Bacillus thuringiensis - Applied and Environmental Microbiology

7 downloads 0 Views 78KB Size Report
Jan 6, 2006 - 24/7. Pineapples. Honduras. 10. 17029. Bornholm. 6/9. Sugar peas. Zambia. 200. 16629. Esbjerg. 28/8. Cherry tomatoes. The Netherlands. 20.
APPLIED AND ENVIRONMENTAL MICROBIOLOGY, May 2006, p. 3435–3440 0099-2240/06/$08.00⫹0 doi:10.1128/AEM.72.5.3435–3440.2006 Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Vol. 72, No. 5

Occurrence of Natural Bacillus thuringiensis Contaminants and Residues of Bacillus thuringiensis-Based Insecticides on Fresh Fruits and Vegetables Kristine Frederiksen,1 Hanne Rosenquist,1 Kirsten Jørgensen,2 and Andrea Wilcks1* Department of Microbiology and Risk Assessment, Danish Institute for Food and Veterinary Research, 2860 Søborg, Denmark,1 and Department of Veterinary Pathobiology, The Royal Veterinary and Agricultural University, 1870 Frederiksberg C, Denmark2 Received 6 January 2006/Accepted 8 March 2006

A total of 128 Bacillus cereus-like strains isolated from fresh fruits and vegetables for sale in retail shops in Denmark were characterized. Of these strains, 39% (50/128) were classified as Bacillus thuringiensis on the basis of their content of cry genes determined by PCR or crystal proteins visualized by microscopy. Random amplified polymorphic DNA analysis and plasmid profiling indicated that 23 of the 50 B. thuringiensis strains were of the same subtype as B. thuringiensis strains used as commercial bioinsecticides. Fourteen isolates were indistinguishable from B. thuringiensis subsp. kurstaki HD1 present in the products Dipel, Biobit, and Foray, and nine isolates grouped with B. thuringiensis subsp. aizawai present in Turex. The commercial strains were primarily isolated from samples of tomatoes, cucumbers, and peppers. A multiplex PCR method was developed to simultaneously detect all three genes in the enterotoxin hemolysin BL (HBL) and the nonhemolytic enterotoxin (NHE), respectively. This revealed that the frequency of these enterotoxin genes was higher among the strains indistinguishable from the commercial strains than among the other B. thuringiensis and B. cereus-like strains isolated from fruits and vegetables. The same was seen for a third enterotoxin, CytK. In conclusion, the present study strongly indicates that residues of B. thuringiensis-based insecticides can be found on fresh fruits and vegetables and that these are potentially enterotoxigenic. The gram-positive spore-forming bacterium Bacillus thuringiensis is ubiquitous in the environment and is closely related to the food-borne pathogen Bacillus cereus. The only difference between the two species is the ability of B. thuringiensis to produce parasporal crystalline inclusions, the so-called crystal proteins (Cry proteins) or ␦-endotoxins, which are plasmid encoded. These toxins have highly specific activity against certain insects (6, 23), especially within the orders Lepidoptera, Diptera, and Coleoptera (11), and B. thuringiensis is therefore of commercial interest. Plant protection products based on selected strains of B. thuringiensis are used worldwide in, e.g., the production of fruits and vegetables in greenhouses and in the field. The high specificity of the Cry proteins against insects is mainly due to specific receptors in the insect gut, which are not present in the mammalian gut. These toxins are therefore considered harmless to humans. However, B. thuringiensis strains are capable of producing a variety of other toxins and virulence factors that can affect humans, including the same diarrhea-causing enterotoxins as produced by B. cereus (12, 13, 20). The two best-characterized enterotoxins are hemolysin BL (HBL) and nonhemolytic enterotoxin (NHE), which are both three-component toxins requiring expression of all three genes for full virulence (17, 19). A third enterotoxin, cytotoxin K (CytK), is a single-component toxin once reported to be involved in a severe food poisoning case that caused the deaths of three individuals (18). However, the role of CytK is not yet

fully understood, since recent studies have indicated the existence of two variants of CytK (4, 10), the one (CytK2) having a lower toxicity than CytK1, which caused the severe food poisoning case. Also, commercially used B. thuringiensis strains have been shown to harbor genes for HBL, NHE, and CytK, and the expression of components of the three-component enterotoxins HBL and NHE has also been established (8, 16, 22). Since methods for identification of B. cereus-like bacteria in food and clinical settings do not distinguish between B. cereus and B. thuringiensis, the presence of B. thuringiensis in food and the role of this organism in food poisoning are not well described. The bacterium B. thuringiensis has only in one case been associated with food poisoning (15), although the bacterium has the same genetic potential for producing enterotoxins as B. cereus. A recent study in our laboratory has shown that more than half of the isolated B. cereus-like strains from 40 ready-to-eat products at a Danish retail market in fact belonged to B. thuringiensis (22), indicating that this species could be the causative agent of some of the food-borne outbreaks earlier ascribed to B. cereus. Counts of B. thuringiensis above 104 CFU g⫺1 were found in some fruits and vegetables (22). Whether the high counts originated from natural contaminants or residues of B. thuringiensis insecticides was not shown. Nevertheless, we hypothesize that residues of B. thuringiensis insecticides account for some of the B. thuringiensis strains present on fresh fruits and vegetables. Fresh fruits and vegetables are normally not associated with B. cereus-related diarrhea. However, used as ingredients, these products may contaminate complex food dishes, e.g., starchy dishes, in which there are good conditions for growth, especially if the final dishes are improperly cooled after heat treatment. The main purpose of this study was to determine the occur-

* Corresponding author. Mailing address: Department of Microbiological Food Safety, Danish Institute for Food and Veterinary Research, Mørkhøj Bygade 19, DK-2860 Søborg, Denmark. Phone: 45 7234 7185. Fax: 45 7234 7698. E-mail: [email protected]. 3435

3436

FREDERIKSEN ET AL.

rence of B. thuringiensis on fresh fruits and vegetables for sale in Danish retail shops, including natural contaminants, as well as residues of B. thuringiensis-based insecticides. Another aim was to compare the enterotoxigenic potential of these two groups of B. thuringiensis with that of other B. cereus-like organisms isolated from fresh fruits and vegetables.

APPL. ENVIRON. MICROBIOL. expected number of commercial strains in the samples from tomatoes is 8 ⫻ 0.18 ⫽ 1.4 ⫾ the square root of 1.4, that expected in the samples from cucumbers is 4 ⫻ 0.18 ⫽ 0.7 ⫾ the square root of 0.7, and that expected in the samples from peppers is 7 ⫻ 0.18 ⫽ 1.3 ⫾ the square root of 1.3. The observed numbers in the tomato, cucumber, and pepper samples were 6, 4, and 5, respectively. These numbers differ from the expected average by more than 3 standard deviations, meaning that there is a 99% probability (P ⬎ 0.99) that the prevalence of commercial strains among the three products was different from the overall distribution of commercial strains in the 128 food samples.

MATERIALS AND METHODS Strains. A total of 128 B. cereus-like strains were randomly selected from the positive samples of 991 fresh fruit and vegetable products collected and enumerated for content of B. cereus-like bacteria in a previous study (22). The 128 isolates originated from lettuce (n ⫽ 32), tomatoes (n ⫽ 8), cucumbers (n ⫽ 4), peppers (n ⫽ 7), berries (n ⫽ 17), grapes (n ⫽ 6), herbs (n ⫽ 15), apples (n ⫽ 10), root vegetables (n ⫽ 19), and other products (n ⫽ 10). These food samples had been randomly collected nationwide from local retail establishments over a period of 1 year and differed with regard to sampling place and time, origin, and/or product type. The following strains of B. thuringiensis from six commercial products were also included in the study: B. thuringiensis subsp. kurstaki present in Dipel, Biobit, and Foray; a transconjugant between B. thuringiensis subsp. aizawai and subsp. kurstaki from Turex; and B. thuringiensis subsp. israelensis from Vectobac and Bactimos. Isolation of DNA. Genomic DNA was extracted by boiling a bacterial colony for 10 min in TE buffer (10 mM Tris, 1 mM EDTA, pH 8.0). After centrifugation, the supernatant was used for PCR and RAPD (random amplified polymorphic DNA) analysis. For isolation of plasmid DNA from overnight cultures growing in LuriaBertani medium (Oxoid, Basingstoke, United Kingdom), the QIAprep Spin Miniprep kit (QIAGEN, Hilden, Germany) was used. After harvesting, the cells were incubated in Buffer P1 containing lysozyme (20 mg/ml) for 90 min at 37°C before the remaining steps were performed as described by the manufacturer. Detection of protein crystals and selected insecticide toxin genes. The method for examination of crystalline inclusions by phase-contrast microscopy was the same as that described in a previous report (22). PCR analyses were carried out to detect various groups of insecticidal toxin genes from B. thuringiensis. Four general primer sets were employed, detecting cry-1, cry-3, cry-11, and cyt1A, as described earlier (16). The reaction for detection of cry-1 consisted of one Ready-To-Go PCR bead (Amersham Pharmacia Biotech, Hillerød, Denmark), 10 pmol of each primer, and 5 ␮l of DNA. The PCR conditions were the following: a single denaturation step of 10 min at 94°C, a step cycle program set for 30 cycles (with a cycle consisting of denaturation at 94°C for 40 s, annealing at 52°C for 40 s, and extension at 72°C for 1.5 min), and a final extension at 72°C for 7 min. The PCR for cry-3, cry-11, and cyt1A consisted of 10 ␮l of Eppendorf MasterMix 2.5⫻ (Eppendorf, Hamburg, Germany), 20 pmol of each primer, and 5 ␮l of DNA. PCR conditions were the same as for cry-1, but with annealing temperatures of 48°C for cry-3 and 51°C for cry-11 and cyt1A. RAPD analysis. Four 9- or 10-base primers, OPA-02, OPA-03, OPA-09 (Operon Technologies, Alameda, CA), and 0940-12 (5), were used. The PCR mixture consisted of one Ready-To-Go PCR bead (Amersham Pharmacia Biotech), 20 pmol of primer and 5 ␮l of DNA. The PCR conditions were as described by Daffonchio et al. (7). Detection of genes for emetic toxin and enterotoxins. For detection of genes encoding the production of the emetic toxin and CytK, the previously described primers and conditions were used (22). For simultaneous detection of all three genes for each of the enterotoxins HBL and NHE, a multiplex PCR method was developed. For detection of the three genes hblA, hblD, and hblC encoding HBL, and the three genes nheA, nheB, and nheC encoding NHE, the primer sequences described by Hansen and Hendriksen (13) were used. The PCR for detection of genes of the HBL complex consisted of 15 ␮l of Eppendorf MasterMix 2.5⫻, primer amounts of 30 pmol for hblA and 5 pmol each for hblD and hblC, and 5 ␮l of DNA. The PCR conditions were as follows: a single denaturation step of 10 min at 94°C, a step cycle program set for 35 cycles (with a cycle consisting of denaturation at 92°C for 40 s, annealing at 52°C for 40 s, and extension at 68°C for 1.5 min), and a final extension at 65°C for 7 min. The mixture for detection of genes of the NHE component consisted of 10 ␮l of Eppendorf MasterMix 2.5⫻, 20 pmol of each primer, and 5 ␮l of DNA. The PCR conditions were the same as for cry-1 with an annealing temperature of 55°C. All PCR amplifications were performed on a Peltier Thermal Cycler PTC-225 (MJ Research, Bio-Rad, Waltham, MA). Statistics. The observed prevalence of commercial B. thuringiensis strains among the total food isolates was 23/128, or 0.18. Assuming that the prevalence of the commercial strains is equally distributed among the food isolates, the

RESULTS Identification of Bacillus thuringiensis strains. Fifty of the 128 isolates belonging to the B. cereus group were classified as B. thuringiensis because of their content of either crystal proteins visualized by phase-contrast microscopy or cry genes as detected by PCR. Among the B. thuringiensis isolates, 38 strains contained visible protein crystals, whereas 12 strains were positive for a cry gene but had no visible protein crystals. The majority of the 50 isolates, namely, 31 strains, were positive for cry-1, which is also present in the commercially used strains B. thuringiensis subsp. kurstaki HD1 and B. thuringiensis subsp. aizawai. In addition, 1 strain out of the 50 B. thuringiensis isolates was positive for cry-11, 1 was positive for cry-3, 1 was positive for cyt1A, and 1 strain was positive for both cry-11 and cyt1A, whereas 15 strains were negative for the cry genes tested. Identification of strains indistinguishable from commercial B. thuringiensis strains. On the basis of the content of protein crystals, crystal protein genes, and enterotoxigenic profiles, the 128 isolates were divided into 55 different groups (Table 1). Most of these comprised only one strain, revealing a high diversity among the isolates. The largest group of isolates comprised 20 strains grouping with B. thuringiensis strains isolated from the commercial products Biobit, Dipel, Foray, and Turex. Furthermore, there was a group of eight isolates harboring the same genes as the largest group but where no protein crystals could be detected by microscopy. None of the isolates grouped with the B. thuringiensis subsp. israelensis strain present in the commercial products Bactimos and Vectobac (Table 1). RAPD analysis (data not shown) and plasmid DNA profiling performed on the 28 isolates harboring the same crystal protein and enterotoxin genes as the B. thuringiensis strains present in Biobit, Dipel, Foray, and Turex revealed that 23 strains were indistinguishable from the active organisms in these products. Plasmid profiling (Fig. 1) divided the strains into two groups, one group of 14 isolates indistinguishable from the B. thuringiensis subsp. kurstaki strain in Dipel, Biobit, and Foray and another group of 9 isolates indistinguishable from the B. thuringiensis subsp. aizawai strain in Turex (Table 2). A statistically significantly (P ⬎ 0.99) high proportion of these isolates were from tomatoes, cucumbers, and peppers (Table 3). Enterotoxigenic profiles. Multiplex PCR revealed that the B. thuringiensis strains in six commercial products tested (Dipel, Biobit, Foray, Turex, Vectobac, and Bactimos) harbored all three genes for HBL and NHE, as well as the gene encoding CytK (Table 1). The same result was obtained for the 23 food isolates indistinguishable from the commercial strains. A high frequency of genes involved in human diarrhea was also found in the two groups of organisms differing from the commercial strains. In total, 70% of the noncommercial B. thuringiensis strains and 63% of the other B. cereus-like strains

B. THURINGIENSIS INSECTICIDES ON FRUITS AND VEGETABLES

VOL. 72, 2006

3437

TABLE 1. Occurrence of human toxigenic genes, crystal toxin genes, and visual crystals in B. cereus-like isolates from fruits and vegetables NHE genes

HBL genes cytK

Strain(s) or parameter nheA

nheB

nheC

hblA

hblD

hblC

Crystal toxin genes

Emetic-toxin gene

cry-1

cry-3

cry-11

cyt1A

Visual crystal

Vectobac, Bactimos Dipel, Foray, Biobit, Turex, 16319, 16377, 16387, 16391, 16392, 16393, 19394, 16401, 16412, 16471, 16545, 16630, 16631, 17007, 17302, 17385, 17399, 17530, 17535, 17539 16380, 16468, 16629, 16818, 17015, 17029, 17531, 17553 16195, 16373, 16390, 16402, 16470, 16546, 17030, 17299, 17540, 17542 16194, 16232, 16266, 16395, 17009, 17532, 17546, 17547 16193, 16272, 16378, 16400, 16411, 16557, 16638, 17528 16229, 16636, 16637, 17011, 17012, 17536, 17551 16233, 16237, 16375, 16547 16196, 16397, 16467 16265, 17010, 17529 16384, 16632, 16633 16472, 16817, 17538 16549, 16550, 17548 16228, 16264 16240, 16635 16269, 16321 16379, 17534 17533, 17544 16192 16197 16198 16230 16231 16234 16238 16239 16245 16246 16267 16270 16273 16274 16275 16276 16374 16376 16386 16389 16396 16398 16399 16634 16938 16940 17008 17013 17014 17031 17032 17033 17301 17537 17543 17545 17549 17550

⫹ ⫹

⫹ ⫹

⫹ ⫹

⫹ ⫹

⫹ ⫹

⫹ ⫹

⫹ ⫹

⫺ ⫺

⫺ ⫹

⫺ ⫺

⫹ ⫺

⫹ ⫺

⫹ ⫹









































































































⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫺ ⫺ ⫹ ⫹ ⫹ ⫹ ⫹ ⫺ ⫺ ⫹ ⫹ ⫺ ⫹ ⫺ ⫹ ⫺ ⫺ ⫹ ⫹ ⫹ ⫺ ⫺ ⫹ ⫹ ⫺ ⫺ ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫺ ⫹ ⫹ ⫺

⫺ ⫺ ⫺ ⫺ ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹ ⫺ ⫺ ⫺ ⫹ ⫺ ⫹ ⫹ ⫺ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹ ⫹ ⫺ ⫺ ⫹ ⫹ ⫺ ⫺ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫺ ⫺ ⫹ ⫹

⫹ ⫺ ⫺ ⫺ ⫹ ⫺ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹ ⫺ ⫹ ⫺ ⫹ ⫺ ⫺ ⫹ ⫺ ⫹ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫹ ⫹ ⫺ ⫺ ⫺ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫺ ⫺

⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫺ ⫺ ⫹ ⫺ ⫺ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫺ ⫺ ⫺ ⫺ ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹ ⫹ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫹

⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫺ ⫹ ⫺ ⫹ ⫺ ⫹ ⫺ ⫺ ⫺ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹

⫹ ⫺ ⫹ ⫹ ⫺ ⫹ ⫹ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫹ ⫹ ⫹ ⫺ ⫹ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫹ ⫺ ⫺ ⫺ ⫹ ⫺ ⫹ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹

⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫹ ⫹ ⫹ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫹ ⫹ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫹ ⫺ ⫺ ⫹ ⫹ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺

⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺

⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺

⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺

⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺

⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫹

⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫹ ⫺ ⫺ ⫺ ⫺ ⫹ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ ⫺ ⫹ ⫺ ⫺ ⫹ ⫺ ⫹ ⫹ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺

Total no. of isolates (n ⫽ 128)

105

93

81

108

110

79

54

2

31

1

2

2

38

3438

FREDERIKSEN ET AL.

APPL. ENVIRON. MICROBIOL. TABLE 3. Distribution among different food types of the three defined groups of organisms tested Strains indistinguishable from B. thuringiensis insecticides

Food type

Other B. cereus-like organisms

Total

Dipel, Biobit, Foray

Turex

2 3 3 0 0 2 0 0 0 4

0 3 1 5 0 0 0 0 0 0

9 1 0 1 4 2 5 1 4 0

21 1 0 1 13 2 10 9 16 5

32 8 4 7 17 6 15 10 20 9

14

9

27

78

128

Lettuce Tomatoes Cucumbers Peppers Berries Grapes Herbs Apples Root vegetables Other products

FIG. 1. Agarose gel electrophoresis of plasmid DNA. Shown are B. thuringiensis subsp. kurstaki HD1 from Dipel (lane 2), B. thuringiensis subsp. aizawai from Turex (lane 5), isolates indistinguishable from the Dipel strain (lanes 3, 4, 6, 8, 10, and 11), and isolates indistinguishable from the Turex strain (lanes 7, 9, and 12). Lane M, EcoRI- and HindIII-digested ␭ DNA (MBI Fermentas).

Noncommercial B. thuringiensis strains

Total

B. thuringiensis (Table 1). In addition to the gene for emetic toxin, both strains contained all three genes for the NHE complex and hblD of enterotoxin HBL (Table 1).

harbored at least one gene involved in human disease (Table 4). The gene cytK was more frequently found in the noncommercial B. thuringiensis strains (in 56% of the strains) than in the other B. cereus-like strains (in 21% of the strains), while the occurrence of genes encoding enterotoxins HBL and NHE was relatively high in both groups. Genes encoding the emetic toxin were only detected in two strains that could not be classified as

DISCUSSION High counts of B. cereus-like organisms are generally considered to be associated mainly with heat-treated, starchy products such as rice and pasta, where growth of this organism

TABLE 2. Characteristics of strains indistinguishable from commercial B. thuringiensis strains Strain

Sampling region

Sampling date (day/mo)

Type

Country of origin

Concn in product (CFU g⫺1)a

Strains similar to Dipel, Biobit, and Foray 16387 16392 16545 17015 17385 17530 17302 17399 16394 16401 17029 16629 17539 17553

Viborg Vejle Esbjerg Herning Aalborg Copenhagen Herning Herning Ringsted Ringsted Bornholm Esbjerg Copenhagen Copenhagen

8/8 13/8 21/8 2/10 13/11 22/8 8/11 21/11 24/7 24/7 6/9 28/8 15/8 15/8

Cucumbers Cucumbers Cucumbers Figs Grapes Grapes Lettuce Lettuce Melons Pineapples Sugar peas Cherry tomatoes Cherry tomatoes Beefsteak tomatoes

Denmark Denmark Denmark Turkey Italy Chile Unknown Italy Spain Honduras Zambia The Netherlands Denmark Denmark

ND b 11,000 550 1,000 9,400 100 100 1,500 100 10 200 20 12,000 100

Strains similar to Turex 16412 16380 16391 16393 16471 16818 16630 16631 17535

Odense Esbjerg Vejle Vejle Ringsted Odense Esbjerg Esbjerg Copenhagen

16/8 31/7 13/8 13/8 7/8 11/9 28/8 28/8 17/9

Cucumbers Pepper mixture Peppers Pepper mixture Peppers Peppers Cherry tomatoes Stalk tomatoes Stalk tomatoes

Denmark The Netherlands The Netherlands The Netherlands The Netherlands Turkey The Netherlands The Netherlands The Netherlands

10 4,600 300 300 ND 50 20 20 1,600

a b

The counts presented are a subset of previously published concentrations of B. cereus-like organisms in ready-to-eat food (22). ND, not determined.

B. THURINGIENSIS INSECTICIDES ON FRUITS AND VEGETABLES

VOL. 72, 2006

TABLE 4. Percentage of toxigenic genes among strains different from the commercial B. thuringiensis strains % of toxigenic genes (no. of positive strains) B. thuringiensis noncommercial strainsa (27 isolates)

Other B. cereus-like organisms (78 isolates)

Total (105 isolates)

HBL (all three genes) hblA hblD hblC

52 (14) 85 (23) 100 (27) 52 (14)

46 (36) 79 (62) 90 (70) 54 (42)

48 (50) 81 (85) 92 (97) 53 (56)

NHE (all three genes) nheA nheB nheC

44 (12) 70 (19) 81 (22) 52 (14)

31 (24) 81 (63) 63 (49) 62 (48)

34 (36) 78 (82) 68 (71) 59 (62)

cytK

56 (15)

21 (16)

30 (31)

2.6 (2)

1.9 (2)

63 (49)

65 (68)

Gene(s) involved in human disease

Emetic-toxin gene At least one of the human toxin genes

0 70 (19)

a Twenty-three of the 50 identified B. thuringiensis isolates were indistinguishable from strains in commercial B. thuringiensis products and harbored the genes for all components of the three enterotoxins HBL, NHE, and CytK.

may have occurred because of improper cooling after cooking. Recently, though, we found high counts (⬎104 CFU g⫺1) in fresh cucumbers and tomatoes (22), and others have also demonstrated high counts of B. cereus-like organisms in fresh vegetables (25). It was suggested that these findings were not attributed to growth on the vegetables but could be a result of spraying with B. thuringiensis-based insecticides. The present study revealed that B. thuringiensis strains indistinguishable from the commercial strains in the microbial insecticides Biobit, Dipel, Foray, and Turex are present on fresh vegetables for sale in Danish retail shops, particularly tomatoes, cucumbers, and peppers originating from Denmark, as well as from other countries. According to previously published data on the counts of B. cereus-like organisms on these vegetables (22), two of the samples with B. thuringiensis strains indistinguishable from the commercial strains fell into the group with high counts; i.e., spraying with B. thuringiensis insecticides may result in residues reaching levels above 104 CFU g⫺1. The majority of the tomatoes and cucumbers with isolates indistinguishable from the strain in Dipel originated from Denmark, where use of Dipel is allowed for plant protection (http: //www.mst.dk). Use of the product Turex is not allowed in Denmark, but it is in The Netherlands (http://www.ctb.agro.nl), and actually the majority of the isolates indistinguishable from the strain in Turex were from products produced in The Netherlands. Although one could argue that the finding of commercial-strain-like strains on these products is due to natural isolates indistinguishable from commercial strains, there are strong indications that the findings are actual residues of commercial products. This is substantiated by the fact that the products (tomatoes, cucumbers, and peppers) are grown in greenhouses in Denmark and The Netherlands, where contamination with natural isolates must be considered minimal, and the fact that Turex-like strains are primarily found on Dutch products. While B. thuringiensis strains have previously been

3439

isolated from vegetables (22, 25), grapes (2, 3), milk, pasta, and bread (9, 21), this is one of the first studies revealing that many B. thuringiensis strains isolated from fresh produce are in fact indistinguishable from commercial B. thuringiensis strains. Recently, Hendriksen and Hansen isolated HD1 strains indistinguishable from the commercial strains from cabbage for human consumption (14). Our study demonstrated that genes encoding enterotoxins were more frequently found in the strains indistinguishable from the commercial strains than in the strains different from these. The presence of enterotoxin-encoding genes in commercial B. thuringiensis strains was also found in previous studies (12, 13, 20). Still, additional investigations are needed to clarify whether the genes are expressed in the human gut after ingestion of the bacteria or spores, although several studies have shown that enterotoxin genes in commercial strains are not only present but also expressed in vitro (8, 16, 22, 24). Since all commercial strains harbor genes for all of the three known enterotoxins, HBL, NHE, and CytK, there is a risk that high levels of these organisms may cause human disease. Taking this enterotoxigenic potential into account, as well as the fact that B. thuringiensis cannot be separated from B. cereus at the chromosomal level, vegetable producers and food authorities responsible for food safety should consider the amount of B. thuringiensis insecticide residue left on products after harvest. The European Food Safety Authority has recommended that processors should ensure that levels of B. cereus bacteria between 103 and 105/g are not reached at the day of consumption (1). We recommend that this statement should apply also to residues of commercial, enterotoxin-encoding B. thuringiensis strains. ACKNOWLEDGMENTS We thank the National Environmental Research Institute (Bjarne Munk Hansen) for providing strains from commercial products. Thanks to the Regional Veterinary and Food Authorities for collecting and analyzing the food samples and Bodil Madsen and Rikke Kubert for technical assistance. The Danish Veterinary and Food Administration and the Danish Environmental Protection Agency partly funded this study. REFERENCES 1. Anonymous. 2005. Opinion of the Scientific Panel on Biological Hazards on Bacillus cereus and Other Bacillus spp. in Foodstuffs. Question no. EFSAQ-2004-010. Eur. Food Saf. Authority J. 175:1–48. 2. Bae, S., G. H. Fleet, and G. M. Heard. 2004. Occurrence and significance of Bacillus thuringiensis on wine grapes. Int. J. Food Microbiol. 94:301–312. 3. Bidochka, M. J., L. B. Selinger, and G. G. Khachatourians. 1987. A Bacillus thuringiensis isolate found on grapes imported from California. J. Food Prot. 50:857–858. 4. Brillard, J., and D. Lereclus. 2004. Comparison of cytotoxin cytK promoters from Bacillus cereus strain ATCC 14579 and from a B. cereus food-poisoning strain. Microbiology 150:2699–2705. 5. Brousseau, R., A. Saint-Onge, G. Prefontaine, L. Masson, and J. Cabana. 1993. Arbitrary primer polymerase chain reaction, a powerful method to identify Bacillus thuringiensis serovars and strains. Appl. Environ. Microbiol. 59:114–119. 6. Crickmore, N., D. R. Zeigler, J. Feitelson, E. Schnepf, J. Van Rie, D. Lereclus, J. Baum, and D. H. Dean. 1998. Revision of the nomenclature for the Bacillus thuringiensis pesticidal crystal proteins. Microbiol. Mol. Biol. Rev. 62:807–813. 7. Daffonchio, D., S. Borin, G. Frova, P. L. Manachini, and C. Sorlini. 1998. PCR fingerprinting of whole genomes: the spacers between the 16S and 23S rRNA genes and of intergenic tRNA gene regions reveal a different intraspecific genomic variability of Bacillus cereus and Bacillus licheniformis. Int. J. Syst. Bacteriol. 48:107–116. 8. Damgaard, P. H. 1995. Diarrhoeal enterotoxin production by strains of Bacillus thuringiensis isolated from commercial Bacillus thuringiensis-based insecticides. FEMS Immunol. Med. Microbiol. 12:245–250.

3440

FREDERIKSEN ET AL.

9. Damgaard, P. H., H. D. Larsen, B. M. Hansen, J. Bresciani, and K. Jorgensen. 1996. Enterotoxin-producing strains of Bacillus thuringiensis isolated from food. Lett. Appl. Microbiol. 23:146–150. 10. Fagerlund, A., O. Ween, T. Lund, S. P. Hardy, and P. E. Granum. 2004. Genetic and functional analysis of the cytK family of genes in Bacillus cereus. Microbiology 150:2689–2697. 11. Feitelson, J. S., J. Payne, and L. Kim. 1992. Bacillus thuringiensis: insects and beyond. Bio/Technology 10:271–275. 12. Gaviria Rivera, A. M., P. E. Granum, and F. G. Priest. 2000. Common occurrence of enterotoxin genes and enterotoxicity in Bacillus thuringiensis. FEMS Microbiol. Lett. 190:151–155. 13. Hansen, B. M., and N. B. Hendriksen. 2001. Detection of enterotoxic Bacillus cereus and Bacillus thuringiensis strains by PCR analysis. Appl. Environ. Microbiol. 67:185–189. 14. Hendriksen, N. B., and B. M. Hansen. 2006. Detection of Bacillus thuringiensis kurstaki HD1 on cabbage for human consumption. FEMS Microbiol. Lett. 257:106–111. 15. Jackson, S. G., R. B. Goodbrand, R. Ahmed, and S. Kasatiya. 1995. Bacillus cereus and Bacillus thuringiensis isolated in a gastroenteritis outbreak investigation. Lett. Appl. Microbiol. 21:103–105. 16. Jensen, G. B., P. Larsen, B. L. Jacobsen, B. Madsen, A. Wilcks, L. Smidt, and L. Andrup. 2002. Isolation and characterization of Bacillus cereus-like bacteria from faecal samples from greenhouse workers who are using Bacillus thuringiensis-based insecticides. Int. Arch. Occup. Environ. Health 75: 191–196. 17. Lindback, T., A. Fagerlund, M. S. Rodland, and P. E. Granum. 2004. Char-

APPL. ENVIRON. MICROBIOL.

18. 19. 20. 21. 22. 23. 24.

25.

acterization of the Bacillus cereus Nhe enterotoxin. Microbiology 150:3959– 3967. Lund, T., M. L. De Buyser, and P. E. Granum. 2000. A new cytotoxin from Bacillus cereus that may cause necrotic enteritis. Mol. Microbiol. 38:254–261. Lund, T., and P. E. Granum. 1997. Comparison of biological effect of the two different enterotoxin complexes isolated from three different strains of Bacillus cereus. Microbiology 143:3329–3336. Perani, M., A. H. Bishop, and A. Vaid. 1998. Prevalence of beta-exotoxin, diarrhoeal toxin and specific delta-endotoxin in natural isolates of Bacillus thuringiensis. FEMS Microbiol. Lett. 160:55–60. Phillips, J. D., and M. W. Griffiths. 1986. Factors contributing to the seasonal variation of Bacillus spp. in pasteurized dairy products. J. Appl. Bacteriol. 61:275–285. Rosenquist, H., L. Smidt, S. R. Andersen, G. B. Jensen, and A. Wilcks. 2005. Occurrence and significance of Bacillus cereus and Bacillus thuringiensis in ready-to-eat food. FEMS Microbiol. Lett. 250:129–136. Schnepf, E., N. Crickmore, J. Van Rie, D. Lereclus, J. Baum, J. Feitelson, D. R. Zeigler, and D. H. Dean. 1998. Bacillus thuringiensis and its pesticidal crystal proteins. Microbiol. Mol. Biol. Rev. 62:775–806. Tayabali, A. F., and V. L. Seligy. 2000. Human cell exposure assays of Bacillus thuringiensis commercial insecticides: production of Bacillus cereuslike cytolytic effects from outgrowth of spores. Environ. Health Perspect. 108:919–930. Valero, M., L. A. Hernandez-Herrero, P. S. Fernandez, and M. C. Salmeron. 2002. Characterization of Bacillus cereus isolates from fresh vegetables and refrigerated minimally processed foods by biochemical and physiological tests. Food Microbiol. 19:491–499.