Genotypes and the persistence survival phenotypes of Bacillus cereus isolated from UHT milk processing lines

Food Control 82 (2017) 48e56

Contents lists available at ScienceDirect

Food Control journal homepage: www.elsevier.com/locate/foodcont

Genotypes and the persistence survival phenotypes of Bacillus cereus isolated from UHT milk processing lines Yingying Lin a, b, Fazheng Ren a, c, Liang Zhao a, d, Huiyuan Guo a, b, * a Beijing Advanced Innovation Center for Food Nutrition and Human Health, College of Food Science & Nutritional Engineering, China Agricultural University, Beijing 100083, China b Key Laboratory of Functional Dairy, China Agricultural University, Beijing 100083, China c Beijing Laboratory of Food Quality and Safety, China Agricultural University, Beijing 100083, China d Hebei Engineering Research Center of Animal Product, Sanhe 065200, China

a r t i c l e i n f o

a b s t r a c t :

Article history: Received 19 April 2017 Received in revised form 15 June 2017 Accepted 16 June 2017 Available online 19 June 2017

Bacillus cereus is an important foodborne pathogen that can cause emesis and diarrhea. The aim of this study was to find the phenotypic properties of Bacillus cereus enabling its persistence in dairy plants, and identify the corresponding genotypic features. The toxin gene profile, multilocus sequence typing (MLST), biofilm-forming capability and the alkali and acid tolerance of the biofilms of the strains were analyzed. All strains were positive for one or more toxin genes tested, except for A20. The three main toxin genes were nheABC (74.07%), bceT (73.37%), and ces (48.15%), followed by entFM (40.74%) and cytK (33.33%), but no strain harbored hblABD. A total of 17 ST-types were generated among 27 isolates by MLST, and clustered into four groups. Although no significant difference was found in the tolerance of biofilm to acid and alkali among four MLST groups, the strains of MLST groupⅠhad stronger biofilm formation capability. The biofilm was formed on stainless steel coupons, of which the OD values of A1, B16 and B18 were more than 1.0, and close to each other in the phylogenetic tree. Persistence survival strategies analysis showed that strains can be divided into three groups: 15 strains without persistence strategies; two strains were capable of forming biofilm may be effectively inactivated by hot-acid or hotalkali; ten isolates were capable of forming biofilm and possessing the tolerance of biofilm to acid and alkali, which should be under strict supervision and control in the future. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Bacillus cereus Multilocus sequence typing Biofilm Alkali tolerance Toxin gene

1. Introduction Bacillus cereus is a Gram-positive, spore-forming bacterium and ubiquitous in environment. As its heat stable characteristic, Bacillus cereus is commonly contaminating raw milk and dairy products, but an underestimated pathogen in the dairy industry. It can cause two types of gastrointestinal disease, emesis and diarrhea (Bremer, Fillery, & McQuillan, 2006; Ceuppens, Boon, & Uyttendaele, 2013; Kim et al., 2009), and its growth may result in various dairy defects. Several studies performed in Ethiopia, Irish, Tunisa, and Turkey found that the prevalence of Bacillus cereus in raw milk reached 12.86%, 23%, 47.5%, and 90.0%, respectively (Aouadhi, Maaroufi, & Mejri, 2014; Gundogan & Avci, 2014; Garedew,

* Corresponding author. College of Food Science and Nutritional Engineering, China Agricultural University, P.O. Box 287, No. 17 Qinghua East Road, Haidian, Beijing 100083, China. E-mail address: [email protected] (H. Guo). http://dx.doi.org/10.1016/j.foodcont.2017.06.025 0956-7135/© 2017 Elsevier Ltd. All rights reserved.

Mengesha, Birhanu, & Mohammed, 2015; O'Connell et al., 2016). The occurrance of the pathogen in dairy products in different studies in USA, Turkey and Brazil reached 17.64%, 20% and 24.23%, respectively (Gundogan & Avci, 2014; Montanhini & Bersot, 2013; Reis, Montanhini, Bittencourt, Destro, & Bersot, 2013). A study conducted in 10 local dairy farms in Beijing showed that the total occurrance of Bacillus cereus in raw milks was as high as 9.8%, and the nhe, hbl, and ces genes were detected at the rate of 100%, 55.0%, and 5.0%, respectively(Cui et al., 2016). In pasteurized milk, the contamination rate of Bacillus cereus was 33.3e71.4% in China (Zhou, Liu, He, Yuan, & Yuan, 2008). These previous data provided information that occurrence of Bacillus cereus in raw milk and final products was high, especially in China. This is a matter of concern as it can lead to foodborne outbreaks. However, the genotype diversity of Bacillus cereus isolates from dairy plants in China remains unclear, especially the ultra-high temperature sterilization (UHT) milk. In modern dairy plant environment, Bacterium can negatively

Y. Lin et al. / Food Control 82 (2017) 48e56

survive and colonise. Indeed, the endospores and biofilm formation properties of Bacillus cereus are considered as its survival strategies (Ehling-Schulz, Fricker, & Scherer, 2004a), which is the main cause of Bacillus cereus contamination in dairy industry (Flach, ~o, 2014). The resistance towards Grzybowski, Toniazzo, & Corça heat and disinfectants of Bacillus cereus endospores allow them survive cleaning procedures, and then they attach to equipment surface depending on their hydrophobic properties (Ryu & Beuchat, 2005). Adherence to stainless steel surfaces of dairy plants can result in biofilm formation, which is more resistant to antimicrobials and cleaning regimes compared to planktonic cells. This makes elimination of Bacillus cereus from dairy industry a big challenge, and induces recurrent contamination of dairy products (Kumari & Sarkar, 2014; Shaheen, Svensson, Andersson, Christiansson, & Salkinoja-Salonen, 2010). In recent year, the structure and the mechanism of resistance exhibited by Bacillus spores are investigated (Soni, Oey, Silcock, & Bremer, 2016), and a variety of sanitizers and biocide are evaluated for use in CIP systems to promote its biofilm removal (Lemos, Gomes, Mergulhao, Melo, & Simoes, 2015). Besides, biofilm-producing ability of Staphylococcus aureus and Listeria monocytogenes isolated from diary plants were investigated (In Lee et al., 2017; Lee et al., 2014). Thus hygiene practices must be improved to prevent biofilm formation, and appropriate treatment should be taken for removing biofilms (Lee, Cappato, Corassin, Cruz, & Oliveira, 2016).However, to develop novel control strategy of Bacillus cereus in dairy processing environment, there is a strong need to gain a better insight into the phenotypic properties enabling its persistence under the dairy conditions, and identify the corresponding genotypic features. Therefore, the aim of this study was to isolate the persistent Bacillus cereus from the UHT milk processing lines, and reveal the connections of the genotypes and survival phenotypes of the strains. First, the intraspecific genotype diversity of 27 Bacillus cereus strains isolated from UHT milk processing lines were characterized by Multilocus sequence typing (MLST)-based phylogenetic analysis, and the potential hazard based on toxin genes traits of the strains were evaluated. Then, the abilities of the Bacillus cereus strains to form biofilms were investigated, and the threedimensional structure of the biofilms were observed by confocal laser scanning microscope. Third, we exposed the Bacillus strains biofilms to highly alkaline and acid liquids at high temperature applied during the cleaning-in-place (CIP) procedures, to assess the alkali and acid tolerance of the biofilms These data is of importance for further understanding the persistent mechanism of Bacillus cereus and might then be used for providing effective strategies to minimize it. 2. Materials and methods 2.1. Sample collection and Bacillus cereus isolation In this study, two local dairy plants in China were under surveillance, and investigated at the critical control points throughout the production chain every production day, from November 2014 to May 2015. The isolates were all obtained in UHT milk processing lines, including silo tanks, ingredients tanks, pasteurization tanks, post-pasteurization (air, homogenizer), filling room air and UHT milk products. Bacillus cereus was isolated according to the standard procedures described in the National Standards of the People's Republic of China (GB.4798.14-2014). Briefly, for each serial dilution (101 to 104), 100 mL were plated on MYP agar (Qingdao Hope Bio-Technology Co., Ltd, China), a selective medium for isolating members of the Bacillus cereus group, and incubated at 37
C for 48 h. Those bacteria forming rough and dry

49

colonies with a violet pink background surrounded by egg yolk precipitation on the MYP agar and with parasporal crystals observed under phase-contrast microscopy, were identified as presumptive Bacillus cereus. These presumptive colonies were transferred into 2 mL brain heart infusion (BHI, AOBOX Technology) broth and incubated at 37
C for 16 h. Then, the bacterial cultures were spread on BHI agar, and incubated at 37
C for another 24 h Single colony was chosen for further study, and all isolates were stored at 80
C until use. 2.2. DNA extraction and identification of Bacillus cereus-like strains Genomic DNA was extracted from overnight cultures grown in LuriaeBertani (LB) broth using the TIANamp Bacteria DNA Kit (Tiangen Bio-Tech Co., Ltd, China) in accordance with the manufacturer's instructions. 16S rRNA sequence analysis was used to further characterize the Bacillus cereus-like strains, using primers 27F and 1492R (Lane, 1991). The amplified sequences of each strain were analyzed by nucleotide blast (http://blast.ncbi.nlm.nih.gov/ Blast.cgi). The strains with high similarity to 16S rRNA sequence of the Bacillus cereus reference strain (Evalue ¼ 0 and Max identity  98%) were regarded as Bacillus cereus strains (Zhou, Zheng, Dou, Cai, & Yuan, 2010). 2.3. Detection of toxin genes For detection of Bacillus cereus ten toxin genes (hblA, hblB, hblC, nheA, nheC, nheD, cytK, bceT, emtF, ces), the primer pairs and PCR reaction conditions were in accordance with a previous report (Seong, Lim, Lee, Lee, & Hong, 2008). Bacillus cereus ATCC 14579 provided by CGMCC was used as control strain. 2.4. MLST-based phylogenetic analysis All the isolates were characterized by the MLST scheme using primers and conditions reported in the primers section of the Bacillus cereus MLST database (http://pubmlst.org/bcereus/). Portions of glpF, gmk, ilvD, pta, pur, pycA, and tpi housekeeping genes were amplified. The PCR programs were performed by using the given conditions and cycle. Amplification products were purified using the QiaAmp PCR purification kit (Qiagen, Germany). The sequences of the seven housekeeping genes of each stain were assigned allele numbers based on the locus queries at http://pubmlst.org/bcereus/, and sequence types (ST) were numbered based on the combination of seven alleles. To assess the relationship of the Bacillus cereus isolates under this study, a phylogenetic tree was generated from their allelic profiles using the Unweighted Pair Group Method with Arithmetic mean (UPGMA) algorithm with the aid of MEGA7.0 software. Genetic distances, based on the nucleotide polymorphism in housekeeping alleles, were calculated using the UPGMA method and Kimura 2-parameter mathematical model. The correctness of the results was evaluated using a 1000-step bootstrap test. 2.5. Assay of biofilms formation on stainless steel coupons Biofilms were grown on stainless steel coupons of AISI 304 with 2B finished (1  1 cm2) vertically placed in 24-well polystyrene plates which were half filled with 2 ml broth and inoculated with 1.5% pre-culture, following the method described by Hayrapetyan (Hayrapetyan, Abee, & Nierop Groot, 2016). The plates were incubated at 30
C for 48 h. Biofilm on the coupons were quantified using the crystal violet (CV) assay for total biofilm biomass formed as described in Hayrapetyan (Hayrapetyan, Muller, Tempelaars,

50

Y. Lin et al. / Food Control 82 (2017) 48e56

Abee, & Nierop Groot, 2015). Coupons with biofilms were gently washed three times with distilled water to remove non-biofilm cells, and left in 0.1% w.v1 of CV for 30 min to stain. After staining, the coupons were washed three times with distilled water and subsequently de-stained in 2 ml 75% ethanol for 30 min. Intensity of the stain was measured by taking optical density (OD) readings at 595 nm (Tecan M200pro Hydrolex, Switzerland). To correct background staining, the mean OD value obtained for control (without biofilm) was subtracted from the OD value obtained from each condition. Biofilm formation assay was carried out in triplicate for all the 27 Bacillus cereus isolates with the reference strains ATCC 10987 and ATCC 14579. Observations were also performed using confocal laser scanning microscopy (CLSM) following the methods described by Faille (Faille et al., 2014) with a few modification. Biofilms were stained with the propidium iodide (PI) for 15 min, and subsequently analysed on an inverse confocal laser scanning microscope LSM780 with a 40 air objective (Carl Zeiss AG, Jena, Germany) at 488 nm excitation by an argon laser. An area of ca. 220 mm (xaxis)  220 mm (y-axis) was screened in 0.92 mm intervals in the zaxis (z-stack) in the green (emission 488 nm) and red (emission 546 nm) channels, respectively. 2.6. Alkali and acid tolerance of Bacillus cereus biofilm The tolerance of biofilm towards solutions of aqueous hot alkali and acid was investigated as described (Shaheen et al., 2010) with a few modification. The coupons with biofilm were prepared as described, and washed by dipping 3 times in phosphatebuffered-saline (PBS). For measurement of hot-alkaline tolerance, 10 ml of 2.0% (w/v) aqueous NaOH (pH 13.1) in the tube was heated at 80
C in a water bath. When the temperature in a parallel tube containing 10 ml of water reached 80
C, coupons were placed in the tube and heated for 15 min, then sample tubes were removed and immediately cooled in an ice water bath. After cooling, coupons were placed in 50 ml tubes filled with 3 ml PBS and 0.5 g glass beads (D ¼ 100 mm, SIGMA). Tubes were vortexed at maximum speed for 1 min to detach the cells from the coupons. Serial dilutions were made and spread plated on BHI-agar plates and colony forming units (CFU) were counted after 24 h incubation at 30
C. Resistance of the biofilm to hot acid was measured similarly, except that 1.5 %w/v aqueous HNO3 (pH 0.8) was used instead of 2.0% NaOH. Log10 CFU ml1 was plotted vs time, and standard linear regression was performed using the Excel (Microsoft Inc., 2013). The D-value was determined by taking the negative reciprocal of the slope reported for each regression line. 3. Results 3.1. Isolation and distribution of Bacillus cereus 43 positive strains were identified as Bacillus cereus-like stains. 27 strains were subsequently identified as Bacillus cereus based on the detection of 16S rRNA sequence analysis. In these stains, 5 (18.52%) strains were isolated from silo tanks, 8 (29.63%) from ingredients tanks, 7 (25.93%) from pasteurization tanks, 3 (11.11%) from post-pasteurization processing step, 1 (3.70%) from filling room air and 3 (11.11%) from the final UHT milk products. The distribution of Bacillus cereus is uneven in the UHT milk processing lines, but there is a decrease trend in occurrance of Bacillus cereus from raw milk and ingredient to final products. Up to 25.9% of the isolated strains were from the pasteurization tanks, indicating Bacillus cereus could not be inactivated by pasteurization. After UHT treatment, though few, Bacillus cereus can still be

detected in the products. 3.2. Distributions of toxin genes in Bacillus cereus isolates Foodborne illness caused by Bacillus cereus is related to the production of diarrheal toxin and emetic toxin. The distributions of the diarrheal enterotoxin genes and emetic toxin cereulide genes in 27 Bacillus cereus strains isolates from UHT milk processing lines are shown in Table 1. HBL consists of a binding component (B) and two lytic components (L1 and L2), which are encoded by the hblA, hblD, and hblC genes, respectively. NHE is a complex pore-forming toxin consisting of three proteins, NheA, NheB, and NheC, encoded by one operon containing three genes, nheA, nheB, and nheC, respectively. HBL and NHE show toxicity only when all three genes are present and expressed. Strains were positive for hblA and hblC gene with the frequencies of 55.6% and 77.8%, respectively. By contrast, hblD gene was not found in any isolates, showing the lowest percentage out of all enterotoxins. 74.1%, 88.9% and 100% strains possess the nheA, nheB and nheC genes, respectively, and the occurrance of nheABC in isolated strains was 74.1%. 73.4% of the strains harbored bceT gene, and the percentage of Bacillus cereus strains carrying the cytK and entFM gene was low (only 33.3% and 40.7%). 48.2% of the Bacillus cereus strains carried the ces gene, which was related to vomit. Besides, Bacillus cereus in the three final products all contained the ces gene, which will be a potential food risk factor. 3.3. MLST-based phylogenetic analysis The genetic relationship among the Bacillus cereus strains was analyzed using MLST. Fig. 1 shows the dendrogram constructed using UPGMA with allelic profiles of ST. Basing on sequence alignment, these strains clustered into four main groups, designated Groups I, II, III, and IV (Fig. 1). The dendrogram shows that the population of isolates evolved initially by separating in two different branches: one corresponding to group IV and one corresponding to the rest of the population (Fig. 1). 27 total isolates belonged to 18 different ST subgroups, and 11 of the ST (48%) were identified only once. It indicated that the MLST profiles of Bacillus cereus isolates from UHT milk processing lines in China were highly diverse. The most common ST subgroup was ST857 (4/27, 14.8%), which was detected in pasteurization tanks (two samples), silo tank (one sample) and ingredient tank (one sample). Another two ST subgroups (ST 378 and ST427) each contained three isolates, of which three ST427 strains were all isolated from the final milk products. Two ST1140 strains were from silo tanks, and two ST144 and two ST760 strains all from ingredients tanks. On the one hand, the strains from the same processing step may have identical ST. Two strains (A4, A5) with the same ST were both isolated from ingredient tanks. On the other hand, the same ST could be detected in strains with different sources, such as A4, A16 and A20 (ST-857), which were isolated from the pasteurization tank, ingredients tank, and silo tank, respectively. Besides, no identical ST subgroups were both found in the two dairy plants, indicating that stains from different dairy plant differ from each other in the phylogenetic tree. The three isolates found in the final product belonged to ST427 and were different from the other strains. It is interesting that these three strains were closely related to the strain B21 in the phylogenetic tree, with 99% similarity (Fig. 1) 3.4. Biofilm formation by Bacillus cereus The data in Fig. 2 showed biofilm forming capacity within the tested Bacillus cereus strains grown on stainless steel coupons, and

Y. Lin et al. / Food Control 82 (2017) 48e56

51

Table 1 Bacillus cereus food isolates used in this study and PCR analysis of Bacillus cereus toxin genes. B. cereus stain

Sources

HBL complex hblA

a

b

NHE complex

Other enerotoxin

Cereulide gene

hblC

hblD

nheA

nheB

nheC

entFM

bceT

cytK

ces

ATCC 14579 A3c A8 A12 A20 B2

CGMCC Silo tanks (n ¼ 5)

þ þ e e þ e

þ þ þ þ e e

þ e e e e e

þ þ e e e e

þ þ þ þ e e

þ þ þ þ þ þ

þ e e þ e e

þ e þ þ e e

þ e þ þ e e

e e e þ e þ

A6 A11 A15 A16 B15 B16 B18 B24

Ingredients tanks (n ¼ 8)

e e þ þ þ e e þ

þ þ þ þ þ e e þ

e e e e e e e e

þ þ þ þ þ þ þ þ

þ þ þ þ þ e þ þ

þ þ þ þ þ þ þ þ

e e e þ þ e þ þ

þ þ þ þ e e þ e

e þ þ e e e e e

e þ e þ e þ þ þ

þ þ e e þ þ þ

e þ þ þ þ þ þ

e e e e e e e

þ þ e e e þ þ

þ þ þ þ þ þ þ

þ þ þ þ þ þ þ

e e e þ þ e þ

e þ þ þ þ þ e

e e e e e þ þ

e e e e e e þ

A2 A4 A5 A18 A19 B21 B23

Pasteurization tanks(n ¼ 7)

A7 A14 A17

Post-pasteurization(n ¼ 3)

e e þ

þ þ e

e e e

þ þ þ

þ þ þ

þ þ þ

e þ þ

þ þ þ

e e þ

e þ þ

A1

Filling room air (n ¼ 1)

þ

þ

e

þ

þ

þ

e

þ

e

e

UHT final milk products(n ¼ 3)

þ þ e

þ þ þ

e e e

þ þ þ

þ þ þ

þ þ þ

e e þ

þ þ þ

þ þ e

þ þ þ

55.56%

77.78%

0

74.07%

88.89%

100%

40.74%

73.37%

33.33%

48.15%

C1 C2 C3 Rate (%) a

ATCC14579 is the positive strain, it's from CGMCC (China General Microbiological Culture Collection Center). Result(±)indicate positive and negative signals. In front of isolate number, A, B or C was added. “A” indicated the strain was collected in A dairy plant, “B” indicated the strain was collected in B dairy plant, “C” indicated the strain was collected in UHT milk products. b c

were listed based on the MLST group. We classified the formation of a biofilm by the Bacillus cereus isolates as strong (OD595 > 1.0), intermediate (0.5 < OD595 < 1.0), and weak (OD595 < 0.5). First, the ability of the reference strains Bacillus cereus ATCC14579 and Bacillus cereus 10987 to form biofilms on stainless steel coupons was tested. We confirmed that ATCC 10987 formed strong biofilms while ATCC14579 formed weak biofilms on the stainless steel surfaces, which is in agreement with previously published reports. In 27 Bacillus cereus isolates, OD595 ranged from 0.27 to 2.58, with a large diversity. Three strains showed weak biofilm formation capacity with an OD595 of 0.263e0.472, and 15 strains showed intermediate capacity with an OD of 0.554e0.985. The remaining strains showed strong capacity with an OD of 1.07e2.58, mainly isolated from silo tanks and ingredient tanks. The results demonstrated that the biofilm formation capacity varied from strain-tostrain with no obvious connection to MLST group. However, it is interesting that 27% strains of GroupⅠshowed relatively strong biofilm capability. The OD values of A1, B16 and B18 were more than 1.0, and these three strains were very close to each other in the phylogenetic tree. B16 and B18 belonged to the same ST, and A1 showed 99% similarity to them. Capacity of the Bacillus cereus strains to form biofilm was also measured in polystyrene 96-wellplate in BHI, and the results (Fig. S1) were similar to the biofilmformation capacity on stainless steel. As shown in Fig. 3, Bacillus cereus biofilms were not composed of homogeneous monolayers of microbial cells, and the significant differences in three-dimensional structure were found among strains. First, the three-dimensional structures of the reference strains Bacillus cereus ATCC10987 and Bacillus cereus 14579 were

tested. As clearly shown on the cross-sections (Fig. 3A-b and 3A-g) and on the biofilm profile (Fig. 3A-d), ATCC 10987 formed a net-like pattern biofilm, which was continuous and thick on the biofilm profile. Differently, ATCC14579 formed a flat cluster-like pattern biofilm. The results demonstrate that the difference among biofilm structures can be evidenced CLSM. The biofilms of isolated Bacillus cereus A1 and A20 (Fig. 3C and D) were organized in a continuous net-like pattern. In contrast, a patchy biofilm distribution was observed for Bacillus cereus C1 and B23 (Fig. 3E and F). These two Bacillus cereus strains were mostly composed of small flat clusters and of single cells, and the biofilms were not continuous. The threedimensional structures of Bacillus cereus were in agreement with results of the OD value. 3.5. Alkali and acid tolerance testing The results for the resistance of the 27 Bacillus cereus biofilm to hot-alkali and hot-acid are shown as D-values in Table 2. Dvalue was measured to model the duration of the hot washing of CIP procedure in the dairy plants, ranged from 1.79 min to 9.52 min in this study. The strains were listed based on the MLST group. The results demonstrated that the D-values varied from strain-to-strain with no obvious connection to MLST group. As summarized in Table 3, biofilms of 11 strains were effectively destroyed by hot alkali and acid washing, with low D-values. Interestingly, biofilms of four strains were moderately resistant to hot alkali, with D-values of 5.14 (A5), 7.83 (A11), 4.60 (A12) and 5.44 (C3) min, but they were almost effectively killed by hot acid washing. Moreover, two strains (A15 and A17) with low alkali D-

52

Y. Lin et al. / Food Control 82 (2017) 48e56

Fig. 1. UPGMA-based dendrogram and the allelic profiles of the 27 strains used in this study. The phylogenetic tree was generated in MEGA v7.0 with the UPGMA (Unweighted Pair Group Method with Arithmetic mean) method on categorical numeric data based on the internal fragments of seven housekeeping genes. Cophenetic correlation are shown next to the branches. These strains clustered into four main groups, designated Groups I, II, III, and IV. *: new ST.

values, were resistant to acid. Ten of the 27 isolates were highly or moderately alkali and acid resistant, which could survive the treatment of hot-alkaline and acid solution, posing a threat to the UHT milk production.

4. Discussion The results presented in this study reported the toxin gene profile, spore resistance and biofilm-forming capacity of Bacillus

Fig. 2. Biofilm formation of Bacillus cereus food isolates and reference strains ATCC14579 and ATCC 10987 on stainless steel coupons. The biofilm was formed on stainless steel coupons of AISI 304 with 2B finished (1  1 cm2) vertically placed in 24-well polystyrene plates in BHI. The biofilm was measured with the CV assay after 72 h incubation at 30
C. Presented values are the averages of 3 replicates performed on three independent occasions with standard deviation. The threshold, indicated by a solid line, values higher than the threshold level were considered positive for strong biofilm formation ability.

Y. Lin et al. / Food Control 82 (2017) 48e56

53

Fig. 3. Confocal laser scanning microscopic images of Bacillus cereus biofilms. Biofilms were stained with propidium iodide. a ¼ horizontal projection (xy). b and g: vertical projections (xz and yz, respectively). The cross-sections were taken along transects indicated by the green and red lines in the two-dimensional image. d: Biofilm profile. A: Bacillus cereus ATCC 10987; B: Bacillus cereus ATCC 14579; C: Bacillus cereus A1; D: Bacillus cereus A20; E: Bacillus cereus C1; F: Bacillus cereus B23. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

54

Y. Lin et al. / Food Control 82 (2017) 48e56

Table 2 Tolerance of the biofilms of the 27 dairy plant isolates of Bacillus cereus and reference strains ATCC1 4579 and ATCC 10987 to 2.0% w/v NaOH and 1.5% w/v HNO3 at 80
C. Strains

D-alkali (min)

D-acid (min)

MLST group

A17 A18 B24 A11 A19 A12 A1 B16 B18 B15 B23

1.77 3.19 3.29 7.83 3.83 4.60 9.52 6.10 6.89 3.03 2.72

± ± ± ± ± ± ± ± ± ± ±

0.31 1.18 0.86 0.32 0.87 1.08 0.19 0.27 0.07 1.16 0.68

6.21 3.25 3.93 3.08 1.79 2.07 3.34 4.87 2.95 2.16 2.00

± ± ± ± ± ± ± ± ± ± ±

0.71 0.11 0.06 1.26 0.11 0.94 1.09 0.07 0.06 0.05 0.03

Ⅰ

A3 A8 A2 A7 A14

3.59 2.59 4.25 8.00 1.92

± ± ± ± ±

0.07 0.08 0.27 0.05 1.18

2.01 3.19 4.99 4.41 2.59

± ± ± ± ±

0.78 0.07 2.17 0.05 1.08

Ⅱ

C2 C1 B21 C3 A6 A15 A20 A5 A4 A16

4.45 3.89 4.02 5.44 5.00 2.00 4.12 5.14 2.77 2.01

± ± ± ± ± ± ± ± ± ±

0.77 0.89 0.85 0.86 0.68 0.79 0.81 0.93 1.18 0.12

4.35 3.97 5.35 3.99 4.66 3.35 5.69 2.03 2.02 3.35

± ± ± ± ± ± ± ± ± ±

1.27 0.05 1.44 0.56 0.01 0.39 0.06 0.18 0.23 0.12

Ⅲ

B2

2.94 ± 0.05

2.94 ± 0.05

Ⅳ

ATCC10987 ATCC14579

6.36 ± 0.01 2.31 ± 0.67

3.44 ± 0.01 2.00 ± 0.24

reference reference

cereus in UHT milk processing lines and final products in two dairy plants in China, and identified their corresponding genotypic feature. The presence of toxin genes sometimes can reflect the risk levels of different Bacillus cereus strains. Previous studies revealed that the HBL, NHE, and cytotoxin-K proteins were the primary virulence factors in the diarrheal type of food poisoning (Granum, O'Sullivan, & Lund, 1999; Ngamwongsatit et al., 2008). In total, there was no clear correlation between toxigenic profiles with isolation origin of the 27 strains. The high occurrence of nheABC, observed in this study is consistent with most previous studies, such as those performed by Arslan (Arslan, Eyi, & Küçüksari, 2014), but higher than a study conducted in pasteurized milk in China (46.7%) (Zhou et al., 2008). Surprisingly, no hblACD toxin gene producing strains was found in this study, which is responsible for the diarrheal type of Bacillus cereus. Half of the Bacillus cereus strains (48.15%) carried the ces gene, which was associated with vomit. The distribution of ces gene was found to be relatively high in the isolated strains, compared to previous studies (6.5%) (Karagoz, Adiguzel, & Dikbas, 2015). Thus the percentage of strains simultaneously carrying diarrhea and vomit genes was high (37%), if contaminated food, which is likely to cause consumers to diarrhea and vomiting symptoms. The 27 strains were genotyped by MLST and grouped into 17 different STs, thus demonstrating great genetic diversity in isolates from UHT milk processing lines. ST857 (14.8%) was the predominant allelic profile and was followed by ST378, ST427 (11.1%) and ST144, ST760, ST1140 (7.4%), as shown in Table 2. The predominant subtypes were mainly isolated from silo tanks, pasteurization tanks, and filling room collected from A dairy plant, revealing the wide presence of the individual subtypes in different locations in the UHT milk processing lines. ST 378 was also isolated from farm dairy product in Italy (Cardazzo et al., 2008), harboring nheABC,

Table 3 Characteristics of Bacillus cereus isolates from UHT milk processing lines. B. cereus stain

Biofilm

D-values

Virus genes

ST

A1 A7 B18 A20 A6 B2 B16 B21 C2 A2

2.58 1.15 1.49 1.37 1.22 1.18 1.08 0.99 0.80 0.80

Hm Hm Mm Mm Mm Mm Mm Mm Mm Mm

NHE, bceT NHE, bceT NHE, entFM, bceT, ces

92 378 144 857 2473 2112 144 1236 427 378

A11 C3

1.63 1.07

Ml Ml

NHE, bceT, cytK, ces NHE, entFM, bceT, ces

205 427

A18 A3 A15 A12 C1 A16 A14 A17 A8 B24 A19 A5

0.93 0.91 0.91 0.87 0.86 0.80 0.76 0.74 0.58 0.58 0.56 0.55

Ll Ll Lm Ml Ll Ll Ll Lm Ll Ll Ll Ml

entFM, bceT NHE bceT, cytK entFM, bceT, cytK, ces NHE, bceT, cytK, ces NHE, entFM, bceT, ces NHE, entFM, bceT, ces NHE, entFM, bceT, cytK, ces bceT, cytK NHE, entFM entFM, bceT bceT

2642 1140 760 371 427 857 378 2038 1140 460 462 857

B15 A4 B23

0.47 0.47 0.27

Ll Ll Ll

NHE, entFM NHE, bceT NHE, entFM, cytK

1137 857 *

NHE, ces NHE, NHE, NHE, NHE

bceT ces bceT, cytK bceT, cytK, ces

H: high alkali tolerance, M: moderate alkali tolerance, L: low alkali tolerance, h: high acid tolerance, m: moderate acid tolerance, l: low acid tolerance.

hblACD and entFM genes. The ST144 was first isolated from food and tools in the outbreak for emetic toxin in Germany (Ehling-Schulz, Fricker, & Scherer, 2004b), and also isolated from infant formula in China (Yang et al., 2017). In our study, isolates bearing toxin genes did not belong to a single phylogenetic group, and the stains belonging to the same ST harbored different toxin genes, such like three ST378 strains. This indicated there was no significant correlation between the ST and toxin characteristics, which is in agreement with previous study (Cardazzo et al., 2008). Furthermore, regulation of toxin expression is complex, and the presence of toxin-encoding genes does not automatically equate with pathogenicity (Cardazzo et al., 2008). So with the currently available evidence, detection of the toxin proteins produced in particular strains is very necessary for their pathogenicity prediction. The survival properties of Bacillus cereus strains isolated from dairy processing lines were investigated, including biofilm-forming ability as well as tolerance of biofilms to hot-alkali and hot-acid. We classified the D-values of the Bacillus cereus isolates as high (Dvalues >8.0 min), moderate (4 min < D-values < 8 min), and low (Dvalues < 4 min). Two out of the 27 isolates had biofilms with high D-value, which was more than 8 min in hot alkali (pH > 13). The alkali tolerance strain (A1), with 1.57 log reduction of the biofilm viability when suspended in 2.0% NaOH at 80
C for 15 min, may be the most tolerant strain reported for Bacillus cereus. And it is interesting that some biofilm with highest resistance to hot-alkali were effectively killed by hot acid-washing, such as A5, A11. The results suggest that hot acid washing sometimes could be used to reduce colonization of dairy processing lines by hot-alkali resistant Bacillus cereus. Another strategy for Bacillus cereus survival was represented by isolates with biofilm forming ability. 9 out of the 27 strains tested in this study were able to form biofilms strongly, suggesting that the ability of Bacillus cereus to produce biofilms is a widespread feature

Y. Lin et al. / Food Control 82 (2017) 48e56

among strains isolated from dairy plants. Combining the results of biofilm formation ability with the biofilm resistance capacity (Table 3), we found that some strains with strong biofilm formation ability were more resistant to hot-alkali and hot-acid, and the tolerance of biofilm almost automatically equate with biofilmforming ability. First,15 strains showed low biofilm resistance and weak or moderate biofilm formation ability, predicting they could be inactivated by CIP. Besides, we can also find that only 2 out of 27 isolates (A11, C3) can form biofilm strongly but with low D-value of biofilm. Another ten strains showed both strong biofilm-forming ability and moderate biofilm resistance. The presence of the ten Bacillus cereus strains poses a threat to the UHT milk production, and effective targeted strategies should be used to minimize them. The strict assurance quality systems for the dairy industry should be adopted, as previous studies have demonstrated that the application of GMP and HACCP overall change the quality and safety of products manufactured (Costa Dias et al., 2012; Cusato et al., 2013, 2014). When comparing the persistence abilities and their MLST results (Table 3), we found large strain-to-strain variations in the genotype of the Bacillus cereus with similar survival properties. However, it is interesting that the isolates harboring strong and moderate biofilm-forming ability were most resistant to hot-alkali and hot-acid. Except for this, other relationships between biofilm formation and toxin production were not observed in this study. Previews study also reported that Bacillus cereus isolates are not related to the genetic background for the enterotoxin or emetic toxin production(Oh, Chang, Choi, Ok, & Lee, 2015). 5. Conclusion Overall, there was no obvious connection between the survival phenotypes and genotypes of the persistent Bacillus cereus isolated in this study. But the strains of MLST groupⅠhad stronger biofilm formation capability. The isolates harboring strong and moderate biofilm-forming ability had higher tolerance of alkali and acid. This study focused on a few strains with very strong survival ability, and revealed several properties explaining their successful colonization in the dairy processing lines. Ten persistent strains were identified with two strategies for survival: strong biofilm forming capability and high survival in the CIP washing liquid, which should be under strict supervision and control in the future. In addition, two strains only harbored ability to form biofilm but without biofilm resistance to CIP. These findings will contribute to improving strategies for minimizing Bacillus cereus in milk production. Acknowledgements This work was supported by the Beijing Municipal Commission of Education Co-constructed program, and Beijing Science and Technology Project (D141100004814001). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.foodcont.2017.06.025. References Aouadhi, C., Maaroufi, A., & Mejri, S. (2014). Incidence and characterisation of aerobic spore-forming bacteria originating from dairy milk in Tunisia. International Journal of Dairy Technology, 67(1), 95e102. http://dx.doi.org/10.1111/ 1471-0307.12088. Arslan, S., Eyi, A., & Küçüksari, R. (2014). Toxigenic genes, spoilage potential, and antimicrobial resistance of Bacillus cereus group strains from ice cream. Anaerobe, 25, 42e46. http://dx.doi.org/10.1016/j.anaerobe.2013.11.006.

55

Bremer, P. J., Fillery, S., & McQuillan, A. J. (2006). Laboratory scale clean-in-place (CIP) studies on the effectiveness of different caustic and acid wash steps on the removal of dairy biofilms. International Journal of Food Microbiology, 106(3), 254e262. http://dx.doi.org/10.1016/j.ijfoodmicro.2005.07.004. Cardazzo, B., Negrisolo, E., Carraro, L., Alberghini, L., Patarnello, T., & Giaccone, V. (2008). Multiple-locus sequence typing and analysis of toxin genes in Bacillus cereus food-borne isolates. Applied and Environmental Microbiology, 74(3), 850e860. http://dx.doi.org/10.1128/AEM.01495-07. Ceuppens, S., Boon, N., & Uyttendaele, M. (2013). Diversity of Bacillus cereus group strains is reflected in their broad range of pathogenicity and diverse ecological lifestyles. Fems Microbiology Ecology, 84(3), 433e450. http://dx.doi.org/10.1111/ 1574-6941.12110. Costa Dias, M. A., Sant’Ana, A. S., Cruz, A. G., Faria, J.de A. F., Fernandes de Oliveira, C. A., & Bona, E. (2012). On the implementation of good manufacturing practices in a small processing unity of mozzarella cheese in Brazil. Food Control, 24(1e2), 199e205. http://dx.doi.org/10.1016/j.foodcont.2011.09.028. €rtlbauer, E., Cao, J., Ding, S., et al. (2016). CharacterCui, Y., Liu, X., Dietrich, R., Ma ization of Bacillus cereus isolates from local dairy farms in China. FEMS Microbiology Letters, 363(12), 1e6. http://dx.doi.org/10.1093/femsle/fnw096. Cusato, S., Gameiro, A. H., Corassin, C. H., Sant’Ana, A. S., Cruz, A. G., Faria, J.de A. F., et al. (2013). Food safety systems in a small dairy Factory: Implementation, major challenges, and assessment of systems' performances. Foodborne Pathogens and Disease, 10(1), 6e12. http://dx.doi.org/10.1089/fpd.2012.1286. Cusato, S., Gameiro, A. H., Sant’Ana, A. S., Corassin, C. H., Cruz, A. G., & de Oliveira, C. A. F. (2014). Assessing the costs involved in the implementation of GMP and HACCP in a small dairy factory. Quality Assurance and Safety of Crops and Foods, 6(2), 135e139. http://dx.doi.org/10.3920/QAS2012.0195. Ehling-Schulz, M., Fricker, M., & Scherer, S. (2004a). Bacillus cereus, the causative agent of an emetic type of food-borne illness. Molecular Nutrition and Food Research. http://dx.doi.org/10.1002/mnfr.200400055. Ehling-Schulz, M., Fricker, M., & Scherer, S. (2004b). Identification of emetic toxin producing Bacillus cereus strains by a novel molecular assay. FEMS Microbiology Letters, 232(2), 189e195. http://dx.doi.org/10.1016/S0378-1097(04)00066-7. n e zech, T., Midelet-Bourdin, G., Lequette, Y., Clarisse, M., Ronse, G., et al. Faille, C., Be (2014). Sporulation of Bacillus spp. within biofilms: A potential source of contamination in food processing environments. Food Microbiology, 40, 64e74. http://dx.doi.org/10.1016/j.fm.2013.12.004. ~o, G. (2014). Adhesion and production Flach, J., Grzybowski, V., Toniazzo, G., & Corça of degrading enzymes by bacteria isolated from biofilms in raw milk cooling tanks. Food Science and Technology (Campinas), 34(3), 571e576. http:// dx.doi.org/10.1590/1678-457x.6374. Granum, P. E., O'Sullivan, K., & Lund, T. (1999). The sequence of the non-haemolytic enterotoxin operon from Bacillus cerceus. Fems Microbiology Letters, 177(2), 225e229. http://dx.doi.org/10.1016/s0378-1097(99)00312-2. Gundogan, N., & Avci, E. (2014). Occurrence and antibiotic resistance of Escherichia coli, Staphylococcus aureus and Bacillus cereus in raw milk and dairy products in Turkey. International Journal of Dairy Technology, 67(4), 562e569. http:// dx.doi.org/10.1111/1471-0307.12149. Hayrapetyan, H., Abee, T., & Nierop Groot, M. (2016). Sporulation dynamics and spore heat resistance in wet and dry biofilms of Bacillus cereus. Food Control, 60, 493e499. http://dx.doi.org/10.1016/j.foodcont.2015.08.027. Hayrapetyan, H., Muller, L., Tempelaars, M., Abee, T., & Nierop Groot, M. (2015). Comparative analysis of biofilm formation by Bacillus cereus reference strains and undomesticated food isolates and the effect of free iron. International Journal of Food Microbiology, 200, 72e79. http://dx.doi.org/10.1016/ j.ijfoodmicro.2015.02.005. In Lee, S. H., Barancelli, G. V., de Camargo, T. M., Corassin, C. H., Rosim, R. E., da Cruz, A. G., … de Oliveira, C. A. F. (2017). Biofilm-producing ability of Listeria monocytogenes isolates from Brazilian cheese processing plants. Food Research International, 91, 88e91. http://dx.doi.org/10.1016/j.foodres.2016.11.039. Karagoz, K., Adiguzel, G., & Dikbas, N. (2015). Emetic and enterotoxin profiles of food-borne Bacillus cereus strains from Turkey. Research Journal of Biotechnology, 10(6), 44e49. Kim, S. K., Kim, K.-P., Jang, S. S., Shin, E. M., Kim, M.-J., Oh, S., et al. (2009). Prevalence and toxigenic profiles of Bacillus cereus isolated from dried red peppers, rice, and sunsik in Korea. Journal of Food Protection, 72(3), 578e582. Kumari, S., & Sarkar, P. K. (2014). Prevalence and characterization of Bacillus cereus group from various marketed dairy products in India. Dairy Science and Technology, 94(5), 483e497. http://dx.doi.org/10.1007/s13594-014-0174-5. Lane, D. J. (1991). 16S/23S rRNA sequencing. In E. Stackebrandt, & M. Goodfellow (Eds.), Nucleic acid techniques in bacterial systematics (pp. 115e147). Chichester: John Wiley and Sons. Lee, S. H. I., Cappato, L. P., Corassin, C. H., Cruz, A. G., & Oliveira, C. A. F. (2016). Effect of peracetic acid on biofilms formed by Staphylococcus aureus and Listeria monocytogenes isolated from dairy plants. Journal of Dairy Science, 99(3), 2384e2390. http://dx.doi.org/10.3168/jds.2015-10007. Lee, S. H. I., Mangolin, B. L. C., Gonçalves, J. L., Neeff, D. V., Silva, M. P., Cruz, A. G., et al. (2014). Biofilm-producing ability of Staphylococcus aureus isolates from Brazilian dairy farms. Journal of Dairy Science, 97(3), 1812e1816. http:// dx.doi.org/10.3168/jds.2013-7387. Lemos, M., Gomes, I., Mergulhao, F., Melo, L., & Simoes, M. (2015). The effects of surface type on the removal of Bacillus cereus and Pseudomonas fluorescens single and dual species biofilms. Food and Bioproducts Processing, 93, 234e241. http://dx.doi.org/10.1016/j.fbp.2014.08.009. Montanhini, M. T. M., & Bersot, L. D. S. (2013). Evaluation of psychrotrophic behavior

56

Y. Lin et al. / Food Control 82 (2017) 48e56

and lipolytic and proteolytic activity of Bacillus cereus isolated from refrigerated dairy products. Acta Scientiarum Technology, 35(1), 163e167. http:// dx.doi.org/10.4025/actascitechnol.v35i1.13752. Ngamwongsatit, P., Buasri, W., Pianariyanon, P., Pulsrikarn, C., Ohba, M., Assavanig, A., et al. (2008). Broad distribution of enterotoxin genes (hblCDA, nheABC, cytK, and entFM) among Bacillus thuringiensis and Bacillus cereus as shown by novel primers. International Journal of Food Microbiology, 121(3), 352e356. http://dx.doi.org/10.1016/j.ijfoodmicro.2007.11.013. Oh, S. K., Chang, H.-J., Choi, S.-W., Ok, G., & Lee, N. (2015). Toxin profile, biofilm formation, and molecular characterization of emetic toxin-producing Bacillus cereus group isolates from human stools. Foodborne Pathogens and Disease, 12(11), 914e920. http://dx.doi.org/10.1089/fpd.2015.2004. O'Connell, A., Lawton, E. M., Leong, D., Cotter, P., Gleeson, D., & Guinane, C. M. (2016). Detection of presumptive Bacillus cereus in the Irish dairy farm environment. Irish Journal of Agricultural and Food Research, 55(2), 145e151. http:// dx.doi.org/10.1515/ijafr-2016-0014IJAFR. Reis, A. L. S., Montanhini, M. T. M., Bittencourt, J. V. M., Destro, M. T., & Bersot, L. S. (2013). Gene detection and toxin production evaluation of hemolysin BL of Bacillus cereus isolated from milk and dairy products marketed in Brazil. Brazilian Journal of Microbiology, 44(4), 1195e1198. Ryu, J. H., & Beuchat, L. R. (2005). Biofilm formation and sporulation by Bacillus cereus on a stainless steel surface and subsequent resistance of vegetative cells and spores to chlorine, chlorine dioxide, and a peroxyacetic acid-based sanitizer. Journal of Food Protection, 68(12), 2614e2622.

Seong, S. J., Lim, J. S., Lee, K. G., Lee, S. J., & Hong, K. W. (2008). Toxin gene profiling of Bacillus cereus food isolates by PCR. Journal of the Korean Society for Applied Biological Chemistry, 51(4), 263e268. http://dx.doi.org/10.3839/jksabc.2008. 046. Shaheen, R., Svensson, B., Andersson, M. A., Christiansson, A., & SalkinojaSalonen, M. (2010). Persistence strategies of Bacillus cereus spores isolated from dairy silo tanks. Food Microbiology, 27(3), 347e355. http://dx.doi.org/10.1016/ j.fm.2009.11.004. Soni, A., Oey, I., Silcock, P., & Bremer, P. (2016). Bacillus spores in the food industry: A review on resistance and response to novel inactivation technologies. Comprehensive Reviews in Food Science and Food Safety, 15(6), 1139e1148. http:// dx.doi.org/10.1111/1541-4337.12231. Yang, Y., Yu, X., Zhan, L., Chen, J., Zhang, Y., Zhang, J., … Mei, L. (2017). Multilocus sequence type profiles of Bacillus cereus isolates from infant formula in China. Food Microbiology, 62, 46e50. http://dx.doi.org/10.1016/j.fm.2016.09.007. Zhou, G., Liu, H., He, J., Yuan, Y., & Yuan, Z. (2008). The occurrence of Bacillus cereus, B. thuringiensis and B. mycoides in Chinese pasteurized full fat milk. International Journal of Food Microbiology, 121(2), 195e200. http://dx.doi.org/10.1016/ j.ijfoodmicro.2007.11.028. Zhou, G., Zheng, D., Dou, L., Cai, Q., & Yuan, Z. (2010). Occurrence of psychrotolerant Bacillus cereus group strains in ice creams. International Journal of Food Microbiology, 137(2e3), 143e146. http://dx.doi.org/10.1016/j.ijfoodmicro.2009. 12.005.