Development of real-time PCR tests for detecting botulinum neurotoxins A, B, E, F producing Clostridium botulinum, Clostridium baratii and Clostridium butyricum

Authors

  • P. Fach,

    1.  Agence Française de Sécurité Sanitaire des Aliments (AFSSA), Laboratoire d’Etudes et de Recherches sur la Qualité des Aliments et les Procédés Agro-alimentaires (LERQAP), Maisons-Alfort, France
    Search for more papers by this author
  • P. Micheau,

    1.  Agence Française de Sécurité Sanitaire des Aliments (AFSSA), Laboratoire d’Etudes et de Recherches sur la Qualité des Aliments et les Procédés Agro-alimentaires (LERQAP), Maisons-Alfort, France
    Search for more papers by this author
  • C. Mazuet,

    1.  Institut Pasteur, Centre de référence des bactéries anaérobies, Paris, France
    Search for more papers by this author
  • S. Perelle,

    1.  Agence Française de Sécurité Sanitaire des Aliments (AFSSA), Laboratoire d’Etudes et de Recherches sur la Qualité des Aliments et les Procédés Agro-alimentaires (LERQAP), Maisons-Alfort, France
    Search for more papers by this author
  • M. Popoff

    1.  Institut Pasteur, Centre de référence des bactéries anaérobies, Paris, France
    Search for more papers by this author

Patrick Fach, Agence Française de Sécurité Sanitaire des Aliments (AFSSA), Laboratoire d’Etudes et de Recherches sur la Qualité des Aliments et les Procédés Agro-alimentaires (LERQAP), 23 avenue du général De Gaulle, 94700 Maisons-Alfort, France.
E-mail: p.fach@afssa.fr

Abstract

Aims:  To develop real-time PCR assays for tracking and tracing clostridia responsible for human botulism.

Methods and Results:  Real-time PCR assays based on the detection of the genes ntnh encoding the nontoxin-nonhaemagglutinin (NTNH) proteins or the most homologous regions of the botulinum neurotoxin (bont) genes have been developed together with four real-time PCR assays, each being specific of the genes bont/A, bont/B, bont/E, bont/F and enables a toxin type-specific identification. The specificity of the assays was demonstrated using a panel of botulinum toxin producing clostridia (29 strains), nonbotulinum toxin producing clostridia (21 strains) and various other bacterial strains. The toxin type-specific assays had a sensitivity of 100 fg–1000 fg of total DNA in the PCR tube (25–250 genome equivalents) which correspond to 103 to 104 cells ml−1. After a 48 h enrichment in anaerobic conditions, these PCR assays allowed the detection of Clostridium botulinum type A in a naturally contaminated sample of ‘foie gras’ suspected in a C. botulinum outbreak.

Conclusion:  These PCR tests are specific and reliable for detection of heterogeneous BoNT producing clostridia responsible for human botulism.

Significance and Impact of the Study:  Adoption of these PCR assays is a step forward a reliable and rapid detection of these clostridia in food samples.

Introduction

Botulinum neurotoxins (BoNT) are produced by phenotypically and genetically different Clostridium species including Clostridium botulinum, and some strains of Clostridium baratii (McCroskey et al. 1991) and Clostridium butyricum (McCroskey et al. 1986). BoNTs are the most potent biological and chemical substances known and are responsible for botulism, which is characterized by severe flaccid paralysis. BoNTs are divided into seven toxin types (A–G) according to their antigenic properties. Toxin types A, B, E and more rarely F cause human botulism, whereas toxin types C and D are mainly responsible for animal botulism (Popoff 1995; Herreros et al. 1999; Schiavo et al. 2000). Although less common, bivalent strains that express two different BoNT types exist and are designated by the predominant toxin produced (Ab, Ba, Af and Bf) (Gimenez and Gimenez 1993). Other bivalent strain variants such as the A1(B) strains contain both BoNT/A and BoNT/B genes but express only BoNT/A (Rodriguez Jovita et al. 1998). Sequencing of the BoNT genes from multiple strains of serotypes A, B, E and F showed significant sequence variation within each serotype. While serotypes differ by about 35–70%, subtype differences range from approximately 2–32%. Four distinct lineages within each of the BoNT A and B serotypes (Hill et al. 2007) and six distinct lineages of serotype E strains (Chen et al. 2007) were identified. The nucleotide sequences of the seven toxin genes of the serotypes were compared and showed various degrees of interrelatedness and recombination, as was previously noted for the nontoxin-nonhaemagglutinin (NTNH) gene, which is linked to the BoNT gene (Raphael and Andreadis 2007). The NTNH protein that is closely associated with the BoNT protein is present in all toxin clusters but its function is not fully understood. It is thought that this protein assists in stability of the neurotoxin within the acidic and protease-rich environment of the stomach and assists in transport of the toxin from the intestinal area to the bloodstream (Maksymowych et al. 1999).

Currently, the identification of C. botulinum (and other BoNT producing clostridia) requires the production of toxin in culture and testing by mouse bioassay with neutralization of the toxin with type-specific antitoxin. Since this process is time and labour consuming, we sought to develop rapid nucleic acid based assays to presumptively identify BoNT producing clostridia. The nucleotide diversity (identity ranges from 56% to 76%) among the BoNT genes (Hauser et al. 1995) present significant challenges in designing universal primers and/or probes that enable detection of these genes by real-time PCR. Nonetheless, BoNTs are generated as part of a progenitor toxin complex and a conserved component among serotypes is the NTNH. East and Collins (1994) demonstrated that the gene encoding NTNH is present in all strains that produce BoNTs and absent from strains that are nontoxic. Nucleotide sequence analysis of the cluster of genes associated with the BoNT gene demonstrates the presence of the NTNH gene directly upstream of the BoNT gene in all toxin types tested. Comparisons of the NTNH amino acid sequence from toxic strains of C. botulinum reveal a high level of similarity (amino acid identity ranges from 70% to 99%). Moreover, the BoNT and NTNH genes are likely cotranscribed as demonstrated by the identification of a transcript with a size that is approximately the sum of both genes (Henderson et al. 1996).

For tracking BoNT producing clostridia we designed two real-time PCR assays that target respectively the highly conserved regions of the BoNT and NTNH genes. Each test use a single fluorescently labelled probe to detect bacterial strains harbouring the BoNT gene cluster. Owing to the very low G+C content of C. botulinum and due to the high genetic diversity within the BoNT and NTNH genes we used some degenerate and locked nucleic acid (LNA) oligonucleotides. These PCR tests based on identification of the most conserved region of the BoNT and NTNH genes permit the simultaneous detection of C. botulinum A, B, E or F (and other BoNT producing clostridia). Tracing BoNT/A to BoNT/F producing clostridia, could be achieved with four different real-time PCR assays, each being specific for C. botulinum types A, B, E and F, and enables a toxin type-specific identification.

Materials and methods

Bacterial strains

The bacterial strains used in this study are listed in Table 1. Clostridium strains were maintained on tryptone-glucose-yeast extract (TYG) and stored at −80°C. Spores of Bacillus strains were produced as previously described (Claus and Berkeley 1986; Choma et al. 2000) and kept in a 30% (v/v) glycerol solution at −20°C.

Table 1.   Results of PCR-assays on pure cultures of Clostridium botulinum and other species
SpeciesToxin typeNumber of strains testedNTNH PCRBoNT PCRBoNT/ A PCRBoNT/ B PCRBoNT/ E PCRBoNT/ F PCR
  1. nt, not tested; NT, non BoNT producing strains.

Clostridium botulinum type AA8+++
Clostridium botulinum type BB8+++
Clostridium botulinum type ABAB2++++
Clostridium botulinum type EE4+++
Clostridium butyricum type EE2+++
Clostridium botulinum type FF1+++
Clostridium baratii type FF1+++
Clostridium botulinum type CC1nt
Clostridium botulinum type DD1nt
Clostridium botulinum type GG1+
Clostridium TetaniNT1
Clostridium ButyricumNT2
Clostridium BaratiiNT1
Clostridium SporogenesNT1
Clostridium SpirogenesNT1
Clostridium SubterminaleNT1
Clostridium SordelliiNT1
Clostridium SepticumNT1
Clostridium OedematiensNT1
Clostridium MangenottiNT1
Clostridium ChauvoeiNT1
Clostridium. BifermentansNT1
Clostridium BeijerinckiiNT1
Clostridium DifficileNT2
Clostridium PerfringensNT5
Bacillus cereusNT1
Bacillus ThuringiensisNT1
Streptococcus faecalisNT1

Cultures

Strains of Clostridium spp. were anaerobically grown overnight at 37°C by transfer of 100 μl TYG stock cultures to 10 ml of TYG broth. Strains were then subcultured for 9 h at 30°C by transferring 100 μl of the cultures to 10 ml of TYG broth. Cultures of Bacillus spp. strains were grown in J agar [J broth containing tryptone 5 g (Biokar, Beauvais, France), yeast extract 15 g (Biokar), K2HPO4 3 g, glucose 2 g, water 1 l, pH 7·4, with agar 15 g (Biokar)] (Choma et al. 2000). They were grown by transfer of a single 24–48 h colony into 10 ml of J broth.

Procedure for testing the specimen of ‘Foie gras’

Samples of 25 or 50 g of ‘foie gras’ were diluted 10-fold (w/v) in prereduced TYG broth containing 200 mg l−1 of d-cycloserine, as described by Sebald and Petit (1997) and under a gas flow of N2/H2 (95 : 5, v/v). After 48 h incubation at 37°C, a 40 ml aliquot of the enrichment broths was collected and kept frozen at −20°C for further mouse bioassay, and a 2 ml aliquot of the same enrichment broths was collected, centrifuged at 9000 g for 5 min, and the supernatant discarded. The cell pellet was mixed with the 1 ml of the EasyMag lysis buffer (Biomérieux, Marcy l’étoile, France) and store at 4°C until DNA extraction. DNA extraction was performed with the Nuclisens EasyMag bio-robot (Biomérieux) and following the instructions of the manufacturer. Extracted DNA was stored at −20°C until 5 μl were tested in the PCR reactions.

Mouse bioassay

The samples of ‘foie gras’ were characterized at Pasteur Institute (Reference Centre for anaerobic bacteria, Paris, France) by the standard mouse bioassay. Two ml of the enrichment culture (48 h) were centrifuged and 1 ml of the culture supernatant was incubated with 200 μg ml−1 trypsin for 20 min at room temperature. A volume of 0·5 ml of 10-fold serial dilutions was then injected intraperitoneally into Swiss male mice (two mice per sample) and the mice were watched for the characteristic symptoms of botulism (laboured breathing, pinching of the waist and paralysis) and for death for 4 days. Botulinum toxins were confirmed and types were identified by a seroneutralization test on mice using specific botulinum antitoxins (Pasteur Institute, Paris, France) (Fach et al. 1995).

Primers and probes

Primers and probes were designed by alignment of the BoNT gene sequences from C. botulinum, C. baratii and C. butyricum using the multalin program ClustalW (http://align.genome.jp/). The most highly homologous regions of the sequences allowed a set of universal primers and one probe to be designed for simultaneous detection of the BoNT/A, BoNT/B, BoNT/E, BoNT/F and BoNT/G genes in real-time PCR (Table 2). Four primer pairs and four probes with each being specific for either BoNT/A, BoNT/B, BoNT/E or BoNT/F were designed (Table 2). They were selected from the nonhomologous regions of the BoNT/A, BoNT/B, BoNT/E and BoNT/F genes to enable a toxin type-specific identification. Finally, one primer pair and one probe were designed in the most highly homologous regions of the NTNH gene sequences (Table 2). In order to permit mismatches within some primers or probes site, we used LNA bases which bracket the bases in some oligonucleotide sequences that showed degeneracy among strains in order to increase specificity for the conserved nucleotides. All probes were 5′-labelled with 6-carboxylfluorescein (FAM) and 3′-labelled with Black Hole Quencher (BHQ), except for the NTNH probe which was 3′-labelled with a minor groove binder ligand (MGB).

Table 2.   Primers and probes
Oligonucleotide* DNA sequences (5′→3′)Annealing temperature (°C)Size of PCR product (bp)
  1. *In the sequence of oligonucleotides W is (A,T); R is (A,G); K is (G,T); Y is (C,T); H is (A,T,C); D is (G,A,T) and M is (A,C). LNA monomers are underlined. FAM, 6-carboxylfluorescein; BHQ, Black Hole Quencher; MGB, Minor Groove Binder.

A1GGAGTCACTTGAAGTTGATACAAATC6075
A2GCTAATGTTACTGCTGGATCTGTAG60
A3FAM-TCTTTTAGGTGCAGGCAAATTT-BHQ60
B1GATGAACAGCCAACATATAGTTGTCA60172
B2GTTTCCTTTTTACCTCTTTTAAGTACCATT60
B3FAM-TGATGAKATAGGATTGATTGGTATTCA-BHQ60
E1CTATCCAAAATGATGCTTATATACCAAA60115
E2GGCACTTTCTGTGCATCTAAATA60
E3FAM-ATGATTCTAATGGAACAAGTGATATAGAACAACATGATGT-BHQ60
F1GCAATATAGGATTACTAGGTTTTCATTC60112
F2GAAATAAAACTCCAAAAGCATCCATT60
F3FAM-TTGGTTGCTAGTAGTTGGTATTATAACAA-BHQ60
U1AATAATTCDGGMTGGAAARTATC50187–200
U2CCATTDATRTAAAKTYTAG50
U3FAM-ATYATTAGTDATAGTTACAAAAAWCCA-BHQ50
NT1ARTGGWATGGGAACHATG5080
NT2GCWGGATCAAYATARAATT50
NT3FAM-CAACCATTTYTAACMYATAAATA-MGB50

PCR conditions

Universal PCR tests targeting the most highly homologous regions of either the BoNT or NTNH genes were performed in a total volume of 25 μl of 1 × concentration of a LightCycler-Faststart DNA master hybridization probes mix (Roche Diagnostics, Meylan, France), containing, 4·2 mmol l−1 MgCl2, 0·9 μmol l−1 of each primer, 0·4 μmol l−1 of probe and 5 μl of template DNA. Amplifications were carried out on the ABI Prism 7700 Sequence Detection System (Applied Biosystems) according to the following temperature profile: one cycle of 95°C for 10 min, 5 cycles of 95°C for 15 s, 45°C for 20 s with 2·5 min transition time between 45°C and 95°C, and 35 cycles of 15 s at 95°C and 20 s at 50°C with 2·5 min transition time between 50 and 95°C.

BoNT type-specific amplifications were performed in a total volume of 25 μl of 1 × qPCR master mix buffer from Eurogentec (Liege, Belgium), containing, 0·3 μmol l−1 of each primer, 0·1 μmol l−1 of probe and 5 μl of template DNA. Amplifications were carried out on an ABI Prism 7700 Sequence Detection System (Applied Biosystems, Courtaboeuf, France) according to the following temperature profile: one cycle of 95°C for 10 min and 35 cycles of 15 s at 95°C and 1 min at 60°C.

Positive controls and two negative controls containing all reagents except DNA template were included with each amplification set. To avoid contamination, sample preparation, PCR amplification and PCR detection were performed in different rooms. For each PCR cycle, the fluorescence signal of each reporter dye was reported as Rn. ΔRn was Rn minus the baseline reporter dye intensity established in the first few cycles. At the end of the PCR, a reaction was considered positive if its ΔRn curve exceeded the threshold, defined as 10 times the standard deviation of the mean baseline emission calculated between the 3rd and 15th cycles. The cycle threshold (Ct) was defined as the cycle number at which a sample’s ΔRn fluorescence crossed the determined threshold value of 0·5.

Sensitivity of the real-time PCR

The limit of detection (LOD) of the PCR test based on universal BoNT primers and probe was determined with either total genomic DNA or recombinant plasmids. Genomic DNA from C. botulinum types A (strain Hall A), B (strain IPBL6), E (strain HV) and F (strain NCIMB 10658) was extracted and purified as previously described by Popoff et al. (1985) and quantified by spectrophotometry. The recombinant plasmids which contained amplified fragments from bont/A, bont/B, bont/E and bont/F genes were those described by Fach et al. (2002). Plasmid DNA concentrations were determined by fluorimeter and the copy number of each plasmid was calculated according to its size.

The LOD of the four toxin type-specific PCR tests, each being specific for either BoNT/A, BoNT/B, BoNT/E or BoNT/F was determined with either serial dilutions of total purified genomic DNA or with of serial dilutions of C. botulinum broth cultures which were counted microscopically in a Petrov chamber before DNA extraction with the InstaGene™ Matrix (Bio-Rad Laboratories, Marnes-La-Coquette, France), as previously described by Fach et al. (2002).

The LOD of the universal PCR test targeting the NTNH genes was determined with of serial dilutions of C. botulinum broth cultures which were counted microscopically in a Petrov chamber before DNA extraction with the InstaGene™ Matrix (Bio-Rad), as previously described by Fach et al. (2002).

The LOD of the different PCR assays was determined by performing at least three triplicate for each dilution level. In addition, each DNA extract corresponding to one dilution level has been tested in PCR in duplicate, so that at least 6 Ct values were obtained for each concentration level.

Results

Specificity

Purified genomic DNA templates from C. botulinum type A (n = 8), C. botulinum type B (n = 8), C. botulinum type AB (n = 2), C. botulinum type E (n = 4), and C. botulinum type F (n = 1) yielded expected specific positive signals by the four toxin type-specific PCR tests (Table 1). The assays tested negative strains expressing toxin types C, D, and G. It should be noted that type G strains are extremely rare and only one of two reported strains was incorporated in our panel. The toxigenic nonbotulinum Clostridium spp., C. butyricum type E (n = 2) and C. baratii type F (n = 1) were tested positive for the presence of BoNT/E and BoNT/F genes respectively. No amplification was observed when nonbotulinum toxin expressing Clostridium spp. (21 strains) or other bacteria (three strains) DNA was used as template. To confirm that botulinum toxin was expressed by the toxic clostridial strains used in this study, mouse neutralization bioassay of culture supernatants was performed in Pasteur Institute (data not shown).

The universal PCR tests targeting the most highly homologous regions of either the BoNT or NTNH genes yielded the expected positive signals for C. botulinum types A, B, AB, E or F and for the toxigenic nonbotulinum Clostridium spp., C. butyricum type E and C. baratii type F. Additionally, the NTNH PCR assay tested positive strains expressing toxin G. None of the C. sporogenes or C. botulinum-like strains or other bacterial species yielded a positive signal by these two PCR assays.

Sensitivity

Purified genomic DNA from C. botulinum types A (strain Hall A), B (strain IPBL6), E (strain HV) and F (strain NCIMB 10658) was quantified by spectrophotometry and converted to genomic copy number by assuming that the size of each strain’s genome is equal to that of the sequenced C. botulinum type A Hall strain (3·9 Mbp) (http://www.sanger.ac.uk/Projects/C_botulinum/). Using purified genomic DNA as template, limits of detection was determined by testing serial 10-fold dilutions of DNA in duplicate. The LOD for each toxin type ranged between 250 and 2500 genome copies by using the universal BoNT PCR test (Table 3). These data correlate with the LOD obtained with the recombinant plasmids which was 500 copies for BoNT/A, 250 copies for BoNT/B, 350 copies for BoNT/E and 450 copies for BoNT/F (data not shown). The LOD as determined with genomic DNA from C. botulinum strains was 10-fold improved and ranged between 25 and 250 genome copies when using the four toxin type-specific PCR tests. The LOD was lowest for type A strain Hall (25 genome copies) and highest for type E strain HV (250 genome copies) (Table 3).

Table 3.   Sensitivity of the real-time PCR assays using dilutions of purified genomic DNA as template
StrainToxin typeToxin type-specific PCRUniversal BoNT PCR
Purified genomic DNAPurified genomic DNA
fg DNAGenome equivalents*fg DNAGenome equivalents*
  1. *Genome equivalents are calculated based on the mass of the C. botulinum type A Hall strain as described in the text.

Clostridium botulinum A (strain Hall)A100251000250
Clostridium botulinum B (strain BL6)B100–100025–2501000250
Clostridium. botulinum E (strain HV)E100025010 0002500
Clostridium botulinum F (strain NCIMB 10 658)F100–100025–25010 0002500

The LOD of the four toxin type-specific PCR tests was further determined with triplicates of serial dilutions of C. botulinum broth cultures which were counted microscopically in a Petrov chamber before DNA extraction. Each DNA triplicate was tested in double in the PCR tests so that 6 Ct values were obtained for each bacterial dilution. This allowed to determine mean Ct values for each bacterial dilution, the LOD and the probability of detection of the methods (Fig. 1). Concentrations of 103 cells ml−1 of C. botulinum type A, B and F (approximately 25 copies in the PCR tube) and higher concentrations yielded positive results and 100% probability of detection. Usually, the mean Ct values were around 30 for the concentration of 103 cells ml−1. For concentrations under 103 cells ml−1 (less than 25 copies in the PCR tube) the mean Ct values were higher than 30 and the probability of detection was lower than 100%. For C. botulinum type E, 104 cells ml−1 and higher concentrations tested positive with 100% probability of detection (Fig. 1).

Figure 1.

 Sensitivity of the real-time PCR assays using dilutions of C. botulinum broth cultures as template. The LOD of the four toxin type-specific PCR tests was determined with at least triplicates of serial dilutions of C. botulinum broth cultures which were counted microscopically in a Petrov chamber before DNA extraction. Each DNA extract corresponding to one dilution level has been tested in PCR in duplicate, so that at least 6 Ct values were obtained for each concentration level. This allowed to determine mean Ct values (Axis Y, left) and standard deviation for each bacterial dilution and the LOD which is expressed as a probability of detection (Axis Y, right) for each concentration level (Axis X: number of Clostridia per ml). (inline image, Probability; inline image, Ct values).

The LOD of the NTNH PCR test was determined using serial 10-fold dilutions of C. botulinum broth cultures as template. Each DNA extract was tested in duplicate. The LOD for each toxin type ranged between 3 and 25 genome copies in the PCR tube. The LOD was lowest (three genome copies) for strains of C. botulinum type A, B and E and highest for C. botulinum type F (25 genome copies). Taken together, our results indicate that the sensitivity level of the NTNH PCR allows the detection of few ntnh gene copies (less than 25 gene copies) in the PCR tube (data not shown).

Detection of C. botulinum in a specimen of ‘Foie gras’

In order to test the applicability of the PCR assays in investigation of food we have tested one specimen of ‘foie gras’ which has been collected during a food outbreak occurring in France in December 2007. After a 48 h enrichment in anaerobic conditions and DNA extraction using the Nuclisens EasyMag bio-robot, the sample tested positive by universal PCR tests targeting the most highly homologous regions of either the BoNT or NTNH genes. In addition, it yielded also a positive signal with BoNT/A type-specific PCR test (Fig. 2). The specimen has been confirmed positive for C. botulinum type A with the reference lethality test on mice. After anaerobic enrichment, the estimated titre of toxin in the specimen of ‘foie gras’ was around 10 000 LD ml−1.

Figure 2.

 Detection of Clostridium botulinum type A in naturally contaminated Foie gras enriched in anaerobic conditions and tested in duplicate with BoNT real-time PCR, BoNT/A real-time PCR and NTNH real-time PCR assays. Results in graphics are expressed in fluorescence intensity (ΔRn) against the number of PCR cycles.

Discussion

BoNT are produced by phenotypically and genetically different clostridia belonging to the species Clostridium botulinum, C. butyricum, and C. baratii. Laboratory detection and identification of BoNT-producing clostridia on the basis of bacteriological characteristics are difficult. Currently, the most sensitive standard method of BoNT detection is the mouse bioassay. However, this technique is time consuming and requires handling of laboratory animals. A rapid and specific test for BoNT-producing Clostridium detection is needed for food quality control and investigations of suspected food-borne outbreaks.

In order to develop real-time PCR assays targeting the BoNT and the NTNH genes a data bank collecting all available DNA sequences encoding for bontA, bontB, bontE and bontF and ntnh has been achieved. It includes the sequences of 4 BoNT/A variants (A1, A2, A3 and A4), 7 BoNT/B variants (proteolytic group: B1, B2, B3, Ba, Bf and Ab; nonproteolytic group : B), 6 BoNT/E variants (4 for C. botulinum : E1, E2, E3, E6 and 2 for C. butyricum :E4, E5), 4 BoNT/F variants (including the variant from C. baratii) and all available NTNH gene sequences. Alignment of the BoNT gene sequences and identification of the most conserved regions within the BoNT genes allowed a set of primers and one single probe to be designed for screening any C. botulinum types A, B, AB, E or F and the toxigenic nonbotulinum Clostridium spp., C. butyricum type E and C. baratii type F. The real-time PCR assay based on the highly homologous regions of the NTNH gene tested positive all these strains and additionally the strain expressing toxin G. None of the C. sporogenes or C. botulinum-like strains or other bacterial species yielded a positive signal by these two PCR assays.

Four toxin type-specific PCR assays with each being specific for either BoNT/A, BoNT/B, BoNT/E or BoNT/F were designed from the nonhomologous regions of the BONT genes and yielded expected specific positive signals with C. botulinum A, B, AB, E or F and C. butyricum type E and C. baratii type F. They tested negative strains expressing toxin types C, D or G and the nonbotulinum toxin expressing Clostridium spp. or the other bacteria analysed in this study.

The sensitivity of the PCR assays was variable between each real-time PCR test. Using purified genomic DNA from C. botulinum types A (strain Hall A), B (strain IPBL6), E (strain HV) and F (strain NCIMB 10658) as template, the LOD of the universal BoNT PCR test ranged between 250 and 2500 genome copies. It was 10-fold improved and ranged between 25 and 250 genome copies when using the four- toxin type-specific PCR tests. The LOD was lowest for type A strain Hall (25 genome copies) and highest for type E strain HV (250 genome copies). The LOD of the four toxin type-specific PCR tests was further determined with serial dilutions of C. botulinum calibrated broth cultures. For C. botulinum types A, B and F, concentrations of 103 cells ml−1 (approximately 25 copies in the PCR tube) and higher concentrations yielded positive results and 100% probability of detection whereas for C. botulinum type E, 104 cells ml−1 and higher concentrations tested positive with 100% probability of detection.

The sensitivity of the NTNH PCR test as determined using serial dilutions of C. botulinum calibrated broth cultures as template ranged between 3 and 25 genome copies in the PCR tube. The LOD was lowest (three genome copies) for strains of C. botulinum type A, B and E and highest for C. botulinum type F (25 genome copies).

In order to test the applicability of the PCR assays in investigation of food we tested one specimen of ‘foie gras’ collected during a food outbreak occurring in France in December 2007. After enrichment in anaerobic conditions and DNA extraction, the sample tested positive by universal PCR tests targeting the most highly homologous regions of either the BoNT or NTNH genes. In addition, it yielded also a positive signal with BoNT/A type-specific PCR test and when tested with the reference lethality test on mice. These results indicate that the NTNH and BoNT real-time PCR assays can be used to screen enrichment cultures of primary food specimens. The strategy which combines detection of BoNT and NTNH genes involved, respectively in BoNT and NTNH proteins synthesis is an innovative and efficient approach for the development of a rapid and optimal screening/typing methods for testing BoNT producing clostridia. Such method should be helpful for the routine monitoring of BoNT producing Clostridia contamination into food or water and for the risk evaluation of these clostridia pollution along the food chain.

Acknowledgements

This work was partially supported by the BIOTRACER European project (Contract no 036272) under the sixth framework program priority 5 ‘Food Quality and Safety’.

Ancillary