Background

The Bacillus cereus group, also known as B. cereus sensu lato, is a species complex which shows high degree of phenotypic and genotypic similarity. The group classically consists of Gram-positive, rod-shaped, spore-forming aerobic bacteria that are widespread in natural environments. The genetic similarity within the B. cereus group has been widely studied [1–6].

Bacillus cereus s. l. has been found to have significant impact on human health, agriculture, and food processing [5]. B. cereus commonly cause spoilage in food products [7]. Additionally, it is an opportunistic pathogen which can cause two types of food poisoning in humans, characterized by either nausea and vomiting or abdominal pain and diarrhea [8, 9].

The virulence of B. cereus sensu lato is attributed to different factors. Diarrheal disease is associated with the production of enterotoxins such as hemolysin BL (HBL), non-hemolytic enterotoxin (NHE), cytotoxin K, and enterotoxin FM [10–15], whiles the virulence of emetic strains is attributed to the production of a heat stable cereulide, synthesized by a non-ribosomal peptide synthetase encoded by ces genes [16]. The emetic toxin, usually preformed in food, is not inactivated during food processing or gastrointestinal passage because it is highly resistant to heat treatments, extreme pH conditions and protease activities [17–19]. Therefore, ingestion of living B. cereus is not necessary for illness of this type to occur. On the other hand, diarrheal food poisoning is not caused by preformed toxins in food, but by viable vegetative B. cereus cells (not spores) producing enterotoxins in the small intestine, because spores do not produce enterotoxins. Additionally, spores are easily degraded under gastrointestinal conditions by the host’s digestive enzymes [20–22].

Ghanaian traditional milk products including nunu (yoghurt-like) and woagashie (cheese-like) are produced mainly by spontaneous fermentations of raw cow milk [23, 24]. These products are consumed widely throughout Ghana. However, there has been increasing public concern about the safety of consuming the Ghanaian milk products due to the crude methods of milk production, handling and processing which expose the products to possible contamination with various potential foodborne pathogens such as B. cereus. To ensure consumer safety and increase patronage of the Ghanaian traditional dairy products, attempts have been made to characterize the dominant microorganisms and to select starter cultures for the production of quality and safe products [24, 25]. The incidence of Bacillus cereus on dairy farms and in milk products has been reported elsewhere, particularly in Scandinavia, the Netherlands, Australia and Brazil [26–29]. However, there is currently no reported study on the prevalence of Bacillus cereus in dairy farms and milk products in Ghana, and their associated virulence and antibiotic resistance profile. This study therefore sought to assess the prevalence, virulence factor genes and phenotypic antibiotic resistance of Bacillus cereus sensu lato isolated from cattle grazing farms, raw milk, and traditional dairy products in Ghana.

Methods

Sampling

A total of 114 samples made up of 25 soil collected from cattle grazing farm land, 30 raw milk, 28 nunu (yoghurt-like product) and 31 woagashie (West African soft cheese) were used for the isolation of B. cereus. Raw milk, nunu and woagashie samples were purchased from retail markets in Tamale in the Northern region of Ghana. Raw unpasteurized milk and its products sampled were originally sourced from dairy farms which do not rely on supplementary feeding or routine antibiotics use. There was also no record of B. cereus infection among the cattle herds. Soil samples were collected from cattle grazing fields located within 10 mile radius in farming communities in the Northern Region of Ghana. Soil samples were taken from sites at least 500 m apart with a sterilized spatula down to a depth of about 20 cm from the ground surface into sterile stomacher bags and transferred to the laboratory for analysis. Sampling was done between January and October, 2015.

Isolation of and identification of B. cereus s. l.

For the isolation of Bacillus cereus s. l., 25 g of each sample was transferred into 225 ml of sterile phosphate buffered saline (PBS) in a sterile stomacher and homogenized for 2 min using BagMixer stomacher (Inter science, St Nom, France). The homogenate was serially diluted (10-fold) in sterile PBS and 0.1 mL of each dilution was inoculated onto duplicated agar plates containing B. cereus agar base (Oxoid, UK) supplemented with 100 ml/l of Egg Yolk Emulsion (Oxoid,) and 5 ml/l of Polymyxin B Selective Supplement (Oxoid). Plates were incubated at 30 °C for 24 h and observed for growth. Suspected B. cereus colonies with blue appearance (typically mannitol-negative) and lecithinase positive (zone of precipitation around colonies) were selected from each plate and sub-cultured on nutrient agar (Oxoid). Suspected colonies were further identified by phenotypic and biochemical tests [30] including cell shape and motility, hemolysis, production of catalase, oxidase, urease and lecithinase, nitrate reduction, fermentation of D-glucose, maltose, D-xylose, lactose and D-mannitol, and growth at a temperature of 10 °C. Bacillus cereus ATCC 11778 and B. cereus ATCC 14579 were used as reference strains for phenotypic tests.

PCR detection of virulence factor genes in B. cereus s. l.

Prior to DNA extraction, bacterial cultures were grown by streaking on nutrient agar and incubating at 30 °C for 24 h for preparation of template DNA for PCR screening. The bacterial genomic DNA of was extracted using the InstaGene Matrix DNA extraction kit following the instructions of the manufacturer (Bio-Rad, Hercules, CA, USA).

PCR screening was done to detect the presence of 8 enterotoxigenic genes (hblA, hblC, hblD, nheA, nheB, nheC, cytK and entFM), and one emetic gene (ces). The primer pair sequences used for the amplification of virulence factor genes of B. cereus in this study are shown in Table 1. The PCR reaction was carried out as described by Kim et al. [31]. Briefly, the PCR reaction mixture contained 25 ng of template DNA, 0.5 U dreamTaq DNA polymerase (Fermentas GmbH, St. Leon-Rot, Germany), 10 mM Tris-HCl (pH 8.3), 10 mM KCl, 0.2 mM each deoxynucleoside triphosphate (dNTPs), 2.5 mM MgCl2, and 1 μM each primer. Sterile MilliQ water was used for the preparation of the PCR mixture and for all negative control reactions. The reaction was performed in an automatic thermal cycler (Biotron, Göttingen, Germany) under the following optimized cycling program: an initial denaturation step of 3 min at 95 °C; 35 cycles of denaturation at 94 °C for 30 s, annealing at 58 °C for 45 s, extension at 72 °C for 1.5 min; and a final extension at 72 °C for 5 min.

Table 1 Sequences of PCR primers targeting various targeting various virulent factor genes in this study
Full size table

The amplified PCR fragments were analysed by submerged 1.5% agarose gel electrophoresis in 1x buffer (108 g Trisbase/l, 55 g boric acid/l and 40 ml of 0.5 M EDTA, pH 8.0). Following electrophoresis, gels were stained with ethidium bromide, photographed under UV illumination. A reaction mixture without DNA template served as a general control for extraneous nucleic acid contamination. Other controls including sterile MilliQ water and template DNA were used to detect false-positive and false-negative reactions. PCR amplification and electrophoresis experiments were all carried out in triplicates.

Antibiotics susceptibility testing

Resistance/susceptibility to antimicrobials by B. cereus sesu lato isolates were determined in Mueller-Hinton (MH) broth using the broth dilution method recommended by the standard criteria of the Clinical and Laboratory Standards Institute (CLSI) guide [32–35]. The 15 antibiotics including Amoxicillin, Ampicillin, Cefepime, Chloramphenicol, Ciprofloxacin, Clindamycin, Erythromycin, Gentamycin, Oxacillin, Penicillin, Quinupristin/Dalfopristin, Rifampin, Tetracycline, Trimethoprim /sulfumethoxazole and Vancomycin were each diluted in two-fold in the range of 64 to 0.015 mg/L of MH-broth. The final inoculum of B. cereus sesu lato suspension in the broth media were equivalent to approximately 1 × 103 CFU/ml. Growth was carried out at 35 ± 2 °C for 18–20 h incubation period and examined in a microplate reader (OD at 610 nm). The breakpoints against B. cereus in the CLSI guideline M45A2E (2010) and M45-P (2005) were used for all antimicrobial agents except oxacillin and quinupristin/dalfopristin for which the breakpoints for Staphylococcus spp. in the CLSI guideline M 100-S22 (2012) and S24 (2014) were used according to Luna et al. [35].

Results and discussion

Prevalence and phenotypic characteristics of Bacillus cereus s. l.

The prevalence of B. cereus sensu lato in soil from cattle grazing farms, raw milk, nunu, and woagashie are shown in Table 2. Eighteen of 25 (72.0%) soil, 14 of 30 (46.6%) raw milk, 10 of 28 (35.7%) nunu and 12 of 31 (38.7%) woagashi samples were positive for B. cereus sensu lato with mean counts (log10 CFU/g) of 4.2 ± 1.8, 3.3 ± 2.0, 1.8 ± 1.4 and 2.6 ± 1.8 respectively (Table 2). All the isolates showed common phenotypic and biochemical characteristics that are consistent with the identification of Bacillus cereus sensu lato. The isolates were motile rods with peritrichous flagella, and haemolytic with β-haemolysis or lavender-green coloration under heavy growth which is an indication of proteolytic activity. Additionally, they produced catalase, lecithinase, and reduced nitrate. The isolates fermented D-glucose and maltose but were variable in their ability to ferment sucrose and lactose. None of the isolates fermented of D-xlyose and D-mannitol. Production of oxidase and urease was variable while there was no production of indole. All the isolates were able to grow at 10 °C. Thus the phenotypic and biochemical characteristics suggest that the pool of isolates selected did not include B. anthracis which is non-hemolytic, and B. cytotoxicus which has a minimum growth temperature of 20 °C [36]. Previous reports suggests that B. cereus strains isolated from dairy products adapt to environmental culture conditions [37] which might explain the ability of some B. cereus isolated from milk to ferment lactose. In analysing 334 samples of pasteurized milk, TeGiffel et al. [37] found that 40% of the samples were contaminated with B. cereus, out of which 53% of the B. cereus isolates could grow at 7 °C, and 20% fermented lactose which is an uncommon carbon source for B. cereus [37]. Similarly, 17 B. cereus s. l. isolated from milk and dairy products in the current work fermented lactose, whiles no isolate from soil could ferment lactose.

Table 2 Prevalence of B. cereus sensu lato in soil, raw milk and milk products
Full size table

B. cereus is widespread in the environment and shows great ecological diversity, enhancing their ability to contaminate many raw and finished food products including milk and milk products, although usually at low levels of ≤ 103 CFU/g [38–42]. In the present study, raw milk and traditional milk products (nunu and woagashie) all had B. cereus counts of <4 log CFU/g. In general, it is estimated that consumption of food containing B. cereus cells and/or spores between 105 and 108 can cause disease [9, 43]. Therefore, the load of B. cereus sensu lato in milk and milk products in the present study are within acceptable limits for consumption according to the EFSA recommended level of <105 CFU/g at the point of consumption. However, there are reported cases of both emetic and diarrhoeal diseases involving lower levels (below 105 cfu/g) of B. cereus [43]. Therefore the potential for B. cereus infections through the consumption of unpasteurized milk and milk products in Ghana cannot be underestimated. Additionally, Ghana lacks proper foodborne diseases surveillance systems to provide reliable data on the burden of foodborne illnesses involving B. cereus in milk and milk products. Thus a number of illnesses or sporadic outbreaks of B. cereus infections resulting from the consumption of unpasteurized milk and milk products may go unreported. Because B. cereus are generally resident flora of soil and frequently associated with farm environments and the fecal shedding of cattle, there is a higher risk of contamination of milk, and subsequent entry into the dairy food chain where they can cause spoilage and/or diseases. It is therefore important for dairy farmers and processors of traditional milk products to practice high level of good hygienic practices (GHP) and Good manufacturing practices (GMP), as well as implement the use of starter cultures for the production fermented dairy products.

Distribution of virulence factor genes among B. cereus s. l. isolates

PCR based detection of Virulence factors in B. cereus sensu lato targeted genes encoding enterotoxins and emetic toxin. These included genes encoding haemolytic (hblA, hblC, and hblD) and non-haemolytic (nheA, nheB, and nheC) enterotoxin complexes, cytotoxin K (cytK), enterotoxin FM (entFM), and cereulide (ces) as shown in Fig. 1. All primers used produced amplicons of the expected size from their respective target virulence genes with reproducible results in repeated experiments. The distribution of virulence genes among 96 B. cereus s. l. isolates are shown in Table 3. For HNE encoding genes, about 60% (58/96) of B. cereus s. l. isolates were found to harbour simultaneously the nheABC genes, 13% (12/96) harboured only one gene, 19% (18/96) harboured simultaneously two genes, and 8% possessed none of the NHE encoding genes. For HBL encoding genes, 38% (37/96) possessed only one gene, 24% (23/96) possessed simultaneously two genes, 13% (12/96) possessed simultaneously all three bhlACD genes, and 25% (24/96) possessed no HBL encoding gene at all. The prevalence of cytk, entFM, and the emetic gene ces among B. cereus s. l. isolates were 75, 67 and 9% respectively. The emetic gene was only detected in B. cereus s. l. isolated from milk and milk products but not from soil samples.

Fig. 1
figure 1

Representative PCR products detecting various virulence factor genes in B. cereus s. l. isolated from soil and various dairy products. Lane M, 100 bp molecular size DNA marker; lane 1, nheA; lane 2, nheB; lane 3, nheC; lane 4, hblA; lane 5, hblC; lane 6, hblD; lane 7, cytK; lane 8, entFM; lane 9, Ces

Full size image
Table 3 Distribution of enterotoxin and emetic toxin genes in B. cereus senso lato isolated from dairy farm and milk product
Full size table

Virulence factors HBL, NHE and cytotoxin K are primarily responsible for the production of B. cereus s. l. enterotoxins [12, 14, 44]. The isolated B. cereus s. l. commonly possessed cytk (75%) as the most prevalent toxin gene followed by entFM (67%). Similar results of high prevalence of nheABC and entFM gene complexes have previously been reported to be widespread among wild B. cereus isolates from various food and environmental sources [30, 45–48], as well as some reference strains [31]. Similarly, various studies have reported higher prevalence rates, usually between 40 and 70.6% of the HBL gene complex in B. cereus isolated from milk and dairy products [28, 49–51]. The prevalence rate of the emetic toxin gene ces was 9%. They were however detected only in B. cereus s. l. isolated milk and milk product but not from soil samples. Emetic toxin producing genes have previously been detected at different low rates (1.5 to 17.2%) in isolated B. cereus strains isolated from various food sources [42, 48, 52]; the different prevalence rates being attributed to the differences in food property [53–55]. Kim et al., [31], did not successfully generate amplicons for the emetic gene, ces, in both reference and commercial strains of B. cereus. These findings therefore seems to suggest that emetic toxin genes are not highly prevalent or are rare among B. cereus isolates.

Resistance of B. cereus s. l. to antibiotics

Resistance/susceptibility to different antimicrobial agents by B. cereus s. l. are shown in Table 4. Irrespective of the origin (soil, milk or milk products) of the isolates, they were generally resistant to ampicillin (98%), oxacillin (92%), penicillin (100%), amoxicillin (100%), cefepime (100%) and trimethoprim/sulfamethoxazole (80% with 20% intermediate resistant strains). They were however susceptible to other antimicrobials such as Chloramphenicol (99%), Ciprofloxacin (100%), Clindamycin (100%), Erythromycin (92%), Gentamicin (100%), Quinupristin/dalfopristin (100%), Rifampin (100%), Tetracycline (97%) and Vancomycin (100%).

Table 4 Resistance to antimicrobials by B. cereus sensu lato isolated from dairy farms and milk products
Full size table

Because B. cereus have clinical significance, determining their resistance or otherwise to antimicrobial agents is critical for treatment during outbreaks. Previous reports have shown that B. cereus is susceptible to imipenem and vancomycin, and most strains are sensitive to chloramphenicol, aminoglycosides, ciprofloxacin, erythromycin, and gentamicin [56–59]. Some strains of B. cereus are moderately sensitive to clindamycin and tetracycline [58]. In this report, B. cereus s. l. isolated from soil from cattle grazing fields and milk and dairy products were predominantly resistant to β-lactam antibiotics. The abundant production of β-lactamases by bacteria including Bacillus species is a common cause of antibiotic resistance in bacteria [39, 60]. Wild-type genomes of many bacteria, including Bacillus species have been found to possess genes encoding the production of β-Lactamase. However, these chromosomal β-lactamases do not generally provide effective antibiotic resistance in wild-type bacilli, despite evidence that the genes are not completely silenced [61, 62].

Conclusions

Bacillus cereus sensu lato is prevalent in soil from cattle grazing fields, raw milk and traditional dairy products, but load in milk and traditional dairy products are at levels considered to be safe for consumption. However, various enterotoxin genes associated with the virulence of B. cereus are widespread among isolates. Additionally, the B. cereus s. l. isolates were generally resistant to β-lactam antibiotics but susceptible to other antibiotics. There is therefore the need to observe good hygienic and manufacturing practice by milk producers and traditional dairy processors to prevent contamination and subsequent potential disease outbreak by B. cereus.