Novel Structural Elements within the Nonproteolytic Clostridium botulinum Type F Toxin Gene Cluster

Abstract

We sequenced for the first time the complete neurotoxin gene cluster of a nonproteolytic Clostridium botulinum type F. The neurotoxin gene cluster contained a novel gene arrangement that, compared to other C. botulinum neurotoxin gene clusters, lacked the regulatory botR gene and contained an intergenic is element between its orfX2 and orfX3 genes.

Botulinum neurotoxin (BoNT), a Zn²⁺-proteinase produced by Clostridium botulinum and rare neurotoxigenic strains of Clostridium butyricum and Clostridium baratii, is the most poisonous substance known (1). It causes animal and human botulism worldwide, is a widely used licensed therapeutic, and has been classified as a category A (maximum-threat) bioweapon (1). BoNT exists in seven toxin types (types A to G) that are distinguished by the inability of polyclonal antibodies specific for one toxin type to neutralize any of the other six toxin types. Variants or subtypes within BoNT type A (BoNT/A), BoNT/B, BoNT/E, and BoNT/F have been identified and designated with the serotype followed by an Arabic number, e.g., A1, A2, etc. (9). Clostridium botulinum is a heterogeneous species that has been subdivided into group I (proteolytic types A, B, and F), group II (nonproteolytic types B, E, and F), group III (types C and D), and group IV (type G) strains (9).

More than 40 years ago in 1966, an outbreak of food-borne botulism in California from home-prepared venison jerky resulted from a rare nonproteolytic C. botulinum type F strain, here designated IBCA66-5436 (8, 14), whose toxin was recently categorized by Raphael et al. as subtype F6 (16). However, the complete structure of, and the gene content within, the toxin gene cluster of strain IBCA66-5436 remained uncharacterized.

Two main bont cluster structures are known. The hemagglutinin (ha) toxin gene cluster is found in type A1, A5, B, C, D, and G strains, while the orfX toxin gene cluster is found in type A1 to A4, E, and F strains (6, 10, 17). Here we report the structure of the first complete toxin gene cluster for nonproteolytic C. botulinum type F6. We found that the toxin gene cluster has a novel gene arrangement which unexpectedly lacks the regulatory botR gene and that it contains a type E-like orfX1 gene and an intergenic is element between its orfX2 and orfX3 genes.

A high-quality draft genome sequence of strain IBCA66-5436 was generated from standard and paired-end libraries that were sequenced on both Illumina (Solexa) and 454 Titanium (Roche Diagnostics) platforms (3). The 454 data (30-fold coverage) were assembled using Newbler version 2.3, and the Illumina data (288-fold coverage) were assembled with VELVET version 0.7.63. The two assembly results were integrated using parallel Phrap version SPS 4.24, and the integrated assembly was examined with Consed.

An orfX toxin cluster was identified by BLAST (GenBank accession number HQ441176) (Fig. 1), as has been found in other bont/F gene clusters (16). Like the nonproteolytic C. botulinum and C. butyricum bont/E gene clusters, the nonproteolytic bont/F6 gene cluster lacked the botR gene. The botR gene is a putative sigma 70 factor that is believed to positively regulate the expression of BoNT and its accessory proteins (7, 10, 13, 15). However, unlike bont/E clusters, the bont/F6 cluster contained an additional 0.5-kb intergenic spacing between its orfX2 and orfX3 genes (Fig. 1) (4, 11). A BLAST search determined that the orfX2-orfX3 intergenic spacing sequence in the bont/F6 cluster has approximately 50% homology to an insertion sequence element (GenBank accession number CP001056; region 1390960 to 1391208) found in the nonproteolytic C. botulinum type B strain Eklund 17B (ATCC 25765). Interestingly, the type A1 HA⁻ OrfX⁺ gene clusters of C. botulinum strains NCTC 2916 A(B) and CDC5328A also contain an orfX2-orfX3 intergenic spacing (12) (Fig. 1) that however lacks homology with the type F6 orfX2-orfX3 intergenic spacing. Parenthetically, it should be noted that the nonproteolytic Eklund 17B strain contains a hemagglutinin (ha), rather than an orfX, toxin gene cluster (2).

FIG. 1.

Botulinum neurotoxin gene cluster arrangements of orfX clusters. Note that the nonproteolytic F6 and E clusters lack the botR regulatory gene, which is present in the proteolytic F1 and A1 to A4 clusters. Although the F6 and A1 clusters both contain an orfX2-orfX3 intergenic spacing that consists of a degenerated is element, the two is sequences are not homologous. The subtype of the bontA gene of strain CDC41370 was designated A1-2 in Fig. 1 of reference (12). The complete neurotoxin gene cluster sequences of bont/F2 to bont/F5 and bont/F7 are not available in GenBank and so could not be included in this report.

To characterize the nucleic acid and predicted amino acid sequences of the genes of the bont/F6 cluster, we aligned (CLUSTALW) each gene sequence with its selected homologs from orfX gene clusters in type A, E, and F strains and computed the pairwise identities with the MEGA4 software (http://www.megasoftware.net). A comparison of the non-bont genes found that most genes of the bont/F6 cluster shared a low percentage identity with their orfX gene cluster homologs in type A, E, and F strains (Table 1). We calculated the pairwise identities of the non-bont genes, which ranged from 65.0% to 92.5% for nucleotides and from 51.5% to 87.3% for amino acids. Interestingly, the p47, orfX2, and orfX3 genes of the bont/E cluster were the least identical to their bont/F6 homologs (65.0% to 79.2% for nucleotides and 51.4% to 71.6% for amino acids). In contrast, the orfX1 gene of the bont/E cluster was most identical to the orfX1 gene of the bont/F6 cluster (92.5% nucleotides and 87.3% amino acids). Table 1 includes only one representative of the several known bont/E gene clusters (bont/E3 of strain Alaska E43), because the bont/E subtype gene cluster sequences are known to be conserved (4). The complete neurotoxin gene cluster sequences of bont/F2 to bont/F5 and bont/F7 are not available in GenBank and so could not be included in this report.

TABLE 1.

Comparison of the nucleotide and amino acid identities of the genes of the bont/F6 cluster of strain IBCA66-5436 to the genes of the C. botulinum type A, E, and F orfX clusters

a The accession numbers are for the GenBank database(s).

b The values in red are the most identical. The values in blue were the least identical. NA, not available.

In conclusion, the nonproteolytic C. botulinum type F6 strain IBCA66-5436 contained a novel bont gene cluster arrangement and novel gene sequences within it. Like the nonproteolytic bont/E gene clusters, the nonproteolytic bont/F6 gene cluster lacked the regulatory botR gene. The absence of the botR gene in the nonproteolytic F6 orfX toxin gene cluster further suggests that in nonproteolytic clostridia with an orfX cluster, the regulation of botulinum toxin expression may differ from its regulation in proteolytic orfX clostridia that have botR in their toxin gene cluster.

Differences in the structure of botulinum toxin gene clusters among different toxin types and subtypes may enable the improved design of DNA-based methods for detection of C. botulinum and related neurotoxigenic clostridia and for differentiation between proteolytic and nonproteolytic neurotoxigenic strains (5).