Assessment of a New Selective Chromogenic Bacillus cereus Group Plating Medium and Use of Enterobacterial Autoinducer of Growth for Cultural Identification of Bacillus Species

Abstract

A new chromogenic Bacillus cereus group plating medium permits differentiation of pathogenic Bacillus species by colony morphology and color. Probiotic B. cereus mutants were distinguished from wild-type strains by their susceptibilities to penicillin G or cefazolin. The enterobacterial autoinducer increased the sensitivity and the speed of enrichment of B. cereus and B. anthracis spores in serum-supplemented minimal salts medium (based on the standard American Petroleum Institute medium) and buffered peptone water.

Bacillus anthracis, the etiological agent of anthrax, belongs to the B. cereus group of aerobic, saprophytic, spore-forming gram-positive rods. Virulence depends on two plasmids: pXO1 encodes protective antigen, edema factor, and lethal factor, which comprise anthrax toxin (20), and pXO2 codes for a protective capsule (28). The B. cereus group also includes other pathogenic species, including B. thuringiensis, which synthesizes crystalline protein inclusions that are toxic to insects (25), as well as B. cereus, B. mycoides, B. pseudomycoides, and B. weihenstephanensis (10, 12, 17, 27), which are recognized as human food-borne pathogens. Symptoms of cramp-like abdominal pains, watery diarrhea, and severe acute nausea and vomiting result from enterotoxin production by these species and also from secretion of an emetic toxin for B. cereus (1). It is now generally accepted that B. anthracis, B. cereus, B. mycoides, B. pseudomycoides, B. thuringiensis, and B. weihenstephanensis should be regarded as a single species due to their close genomic similarity (5, 10, 13).

B. anthracis has long been regarded as a potential bioterrorist weapon; in a scenario in which deliberate malicious dispersal of spores might be threatened, therefore, techniques for rapid and sensitive environmental testing for B. anthracis spores are urgently needed. Furthermore, the incidence of B. cereus food poisoning is increasing in industrialized countries, and so improved methods for discrimination among members of the group are required for early diagnosis. It is also necessary to be able to differentiate attenuated B. cereus mutants used as probiotics (e.g., strains Paciflor and Toyocerin [19; H. Mietke, W. Beer, W. Voigt, B. Zucker, G. Schabert, L, Restaino, and R. Reissbrodt, Abstr. 101st Gen. Meet. Am. Soc. Microbiol., poster P-58, 2001]) from toxigenic wild-type strains. This paper describes the assessment of the recently launched chromogenic Bacillus cereus group plating medium (Biosynth AG, Staad, Switzerland), which has been claimed to be able to permit cultural identification of pathogenic species of the B. cereus group (3, 12, 22). We also propose a method for the enrichment of samples with low-level contamination with Bacillus spores that involves the enterobacterial autoinducer of growth (AI) described by Lyte and coworkers (7, 8, 15).

Bacillus cereus group plating medium (formerly BCM Bacillus cereus/Bacillus thuringiensis plating medium [Biosynth AG]; also commercially available as Cereus-Ident-Agar [Heipha, Eppelheim, Germany]) contains 5-bromo-4-chloro-3-indoxyl-myoinositol-1-phosphate, which changes from colorless to turquoise upon enzymatic cleavage. B. cereus, B. mycoides, B. thuringiensis, and B. weihenstephanensis secrete phosphatidylinositol phospholipase C and so grow as turquoise colonies with species-specific morphologies (Table 1). Probiotic strain Toyocerin (19; Mietke et al., Abstr. 101st Gen. Meet. Am. Soc. Microbiol.) was distinguished by its susceptibilities to penicillin G and cefazolin (Fig. 1a); probiotic strain Paciflor (19; Mietke et al., Abstr. 101st Gen. Meet. Am. Soc. Microbiol.) behaved identically (data not shown). These data confirm the results of an extensive study with more than 500 B. cereus isolates from various feeds, foods, spices, etc., as well as American Type Culture Collection (ATCC) strains (24). All except 2 of 427 wild-type isolates tested grew as typical turquoise colonies that were resistant to these antibiotics, while all 41 Paciflor strains and 47 of 48 Toyocerin strains tested on this medium grew as turquoise colonies that were sensitive to penicillin G and cefazolin. Although B. anthracis was also sensitive to penicillin G, it was distinguishable from B. cereus wild-type and probiotic strains by growth as white colonies on Bacillus cereus group plating medium. The gene encoding phosphatidylinositol phospholipase C appears to be present in the B. anthracis genome (14), but either the gene is not expressed or the gene product is not active under the growth conditions of our assay. Other Bacillus species, as well as gram-negative and other gram-positive species, were partially or completely inhibited by this medium (Table 1).

TABLE 1.

Selective and differential properties of chromogenic Bacillus cereus group plating medium (Biosynth AG)

Speciesa	No. of strains tested	Colony characteristic(s)b
B. anthracis	16	Dull white; diam, 2-5 mm
B. cereus wild-type strains	19	Dull turquoise; with or without turquoise halos; diam, 2-5 mm
B. cereus probiotic strains Paciflor and Toyocerin	10	Turquoise; diam, 2-4 mm
B. mycoides	1	Turquoise; irregular mycoid growth
B. thuringiensis	4	Dull turquoise, with or without halos; diam, 2-5 mm
B. weihenstephanensis	3	Turquoise; diam, 1-2 mm
B. circulans, B. licheniformis	1 each	White; diam, 1 mm
B. brevis, B. lentus, B. megaterium, B. pumilus, B. sphaericus, B. subtilis	1 each	No growth
Listeria monocytogenes	4	Turquoise; diam, 1 mm
Listeria innocua, Listeria seeligeri, Listeria welshimeri	1 each	White; diam, 1 mm
Enterococcus spp.	12	No growth or pinpoint colonies
Staphylococcus aureus	6	No growth
Staphylococcus epidermidis, Staphylococcus saprophyticus, Micrococcus spp.	1 each	No growth

^aFrom the culture collections of the Robert Koch-Institut, Wernigerode, Germany; the Bundesforschungsanstalt für Viruskrankheiten der Tiere, Jena, Germany; and the Sächsische Landesanstalt für Lanwirtschaft, Leipzig, Germany.

^bPlates were streaked with fresh cultures of test bacteria (approximately 10⁵ cells/loop) and incubated overnight at 36 ± 1°C for 24 h.

FIG. 1.

(a) Antibiotic sensitivity of probiotic B. cereus strain Toyocerin. Chromogenic Bacillus cereus group plating medium was streaked with a fresh culture (approximately 10⁵ cells/loop), filter paper disks impregnated with 10 μg of penicillin G (P) or 30 μg of cefazolin (C) were placed on the agar, and the plate was incubated overnight at 36 ± 1°C. Inhibition zones of >20 mm indicated antibiotic sensitivity. (b) Promotion of B. cereus ATCC 1778 spore growth on egg-white agar medium by AI. Filter paper disks impregnated with (clockwise, from the top) 50, 20, 10, and 2 U of AI per ml or with 0.5 μg of 2,3-dihydroxybenzoyl-serine (DHBS; center) were placed on the agar, and the plate was incubated overnight at 36 ± 1°C.

While the chromogenic plating medium provides selectivity for pathogenic members of the B. cereus group, the sensitivity and speed of detection depend upon the prior enrichment of samples that may contain very few microorganisms. We previously demonstrated that AI prepared from Escherichia coli cultures in a physiologically relevant nutrient-poor, serum-supplemented minimal salts medium based on the Standard American Petroleum Institute (SAPI) medium and containing 50 μM l-norepinephrine (NE) (7, 8, 15) resuscitated highly stressed cells of Salmonella enterica and enterohemorrhagic E. coli (23). In the present study we show that AI, now commercially available as Bacxell (BioNutrix LLC, Minneapolis Minn.), can also enhance the growth of B. cereus and B. anthracis spores. Initial experiments were performed in a Bioscreen C apparatus (Labsystems, Helsinki, Finland) with serum-SAPI medium (Fig. 2). Spores of B. cereus strain ATCC 1778 prepared by standard methods (4) did not grow in this medium, but addition of AI stimulated growth in a dose-dependent manner. Interestingly, supplementation with NE further enhanced the effectiveness of AI, although NE alone had no effect (Fig. 2). Similar data were obtained with two other wild-type isolates of B. cereus (data not shown). The nature of this synergy is unknown, but it may relate to the ability of NE to facilitate the removal of iron from serum transferrin for uptake by bacteria (8). Growth stimulation by AI is also seen in a simple bioassay on nutrient-rich agar medium containing egg white from fresh hen's eggs (Fig. 1b), a medium that was developed to assay the effects of exogenous siderophores on siderophore-proficient bacteria (11). AI allows B. cereus to overcome the growth-inhibitory effects of iron limitation, possibly due to the presence of the enterobactin complex in AI preparations, since the enterobactin breakdown product 2,3-dihydroxybenzoyl-serine promotes the growth of B. cereus on egg-white agar (Fig. 1b). The extent to which other undefined components of AI preparations may contribute to this effect is under investigation in our laboratories. In addition, we are testing whether AI may function by modulating iron acquisition systems, such as the reported B. anthracis homologs of the Staphylococcus aureus IsdG and IsdI proteins that are proposed to be involved in the facilitation of iron uptake iron from heme (26). Another possibility is that AI interacts with an endogenous quorum-sensing system like those recently described in other gram-positive organisms (9).

FIG. 2.

Bioscreen C analysis of the effects of AI on B. cereus ATCC 1778 spores. Approximately 10 spores of B. cereus strain ATCC 1778 were inoculated into 0.29 ml of serum-SAPI medium in each well of a 100-well microtiter plate. Growth at 36 ± 1°C was monitored by measurement of the OD₆₀₀ over 24 h (t, time); each line represents the average of data from 20 replicate wells supplemented as follows: no supplement (line A), 50 μM NE (line B), 50 U of AI per ml (line C), 100 U of AI per ml (line D), and 50 μM NE plus 100 U of AI per ml (line E).

It was not possible to use the Bioscreen C for the work with B. anthracis because of the need for category 3 containment facilities. We therefore used our detailed analysis of B. cereus strains as a model for the effect of AI on Bacillus species to design experiments that could be performed under containment conditions to determine whether B. anthracis strains behave similarly. Thus, we were able to demonstrate that AI had similar effects with batch cultures in which growth was monitored by measurement of the optical density at 600 nm (OD₆₀₀) directly in the culture tubes. Figure 3a shows that spores of three representative B. anthracis strains grew very poorly in serum-SAPI medium, but as in the case of B. cereus, supplementation with AI stimulated culture growth significantly. Indeed, although data obtained by the different methods are not absolutely comparable, it appears that the growth-stimulatory effect of AI on B. anthracis strains was greater, with a shorter lag phase, than that on B. cereus. The pathogenic Bacillus species can therefore be added to the list of clinical strains for which AI stimulates growth in the stressful environment of serum-SAPI medium (7, 18). It should be noted, however, that growth stimulation by AI is not confined to nutrient-poor, iron-restricted environments. Addition of AI to the nutrient-rich medium buffered peptone water (BPW) inoculated with approximately 20 B. anthracis spores/ml resulted in a marked enhancement in culture density that was detectable after as little as 4 h of incubation (Fig. 3b). Supplementation with NE did not result in any significant enhancement in growth, and there was no synergy between NE and AI in BPW (data not shown). Another supplement known to resuscitate highly stressed enteric bacteria, the commercial antioxidant Oxyrase (Oxyrase Inc., Mansfield, Ohio) (23) also enhanced the growth of B. anthracis spores (Fig. 3b). It may be that AI, like Oxyrase, prevents the generation of damaging free radicals at a time when stressed bacteria are in a particularly vulnerable state (6).

FIG. 3.

Batch culture analysis of the effects of AI on B. anthracis spores. (a) Five-milliliter batches of serum-SAPI medium containing 50 μM NE with (+) or without (−) supplementation with 100 U of AI per ml were inoculated with 68 spores of B. anthracis strain Spike(pOX1) (A), 44 spores of B. anthracis strain 340(pOX1 pOX2) (B), or 58 spores of B. anthracis strain 369(pOX1 pOX2) (C). Cultures were incubated at 36 ± 1°C with shaking (at 140 rpm longitudinally) for 24 h; growth was determined by measurement of the OD₆₀₀. (b) Five-milliliter batches of unsupplemented BPW (A) or BPW supplemented with 0.2 U of Oxyrase per ml (B) or with 100 U of AI per ml (C) were inoculated with approximately 100 spores of B. anthracis strain 369 and incubated at 36 ± 1°C with shaking. OD₆₀₀ measurements were taken at 4 and 6.5 h.

We propose the use of AI-supplemented enrichment cultures in combination with Bacillus cereus group plating medium for the detection of pathogenic Bacillus species in environmental, food, or clinical samples. Not only may AI resuscitate highly stressed bacteria that would not otherwise grow in culture, but AI also enhances enrichment culture growth, thus improving the sensitivity and speed of detection. To demonstrate the effectiveness of AI, 4-h BPW cultures from the experiment whose results are shown in Fig. 3b were streaked onto Bacillus cereus group plating medium, and a lysate of B. anthracis-specific γ phage (2) was spotted at several positions on each inoculated plate. Following incubation at 36 ± 1°C for approximately 4 h, no growth was observed on plates inoculated with the unsupplemented BPW culture, but pinpoint colonies were visible on plates inoculated with the AI-supplemented culture, among which lytic zones at sites of inoculation with γ phage indicated the presence of B. anthracis. This method therefore provides a reliable early warning of the presence of anthrax spores in a test sample that can be confirmed following further incubation by the presence of white colonies with lytic zones. Cultures or colonies may be further characterized by species-specific PCR-based tests (16) or for the detection of B. anthracis virulence plasmids (21).