Food Microbiology journal homepage: www.elsevier.com/locate/ijfoodmicro

Reduction of Salmonella enterica serotype Poona and background microbiota on fresh-cut cantaloupe by electron beam irradiation Mangesh P. Palekar a,1, T. Matthew Taylor a, Joseph E. Maxim b,2, Alejandro Castillo a,⁎ a b

Texas A&M University, Department of Animal Science, College Station, TX 77843-2471, USA Texas A&M National Center for Electron Beam Food Research, College Station, TX 77845-2259, USA

a r t i c l e

i n f o

Article history: Received 19 May 2014 Received in revised form 28 January 2015 Accepted 2 February 2015 Available online 8 February 2015 Keywords: Electron beam Irradiation Salmonella Cantaloupe Food safety Fresh-cut produce

a b s t r a c t The efficacy of electron beam (e-beam) irradiation processing to reduce Salmonella enterica serotype Poona on surfaces of fresh-cut cantaloupe, and the impact of e-beam irradiation processing on the numbers of indigenous microorganisms were determined. Additionally, the D10-value for S. Poona reduction on the cut cantaloupe was also determined. Fresh-cut cantaloupe pieces, inoculated with S. Poona to 7.8 log10 CFU/g, were exposed to 0.0, 0.7, or 1.5 kGy. Surviving S. Poona, lactic acid bacteria (LAB), and fungi (yeasts, molds) were periodically enumerated on appropriate media over 21 days of storage at 5 °C. Cantaloupe surface pH was measured for irradiated cantaloupe across the 21 day storage period. To determine the D10-value of S. Poona, cantaloupe discs were inoculated and exposed to increasing radiation dosages between 0 and 1.06 kGy; surviving pathogen cells were selectively enumerated. S. Poona was significantly reduced by irradiation; immediate reductions following exposure to 0.7 and 1.5 kGy were 1.1 and 3.6 log10 CFU/g, respectively. After 21 days, S. Poona numbers were between 4.0 and 5.0 log10 CFU/g less than untreated samples at zero-time. Yeasts were not reduced significantly (p ≥ 0.05) by e-beam irradiation and grew slowly but steadily during storage. Counts of LAB and molds were initially reduced with 1.5 kGy (p b 0.05) but then LAB recovered grew to high numbers, whereas molds slowly declined for irradiated and control samples. Cantaloupe pH declined during storage, with the greatest decrease in untreated control cantaloupe (p b 0.05). The D10-value for S. Poona was determined to be 0.211 kGy, and this difference from the reductions observed in the cut cantaloupe studies may be due to the more precise dose distribution obtained in the thin and flat cantaloupe pieces used for the D10-value experiments. The effect of e-beam irradiation at the same doses used in this study was determined in previous studies to have no negative effect in the quality of the cut cantaloupe. Therefore, incorporation of low dosage ionizing irradiation and consistent application of irradiation processing can significantly improve the microbiological safety of fresh-cut cantaloupe. © 2015 Elsevier B.V. All rights reserved.

1. Introduction The importance of fresh fruits and vegetables as sources of multiple nutrients and their role in promoting good health have stimulated increased demand and consumption in recent years (Mandrell et al., 2006). Cantaloupe is a popular fruit with an increasing demand that belongs to the family Cucurbitaceae alongside honeydew, watermelon, cucumber, and multiple squash varieties (Parnell and Harris, 2003; Parnell et al., 2005). The United States Department of Agriculture Economic Research Service (USDA-ERS) reports that per capita usage of commercially grown cantaloupe has risen steadily since 1970, growing to 11.4 lb in 1999, although a steady decline has been observed since

⁎ Corresponding author at: Department of Animal Science, Texas A&M University, 474 Olsen Boulevard, College Station, TX 77843-2471, USA. Tel.: + 1 979 845 4425; fax: + 1 979 862 3475. E-mail address: [email protected] (A. Castillo). 1 Current affiliation: Kraft Foods, 555 South Broadway, Tarrytown, NY 10591, USA. 2 Current affiliation: Not currently professionally affiliated.

http://dx.doi.org/10.1016/j.ijfoodmicro.2015.02.001 0168-1605/© 2015 Elsevier B.V. All rights reserved.

2004, reaching 8.5 lb per capita in 2010 (USDA-ERS, 2012). This decline in cantaloupe consumption does not seem to be resulting from the recent cantaloupe-linked outbreaks of listeriosis (CDC, 2012a,b) and salmonellosis (CDC, 2012a). Although not a causal relationship, the increase in consumption and global trade of fresh produce, including cantaloupe, has been accompanied by a rise in the number of foodborne disease outbreaks in recent years. Multiple outbreaks of illness caused by different serotypes of Salmonella enterica subsp. enterica have been epidemiologically linked to cantaloupes over the years, mainly to imported cantaloupes (CDC, 1991, 2002; Deeks et al., 1998; Mohle-Boetani et al., 1999). Some of these outbreaks were instrumental in the U.S. Food and Drug Administration (FDA) executing policies halting the importation of Mexicogrown cantaloupes into the United States (FDA, 2011). These policies are still in effect. The import alert against Mexican cantaloupes opened opportunities for other Latin American markets, and in 2008 and 2011, two multi-state outbreak of Salmonella illness were linked to consumption of cantaloupes imported from Honduras and Guatemala, respectively (CDC, 2008, 2011). Most recently, an outbreak of disease caused

M.P. Palekar et al. / International Journal of Food Microbiology 202 (2015) 66–72

by S. Newport and S. Typhimurium was linked to domestic cantaloupes from an Indiana cantaloupe production and packing operation. Lack of proper packing facility sanitation and failure to ensure effective sanitizer levels in the wash water were identified as significantly contributing to transmission of contaminated cantaloupes to consumers (CDC, 2012a,b). The FDA has provided guidance to the melon industry for the reduction of microbial hazards on fresh-cut melons, including cantaloupes (FDA, 2008). In addition to multiple production-level processes to reduce transmission of pathogens from the field to processing, FDA recommended the use of procedures to reduce microbial loads by at least 2.0–3.0 log10 cycles on melons during postharvest operations, to reduce the potential presence of bacterial pathogens on the melons prior to cutting or slicing. Additionally, FDA recommended the use of a surface decontamination process, citing the inability of washing procedures to reliably eliminate pathogens from melon surfaces during even extended washing in disinfectant (e.g., chlorine)-containing water (FDA, 2008). Thus, effective process strategies are needed by melon processors to disinfect the cantaloupe surfaces from microbial pathogens prior to slicing or cutting that do not result in significant loss of quality and shelf-life. Although still not approved for use in the US for reducing pathogens in produce other than spinach and iceberg lettuce, ionizing irradiation has shown promise as a non-thermal processing technology capable of reducing pathogens on melon surfaces (Castillo et al., 2009). This technology can be applied even on melons after cutting, provided that strict adherence to good manufacturing practices is observed during and after cutting. Palekar et al. (2004) reported that 0.7 kGy electron beam (e-beam) irradiation application to cantaloupe surfaces, followed by storage at 5 °C, resulted in significantly lower numbers of aerobic organisms on melon surfaces as compared to counts on melons washed with water or chlorine (200 ppm hypochlorite, 3 min). Application of hypochlorite followed by 0.7 or 1.4 kGy irradiation resulted in even greater suppression of aerobic microbes, as chlorine washing followed by irradiation (1.4 kGy) effectively inhibited growth of aerobes over 21 days of refrigerated storage (Palekar et al., 2004). Castell-Perez et al. (2004) reported that exposure of cantaloupes to irradiation up to 3.1 kGy did not significantly alter several quality indices, including sugar content, carotene content, or texture during storage, as compared to untreated controls. Irradiation (0.5 kGy) of washed (76 °C, 3 min), cubed cantaloupes resulted in reductions of aerobic bacteria of 0.5– 0.6 log10 CFU/g versus washed, non-irradiated control samples over 7 days of refrigerated storage (Fan et al., 2006). Reports on the use of e-beam irradiation for microbial reduction on cut cantaloupes are scarce. Fan et al. (2006) reported the use of gamma irradiation to reduce total aerobic microorganisms on cantaloupe cubes. Previously, Palekar et al. (2004) reported the effect of ebeam irradiation on total aerobic microorganisms on cut cantaloupe. However, studies on the use of irradiation to reduce bacterial pathogens on cut cantaloupes were not found. Therefore, the objectives of this study were to: (i) study the effect of electron beam irradiation processing for the reduction of a strain of S. Poona and background microbiota on sliced cantaloupes, and; (ii) determine the D10-value for a strain of S. Poona on sliced cantaloupe irradiated with electron beam. The S. Poona used was a specific strain that was involved with a salmonellosis outbreak linked to cantaloupes.

67

When needed, the frozen culture was revived by transferring to tryptic soy broth (TSB; Becton, Dickinson and Co., Sparks, MD, USA) and incubating for 24 h at 37 °C. One loopful of the broth culture was streaked onto tryptic soy agar (TSA, Becton, Dickinson and Co.) and incubated at 37 °C for 24 h. A single colony was picked from the plate, transferred to a TSA slant and incubated at 37 °C for 24 h. The Rif+ S. Poona maintained on TSA slant was transferred to a flask containing 300 mL of TSB and grown by incubating at 37 °C for 24 h. Cells were then harvested by centrifugation (Centrifuge B4i, Jouan, Thermo-Fisher Scientific, Waltham, MA, USA) at 2346 ×g in sterile tubes and then washed three times with equal volumes of sterile 0.1% peptone water. The mean population of S. Poona in peptone water suspension was 8.3 ± 0.3 log10 CFU/mL. 2.2. Fruit preparation and inoculation

2. Methods

Cantaloupes (Cucumis melo L. var. reticulatus) were purchased directly from a producer in South Texas and were not subjected to any wash or other treatment by the supplier. The cantaloupes were boxed in the field by the supplier and shipped to the Food Microbiology laboratory at Texas A&M University. In the laboratory, all melons were washed with sterile distilled water while gently scrubbing with a brush for 2 min. Washed melons were then peeled, cut and cored into cylinders 2.54 cm in diameter using stainless steel core borers. All materials used for cutting and handling were continually sanitized by dipping in 70% ethanol and flaming to prevent cross contamination. Each piece was individually dipped in the suspension of S. Poona and left at 25 °C for 1 h on a sterilized rack to allow the excess suspension to drain off. The average count of S. Poona on cantaloupe pieces was 7.8 log10 CFU/g. The cylindrical pieces of cantaloupe were then packed in 15 cm × 19 cm Whirl-Pak™ bags (Nasco, Fort Atkinson, WI), each bag containing 100 g of inoculated fruit. The bags were then double sealed using an Impulse Heat Sealer (American International Electric, New York, NY, USA) to ensure hermetic conditions. The oxygen level in the bags was 9% and the carbon dioxide level was 3%. These packaging conditions were chosen after preliminary experiments to determine the optimal packaging film and weight of fruit to be packaged was determined during preliminary experiments so that the gas permeation rate of the film balanced the respiration rate of the fruit at 5 °C. For D10-value determination, cored cylinders of cantaloupe were further sliced into 4 mm thick sections. This was done to match the thickness of the cantaloupe piece with the 4 mm thickness of alanine dosimeters so that the absorbed dose on the dosimeter would be an accurate measure of the dose that would be absorbed by the cantaloupe disc, all other parameters being constant. The cantaloupe discs were inoculated with a suspension of Rif+ S. Poona prepared as described above, and then placed on sterile Petri dishes. Six discs of inoculated cantaloupe were prepared per Petri plate and each Petri plate was subjected to each particular dose. Since the entire process from cutting the cantaloupes to sealing the bags or preparing the Petri plates with thin pieces of cantaloupe took a maximum time of 6 h, all bags were kept in a refrigerated container (~ 5 °C) to prevent microbial outgrowth during fruit preparation and inoculation. The refrigerated container with all samples was then transported to the National Center for Electron Beam Food Research at Texas A&M University, located in the same campus, for immediate irradiation processing. This container was only opened to remove the samples as they were being irradiated. The time between inoculation and irradiation did not exceed 9 h.

2.1. Bacterial cultures and inoculum preparation

2.3. Dosimetry

A rifampicin-resistant (Rif+) strain of S. enterica serotype Poona isolated from a cantaloupe-associated outbreak was kindly provided by Dr. Linda J. Harris of the University of California at Davis (Davis, CA, USA). The S. Poona organism was maintained at − 80 °C on Protect™ Bacterial Preservers (Key Scientific Products, Round Rock, TX, USA).

Dose mapping of the cantaloupe cylinders was carried out using alanine pellets (Harwell Dosimeters, Oxfordshire, United Kingdom) at the dosimetry laboratory of the National Center for Electron Beam Food Research at Texas A&M University (College Station, TX, USA). A dual electron beam configuration, from the top and bottom of the

68

M.P. Palekar et al. / International Journal of Food Microbiology 202 (2015) 66–72

sample, was used for this experiment to minimize the dose variation throughout the sample. High-density polyethylene (HDPE) sheets were used as attenuators for reducing the energy of incident electrons in order to achieve the low target doses of 0.7 and 1.5 kGy on the cylindrical cantaloupe pieces. The alanine pellets placed on the cantaloupe pieces were analyzed for the absorbed dose using Electron Paramagnetic Resonance (EPR) spectroscopy (Bruker EMS 104 EPR Analyzer, Bruker Instruments, Billerica, MA, USA). Multiple dosimetry runs were carried out to finalize the exact attenuation scheme and conveyor speed required to achieve the target doses for the experiment. For D10-value determination, alanine pellets were used to map the irradiation dosage for the 4 mm cantaloupe slices. Dose mapping was aimed at establishing a range of doses between 0 and 1 kGy at target increments of 0.1 kGy. The dosimeters were placed in holes punched in the center of cantaloupe discs placed in a Petri dish. This was done to take into account the effect of cantaloupe mass surrounding the alanine in dose absorption. HDPE sheets were used as attenuators as previously described. Extensive dosimetry work along with various combinations of attenuation led to the finalization of the following doses for the experiment: 0, 0.12, 0.19, 0.29, 0.41, 0.54, 0.63, 0.79, 0.85, and 1.06 kGy. 2.4. Irradiation treatment Cantaloupe-containing packets were placed in a single layer in cardboard boxes on a conveyor and exposed to either 0.7 or 1.5 kGy e-beam irradiation using dual beams directed from the top and bottom of the packets as described above. The same treatment conditions and similar doses were tested previously and found to have little or no negative effect on the quality and sensory characteristics of cut cantaloupes (Palekar et al., 2004). Processing was carried out at the National Center for Electron Beam Food Research at Texas A&M University. The postirradiation handling and storage processes were similar to those described previously (Palekar et al., 2004). On days 0, 3, 6, 9, 12, 15, 18 and 21 post-irradiation, packets were removed from the cold room for microbiological and pH analyses. For the D10-value study, ten Petri dishes containing six cantaloupe discs each were prepared and inoculated with S. Poona. Each plate was covered with Saran wrap and then placed in a Whirl-Pak™ (Nasco, Fort Atkinson, WI, USA) bag to prevent any leakage and contamination. The plates were then irradiated at the doses indicated above. After irradiation, the samples along with the non-irradiated control were immediately analyzed for microbial counts. The use of thin pieces of cantaloupe for D10 -value determinations was to enable the use of precision dosing, which would reduce dose variation and allow the use of 0.1 kGy dose increments, which may not be possible with the dose variations naturally obtained during irradiation of cut cantaloupe packets, which was intended to simulate potential process conditions in an industrial setting. 2.5. Microbiological analysis Enumeration of Rif+ S. Poona was carried out on TSA plates containing rifampicin (100 mg/L; Sigma Aldrich Co., St. Louis, MO, USA) and cycloheximide (100 mg/L; Sigma Aldrich Co.) (TSA-RC). Cycloheximide served as an antifungal agent to prevent indigenous fungi from spreading over the solid medium before bacterial colonies could develop. Rifampicin (0.1 g) was dissolved in 5 mL methanol, whereas cycloheximide (0.1 g) was dissolved in 5 mL of sterile distilled water. Filtered (0.45 μm) solutions of antimicrobials were added to 1 L of sterile, tempered (45 °C) TSA; plates were then poured. Enumeration of lactic acid bacteria (LAB) was carried out using de Man, Rogosa and Sharpe lactobacilli agar (MRS; Becton, Dickinson and Co.) overlaid with All Purpose Tween agar (APT; Becton, Dickinson and Co.), adjusted to pH 4.0 with 10% tartartic acid (Mallinckrodt Chemical Works, St. Louis, MO) (Downes and Ito, 2001). A 25 g sample was taken from each packet and aseptically placed into a sterile stomacher bag with 225 mL of 0.1% (w/v) peptone water

(Becton, Dickinson and Co.). The mixture was pummeled in a laboratory masticator (Stomacher 400, Seward, London, UK) at high speed for 2 min. Appropriate serial dilutions were made from this homogenate and were spread onto TSA-RC (limit of detection by plating: 1.0 log10 CFU/g). Plates were then incubated at 37 °C for 24 h before counting. All Salmonella counts (CFU/g) were then transformed into log10 CFU/g. For enumeration of LAB, appropriate dilutions were spread onto MRS agar-containing plates and then overlaid with molten APT agar. The overlaid agar was allowed to set, after which the plates were incubated at 35 °C for 3–5 days. Colonies of LAB embedded between the two agar layers were counted and calculated as CFU/g. For each round of sample analysis, 15 characteristic LAB colonies were randomly picked from the MRS + APT agar plates and transferred to TSA slants for further characterization by Gram staining, catalase production and fermentative metabolism (O-F Glucose; O-F Basal Medium + Glucose + 0.1% Yeast Extract; Becton, Dickinson and Co.) testing. Yeasts and molds were enumerated by inoculating appropriate dilutions onto a Yeast and Mold Petrifilm™ (3M® Microbiology Products, St. Paul, MN, USA) and incubating for 5 days at 25 °C. The observed growth after incubation was recorded separately as yeast and mold counts per gram. For the D10-value study, after irradiation of samples, each disc of cantaloupe was mixed separately with 10 mL of sterile 0.1% peptone water in the Whirl‐Pak™ bags, 7.6 × 12.7 cm in size, and pummeled at a high speed for 2 min in the lab stomacher. An aliquot of the homogenate was serially diluted 10-fold and spread onto TSA-RC; developing colonies were reported as CFU/g and then converted to log10 CFU/g. The counts were reported as an average of six replicates for each dose of irradiation. 2.6. Confirmation of isolates Five representative colonies were transferred from each plate onto TSA slants and confirmation for Salmonella was carried out biochemically using triple sugar iron (TSI; International Bioproducts, Bothell, WA, USA) and lysine iron (LIA; International Bioproducts) agar slants, and serologically using the agglutination reaction with Salmonella O Poly A-I and Vi antiserum (Becton, Dickinson and Co.). 2.7. Measurement of surface pH of cut cantaloupe Surface pH on the cantaloupe pieces from each package was measured immediately after opening and prior to conducting the microbiological analysis. A Markson 612 portable pH meter (Markson Science, Inc., Phoenix, AZ, USA) with a flat bulb design electrode (Markson Science, Inc.) was used for measurements. The pH meter was calibrated and sanitized prior to use on each day of analysis. Three pH readings were taken for each sample and the average was calculated. 2.8. Analysis of data All experiments were replicated three times. All microbial count data were converted to log10 before analysis. To determine log reductions, the mean log count of Salmonella or background microbiota after irradiation at 0.7 or 1.5 kGy was subtracted from the mean log count of the non-treated controls. To facilitate statistical data analysis, all counts that were below the detection limit of the plate count method (1.0 log10 CFU/g) were assigned a value of 0.5 log10 CFU/g, which is the level between 0 and this detection level. Data recorded for each parameter tested in three trials were analyzed by one-way analysis of variance (ANOVA) using Statistical Analysis Software v.9.0 (SAS Institute, Cary, NC, USA). For the pathogen and normal flora reduction study, for each organism enumerated, the average cell number (log10 CFU/g) was plotted against irradiation dose to display the reduction achieved in the sample due to irradiation on day 0 and the effect of storage over

M.P. Palekar et al. / International Journal of Food Microbiology 202 (2015) 66–72

21 days at 5 °C. The effect of irradiation dose on the reduction of the studied organisms was analyzed by ANOVA (SAS v.9.0), with mean separation using Duncan's multiple range test. The level of significance was set at b 0.05. For the D10-value study, numbers of surviving S. Poona were plotted against increasing doses of irradiation and a linear regression chart was established. The D10-value was determined from the reciprocal of the regression line as the dose in kGy required to reduce the S. Poona population by one log10 cycle. 3. Results and discussion 3.1. Reduction of microbiota on irradiated sliced cantaloupe 3.1.1. S. Poona As seen in Fig. 1, electron beam irradiation reduced the strain of S. Poona used in this study by 1.1 log10 CFU/g at 0.7 kGy and 3.6 log10 CFU/g at 1.5 kGy in comparison with the non-irradiated control, immediately after irradiation treatment (day 0). The counts at day 0 indicate the immediate effect of irradiation treatment on reduction of salmonellae. Storage of the irradiated cantaloupe for 21 days at 5 °C resulted in gradual reduction in Salmonella numbers with time, although a complete elimination was not observed up to the end of the study (21 days). Cantaloupe, being a nutritionally rich medium, can support the survival and proliferation of Salmonella (Golden et al., 1993). Up to day 15 of storage, the decline in numbers was greater for 0.7 and 1.5 kGy-irradiated samples in comparison to the control (Fig. 1). However, after day 15, salmonellae in the control were reduced at a faster rate than the 0.7 and 1.5 kGy samples. At the end of 21 days, the non-irradiated samples had ~7.0 log10 CFU/g while those irradiated at 0.7 kGy and 1.5 kGy had 3.4 and 2.2 log10 CFU/g, respectively. According to these data, electron beam irradiation of sliced cantaloupe was effective in reducing Salmonella. As expected, the subsequent storage of irradiated fruit at refrigeration temperature prevented proliferation of survivors. Prakash and Foley (2004) reported the levels of producecontaminating pathogens to be usually low (≤ 3 log10 CFU/g) and hence low doses of irradiation can be effective in eliminating the threat that they may pose to consumer health. Golden et al. (1993) reported a rapid growth of Salmonella in sliced cantaloupes, honeydew and watermelons when stored at 23 °C; but no increase when stored at 5 °C, in spite of having viable survivors in the samples (Golden et al., 1993). Our study was carried out over 21 days at 5 °C and data demonstrate a slow reduction in salmonellae over time. Thus, there appear to be factors in addition to temperature of storage that contribute towards the reduction in salmonellae on cantaloupes after irradiation. One likely factor would be the proliferation of surviving native biota such as LAB and yeasts that could compete with the pathogen for available nutrients and water. Thus, Salmonella may have been effectively antagonized by

Fig. 1. Survival of S. Poona on irradiated sliced cantaloupe stored at 5 °C. Symbols indicate mean S. Poona on cantaloupes treated by 0.0 (●), 0.7 (■), or 1.5 (▲) kGy electron beam irradiation from three independent replicates (n = 3); error bars indicate one standard deviation from the mean. Dashed line indicates limit of detection (1.0 log10 CFU/g).

69

native microbiota as seen in the survival curves (Fig. 1). There has not been much work done to study the interaction of pathogens with native microorganisms on irradiated fresh produce. Various LAB are known to produce antimicrobials such as diverse organic acids, antimicrobial polypeptides, and some peroxides, and are effective biocontrol agents against a variety of Gram-negative as well as Gram-positive pathogens (Kostrzynska and Bachand, 2006). It is possible that, in our study, indigenous irradiation-surviving LAB in the sliced cantaloupe secreted various antimicrobial substances that functioned to inhibit S. Poona growth over 21 days of storage. An interesting observation in our study was the comparative reduction of Salmonella at different doses of irradiation (Fig. 1). While it would be expected that a proportionate reduction would be observed at twice the dosage, this was not the case. Matic et al. (1990) determined the radiation resistance of differing inoculum levels of S. Enteriditis, S. Typhimurium and S. Lille in whole egg powder as a function of dose and storage time. They reported a similar trend in reduction as observed herein. At 1 kGy, they reported a reduction in salmonellae b1.0 log10, but at 2 kGy a 3.0 log10 reduction was observed. When the egg powder was inoculated at a level of 4.7 log10 CFU/g, they reported a reduction of around 0.5 log10 at 1 kGy and almost 2.0 log10 at 2 kGy (Matic et al., 1990). Such disproportionate reduction was not seen at doses higher than 2 kGy in their study. This indicates the possibility that irradiation results in varying levels of bacterial killing at differing dosages, with the consequent challenges in how reliable published D10-value data are as a guide for irradiation processing of foods. Further research needs to be done to investigate the effect of varying inoculum levels and irradiation doses on the efficiency of pathogen reduction on various commodities. Comparing various inoculation methods to replicate different contamination mechanism resulting from various growing, harvesting and postharvest processing conditions would also provide a better understanding on how produce can be processed to reduce pathogens. 3.1.2. Lactic acid bacteria The LAB counts on cut cantaloupes exposed to e-beam irradiation were reduced by 0.2 log10 CFU/g with 0.7 kGy and N2.4 (below the detection limit) log10 CFU/g with 1.5 kGy on day 0 (Fig. 2). There was no significant difference in counts between the control and samples irradiated at 0.7 kGy throughout the duration of the study (p ≥ 0.05), whereas at 1.5 kGy the LAB counts were significantly lower (p b 0.05) than the control up to day 15. This is consistent with previous work on tomato irradiation (Schmidt et al., 2006). Similarly, Kim et al. (2012) described a minimum gamma dose of 4.0 kGy required to achieve reduction in LAB numbers on fermented sausages, while doses up to 2.0 kGy did not reduce LAB numbers at 1 day of refrigerated

Fig. 2. Growth of lactic acid bacteria on irradiated sliced cantaloupe stored at 5 °C. Symbols indicate mean lactic acid bacteria on cantaloupes treated by 0.0 (●), 0.7 (■), or 1.5 (▲) kGy electron beam irradiation from three independent replicates (n = 3); error bars indicate one standard deviation from the mean. Dashed line indicates limit of detection (1.0 log10 CFU/g).

70

M.P. Palekar et al. / International Journal of Food Microbiology 202 (2015) 66–72

storage versus untreated controls. Yeasts, molds, and Gram-positive organisms such as the LAB are more resistant to irradiation than Gram-negative enteric organisms such as Salmonella (Monk et al., 1995). However at the higher dose (1.5 kGy), the decline in counts was greater than that observed at 0.7 kGy. This difference in bacterial reduction at different dose ranges follows the same trend that was observed in our results with S. Poona (Fig. 1). At 5 °C, the numbers of LAB increased steadily over 21 days of storage for both irradiation doses. Studies have shown that LAB grow well at refrigeration temperatures, especially in a sugar-rich medium like cantaloupe. They have been shown to increase by 2.3 log10 CFU/g on fresh-cut cantaloupe stored at 4 °C for 11 days and can also reach counts as high as 7.0 log10 CFU/g after storage for 5 days at 20 °C (Lamikanra et al., 2000). As shown in Fig. 2, LAB counts in samples irradiated at 1.5 kGy were below detection limits until day 3, after which they rose sharply to reach 7.4 log10 CFU/g by day 21 (Fig. 2). Their ability to grow rapidly at low temperatures gives them a significant advantage to proliferate and cause spoilage in fruit. Thus, reducing LAB initial numbers can significantly increase the shelf-life of fresh-cut fruit. At low temperatures, LAB have a competitive advantage over many Gram-negative bacteria. Bacteriocin production and reduction of pH by fermentation of sugars are two factors that provide them a competitive advantage over other bacterial populations. Further research is recommended to establish any antimicrobial effect exerted by lactic acid bacteria on Salmonella in irradiated cantaloupe. 3.1.3. Yeasts and mold Mean yeast populations on the controls and on cut cantaloupes subjected to 0.7 and 1.5 kGy were 2.1, 1.6 and 1.8 log CFU/g, respectively. This accounted for reduction values of 0.5 and 0.3 log cycles for 0.7 and 1.5 kGy, respectively. Neither the differences in yeast populations nor the log reductions were significant (p ≥ 0.05). These results are shown in Fig. 3. Yeasts have previously demonstrated higher resistance to radiation processing on some produce surfaces than bacteria (Neal et al., 2010), though reductions of fungi have been reported to be greater than bacteria following irradiation on smooth produce items (e.g., tomatoes) (Mahmoud, 2010). Over 21 days of storage, yeasts on control samples increased in numbers to a greater extent than on cantaloupes irradiated at 0.7 or 1.5 kGy, both of which exhibited no differences in counts through the entire storage period (p ≥ 0.05) (Fig. 3). At all doses, storage over 21 days at 5 °C did not result in an increase in numbers of yeasts equivalent to that observed for the LAB. Over 21 days of storage, the increase in yeast counts in the control, 0.7 and 1.5 kGy cantaloupes were 1.5, 1.1 and 0.9 log10 CFU/g respectively. This may be a result of slower growth at 5 °C or competing flora. Molds are known to be as sensitive to irradiation as some vegetative bacteria (Mahmoud, 2010). A similar observation was made in this

Fig. 3. Growth of yeasts on irradiated sliced cantaloupe stored at 5 °C. Symbols indicate mean yeasts on cantaloupes treated by 0.0 (●), 0.7 (■), or 1.5 (▲) kGy electron beam irradiation from three independent replicates (n = 3); error bars indicate one standard deviation from the mean. Dashed line indicates limit of detection (1.0 log10 CFU/g).

study with day 0 counts, where exposure to doses of 0.7 and 1.5 kGy resulted in a 0.5 and N1.5 (below detection) log10 CFU/g reduction in molds, respectively (Fig. 4). As in the case of LAB, mold counts in cut cantaloupes after irradiating at 1.5 kGy were significantly (p b 0.05) lower than those of the controls and the cut cantaloupes irradiated at 0.7 kGy, whereas at 0.7 kGy counts were not different from those of the controls (p ≥ 0.05). Unlike the growth patterns of LAB and to a lesser degree yeasts, mold counts in the control and irradiated melons showed a survival or very slow decline pattern up to 12 days of storage, and dropped below detection limits by day 15. The eventual drop in counts to levels close to or under the detection limit of the counting method may be due to competition for sugars and other nutrients that depleted with time due to activity of other competing organisms. Molds grow best under aerobic conditions and hence low levels of oxygen in the package may have contributed towards their reduction in counts over time. Aziz and Moussa (2002) irradiated a variety of fruits at doses of 1.5 and 3.5 kGy and enumerated surviving molds over time under refrigeration. They reported that on day 0, fruits irradiated with 1.5 kGy and 3.5 kGy had an average reduction in mold counts of around 2.0 and 3.0 log10 CFU/g, respectively, and even after 14 days under refrigeration they reported no increase in counts. In spite of the downward trend in our observations, there was a large fluctuation in counts on certain days of analysis; no visual mold was observed in control or irradiated melons after 21 days. O'Connor and Mitchell (1991) irradiated strawberries using gamma rays at 1.2 kGy, enumerating surviving yeasts and mold on day 1 and day 5 after storing the fruit at 8 °C. Because of the large variations in the counts obtained, these researchers concluded that yeast and mold counts are not useful as indicators for determining efficacy of irradiation. Our findings suggest that e-beam irradiation of cantaloupe is effective for reducing molds but not yeasts, and that storage at low temperature can keep surviving flora from growing to undesirable levels. The differences in resistance to irradiation in LAB, yeast, molds and Salmonella may be due to structural differences among these microorganisms, including cell wall thickness and composition, as well as possible physiological differences in resistance to irradiation between Grampositive and Gram-negative bacteria (Lenhart et al., 2012). 3.2. pH measurement pH often changes as a result of the microbiological activity in foods. A high sugar commodity like cantaloupe has a pH close to neutral when fresh. In this study, differences in cantaloupe pH between the control, 0.7 kGy and 1.5 kGy-exposed melon tissues over time were associated with the activity of LAB (Fig. 5). Initial pH values on samples after irradiation were 5.46, 5.79 and 6.10 for controls and cut cantaloupes

Fig. 4. Growth of molds on irradiated sliced cantaloupe stored at 5 °C. Symbols indicate mean molds on cantaloupes treated by 0.0 (●), 0.7 (■), or 1.5 (▲) kGy electron beam irradiation from three independent replicates (n = 3); error bars indicate one standard deviation from the mean. Dashed line indicates limit of detection (1.0 log10 CFU/g).

M.P. Palekar et al. / International Journal of Food Microbiology 202 (2015) 66–72

Fig. 5. Surface pH of irradiated sliced cantaloupe stored at 5 °C. Symbols indicate mean pH on cantaloupes treated by 0.0 (●), 0.7 (■), or 1.5 (▲) kGy electron beam irradiation from three independent replicates (n = 3); error bars indicate one standard deviation from the mean.

irradiated at 0.7 and 1.5 kGy, respectively. The pH of cut cantaloupes irradiated at 1.5 kGy was significantly different (p b 0.05) from the controls and 0.7 kGy. This difference may not be explained by the effect of fermentative microbiota, since these values were obtained within less than 2 h after processing. The effect of irradiation at 1.5 kGy on the pH of cut cantaloupes warrants further investigation to determine how this observation could affect the quality of cut cantaloupes. This effect may be either positive or negative. In fact, in a previous study sour taste was found to be significantly stronger on the control than on irradiated cut cantaloupes. In addition, aerobic plate counts, which may include LAB, remained significantly (p b 0.05) lower for the irradiated product over the storage time (Palekar et al, 2004). Throughout the storage period, control samples maintained a lower pH than irradiated cantaloupes, and this trend was more apparent when comparing controls and cantaloupes irradiated at 1.5 kGy. This may indicate maximum fermentation occurring in control samples. This is also supported by significantly lower counts observed for LAB throughout storage (see Fig. 2). Thus, e-beam irradiation was capable of reducing the numbers of fermentative microorganisms and the magnitude of the reduction was related to the irradiation dose. At the completion of storage (21 days) there was no significant difference between pH of the control (4.63) and cantaloupe irradiated at 0.7 kGy (4.83). The cantaloupe irradiated at 1.5 kGy maintained significantly (p b 0.05) higher pH through the end of storage (5.61) in comparison to the other treatments. Temperature of storage may also have played a role in limiting the decrease in pH during storage, by depressing the fermentation of sugars (Manvell and Ackland, 1986). Over 21 days at 5 °C, cantaloupe irradiated with 0.7 kGy showed the highest pH reduction of 0.96 while the control dropped by 0.83 and 1.5 kGy by 0.49. 3.3. D10-value for S. Poona in irradiated sliced cantaloupe The death curve for S. Poona on cantaloupe discs as affected by ebeam dose is shown in Fig. 6. There was a 4.8 log10 CFU/g reduction from the non-irradiated control (7.6 log10 CFU/g) on discs irradiated at 1.06 kGy. From this curve, a D10-value of 0.211 kGy for S. Poona in cantaloupe is proposed. Multiple factors play a role in determining the radiation sensitivity of organisms and hence their D10-value is specific to certain parameters under which it is determined. Composition of the medium is the most significant factor. According to Farkas (2007), cells irradiated in phosphate buffer are more sensitive compared to those irradiated in foodstuffs. According to Pillai (2004), presence of shielding or “quenching” molecules in foods can decrease nucleic acid damage in targeted microorganisms. This is especially significant in low-dose irradiated foods, where such scavenger molecules reduce the desired dose and render previously established D10-values ineffective. Other parameters that significantly influence irradiation efficacy

71

Fig. 6. Linear best-fit death curve of S. Poona with increasing doses of irradiation. The lines drawn serve the purpose of illustrating the D10-value but not for calculation.

are temperature of the product and processing, and moisture content (Farkas, 2007). Thus, it is important to understand the environmental conditions while determining D10-values for pathogens. Research initiatives in the past have focused on the irradiation of meats, with little work completed on produce. Hence, most of the literature cited relates pathogen reduction in meats. In a study conducted by Mulder (1984) D10-values were different for differing serotypes of Salmonella, ranging from 0.77 kGy in whole egg for S. Give to 0.33 kGy within the same product for S. Enteritidis. In ready-to-eat meats, Listeria monocytogenes displayed significant variability in radiation resistance. D10-values for L. monocytogenes inoculated on commercially available beef, mixed meat and poultry frankfurters ranged from 0.49 to 0.71 kGy (Sommers and Thayer, 2000). In further studies, these researchers reported that the D10-value of Salmonella serovars on beef/pork frankfurters was 0.70 kGy in 100% nitrogen atmosphere, but decreased to 0.61 kGy on frankfurters stored under 100% carbon dioxide (Sommers and Boyd, 2006). As shown in Fig. 1, a 1.1 log10 CFU/g reduction of S. Poona occurred at 0.7 kGy and 3.6 log10 CFU/g reduction at 1.5 kGy. Comparing these results with the D10-value of 0.211 kGy proposed, observed reductions on cut cantaloupe are less than those expected. A 3.0 log10 reduction should have occurred at 0.7 kGy and a 7.0 log10 reduction at 1.5 kGy, according to the D10-value. However, due to several differences in the way that the experiments were designed, this may amount to an invalid comparison. The effect of difference in the shape and size of the two samples on the irradiation efficiency was also studied. When the cylindrical pieces of cantaloupe were laid down and irradiated, the dual electron beam was incident on a curved surface from the top and bottom of the cylinder. When the 4 mm flat discs of cantaloupe were used, the electron beam was incident on a flat surface. The inability of e-beams to deliver a consistent dose over irregular surfaces is well documented and was conformed during dosimetry studies (data not shown). In addition, the small doses required for the D10-value studies required that this approach be used, especially with commercial scale accelerators as the accelerator used in this study. The difference in packaging environment may also be a factor to consider. According to Farkas (2007), the lethal effect of irradiation on microbial cells increases in the presence of oxygen. In an oxygen-free environment and in the presence of moisture, radiation resistance usually increases by a factor of 2 to 4. However, in dry conditions without oxygen, resistance can increase by a factor of 8 to 17. The cylinders were irradiated in packages containing 9% oxygen, whereas the discs were irradiated in air (20.9% oxygen). Low oxygen levels in the packet may have contributed to the increase in resistance of S. Poona to irradiation treatment. D10-values for any foodstuff should be established by taking into consideration all possible physical and environmental parameters beyond the characteristics of the target pathogen.

72

M.P. Palekar et al. / International Journal of Food Microbiology 202 (2015) 66–72

The bulk of studies on the benefits that ionizing irradiation can offer to reduce the risk of foodborne diseases seems to continue to be outweighed by consumers' concerns about safety and quality issues. While gamma irradiation uses radioactive sources of energy for applying the treatment, e-beam uses electricity to power the linear accelerators. The limitations of e-beams to deliver a uniform dose over materials with irregular surfaces, such as carcasses, or whole cantaloupes seems to be resolved by the use of the Maxim Electron Scatter Chamber, which takes advantage of the electron scattering to distribute the energy evenly over any type of surface (Cuervo et al., 2009; Maxim et al., 2014). This could reduce some sources of concern among consumers and would contribute to expand the awareness of the advantages of ebeam irradiation in improving food safety. As a conclusion, e-beam irradiation may be a viable process for reducing foodborne bacterial pathogens in cut melons such as cantaloupes, with the resulting reduction in the risk of foodborne diseases. This practice should not be used as a replacement for adequate food safety practices in growing and packing, but as a complementary measure to further reduce pathogens potentially present. E-beam irradiation has also been demonstrated to have a significant increase in shelf life of cut cantaloupes (Palekar et al., 2004), therefore, quality benefits with the expected parallel economic benefit can also be obtained if ebeam irradiation is approved and used for produce processing, provided that this treatment do not result in damage in the quality of the product. Acknowledgments This research was supported by a grant from the U.S. Department of Agriculture National Integrated Food Safety Initiative within the TriState Fruit and Vegetable Safety Consortium (00-52102-9637). References Aziz, N.H., Moussa, L.A.A., 2002. Influence of gamma-radiation on mycotoxin producing moulds and mycotoxins in fruits. Food Control 13, 281–288. Castell-Perez, E., Moreno, M., Rodriguez, O., Moreira, R.G., 2004. Electron beam irradiation treatment of cantaloupes: effect on product quality. Food Sci. Technol. Int. 10, 383–390. Castillo, A., Martínez-Téllez, M.A., Rodríguez-García, M.O., 2009. Melons. In: Sapers, G.M., Solomon, E.B., Matthews, K.R. (Eds.), The Produce Contamination Problem: Causes and Solutions. Academic Press, New York, NY, pp. 189–221. CDC, 1991. Epidemiologic notes and reports multistate outbreak of Salmonella poona infections — United States and Canada, 1991. Morb. Mortal. Wkly Rep. 40, 549–552. CDC, 2002. Multistate outbreaks of Salmonella serotype Poona infections associated with eating cantaloupe from Mexico—United States and Canada, 2000–2002. Morb. Mortal. Wkly Rep. 51, 1044–1047. CDC, 2008. Investigation update: outbreak of Salmonella Litchfield infections. [Online]. Available at, http://www.cdc.gov/salmonella/litchfield/ (Accessed: May 10, 2014). CDC, 2011. Investigation update: multistate outbreak of Salmonella Panama infections linked to cantaloupe. [Online]. Available at, http://www.cdc.gov/salmonella/ panama0311/032911/ (Accessed: May 10, 2014). CDC, 2012a. Multistate Outbreak of Listeriosis Linked to Whole Cantaloupes from Jensen Farms, Colorado. [Online]. Available at, http://www.cdc.gov/listeria/outbreaks/ cantaloupes-jensen-farms/index.html/ (Accessed: November 24, 2014). CDC, 2012b. Multistate outbreak of Salmonella Typhimurium and Salmonella Newport infections linked to cantaloupe (final update). [Online]. Available at, http://www.cdc.gov/salmonella/typhimurium-cantaloupe-08-12/ (Accessed: May 14, 2014). Cuervo, M.P., Rodrigues-Silva, D., Maxim, J., Castillo, A., 2009. Use of a novel device to enable irradiation of fresh cantaloupes by electron beam irradiation. International Association for Food Protection's 5th European Symposium on Food Safety, 7–9 October 2009, Berlin, Germany. Deeks, S., Ellis, A., Ciebin, B., Khakhria, R., Naus, M., Hockin, J., 1998. Salmonella Oranienburg, Ontario. Can. Commun. Dis. Rep. 24, 177–179.

Downes, F.P., Ito, K. (Eds.), 2001. Compendium of Methods for the Microbiological Examination of Foods, fourth ed. American Public Health Association, Washington, D.C. Fan, X., Annous, B.A., Sokorai, K.J.B., Burke, A., Mattheis, J.P., 2006. Combination of hotwater surface pasteurization of whole fruit and low-dose gamma irradiation of fresh-cut cantaloupe. J. Food Prot. 69, 912–919. Farkas, J., 2007. Physical methods of food preservation. In: Doyle, M.P., Beuchat, L.R. (Eds.), Food Microbiology: Fundamentals and Frontiers, third ed. ASM Press, Washington, DC, pp. 685–712. FDA, 2008. Guidance for industry: guide to minimize microbial food safety hazards of fresh-cut fruits and vegetables. [Online]. Available at, http://www.fda.gov/Food/ GuidanceRegulation/GuidanceDocumentsRegulatoryInformation/ ProducePlantProducts/ucm064458.htm#ch8 (Accessed: April 29, 2014). FDA, 2011. Detention without physical examination of cantaloupes from Mexico: Import Alert # 22-01. [Online]. Available at, http://www.accessdata.fda.gov/cms_ia/ importalert_67.html (Accessed: May 10, 2014). Golden, D.A., Rhodehamel, E.J., Kautter, D.A., 1993. Growth of Salmonella spp. in cantaloupe, watermelon, and honeydew melons. J. Food Prot. 56, 194–196. Kim, I.S., Jo, C., Lee, K.H., Lee, E.J., Ahn, D.U., Kang, S.N., 2012. Effects of low-level gamma irradiation on the characteristics of fermented pork sausage during storage. Radiat. Phys. Chem. 81, 466–472. Kostrzynska, M., Bachand, A., 2006. Use of microbial antagonism to reduce pathogen levels on produce and meat products: a review. Can. J. Microbiol. 52, 1017–1026. Lamikanra, O., Chen, J.C., Banks, D., Hunter, P.A., 2000. Biochemical and microbial changes during the storage of minimally processed cantaloupe. J. Agric. Food Chem. 48, 5955–5961. Lenhart, J.S., Schroeder, J.W., Walsh, B.W., Simmons, L.A., 2012. DNA repair and genome maintenance in Bacillus subtilis. Microbiol. Mol. Biol. Rev. 76, 530–564. Mahmoud, B.S.M., 2010. The effects of X-ray radiation on Escherichia coli O157:H7, Listeria monocytogenes, Salmonella enterica, and Shigella flexneri inoculated on whole Roma tomatoes. Food Microbiol. 27, 1057–1063. Mandrell, R.E., Gorski, L., Brandl, M.T., 2006. Microbiology of Fruits and Vegetables. CRC Press Inc., Albany, CA. Manvell, P.M., Ackland, M.R., 1986. Rapid detection of microbial growth in vegetable salads at chill and abuse temperatures. Food Microbiol. 3, 59–65. Matic, S., Mihokovic, V., Katusinrazem, B., Razem, D., 1990. The eradication of Salmonella in egg powder by gamma-irradiation. J. Food Prot. 53, 111–114. Maxim, J.E., Neal, J.A., Castillo, A., 2014. Development of a novel device for applying uniform doses of electron beam irradiation on carcasses. Meat Sci. 96, 373–378. Mohle-Boetani, J., Reporter, R., Werner, S.B., Abbott, S., Farrar, J., Waterman, S.H., Vugia, D.J., 1999. An outbreak of Salmonella serogroup Saphra due to cantaloupes from Mexico. J. Infect. Dis. 180, 1361–1364. Monk, J.D., Beuchat, L.R., Doyle, M.P., 1995. Irradiation inactivation of food-borne microorganisms. J. Food Prot. 58, 197–208. Mulder, R.W.A.W., 1984. Ionizing energy treatment of poultry. Food Technol. Aust. 36, 418–420. Neal, J.A., Booren, B., Cisneros-Zevallos, L., Miller, R.K., Lucia, L.M., Maxim, J.E., Castillo, A., 2010. Shelf life and sensory characteristics of baby spinach subjected to electron beam irradiation. J. Food Sci. 75, S319–S326. O'Connor, R.E., Mitchell, G.E., 1991. Effect of irradiation on microorganisms in strawberries. Int. J. Food Microbiol. 12, 247–256. Palekar, M.P., Cabrera-Diaz, E., Kalbasi-Ashtari, A., Maxim, J.E., Miller, R.K., CisnerosZevallos, L., Castillo, A., 2004. Effect of electron beam irradiation on the bacterial load and sensorial quality of sliced cantaloupe. J. Food Sci. 69, M267–M273. Parnell, T.L., Harris, L.J., 2003. Reducing Salmonella on apples with wash practices commonly used by consumers. J. Food Prot. 66, 741–747. Parnell, T.L., Harris, L.J., Suslow, T.V., 2005. Reducing Salmonella on cantaloupes and honeydew melons using wash practices applicable to postharvest handling, foodservice, and consumer preparation. Int. J. Food Microbiol. 99, 59–70. Pillai, S.D., 2004. Food irradiation. In: Beier, R.C., Pillai, S.D., Phillips, T.D. (Eds.), Preharvest and Postharvest Food Safety. Blackwell Publishing, Ames, IA, pp. 375–387. Prakash, A., Foley, D., 2004. Improving safety and extending shelf life of fresh-cut fruits and vegetables using irradiation. In: Komolprasert, V., Morehouse, K. (Eds.), Irradiation of Food and Packaging: Recent Developments. American Chemical Society, Washington, DC, pp. 90–106. Schmidt, H.M., Palekar, M.P., Maxim, J.E., Castillo, A., 2006. Improving the microbiological quality and safety of fresh-cut tomatoes by low-dose electron beam irradiation. J. Food Prot. 69, 575–581. Sommers, C.H., Boyd, G., 2006. Radiation sensitivity and postirradiation growth of foodborne pathogens on a ready-to-eat frankfurter on a roll product in the presence of modified atmosphere and antimicrobials. J. Food Prot. 69, 2436–2440. Sommers, C.H., Thayer, D.W., 2000. Survival of surface-inoculated Listeria monocytogenes on commercially available frankfurters following gamma irradiation. J. Food Saf. 20, 127–137. USDA-ERS, 2012. Fruit & Tree Nuts. Available at, http://www.ers.usda.gov/topics/crops/ fruit-tree-nuts/background.aspx (Accessed: Nov 24, 2014).