Insight into Bacillus cereus Associated with Infant Foods in Beijing

This study was undertaken to investigate the prevalence, antimicrobial resistance, and virulence gene profiles of Bacillus cereus in different brands of infant formula in Beijing supermarkets. Eighty-eight Bacillus cereus isolates were recovered in sixty-eight infant formulas of five domestic brands and fourteen imported brands. The prevalence rate in domestic and imported samples were 70.6% and 52.9%, respectively. Lower mean prevalence level was found in domestic samples (1.17 MPN/g) compared with the imported samples (3.52 MPN/g). Twenty-four virulence gene profiles were found, and most strains carried at least one virulence gene. The prevalence of nheA, nheB, nheC, cytK, bceT, and entFM in domestic and imported brand samples was similar. The occurrence of enterotoxin genes hblA, hblC, and hblD in domestic samples were 22.2%, 27.8%, and 22.2%, respectively, which was significantly higher than imported samples. Antimicrobial drugs-susceptibility analysis showed that all isolates were susceptible to gentamincin, amikacin, and ciprofloxacin; 38%, 7%, and 2.3% were resistant to rifampin, tetracycline, and chloramphenicol, respectively; and only one isolate was resistant to trimethoprim-sulfamethoxazole. Moreover, the cell numbers of Bacillus cereus in prepared infant formula increased rapidly at room temperature. Thus, monitoring guidelines are needed for accepted levels of Bacillus cereus in infant formula.


Introduction
Infant formula is a major source of nutrition for children before they can digest other foods. The immune system of babies is weak, and any pathogen present in their food may cause illness. Therefore, the hygienic quality of infant formula is important to protect the health of infants and to diminish risks and associated with its consumption. As one of the high-protein foods, dairy products are conducive to foodborne pathogen proliferation. A growing research has suggested that Bacillus cereus is a common pathogen in milk or milk-related products [1,2].
Bacillus cereus, a spore-forming bacterium, is an opportunistic human food-borne pathogen, which is widely distributed in environment and frequently isolated as a contaminant of cereals, processed milk products, and other foods [3,4]. The spores of B. cereus are specifically troublesome in the food industry because they can be intractable to pasteurization, radiation, disinfectants, and desiccation, and their hydrophobic nature allows them to adhere to the surface of solid materials [5,6]. In addition to gastroenteritis, B. cereus can also cause systemic and local infection in immunologically compromised individuals [7].
Bacillus cereus is capable of producing emesis toxin (ETE) and three enterotoxins, including hemolytic enterotoxin (HBL), nonhemolytic enterotoxin (NHE), and enterotoxin K (EntK), of which HBL and NHE are important enterotoxins that cause diarrhea-based food poisoning. There are five virulence genes associated with the production of these enterotoxins: hemolysin BL gene (hbl), non-hemolytic enterotoxin gene (nhe), enterotoxin FM gene (entFM), enterotoxin T gene (becT), and cytotoxin K gene (cytK). HBL requires all three genes positive for hblA, hblC, and hblD to be toxic, while NHE is most toxic when all three genes are positive for nheA, nheB, and nheC. The virulence gene expressing vomitoxin is ces gene, which encodes a heat-resistant toxin that is not easily cleaved [4,8]. A total of 24 virulence gene carriage patterns of B. cereus were identified, among which the nhe gene had the highest carriage rate of 92.98%, followed by the entFM gene (71.93%), and 70.18% of strains carried both nhe and entFM genes. The subtyping results showed that the carriage rates of nheA, nheB, and nheC genes were 88.72%, 88.72%, and 49.12%, respectively. The hemolysin BL gene carriage rates were 24.56% for hblA, 22.81% for hblC, 17.54% for hblD, and 22.81% for cytK [8,9].
Besides the enterotoxic toxins, it has been found that some B. cereus isolates show resistance to antimicrobial drugs [10]. A variety of antimicrobial drugs are widely used, and the problem of bacterial resistance is getting worse, causing great threat to human health. With antimicrobial agents widely used in farmed-animal industries, the food chain constitutes an important source of antimicrobial resistance [11]. There is evidence that resistant microorganisms can spread to humans via the food chain or indirect contact from farm animal waste [12,13]. Therefore, it is necessary to evaluate the antibiotic resistance of B. cereus in infant formula.
Up to now, many countries have stipulated the limit standard of B. cereus in infant formula to control this bacterium. In Australia and New Zealand, it was (n = 5, c = 0, m = 100 cfu/g (mL)); in Canada, it was (n = 10, c = 1, m = 100 cfu/g (mL), M = 10, 000 cfu/g (mL)); and in the European Union, it was (n = 5, c = 1, m = 100 cfu/g (mL), M = 500 cfu/g (mL)). While in China, the national standard of infant formula, GB 10765-2010, does not stipulate the limit standard of B. cereus in infant formula, the control and inspection of B. cereus in infant formula is mainly according to a prevalence level less than 100, 000 cfu/g (mL). Therefore, it is important to determine the prevalence and also enumerate B. cereus in babies' food in order to assess its safety. Therefore, the present study was undertaken to evaluate the prevalence and potential risk of B. cereus in infant formulas in Beijing. The research was further extended to carry out the virulence genes profile, and antimicrobial-resistance profiles of the isolates were evaluated.

Sample Collection
A total of 68 infant formula samples were collected from supermarkets in Beijing during the period from June to July 2014. Among the 68 samples, 17 were from different domestic brands, and the other 51 samples involved 14 different imported brands. The samples of every brand were from different batches (1 to 5). The infant formulas selected in this study included most of the popular brands in the Chinese retail market.

Microbiological Analysis
The B. cereus in infant formula was detected by using a most probable number (MPN) procedure method described by Tallent et al. [14]. Briefly, 25 g of each sample was suspended in 225 mL of 0.85% saline and beaten for 1 min by a homogenizer (BAG-MIXER100, interscience, France). Then, 10 mL of homogenate was serially diluted (10-fold) in 0.85% physiological saline with selected 3 continuous dilution, and then, 3 × 10 mL of each dilution was inoculated into 3 tubes with 10 mL double trypticase soy polymyxin broth (TSPB, Difco, 7 Loveton Circle, Sparks, MD, USA), followed by incubation at 30 • C for 48 h. The culture was then taken from tubes streak and applied onto mannitol yolk polymyxin (MYP) agar (Difco, 7 Loveton Circle, Sparks, MD, USA) and then incubated at 30 • C for 24 h. The typical colonies from plates that contain suspicious colonies were cultured on blood plates (Oxoid, Hampshire, UK) incubated at 30 • C for 24 h. The colony on the blood plates were confirmed by BD PhoenixTM-100 (Becton, Dickinson and Company, 7 Loveton Circle Sparks, USA), and the gyrB gene was detected using a previously described PCR method for the identification of Bacillus cereus [15]. The strains were conserved in Inspection and Quarantine of China microbial culture collection management center (IQCC).

Detection of Virulence Genes
The strains isolated from infant formula samples were screened by PCR for the ten virulence genes, including hblA/C/D, nheA/B/C, cytK, bceT, ces, and entFM [9]. For extraction of DNA template, bacteria were plated on Tryptone soya agar (TSA, Oxoid, Hampshire, England) and incubated at 30 • C for 24 h; then, 5 colonies were taken in 300 µL of Tris-EDTA buffer. The lysis of bacteria was performed by incubation at 100 • C for 10 min, and debris was removed by centrifugation at 15,000× g for 3 min. The DNA-containing supernatant was transferred to a new tube and stored at −20 • C.
For amplification, Ex Taq kit (TaKaRa, Otsu City, Japan) was used, and 25 µL mixtures were as follows: 2.5 µL of 10×Ex Taq buffer, 0.25 µL Ex Taq DNA polymerase (5 U/µL), 2 µL dNTP mixture (2.5 mM), 2 µL template DNA, 0.5 µL of each primer (10 µM), and 17.25 µL ddH 2 O. The reactions were performed on a Veriti 96-Well Thermal Cycler (Applied Biosystems, Carlsbad, CA, USA) under the following conditions: 95 • C for 10 min; 35 cycles at 95 • C for 30 s, annealing temperature for 30 s, and 72 • C for 30 s; and a final extension at 72 • C for 10 min. The B. cereus ATCC 11,778, ATCC 33,019 strains were used as positive control, and Salmonella typhimurium IQCC 10,503 without virulence genes as detected in this study was used as negative control. PCR products were analyzed by electrophoresis in 2% agarose and stained with ethidium bromide (EB). Gel images were captured on a Versadoc Imager (Bio-Rad, Hercules, CA, USA).

Simulated Survive Test
Two strains, both isolated from infant formula samples, were randomly selected and used to investigate whether formula supports B. cereus survival, and simulated samples were prepared. Briefly, two B. cereus isolates were respectively inoculated onto infant formula powder (25 g) at finial concentration at 10 MPN/g, and 225 mL DDW were added and then incubated at room temperature (22 • C) and 37 • C for 24 h. During incubation, simulated samples were tested every hour by the MPN procedure method described above.

Statistical Analysis
The rates of recovery of B. cereus among the domestic brand and the imported brand formula were compared using Student's t-test. Student's t-test was also performed to compare the prevalence of virulence genes.

Prevalence of B. cereus in Infant Formula Samples
A total of 68 infant formula samples, including 51 imported brands samples and 17 domestic brands samples, were analyzed. Among 51 imported brands samples, 27 samples were positive for B. cereus, with the prevalence rate of 52.9% and the mean prevalence level of 3.52 MPN/g (Table 1), while in 17 domestic brands samples, 12 samples were positive, with prevalence rate of 70.6% and the mean prevalence level 1.17 MPN/g. Among the samples positive for B. cereus, thirty-three had prevalence level < 3 MPN/g, four of them measured between 3~10 MPN/g, and only two of samples were measured to a level of 11~50 MPN/g. Despite the prevalence rate of B. cereus in domestic brands, the sample was higher than the imported brands sample; however, the difference was statistically not significant at a 5% level.

Virulence Genes Profiles of B. cereus
Eighty-eight isolates were identified as B. cereus, the multiplex PCR was carried to detect the virulence genes, and most isolates carried nheC gene, followed by nheB and nheA, which were 84.1%, 63.6%, and 54.5%, respectively. The occurrence of three genes of hemolytic enterotoxin complex (HBL) of hblA, hblC, and hblD were 5.68%, 9.09%, and 5.68%, respectively. The frequency of cytK gene was 21.6%. The ces gene encoding emetic toxin was only found in three (3.41%) isolates (Table 2). About 24 virulence gene profiles of B. cereus were detected (Table 3), and the predominant profile was XII (22.73%), in which NHE complex genes were positive and the other virulence genes were negative by PCR test. In profile XII, seventeen strains were isolated from imported formula samples. Additionally, there were 53.4% (47/88) B. cereus isolates with nheA, nheB, and nheC simultaneously, while 4.55% (4/88) strains carried three hbl genes. Three isolates include all three types of both nhe and hbl. However, thirteen strains were without any of the virulence genes. Statistical analysis suggested the prevalence of nheA, nheB, nheC, cytK, bceT, and entFM in domestic brand samples were similar with imported samples (p > 0.05), ranging from 18.6% to 64.3%, while hblA, hblC, and hblD in domestic samples were 22.2%, 27.8%, and 22.2% significantly higher than in imported samples (p < 0.05).

Number of Isolates
Result of samples tested by virulence gene primer pairs: positive +; negative -.

Resistance to Antimicrobials
All of the 88 isolates were susceptible to GM, AMK, and CIP; seven strains were resistant to TET; two strains resistant against CHL; and only one isolate resistant against SXT. The strains had low sensitivity to RIF (Table 4). Furthermore, seven TET resistant isolates were also resistant to RIF. Seven strains with both TET resistance and RIF resistance were detected in this study, and no multiple antibiotic-resistant isolates were detected.

Simulated Survive Test
To investigate whether prepared infant formula supports B. cereus growth, two strains were randomly selected for the simulated survive test. B. cereus MPN levels were tested every hour for 24 h. At room temperature, the MPN reached 139/g and 127/g in two prepared simulated infant formula samples within ten hours. After 24 h incubation, B. cereus levels reached 1000 MPN/g, while increasing the temperature to 37 • C resulted in increased of B. cereus levels to 1000 MPN/g following only 6 h of incubation and 10,000 MPN/g after 24 h incubation.

Discussion
Foodborne diseases caused by foodborne pathogens have become one of the major safety issues that threaten human health. In particular, Salmonella, E. coli, Staphylococcus aureus, and Listeria monocytogenes are considered to be the most important and serious foodborne pathogenic bacteria [17]. Bacillus cereus is a neglected foodborne pathogen because of its strong environmental tolerance, spore-forming ability, and ability to produce toxins. It is often found in raw milk and dairy products.
In China, it is generally considered that imported brands of infant food are better than the domestic. However, the results of the present study proved that infant formulas were frequently found to include B. cereus although there was no significant difference between the imported and domestic infant formula samples (p > 0.05) for the prevalence of B. cereus. The result of the overall prevalence of B. cereus in infant formula was higher than a previously study [18], which reported that 14.08% of samples were contaminated with B. cereus from infant formula samples. The reason may be that B. cereus was examined by the plate count method in their study, while the MPN method was adopted in our study. The MPN method has lower detection limit than the plate count method and frequently was used when the level of Bacillus cereus ≤ 1000 CFU/g.
Infant formulas are prepared with warm water before consumption, and this organism may be activated by a normal process. If the temperature is adequate, one can assume that spores of B. cereus will be activated. The temperature of the water used to prepare the infant formula is advised to be 35 • C~60 • C, and this temperature may facilitate the activation of spores. Therefore, it is not surprising that some cases of B. cereus poisoning were linked to the infant formula [19,20]. Investigation was also conducted to study the possibility that large number of B. cereus could be ingested through the consumption of contaminated infant formula. Although most infant formula products in China have directions for use, it is possible that they may be consumed outside the instructions. Results of room temperature (22 • C) were examined to reflect situations such as those in some villages and towns or small childcare centers, where remaining infant formula food may not remain under refrigeration temperature. At room temperature, in simulated prepared infant formula samples, with an increase in cell number to an MPN of 100/g for 10 h after 24 h incubation, B. cereus numbers reached an MPN of 1000/g. However, increasing the temperature to 37 • C resulted in increases of B. cereus to an MPN of 1000/g after only 6 h of incubation and an MPN of 10,000/g after 24 h incubation. These data raise the concern that contaminated infant formula products available in China pose a potential risk to infants and raise the possibility that these products have already been a cause of illness in the past. Considering the prevalence level of B. cereus in foods that caused illness in the past, mostly ≥100,000 cfu/g, the level in determined in this study was quite low, indicating that the samples were relatively safe in terms of prevalence level at the time of purchase. However, the infant formula food can be safe only if it is consumed as per the guidelines given by the brands. In prepared infant formula that supports growth, the cell number is expected to increase at the time of consumption, especially when unfinished, prepared products are not properly maintained under refrigeration.
Virulence genes are thought to be linked with food spoilage, diarrhea, emesis, and other complications caused by B. cereus [21]. The diarrheal form of the syndrome has been associated mainly with hemolysin BL (Hbl), non-hemolytic enterotoxin (Nhe), and the cytotoxin K (CytK) [22,23]. For the detection virulence genes, we observed 75 isolates carried at least one virulence gene, and of them, nhe genes were the greatest (98.67%). This was consistent with what was reported in previous studies [9], which showed that B. cereus isolated from foods showed more frequent detection of NHE complex genes. HBL enterotoxin complex consists of B, L1, and L2, and its enterotoxic activity appears when all these components of the HBL complex are present [4,24]. Regarding the occurrence of the hemolytic BL genes, hblCDA among food strains isolated in Korea was 81.8% [24], while hblC, hblD, and hblA genes among B.cereus isolated in Chinese pasteurized, full-fat milk occurred with frequencies between 37.0% and 71.7% [25]. In this study, we found these three genes detection rates was 4.55%, lower than those mentioned above. Three isolates harbored all detected virulence genes; according to the labels, we found these strains isolated from two domestic-brand infant formulas. It is suspected that these isolates could be more virulent to humans, and the possibility of severe food poisoning case caused by these virulence genes might exist in China.
Foodborne pathogens that are resistant to a variety of antibiotics have become a major health concern [10]. Aminoglycosides, macrolides, and chloramphenicol antibiotics are usually recommended as the drugs of choice against B. cereus infections. In this study, some isolates were found to be resistant to macrolide and chloramphenicol antibiotics, which may be due to their identical action, acting on the 50S subunit of the bacterial ribonucleoprotein bodies and blocking protein synthesis. Seven multiple antibiotic isolates were also found, a result similar to previous studies, and B. cereus isolated from food was also found to be resistant to multiple antibiotics [9,11]. The isolation of these resistant B.cereus strains from infant formula is worrying. It is suggested that the inappropriate use of antibiotics in veterinary medicine contribute to a potential prevalence of raw materials of infant formula.
It is also found that strains isolated from one sample had different virulence gene profiles and antibiotic resistance. This might indicate the natural environment is reservoir for B. cereus, and food products are easily contaminated, and there is more than one contaminant source of infant formula. More isolated from infant formula, other foods and even some samples from the production environment are needed for subtyping and for infectious resource tracing and control.

Conclusions
The current findings suggest that the prevalence of Bacillus cereus in infant formula remains high and that antibiotic-resistance genes and virulence genes are present. Moreover, our data suggest that B. cereus may be an important pathogen of infant formula food poisoning in Beijing and needs to be controlled in some way. While in China, the regulation of B. cereus is currently insufficient in dairy products, these data could be useful for establishing microbiological safety rules for food, including infant formulas. Due to sampling volume and regional limitations, our study could not cover a larger area, but this study can be used as a basis In the future, the research on Bacillus cereus prevalence can be carried out on a large scale, such as increasing the number of samples and collection sites and conducting more in-depth research on its pathogenicity.