sign in sign up
<<
Home , Bacillus, Bacillus cereus, Clostridium botulinum

1178

Journal of Food Protection, Vol. 71, No. 6, 2008, Pages 1178–1185 Copyright ᮊ, International Association for Food Protection

Levels and Toxigenicity of cereus and perfringens from Retail Seafood

T. RAHMATI AND R. LABBE*

Food Laboratory, Department of Food Science, University of Massachusetts, Amherst, Massachusetts 01003, USA

MS 07-575: Received 26 October 2007/Accepted 2 February 2008

ABSTRACT Downloaded from http://meridian.allenpress.com/jfp/article-pdf/71/6/1178/1682005/0362-028x-71_6_1178.pdf by guest on 03 October 2021

For the period 1990 through 2003, seafood was the most commonly identified food linked to foodborne outbreaks in the United States. as a commodity has rarely been examined for the presence of in particular. For the present study, 347 fresh and processed retail seafood samples were examined for the presence of , , and B. cereus. The presence of C. botulinum was not confirmed in any of the isolates, but C. perfringens was confirmed in 17 samples. One of the C. perfringens isolates possessed the gene, as determined by PCR. In contrast, 62 confirmed B. cereus isolates were obtained from separate samples at levels ranging from 3.6 to Ͼ1,100 CFU/g. Thirty (48%) of 62 isolates produced both the BL (HBL) and nonhemolytic (NHE) , and 58 (94%) and 31 (50%) produced NHE or HBL , respectively. The presence of at least one of the three genes of the NHE complex was detected in 99% of the isolates; 69% of the isolates possessed all three genes. In contrast, 71% of the isolates possessed at least one of the three genes of the HBL complex, and 37% possessed all three HBL gene components. Fifty of the 62 B. cereus isolates were from imported seafood, and 19 (38%) of these samples were at levels Ͼ100 CFU/g. Twelve of the 14 highest enterotoxin assay results were from isolates from imported food. Only one B. cereus isolate possessed the synthetase gene, ces; this isolate also possessed the genes for the three-component HBL and NHE complexes. A majority of enterotoxin-producing isolates were resistant to 2 of 10 antibiotics tested, ceftriaxone and clindamycin. Our results demonstrate the potential of seafood as a vehicle for caused by B. cereus, in particular the enterotoxin-producing genotype.

In the United States from 1990 through 2003, seafood cases, the complex is produced by vegetative growth in the was the most commonly identified food linked to foodborne . The NHE complex is also composed of outbreaks with identified vehicles (41). Bacillus cereus and three different proteins, NheA, NheB, and NheC, coded for Clostridium perfringens are gram-positive -forming by the three genes nheA, nheB, and nheC, also organized foodborne . Historically they have been associ- in one operon (16). Both NHE and HBL complexes require ated with food service operations in particular because of all three components for biological activity. There is some the heat resistance of their . The potential role of disagreement as to whether individual bacterial strains may Bacillus and B. cereus in particular in the safety of possess both the emetic and diarrheal toxins (2, 13). Mo- minimally processed convenience foods has been noted (19, ravek et al. (30) claimed that for most strains, cytotoxicity 32). B. cereus was the most common foodborne is largely due to Nhe. Commercial assay kits are available isolated from airline food from 1991 through 1994 (20). for the detection of the enterotoxins, but because these kits Foodborne illness caused by B. cereus and C. perfringens measure only one component of each complex, positive iso- is highly underreported. Historically, fish samples have lates can be considered only potentially enterotoxigenic. A rarely been examined for the presence of these two patho- third one-component enterotoxin was reported by Lund et gens. al. (28). Although infrequently reported thus far, the path- B. cereus causes two types of food poisoning symp- ologic change it causes, necrotic , is similar to that toms, emesis and . The emetic disease is caused by caused by the ␤- of C. perfringens. PCR primers spe- cereulide, a 45-kDa cyclic heat-stable dodecadepsipeptide, cific for the enterotoxin genes have been described, and and is typically associated with ; emesis begins within primers specific for the cereulide gene (ces) were developed 1 to 5 h of consumption. In diarrheal cases, at least two recently (12). three-component heat-labile enterotoxins are produced by The role of C. perfringens enterotoxin in foodborne food poisoning strains, hemolysin BL (HBL) and nonhe- illness has been well established. The clinical signs are sim- molytic enterotoxin (NHE) (15). The HBL complex is com- ilar to those caused by B. cereus enterotoxin. However, un- posed of three proteins, B, L1, and L2, transcribed from the like the B. cereus toxins (14), the production of C. perfrin- genes hblC (encoding L2), hblD (encoding L1), and hblA gens enterotoxin is linked to the sporulation process, which (encoding B) organized in an operon (37, 39). In clinical occurs in the small intestine following consumption of large numbers of cells from temperature-abused foods (26). * Author for correspondence. Tel: 413-545-1021; Fax: 413-545-1262; During the past decades, a number of surveys have E-mail: [email protected]. been conducted concerning the incidence of B. cereus in J. Food Prot., Vol. 71, No. 6 C. PERFRINGENS AND B. CEREUS IN SEAFOOD 1179

TABLE 1. Oligonucleotide primers used in this study Primera Gene Amplified fragment size (bp) Sequence 5Ј→3Ј Reference

HBLC F hblC 740 GAT ACT AAT GTG GCA ACTGC 17 HBLC R TTG AGA CTG CTC GTTAGT TG HBLD F hblD 829 ACC GGT AAC ACT ATT CAT GC 17 HBLD R GAG TCC ATA TGC TTA GAT GC HBLA F hblA 1,154 AAG CAA TGG AAT ACA ATG GG 17 HBLA R AGA ATC TAA ATC ATG CCA CTG C NHEA F nheA 755 GTT AGG ATC ACA ATC ACC GC 17 NHEA R ACG AAT GTA ATT TGA GTC GC NHEB F nheB 743 TTT AGT AGT GGA TCT GTA CGC 17 NHEB R TTA ATG TTC GTT AAT CCT GC NHEC F nheC 683 TGG ATT CCA AGA TGT AAC G 17 Downloaded from http://meridian.allenpress.com/jfp/article-pdf/71/6/1178/1682005/0362-028x-71_6_1178.pdf by guest on 03 October 2021 NHEC R ATT ACG ACT TCT GCT TGT GC CES F1 ces 1,271 GGT GAC ACA TTA TCA TAT AAG GTG 12 CES R2 GTAAGCGAACCTGTCTGTAACAACA CPE F cpe 233 GGA GAT GGT TGG ATA TTA GG 29 CPE R GGA CCA GCA GTT GTA GAT A NA2 F nhe 766 AAG CIG CTC TTC GIA TTC 11 NB1 R ITI GTT GAA ATA AGC TGT GG HD2 F hbl 1,091 GTA AAT TAI GAT GAI CAA TTT C 11 HA4 R AGA ATA GGC ATT CAT AGA TT a F, forward primer; R, reverse primer. HBLC, HBLD, and HBLA were used for hemolysin BL components; NHEA, NHEB, and NHEC were used for nonhemolytic enterotoxin components; CES was for cereulide peptide synthetase; CPE was for C. perfringens enterotoxin; NA2 was for nonhemolytic enterotoxin A; NB1 was for nonhemolytic enterotoxin B; HD2 was for hemolytic enterotoxin D; HA4 was for hemolytic enterotoxin A. foods (3, 19, 22, 23, 33, 36, 40, 43, 48, 49, 51), but only The control strain for C. perfringens was NCTC 8238. Isolates one targeted seafood for several foodborne pathogens (1). and controls were maintained in the spore state on nutrient agar Only the most recent surveys have determined the entero- slants for B. cereus or in Cooked medium (Difco, Becton toxin-producing ability of isolates, and no such surveys Dickinson, Sparks, Md.) for C. perfringens. have been published for U.S. retail foods. Numerous sur- Enumeration and isolation. Samples obtained from 2001 veys of the incidence of C. perfringens in foods have been through 2004 were examined for B. cereus and C. botulinum type reported (24), but few of these have focused on fish (21, E by the most-probable-number (MPN) method as described in 42). As with B. cereus, until recently such surveys did not the U.S. Food and Drug Administration Bacteriological Analytical determine the enterotoxin-producing potential of isolates. Manual (BAM) (46) and were examined for C. perfringens using Few nonoutbreak isolates possess the C. perfringens en- a previously described method (27). Samples were homogenized with a stomacher, and an initial three-tube MPN series was created terotoxin gene cpe (27, 49), even though from 1983 through with 1-ml dilutions of inoculum at 10Ϫ1,10Ϫ2, and 10Ϫ3. For B. 2002 this organism was second in terms of confirmed cases cereus, Trypticase soy broth was used, and culture and third in terms of outbreaks of bacterial foodborne dis- from tubes was plated onto –egg yolk–polymyxin agar ease in the United States (4–7). and incubated at 32ЊC for 24 to 48 h. For all isolates, confirmatory In the present study, we evaluated the levels and char- tests as described in the BAM were performed: , anaerobic acteristics of B. cereus and C. perfringens in retail seafood. utilization of glucose, nitrate reduction, Voges-Proskauer reaction, Where possible, a distinction was noted between results decomposition, absence of inclusion bodies, of from domestic versus imported seafood; the majority of the sheep blood agar, and resistance. edible seafood consumed in the United States is imported Toxin assays. B. cereus enterotoxin production was mea- (31). We also intended to evaluate Clostridium botulinum sured with the B. cereus reversed passive latex agglutination type E in retail seafood, but this organism was not detected (RPLA) test kit (Unipath-Oxoid, Columbia, Md.), which detects in any of the samples. the L2 component of the tripartite toxin HBL (39), and with the Bacillus diarrheal enterotoxin visual immunoassay (Tecra Bioen- MATERIALS AND METHODS terprises, Pty, Ltd., Roseville, Australia), which measures the NheA antigen from the Nhe complex. The tests were performed Strains used. Isolates in this study were obtained from 347 according to the manufacturers’ instructions. Amounts of entero- fresh and processed seafood samples collected from small grocery toxin produced were evaluated by comparison with index values stores, large supermarkets, and ethnic food stores in western Mas- derived from the Oxoid and Tecra readings. The lowest reproduc- sachusetts from February 2001 to July 2004 and were analyzed ible titer for the Oxoid test kit was 1:2. on the day of purchase. Control strains for B. cereus possessing HBL and NHE complexes were ATCC 14579 and ATCC 1230/ C. botulinum identification. Presumptive C. botulinum iso- 88, respectively, and the control for the emetic strain was ATCC lates were further identified using l6S rRNA gene sequencing (Ac- F4810/72 (A. Wong, University of Wisconsin, Madison, Wis.). cugenix, Newark, Del.). 1180 RAHMATI AND LABBE J. Food Prot., Vol. 71, No. 6 C Њ ϩ Growth at 12 ϩ Downloaded from http://meridian.allenpress.com/jfp/article-pdf/71/6/1178/1682005/0362-028x-71_6_1178.pdf by guest on 03 October 2021 HBL complex ϩϩϩ ϪϪϪ ϩϩϩϩ ϩϩϩ ϩϪϪ ϪϪϪ ϩ ϩϩϪ ϩ ϩϩϩ ϩϩϩ ϩϩϩ ϩ ϩϩϩ ϩϩϩ ϩϩϪ ϩϩϩ ϩϩϩϩϪϩ ϩ ϩϪϪ ϩϩϩ ND ND ND hblC hblD hblA ces b c 1:64 1:4 1:64 1:128 1:32 1:2 1:128 1:128 1:16 1:8 1:64 1:4 1:64 1:16 1:8 1:64 1:64 ND kit titer Oxoid test NHE complex ϩϩϩ ϩϩϩ ϩϩϩϪ ϩϪϪ ϩϩϩϪ ϪϪϪ ϩϩϩϪ ϩϪϪ ϩϩϩϪ ϪϪϪ ϩϩϩϪ ϩϪϪ ϩϩϩ ϩϩϩϪ ϩϪϪ ϩϩϩϪ ϪϩϪ ϩϩϩ ϩϩϪϪ ϪϪϪ ϩϪϩϪ ϩϪϪϩϩϩϪ ϩ ϪϪϪϩϪϪ ϩ ϩϩϩϪ ϩϪϪ ϪϩϩϪ ϪϪϪ ϩϩϩϪ ϪϪϪ ϩϩϪ ϩϩϩϪ ϪϩϪ ϩϩϩϪ ϪϪϪ ϩϩϩ ϩϩϩ ϩϩϩ ϩϩϩϪ ϩϪϪ ϩϩϩϪ ϪϩϪ ϩϩϩ ϩϪϩ ϩϩϩ ϩϩϩ ϩϪϩ ϩϩϩ ϩϩϪ ϩϩϩ nheA nheB nheC a ϪϩϩϩϪ ϩϪϪ Ϫϩϩϩ ϪϪϩϩϪ ϪϪϪ ϪϪϩϪϪ ϩϩϩ Tecra test kit index 1,100 5 1,100 3 B. cereus Ͼ Ͼ level (MPN/g) isolates B. cereus EelMackerelFish cakeCrabMackerelMusselPrawnFish noodleMussel 75Squid 43Dried 9.2 fishShrimp 240Dried 75 fish 240 5 Shrimp 4 Crab 43 5 23Squid 4 Cuttlefish 3.6Squid 4 23 23Cuttlefish 4 3 Shrimp 3.6Shrimp 3.6 5 Fish cracker 23 Cuttlefish 4 Squid 4 3.6 4 9.2Fish 7.4 pasteShrimp powder 23Fish sauceCrab 3 sauce 29 5 240 3 Shrimp paste 3.6Fish sauce 15Shrimp 4 sauceSquid 3 15Anchovy 4 9.2 3 1,100Shrimp 4 Anchovy 460 9.2Smoked fish 3 460 460 4 4 4 4 1,100 3.6 4 4 460 3.6 23 4 4 4 4 4 Characteristics of d d d d d d d d d d d d d d d d d d d d d d d d d d d d Sample no. Source d d d d d d d d 5 6 7 8 9 10 12 14 16 17 18 19 21 22 23 25 26 30 31 3 4 11 13 15 20 24 27 28 29 32 33 34 35 36 37 TABLE 2. 1230/88ATCC 14579F4810/72 HBL1 positive NHE control positive control2 Emetic control Fish stick ND 4 ND ND 23 ND ND ND ND 1:128 ND ND ND ND ND J. Food Prot., Vol. 71, No. 6 C. PERFRINGENS AND B. CEREUS IN SEAFOOD 1181 C Њ Growth at 12 Downloaded from http://meridian.allenpress.com/jfp/article-pdf/71/6/1178/1682005/0362-028x-71_6_1178.pdf by guest on 03 October 2021 HBL complex ϩϩϩ ϩϪϩ ϩϩϩ ϩϩϩ ϩϩϩ ϩϩϩ ϩϩϩ ϩϩϪ ϩϩϩ ϩϩϩ ϩϩϩ ϩϩϩ ϩϩϩ ϩϩϪ hblC hblD hblA ces b 1.32 1:4 1:2 1:64 1:32 1:128 1:8 1:2 1:32 1:16 1:8 1:16 1:32 1:128 kit titer Oxoid test ϪϪ ϪϪϪ e Ϫ NHE complex e ϩϩϩ ϩϩϩϪ ϪϪϪ ϩϪϩ ϩϩϩ ϩϩϩ ϩϩϩϪ ϩϩϩ ϩϩϩ ϩϪϩϪ ϪϪϪ ϩϪϩ ϩϩϩ ϩϩϩϪ ϪϪϪ ϩϩϩϪ ϩϩϪ ϩϩϩϪ ϪϪϩ ϩϪϩ ϩϪϩ ϩϪϩ ϩϪϩϪ ϪϪϪ ϩϩϩϪ ϪϪϪ ϩϩϩϪ ϪϪϪ ϩϩϩ ϩϩϩ ϩϪϩ ϩϩϩϪ ϪϩϪ ϩϩϩ Ϫ nheA nheB nheC a Tecra test kit index 1,100 4 1,100 4 1,100 4 1,100 4 B. cereus Ͼ Ͼ Ͼ Ͼ level (MPN/g) 3 were considered negative. Ͻ 1:2 were considered negative. Ͻ SquidFish pasteCuttlefishSheatfishAnchovyEelCroakerSquidSheatfish 43 3.6Carp 23Sheatfish 43Mussel 9.2Shrimp 4 3 4 9.2 21 3 23 4 3.6 240 23 4 4 3.6 5 4 9.2 5 4 4 5 Continued d d d d d d d d d d d d d Sample no. Source Strains with an indexStrains of with a titer of Sample from seafood known to be imported. ND, not determined. A positive PCR product was obtained using the NA2 and NB1 primer pairs. 40 44 46 49 50 5152 Cod Bluefish 150 5 39 41 42 43 45 47 48 53 Sole 5455565758 Trout Tilapia Scallop Salmon Trout 36 460 1,100 7.4 4 4 4 4 5960 Monkfish Salmon 9.2 4 6162 Hake Salmon 11 3.6 4 4 TABLE 2. 38 a b c d e 1182 RAHMATI AND LABBE J. Food Prot., Vol. 71, No. 6

Starch hydrolysis. The ability of B. cereus isolates to hy- and the diameter of the zone of clearance around each disk was drolyze starch was tested on nutrient agar plates containing 1% measured with reference to the zone around the control disks pro- soluble starch. Plates were incubation at 30ЊC for 48 h and then vided by BBL. flooded with Lugol’s iodine. RESULTS AND DISCUSSION Low-temperature growth studies. Tubes containing 5 ml of Trypticase soy broth (TSB; Difco, Becton Dickinson) were inoc- Consumer interest in minimally processed chilled prod- ulated with a loopful of each B. cereus isolate from the nutrient ucts with long shelf lives raises concerns about the survival agar slants. Tubes were incubated overnight at 32ЊC, and then 100 and growth of foodborne pathogens in general and survival ␮l was inoculated into another 5-ml tube of TSB and incubated of the spore-forming subgroup in particular. We investigat- at 12ЊC in a circulating low-temperature water bath (Isotemp ed the levels and distribution of certain virulence factors of 1016S, Fisher Scientific, Pittsburgh, Pa.). Tubes were examined three such species in 347 seafood samples. each day for 14 days for visible growth. Strains that grew at 12ЊC also were tested for the ability to grow at 7ЊC in TSB and on C. botulinum and C. perfringens. Of five presumptive

Trypticase soy agar. C. botulinum isolates, none were confirmed as C. botulinum Downloaded from http://meridian.allenpress.com/jfp/article-pdf/71/6/1178/1682005/0362-028x-71_6_1178.pdf by guest on 03 October 2021 by DNA sequence analysis. These isolates were not studied PCR. The presence of toxin-encoding genes in isolates and further. Seventeen (5%) confirmed isolates of C. perfrin- controls was determined using single primer pairs. The presence gens were obtained at levels of 3.6 to 240 MPN/g (not of the C. perfringens enterotoxin gene was determined as previ- shown). The presence of cpe was detected in one of these ously described using strain NCTC 8239 as a positive cpe control. isolates. Previous surveys have indicated the low incidence (27). For B. cereus, genomic DNA was isolated from overnight growth at 32ЊC on nutrient agar slants using Gene Releaser (Bio- of enterotoxin-producing C. perfringens in meat and fish ventures, Murfreesboro, Tenn.) according to the manufacturer’s (27, 38, 50). This finding is surprising given the role of this protocol. To ensure that cultures were in the vegetative cell stage, pathogen in the incidence of foodborne illness. short incubation times, an incubation temperature of 20ЊC, or both B. cereus. Sixty-two seafood samples contained B. ce- were used for culture growth. In cultures that reached sporulation Ͼ stage VII, free spores (25) were negative for the toxin genes. reus at levels between 3.6 and 1,100 MPN/g (Table 2). While working to determine the presence of B. cereus enterotoxin The distribution of levels of B. cereus was evenly divided emetic genes, we investigated the use of boiling for preparation among the three ranges: 1 to 10, 10 to 100, and 100 to of template DNA. In a screening study, additional PCR products Ͼ1,100 MPN/g. For each isolate, the presence of the genes were detected that had not been obtained with Gene Releaser. for the NHE and HBL complexes was determined by the Therefore, all samples with previous negative PCR results were PCR assay and the presence of the associated enterotoxins retested for the six enterotoxin gene targets. The boiling procedure in the sample was determined with the commercial test kits. also was employed for preparation of template DNA for ces by All isolates except no. 49 possessed one or more genes of ␮ suspending a 2-mm loopful of culture from nutrient agar in 60 l the NHE complex: nheA, nheB, and/or nheC. The presence of distilled water and heating in a boiling water bath for 20 min, of all three was detected in 43 (69%) of the 62 isolates, after which the sample was cooled and then centrifuged to remove and 16 (31%) and 2 (5%) of the isolates possessed two and insoluble material. The primers used for the detection of the var- ious genes are listed in Table 1. The PCR mixture (50 ␮l) con- one of the three genes, respectively. The common occur- sisted of 20 ␮l of DNA template, 800 ␮M deoxynucleoside tri- rence of nhe in B. cereus has previously been reported (8, 17, 18, 30, 53) for both foodborne and food-associated B. phosphate mix (Promega, Madison, Wis.), 1.5 mM MgCl2, 500 nM concentrations of each primer, 0.1 U of Eppendorf Taq DNA cereus and isolates (18, 35). The polymerase (Brinkmann, Westbury, N.Y.), and 5 ␮lofTaq DNA nheA gene was detected in 58 of 62 isolates by using the polymerase assay buffer. Thermal cycling was carried out in a NHEA primer pair. In all but two samples (no. 5 and 12), Techgene thermocycler (Techne, Cambridge, UK) with initial de- the corresponding gene product was present as determined naturation at 95ЊC for 5 min followed by 35 cycles of 5 min at by the commercial diarrheal enterotoxin immunoassay kit Њ Њ Њ 94 C, 1.5 min at 55 C, and 1.5 min at 72 C and then a final (Tecra, Table 2). In total, 58 (94%) of the 62 isolates pro- Њ extension for 4 min at 72 C. The results were determined by elec- duced the NHE toxin, similar to results from other studies trophoresis of 20 ␮l of PCR products in a 1.5% agarose gel for (36). 30 min at 80 V and staining with ethidium bromide with appro- priate size PCR markers (Promega). In this study, ENVISION Only two isolates (no. 21 and 49) produced detectable DNA dye (MidSci, St. Louis, Mo.), a DNA dye without ethidium NheA in the absence of nheA, an observation occasionally bromide, was used in place of ethidium bromide and provided reported by others (18) and perhaps due to sequence dif- equally good results. ferences among nheA copies. For isolate 49, the PCR was repeated using nhe primers in which the forward primer is Antibiotic resistance screening. Isolates were grown to an located in nheA and the reverse primer is located in nheB A600 density of 0.2 to 0.3 (early exponential phase) in nutrient and bases at known variable regions are replaced by inosine broth and then spread plated (0.1 ml) onto nutrient agar. Antimi- (11) (Table 1). When these primers were used, a PCR prod- crobial susceptibility test disks (Sensi Discs, BBL, Becton Dick- uct was obtained. However, for isolate 21 no PCR product inson) were placed on each plate. Two disks were used for each of the 10 antibiotics. The disks contained 10 ␮g of ampicillin, 30 was obtained using this composite primer. ␮g of ceftriaxone, 30 ␮g of chloramphenicol, 2 ␮g of clinda- Compared with the almost ubiquitous presence of the mycin, 15 ␮g of erythromycin, 30 ␮g of nalidixic acid, 300 ␮g NHE genes in the isolates, the distribution of the HBL of streptomycin, 30 ␮g of tetracycline, 5 ␮g of trimethoprim, or genes was much more variable, as determined by PCR as- 30 ␮g of . Plates were incubated at 32ЊC for 24 h, say, with various combinations of single or multiple HBL J. Food Prot., Vol. 71, No. 6 C. PERFRINGENS AND B. CEREUS IN SEAFOOD 1183 components. Forty-four (71%) of the 62 isolates possessed TABLE 3. Antibiotic sensitivity of 33 enterotoxin-positive B. ce- one or more of the three HBL components, and all three reus isolatesa components were detected in 23 (37%) of the isolates. Forty No. (%) of isolates: (65%) of the isolates tested positive with the primer pair Antimicrobial Concn agent (␮g/disk) Resistant Susceptible specific for hblC, which codes for the L2 component de- tected by the Oxoid RPLA kit. Guinebretie`reetal.(17) Ampicillin 10 1 (3) 2 (6) detected hblC in 61% of foodborne B. cereus isolates by Ceftriaxone 30 18 (55) 0 using the same primer pairs as used in the present study, Chloramphenicol 30 0 22 (67) and an additional 31% of their isolates were positive based Clindamycin 2 25 (76) 1 (3) on Southern blotting, In other studies, the presence of hblC Erythromycin 15 7 (21) 1 (3) has been reported in 29 to 41% of isolates (30, 52). The Nalidixic acid 30 0 12 (36) complete HBL complex also has been detected in B. thu- Streptomycin 10 0 31 (94) ringiensis (34). The 28 isolates that were negative by PCR Tetracycline 30 12 (36) 14 (42) Downloaded from http://meridian.allenpress.com/jfp/article-pdf/71/6/1178/1682005/0362-028x-71_6_1178.pdf by guest on 03 October 2021 assay for both hblD and hblA were reexamined using a Trimethoprim 5 0 1 (3) single set of composite primers targeting these two genes Vancomycin 30 0 31 (94) (11) (Table 1). No additional PCR products were detected a Positive for HBL or NHE. using this composite primer pair, suggesting minimal or no polymorphism in hblD and hblA among these 26 isolates. For most isolates (31 of 40) in which hblC was de- 58 isolates, 32 (73%) were from imported seafood. Simi- tected by PCR, its gene product, L , also was present, as 2 larly, for HBL, 30 of the 62 isolates were positive by the determined with the BCET-RPLA (Oxoid) test kit. For only Oxoid RPLA test, and of these 23 (77%) were from im- two isolates (no. 3 and 19) was the reverse true, i.e., the ported seafood. Except for two isolates (no. 52, Tecra assay; presence of enterotoxin in the absence of detectable hblC. and no. 62, Oxoid assay), all the isolates producing entero- In total, 31 (50%) of B. cereus isolates gave positive results toxins at the highest concentrations (index no. 5 for Tecra for L with this test. 2 and dilution 1:128 for Oxoid) were from imported seafood. Unlike nhe and hbl, the presence of ces, the cereulide The relatively high levels and toxin-forming potential of B. synthetase gene, was detected in only one (2%) isolate. cereus isolates in the imported seafood in this study is of However, this isolate produced HBL and NHE and pos- concern because most seafood consumed in the United sessed the three hbl and nhe genes, contrary to the results States is imported. of Ehling-Schulz et al. (12), who did not find hbl in 24 food and clinical strains possessing ces. Finley et al. (13) Antibiotic resistance. Thirty-three randomly selected also detected NHE in an emetic strain. However, this isolate B. cereus isolates that produced HBL, NHE, or both were and 15 others (26%) failed to hydrolyze starch, a phenotype screened for susceptibility to 10 antibiotics (Table 3). These common to but not indicative of the emetic biotype (2, 12, antibiotics were previously used in a study of antibiotic- 47). The results obtained in the present study support the resistant in shrimp (10). For only two antibiotics, view that emetic-type strains are uncommon (53) and may ceftriaxone and clindamycin, were more than half the iso- be more commonly associated with rice and the of rice lates resistant. B. cereus is usually susceptible to clinda- paddies (3). mycin (44, 45, 49) and resistant to third-generation ceph- Six of the isolates grew at 12ЊC, but none were able alosporins such as ceftriaxone (9). to grow at 7ЊC, which is surprising given the type of prod- Our results support the hypothesis that nhe is present uct tested. in virtually all B. cereus strains but the presence of hbl is Ninety-five percent of the B. cereus isolates produced less common. The fact that 95% of these non–food poison- one or more enterotoxins. However, the roles of HBL and ing strains possessed one or both of the enterotoxins has NHE in foodborne illness remain unclear. In particular, the significant implications for microbial . Never- results of the semiquantitative commercial test kits are not theless, it remains to be determined which of the HBL or always correlated with the presence of the gene whose NHE toxin components (and what concentrations) are re- product they measure. Moravek et al (30) recently reported sponsible for illness and what significance the results from the development of a quantitative immunoassay that re- the commercially available assay kits have. vealed that B. cereus strains are widely variable in the ACKNOWLEDGMENTS amount of enterotoxin produced. These authors concluded that quantitative detection of enterotoxin, in particular We thank A. Wong for providing the control strains of B. cereus. NHE, allows improved assessment of the toxic potential of This study was supported by the U.S. Department of Agriculture Seafood Safety grant. isolates. REFERENCES Imported versus domestic seafood. Fifty of the 62 B. cereus isolates were from imported seafood. Of these 50 1. Abeyta, C., Jr. 1983. Bacteriological quality of fresh seafood prod- Ͼ ucts from Seattle retail markets. J. Food Prot. 46:901–909. isolates, 19 (38%) were at levels 100 MPN/g (Table 2). 2. Agata, N., M. Ohta, and M. Mori. 1996. Production of an emetic For enterotoxins, 58 of the 62 isolates produced NHE, as toxin, cereulide, is associated with a specific class of Bacillus cereus. measured by the Tecra visual immunoassay, and of these Curr. Microbiol. 33:67–69. 1184 RAHMATI AND LABBE J. Food Prot., Vol. 71, No. 6

3. Altayar, M., and A. Sutherland. 2006. Bacillus cereus is common in 26. Labbe, R., and V. Juneja. 2006. Clostridium perfringens gastroen- the environment but emetic toxin producing isolates are rare. J. Appl. teritis, p. 136–184. In D. Cliver and H. Rieimann (ed.), Foodborne Microbiol. 100:7–14. and intoxications, 3rd ed. Elsevier, New York. 4. Centers for Disease Control. 1990. Foodborne disease outbreaks, 5- 27. Lin, Y.-T., and R. Labbe. 2003. Enterotoxigenicity and genetic re- year summary, 1983–1987. Morb. Mortal. Wkly. Rep. 39(SS01):15– latedness of Clostridium perfringens isolates from retail foods in the 23. United States. Appl. Environ. Microbiol. 69:1642–1646. 5. Centers for Disease Control and Prevention. 1996. Surveillance for 28. Lund, T., M. DeBuyser, and P. E. Granum. 2000. A new cytotoxin foodborne-disease outbreaks—United States, 1988–1992. Morb. from Bacillus cereus that may cause necrotic enteritis. Mol. Micro- Mortal. Wkly. Rep. 45(SS05):1–66. biol. 38:254–261. 6. Centers for Disease Control and Prevention. 2000. Surveillance for 29. Meer, R., and J. G. Songer. 1997. Multiplex polymerase chain re- foodborne disease outbreaks—United States, 1993–1997. Morb. action assay for genotyping Clostridium perfringens. Am. J. Vet. Res. Mortal. Wkly. Rep. 49(SS01):1–51. 58:702–705. 7. Centers for Disease Control and Prevention. 2006. Surveillance for 30. Moravek, M., R. Dietrich, C. Buerk, V. Broussolle, M.-H. Guine- foodborne-disease outbreaks—United States, 1998–2002. Morb. bretie`re, P. E. Granum, C. Nguyen-The, and E. Ma¨rtlbauer. 2006. Determination of the toxic potential of Bacillus cereus isolates by Mortal. Wkly. Rep. 55(SS10):1–34. quantitative enterotoxin analysis. FEMS Microbiol. Lett. 257:293– Downloaded from http://meridian.allenpress.com/jfp/article-pdf/71/6/1178/1682005/0362-028x-71_6_1178.pdf by guest on 03 October 2021 8. Dietrich, R., M. Moravek, C. Burk, P. E. Granum, and E. Martbauer. 298. 2005. Production and characterization of antibodies against each of 31. National Marine Fisheries Service. 2005. Imports and exports of the three subunits of the Bacillus cereus nonhemolytic enterotoxin fishery products: annual summary 1990–2004. Office of Science and complex. Appl. Environ. Microbiol. 71:8214–8220. Technology, Fisheries Statistics and Economics Division, National 9. Drobniewski, F. 1993. Bacillus cereus and related species. Clin. Mi- Marine Fisheries Service, Washington, D.C. crobiol. Rev. 6:324–338. 32. Nissen, H., J. Rosnes, J. Brendehaug, and G. Kleiberg. 2002. Safety 10. Duran, G., and D. Marshall. 2005. Ready-to-eat shrimp as an inter- evaluation of –processed ready meals. Lett. Appl. Micro- national vehicle of antibiotic-resistant bacteria. J. Food Prot. 68: biol. 35:433–438. 2395–2401. 33. Nygren, B. 1962. The occurrence of Clostridium perfringens and 11. Ehling-Schulz, M., M.-H. Guinebretie`re, A. Monthan, O. Berge, M. Bacillus cereus in foodstuff containing lecithin and in foodstuff ad- Fricker, and B. Svensson. 2006. Toxin gene profiling of enterotoxic ditives. Acta Pathol. Microbiol. Scan. Suppl. 160:1–89. and emetic Bacillus cereus. FEMS Microbiol. Lett. 260:232–240. 34. Pru¨ß, B. M., R. Dietrich, B. Nibler, E. Ma¨rtlbauer, and S. Scherer. 12. Ehling-Schulz, M., N. Vukov, A. Schulz, S. Shaheen, M. Andersson, 1999. The hemolytic entertotoxin HBL is broadly distributed among E. Ma¨rtlbauer, and S. Scherer. 2005. Identification and partial char- species of the Bacillus cereus group. Appl. Environ. Microbiol. 65: acterization of the non-ribosomal peptide synthetase gene responsi- 5436–5442. ble for cereulide production in emetic Bacillus cereus. Appl. Envi- 35. Rivera, A. M. G., P. E. Granum, and F. Priest. 2000. Common oc- ron. Microbiol. 71:105–114. currence of enterotoxin genes and enterotoxigenicity in Bacillus thu- 13. Finlay, W. J. J., N. A. Logan, and A. D. Sutherland. 1999. Semi- ringiensis. FEMS Microbiol. Lett. 190:151–155. automated metabolic staining assay for Bacillus cereus emetic toxin. 36. Rusul, G., and N. H. Yaacob. 1995. Prevalence of Bacillus cereus Appl. Environ. Microbiol. 65:1811–1812. in selected foods and detection of enterotoxin using TECRA-VIA 14. Finlay, W. J. J., N. A. Logan, and A. D. Sutherland. 2000. Bacillus and BCET-RPLA. Int. J. Food Microbiol. 25:131–139. cereus produces most emetic toxin at lower temperatures. Lett. Appl. 37. Ryan, P., J. Macmillan, and B. Zilinskas. 1997. Molecular cloning Microbiol. 31:385–389. and characterization of the genes encoding the L1 and L2 components 15. Granum, P. E. 2007. Bacillus cereus, p. 445–455. In M. P. Doyle of hemolysin BL from Bacillus cereus. J. Bacteriol. 179:2551–2556. and L. R. Beuchat (ed.), : fundamentals and fron- 38. Saito, M. 1990. Production of enterotoxin by Clostridium perfrin- tiers, 3rd ed. ASM Press, Washington, D.C. gens derived from humans, animals, foods and the natural environ- 16. Granum, P. E., K. O’Sullivan, and T. Lund. 1999. The sequence of ment in Japan. J. Food Prot. 53:115–118. the non-hemolytic enterotoxin operon from Bacillus cereus. FEMS 39. Schoeni, J., and A. Wong. 2005. Bacillus cereus food poisoning and Microbiol. Lett. 67:185–189. its toxins. J. Food Prot. 68:636–648. 17. Guinebretie`re, M.-H., V. Broussolle, and C. Nguyen-The. 2002. En- 40. Smith, D., M. Berrang, P. Feldner, R. Phillips, and J. Meinersmann. terotoxigenic profiles of food-poisoning and food-borne Bacillus ce- 2004. Detection of Bacillus cereus on selected retail prod- reus strains. J. Clin. Microbiol. 40: 3053–3056. ucts. J. Food Prot. 67:1770–1773. 18. Hansen, B., and N. Hendriksen. 2001. Detection of enterotoxigenic 41. Smith DeWaal, C., G. Hicks, K. Barlow, L. Alderton, and L. Ve- gosen. 2006. Foods associated with food-borne illness outbreaks Bacillus cereus and Bacillus thuringiensis strains by PCR analysis. from 1990 through 2003. Food Prot. Trends 26:466–473. Appl. Environ. Microbiol. 67:185–189. 42. Taniguti, T., and B. Zenitani. 1969. Incidence of Clostridium per- 19. Harmon, S., and D. Kautter. 1991. Incidence and growth potential fringens in fish. I. On the application of LAS medium to detection of Bacillus cereus in ready-to-serve foods. J. Food Prot. 54:373– of Clostridium perfringens. J. Food Hyg. Soc. Jpn. 10:199–203. 374. 43. te Giffel, M., R. Beumer, S. Leijendekkers, and F. Rombouts. 1996. 20. Hatakka, M. 1998. Microbiological quality of hot meals served by Incidence of Bacillus cereus and in foods in The airlines. J. Food Prot. 61:1052–1056. Netherlands. Food Microbiol. 13:53–58. 21. Hobbs, G., D. Cann, B. Wilson, and J. Shewan. 1965. The incidence 44. Tuazon, C., H. Murray, C. Levy, M. Solny, J. Curtin, and J. Shea- of organisms of the genes Clostridium in vacuum-packed fish in the gren. 1979. Serious infections from Bacillus spp. JAMA 241:1137– United Kingdom. J. Appl. Bacteriol. 28: 265–270. 1140. 22. Johnson, K. 1984. Bacillus cereus foodborne illness—an update. J. 45. Turnbull, P., and J. Kramer. 1995. Bacillus, p. 349–356. In P. Murray, Food Prot. 47:145–153. E. Baron, F. Tenover, and R. Yolken (ed.), Manual of clinical micro- 23. Kramer, J., and R. Gilbert. 1989. Bacillus cereus and other Bacillus biology. ASM Press, Washington, D.C. species, p. 21–70. In M. Doyle (ed.), Foodborne bacterial pathogens. 46. U.S. Food and Drug Administration. 1998. Bacteriological analytical Marcel Dekker, New York. manual, 8th ed. AOAC International, Gaithersburg, Md. 24. Labbe, R. 1989. Clostridium perfringens, p. 191–234. In M. Doyle 47. Valero, M., L. A. Herna´ndez-Herrero, P. S. Ferna´ndez, and M. (ed.), Foodborne bacterial pathogens. Marcel Dekker, New York. C. Salmero´n. 2002. Characterization of Bacillus cereus isolates 25. Labbe, R. 2005. Sporulation (morphology) of the , p. 647– from fresh vegetables and refrigerated minimally processed foods 659. In P. Durre (ed.), Handbook of the clostridia. CRC Press, Boca by biochemical and physiological tests. Food Microbiol. 19:491– Raton, Fla. 499. J. Food Prot., Vol. 71, No. 6 C. PERFRINGENS AND B. CEREUS IN SEAFOOD 1185

48. Van Netten, A., A. van de Moosdijk, P. van Hoensel, D. Mossel, and and F. M. van Leusden. 2006. Prevalence of potentially pathogenic I. Perales. 1990. Psychrotrophic strains of Bacillus cereus producing Bacillus cereus in food commodities in The Netherlands. J. Food enterotoxin. J. Appl. Bacteriol. 69:73–79. Prot. 69:2587–2594. 49. Weber, D., S. Saviteer, W. Rutala, and C. Thomann. 1988. In vitro 52. Wong, H., M. Chang, and J. Fan. 1988. Incidence and characteriza- susceptibility of Bacillus spp. to selected antimicrobial agents. An- tion of Bacillus cereus isolates contaminating dairy products. Appl. timicrob. Agents Chemother. 32:642–645. Environ. Microbiol. 54:699–702. 50. Wen, R., and B. McClane. 2004. Detection of enterotoxigenic Clos- 53. Yang, I. C., D. Shih, T.-P. Huang, Y.-P. Huang, J.-Y. Wang, and T. tridium perfringens type A isolates in American retail foods. Appl. M. Pan. 2005 Establishment of a novel multiplex PCR assay and Environ. Microbiol. 70:2685–2691. detection of toxigenic strains of the species in the Bacillus cereus 51. Wijnands, L. M., J. B. Dufrenne, F. M. Rombouts, P. H. in ’t Veld, group. J. Food Prot. 68:2123–2130. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/71/6/1178/1682005/0362-028x-71_6_1178.pdf by guest on 03 October 2021