INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Type E botulinum toxin (BoNT/E)-producing Clostridium butyricum was first isolated from two cases of infant botulism in Italy in 1984 (1, 8). In 1997, we isolated BoNT/E-producing C. butyricum from the food implicated in food-borne botulism in China (10). Because our results indicated that type E food-borne botulism can be caused by BoNT/E-producing C. butyricum, we reexamined the cultural and biochemical properties of BoNT/E-producing organisms that had previously been isolated from type E food-borne botulism cases and found that two isolates were identifiable as C. butyricum (9). In addition, we isolated several strains of BoNT/E-producing C. butyricum from soil specimens of China (9). In 1998, an outbreak of food-borne botulism was reported in India and was strongly suggested to be caused by BoNT/E-producing C. butyricum (2). These studies indicate that soil is the principal habitat of BoNT/E-producing C. butyricum and that this organism may be widely distributed throughout the world (9). For improved surveillance of BoNT/E-producing C. butyricum, biochemical and genetic analysis of this organism is required.

In this study, we performed molecular analysis of the strains isolated in Italy and China, by using nucleotide sequencing of the bont/E gene, random amplified polymorphic DNA (RAPD) assay, pulsed-field gel electrophoresis (PFGE), and Southern blot hybridization for the bont/E gene.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains. Thirteen strains of BoNT/E-producing C. butyricum (BL 5262, BL 6340, LCL 063, LCL 095, LCL 155, KZ 1899, KZ 1897, KZ 1898, KZ 1886, KZ 1887, KZ 1889, KZ 1890, and KZ 1891) (see Table 2) and two strains of nontoxigenic C. butyricum (IFO 13949 and IFO 3315) were used in this study. BL 5262 and BL 6340 were isolated from two cases of infant botulism reported in Rome, Italy (8). BL 5262 is equivalent to BL 5839 and ATCC 43181, and BL 6340 is equivalent to BL 5520 and ATCC 43755 (C. L. Hatheway, personal communication). LCL 063 and LCL 095 were isolated from two cases of food-borne botulism in Jining, Shandong province, and Peixian, Jiangsu province, respectively, in China (9). KZ 1899 and LCL 155 were isolated from the food implicated in a case of food-borne botulism in Guanyun, Jiangsu province, in China (9, 10). KZ 1897 and KZ 1898 were isolated from soil specimens collected from a site around the home of the patients in the Guanyun case (9). KZ 1886, KZ 1887, KZ 1889, KZ 1890, and KZ 1891 were isolated from soil specimens from the Weishan lake area in China (9). Guanyun, Jining, and Peixian are, in a broad sense, located in the Weishan lake area. A neurotoxigenic C. butyricum strain from the Indian outbreak (2) could not be obtained.

Extraction of whole-cell DNA. All test strains were inoculated in 10 ml of brain heart infusion (BHI) broth (BBL Becton Dickinson and Company, Cockeysville, Md.) and cultured at 37°C overnight. The cultures were centrifuged at 15,000 × g for 15 min to collect cells. The cells were resuspended with 400 µl of TE buffer (10 mM Tris [pH 7.4], 1 mM EDTA), incubated at 37°C for 15 min with 25 U of mutanolysin (Nacalai Tesque, Kyoto, Japan), and subsequently digested with 25 µl of proteinase K (20 mg/ml) for 15 min. The cells were then incubated with 1% sodium dodecyl sulfate and 1 µl of RNase (10 mg/ml) at 37°C for 15 min. The cell lysate was treated with an equal volume of phenol and subsequently with an equal volume of chloroform-isoamyl alcohol (24:1). The DNA was precipitated with isopropanol, rinsed with 70% ethanol, and finally resolved with 200 µl of TE buffer.

Sequencing of the bont/E gene. The nucleotide sequences of the bont/E genes were determined for 11 strains isolated in China and C. butyricum BL 6340. PCR primers KAG165 (5' CAAGATTACAATTGGGTTATATGTGATCTTAATCATGA 3') and KAG166 (5' CTAAGTCCTTTGGAATTTATGACTTTAGCCGT 3') were designed to amplify the whole open reading frame of the bont/E gene based on data for the bont/E gene sequences (13, 14). The PCR mixture consisted of 0.2 mM (each) deoxynucleoside triphosphates, 50 pmol of each primer, 1 µg of whole-cell DNA, and 2.5 U of TaKaRa Ex Taq polymerase (Takara Shuzo, Otsu, Japan) in a 50-µl volume of Ex Taq buffer (Takara Shuzo). The PCR was carried out by using a GeneAmp PCR System 9700 (PE Applied Biosystems, Foster City, Calif.) and a two-step procedure of 30 cycles at 94°C for 20 s and 70°C for 2 min preceded by preheating at 94°C for 1 min. The amplified DNA fragments were purified with a QIAquick PCR Purification kit (Qiagen GmbH, Hilden, Germany). The purified PCR products were sequenced in both directions by using a BigDye Terminator Cycle Sequencing Ready Reaction kit (PE Applied Biosystems) with synthetic primers and electrophoresed on an ABI Prism 310 Genetic Analyzer (PE Applied Biosystems).

RAPD assay. RAPD assay was performed by using Ready-To-Go RAPD Analysis Beads (Amersham Pharmacia Biotech Inc., Piscataway, N.J.). Primer 1 (5'-GGTGCGGGAA-3') and primer 6 (5'-CCCGTCAGCA-3') were selected by comparing the amplification patterns in each of six primers offered by the manufacturer. The PCR was carried out in the tube supplied using 25 pmol of primer and 10 ng of the purified DNA in a total volume of 25 µl, which was overlaid with 50 µl of mineral oil. A PCR profile (preheating at 95°C for 5 min followed by 45 cycles at 95°C for 1 min, 36°C for 1 min, and 72°C for 2 min) was performed on a TouchDown Thermocycler (Hybaid Ltd., Ashford, United Kingdom). Ten microliters of PCR product was analyzed by 3% agarose gel electrophoresis on a horizontal electrophoresis unit (Mupid-2; Cosmo Bio. Co. Ltd., Tokyo, Japan).

PFGE. All test strains were cultured in 10 ml of BHI broth at 37°C for 12 h. Cells were collected from 1 ml of BHI cultures by centrifugation at 15,000 × g for 3 min. The cells were resuspended in 100 µl of a suspension buffer (10 mM Tris [pH 7.2], 50 mM EDTA, 20 mM NaCl) and mixed with 100 µl of 1.2% low-melting-temperature agarose (FMC BioProducts, Rockland, Maine). One hundred microliters of the mixture was allowed to solidify in a plug mold (Bio-Rad Laboratories, Hercules, Calif.). The embedded cells were lysed at 37°C for 5 h in 500 µl of a lysing buffer (10 mM Tris [pH 7.2], 100 mM EDTA, 50 mM NaCl, 0.2% sodium deoxycholate, 0.5% sodium laurylsarcosine, 1 mg of lysozyme per ml, and 20 U of mutanolysin per ml). The plugs were rinsed with 1 ml of a wash buffer (20 mM Tris [pH 8.0], 50 mM EDTA) and were digested with 1 mg of proteinase K per ml in a proteinase K buffer (100 mM EDTA [pH 8.0], 0.2% sodium deoxycholate, 1% sodium laurylsarcosine) at 50°C overnight. To inactivate the proteinase K, the plugs were washed with 1 ml of the wash buffer containing 1 mM phenylmethylsulfonyl fluoride for 1 h with gentle shaking and subsequently washed with 1 ml of the wash buffer for 30 min three times. Before digestion by restriction endonucleases, the plugs were washed with each restriction endonuclease buffer for 30 min with gentle shaking. The plugs were digested with SmaI or XhoI (Takara Shuzo). The digestion was performed for 20 h in 400 µl of the optimal buffer and at the optimal temperature recommended by the manufacturer. All samples were electrophoresed with a CHEF-DR II (Bio-Rad Laboratories) apparatus through a 1% Pulsed-Field Certified Agarose gel (Bio-Rad Laboratories) in 0.5× Tris-borate-EDTA buffer at 14°C and 6 V/cm (200 V). Switch times were ramped from 3 to 20 s for the SmaI digestion and from 3 to 25 s for the XhoI digestion. The gel was stained with 1 µg of ethidium bromide per ml for 30 min and destained in distilled water for 30 min.

Southern blot analysis for the bont/E gene. After PFGE, DNA fragments in the gel were transferred onto Hybond N⁺ nylon membranes (Amersham Pharmacia Biotech Inc.) by an alkaline transfer procedure according to the manufacturer's instructions. A DNA fragment containing the whole open reading frame of the bont/E gene was amplified from LCL 155 with PCR by using primers KAG165 and KAG166 and was purified as described in "Sequencing of the bont/E gene" above. The purified fragment was labeled with alkaline phosphatase by using a Gene Images AlkPhos Direct Labeling and Detection System (Amersham Pharmacia Biotech Inc.) and was used as a DNA probe. Hybridization was performed at 55°C for 16 h. Detection of the hybrids was carried out by using an ECF chemifluorescent signal generation system (Amersham Pharmacia Biotech Inc.) according to the manufacturer's instructions. A FluorImager SI (Amersham Pharmacia Biotech Inc.) was used to obtain the chemifluorescent image.

Nucleotide sequence accession number. The sequence data reported in this paper have been submitted to the DDBJ/EMBL/GenBank nucleotide sequence databases under the accession no. AB037704 to AB037714 and AB039264.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

The bont/E gene from BoNT/E-producing C. butyricum. The nucleotide and deduced amino acid sequences of BoNT/E products of 11 C. butyricum isolates from China were identical but differed slightly from those of BoNT/E from either Clostridium botulinum NCTC 11219 (14) or C. butyricum BL 6340 (Fig. 1). The deduced amino acid sequence of BoNT/E from the Chinese isolates showed 96.9 and 95.0% identity with those of BoNT/E products from NCTC 11219 and BL 6340, respectively. Between NCTC 11219 and BL 6340, identity of the deduced amino acid sequences of BoNT/E was 97.4% (Table 1).

View larger version (71K): [in this window] [in a new window]   FIG. 1.   Comparison of the deduced amino acid sequences of BoNT/E derived from C. butyricum LCL 155 and BL 6340 and from C. botulinum NCTC 11219 and Beluga. Dots in the sequences of BL 6340 (accession no. AB039264), NCTC 11219 (accession no. X62683) (14), and Beluga (accession no. X62089) (13) indicate that these residues are identical to those of LCL 155 (accession no. AB037704). The light (L) and heavy (H) chains are marked by arrows.

RAPD assay. The BoNT/E-producing C. butyricum strains fell into three clusters on visual inspection according to the results of RAPD assay performed by using either primer 1 or primer 6 (Fig. 2). The first cluster comprised two strains isolated from infant botulism patients in Italy (BL 5262 and BL 6340), and these two strains showed an identical band pattern when either primer was used: profiles were designated I for primer 1 and i for primer 6 (Table 2). The second cluster comprised isolates associated with three outbreaks of food-borne botulism in China (LCL 155, KZ 1899, KZ 1897, KZ 1898, LCL 063, and LCL 095), and these isolates also showed an identical band pattern using either primer: profiles were designated II for primer 1 and ii for primer 6 (Table 2). The third cluster comprised strains isolated from soil specimens of the Weishan lake area, which exhibited an identical band pattern with primer 6 (Fig. 2B) (RAPD profile iii in Table 2) but slightly different patterns with primer 1 (Fig. 2A) (RAPD profiles III-1 and III-2 in Table 2). Thus, the third cluster was divided into subclusters 1 (KZ 1886, KZ 1889, and KZ 1891) and 2 (KZ 1887 and KZ 1890). Nontoxigenic C. butyricum strains showed different band patterns with both primers and no similarity with any strains of BoNT/E-producing C. butyricum.

View larger version (137K): [in this window] [in a new window]   FIG. 2.   RAPD assay. (A) RAPD assay with primer 1; (B) RAPD assay with primer 6. Lanes 1 and 2, nontoxigenic C. butyricum IFO 13949 and IFO 3315, respectively. Lanes 3 and 4, BoNT/E-producing C. butyricum isolated from two cases of infant botulism (BL 5262 and BL 6340, respectively). Lanes 5 to 10, BoNT/E-producing C. butyricum associated with food-borne botulism in China (LCL 155, KZ 1899, KZ 1897, KZ 1898, LCL 063, and LCL 095, respectively). Lanes 11 to 15, BoNT/E-producing C. butyricum isolated from the Weishan lake area (KZ 1886, KZ 1887, KZ 1889, KZ 1890, and KZ 1891, respectively).

PFGE. In preliminary experiments, four kinds of restriction endonucleases (ApaI, SacII, SmaI, and XhoI) were used for digestion of chromosomal DNA in PFGE. ApaI and SacII were not used in the subsequent experiments because the number of bands resulting from the digestion with these endonucleases was so low that the test strains could not be distinguished clearly (data not shown). However, distinguishable PFGE patterns could be obtained by digestion with SmaI or XhoI (Fig. 3).

View larger version (99K): [in this window] [in a new window]   FIG. 3.   PFGE and Southern blot analysis for the bont/E gene. (I) PFGE of genomic DNA with SmaI digestion (A), XhoI digestion (B), and nondigestion (C). (II) Southern blotting after PFGE with SmaI digestion (A), XhoI digestion (B), and nondigestion (C). Lanes 1 and 2, nontoxigenic C. butyricum IFO 13949 and IFO 3315, respectively. Lanes 3 and 4, BoNT/E-producing C. butyricum isolates from two cases of infant botulism (BL 5262 and BL 6340, respectively). Lanes 5 to 10, BoNT/E-producing C. butyricum isolates associated with food-borne botulism in China (LCL 155, KZ 1899, KZ 1897, KZ 1898, LCL 063, and LCL 095, respectively). Lanes 11 to 15, BoNT/E-producing C. butyricum isolates from the Weishan lake area (KZ 1886, KZ 1887, KZ 1889, KZ 1890, and KZ 1891, respectively). The switch times were ramped from 3 to 20 s for 18 h at 200 V for the SmaI digestion and nondigestion and from 3 to 25 s for 20 h at 200 V for the XhoI digestion.

Southern blot analysis for the bont/E gene. Nondigested chromosomal DNAs of the toxigenic strains tested hybridized with a DNA probe for the bont/E gene (Fig. 3I-C and II-C). However, some other smaller DNAs, which were probably plasmids, shown in some strains (about 120 kbp in IFO 13949, BL 5262, and BL 6340 and 170 kbp in KZ 1887 and KZ 1890) did not hybridize with the probe (Fig. 3I-C and II-C). A 220-kbp fragment generated by SmaI digestion hybridized with the probe in the cluster of infant botulism isolates (Fig. 3I-A and II-A). In strains isolated from food-borne botulism outbreaks and soil specimens in China, a large identical fragment (larger than 388 kbp) generated by SmaI digestion hybridized with the probe (Fig. 3I-A and II-A). A 160-kbp fragment generated by XhoI digestion hybridized with the probe in the two clusters of infant botulism isolates and of soil specimen isolates from the Weishan lake, but a fragment of 380 kbp (LCL 155, KZ 1899, KZ 1897, KZ 1898, and LCL 063) or 420 kbp (LCL 095) hybridized in the cluster associated with the food-borne botulism in China (Fig. 3I-B and II-B).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The complete nucleotide sequence of the bont/E gene was determined previously for two strains of C. botulinum type E, NCTC 11219 (14) and Beluga (13), and for two strains of BoNT/E-producing C. butyricum, BL 5262 and BL 6340 (13). Regarding the bont/E gene from BL 6430, we detected three nucleotide insertions leading to one amino acid change (N₁₁₉₅, in Fig. 1) that were not found in the previous study (13). The deduced amino acid sequences of C. botulinum NCTC 11219 and Beluga are 99.4% homologous (Table 1). In contrast, the identity is 95.0% between C. butyricum Chinese isolates and C. butyricum BL 6340. The botulinum neurotoxin is composed of a light chain and a heavy chain linked by a disulfide bond (12). The light chain possesses zinc protease activity, while the N-terminal and C-terminal regions of the heavy chain are responsible for translocation and receptor binding, respectively (12). The nucleotide and deduced amino acid sequences of the light chain from the Chinese isolates were completely identical to those from NCTC 11219 (Fig. 1), although the amino acid sequence of the light chain from BL 6340 was different by 17 residues from the sequence of that from NCTC 11219. However, in the heavy chain, mainly in the C-terminal region, amino acid changes were found between the Chinese isolates and NCTC 11219; 39 residues, far more than were different (16 residues) between BL 6340 and NCTC 11219, were changed throughout the heavy chain (Fig. 1).

There is controversy regarding the location of the bont/E gene of BoNT/E-producing C. butyricum. Fujii et al. (4) and Zhou et al. (15) demonstrated that the bont/E gene of BoNT/E-producing C. botulinum was chromosomally located. However, Hauser et al. reported that the bont/E gene was located on a large plasmid (up to 100 kb), based on the preferential PCR amplification of the bont/E gene in plasmid DNA compared with chromosomal DNA (5). In the present study, extrachromosomal DNAs, which were likely to be plasmids, were indeed found: they were electrophoresed at around 120 to 170 kb in the absence of digestion in IFO 13949, BL 5262, BL 6340, KZ 1887, and KZ 1890 (Fig. 3I-C). However, the bont/E gene probe hybridized with chromosomal, but not with extrachromosomal, DNAs (Fig. 3II-C). Our results support the chromosomal localization of the bont/E gene in BoNT/E-producing C. butyricum.

In this study, 13 BoNT/E-producing C. butyricum strains were divided into three genetic clusters based on the profiles produced by RAPD assay (Fig. 2), PFGE (Fig. 3I-A and 3I-B), and Southern hybridization with the bont/E gene probe (Fig. 3II-A and 3II-B). Interestingly, the division into three clusters could also be used for the fermentation patterns of arabinose and inulin (Table 2). In a previous study, PFGE revealed extensive genetic diversity among C. botulinum type E isolates derived from trout farms, even when the isolates came from the same farms (7). And indeed, genetic diversity was shown among BoNT/E-producing C. butyricum strains in this study, but it was not as extensive as that among the C. botulinum type E isolates. Genetic profiles of BoNT/E-producing C. butyricum strains in the same cluster were fairly homologous, suggesting that each cluster consists of a clone of BoNT/E-producing C. butyricum. Therefore, the geographical distribution of BoNT/E-producing C. butyricum suggested from the present study is as follows: (i) two clones are distributed over the Weishan lake area, and one of them appears to be tightly associated with food-borne botulism; and (ii) one clone is distributed in Italy. Recently, two unrelated cases of intestinal toxemia botulism caused by C. butyricum were reported in Italy (3). Based on PFGE analysis and analysis of antibiotic susceptibility, the two BoNT/E-producing C. butyricum strains isolated from these two cases were indistinguishable from the two C. butyricum strains that had caused type E infant botulism in Italy (3). These findings also support the idea that BoNT/E-producing C. butyricum has a wide clonal distribution. It should be noted that the division of the Chinese strains into two genetic clusters could also be used for the presence or absence of the association with food-borne botulism. Further analysis of differences between the natures of these two clusters (clones) should contribute to elucidation of the pathogenesis of BoNT/E-producing C. butyricum.

The present study showed that genetic profiling is useful for epidemiological surveys of BoNT/E-producing C. butyricum. Previously, we examined soil specimens of the Weishan lake area for type E botulinum toxicity and isolated the responsible organisms. All organisms that exhibited type E toxicity were shown to be C. butyricum, and we were unable to isolate C. botulinum type E (9). Moreover, two stock strains with type E botulinum toxicity, which had been isolated from food-borne botulism cases in the Weishan lake area, were also shown to be C. butyricum (9). In this area, type E botulism is predominant among human botulism cases (data not shown). It seems reasonable to conclude that almost all of the type E botulism cases in this area were caused by BoNT/E-producing C. butyricum.

RAPD assay and PFGE are two powerful tools that have been extensively used in the molecular typing of causative bacteria in outbreaks and in epidemiological surveys (11). In this study, both RAPD assay and PFGE gave reproducible results. PFGE was more sensitive for typing of the isolates than was RAPD assay. However, some nontoxigenic strains could be analyzed by RAPD assay but not by PFGE, probably due to their high DNase activities (data not shown). DNase production is a persistent problem in the typing of clostridial species (6). In such cases, RAPD assay and/or other methodology should be applied.

Botulism cases caused by BoNT/E-producing C. butyricum strains other than those described in this paper will probably be found somewhere in the world in the future. Genetic analysis of such causative C. butyricum strains should provide more detailed and useful information on the epidemiology of this organism.