2015 Volume 21 Issue 2 Pages 275-279
The incidence, intraspecific diversity and toxigenic profile of Bacillus cereus in the Yellow-Water were investigated. Of 50 samples tested, 17 (34%) were contaminated by B. cereus with 3.7 × 103 − 4.9 × 103 CFU/mL. The intraspecific diversity of B. cereus was observed by repetitive (GTG)5 sequence polymorphism-PCR, with six subtypes, and were further discriminated into seven subtypes by 16S-23S rRNA intergenic transcribed spacer. 94% of isolates contained at least 2 toxin genes of nheABC, hblCDA, cytK, entFM and EM1, but 6.0% carried none of them. The detection rates of nheABC, hblCDA, cytK and entFM were 94%, 88%, 46% and 70%, respectively. All isolates with nheABC could produce NHE, but of the isolates harbored hblCDA, 52% was positive for HBL. Only 6.0% isolates possessed EM1 and they were positive for emetic toxin assay. The results suggested that Yellow-Water needs stricter microbial standard as a fermented food flavor enhancer.
Bacillus cereus, a widespread pathogen, can be found in various types of food or foodstuffs and is increasingly reported as a major agent in cases of bacterial diarrheal and emetic food poisoning (Park et al., 2009). Diarrheal syndrome is caused by any one or a combination of heat-labile toxins hemolysin BL (HBL, encoded by hblA, hblC and hblD), non-hemolytic enterotoxin (NHE, encoded by nheA, nheB and nheC), haemolytic single protein enterotoxin cytotoxin K (cytK), and enterotoxin FM (entFM) (Schoeni and Wong, 2005). The emetic type is caused by the cereulide (EM1), a small peptide toxin which is not inactivated by any type of food processing including heat, acid and protease (Rajkovic et al., 2008). Although B. cereus is an opportunistic pathogen, data available shows that B. cereus has been implicated in about 33% of all bacterial foodborne diseases in Norway (1988–1993), 47% in Iceland (1987–1992), 22% in Finland (1992) and 8% in the Netherlands (1991) (Adams and Moss, 2000). The high risk of B. cereus food poisoning has already given rise to concern in the world food industry.
The Yellow-Water is an important metabolic product of Chinese traditional solid state brewing liquor, and it is from the fermenting liquor lees which consist of the mixture grains such as sorghum, corn, wheat, rice, cavings, the fermentation starter Qu and the mother lees (Li et al., 2013). During the natural fermenting, the microorganisms, which come from the cellar mud, the Qu, the mother lees and the local natural environment, take part in the metabolic activities of liquor lees and the metabolic water is produced associated with these activities. When the metabolic water circulates in the fermenting lees, many flavors resolve in it and gradually deposit in the bottom of soil cellar. The brown metabolic water in the bottom of soil cellar is named as Yellow-Water. In Chinese traditional solid state brewing liquor industry, the Yellow-Water was used to act as the liquor lees-filling fermentation, the culture of manmade cellar mud and the cross-distillation of liquor for improving the liquor flavor (Peng, 2011). In recent years, the Yellow-Water was exploited into the fermented food flavor enhancer by some food condiment producers in China. Unfortunately, some fermented foods, such as fermented soybean, fermented rice milk and fermented rice and vegetable roll, are often sporadically contaminated by B. cereus when the Yellow-Water flavor enhancer is incorporated into the fermentation. Although some researchers have already thrown light on the B. cereus in the Yellow-Water (Peng, 2009), few studies have been conducted on the incidence, intraspecific diversity and toxigenic profile of B. cereus in the Yellow-Water.
Therefore, the major objective of this study was to elucidate the incidence, intraspecific diversity and toxigenic profile of B. cereus in the Yellow-Water, based on genotypic (16S rRNA, gyrB, rpoB, 16S – 23S rRNA intergenic transcribed spacer (16S – 23S rRNA ITS), repetitive (GTG)5 sequence polymorphism-PCR (GTG)5 Rep-PCR), hbl, nhe, cytK, entFM and EM1) and phenotypic methods. Concurrently, the ability of B. cereus strains to HBL, NHE and emetic toxin were also assessed by blood agar plate diffusion assay, immunological kits BDE-VIA and HEp-2 human larynx carcinoma cells, respectively. To our knowledge, this is the first report on the incidence, intraspecific diversity and safety issues of B. cereus recovered from the commercial Yellow-Water.
Isolation of B. cereus A total of 50 Yellow-Water samples were collected from different brands or different fermentation batches from 2011 to 2012. 10 mL sample was suspended in 90 mL PBS buffer (pH 7.0) and vortexed for 2 min, and then centrifuged at 8,000 revolutions per minute for 1 min at 20°C. The resulting pellet was re-suspended in 5 mL PBS buffer and serially diluted (10-fold) in 0.85% (w/v) NaCl solution. Then 0.1 mL of suitable dilution was inoculated onto Mannitol-Egg Yolk-Polymyxin B agar (Oxoid, Basingstoke, Hampshire, UK) and followed by incubation at 30°C for 24 h. The suspicious colonies were selected to sub-culture on nutrient agar. All colonies on the sub-cultures were screened with 16S rRNA analysis, rpoB and gyrB according to Li et al. (2013) and Ahaotu et al. (2013). The presumptive B. cereus strains were further confirmed by using API 50 CHB and API 20 E kit (BioMérieux, Marcy LÉtoile, France). The isolates of B. cereus were maintained in 30% glycerol at −30°C.
B.cereus ATCC 14579 from the Chinese Center for Disease Control and Prevention (CCDCP, Beijing, China) was used as control strain. It harbores the hblACD, nheABC, cytK and entFM genes (Chon et al., 2012).
Intraspecific diversity Genomic DNA was extracted and purified from B. cereus in the mid-logarithmic growth phase in the nutrient broth according to Xiang et al. (2011). The (GTG)5 Rep-PCR and the 16S – 23S rRNA ITS were employed to investigate intraspecific diversity as described by Ahaotu et al. (2013) and Xiang et al. (2011), respectively. The amplification fingerprint were visualized by electrophoresis on PAGE and converted to a binary (0/1) of the matrix. Then, the data was statistically analyzed by the software program NTSYS PC 2.11.
Detection of enterotoxin and emetic toxin genes The genes for enterotoxin HBL (hblA, hblC and hblD) and NHE (nheA, nheB and nheC) were amplified as described by Ouoba et al. (2008). The presence of enterotoxin genes cytK and entFM, and cereulide gene EM1 were detected by the method described by Thorsen et al. (2011) and Chon et al. (2012). All amplified genes were respectively cloned into the pGEM-T plasmid vector (Promega, USA) and transformed into the chemically competent E. coli DH 5α cells for sequencing. The amplified gene sequences were further evaluated by blastx program in the NCBI.
Determination of haemolytic and non haemolytic activity The haemolytic activity was detected by blood agar plate diffusion assay. The Luria-Bertani (LB) agar base was autoclaved at 121°C for 15 min and the sterile defibrinated sheep blood (5%, w/v) was mixed in it after cooling to 50°C and then poured into the plates. Aliquots (100 µL) of the culture supernatant of B. cereus isolates were added into the 8 mm well in the plates and incubated at 37°C for 12 h. The haemolytic activity was indicated by the zone of clearing around the wells. The commercial immunological kit BDE-VIA (Tecra Diagnostics, Reading, UK) was used with cell-free culture supernatants to assess NHE production. Strains with an index of less than 3 in BDE-VIA test were considered negative (Chon et al., 2012).
Detection of the emetic toxin The HEp-2 human larynx carcinoma cells were employed to detect the emetic toxin production by vacuolation assay. 10 mL overnight cultures of B. cereus in 10% skim milk medium were transferred into the 100 mL Mannitol-Egg Yolk-Polymyxin B medium and incubated at 30°C with 200 revolutions per minute continuous agitation for 48 h. 50 mL cultures were centrifuged at 3,000 × g under 4°C for 10 min and the supernatants were filtered through 0.45 µm disposable filters. The fluids were further autoclaved at 121°C for 15 min to sterilize the samples and to denature heat-labile toxins. Then, the heat-treated fluids were added to a HEp-2 tissue culture to perform toxicity assay (Chon et al., 2012).
Incidence of B. cereus During the solid state liquor brewing, those microorganisms which are obligate and facultative anaerobes or can form endogenous spores can inhabit in the Yellow-Water because of high acid, high alcohol and anaerobic environment, such as Clostridium, Lactobacillus, Serratia, Bacillus, etc (Peng, 2011). While, only Bacillus may survive in the Yellow-Water after Yellow-Water is processed into the flavor enhancer (Peng, 2009). In the Bacillus genera, the B. cereus spores and vegetative forms are widely distributed in various natural environments and food types, specially cereals and cereal derivative food (King et al., 2007). And thus they are commonly isolated from the traditional cereal-based fermentedonal cereal-based fermented food or additives (Thorsen et al., 2010). In current study, the isolates belonging to the B. cereus were detected in 34% (17 out of 50) of the samples analyzed, and their counts were about 3.7 × 103 − 4.9 × 103 CFU/mL in the contaminated samples. Peng (2009) reported that 24.6 − 37.4% of Yellow-Water samples were inhabited by B. cereus and the major contamination reason is that the brewing materials were contaminated by B. cereus. This contaminated ratio was consistent with our current results. It is important to mention that even though B. cereus was often found in the Yellow-Water and the Yellow-Water was used to act as cross-distillation of liquor to improve liquor quality, no reports on B. cereus toxin poisoning from the consumption of Chinese Liquor exists. This may be due to the heat-labile characteristics of enterotoxins and the non-distillate characteristics of emetic toxin when the Yellow-Water was distillated. In Europe, B. cereus at levels above 103 – 105 CFU/g or mL food is considered unsafe (Thorsen et al., 2010). Although the B. cereus was about 3.7 × 103 − 4.9 × 103 CFU/mL in the contaminated samples, its multiplication was usually highly promoted when it was incorporated into the fermentation of some fermented foods (Peng, 2009). Therefore, the high incidence of B. cereus in the Yellow-Water would bring potential high food poisoning risk of some fermented foods.
Intraspecific diversity of B. cereus A total of 431 B. cereus isolates were obtained from 17 contaminated samples. 150 isolates were randomly selected to further analyze. According to 16S rRNA gene sequences (GenBank ID: KJ 720020 − KJ 720026), these isolates can be identified as B. cereus (≥ 99% similarity) (Table 1). The same identity was also confirmed by the gyrB (GenBank ID: KJ 720006 – KJ 720012) and rpoB genes (GenBank ID: KJ 720013 − KJ 720019) (Table 1). Interestingly, their biochemical profiles exhibited differences in the API 50 CHB analysis to some degree, including salicin hydrolysis, starch fermentation, hemolysis and lecithinase hydrolysis (Table 2). This finding implied that the intraspecific diversity has happened although with ≥ 99% 16S rRNA and ≥ 95% gyrB and rpoB similarity each other. Xiang et al (2011) demonstrated that 16S rRNA may not always provide sufficient variability to discriminate the different strains of the same species. Recently, the 16S – 23S rRNA ITS and (GTG)5 Rep-PCR length polymorphism have been utilized to discriminate the different strains of the same species but with some differences in biochemical characterization (Xiang et al., 2011; Chon et al., 2012). In current study, the (GTG)5 Rep-PCR fingerprints, a useful tool to describe the genetic diversity of members of B. cereus species, have shown 6 subtypes of isolates (Fig. 1A). In general, their fingerprints and band intensity distributions were clearly different each other. However, (GTG)5 Rep-PCR fingerprints could not discriminate YB17 group and YB43 group. Therefore, 16S – 23S rRNA ITS was employed to further analyze these closely related isolates. The 16S – 23S rRNA ITS fingerprints generated seven subtypes and YB17 and YB43 groups were also discriminated (Fig.1B). These molecular evolutionary evidences, as well as their biochemical difference, implied that B. cereus isolates from the Yellow-Water have fell into 7 different subtypes under the levels of B. cereus species, suggesting intraspecific diversity.
| Group | Number | 16S rRNA (≥99%) | rpoB | gyrB |
|---|---|---|---|---|
| YB05 | 42 (28%) | B. cereus | B. cereus (99%) | B. cereus (100%) |
| YB17 | 9 (6%) | B. cereus | Bacillus sp (95%) | B. cereus (97%) |
| YB18 | 18 (12%) | B. cereus | B. cereus (99%) | B. cereus (98%) |
| YB21 | 9 (6%) | B. cereus | B. cereus (99%) | B. cereus (97%) |
| YB37 | 27 (18%) | B. cereus | B. cereus (99%) | B. cereus (97%) |
| YB43 | 36 (24%) | B. cereus | Bacillus spp (95%) | B. cereus (96%) |
| YB50 | 9 (6%) | B. cereus | B. cereus (99%) | B. cereus (97%) |
| Group | Salicin | Starch | Hemolysis | Lecithinase |
|---|---|---|---|---|
| YB05 | +a | + | + | + |
| YB17 | − | − | − | + |
| YB18 | + | − | − | + |
| YB21 | + | + | + | + |
| YB37 | + | + | + | − |
| YB43 | + | − | + | − |
| YB50 | − | + | − | − |
| 14579 | + | + | + | + |
Cluster analysis of (GTG)5 Rep-PCR (A) and 16S – 23S rRNA its-PCR (B) of B. cereus strains from the Yellow-Water.
Distribution profiles of diarrheal and emetic toxin genes In B. cereus strains of 7 groups, the toxin genes hbl A, C and D, nhe A, B and C, entFM, cytK and EM1 were detected and the profiles were displayed in Table 3. In all isolates, only YB50 group (6.0%) did not carry all toxin genes in current investigation, the others (94%) contained at least 2 toxin genes. The nhe genes were highly prevalent in the isolates and 94% of strains carried nhe genes, which agreed with 90 – 100% of food related strains possessing these genes (Thorsen et al., 2011; Chon et al., 2012). Concurrently, the hbl genes were also highly prevalent in the isolates. 52% of isolates possessed complete set hbl genes and 36% of isolates have only one or two of hblA, hblC and hblD gene (Table 3). In general, the percentage of hbl positives in this study (88%) is consistent with previous works that reported 84 − 92% of incidence in food related B. cereus isolates (Park et al., 2009; Kim et al., 2011; Chon et al., 2012). The cytK and entFM were 46% and 70% of prevalence in the Yellow-Water, respectively. Variable detection rates of cytK gene in food related isolates have been reported in previous studies ranging from 13 to 57% of prevalence (Guinebretiere et al., 2002; Kim et al., 2011). Interestingly, 70% prevalence of entFM in the Yellow-Water is not in good agreement with previous investigates, 90 − 100% (Guinebretiere et al., 2002; Kim et al., 2011). The proportion of emetic toxin isolates varied in different foods (Chon et al., 2012). Svensson et al. (2006) reported that 0.05% of isolates from dairy production chain was positive for emetic toxin genes, while Altayar and Sutherland (2006) found that the prevalence was 16% in potatoes. In current study, the emetic toxin EM1 gene has been detected in only 6.0% of prevalence. The emetic toxin gene has been shown to be highly conserved in cereulide producing B. cereus isolates regardless of origin (food and geographical origin) (Chon et al., 2012), suggesting that the current results are reliable. Emetic B. cereus strains usually have common phenotypic traits, such as inability to hydrolyze starch or salicin, and no hemolysis (Ehling-Schulz et al., 2005). The biochemical tests of emetic B. cereus strains in the Yellow-Water well corresponded with all these characteristics (Table 2).
| Group | hbl gene | nhe gene | Other enterotoxin | Cereulide | toxin detection | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| A | C | D | A | B | C | cytK | entFM | EM1 | HBL | NHE | emetic | |
| YB05 | +a | + | + | + | + | + | + | + | − | 17.0 | 4 | − |
| YB17 | − | − | − | + | + | − | + | − | + | − | 3 | + |
| YB18 | − | − | + | + | + | + | + | + | − | − | 5 | − |
| YB21 | + | + | + | + | + | − | − | + | − | 10.0 | 3 | − |
| YB37 | + | + | + | − | + | + | − | − | − | 7.0 | 4 | − |
| YB43 | + | − | + | + | − | + | − | + | − | − | 3 | − |
| YB50 | − | − | − | − | − | − | − | − | − | − | − | − |
| 14579 | + | + | + | + | + | + | + | + | − | 17.0 | 4 | − |
Detection of toxin production All strains harbored complete or part set nhe genes could produce the NHE toxin as determined using the Tecra BDE-VIA kit (Table 3). However, in isolates which carry hbl genes, only those with complete hbl genes displayed the haemolytic activity on blood agar, the other strains possessed one or two of hblA, hblC and hblD genes were negative for the HBL toxin production reaction (Table 3). In fact, the expression of toxin genes may not always occur although the majority of B. cereus harbors the diarrhoeal toxin genes (Samapundo et al., 2011). HBL has been reported to be generally produced by 45 – 65% of B. cereus from the environment or food (Thaenthanee et al., 2005). In the HBL toxin positive strains from the Yellow-Water, 53.8% strains produced high levels of the toxin (with an index of 17 mm), while 61.7% of NHE toxin positive strains produced high levels of the toxin (production at an index of ≥ 4) according to the criteria presented in Guinebretiere et al. (2002). In emetic toxin study, the YB17 group was positive for EM1 gene (Table 3) and it was also positive for HEp-2 cell assay.
In the Yellow-Water samples, the B. cereus contaminated incidence was about 34%. And the polymorphism of (GTG)5 Rep-PCR and 16S – 23S rRNA ITS indicated the intraspecific diversity, respectively with six and seven subtypes. The detection rates of nheABC, hblCDA, cytK and entFM diarrheal toxin genes in isolates were 94%, 88%, 46% and 70%, respectively. All isolates with nheABC genes could produce NHE, but 52% of isolates with hblCDA genes was positive for HBL. Only 6.0% isolates harbored EM1 genes and were positive for emetic toxin assay. The results suggested that Yellow-Water has a potential risk of food poisoning when it is employed to act as a flavor enhancer and is incorporated into the fermentation. Therefore, it needs stricter microbial standard and effective methods to protect the contamination, such as washing of cereals materials, heating of cereals materials, filtration of Yellow-Water by 0.45 µm membrane, etc. Furthermore, the actual levels of B. cereus in fermented food incorporated by Yellow-Water should also receive more attention and further studies.
Acknowledgements This work was financially supported by the applied basic research program of the science and technology department of Sichuan, China (2014JY0045), the key research program of the education department of Sichuan, China (14ZA0110), the Chunhui Program of ministry of education of the people's republic of China (Z2014061) and the key laboratory of food biotechnology of Sichuan, Xihua University, Sichuan, China (SZJJ2014-007).
