Recombination and insertion events involving the botulinum neurotoxin complex genes in Clostridium botulinum types A, B, E and F and Clostridium butyricum type E strains

Abstract

Background: Clostridium botulinum is a taxonomic designation for at least four diverse species that are defined by the expression of one (monovalent) or two (bivalent) of seven different C. botulinum neurotoxins (BoNTs, A-G). The four species have been classified as C. botulinum Groups I-IV. The presence of bont genes in strains representing the different Groups is probably the result of horizontal transfer of the toxin operons between the species.

Results: Chromosome and plasmid sequences of several C. botulinum strains representing A, B, E and F serotypes and a C. butyricum type E strain were compared to examine their genomic organization, or synteny, and the location of the botulinum toxin complex genes. These comparisons identified synteny among proteolytic (Group I) strains or nonproteolytic (Group II) strains but not between the two Groups. The bont complex genes within the strains examined were not randomly located but found within three regions of the chromosome or in two specific sites within plasmids. A comparison of sequences from a Bf strain revealed homology to the plasmid pCLJ with similar locations for the bont/bv b genes but with the bont/a4 gene replaced by the bont/f gene. An analysis of the toxin cluster genes showed that many recombination events have occurred, including several events within the ntnh gene. One such recombination event resulted in the integration of the bont/a1 gene into the serotype toxin B ha cluster, resulting in a successful lineage commonly associated with food borne botulism outbreaks. In C. botulinum type E and C. butyricum type E strains the location of the bont/e gene cluster appears to be the result of insertion events that split a rarA, recombination-associated gene, independently at the same location in both species.

Conclusion: The analysis of the genomic sequences representing different strains reveals the presence of insertion sequence (IS) elements and other transposon-associated proteins such as recombinases that could facilitate the horizontal transfer of the bonts; these events, in addition to recombination among the toxin complex genes, have led to the lineages observed today within the neurotoxin-producing clostridia.

Figure 1

A 16S rrn dendrogram of clostridial species. The 16S rrn genes from the 15 strains examined in this study (13 C. botulinum indicated in red, one BoNT/E-producing C. butyricum in red and one C. sporogenes in blue) were aligned to the 16S rrn genes of different Clostridium species identified within Genbank via BLAST searches. A maximum likelihood tree using 78 sequences with four outgroup sequences from the Alkaliphilus genus (removed) was generated from 1,208 nucleotides. The scale bar of 0.03 represents three point mutations per 100 bases or 3% diversity between sequences. Two 16S rrn gene sequences from C. sporogenes ATCC 15579 are included. The 16S rrn dendrogram illustrates the genetic diversity within the Clostridium genus and among strains within the Group I-VI designations.

Figure 2

Chromosomal and plasmid synteny plots. Panels 1a-d or 2 a-d show four synteny plots of either chromosomal or plasmid sequence alignments, respectively. The reference sequence listed on the x-axis was queried with the strain sequence listed on the y-axis. The red dots indicate forward matches of the sequence comparisons: the blue dots indicate reverse compliment matches. The continuous diagonal line in the plot in panel 1a illustrates the overall chromosomal organization or synteny shared between the proteolytic strains of Hall and either the Kyoto-F, Loch Maree, 657, Okra or Langeland strains. Panel 1b and 1c plots compare Hall and C. butyricum BL 5262 to the BoNT/E-producing Alaska E43 strain, where little synteny is observed. In panel 1d four contigs of C. sporogenes ATCC 15579 are compared to the Hall strain and reveal genomic synteny and a 701 kb inversion between the two species. Panels 2a-d examine plasmid synteny. The diagonal lines in panel 2a illustrate that the Loch Maree pCLK has a similar organization to pCLJ with a small 16.7 kb inversion that includes the bont/a3 relative to the bont/a4. Panels 2b and 2c show that pCLL within Eklund 17B does not share synteny either to pCLK or pE88 that contains the tetanus toxin. In panel 2d four contigs of the Bf strain show synteny to pCLJ and the 16.7 kb inversion of bont/a4 relative to the bont/f.

Figure 3

BoNT complex and flanking regions in different strains. The bont gene cluster, flanking regions and location (chromosome or plasmid) are indicated for the different strains. The orfX cluster (orfX3-orfX2-orfX1-(botR)-p47-ntnh-bont complex genes) is present in the BoNT/E-producing strains (C. botulinum and C. butyricum), the BoNT/A1 of the A1(B) strain, serotype F (BoNT/F and BoNT/bvF) and the BoNT/A2-A4 subtypes. The ha cluster (ha70-ha17-ha33-botR-ntnh-bont complex genes) is present in the serotype B strains containing BoNT/bvB, BoNT/B1, npBoNT/B, BoNT/(B) and BoNT/A1 of the Hall strain. The flanking regions consist of IS elements, flagellin (fla), lycA and hypothetical (hypo) proteins. The prime symbol indicates a partial gene.

Figure 4

Relative locations of the different bonts within the chromosome or plasmid. Three operons (designated arsC, oppA/brnQ and rarA) within the chromosome of the ATCC 3502 or Beluga strain show where the various bonts are located within the different strains. The bont/a1 of the A1(B) strain, bont/a2 of the Kyoto-F and the bont/f of the Langeland strain are located within the arsC operon. The bont/(b) within the A1(B) strain and the bont/a1 within the Hall strain is located within the oppA/brnQ operon. The rarA operon contains the bont/e complex within the Beluga, Alaska E43 or C. butyricum BL 5262 strains. The relative locations of the bonts in the Group I plasmids are indicated in pCLJ. One site contains either the bont/a3 in the Loch Maree strain, the bont/a4 in the bivalent 657 or the bont/f in the Bf strain. Another site contains either the bont/b1 in the Okra strain, the bont/bv b in 657 or bont/bv b in the Bf strain. The bont/np b location within pCLL within the Eklund 17B strain is indicated. This figure shows the common sites of the bonts in different strains providing evidence that the bonts are not randomly located within the chromosome or plasmid.

Figure 5

Comparison of the arsC operon in different strains. The region of the arsC operon within the ATCC 3502 strain was compared to the arsC region in other strains. The horizontal arrows indicate coding sequences (CDSs). Gene designations are labelled above the arrow. GenBank locus IDs are labelled below the arrow. The first CDS was given the full GenBank locus ID followed by an abbreviated ID that uses only the last 2-3 digits. At this site (847 kb - 868 kb) there is no toxin gene cluster within ATCC 3502; however, this site contains the bont/a1 of the A1(B) strain and the bont/f within the Langeland strain. The components in this region are depicted in the Kyoto-F, Loch Maree, 657 and Okra strains. The regions flanking the arsC operon are similar upstream and downstream in each of these strains.

Figure 6

Dendrogram of arsC gene. The 392 nucleotides of arsC (arenate reductase) were compared among C. botulinum strains and other clostridial species. Where multiple copies of the arsC were present within a strain, the copies are designated as C-1, C-2 or C-3 based upon their location within the operon shown in Figure 5. The dendrogram illustrates that the arsC copies within the same strain are different from each other and that the arsC sequences from Groups I, II and VI strains differ.

Figure 7

Comparison of the oppA/brnQ operon in different strains. The region of the oppA/brnQ operon within the ATCC 3502 strain was compared in the different strains. The horizontal arrows indicate coding sequences (CDSs). Gene designations are labelled above the arrow. GenBank locus IDs are labelled below the arrow. The first CDS was given the full GenBank locus ID followed by an abbreviated ID that uses only the last two to three digits. In this region (895 - 915 kb) the bont/a1 within the ATCC 3502 strain and the bont/(b) of the A1(B) strain are located. No bont genes within the other strains of Langeland, Kyoto-F, 657, Okra and Loch Maree are located here. The regions flanking the oppA/brnQ operon are similar upstream and downstream in each of these strains.

Figure 8

(a) Location of the RarA operon within C. botulinum and C. butyricum strains and (b) dendrogram of rarA genes from different clostridial species. The rarA operon in the ATCC 3502 strain is compared to the rarA operon in the Group II and VI strains. In the Eklund 17B strain the rarA gene is intact. However, in the Alaska 43, Beluga and C. butyricum BL 5262 strains, the rarA gene is split and a bont/e gene cluster has been inserted. Note the similarity of the components within the inserted sequence and that it also contains an intact rarA gene. The regions flanking the rarA operon are similar upstream and downstream in the Group II strains. (b) The dendrogram of rarA genes shows that some strains contain two copies of rarA, one that is intact and one that is split from the insertion of the bont/e complex genes. The 1,195 nucleotides of rarA from both intact and split genes were compared; the sequences of split rarA genes were spliced together to make full-length genes. The dendrogram shows that the sequences of the spliced rarA in C. botulinum Alaska E43 and Beluga type E strains are similar to each other but are different from the spliced rarA in C. butyricum BL 5262. This difference indicates that the insertion of the toxin gene cluster occurred as two separate events in each species. The inserted/intact rarA sequences in both of these species are similar indicating a common source.

Figure 9

Plasmid synteny among pCLK, pCLJ, pCLD and pBf. Three fully sequenced plasmids (pCLK, pCLJ and pCLD) are compared to four contigs of the Bf strain that showed identity to pCLJ by BLASTN analysis. Regions of homology among the bivalent pCLJ, pCLK, pCLD and four pBf contigs is indicated in red and the toxin regions containing of bont/a3, bont/a4 or bont/f are coloured in blue or the bont/bv b and bont/b1 in yellow. The comparisons show the similar location of the bont/bv b and bont/b1among the 3 plasmids. The bont/f and bont/a3 also have similar locations but are inverted in relation to bont/a4. The four Bf contigs include ABDP01000018.1 (84.3 kb), ABDP01000023.1 (68.4 kb), ABDP01000034.1 (16.8 kb), ABDP01000069.1 (0.8 kb) and were ordered according to pCLJ. The coloured symbols are expanded in Figure 10 to detail the genes located in these regions.

Figure 10

Plasmid regions containing bonts. This is an expanded image of the regions between the symbols in Figure 9 and provides details of the genes located within the different plasmids in these areas. The horizontal arrows indicate coding sequences (CDSs). Gene designations are labelled above the arrow. GenBank locus IDs are labelled below the arrow. The first CDS was given the full GenBank locus ID followed by an abbreviated ID that uses only the last two to three digits. The figure shows that the bonts within the plasmids in these Group I strains are located in either of two sites. One location (between the yellow and blue symbols) contains either the bont/f of the Bf strain, the bont/a4 of the 657 strain or the bont/a3 of the Loch Maree strain. The numbers in parentheses, such as 23,742 bp in the pBf panel, indicate additional sequence in that region that is not detailed but is shown in Figure 9. The other plasmid site that contains the bont/b in several plasmids is depicted between the green and purple symbols. This region contains the bont/bv b or bont/b1 in the Bf strain, 657 strain or the Okra strain. The bottom panel depicts the region containing the bont/np b within the Eklund 17B strain. This region shares no similarity to the Group I plasmids.

Figure 11

Dendrogram of the ntnh gene in different BoNT-producing strains. The dendrogram of 3,471 nucleotides of the ntnh gene shows the variation within this gene. Some of the ntnh sequence variation in the strains is due to recombination events. The location of the ntnh in ATCC 3502, ATCC 19397 and Hall strain close to the ntnh within the serotype B strains resulted from a recombination event midway within the ntnh gene resulting in a recombinant B/A ntnh, that is a partial B ntnh and partial A ntnh. Another similar recombination event in ntnh of the 7I03-H strain has resulted in a recombinant C/A ntnh, that is a partial C ntnh and a partial A ntnh. The Langeland F ntnh location in the dendrogram near the serotype A strains of Kyoto-F, Loch Maree and 657 resulted from a recombination event near the 3' end or following the ntnh gene where a bont/f was inserted. The dendrogram illustrates the variation in the ntnh genes from multiple serotypes and the location of recombinant ntnh genes.