Division of Food Bioscience and Technology, College of Life Sciences and Biotechnology, Korea University, Seoul, South Korea

Letters in Applied Microbiology ISSN 0266-8254 ORIGINAL ARTICLE Changes in microbial contamination levels and prevalence of foodborne pathogens in alfalfa (Medicago sativa) and rapeseed (Brassica napus) during sprout production in manufacturing plants S.A. Kim, O.M. Kim and M.S. Rhee Division of Food Bioscience and Technology, College of Life Sciences and Biotechnology, Korea University, Seoul, South Korea Significance and Impact of Study: The present study investigated the levels of microbial contamination present in alfalfa and rapeseed sprouts by examining the samples taken at different stages of the manufacturing process in three actual plants. The results provide detailed information regarding the levels of seed and sprout contamination during production. The results may be useful to those involved in the sprout industry and/or academic research in terms of developing hygienic control measures, efficient intervention methods and appropriate guidelines. Keywords alfalfa, foodborne pathogens, microbial contamination, rapeseed, sprout manufacturing process. Correspondence Min Suk Rhee, Division of Food Bioscience and Technology, College of Life Sciences and Biotechnology, Korea University, 5-1 Anam-dong, Sungbuk-gu, Seoul, 136-713, South Korea. E-mail: rheems@korea.ac.kr 2012/1229: received 9 August 2012, revised 21 September 2012 and accepted 2 October 2012 doi:10.1111/lam.12009 Abstract Samples were taken from three sprout processing plants at five different stages of production (a total of 20 investigations). Quantitative analyses comprised aerobic plate counts (APCs) and the measurement of coliforms and Bacillus cereus levels, whereas qualitative analyses involved assessing the levels of Escherichia coli and major foodborne pathogens (E. coli O157:H7, Listeria monocytogenes, Salmonella spp., and Staphylococcus aureus). The APC for alfalfa seeds (371 461 log CFU g 1 ) and rapeseed (425 511 log CFU g 1 ) increased by approximately 3 log CFU g 1 during sprouting, reaching 717 761 and 733 828 log CFU g 1, respectively, by the final stage of production. Similarly, increasing trends were noted in the level of coliforms (058 403 log CFU g 1 at the seed stage, increasing to 552 699 log CFU g 1 by the sprout stage). Bacillus cereus was detected in eight alfalfa (40%) and 14 rapeseed (70%) sprouts, and L. monocytogenes was isolated from one pregermination soaked alfalfa seed. A slight reduction in the level of bacterial contamination was noted after washing the sprouts with water prior to storage, indicating that improvements to the current washing protocol, or other efficient intervention methods, may be needed. Taken together, these results suggest that improved hygiene control during production and processing and a more sanitary environment are needed. The present study provides comprehensive information regarding the microbiological safety of seeds and sprouts during manufacturing. Introduction In many countries, there is an increase in the consumption of freshly produced vegetables and fruits for their health benefits, and because of significant consumer lifestyle changes that increasingly value fresh and natural foods (Viswanathan and Kaur 2001; Abadias et al. 2008; Kwak et al. 2011). Vegetable sprouts are valuable dietary foods and have become very popular natural foods among health-conscious consumers (Jung et al. 2009). Sprouts generally contain rich nutrients including proteins, minerals and vitamins (Weiss and Hammes 2003; Waje et al. 2009). Various types of sprouts including alfalfa, rapeseed, mung bean and radish have had robust market sales in 30 Letters in Applied Microbiology 56, 30--36 2012 The Society for Applied Microbiology

S.A. Kim et al. Microbial change in sprouts during production recent years. In addition to the nutrients and health benefits of sprout consumption, there are potential risks that come from eating contaminated sprouts. Sprouts have increasingly emerged as a recognized source of foodborne disease in many parts of the world, and the U.S. Food and Drug Administration classified raw sprouts as a potentially hazardous food (Bremer and Fielding 2003; Peñas et al. 2008). Various foodborne pathogens including Salmonella spp., Escherichia coli O157:H7, Listeria monocytogenes and Bacillus cereus are associated with outbreaks caused by the consumption of contaminated sprouts (Ponka et al. 1995; Jung et al. 2009). In addition, sprouts are usually consumed as a raw product without any attempts to control micro-organisms. Therefore, the potential risks were high for developing serious illnesses. It is unlikely that vegetables are free from microbial contamination. Sprouts can be contaminated by different types of bacteria from a variety of sources, including the soil in which the seeds are grown and the water used during the production process (Sagoo et al. 2003). When radish seeds were grown in water inoculated with E. coli O157:H7, the edible parts of the sprouts, including the cotyledons and hypocotyls, showed high levels of bacterial contamination (Hara-Kudo et al. 1997). In another study, radish seeds were cultivated with E. coli O157:H7; subsequently, viable bacteria were detected in both external and internal plant tissues (Itoh et al. 1998). Taken together, these results suggest several possible routes of contamination during sprout production. An increase in the microbial load during sprout germination has been reported. Weiss et al. (2007) and Splittstoesser et al. (1983) investigated the changes in the microbial load that occurred during production under laboratory conditions; however, both of these studies only examined germinating sprouts. Neither group performed microbial analysis of samples taken at each of the different stages of production from actual manufacturing plants. Other processing stages, such as seed soaking, germination, sprouting and sprout washing, provide many opportunities for bacterial contamination or reduction in population. At present, limited information is available regarding the changes in the microbial load in seeds and sprouts during the processing steps at individual manufacturing plants. If we are to improve microbiological safety at manufacturing plants, more information is required regarding the types of foodborne pathogen that contaminate sprouts, and how these microbes colonize sprouts at the different stages of production. Effective and efficient methods of reducing or eliminating microbial contamination are also needed. The present study examined the microbiological quality of alfalfa (Medicago sativa) and rapeseed (Brassica napus) sprouts by sequential sampling in three manufacturing plants. Generally, sprouts are produced via several sequential steps: pregermination soaking in water, germination, and sprouting and growth. Product line samples at five stages including seed storage, pregermination soak, germination, sprouting and sprout storage were obtained from sprout processing plants. Quantitative contamination levels of aerobic plate count (APC), coliforms and B. cereus were analysed, and qualitative contamination levels of E. coli and foodborne pathogenic bacteria (E. coli O157:H7, L. monocytogenes, Salmonella spp., and Staphylococcus aureus) were investigated. Analysis of the samples provided information about the pathogens associated with the samples and the change in microbial loads during sequential manufacturing steps. The results identify intervention points and efficient methods for improving microbial safety. Results and discussion Tables 1 and 2 show bacterial counts of aerobic plate and coliforms for product line samples of alfalfa and rapeseed, respectively. The naturally occurring contamination of APC and coliforms was observed in seeds of alfalfa and rapeseed. All tested seeds of alfalfa and rapeseed had APC of 371 461 log CFU g 1 and 425 511 log CFU g 1, respectively. There was no significant difference between the three processing plants (P > 005). Initial levels of coliforms in seed ranged from 204 to 403 log CFU g 1 for alfalfa and 058 151 log CFU g 1 for rapeseed. Coliforms are found everywhere in the environment (Saroj et al. 2006), so they can come from a variety of sources including soil, water and workers. The rich nutrition of sprouts has the potential to be exploited for microbial contamination and growth (Buck et al. 2003), and the temperature and moisture required for production provide favourable conditions for bacterial growth (Taormina et al. 1999). The APC and coliforms in raw seeds dramatically increased during germination. The final products had much higher coliform counts than those of seeds. The initial contamination of APC in seeds was maintained or slightly increased during the pregermination soak for alfalfa, whereas the initial contamination was slightly reduced in rapeseed (Table 1). During germination, the APCs were dramatically increased up to 714 870 log CFU g 1 for alfalfa and 767 866 log CFU g 1 for rapeseed. During the sprouting step, the APCs were maintained with high contamination levels (>8 log CFU g 1 ). The APC of sprouts was slightly reduced by washing, but they still presented high counts of 717 761 CFU g 1 for alfalfa sprouts and 733 828 log CFU g 1 for rapeseed sprouts. Coliform counts also dramatically increased during germination; the counts in the germination step were significantly higher than those of the pregermination soak, which Letters in Applied Microbiology 56, 30--36 2012 The Society for Applied Microbiology 31

Microbial change in sprouts during production S.A. Kim et al. Table 1 Bacterial count of aerobic plate in product line sample of alfalfa (Medicago sativa) and rapeseed (Brassica napus) at five stages including seed storage, pregermination soak, germination, sprouting and final products of sprouts Sample type Sprout processing plant Bacterial count (log CFU g 1 ) Seed storage Pregermination soak Germination Sprouting Sprout storage Alfalfa (M. sativa) A 371 ± 034 C 502 ± 002 Ba 714 ± 018 Ab * 717 ± 074 Ab B 461 ± 012 D 521 ± 040 Ca 870 ± 028 a,a 829 ± 002 Aa 761 ± 024 Ba C 413 ± 071 C 425 ± 057 Cb 837 ± 078 a,a 811 ± 076 Aab 747 ± 073 Bab Rapeseed (Br. napus) A 511 ± 041 B 399 ± 013 C 767 ± 048 A 733 ± 070 Ab B 425 ± 079 B 382 ± 037 B 866 ± 027 A 884 ± 039 Aa 828 ± 028 Aa C 474 ± 078 B 453 ± 075 B 836 ± 086 A 806 ± 059 Aab Results are expressed as means ± SD. Mean values in the same row followed by different superscript letters (A C) represent statistically significant differences (difference within five stages) (P < 005). Mean values in the same column followed by different superscript letters (a, b) represent statistically significant differences (difference within three processing plants) (P < 005). *Not analysed. increased by 308 (plant A), 399 (plant B) and 516 (plant C) log CFU g 1 for alfalfa and 449 (plant A), 542 (plant B) and 541 (plant C) log CFU g 1 for rapeseed (Table 2). These levels were maintained with 694 754 log CFU g 1 in the sprouting step. The sprouts from the final stage had slightly lower numbers of coliforms compared with the prestage, which is attributed to the washing procedure. Sprouts of alfalfa and rapeseed had coliforms >6 log CFU g 1, except for alfalfa from plant A. The contamination levels were 552 665 log CFU g 1 for alfalfa and 644 699 log CFU g 1 for rapeseed. The B. cereus counts for product line samples of alfalfa and rapeseed are shown in Table 3. Bacillus cereus was not detected in any product line samples of alfalfa from plants A and B, except for one seed sample from plant B. In samples from plant C, B. cereus counts in alfalfa seeds were 011 log CFU g 1, and they were maintained at the pregermination soak. The contamination level of B. cereus reached 163 log CFU g 1 during germination and 139 log CFU g 1 from final alfalfa sprouts. For rapeseed, B. cereus was detected in samples from all three processing plants. Rapeseed contamination during production had the same increasing trends, as follows: 073 273 log CFU g 1 for seed storage, 114 246 log CFU g 1 for pregermination soak, 212 323 log CFU g 1 for germination and 204 273 CFU g 1 for final sprouts. This study established a B. cereus prevalence of 400% (8/20) and 700% (14/20) in the final products for alfalfa and rapeseed sprouts, respectively. Bacillus cereus is ubiquitous in nature, especially in soil (Helgason et al. 2000), and contamination of the sprouts may be attributed to the soil used for seed growth. Bacillus cereus is a spore-forming bacteria; this property confers their high resistance to various treatments and poses difficulties for microbial control. If B. cereus forms spores in sprouts, they can survive for a long time during distribution and germinate in Table 2 Bacterial count of coliforms in product line sample of alfalfa (Medicago sativa) and rapeseed (Brassica napus) at five stages including seed storage, pregermination soak, germination, sprouting and final products of sprouts Sample type Sprout processing plant Bacterial count (log CFU g 1 ) Seed storage Pregermination soak Germination Sprouting Sprout storage Alfalfa (M. sativa) A 204 ± 177 B 357 ± 032 B 665 ± 061 Ab * 552 ± 041 A B 403 ± 020 C 408 ± 051 C 807 ± 028 Aa 754 ± 045 Aa 665 ± 048 B C 278 ± 193 B 210 ± 200 B 727 ± 079 Aab 694 ± 050 Aab 646 ± 081 A Rapeseed (Br. napus) A 100 ± 173 B 237 ± 207 B 686 ± 126 A 644 ± 090 A B 151 ± 261 B 193 ± 083 B 736 ± 056 A 748 ± 030 A 699 ± 072 A C 058 ± 151 B 113 ± 187 B 654 ± 205 A 669 ± 065 A Results are expressed as means ± SD. Mean values in the same row followed by different superscript letters (A C) represent statistically significant differences (difference within five stages) (P < 005). Mean values in the same column followed by different superscript letters (a, b) represent statistically significant differences (difference within three processing plants) (P < 005). *Not analysed. 32 Letters in Applied Microbiology 56, 30--36 2012 The Society for Applied Microbiology

S.A. Kim et al. Microbial change in sprouts during production Table 3 Bacterial count of Bacillus cereus in product line sample of alfalfa (Medicago sativa) and rapeseed (Brassica napus) at five stages including seed storage, pregermination soak, germination, sprouting and final products of sprouts Sample type Sprout processing plant Bacterial count (log CFU g 1 ) Seed storage Pregermination soak Germination Sprouting Sprout storage Alfalfa (M. sativa) A ND ND ND * ND B 068 ± 118 ND ND ND ND C 011 ± 039 B 019 ± 071 B 163 ± 155 A 130 ± 179 A 139 ± 158 A Rapeseed (Br. napus) A 073 ± 127 B 117 ± 105 B 323 ± 085 A 243 ± 027 AB B 273 ± 042 114 ± 122 212 ± 199 262 ± 029 273 ± 046 C 224 ± 166 246 ± 130 287 ± 210 204 ± 211 Results are expressed as means ± SD. Mean values in the same row followed by different superscript letters represent statistically significant differences (difference within five stages) (P < 005). ND, not detected. *Not analysed. favourable conditions. The results presented in the present study suggest that criteria regarding the safe level of B. cereus in sprouts should be established. Listeria monocytogenes and E. coli were isolated only from plant C. Listeria monocytogenes was detected in one alfalfa seed sample from the pregermination soak step. Escherichia coli was detected in two samples, including one alfalfa from the germination step and one rapeseed from the seed storage step. Listeria monocytogenes was isolated from one alfalfa sprout sample. Listeria monocytogenes is a psychrotrophic pathogenic bacteria (Wilkins et al. 1972) that have the potential to grow in sprouts during subsequent steps including distribution, storage and placement in markets for retail. The microbiologically contaminated sprouts identified in this study could be a potential risk of foodborne pathogens for consumers. The detection rate of B. cereus and coliforms increased during the production process. Fourteen rapeseed seeds (14/20, 467%) were contaminated with coliforms, whereas no coliforms were detected in eight seeds. However, by the final stage of production (sprouting and growth), all of the samples were contaminated with coliforms. The same trend was noted for B. cereus, which was detected in only 100% (2/20) of alfalfa seed samples but in 350% (7/20) of sprout samples. These results may be due to the rapid growth of micro-organisms during sprouting. Cross-contamination may also be considered to be a contributing factor to the increased detection rate. Each of the different processing stages can be a potential point of microbial contamination. Cross-contamination may occur after exposure of sprouts to contaminated seeds or may be due to the water used for germination and/or washing, improperly sterilized utensils and equipment, or improper handling by workers with poor personal hygiene practices. In particular, the germination step, in which large numbers of seeds are sprouted in a rotary drum sprouter or in sprout trays, may be a major source of microbial cross-contamination. All three processing plants had a water washing procedure right before the storage of the final product. The sprouts were washed in water with gentle agitation in a large water bath. The APC and coliforms and B. cereus counts were slightly reduced in final products compared with those of the prestep. This result indicates that the washing method used in the three plants may have limited utility for microbial decontamination. Improvement of the washing step or introduction of other alternative methods is required for sprout production. Effective sanitation management, hygienic production facilities and appropriate education programmes for those working on the production line are expected to increase microbiological safety. Indeed, a previous study points out the need for improved monitoring and evaluation of sanitation and sterility during the production process (Yung and Ponce 2008); thus, appropriate controls must be in place to ensure the microbiological safety of sprouts produced in processing plants. The industry requires appropriate handling and cleaning of seed, maintenance and control of sanitary facilities, and good personnel hygiene. Development of efficient interventions and regular employee training also are essential. For consumers, special care must be taken to wash raw sprouts well before the consumption. Governments should continuously inform and educate society about the potential hazards of sprouts and proper handling methods at home using education programmes, media campaigns and the distribution of written materials. In conclusion, the microbial loads and the prevalence of foodborne pathogens were investigated from product line samples of a sprout manufacturing process. To our knowledge, this is the first report on the microbiological quality of samples taken directly from all sequential steps Letters in Applied Microbiology 56, 30--36 2012 The Society for Applied Microbiology 33

Microbial change in sprouts during production S.A. Kim et al. in processing plants, from raw material (seed) to final products (sprouts). The results of these analyses provide the following conclusions: (i) seeds had naturally occurring APC, coliforms and B. cereus, which indicated that proper cleaning of seed should be required; (ii) the microbial loads in seeds increased during subsequent steps, especially during germination, and high population levels were maintained up to the final product; (iii) the presence of foodborne pathogens including B. cereus and L. monocytogenes is a potential risk for the consumer; and (iv) the washing step that was conducted in all three plants before packaging of final products was not sufficient for microbial decontamination, and other efficient interventions are needed for the industry. This study provides important basic information on sprout research and production, and related food industries. Materials and methods Sample collection During March to May 2008, product line samples for alfalfa and rapeseed were collected from three sprout processing plants (A, B and C). Evaluation of the microbiological safety of samples was performed three times for each plant. Because the sanitation environment of plant C was relatively poor compared with other plants, and microbial contamination was high in samples from that plant, 11 additional investigations were conducted for plant C during June to mid-december 2008 (there was no seasonal variation). A total of 20 investigations (A = 3, B = 3 and C = 14) were performed. A total of 180 product line samples were evaluated, including alfalfa (n = 97) and rapeseed (n = 83). Each stepwise sample was collected for the analysis of microorganisms from the following steps: (i) seed storage, unwashed raw seeds of alfalfa and rapeseed stored in warehouse (the alfalfa seeds originated in Italy or Pakistan; the origin of rapeseed seed was Korea or Italy); (ii) pregermination soak, seeds are soaked for 6 8 h in water before germination to improve the germination rate; (iii) germination, germination in rotary drum sprouter with intermittent water supply for approximately 48 h (water supply frequency: every 6 h at plants A and C and every hour at plant B); (iv) sprouting, growth of sprouts in sprout trays with intermittent water spray for approximately 24 h; and (v) sprouts storage, subsequent refrigeration storage of sprouts within 24 h of production (before storage, products were washed with water and dehydrated using a spin dryer). There was no sprouting step for alfalfa and rapeseed from plant A and rapeseed from plant C; instead, both germination and sprouting were performed in the rotary drum sprouter. The collected samples were individually placed into a sterile sampling bag (Whirl-Pak Bags; Nasco, Modesto, CA, USA) and moved to the laboratory stored in cooler bags with ice packs within 4 h. Quantitative analysis: APC, coliforms and Bacillus cereus Sample preparation. The contamination levels as determined by APC, coliforms and B. cereus counts were analysed. Samples (25 g) were aseptically taken and homogenized with 225 ml of sterile 02% peptone water using a stomacher (Circulator 400; Seward, Worthing, UK) at 230 rev min 1 for 2 min. One millilitre of homogenized sample was serially diluted with 9 ml of 02% peptone water to obtain tenfold diluted samples. APC and coliforms. For APC and coliforms enumeration, 100 ll of the serial dilutions was spread-plated in duplicate on plate count agar (PCA; Difco, Becton Dickinson, Sparks, MD, USA) and violet red bile agar (VRBA, Difco), respectively. To obtain a lower detection limit (1 log CFU g 1 ), 200 ll of tenfold dilutions (10 1 ) of samples was spread-plated on five plates each of PCA and VRBA. PCA and VRBA plates were incubated at 35 C for 48 h and 37 C for 24 h, respectively. After incubation, APC and coliforms were enumerated by counting the typical colonies on PCA and the violet colonies on VRBA. Bacillus cereus enumeration. Bacillus cereus counts were obtained using the microbiological methods outlined in the Korean Food Code (Korean Food and Drug Administration 2002). Mannitol egg yolk polymyxin agar (MYP, Difco) supplemented with 50% egg yolk emulsion and polymyxin B was used for the detection of B. cereus. Aliquots (100 ll) of dilutions were spread-plated in duplicate on MYP. The 200-ll aliquot of tenfold dilutions (10 1 ) of samples was spread-plated on five MYP plates and incubated at 30 C for 24 h. After incubation, five colonies exhibiting the typical pink colonies surrounded with a precipitation zone were randomly picked and transferred to tryptic soy agar (TSA; Difco). TSA plates were incubated at 30 C for 24 h and then identified using the PCR kit (PowerChek TM Bacillus cereus detection kit; Kogene Biotech Co., Seongnam, Korea). The final count was calculated by dividing the number of confirmed colonies by five. Qualitative analysis. The presence of foodborne pathogens during sprout production was determined in samples from each of the five sequential manufacturing steps defined above. Generic E. coli and E. coli O157:H7, L. monocytogenes, Salmonella spp. and Staph. aureus were isolated according to the microbiological methods of the 34 Letters in Applied Microbiology 56, 30--36 2012 The Society for Applied Microbiology

S.A. Kim et al. Microbial change in sprouts during production Korean Food Code (Korean Food and Drug Administration 2002) with some modifications. Escherichia coli and Escherichia coli O157:H7. Samples (25 g) were transferred to 225 ml of modified E. coli broth (EC broth, Difco) and then homogenized. The mixture was incubated at 37 C for 24 h. The enrichment culture was then streak-plated using a flamed platinum loop (2 mm in diameter) on eosin methylene blue agar (EMB; Difco) and MacConkey sorbitol agar (SMAC; Difco) and incubated at 37 C for 24 h. Purple colonies with green metallic sheen on EMB were transferred to TSA and incubated at 37 C for 24 h before biochemical identification by API 20E for the confirmation of generic E. coli. Typical colourless colonies on both EMB and SMAC were streaked on TSA, incubated at 37 C for 24 h and identified by the PCR kit (PowerChek E. coli O157 Detection Kit). The O and H antigens were serologically classified using slide agglutination tests with E. coli antisera group Osera (Enka Seiken, Tokyo, Japan) following the manufacturer s instructions. Listeria monocytogenes. Samples (25 g) were placed in stomacher bags containing 225 ml of University of Vermont Medium (UVM)-modified Listeria enrichment broth (Difco) and homogenized. The mixture was incubated at 30 C for 24 h. After incubation, 01 mlofuvm broth was inoculated into 10 ml of Fraser Listeria broth (Difco) (supplemented with Fraser Listeria broth supplement) and incubated at 30 C for 24 h. The enrichment culture was then streak-plated using a flamed platinum loop (2 mm in diameter) on modified Oxford agar (Oxoid, Hampshire, UK) and incubated at 30 C for up to 24 h. Black and shiny colonies with a black halo were picked, transferred to TSA supplemented with 06% yeast extract and incubated at 30 C for 24 h. Colonies were confirmed using the CAMP test with a positive for Staph. aureus ATCC 25923 and negative for Rhodococcus equi ATCC 6939, and the PCR kit (PowerChek Listeria monocytogenes detection kit). Salmonella spp. Samples (25 g) were added to 225 ml of buffered peptone water and then homogenized. The mixture was incubated at 37 C for 18 h. After incubation, 01 ml of pre-enriched culture was transferred to 10 ml of Rappaport-Vassiliadis broth (Oxoid) and incubated at 42 C, and 1 ml of the culture was transferred to 10 ml of selenite F broth (Difco) and incubated at 37 C. The enrichment culture was then streak-plated using a flamed platinum loop (2 mm in diameter) on xylose lysine deoxycholate agar (Difco), and the plates were incubated at 37 C for 24 h. Suspected black colonies were transferred to TSA and incubated at 37 C for 24 h. Colonies were identified using the PCR kit (PowerChek Salmonella spp. detection kit). Staphylococcus aureus. Samples (25 g) were placed in stomacher bags containing 225 ml of brain-heart infusion broth (Difco) and homogenized. The mixture was incubated at 35 C for 24 h. The enriched sample was streakplated on Baird-Parker medium (Difco) supplemented with 5% egg yolk tellurite emulsion with supplemented egg yolk and incubated at 35 C for 48 h. Presumptive grey-black to jet-black circular colonies were picked, transferred to TSA and incubated at 35 C for 24 h. They were then subjected to a coagulase test with EDTA (Becton Dickinson, BBL, Franklin Lakes, NJ, USA), and coagulase-positive cultures were confirmed using the PCR kit (PowerChek Staphylococcus aureus detection kit). Statistical analysis The duplicate plate count averages for the three replicates were converted to log CFU g 1 for analysis of variance. Data were analysed by analysis of variance (ANOVA) using the SAS statistical package (SAS 8.2; SAS Institute, Cary, NC, USA). When the analyses of variance indicated significant differences, least-square means were separated by the probability option (PDIFF, a pairwise t-test). Acknowledgements This study was partially supported by the Korea University grant and Miraewon Co. The authors thank the Institute of Biomedical Science and Food Safety, Korea University Food Safety Hall, for providing the equipment and facilities. References Abadias, M., Usall, J., Anguera, M., Solsona, C. and Viñas, I. (2008) Microbiological quality of fresh, minimallyprocessed fruit and vegetables, and sprouts from retail establishments. Int J Food Microbiol 123, 121 129. Bremer, P.J. and Fielding, T. (2003) The safety of seed and bean sprouts: risks and solutions. NZ Food J 3, 33 40. Buck, J., Walcott, R. and Beuchat, L. (2003) Recent trends in microbiological safety of fruits and vegetables [online]. Plant Health Prog, doi: 10.1094/PHP-2003-0121-01-RV. Hara-Kudo, Y., Konuma, H., Iwaki, M., Kasuga, F., Sugita- Konishi, Y., Ito, Y. and Kumagai, S. (1997) Potential hazard of radish sprouts as a vehicle of Escherichia coli O157:H7. J Food Prot 60, 1125 1127. Helgason, E., Økstad, O.A., Caugant, D.A., Johansen, H.A., Fouet, A., Mock, M., Hegna, I. and Kolstø, A.B. (2000) Bacillus anthracis, Bacillus cereus, and Bacillus thuringiensis Letters in Applied Microbiology 56, 30--36 2012 The Society for Applied Microbiology 35

Microbial change in sprouts during production S.A. Kim et al. one species on the basis of genetic evidence. Appl Environ Microbiol 66, 2627 2630. Itoh, Y., Sugita-Konishi, Y., Kasuga, F., Iwaki, M., Hara-Kudo, Y., Saito, N., Noguchi, Y., Konuma, H. et al. (1998) Enterohemorrhagic Escherichia coli O157:H7 present in radish sprouts. Appl Environ Microbiol 64, 1532 1535. Jung, W.Y., Choi, Y.M. and Rhee, M.S. (2009) Potential use of supercritical carbon dioxide to decontaminate Escherichia coli O157:H7, Listeria monocytogenes, and Salmonella typhimurium in alfalfa sprouted seeds. Int J Food Microbiol 136, 66 70. Korean Food and Drug Administration (2002) Korean Food Code. Seoul, Korea: Korean Food and Drug Administration. Kwak, T.Y., Kim, N.H. and Rhee, M.S. (2011) Response surface methodology-based optimization of decontamination conditions for Escherichia coli O157:H7 and Salmonella Typhimurium on fresh-cut celery using thermoultrasound and calcium propionate. Int J Food Microbiol 150, 128 135. Peñas, E., Gómez, R., Frías, J. and Vidal-Valverde, C. (2008) Application of high-pressure treatment on alfalfa (Medicago sativa) and mung bean (Vigna radiata) seeds to enhance the microbiological safety of their sprouts. Food Control 19, 698 705. Ponka, A., Andersson, Y., Siitonen, A., De Jong, B., Jahkola, M., Haikala, O., Kuhmonen, A. and Pakkala, P. (1995) Salmonella in alfalfa sprouts. Lancet 345, 462 463. Sagoo, S.K., Little, C.L., Ward, L., Gillespie, I.A. and Mitchell, R.T. (2003) Microbiological study of readyto-eat salad vegetables from retail establishments uncovers a national outbreak of salmonellosis. J Food Prot 66, 403 409. Saroj, S.D., Shashidhar, R., Dhokane, V., Hajare, S., Sharma, A. and Bandekar, J.R. (2006) Microbiological evaluation of sprouts marketed in Mumbai, India, and its suburbs. J Food Prot 69, 2515 2518. Splittstoesser, D., Queale, D. and Andaloro, B. (1983) The microbiology of vegetable sprouts during commercial production. J Food Saf 5, 79 86. Taormina, P.J., Beuchat, L.R. and Slutsker, L. (1999) Infections associated with eating seed sprouts: an international concern. Emerg Infect Dis 5, 626. Viswanathan, P. and Kaur, R. (2001) Prevalence and growth of pathogens on salad vegetables, fruits and sprouts. Int J Hyg Environ Health 203, 205 213. Waje, C.K., Jun, S.Y., Lee, Y.K., Kim, B.N., Han, D.H., Jo, C. and Kwon, J.H. (2009) Microbial quality assessment and pathogen inactivation by electron beam and gamma irradiation of commercial seed sprouts. Food Control 20, 200 204. Weiss, A. and Hammes, W.P. (2003) Thermal seed treatment to improve the food safety status of sprouts. J Appl Bot 77, 152 155. Weiss, A., Hertel, C., Grothe, S., Ha, D. and Hammes, W.P. (2007) Characterization of the cultivable microbiota of sprouts and their potential for application as protective cultures. Syst Appl Microbiol 30, 483 493. Wilkins, P., Bourgeois, R. and Murray, R. (1972) Psychrotrophic properties of Listeria monocytogenes. Can J Microbiol 18, 543 551. Yung, P.T. and Ponce, A. (2008) Fast sterility assessment by germinable-endospore biodosimetry. Appl Environ Microbiol 74, 7669 7674. 36 Letters in Applied Microbiology 56, 30--36 2012 The Society for Applied Microbiology