P. Schmitt, UMR A408 INRA/Université d‘Avignon ‘Sécurité et Qualité des Produits d'Origine Végétale’, INRA Domaine St Paul, 84914 Avignon Cedex 9, France (e-mail: email@example.com).
Aims: To determine the fate of Bacillus cereus spores or vegetative cells in simulated gastric medium.
Methods and Results: The effects of acidity on the survival of B. cereus in a medium simulating human stomach content was followed on spores at pH 1·0–5·2, and on vegetative cells at pH 2·5–5·7. Gastric media (GM) were prepared by mixing equal volumes of a gastric electrolyte solution with J broth (JB), half-skim milk, pea soup and chicken. At pH 1·0 and 1·4, the number of spores slightly decreased in GM-JB and GM-pea soup and remained stable in GM-milk and GM-chicken. A rapid marked decrease (always higher than 2·0 log CFU ml−1 in 2 h) in vegetative cell counts was observed at pH below 4·2, 4·0, 3·6 and 3·5 in GM-chicken, GM-JB, GM-milk and GM-pea soup, respectively. Between pH 5·0 and 5·3, B. cereus growth was observed in GM-JB (1·2 log CFU ml−1 increase after 4 h) and in GM-pea soup (1·8 log CFU ml−1 increase after 4 h).
Conclusions: Bacillus cereus spores are very much more resistant to gastric acidity than vegetative cells. This resistance strongly depends on the type of food present in the GM.
Significance and Impact of the Study: Our results suggest that the probability that viable B. cereus cells enter the small intestine, where they can cause diarrhoea, strongly depends on the form of the ingested cells (spores or vegetative cells), on what food they are ingested with, and on the level of stomach acidity.
The pathogenic bacteria Bacillus cereus is a major cause of food-borne poisoning in many countries, representing for instance 1% of the food-borne outbreaks in Canada between 1975 and 1984, 22% in the Netherlands between 1977 and 1982, 5% in Sweden between 1992 and 1997, and 3·8% in France in 1999–2000 (Kramer and Gilbert 1989, Todd 1992, Todd 1996, Bean et al. 1997, Granum and Baird-Parker 2000, Lindqvist et al. 2000, Haeghebaert et al. 2002). Most cases are because of the production of a toxin, either an emetic toxin or any of several types of diarrhoeal toxin (Lund and Granum 1997). The emetic syndrome is because of emetic toxin production in foods, while the diarrhoeal syndrome results from toxin production in the small intestine (Kramer and Gilbert 1989, Granum 1994). Large numbers of cells (>105 CFU g−1) probably have to be ingested in foods to cause disease (Kramer and Gilbert 1989), because the spores or vegetative cells have to survive stomach acidity and then multiply in the small intestine. The aim of this work was to describe the fate of B. cereus spores and cells in acid media simulating the human stomach. The pH values tested were in the range of those reported in clinical studies of the acidity of the human stomach and allowed for the variation between food ingestion in an empty stomach (low pH) and in a full one (high pH) (Armand et al. 1994).
Materials and methods
Bacillus cereus strain F4430/73 (B4ac), isolated from pea soup (Spira and Goepfert 1975), strain F4433/73, isolated from meat loaf (both gifts of Professor P.E. Granum, The Norwegian School of Veterinary Science, Oslo, Norway), and strain INRAAV TZ415, isolated from cooked chilled food containing vegetables at the Institut National de la Recherche Agronomique (Avignon, France) (Choma et al. 2000) were used. Stock cultures of spores or vegetative cells were kept at −20°C in water and in a 20% glycerol solution (v/v), respectively.
Working cultures were obtained in J broth (JB) (5 g l−1 peptone, 15 g l−1 yeast extract, 3 g l−1 K2HPO4, 2 g l−1 glucose, adjusted to pH 7·2). This medium is recommended for the culture of Bacillus species (Claus and Berkeley 1986) and has been successfully used by us for B. cereus culture. Glucose was added after autoclaving (20 min, 121°C), using a filter (0·22 μm) sterilized glucose solution. J agar (JA) containing JB plus agar (15 g l−1) was used for plate count. Spores were produced on modified fortified nutrient agar (FNA) (Fernandez et al. 1999).
Inocula of vegetative cells and spores were made. Vegetative cell inocula were prepared from stock culture purified on JA. One purified colony was transferred to 250 ml flasks containing 50 ml of JB or food media and grown at 34°C. Unless otherwise specified, B. cereus cultures were grown for 18 h with vigorous shaking (130 rev min−1) to attain the stationary growth phase.
To produce spores, purified B. cereus stock cultures were grown overnight in JB at 30°C to attain the stationary growth phase. A 0·5-ml aliquot was then spread on modified FNA in 250 ml Roux bottles containing 80 ml modified FNA, and incubated at 30°C for 7 days. Sporulation was checked daily by microscopic examination, and spores were harvested when at least 90% of the cells had produced spores. The agar surface was then flooded with 15 ml of sterile cold demineralized water. The spore suspension was centrifuged, re-suspended in 8 ml of cold distilled water and stored at −20°C until use. Before each experiment spore suspensions were heat-treated (10 min, 72°C). Concentrations of vegetative cells and spores were determined by plating serial dilutions on JA.
Gastric media (GM) were made up using one volume of a sterile (autoclaving 121°C for 20 min) gastric electrolyte solution (4·8 g l−1 NaCl, 1·56 g l−1 NaHCO3, 2·2 g l−1 KCl, 0·22 g l−1 CaCl2) (Gänzle et al. 1999) and 1 volume of JB or each food media. The food media were: commercial UHT sterilized half-skim milk, split pea purée, and chicken meat medium. Split pea purée was obtained by mixing split peas and demineralized water (1/10, w/v), and then autoclaving at 121°C for 20 min. Chicken meat medium was obtained by boiling chicken breast in demineralized water (1/2, w/v), and then blending and filtering through a 0·7-mm nylon mesh. Chicken media aliquots were stored at −20°C until use. GM-chicken was sterilized by autoclaving at 121°C for 20 min. The four GM were named GM-JB, GM-chicken, GM-milk and GM-pea.
After sterilization each GM was supplemented with 500 U l−1 of a filter (0·22-μm) sterile pepsin solution in water (P6887; Sigma-Aldrich, Saint-Quentin-Fallavier, France).
Determination of acid tolerance in GM
The determination of acid tolerance was carried out in GM at target pH values ranging between 1·0 and 5·0. The pH of the GM was adjusted to the desired value with sterile 6 mol l−1 HCl. The pH of each GM was measured at the beginning and end of each experiment, and at regular time intervals for some conditions, with a Schott-Geräte 6820 electrode calibrated using freshly prepared buffers at pH 4·0 and 7·0, and a Schott-Geräte CG825 pH meter (Hofheim, Germany).
Bacillus cereus spores or vegetative cells were added to GM to obtain target initial populations of 106–107Bacillus cereus CFU ml−1. GM were incubated for 6 h at 37°C with shaking (130 rev min−1) to simulate human stomach conditions.
Samples were serially diluted in 0·2 mol l−1 sodium phosphate buffer at pH 7·0 and surface spread on JA duplicate plates using a spiral plate apparatus (Spiral système; Intersciences, Saint-Nom-la-Bretèche, France). We checked that this buffer concentration neutralized the acidity of all the GM at every pH tested. For low bacterial concentrations, samples were manually surface spread on JA. Cell concentrations were expressed as colony forming units per millilitre (CFU ml−1). The limit of detection was taken as one colony on the lowest dilution plate, i.e. 10 CFU ml−1. We have been able to indirectly estimate the reproducibility of our results. Nine curves were performed at an interval of 0·1 pH unit. The behaviour of the bacterium was always the same in all these experiments performed in close conditions, e.g. either slow growth, slow decrease on 6 h or rapid decrease on 2 h.
Spores of B. cereus strain F4430/73 inoculated in GM were tolerant to acidity at pH values between 1·0 and 5·2 (Fig. 1). Decreases in spore counts were lower than 1·5 B. cereus log CFU ml−1 after 6 h of incubation at the lowest tested pH values (between 1·0 and 1·5). No decrease at any pH was observed in GM-milk or GM-chicken. In GM-JB at pH 5·0, outgrowth was observed after 2 h of incubation. The same experiments were carried out with vegetative cells (Fig. 2). The vegetative cells were highly susceptible to acidity, showing a decrease always higher than 3·0 log CFU ml−1 in 2 h at pH lower than 3·5 and a decrease higher than 4·0 log CFU ml−1 in 0·5 h in many instances. The survival of vegetative cells was strongly affected by the nature of the GM. Comparison of the fate of B. cereus at pH values ranging between 4·0 and 4·3 in all the GM, and at pH between 3·5 and 3·6 in GM-JB, GM-milk and GM-pea showed a higher survival in GM-milk and GM-pea than in GM-JB and GM-chicken. At pH values between 5·0 and 5·3, some increases in the B. cereus populations were observed, reaching for instance a 1·2-log CFU ml−1 or a 1·8-log CFU ml−1 increase in GM-JB or GM-pea after 4 h of incubation (Fig. 2).
In addition, two-phase death curves were observed in GM-milk, GM-pea and GM-chicken. Rapid death in GM-milk at pH 3·6 or in GM-chicken at pH 3·2 for instance was followed after 1 h by a slow decrease in viable counts. This suggests the presence of either a small number of highly resistant cells, or of spores in the precultures. To reduce the number of spores that might be present in the inoculum, cells in exponential growth phase (6-h-old culture) were used. In this case, the death curve was linear on a 6-log CFU ml−1 scale (Fig. 3).
Other B. cereus strains were tested for their acid tolerance in GM-JB at pH 4·5 (Fig. 4). Decreases in the number of vegetative cells of strains F4430/73, F4433/73 and TZ4145 were very similar and ranged between 1·5 and 2·1 log CFU ml−1 after 6 h. Changes in the number of spores of strains F4430/73 and TZ4145 in GM-JB at pH 2·5 were also very similar (data not shown).
Bacillus cereus spores were highly resistant to acidity in a range of media simulating the conditions in the human stomach after food ingestion. This result was expected from general knowledge of bacterial spore resistance (Nicholson et al. 2000).
Survival of B. cereus cells was higher in GM-milk than in the other GM. This protective effect of a dairy product against inactivation by low pH was previously observed on Salmonella in cheese (D'Aoust 1985) and might be due to lipids, proteins or particles of the food. Conway et al. (1987) have hypothesized that micro-organisms are trapped in hydrophobic lipid moieties and may consequently survive acid conditions. This effect may be responsible for the strong protective effect on B. cereus that we observed in GM-milk, the fattest GM as estimated from nutrition reference tables (Souci et al. 2000). Milk proteins may also explain the gastric tolerance of some strains of lactic acid bacteria (Charteris et al. 1998). Waterman and Small (1998) showed the survival on beef particles of acid-sensitive bacteria such as Salmonella typhimurium, Shigella flexneri and Escherichia coli and suggested that ground beef could raise the pH of the acidified medium in the microenvironment of the bacteria. Growth of, and toxin production by, Clostridium botulinum at pH 4·3 required proteins from skimmed milk or soya (Smelt et al. 1982). However, in the present study, GM-chicken did not protect B. cereus against acidity. The protective effects of food matrices on bacteria often observed in laboratory conditions are also observed in vivo: in rats, a 100-fold higher Lactococcus lactis survival rate in the stomach was reported with bacteria ingested in feed (Drouault et al. 1999).
The probability of becoming ill after ingestion of B. cereus cells or spores depends on the number of viable B. cereus surviving gastric transit and then entering the small intestine. This work demonstrates that this number strongly depends on the form of the ingested cells (spores or vegetative cells), on the food they are ingested with and on stomach acidity. Stomach acidity is subject to variations with time (in particular that elapsed after eating) and age, and to variability among individuals. Situations (i.e. pH ≥4·5) in which B. cereus cells are not or only weakly affected by stomach transit to the small intestine are frequent: ingestion at the end of a copious meal, when stomach pH is at its maximum (Dressman et al. 1990), or ingestion by elderly people or people suffering from achlorhydria (the absence of gastric secretion) (Russel et al. 1993). In extreme situations with stomach pH ≥5·0, B. cereus can even grow during gastric transit. In addition, it has previously been reported that pre-exposition of cells to nonlethal acid pH induced an acid tolerance response (ATR) in B. cereus (Browne and Dowds 2002, Jobin et al. 2002). Such an inducible ATR is an important component of bacterial survival. Reliable quantitative risk assessment, and establishing minimal infective dose–response or dose–response curves for enterotoxigenic B. cereus must thus allow for these multiple factors affecting survival during gastric transit.
This work was supported by a grant from ‘Ministère de la Recherche et Ministère de l'Agriculture et de la Pêche’ under contract ‘Aliment-Qualité-Sécurité 2000 R0013’‘Caractérisation de la virulence de Bacillus cereus’. We thank Claire Dargaignaratz and Ouafa Rebaï for technical assistance.