Enterotoxin Genes, Antibiotic Susceptibility, and Biofilm Formation of Low-Temperature-Tolerant Bacillus cereus Isolated from Green Leaf Lettuce in the Cold Chain

The prevalence and characteristics of low-temperature-tolerant Bacillus cereus (psychrotolerant B. cereus) in green leaf lettuce collected during cold chain were investigated. Among the 101 isolated B. cereus samples, only 18 were capable of growth at 7 °C, and these isolates shared potential health hazard characteristics with mesophilic isolates. Most psychrotolerant B. cereus isolates contained various combinations of nheA, nheB, nheC, hblA, hblA, hblC, hblD, cytK, and entFM. Most isolates of psychrotolerant B. cereus possessed at least two enterotoxin genes and 28% of isolates harbored tested nine enterotoxin genes. Additionally, the psychrotolerant B. cereus isolates showed resistance to tetracycline and rifampin and intermediate levels of resistance to clindamycin. A total of 23% of isolates among psychrotolerant B. cereus displayed a high level of biofilm formation at 7 °C than at 10 °C or 30 °C. The results of this study indicate that cold distribution and storage for green leaf lettuce may fail to maintain food safety due to the presence of enterotoxigenic, antibiotic-resistant, and strong biofilm forming psychrotolerant B. cereus isolates, which therefore poses a potential health risk to the consumer. Our findings provide the first account of the prevalence and characteristics of psychrotolerant B. cereus isolated from green leaf lettuce during cold storage, suggesting a potential hazard of psychrotolerant B. cereus isolates to public health and the food industry.


Introduction
Temperature is one of the crucial factors affecting the safety and quality of vegetables and fruits [1]. To inhibit the growth of foodborne pathogens, vegetables must be maintained at refrigeration temperatures throughout all steps from harvest to consumption. The temperature of perishable foods such as fruits and vegetables must be controlled throughout the supply chain using cold chain systems. However, some bacteria can grow at refrigeration temperatures, and these bacteria (called phychrotolerant bacteria) have been recognized as spoilage bacteria or foodborne pathogens that cause food safety and food quality problems during refrigerated storage and the distribution process of various food products [2,3].
Bacillus cereus is an endospore-forming foodborne pathogen causing diarrheal or emetic food poisoning [4]. B. cereus is ubiquitous in soil, rive, spices, milk, and vegetables and on human skin [4]. In a previous study on the microbiological quality of fresh vegetables, the most frequent foodborne prevalence and characteristics of psychrotolerant B. cereus strains from foodstuff such as milk, dairy products, and chilled food [17,18,24]. Therefore, the goal of the present study was to evaluate the presence and toxigenic characteristics of B. cereus strains, especially psychrotolerant B. cereus strains in green leaf lettuce as an example of leafy vegetables distributed in cold chain. We investigated the presence of genes encoding enterotoxins, antibiotic susceptibility, and biofilm formation ability at refrigeration temperatures. The results from this study will provide an enhanced understanding of the behavior of psychrotolerant B. cereus strains in agricultural products distributed through the cold chain.

Collection of Samples
We collected the green leaf lettuce from every step of the "farm-to-retail" distributions of leafy vegetables. More detailed sampling steps were as follows: Green leaf lettuce was harvested in an open field on a farm located in Gyeongii Province, South Korea. Harvested vegetables were transported to a packing house located within the same farm where they were hand-packed into boxes at a low temperature. Samples for microbiological quality analysis were collected after harvesting (referred to as 'harvest') and packaging ('packaging'). On the same day, the packaged boxes were loaded onto refrigerated trucks for transportation and storage in a distribution center. The next samples were collected after arrival at the distribution center ('transportation') and indicated storage duration at 9-10 • C in the distribution center, prior to shipment to the retail outlet ('distribution center'). The final sampling site ('retail shop') was the point of sale in the retail shop.
The vegetable samples (100 g) from each sampling site were individually placed in sterilized bags from ten different boxes of the same lot code and were maintained at 5 • C during transport to the laboratory for microbiological quality analysis. The temperature of the green leaf lettuce was monitored with a continuous temperature recorder connected to a reader (BTM-4208SD, LT Lutron, Taipei, Taiwan) with a thermocouple. The thermocouple was completely inserted into the center of the vegetables. The ambient temperature ranged from 0 • C to 5 • C and the temperature of the vegetables was maintained at 5 • C ± 2 • C throughout the entire distribution chain (Figure 1). Foods 2020, 9,249 3 of 14 the presence and toxigenic characteristics of B. cereus strains, especially psychrotolerant B. cereus strains in green leaf lettuce as an example of leafy vegetables distributed in cold chain. We investigated the presence of genes encoding enterotoxins, antibiotic susceptibility, and biofilm formation ability at refrigeration temperatures. The results from this study will provide an enhanced understanding of the behavior of psychrotolerant B. cereus strains in agricultural products distributed through the cold chain.

Collection of Samples
We collected the green leaf lettuce from every step of the "farm-to-retail" distributions of leafy vegetables. More detailed sampling steps were as follows: Green leaf lettuce was harvested in an open field on a farm located in Gyeongii Province, South Korea. Harvested vegetables were transported to a packing house located within the same farm where they were hand-packed into boxes at a low temperature. Samples for microbiological quality analysis were collected after harvesting (referred to as 'harvest') and packaging ('packaging'). On the same day, the packaged boxes were loaded onto refrigerated trucks for transportation and storage in a distribution center. The next samples were collected after arrival at the distribution center ('transportation') and indicated storage duration at 9-10 °C in the distribution center, prior to shipment to the retail outlet ('distribution center'). The final sampling site ('retail shop') was the point of sale in the retail shop.
The vegetable samples (100 g) from each sampling site were individually placed in sterilized bags from ten different boxes of the same lot code and were maintained at 5 °C during transport to the laboratory for microbiological quality analysis. The temperature of the green leaf lettuce was monitored with a continuous temperature recorder connected to a reader (BTM-4208SD, LT Lutron, Taipei, Taiwan) with a thermocouple. The thermocouple was completely inserted into the center of the vegetables. The ambient temperature ranged from 0 °C to 5 °C and the temperature of the vegetables was maintained at 5 °C ± 2 °C throughout the entire distribution chain (Figure 1).

Microbiological Analysis for the B. cereus Group
The microbiological procedures for B. cereus group isolation recommended by the Korean food code [25] were followed. To isolate the B. cereus group from green leaf lettuce collected during the supply chain, 25 g of each sample was transferred to a sterile stomacher bag with 225 mL of 0.85% saline and homogenized for 2 min at 230 rpm in a stomacher 400 (Seward, Norfolk, UK). Next, the suspension solution was inoculated onto mannitol-egg-yolk-polymyxin B agar (MYP, Merck, Darmstadt, Germany) and incubated at 30 °C for 18 h. The B. cereus group appears as dry, pinkish colonies surrounded by a ring of dense precipitate due to the degradation of lecithin. Suspected colonies were selected, and a maximum of five typical colonies from each sample were subcultured

Microbiological Analysis for the B. cereus Group
The microbiological procedures for B. cereus group isolation recommended by the Korean food code [25] were followed. To isolate the B. cereus group from green leaf lettuce collected during the supply chain, 25 g of each sample was transferred to a sterile stomacher bag with 225 mL of 0.85% saline and homogenized for 2 min at 230 rpm in a stomacher 400 (Seward, Norfolk, UK). Next, the suspension solution was inoculated onto mannitol-egg-yolk-polymyxin B agar (MYP, Merck, Darmstadt, Germany) and incubated at 30 • C for 18 h. The B. cereus group appears as dry, pinkish colonies surrounded by a ring of dense precipitate due to the degradation of lecithin. Suspected colonies were selected, and a maximum of five typical colonies from each sample were subcultured on tryptic soy agar (TSA, Merck Darmstadt, Germany). Where there were less than five colonies, all were isolated. Biochemical identification of the selected colonies was conducted by using the Vitek-II system with the BCL card (bioM´erieux, Inc., Marcy l'Etoile, France), according to the manufacturer's directions.

Bacterial DNA Extraction
To obtain the genomic DNA, 1 mL of the 18-20 h bacterial culture was centrifuged (15,000× g, 5 min) and the pellet was washed twice with 1 mL phosphate-buffered saline (PBS). The cell pellet was resuspended in 100 µL sterilized distilled water and heated at 95 • C for 10 min and was then placed on ice. After centrifugation at 14,000× g for 2 min, the supernatant was used as the template for molecular identification and the toxin gene profile.

Identification of B. cereus and B. thuringiensis from the B. cereus Group
To isolate B. cereus and B. thuringiensis, Polymerase chain reaction (PCR) amplification of the gyrase B (gyrB) gene and crystal (cry) gene was carried out as described by Yamada et al. [26]. The primers used for the identification of B. cereus were BC1 (5 -ATTGGTGACACCGATCAAACA-3 ) and BC2r (5 -TCATACGTATGGATGTTATTC-3 ). BTJH-1F (GCTTACCAGGGAAATTGGCAG) and BTJH-R (ATCAACGTCGGCGTCGG) were used for the identification of B. thuringiensis from the B. cereus group. The specific primers for the cry gene detection of B. thuringiensis were K3 (5'-GGCTGTGACACGAAGGATATAGCCAC-3') and K5 (5'-AGGACCAGGATTTACAGGAGG-3'). The amplification of gyrB (365 bp) for B. cereus-specific fragments and gyrB (299 bp) and cry for B. thuringiensis (1600 to 1700 bp) was performed using 30 cycles, each consisting of 60 s at 94 • C, 90 s at 58 • C (gyrB for B. cereus and cry gene for B. thurigiensis), 90 s at 63 • C (gyrB for B. thurigiensis), and 90 s at 72 • C, with a final extension step at 72 • C for 7 min. Bacillus mycoides were identified based on the distinct characteristics of filamentous and rhizoid colonies that formed on the agar plate [27].

Growth Properties
To determine the psychrotolerant properties, all isolates of the B. cereus group were screened for their growth ability at 10 • C, 7 • C, and 5 • C. The B. cereus group isolates were first inoculated on tryptic soy agar and incubated at 30 • C for 18 h. One colony of overnight culture was inoculated on a fresh TSA plate and incubated at 10 • C for 12 days, at 7 • C for 20 days, and at 5 • C for 20 days. After incubation, the plates were examined for observable growth.

Detection of Enterotoxin and Emetic Toxin Genes
To assess the putative pathogenic potential of the B. cereus group isolates, all isolates were screened by PCR for the presence/absence of toxin genes known to encode enterotoxins and emetic toxins. The primer pairs for each toxin gene used in this study were based on our previous study [18]. The primer pair sequences and the conditions of amplification are listed in Table 1. B. cereus ATCC 14579 (diarrheagenic) and NCCP 14796 (emetic) strains were used as the control strains.

Quantification of Biofilm Formation
The biofilm formation ability of the B. cereus group isolates was performed according to the modified method by Singh et al. [29] and was investigated under different temperature conditions (30 • C, 10 • C, and 7 • C) using the microtiter plate method. These temperatures were chosen due to their importance in the distribution chain of agricultural products: 30 • C is the optimum temperature for growth of the B. cereus group, 10 • C is the average surface temperature of the vegetables in the distribution chain, and 7 • C is the growth temperature of the psychrotolerant B. cereus group isolates. To evaluate biofilm formation, all B. cereus group isolates were inoculated into a test tube containing 10 mL of tryptic soy broth (TSB, Merck) and were incubated for 18 h at 30 • C. A total of 100 µL of each bacterial suspension was inoculated into 900 µL of fresh tryptic soy broth in three wells of three sterile 48-well flat-bottomed polystyrene plates. Negative control wells containing only TSB were inoculated in triplicate.
For biofilm evaluation, we used biological triplicates. The plates were covered and incubated aerobically for 72 h at 30 • C, for 12 days at 10 • C, and for 20 days at 7 • C. After incubation, the contents of the microtiter plate were poured off, and the wells were washed three times with 1 mL of phosphate buffered saline. The attached bacteria were then fixed with 1 mL of methanol per well for 15 min, after which the plates were emptied and dried at room temperature. Then, the plates were stained with 500 µL of 1% crystal violet per well for 10 min. The stain was removed, and the plates were washed under running tap water. The plates were air dried. In this study, sterile broth and B. cereus 14579 served as the negative control and B. cereus ATCC 10876 was used as the positive control strain for all incubation temperatures.
The optical density (OD) of each well was measured at 550 nm with a microplate reader (Synergy™ Mx, BioTek, Winooski, VT, USA). The OD of each strain was obtained from the mean value of the three respective wells. The strains were classified as follows. The cut-off OD (ODc) for the microtiter plate test was defined as three standard deviations above the mean OD of the negative control. The strains were classified into four categories: no biofilm producer (ODs ≤ ODc), weak biofilm producer (ODc < ODs ≤ 2 × ODc), moderate biofilm producer (2 × ODc < ODs ≤ 4 × ODc), or strong biofilm producer (4 × ODc < ODs).

Statistical Analysis
All assays were performed in triplicate in two independent experiments, and the results were expressed as the average and log-transformed values. SPSS statistical software (IBM, Armonk, NY, USA) was used to evaluate the results by one-way analysis of variance (ANOVA). To compare the means, Duncan's test was used with a significance level of p < 0.05.

Prevalence and Identification of the B. cereus Group
The prevalence and identification of the B. cereus group from green leaf lettuce throughout the cold chain are illustrated in Table 2. The B. cereus group contaminated all samples collected at each sampling point, and the average contamination level was less than 3 log CFU/g. After harvest, the mean B. cereus number was 1.9 log CFU/g, and the level (2.9 log CFU/g) of contamination at the packaging point was 1 log CFU/g higher than that at the harvest point. During cold storage in the distribution center, the contamination level of B. cereus was maintained, but the average count at the transportation point decreased to 2.1 log CFU/g from 2.6 log CFU/g at the distribution center. Few studies have evaluated B. cereus contamination in vegetables, and approximately 50% of the tested vegetable products were contaminated with B. cereus [21,[30][31][32]. B. cereus contamination of vegetables may be common, and the consumption of vegetables contaminated with B. cereus may result in exposure to foodborne illness [32].  (1) Each data point shows the mean ± standard deviation of triplicated samples. (2) Different superscripts in the same column indicate significant differences (p < 0.05).
In fresh-cut vegetable end products, B. cereus levels between 4 and 5 log CFU/g are considered unsatisfactory for consumption, according to the Health Protection Agency guidelines of the United Kingdom for ready-to-eat foods [33]. All samples collected from each point during the cold chain had values less than 3 log CFU/g, but there is a potential possibility of bacterial contamination in raw fresh vegetables and the risk of foodborne illness outbreak by pathogens is greater when compared to cooked foods as raw vegetables are most often consumed without cooking. Conventional PCR was performed with 101 B. cereus group isolates, and the results supported the identification of B. cereus as all isolates were non-rhizoidal cells and did not harbor the cry gene or the specific gyrB gene for B. thuringiensis.
Low temperatures in the cold chain can effectively control the proliferation of foodborne pathogens [34]. However, the use of long-term cold storage may induce changes in the cellular processes, cell membranes, cell surface protein expression, and ribosome structures that promote efficient survival and proliferation under food environment-associated stress conditions [35]. Psychrotolerant B. cereus isolates, in comparison to non-stress-adapted bacteria, can more easily adapt to other stresses such as pH, heat, and anaerobic conditions [17,18]. Thus, monitoring the prevalence of psychrotolerant B. cereus isolates in vegetables throughout the cold chain could be an important food safety factor for predicting the occurrence of outbreaks caused by psychrotolerant B. cereus in fresh vegetables.

Psychrotolerant Properties of B. cereus Isolated from Vegetables
We collected a total of 101 B. cereus isolates from green leaf lettuce samples. All isolates did not grow at 5 • C: 58.4% among the 101 isolates grew at 12 • C, 50.5% at 10 • C, 17.8% at 7 • C, and none at 5 • C (Table 3). Among the 18 isolates that grew at 7 • C, the majority (17 isolates, 98%) of isolates were collected from green leaf lettuce during transportation from a distribution center to a retail shop. More than 50% of the isolates could grow at 12 • C or 10 • C, and approximately 18% of the isolates were capable of growth at 7 • C; therefore, psychrotolerant B. cereus isolates from green leaf lettuce could increase, depending on the storage condition and transport period of the products under low temperatures, considering that refrigeration is the main conservation technology of these products. The psychrotolerant B. cereus isolates may be important in the food industry due to the fact that these isolates are related to the reduction of the shelf life of food products [36], or the increase in the number of microorganisms, which in turn can become a food poisoning problem. The psychrotolerant B. cereus group is detected in soil, milk, ice cream, dairy products, rice salad, and spices [37,38]. Psychrotrophic B. cereus isolates can become a food poisoning problem when these isolates possess toxin genes to produce diarrheic or emetic syndromes. Recent studies have shown that psychrotolerant B. cereus isolates carry virulence toxin genes associated with diarrheal disease (nheACD and hblACD), and psychrotolerant B. cereus isolates have been shown to produce both HBL and NHE toxins [39][40][41]. Further studies are necessary to determine the prevalence of psychrotolerant B. cereus isolates in various agricultural products supplied throughout the cold chain to evaluate virulence factors, antibiotic resistance, and biofilm formation of these isolates as well as to predict the possible occurrence of foodborne illness.

Enterotoxigenic Potential of Psychrotolerant B. cereus Isolates
HBL, NHE, CytK, EntFM, and cereulide (emetic toxin) are major virulence factors produced by B. cereus [42]. To evaluate the food poisoning potential of psychrotolerant B. cereus isolates, we investigated the distribution and toxigenic profile of toxin genes, and the results are shown in Table 4. NHE is produced only when all three NHE enterotoxin (nheA, nheB, and nheC) complexes are present [43]. We found that 98% of the isolates of non-psychrotolerant B. cereus and 94% of the psychrotolerant B. cereus isolates harbored the NHE complex (the combination of nheA, nheB, and nheC). The production of the HBL enterotoxin composed of B, L2, and L1 subunits depends on the expression of all three genes, namely, hblC, hblD, and hblA. Thirty percent of the non-psychrotolerant B. cereus isolates and 44% of the psychrotolerant B. cereus isolates possessed three coding genes of the HBL complex. The cytK gene, which might cause necrotic enteritis, was detected in 96% of the mesophilic and 57% of the psychrotolerant B. cereus isolates. In addition, 88% of the non-psychrotolerant B. cereus isolates and 78% of the psychrotolerant B. cereus isolates possessed the entFM gene. None of the isolates possessed the ces gene. A total of 23% of the non-psychrotolerant B. cereus isolates and 28% of the psychrotolerant B. cereus isolates possessed all tested enterotoxin genes. We observed that 18 isolates capable of growth at 7 • C among the 101 B. cereus isolates may have the potential to proliferate in food in response to changes in the optimal environmental conditions of these bacteria or in long-term cold storage. Bartoszewicz et al. [37] reported that psychrotolerant B. cereus harbors well-known toxigenic-associated genes such as nheA, hblA, and cytK. Most psychrotolerant B. cereus isolates have been reported to possess the enterotoxin gene and show high hemolytic activity [40,44]. Our data are consistent with those of previous reports showing that psychrotolerant B. cereus can grow at refrigeration temperatures and possesses at least two enterotoxin genes. These results suggest that psychrotolerant B. cereus isolates from green leaf lettuce may pose a health risk if the enterotoxin genes are expressed during prolonged cold storage.

Antibiotic Resistance
The resistance patterns to 12 antibiotics in 83 non-psychrotolerant B. cereus isolates and 18 psychrotolerant B. cereus isolates were determined, and the results are shown in Table 5. No difference in the frequency of antibiotic resistance was observed between non-psychrotolerant B. cereus and psychrotolerant B. cereus isolates. All isolates revealed 100% resistance to penicillin and third-generation cephalosporins and were sensitive to gentamicin, imipenem, ciprofloxacin, erythromycin, and chloramphenicol. This result is consistent with that of a previous report demonstrating that B. cereus is susceptible to imipenem, vancomycin, chloramphenicol, ciprofloxacin, erythromycin, and gentamicin [23,45]. In our study, all non-psychrotolerant B. cereus and psychrotolerant B. cereus isolates showed resistance to beta-lactam antibiotics. B. cereus generally produces beta-lactamase and is uniformly resistant to beta-lactam antibiotics including third-generation cephalosporins. B. cereus isolated from ice cream showed sporadic resistance to tetracycline and erythromycin [22]. Table 5. Antibiotic resistance patterns of the non-psychrotolerant and psychrotolerant B. cereus isolated from green leaf lettuce in the cold chain. In this study, 16% of the non-psychrotolerant B. cereus isolates showed resistance to tetracycline, and 60% of the isolates revealed an intermediate level of resistance to this antibiotic. These findings are consistent with those of other researchers who reported that 7% of the isolates from cereal, ice cream, and soil were resistant to tetracycline [22]. A high number (60%) of psychrotolerant B. cereus isolates showed an intermediate level of resistance to tetracycline. Intermediate level resistance (low-level resistance) is often underestimated. It is not only a gateway to high-level clinical resistance, but is also a gateway to often unsuspected phenomena such as increased bacterial adaptation to antibiotics [46]. Antibiotic resistance in foodborne pathogens has become a serious problem in many parts of the world, emphasizing the need to better understand the mechanisms involved in the emergence and spread of resistant strains [46].

No. (%) of B. cereus Isolates with Indicated
Rifampin is a bactericidal antibiotic effective against a wide variety of bacteria, but B. cereus is frequently resistant to rifampin [47]. In this study, 78% of non-psychrotolerant B. cereus and 22% of psychrotolerant B. cereus isolates were resistant to rifampin. B. cereus isolates from agricultural products showed a high frequency of resistance to rifampicin [21], and more than 60% of B. cereus isolates from cereal and rice have been shown to be resistant to rifampicin [48]. In contrast, isolates from dairy products and ready-to-eat products show sensitivity to rifampin [49,50]. Only one psychrotolerant B. cereus isolate showed resistance to vancomycin. All non-psychrotolerant B. cereus isolates were sensitive to clindamycin, whereas 45% of the psychrotolerant B. cereus isolates displayed intermediate resistance to clindamycin.
Antibiotic therapy is still the primary treatment method for infections of B. cereus. However, antibiotic-resistant B. cereus can easily spread at each stage of the food chain, and may reach humans indirectly through the consumption of contaminated food or food derived products [51]. Under suitable conditions, some bacteria can either induce the onset of a disease or transfer the gene that provides antibiotic resistance to other bacterial pathogens [52]. Furthermore, when bacteria are exposed to environmental stress, they may undergo phenotypic and genotypic changes to enhance their survival in a stressful environment [53]. These changes may give rise to cross-protection when these bacteria are exposed to secondary stresses, which may also subsequently change their antibiotic resistance profiles [54]. McMahon et al. found that when Escherichia coli, Salmonella enterica serovar Typhimurium, and S. aureus were exposed to sublethal food preservation stresses, their antibiotic sensitivity profiles changed [55]. The difference in the antibiotic susceptibility profile between non-psychrotolerant B. cereus and psychrotolerant B. cereus isolates may be due to changes in the phenotype and genotypic properties of -low-temperature-tolerant bacterial isolates.

Biofilm Formation
A biofilm is a multicellular complex formed by microorganisms that are attached to a surface and are embedded in a matrix consisting of exopolymeric substances (EPS). Cells within a biofilm are surrounded by the EPS matrix and cells in the outer layers of the biofilm, protecting them from harsh influences from the environment. Therefore, cells surrounded by biofilms can be resistant to antimicrobial substances or cleaning agents [56]. The ability of non-psychrotolerant B. cereus and psychrotolerant B. cereus isolates to produce biofilms on a microtiter plate at 7 • C, 10 • C, and 30 • C is shown in Figure 2. At 30 • C, five (6.1%) of the non-psychrotolerant B. cereus isolates had moderate biofilm formation abilities, and 19.5% of the isolates had weak biofilm formation abilities. At 10 • C, four isolates (approximately 5%) were moderate biofilm formers, and 25 isolates (30.5%) were weak biofilm formers. Ten percent of non-psychrotolerant B. cereus isolates had a weak biofilm-formation ability, but most of the isolates (90%) did not produce a biofilm at 7 • C. No isolate was a strong biofilm former among the entire non-psychrotolerant B. cereus isolates.

Biofilm Formation
A biofilm is a multicellular complex formed by microorganisms that are attached to a surface and are embedded in a matrix consisting of exopolymeric substances (EPS). Cells within a biofilm are surrounded by the EPS matrix and cells in the outer layers of the biofilm, protecting them from harsh influences from the environment. Therefore, cells surrounded by biofilms can be resistant to antimicrobial substances or cleaning agents [56]. The ability of non-psychrotolerant B. cereus and psychrotolerant B. cereus isolates to produce biofilms on a microtiter plate at 7 °C, 10 °C, and 30 °C is shown in Figure 2. At 30 °C, five (6.1%) of the non-psychrotolerant B. cereus isolates had moderate biofilm formation abilities, and 19.5% of the isolates had weak biofilm formation abilities. At 10 °C, four isolates (approximately 5%) were moderate biofilm formers, and 25 isolates (30.5%) were weak biofilm formers. Ten percent of non-psychrotolerant B. cereus isolates had a weak biofilm-formation ability, but most of the isolates (90%) did not produce a biofilm at 7 °C. No isolate was a strong biofilm former among the entire non-psychrotolerant B. cereus isolates. In contrast, psychrotolerant B. cereus isolates showed a higher biofilm formation ability at 7 °C than at 10 °C or 30 °C. Among psychrotolerant B. cereus isolates, all were non-or weak biofilm formers at 30 °C, and 38.5% of the isolates had moderate biofilm-forming ability at 10 °C. Interestingly, of the psychrotolerant B. cereus isolates, at 7 °C, 23% were strong and moderate biofilm formers, and 15% were weak biofilm formers. Only three psychrotolerant B. cereus isolates (16.6%) were strong biofilm formers at 7 °C, in contrast to the psychrotolerant B. cereus isolates with no strong biofilm former.
Cells in biofilm are more resistant to disinfectant treatment and stress conditions, and the biofilm formation of some psychrotolerant B. cereus isolates under low temperature can be difficult to disinfect foodstuff or food contact surfaces and the possibility of cross-contamination during food processing [57,58]. In this study, psychrotolerant B. cereus isolates had greater biofilm-formation abilities than non-psychrotolerant B. cereus isolates at low temperatures. This result is in agreement with that of Sarita et al. [57], who showed that B. cereus isolates from a chiller tank containing pasteurized milk were able to form a biofilm even at 4 °C. In contrast, B. cereus isolated from food products could grow at low temperatures, but the isolates were unable to form a biofilm at 7 °C [28]. The growth of most B. cereus isolates is generally inhibited when temperature is below 10 °C and forms strong biofilms on a microtiter plate at 30 °C. In this study, however, some psychrotolerant B. cereus isolates could grow at refrigerated temperature and these isolates in biofilm were more resistant to low temperatures than optimal temperature, especially for the transport, retail, and processing of agricultural products such as vegetables, which are typically stored between 0-10 °C, where psychrotolerant B. cereus adapted to cold stress and protected themselves against the cold shift. In contrast, psychrotolerant B. cereus isolates showed a higher biofilm formation ability at 7 • C than at 10 • C or 30 • C. Among psychrotolerant B. cereus isolates, all were non-or weak biofilm formers at 30 • C, and 38.5% of the isolates had moderate biofilm-forming ability at 10 • C. Interestingly, of the psychrotolerant B. cereus isolates, at 7 • C, 23% were strong and moderate biofilm formers, and 15% were weak biofilm formers. Only three psychrotolerant B. cereus isolates (16.6%) were strong biofilm formers at 7 • C, in contrast to the psychrotolerant B. cereus isolates with no strong biofilm former.
Cells in biofilm are more resistant to disinfectant treatment and stress conditions, and the biofilm formation of some psychrotolerant B. cereus isolates under low temperature can be difficult to disinfect foodstuff or food contact surfaces and the possibility of cross-contamination during food processing [57,58]. In this study, psychrotolerant B. cereus isolates had greater biofilm-formation abilities than non-psychrotolerant B. cereus isolates at low temperatures. This result is in agreement with that of Sarita et al. [57], who showed that B. cereus isolates from a chiller tank containing pasteurized milk were able to form a biofilm even at 4 • C. In contrast, B. cereus isolated from food products could grow at low temperatures, but the isolates were unable to form a biofilm at 7 • C [28]. The growth of most B. cereus isolates is generally inhibited when temperature is below 10 • C and forms strong biofilms on a microtiter plate at 30 • C. In this study, however, some psychrotolerant B. cereus isolates could grow at refrigerated temperature and these isolates in biofilm were more resistant to low temperatures than optimal temperature, especially for the transport, retail, and processing of agricultural products such as vegetables, which are typically stored between 0-10 • C, where psychrotolerant B. cereus adapted to cold stress and protected themselves against the cold shift. The results suggest that low temperature may fail to restrain the growth and biofilm formation of some B. cereus isolates in vegetables, therefore presents a potential risk in food safety.
Biofilm formation is enhanced by cell motility, particularly when biofilm is mediated by flagella, and under certain environmental conditions, flagella are necessary for B. cereus biofilm formation [59]. The integration of many diverse signals from the environment, together with other events such as phenotypic and genetic switching during biofilm production and the release of extracellular polymeric substances, might play a role in biofilm formation [60]. Thus, the biofilm-formation ability of psychrotolerant isolates could be due to modified bacterial characteristics (flagella, EPS, growth phase, metabolic activity, etc.) that are temperature dependent. However, further study is necessary to understand the mechanism of biofilm formation by psychrotolerant B. cereus isolates.

Conclusions
B. cereus is the predominant species isolated from agricultural products including green leaf lettuce. Low-temperature-tolerant B. cereus isolates are associated with potential health risks as these isolates share the enterotoxin gene with mesophilic isolates. Psychrotolerant B. cereus isolates can also form strong biofilms, which hampers food industry surface decontamination and increases the possibility of cross-contamination during food processing. The prevalence and pathogenicity of psychrotolerant B. cereus during the cold chain cannot be ignored from the aspect of food safety.