We are searching data for your request:
Upon completion, a link will appear to access the found materials.
Is it theoretically possible that any aerobic bacterium becomes able to produce Botulinum toxin as a result of horizontal gene transfer from Clostridium botulinum to that bacterium?
Mining the botulinum genome
This is a spore of Clostridium botulinum. Credit: IFR(Norwich BioScience Institutes) Scientists at the Institute of Food Research have been mining the genome of C. botulinum to uncover new information about the toxin genes that produce the potent toxin behind botulism.
The toxin that causes botulism is the most potent that we know of. Eating an amount of toxin just 1000 th the weight of a grain of salt can be fatal, which is why so much effort has been put into keeping Clostridium botulinum, which produces the toxin, out of our food.
The Institute of Food Research on the Norwich Research Park has been part of that effort through studying the bacteria and the way they survive, multiply and cause such harm. In new research, IFR scientists have been mining the genome of C. botulinum to uncover new information about the toxin genes.
There are seven distinct, but similar, types of botulinum neurotoxin, produced by different strains of C. botulinum bacteria. Different sub-types of the neurotoxin appear to be associated with different strains of the bacteria. Genetic analysis of these genes will give us information about how they evolved.
Dr Andy Carter, working in Professor Mike Peck's research group, used data generated from sequencing efforts at The Genome Analysis Centre, on the Norwich Research Park. Andy compared the genome sequence of five different C. botulinum strains, all from the same group and all producing the same sub-type of neurotoxin.
Comparison of two closely related C. botulinum strains. Genes shared by both genomes are in green. The newly introduced DNA is shown by the gap in the red bars which connect the two DNA species. Genes in red are neurotoxin cluster associated including two extra ones, numbered 33 and 35, which are remnants of previous neurotoxin gene clusters that have been disrupted during the evolution of the current cluster. Gene number 12 (in black) is the new copy of the DNA replication gene.An initial finding was that the five strains were remarkably similar in the area of the genome containing the neurotoxin gene. This suggests that the bacteria picked up the gene cluster in a single event, sometime in the past. Bacteria commonly acquire genes, or gene clusters, from other bacteria through this horizontal gene transfer. It is a way that bacteria have evolved to share 'weapons', such as antibiotic activity or the ability to produce toxins. To find out more about how C. botulinum acquired its own deadly weapon, Andy delved deeper into the genome sequence.
Like fossils of long lost organisms, Andy found, in the same region of the genome, evidence of two other genes for producing two of the other types of neurotoxin. Although these gene fragments are completely non-functional, finding them in the same place in the genome as the functional neurotoxin gene cluster is significant as it suggests that this region of the genome could be a 'hotspot' for gene transfer.
Looking to either side of the neurotoxin gene cluster uncovered more evidence supporting the hotspot idea. When the gene cluster inserted into the C. botulinum genome, it cut in two another gene. This gene is essential for the bacteria to replicate its DNA, so why does destroying it not prove fatal? C. botulinum was unaffected by this because contained in the segment of imported DNA was another version of the chopped-up gene.
Perhaps this is pointing us to the way C. botulinum first picks up its lethal weapon. This should help us prepare against the emergence of new strains, and may even one day help us disarm this deadly foe.
Introduction
Botulinum neurotoxin is responsible for botulism, a severe and deadly neuroparalytic disease. It is the most potent toxin known, with as little as 30 ng potentially fatal to a human adult (Peck 2009). There are seven types of botulinum neurotoxin (types A–G), which are formed by the Gram-positive anaerobe Clostridium botulinum and some strains of C. butyricum and C. baratii. Types A, B, E, and F neurotoxins have been associated with human pathology (Austin and Dodds 2000 Peck 2006). Clostridium botulinum is a heterogeneous species that comprises four phylogenetically and physiologically distinct bacteria (C. botulinum Groups I–IV), with Groups I and II associated with human botulism (Austin and Dodds 2000 Peck 2009). Strains of Group I (proteolytic) C. botulinum form type A, B, or F neurotoxin, with many strains carrying genes for two neurotoxins, and in some cases also forming two neurotoxins. Strains of Group II (nonproteolytic) C. botulinum form type B, E, or F neurotoxins, but there are currently no reports of strains carrying more than one neurotoxin gene (Macdonald et al. 2011 Peck et al. 2011). One interpretation for this observation would be that horizontal transfer of the neurotoxin genes occurs less frequently in Group II C. botulinum than it does in Group I C. botulinum. Recent genetic analysis has identified a number of neurotoxin subtypes, with particular subtypes often found only in one Group of C. botulinum (Smith et al. 2005 Hill et al. 2007 Macdonald et al. 2011 Peck et al. 2011). For example, strains of Group II C. botulinum type F, which are the subject of this study, exclusively carry the type F6 neurotoxin gene, whereas strains of Group I C. botulinum type F carry a neurotoxin gene, which can be of types F1, and strains of C. baratii all carry the type F7 neurotoxin gene (Raphael et al. 2010).
The neurotoxin has several accessory proteins, the genes for which are adjacent and form the so-called neurotoxin gene cluster. There are two classes the ha-cluster encodes the neurotoxin, a nontoxic-nonhaemagglutinin protein, and three hemagglutinins and the orf-x-cluster encodes the neurotoxin, a nontoxic-nonhaemagglutinin protein, and four other proteins (Orf-X1, Orf-X2, Orf-X3, and P47) of unknown function (Hill et al. 2009 Peck 2009 Peck et al. 2011). In addition, clusters often contain the gene for a regulatory sigma factor, botR. When chromosomally located, botulinum neurotoxin gene clusters are usually flanked by fragments of mobile elements, an observation which has raised the possibility that they may have been acquired by horizontal gene transfer (Sebaihia et al. 2007). Neurotoxin gene clusters can also be located on bacteriophages (Group III C. botulinum) and plasmids (Groups I, II, and IV). Horizontal acquisition of neurotoxin gene clusters must occur relatively rarely on an evolutionary timescale, based on the knowledge that in Group I, C. botulinum neurotoxin gene clusters are found at only three different chromosomal locations, and more dramatically in both Group II C. botulinum type E and in C. butyricum type E, the neurotoxin gene cluster has been inserted directly into rarA, a recombination-associated gene (Hill et al. 2009).
Recent work in our laboratory studied the genetic relationship between members of Group II C. botulinum using a whole-genome DNA microarray. This study extended the observations of others that strains of Group II C. botulinum type B and type F are closely related but that Group II C. botulinum type E strains form a distinct clade (Stringer et al. 2013). Reference genome sequences were already available for Group II C. botulinum type B (Eklund17B) and type E (strains Beluga and Alaska), so we decided to sequence the genome of each type F strain from our microarray study to explain this close relationship with type B strains.
Access options
Get full journal access for 1 year
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Get time limited or full article access on ReadCube.
All prices are NET prices.
Can the C. botulinum neurotoxin gene be transferred to non-clostridia?
My understanding of horizontal gene transfer is that part of the DNA of 1 group of bacteria can be transferred to another group. One example is E. coli O104:H4 acquiring the shiga-like toxin producing gene (presumably from Shigella dysenteriae or something like O157 with help from a virus [prophage]) and the ESBL gene (which is getting more and more common).
Now, while shigatoxin is pretty dangerous, Clostridium botulinum toxin is even more dangerous - but that's an anaerobic bacteria. However, if an E. coli were to acquire the neurotoxin producing gene - we would have a big problem, correct?
Is this possible? If so, under what kind of conditions could this occur?
Botulinum neurotoxin (BoNT) can undergo horizontal gene transfer, and there's some evidence that it has done so in the past -- at least among the Clostridia. Here's a paper on it, available for free through the PubMed Central database.
I suppose that, technically speaking, it's possible that BoNT could one day transfer to another bacterial species. But you should realize that, in nature, horizontal gene transfer is not as easy as you may have been led to believe. Most bacteria have ways of detecting and degrading "foreign" DNA you may have heard of restriction modification systems (which can be a huge problem in the lab if you're trying to shuttle DNA between different bacterial species) or this totally awesome neat-o thing called CRISPR (which I only know a little bit about). The major evolutionary benefit to these systems is likely in protecting the bacteria from viral infection, and maybe from horizontal transfer of "dangerous" genes (e.g., encoding gene products that might be toxic to the bacterium). But the relevant effect here is that horizontal gene transfer ends up being rare, especially between divergent species.
To be honest with you, the most likely way for this to occur would be for somebody to intentionally engineer a non-clostridial BoNT-producing strain. This would be VERY, VERY ILLEGAL and would be considered a potential act of bioterrorism unless the person doing this had very specific permission from the government (in the US, this is regulated by the Department of Justice and the Centers for Disease Control).
Genetic Diversity among Botulinum Neurotoxin-Producing Clostridial Strains
FIG. 1 . Phylogenetic dendrogram of Clostridium species based on 16S rRNA genes. A neighbor-joining tree of 54 sequences reported in GenBank and 36 sequences representative of the strains from this collection is shown. This illustrates the genetic diversity within the clostridia. C. botulinum strains cluster into four distinct groups that follow the group I to group IV designation historically based on physiological characteristics. These groups are interspersed among the 27 other clostridial species in the tree. The tree was constructed using an alignment of 16S rRNA gene sequences that contained 1,329 bases after removal of columns containing more than 80% gap characters and includes sequences from bivalent, nonproteolytic, and proteolytic toxin-producing strains. FIG. 2 . AFLP-based dendrogram of 174 C. botulinum strains. DNA fragments generated from restriction endonuclease digestion of each of the strain DNAs were ligated into linkers and selectively amplified. Forty DNA fragments generated by AFLP experiments were used as a fingerprint to represent each of the strains. If 40 fragments did not exist, fewer fragments were used, as noted in parentheses. The comparison of fingerprints from the 174 strains shows a large separation between the proteolytic (group I) and nonproteolytic (groups II, III, and IV) strains and distinct branches representing groups I to IV. The AFLP groups also contain generally distinct toxin serotypes. The distance measure or genetic distance is the proportion of fragments that two samples do not have in common. FIG. 3 . Comparison of BoNT/A gene sequences. The full-length coding region of the BoNT/A gene in 60 strains and six GenBank sequences were aligned. Four distinct subtypes are apparent. Most strains (54 strains) are of the BoNT/A1 subtype, and four strains are within the BoNT/A2 subtype. Two newly identified subtypes, BoNT/A3 and BoNT/A4, each contain one member: the A254 (Loch Maree) strain and the bivalent Ba207 strain, respectively. These strains show significant sequence variations compared to BoNT/A1 and A2 subtypes. FIG. 4 . Similarity plot comparing BoNT subtype sequences to the BoNT/A2 subtype. BoNT sequences of the BoNT/A1, A3, and A4 subtypes and BoNT/B1 and Chinese C. butyricum BoNT/E were compared to the BoNT sequence of the BoNT/A2 Kyoto-F subtype (GenBank accession number X73423). This plot illustrates that the BoNT/A2 subtype is approximately 99% identical to the BoNT/A1 subtype (A142) through nucleotides 1 to 1146 and approximately 99% identical to the BoNT/A3 subtype (A254) through nucleotides 1147 to 3450. This suggests that the BoNT/A2 subtype is a result of a recombination event between BoNT/A1 and BoNT/A3 lineages of gene sequences. FIG. 5 . Comparison of BoNT/B gene sequences. The full-length coding regions of the BoNT/B gene in 53 strains and seven GenBank sequences were aligned. Four distinct clusters that include the BoNT/B1 and BoNT/B2 and bivalent (Ab149, Ba207, Bf258, and Bf698) and nonproteolytic BoNT/B subtypes are apparent. Most strains are of the BoNT/B2 subtype, with 16 strains being of the BoNT/B1 subtype. Strain B506 is separate from the other BoNT/B2 strains and represents a newly identified variation in this serotype. FIG. 6 . Comparison of BoNT/E gene sequences. The full-length coding regions of the BoNT/E gene in 21 strains and 15 GenBank sequences were aligned, resulting in five clusters labeled E1 to E5. Two clusters contain sequences from C. butyricum BoNT/E strains collected in Italy (E It. butyr.) or China (E Ch. butyr.). The other subtypes include BoNT/E1 and E2 and a newly identified subtype, labeled BoNT/E3, containing four members (E185, E540, E545, and E549). FIG. 7 . Comparison of the seven different serotypes of BoNT gene sequences. Shown is a neighbor-joining alignment of the nucleotide coding regions of the seven BoNT genes (A through G) including the tetanus toxin. The comparison of the BoNT genes shows a different relationship of the serotypes than what is found based on 16S rRNA genes or AFLP analysis. Nonproteolytic and bivalent strains (Ba207 and Ab149) and representatives of the different subtypes are included.Genetic Characterization of the Exceptionally High Heat Resistance of the Non-toxic Surrogate Clostridium sporogenes PA 3679
Clostridium sporogenes PA 3679 is a non-toxic endospore former that is widely used as a surrogate for Clostridium botulinum by the food processing industry to validate thermal processing strategies. PA 3679 produces spores of exceptionally high heat resistance without botulinum neurotoxins, permitting the use of PA 3679 in inoculated pack studies while ensuring the safety of food processing facilities. To identify genes associated with this heat resistance, the genomes of C. sporogenes PA 3679 isolates were compared to several other C. sporogenes strains. The most significant difference was the acquisition of a second spoVA operon, spoVA2, which is responsible for transport of dipicolinic acid into the spore core during sporulation. Interestingly, spoVA2 was also found in some C. botulinum species which phylogenetically cluster with PA 3679. Most other C. sporogenes strains examined both lack the spoVA2 locus and are phylogenetically distant within the group I Clostridium, adding to the understanding that C. sporogenes are dispersed C. botulinum strains which lack toxin genes. C. sporogenes strains are thus a very eclectic group, and few strains possess the characteristic heat resistance of PA 3679.
Keywords: Clostridium botulinum Clostridium sporogenes PA 3679 SpoVA dipicolinic acid food sterilization horizontal gene transfer spore heat resistance.
Results
We first present the population growth sub-model that was previously introduced in [44], and then show how this is coupled to the gene regulation sub-model. The final model resulting from the union of the two sub-models is encoded into COPASI and simulations are used to illustrate the ability to reproduce a selected set of additional experimental results.
A nutrient and quorum-sensing regulated population growth model
Details of a computational model for the growth of a population of C. botulinum Group I type A1 cells in a culture have been reported by [44]. As thoroughly explained in [44], the rationale underlying the need of modelling population dynamics is rooted in the experimentally observed correlation between the bacterial growth phase and the toxin production process. Further evidence supporting this correlation at genetic regulation level is described in the section on the molecular model of BoNT synthesis regulation.
The mathematical modelling of C. botulinum cultures is based on a compartmentalization of the growing population of cells into three distinct groups:
Adapting cells, denoted by AC, which includes the bacterial cells after their addition to the botulinum growth medium. While the metabolic processes involved remain to be established, they may be similar to that reported in Salmonella [63]
Reproducing cells, denoted by RC, formed by the cells that are actively reproducing
Sporulating cells, denoted by SC, which consists of the cells that are committed to sporulation (though not measured in [59]).
The initial population of C. botulinum cells is fully composed of AC cells, which later evolve to RCs and may commit to sporulation and become SCs. These processes are influenced by some biochemical species generically termed “Signal”, as shown in Fig 1 . A future development, not currently included, is to extend the present analysis to start with bacterial spores and to therefore incorporate steps for spore germination and outgrowth [64].
Diagrammatical representation of the best fitting model determined in [44]. The reproduction of cells is controlled by the abundance of nutrients N, and the sporulation is regulated by the concentration of a quorum-sensing signal S.
Different hypotheses relating to the nature of the “signal(s)” (previously described in [44]) led to the discrimination of plausible modelling scenarios and were used to generate corresponding models that were then evaluated on their ability to reproduce the observed pattern of growth observed for C. botulinum type A1 strain ATCC 19397.
We found that a model where two distinct signal sources were considered—the first one determined by the abundance of nutrients essential to C. botulinum cell growth, which we denoted by the abstract species N, and a second one endogenously produced by the bacterial cells and used as a quorum-sensing signal, denoted by S–was most successful at explaining the pattern of growth observed for C. botulinum type A1 strain ATCC 19397. A diagrammatic representation of this modelling option is included in Fig 1 . In this model the rate of cell reproduction increases with the nutrient concentration, N, whilst the rate of sporulation increases with the concentration of the chemical signal S. The model proposed in [44] was encoded using eight reactions. We consider here an updated version, still based on the same rationale, which is encoded by the six reactions listed in Table A of Supporting Information File 1 (Table A in S1 Text). As previously reported in Figure 6 of [44], this model produces a good fit for the experimental growth data generated for strain ATCC 19397.
Molecular model of BoNT synthesis regulation
Several environmental stimuli have been identified with positive and negative regulation of toxin production in C. botulinum Group I type A1. Neurotoxin production has been reported to be associated with the transition from late-exponential to early-stationary phase cultures. This is indicated by a peak in the level of neurotoxin gene cluster expression that is clearly observable in the late-exponential to early-stationary phase of cultures and which drastically decreases during the later stationary phase (as shown in Fig 2 ). Moreover, the expression patterns for all the genes, in both the ntnh/bont and the ha operon, show an equivalent correlation with population dynamics (data available in [26], [21], [22] and [23]). This points to regulatory elements that link population growth to toxigenesis in C. botulinum type A1.
Data from the experimental results published in [26], [21], [22] and [23] for C. botulinum type A1 strains ATCC3502, Hall A-hyper, Hall A and Hall A respectively. Notice that toxin loci of these three strains are genetically identical with each other [9]. Comparison of the time courses measured in optical densities for the cultures (left) and the comparison of the bont gene expression time courses (right). Data normalized to the maximum OD (left) and maximum expression level (right) of the single original time course.
BotR as a positive regulator of BoNT synthesis
Botulinum neurotoxins are produced in the form of a complex containing the neurotoxin itself and one or more non-toxic auxiliary proteins that protect the neurotoxin from environmental stress and assist in absorption [65]. A majority of type A1 toxins are complexed with the non-toxic non-hemagglutinating (NTNH) protein and three hemagglutinins (HA17, HA33 and HA70) [26,40]. The genes coding for these proteins are organized in two operons, namely the ntnh-bont and ha operons [21], and the botR gene can be found between the two.
The botR gene encodes a 21-22kDa protein (BotR), an alternative sigma factor with features of a DNA-binding protein (i.e., highly basic isoelectric point and helix-turn-helix motif [17]). BotR appears as a key positive regulator for the ntnh-bont and ha operons. Indeed, both operons have consensus -10 and core promoter sequences, which are recognized by BotR, which specifically binds to the promoter region of the ntnh-bont and ha operons and directs RNA polymerase (RNAP) to transcribe the two operons [66]. The botR gene is transcribed in the same orientation as bont, and BotR has been characterised as a transcriptional activator of ntnh-bont and ha genes based on botR overexpression or partial inhibition by antisense mRNA in C. botulinum Group I type A1 [17,30,66]. BotR can also target its own promoter, but initiation of transcription could not be observed in vitro [67].
Based on this evidence, BotR is included in the coupled model as a direct positive regulator of toxigenesis as well as a positive regulator of itself.
TCSs as positive and negative regulators of BoNT synthesis
Experiments reported by [30] focused on the toxin regulatory elements in the genome of C. botulinum Group I type A1 strain Hall. In this study, the authors first identified a considerable number (30 in total) of gene pairs coding for two-component systems (TCSs) that affected toxin regulation. TCSs are widely used in bacterial stimulus-response coupling for sensing and relaying a variety of environmental and developmental cues that affect gene activation. A TCS consists of a membrane-bound histidine kinase, that senses a specific stimulus, and a response regulator that typically has the characteristics of a DNA binding protein to mediate the expression of a set of target genes [68]. The signal is relayed from the sensor component to the response regulator via trans-phosphorylation. The role of the TCS candidates were explored by [30] using antisense mRNA silencing to determine which were primarily acting on toxin operons. The search led to the identification of three TCSs that were shown to positively regulate toxin production. These results indicate (please note we will use the CBO equivalent numbers identified for strain ATCC3502):
The three TCSs which positively regulate toxin production, are encoded by the gene pairs cbo_1042/cbo_1041, cbo_1967/cbo_1968, and cbo_0608/cbo_0607
The effects of the three TCSs are independent from that of BotR, since expression of botR is not significantly affected by the mRNA silencing
The CBO0608/CBO0607 TCS was suggested to be homologous to TCSs of the PhoP/PhoR family involved in, but not restricted to, sensing and reacting to phosphate starvation.
These experimental results led us to include two distinct positive regulatory mechanisms in our model: a first one that models the effect of the CBO0608/CBO0607 TCS, which we assume is sensing and reacting to the lack of nutrients, and a second one (consisting of the two species CBO_SHK/CBO_RR) that abstractly represents the two TCSs CBO_1042/CBO_1041 and CBO_1967/CBO_1968, which we assumed to be activated by the increase in concentration of quorum-sensing molecules.
Furthermore, the first reported evidence of negative regulation of C. botulinum Group I type A1 toxin synthesis was provided by Zhang et al. [31], who showed that the CBO_0787/CBO_0786 TCS down-regulates toxin production in strain ATCC 3502. The experimental results [31] most relevant to the coupled model are:
Expression of the TCS components CBO_0787 and CBO_0786 is dependent on the growth phase with a constant level of expression preceding entry to the late exponential phase followed by a subsequent reduction of about 80 percent
The cbo0787 and cbo0786 genes are transcribed polycistronically
Phosphorylated CBO_0786 negatively regulates toxin production, by binding directly to the conserved -10 site of the core promoter regions of ntnh-bont and ha operons so blocking BotR-directed transcription.
Based on this experimental evidence we can infer, and include in the coupled model, a role for the CBO_0787/CBO_0786 TCS as a direct negative regulator of toxigenesis, with phosphorylated CBO_0786 acting as the species exerting the repression by direct binding to the toxin gene promoters.
Nutrition-related metabolic and quorum-sensing pathways as regulators of BoNT
So far, we have identified elements for the model construction without identifying the specific mechanisms for coupling i.e. initiation of response. Experimental evidence indicates that botulinum neurotoxin production is affected by the availability of various carbon and nitrogen sources. Multiple research works [21,26,59] have quantified the effect that nutrients have on the toxin production. Nutrient(s) availability is already included in the population dynamics element of the coupled model and additionally we hypothesize that the abundance of nutrient(s) also regulates toxigenesis directly.
A recent report by Zhang and colleagues [29] demonstrated the role of the global regulator protein CodY in toxin synthesis and elements of this observation provide support for a plausible picture of the nutrition-related effects on toxigenesis in C. botulinum Group I type A1:
CodY is able to bind to the promoter region of the ntnh/bont operon
The binding affinity of CodY for the promoter regions of ntnh/bont operon increases in GTP rich conditions
codY mutant strains show reduced expression levels of bont (approximately 50% less) compared to wild type
The temporal pattern of expression of bont is the same in codY mutant strains and wild type
Two putative binding regions, each one with three mismatches to the consensus CodY-binding motif, are found upstream of the CBO_0787/CBO_0786 operon.
Points 3 to 5 imply that the overall role of CodY is to activate toxin production. However, points 1 and 2 both imply that the effect of CodY is maximal on the operon when the availability of nutrients is high, i.e., when no toxin is produced. Therefore, the binding of CodY to the promoters of ntnh/bont operon must be exerting a repression effect on transcription. That is to say, the activation effect of CodY must be the result of an additional regulation exerted by CodY. Together this means that CodY may be repressing the repressor TCS CBO_0787/CBO_0786 by directly binding to the TCS promoter. For this reason, in a CodY mutant the repression effect of CBO_0787/CBO_0786 would not be released and the expression of the toxin genes is reduced. Point 4 of the sub-section on TCSs as positive and negative regulators of BoNT synthesis indicates that this repression effect of CodY needs to be exerted after the late exponential phase.
We have conducted a sequence analysis of the botR promoter region and found an additional putative binding region for CodY, with some noticeable similarities to sequence motifs and an associated CodY-binding sequence previously identified in the CodY-regulated promoter of another C. botulinum ATCC 3502 gene [69]. This is therefore consistent with the hypothesis that CodY regulates the expression of the alternative sigma factor BotR (see also Supporting Information File 3 for details (S3 Text)). Since we know that botR expression is also phase dependent, we make the assumption that CodY regulates BotR positively, so that when the available nutrient(s) decreases, CodY begins to exert an activation effect on the botR gene transcription. As a consequence, we suppose in our modelling that CodY regulates toxigenesis via two routes, activation and repression, in distinct phases of the population growth.
For modelling, we assume the existence of two distinct forms/behaviours of CodY one we named CodY1, which is prevalent when available nutrient(s) is high, and the other we named CodY2, which accumulates when nutrient(s) are scarce. The transition between the two forms is regulated by the quantity of nutrient(s). In the model, CodY1 represses the ntnh/bont operon, while CodY2 represses the CBO_0787/CBO_0786 operon and upregulates the botR gene transcription. We do not model the mechanism underlying the proposed two behaviours of CodY, but this could involve presence/absence of a bound cofactor or interactions with, or recruitment of different activator/repressor components.
The CBO_0787/CBO_0786 TCS, which has expression regulated by CodY2, is activated via phosphorylation of the CBO_0787 histidine kinase in response to an unknown signal. We assume in the model that the signal is indirectly relayed by the nutrient(s), and therefore by modelled species N. Moreover, since the CBO0608/CBO0607 TCS is assumed to be involved in, but not restricted to, sensing and reacting to phosphate starvation [30], we have placed its regulation under the control of the nutrient(s), by assuming that the phosphorylation of the CBO0608 histidine kinase is repressed by modelled species N.
The molecular details of the quorum-sensing pathway regulating toxin production in C. botulinum Group I type A1 strains has not yet been clarified. What is known from the work reported in [28] is that the genome includes two regions, agrD1 and agrD, which code for homologues of the Staphylococcus aureus agr-like quorum sensing system. Moreover, the authors [28] demonstrated that in the closely related organism C. sporogenes, the pattern of expression of the genes in corresponding regions is strongly correlated with the growth phase: i.e., it increases throughout exponential growth, peaking at late exponential phase, and considerably drops once stationary phase is reached. The authors showed that, the insertional inactivation of the genes in the agrD1 and agrD2 regions in C. botulinum Group I type A1 (strain ATCC 3502) resulted in a reduction in the amounts of toxin produced. More precisely, inactivation of agrD1 led to a marked reduction of the early toxin production, with a return to wild-type levels during late-stationary phase, whereas inactivation of agrD1 led to a more severe restriction of the toxin production that persists throughout the population growth.
Although this experimental evidence clearly indicates a role for quorum-sensing in toxigenesis, the available information is not sufficient to make hypotheses about possible modelling options relating to the pathways that link quorum-sensing with gene expression. We however, decided to include the action of quorum-sensing into the gene expression sub-model in an abstract way. We make a hypothesis that the TCSs CBO_1042/CBO_1041 and CBO_1967/CBO_1968 shown to regulate toxin synthesis in a positive way, sense and react to changes in concentration of a quorum-sensing signal, represented as modelled species S in the population sub-model of the section on nutrient and quorum-sensing regulated population growth model.
Computational model
In the integrated model we include the known regulatory mechanisms controlling toxin production, but not the details of the toxin assembly nor secretion, nor other processes yet to be fully deciphered. Moreover, we limit the model scope to neurotoxin synthesis (i.e., BoNT protein), not as a complex (without the associate proteins, NTNH the HAs, and their interactions). Even though, these simplifications were made in order to prevent the introduction of a large number of unknown kinetic parameters, it is important to note that NTNH, which is transcribed polycistronically from the ntnh-bont operon, is subject to the same regulation as BoNT. As for the ha operon, it is also transcriptionally regulated by BotR, as well as by the three positive regulatory TCSs as shown in [30], and the negative regulatory TCS, as shown in [31]. Thus, we assume that the ANTPs would exhibit the same pattern of expression as BoNT.
The integrated model includes BoNT synthesis and export as a single process, and assumes a delay in export to the culture supernatant.
The model includes the transcription of each species for which the synthesis process is known to be regulated, i.e., the CBO_0787/CBO_0786 proteins, the alternative sigma factor BotR and the bont gene. For these species, transcription and translation are modelled altogether, to avoid introducing too many unknown kinetic parameters into the model. All the synthesis processes are regulated by the abundance of nutrient(s) (modelled species N). As there is no available information on the regulation of the expression for the proteins of the TCSs, CBO_1042/CBO_1041 and CBO_1967/ CBO_1968, we do not include their synthesis processes in the model. Instead we assume a constant concentration of the constituent proteins which change between their unphosphorylated and phosphorylated forms depending on the abundance of regulators. Similarly the model does not include the synthesis process for CodY.
Finally, we include in the integrated model a degradation reaction for each species synthesised, i.e. for CBO_0786, CBO_0787 (and their phospho forms), BotR and BoNT. The integrated computational model for the gene expression network that regulates BoNT production is illustrated in Fig 3 . For the sake of clarity the degradation reactions are not depicted. The gene expression model represents the molecular machinery that regulates toxigenesis inside each bacterial cell. The inner part of the cell is enclosed in the rod-shaped form in Fig 3 , and the N (Nutrients) and S (quorum-signal) modelled species are shared with the population sub-model. We use the same notation of dashed and solid lines as before to distinguish between regulation and mass transfer reactions.
(Top Left) show the role of the BotR sigma factor, and of the three TCSs reported to regulate positively toxigenesis in C. botulinum Group I type A1 strain, along with the negative TCS regulator. (Top Right) our hypothesis of how the availability of nutrients (species N) regulates directly and indirectly (via CodY) toxin production, and how the quorum sensing signal (species S) together with the two positive TCS regulators, recognise the quorum-sensing pathway whose effect on toxin production was experimentally observed in the work of Cooksely and colleagues [28]. (Lower), the dashed arrows represent regulation mechanisms, whereas solid lines model mass transfer reactions. The species N (Nutrients) and S (quorum-sensing signal) are shared with the population sub-model. The state of each bacterial cell is assumed to be the same. Species CBO_0786, CBO_0787 (and their phospho forms), BotR and BoNT are subject to degradation (reactions not graphically depicted).
To complete the definition of the model it is necessary to specify, in terms of molecular interactions, the repression and activation effects on the synthesis processes of the negative regulatory CBO0787/CBO0786 TCS and the alternative sigma factor BotR as well as the impact on the ntnh/bont operon.
We approach this modelling task by explicitly representing as variables of the model the state of the promoters. The promoter of the negative regulatory TCS (named prCBOi) is assumed to be in one of two states: inhibited by the CodY2 species, or active, as illustrated in Fig 4A . The promoter of BotR, called prBR in the model, has three different states of activation: an initial state, (which can express a basal level of synthesis where prBR is not activated by any transcription factor), a second state in which BotR is bound to prBR and acts as a self-activator, and a third state in which CodY2 binds to prBR next to the already bound BotR. In modelling the promoter activity, we assumed that positive regulatory proteins, i.e. CodY2 and the active forms of CBO_0607 and CBO_RR, act as co-factors in transcription, increasing the stability of the transcription machinery and therefore the synthesis rate. Fig 4B illustrates the different levels of activation of the prBR promoter, each one associated with a distinct rate of synthesis.
The synthesis of the negative regulatory TCS, of the alternative sigma factor BotR and the BoNT protein are regulated by inhibitory and activator species. (A) shows the two possible states of prCBOi, the promoter for the polycistronic transcription of proteins CBO0787/CBO0786 (B) illustrates the three possible active states of prBR, the promoter of BotR (C) details the possible states of the ntnh-bont operon promoter prBA: inactive, when not bound, inhibited by CodY1 and/or phosphorylated CBO_0786 inhibits transcription, and activated, by BotR and subsequently by phosphorylated CBO_RR and/or phosphorylated CBO_0607 for increasing levels of activation.
The activity of the promoters of the ntnh/bont operon (prBA) is modelled in a similar way but the multiple positive and negative regulators that affect BoNT synthesis give rise to many more states, as shown in Fig 4C . prBA is modelled as being inactive, i.e. unable to initiate synthesis, if a positive regulator is not bound to it. That is, if the negative regulatory species CodY1 and phosphorylated CBO0786 bind to prBA, synthesis is inhibited (left complex forms illustrated in Fig 4C ). The model also assumes that phosphorylated CBO0786 is a stronger inhibitor than CodY1, and that the inhibition strength is maximum when both inhibitors are bound to prBA. The active forms of the prBA are shown in the right part of Fig 4C . Here, we assume that prBA can be activated in three ways: by the binding of BotR on its own, by the combined binding of BotR, a phosphorylated CBO_RR and phosphorylated CBO_0607 or by the simultaneous binding of BotR and both phosphorylated CBO_RR and phosphorylated CBO_0607, with largest complexes being more active than small complexes. The rationale underlying this modelling is that phosphorylated CBO_RR and phosphorylated CBO_0607 play the role of co-transcription factors, stabilising the transcription machinery and increasing the transcription rate of the prBA which also requires the alternative sigma factor BotR for transcription initiation.
The overall gene expression model corresponds to a set of 49 reactions which are listed in Table B of Supporting Information File 1 (Table B in S1 Text). The initial state of the whole model, as well as the details of the kinetic rates of the gene regulation network and the population sub-models, is provided in Supporting Information File 2 (S2 Text).
The computational model is able to reproduce additional experimental results
In this section we expound the procedure used for calibrating model parameters, and then proceed to validate the model by checking against characteristics of toxigenesis which have been reported previously in sections on nutrient and quorum-sensing regulated population growth model and the molecular model of BoNT synthesis regulation.
To find suitable values for model parameters, we used the experimental data from [59] for type A1 strain ATCC 19397, which we considered as the 'wild type' organism for the purpose of our modelling (WT, hereafter). The experimental time course for the population size (measured in CFU/ml over time in [59]) provides the parameters of the population sub-model, i.e. the kinetic parameters of reactions (1) to (6) provided in Table A of Supporting Information File 1 (Table A in S1 Text). The amount of toxin in the supernatant measured in the same experiment in [59] (measured in MLD50/ml over time) provides the data for fitting the gene expression sub-model, i.e. the kinetics of reactions (1) to (49) listed in Table B of Supporting Information File 1 (Table B in S1 Text). The model parameters are reported in Supporting Information File 2 (S2 Text).
The fitted model is compared with the WT experimental data in Fig 5A and 5B . The experimental data points are shown as empty circles, whereas the computational model is reported as continuous lines. There is an inevitable match between model outcomes and the experimental 'model training' data, which is confirmed by analysis of correlation. For the population an R 2 measure is 0.975 while for the toxin production it is 0.95.
(A) shows the population dynamics, where data measurements are in CFU/ml over time, while (B) illustrates the amount of toxin in the supernatant. In both plots, experimental data points are drawn as circles, while model predicted data are shown as continuous lines.
After tuning model parameters to fit WT observed behaviour, we proceeded to validate the model, by assessing its ability to reproduce the behaviours experimentally observed in the different C. botulinum mutant strains we had considered in the study. We examined four different mutations, which are implemented in the WT model exclusively by changing the initial state of the model, i.e. without any change to the kinetics of the reactions. The mutants we considered for the purposes of our validation are as follows:
The cbo0786 mutant constructed by insertional inactivation in Zhang et al., 2013 [31], is denoted as C786_M model, and addressed by setting the initial value of the prCBOi variable to zero.
The codY mutant constructed by insertional inactivation in Zhang et al., 2014 [29], is denoted as CODY_M model, and addressed by setting the initial values of the CodY1 and CodY2 variables to zero.
The Hall/707 and Hall/714 mutants, constructed by the insertion of DNA anti-sense mRNA strains for the two positive regulatory TCSs CBO_1042/CBO_1041 and CBO_1967/ CBO_1968 in Connan et al., 2012 [30], denoted as RR_M model, and addressed by setting the initial value of the CRR variable to zero.
The Hall/1146 mutant, constructed by the insertion of DNA anti-sense mRNA strains for the positive regulatory TCS CBO_0608/CBO_0607 in Connan et al., 2012 [30], is denoted as C607_M model, and addressed by setting the initial value assigned to the CBO_0607 variable to zero.
For each mutant, we obtain and report the toxigenesis predictions (pattern and amount of BoNT) from both the WT model and the mutant model. Then we examine the relationship between model predictions and the experimental results to determine the ability of the models to reproduce wet-lab evidence.
Comparison with cbo0786 mutant
We summarize in Table 1 the results reported in Fig 4 (upper panel) and 5A of Zhang et al., 2013 [31] for C. botulinum strain ATCC 3502, which we call wild-type (wt), and for the cbo0786 mutant, which we call mut. In the experiments of Zhang et al., the relative expression of the bont gene and the amount of neurotoxin in the supernatant are quantified at three time points: mid-exponential growth phase (ME, approx. 4 hours), late-exponential growth (LE, approx. 7 hours) and at early-stationary phase (ES, approx. 10 hours).
Table 1
Data for bont gene expression and supernatant toxin concentration of C. botulinum ATCC 3502 (wt) and cbo0786 mutant (mut), measured at mid-exponential (ME), late-exponential (LE) and early-stationary (ES) phases.
Relative expression of bont (ELISA, normalized to 16S rn) | Neurotoxin in supernatant (OD at 405 nm) | |||
---|---|---|---|---|
wt | mut | wt | mut | |
ME | 0.85 | 1.60 | 0.65 | 0.88 |
LE | 2.60 | 7.70 | 0.38 | 0.90 |
ES | 2.10 | 4.50 | 0.39 | 1.50 |
We compare our models predictions with the experimental results by showing, in Fig 6A , a graphical representation of the data collected by Zhang et al., 2013 [31] for the amount of toxin in the supernatant (i.e. data in the right columns of Table 1 ) and in Fig 6B the equivalent measures as obtained from our models predictions.
(A): normalized concentration of toxin in the supernatant for C. botulinum ATCC 3502 (wt) and the cbo0786 mutant (mut) as reported in [31] (B): model prediction for toxin concentration in the supernatant (normalized) for wt and for the C786_M mutant (mut).
To make the Zhang et al. data (which reports concentrations as A at 405 nm) comparable to our model results (which predicts concentrations as MLD50/ml), we normalized both the wt and mut data to the maximal measured toxin concentration, which in both Fig 6A and 6B corresponds to the amount of toxin measured at data-point ES for mut. Also, we defined the ME, LE and ES time points for the model simulated cultures to be 14, 16 and 18 hours, respectively.
Comparing Fig 6A with Fig 6B indicates that the experimentally measured and the modelled wt are quite different in terms of the pattern of toxin production. For wt the experimental peak of neurotoxin concentration appears in the culture at the ME measurement time a behaviour that is remarkably distinct from that of the C. botulinum strain ATCC 19397 we considered as the basis of our modelling in this work. Our model is however able to reproduce the increase in toxigenesis induced by the silencing of the cbo0786 gene. Indeed, as can be appreciated from Fig 6B , the model of the mutant (mut) consistently produces higher amounts of toxin in the supernatant.
Comparison with codY mutant
Zhang et al., 2014 [29] measured the amount of toxin in the supernatant in a culture of C. botulinum strain ATCC 3502, which we will consider as wild-type (wt) in this section, and for a codY mutant constructed by insertional inactivation (mut, in this section). Data for the measured concentration of toxin in the supernatant of the cultures of wt and mut as a function of time are summarized in Table 2 . These data have been extracted from Figure 3, page 7654 of [29].
Table 2
Data for the supernatant toxin concentration of C. botulinum ATCC 3502 (wt) and codY mutant (mut), measured in μg/ml at various time points during the culture growth.
Time (hours) | ||||||||
---|---|---|---|---|---|---|---|---|
5 | 6 | 9 | 12 | 24 | 48 | 96 | ||
Toxin concentration in supernatant (μg/ml) | wt | 0.12 | 0.22 | 0.38 | 6.95 | 41.5 | 60.5 | 50.5 |
mut | 0.06 | 0.11 | 0.14 | 2.95 | 17.5 | 30.5 | 29.5 |
To compare our model results with the experimental data reported in Table 2 , we defined a sequence of time points that would match the culture growth phase observation times of Zhang et al., 2014 [29]. In their report the peak of neurotoxin concentration in the wt culture is achieved at time 48 hours and the transition between late exponential and early stationary growth phases occurs at time 9 hours. Therefore, we define the observation time points for the modelled cultures to match those distinctive events (time 17.5 hours for the transition from late-exponential to early-stationary phases, and time 22 hours for the peak of toxin concentration in the supernatant) and we show the comparison between the experimental data and the models predictions in Fig 7 . To facilitate the comparison, we denote the two sequences of time points as t1,t2,…t7. Fig 7A shows the log of the toxin concentration in the supernatant, as obtained in the experimental work of Zhang et al., 2014 [29], and Fig 7B shows the analogous results obtained from our models. As can be observed, there is a good agreement between experiments and model predictions (particularly the relative values for wild type and mutant).
(A): normalized observed concentration of toxin in the supernatant for C. botulinum ATCC 3502 (wt) and the codY mutant (mut), as reported in [29]. (B): model prediction for toxin concentration in the supernatant (normalized) for wt and for the CODY_M mutant (mut).
Comparison with Hall/707, Hall/714 and Hall/1146 mutants
Connan and co-authors report in [30] the results of experimental work investigating the role of various two-component systems in toxigenesis regulation. They compare the amount of toxin in the supernatant of a C. botulinum type A Hall strain culture against the toxin in the supernatant for different mutants in which the two-component systems have been silenced. Of interest for our purposes are the Hall/707 and Hall/714 mutants, for which we have built a model named CRR_M, and the Hall/1146 mutant, for which we constructed a model called C607_M. Since the Hall/707 and Hall/714 mutants provide practically identical results in terms of the amount of toxin produced in the supernatant, we only consider Hall/707 in the following. We denote by wild-type (wt) the original C. botulinum type A Hall strain, and by CRR_M and C607_M the two mutant strains Hall/707 and Hall/1146.
In [30], on page 8, Figure 3D, the authors reported the measured amounts of toxin concentration in the supernatant (A at 405nm), for 3 different time points at 8 hours which corresponds to a point in the exponential growth phase, at 12 hours, in the early stationary phase, and at 24 hours, well inside the stationary phase.
To compare the model predictions with the experimental data reported in Table 3 , we choose three time points in the predicted time courses of the toxin supernatant concentration: time 14.5 hours for the exponential growth phase, time 18.5 hours for the stationary phase and time 24 hours for the stationary phase. In Fig 8 we show the amounts of toxin in the supernatant (normalized with respect to the maximum amount, which in all cases corresponds with data for wt in the stationary phase) coming from the experiments in Connan et. al [30] ( Fig 8A ) and from our model predictions ( Fig 8B ). It can be seen that the models can reproduce the reduced toxigenesis of both mutant phenotypes and can also identify that the C607_M mutant (i.e. the Hall/1146 strain) exhibits a larger reduction in toxin concentration.
Table 3
Data for supernatant toxin concentration of C. botulinum type A Hall (wt) and the mutants Hall/707 (CRR_M) and Hall/1146 (C607_M), measured during the Exponential growth phase (time 8 hours), the early stationary phase (time 12 hours) and the stationary phase (time 24 hours).
Neurotoxin in supernatant(A at 405 nm) | |||
---|---|---|---|
wt | CRR_M | C607_M | |
Exponential | 12 | 4.8 | 1.5 |
Early stationary | 50 | 4.7 | 3 |
Stationary | 250 | 7 | 5 |
(A), experimentally measured amounts of toxin concentration in the supernatant, normalized by the maximal measured concentration (for wt, in the stationary phase) and reported on a log scale. (B), predicted toxin concentrations from our wt, CRR_M and C607_M mutant models, normalized by the maximal predicted concentration (for wt, in the stationary phase), log scale on the vertical axis.
Abstract
Genome sequences of five different Group II (nonproteolytic) Clostridium botulinum type F6 strains were compared at a 50-kb locus containing the neurotoxin gene cluster. A clonal origin for these strains is indicated by the fact that sequences were identical except for strain Eklund 202F, with 10 single-nucleotide polymorphisms and a 15-bp deletion. The essential topB gene encoding topoisomerase III was found to have been split by the apparent insertion of 34.4 kb of foreign DNA (in a similar manner to that in Group II C. botulinum type E where the rarA gene has been disrupted by a neurotoxin gene cluster). The foreign DNA, which includes the intact 13.6-kb type F6 neurotoxin gene cluster, bears not only a newly introduced topB gene but also two nonfunctional botulinum neurotoxin gene remnants, a type B and a type E. This observation combined with the discovery of bacteriophage integrase genes and IS4 elements suggest that several rounds of recombination/horizontal gene transfer have occurred at this locus. The simplest explanation for the current genotype is that the ancestral bacterium, a Group II C. botulinum type B strain, received DNA firstly from a strain containing a type E neurotoxin gene cluster, then from a strain containing a type F6 neurotoxin gene cluster. Each event disrupted the previously functional neurotoxin gene. This degree of successive recombination at one hot spot is without precedent in C. botulinum, and it is also the first description of a Group II C. botulinum genome containing more than one neurotoxin gene sequence.
The work was supported by Norwegian Defence Research Establishment (FFI) and the University of Oslo. We thank Jaran Strand Olsen and Janet Blatny for critically reading the manuscript.
Artin, I., Carter, A. T., Holst, E., Lovenklev, M., Mason, D. R., Peck, M. W., et al. (2008). Effects of carbon dioxide on neurotoxin gene expression in nonproteolytic Clostridium botulinum type E. Appl. Environ. Microbiol. 74, 2391. doi: 10.1128/AEM.02587-07
Austin, J. W., and Leclair, D. (2011). Botulism in the North: a disease without borders. Clin. Infect. Dis. 52, 593. doi: 10.1093/cid/ciq256
Babinszky, G., Csitari, G., and Jozsa, S. (2008). Observations on environmental factors in connection with avian botulism outbreaks in a Hungarian wetland habitat. Acta Microbiol. Immunol. Hung. 55, 455. doi: 10.1556/AMicr.55.2008.4.10
Barash, J. R., and Arnon, S. S. (2014). A novel strain of Clostridium botulinum that produces type B and type H botulinum toxins. J. Infect. Dis. 209, 183. doi: 10.1093/infdis/jit449
Bohnel, H. (2002). Household biowaste containers (bio-bins) – potential incubators for Clostridium botulinum and botulinum neurotoxins. Water Air Soil Pollut. 140, 335. doi: 10.1023/A:1020169520369
Bott, T. L., Johnson, J., Foster, E. M., and Sugiyama, H. (1968). Possible origin of high incidence of Clostridium botulinum type E in an inland bay (Green Bay of Lake Michigan). J. Bacteriol. 95, 1542.
Bradshaw, M., Dineen, S. S., Maks, N. D., and Johnson, E. A. (2004). Regulation of neurotoxin complex expression in Clostridium botulinum strains 62A, Hall A-hyper, and NCTC 2916. Anaerobe 10, 321. doi: 10.1016/j.anaerobe.2004.07.001
Buchanan, R. E., and Gibbons, N. E. (1974). Bergey’s Manual of Determinative Bacteriology, 8th Edn. Baltimore: Williams & Wilkins.
Byappanahalli, M. N., and Whitman, R. L. (2009). Clostridium botulinum type E occurs and grows in the alga Cladophora glomerata. Can. J. Fish. Aquat. Sci. 66, 879. doi: 10.1139/F09-052
Cattaneo, A., Hudon, C., Vis, C., and Gagnon, P. (2013). Hydrological control of filamentous green algae in a large fluvial lake (Lake Saint-Pierre, St. Lawrence River, Canada). J. Great Lakes Res. 39, 409. doi: 10.1016/j.jglr.2013.06.005
Chatla, K., Gaunt, P. S., Hanson, L., Gao, D. X., and Wills, R. (2012). Determination of the median lethal dose of botulinum serotype E in channel catfish fingerlings. J. Aquat. Anim. Health 24, 105. doi: 10.1080/08997659.2012.675935
Chea, F. P., Chen, Y. H., Montville, T. J., and Schaffner, D. W. (2000). Modeling the germination kinetics of Clostridium botulinum 56A spores as affected by temperature, pH, and sodium chloride. J. Food Prot. 63, 1071.
Chun, C. L., Ochsner, U., Byappanahalli, M. N., Whitman, R. L., Tepp, W. H., Lin, G. Y., et al. (2013). Association of toxin-producing Clostridium botulinum with the macroalga cladophora in the Great Lakes. Environ. Sci. Technol. 47, 2587. doi: 10.1021/es304743m
Collins, M. D., and East, A. K. (1998). Phylogeny and taxonomy of the food-borne pathogen Clostridium botulinum and its neurotoxins. J. Appl. Microbiol. 84, 5. doi: 10.1046/j.1365-2672.1997.00313.x
Contreras de Vera, A., Garcia Fernandez, A., Cubero Pablo, M. J., and Leon-Vizcaino, L. (1991). Low prevalence of C. botulinum in Spanish wetlands with a tidal regime. Vet. Rec. 128, 187. doi: 10.1136/vr.128.8.187
Cooksley, C. M., Davis, I. J., Winzer, K., Chan, W. C., Peck, M. W., and Minton, N. P. (2010). Regulation of neurotoxin production and sporulation by a putative agrBD signaling system in proteolytic Clostridium botulinum. Appl. Environ. Microbiol. 76, 4448. doi: 10.1128/AEM.03038-09
Derman, Y., Lindstrom, M., Selby, K., and Korkeala, H. (2011). Growth of group II Clostridium botulinum strains at extreme temperatures. J. Food Prot. 74, 1797. doi: 10.4315/0362-028X.JFP-11-187
Dolman, C. E. (1960). Type E botulism: a hazard of the north. Arctic 13, 230. doi: 10.14430/arctic3704
Duncan, R. M., and Jensen, W. L. (1976). A relationship between avian carcasses and living invertebrates in the epizootiology of avian botulism. J. Wildl. Dis. 12, 116. doi: 10.7589/0090-3558-12.1.116
Eales, C. E., and Turner, A.W. (1952). Description of Clostridium botulinum type-D recovered from soil in South Australia. Austr. J. Exp. Biol. Med. Sci. 30, 295. doi: 10.1038/icb.1952.27
Eklund, M. W., Poysky, F. T., Peterson, M. E., Peck, L. W., and Brunson, W. D. (1984). Type-E botulism in Salmonids and conditions contributing to outbreaks. Aquaculture 41, 293. doi: 10.1016/0044-8486(84)90198-4
Evelsizer, D. D., Clark, R. G., and Bollinger, T. K. (2010). Relationships between local carcass density and risk of mortality in molting mallards during avian botulism outbreaks. J. Wildl. Dis. 46, 507. doi: 10.7589/0090-3558-46.2.507
Fagan, R. P., McLaughlin, J. B., Castrodale, L. J., Gessner, B. D., Jenkerson, S. A., Funk, E. A., et al. (2011). Endemic foodborne botulism among Alaska Native persons-Alaska, 1947. Clin. Infect. Dis. 52, 585. doi: 10.1093/cid/ciq240
Franciosa, G., Maugliani, A., Scalfaro, C., and Aureli, P. (2009). Evidence that plasmid-borne botulinum neurotoxin type B genes are widespread among Clostridium botulinum serotype B strains. PLoS ONE 4:e4829. doi: 10.1371/journal.pone.0004829
Friend, M. (2002). Avian disease at the Salton Sea. Hydrobiologia 473, 293. doi: 10.1023/A:1016570810004
Fu, S.-W., and Wang, C.-H. (2008). An overview of type E botulism in China. Biomed. Environ. Sci. 21, 353. doi: 10.1016/S0895-3988(08)60054-9
Gao, Q. Y., Huang, Y. F., Wu, J. G., Liu, H. D., and Xia, H. Q. (1990). A review of botulism in China. Biomed. Environ. Sci. 3, 326.
Getchell, R. G., and Bowen, P. R. (2006). Ecology of type E botulism within dreissenid mussel beds. Aquat. Invaders 17, 1𠄸.
Giraffa, G., Chanishvili, N., and Widyastuti, Y. (2010). Importance of lactobacilli in food and feed biotechnology. Res. Microbiol. 161, 480. doi: 10.1016/j.resmic.2010.03.001
Girardin, H., Albagnac, C., Dargaignaratz, C., Nguyen-The, C., and Carlin, F. T. (2002). Antimicrobial activity of foodborne Paenibacillus and Bacillus spp. against Clostridium botulinum. J. Food Prot. 65, 806.
Hannett, G. E., Stone, W. B., Davis, S. W., and Wroblewski, D. (2011). Biodiversity of Clostridium botulinum type E associated with a large outbreak of botulism in wildlife from Lake Erie and Lake Ontario. Appl. Environ. Microbiol. 77, 1061. doi: 10.1128/AEM.01578-10
Hauschild, A. H. W., and Gauvreau, L. (1985). Foodborne botulism in Canada, 1971. Can. Med. Assoc. J. 133, 1141.
Heckman, C. W. (1986). The role of marsh plants in the transport of nutrient as shown by a quantitative model for the freshwater section of the Elbe estuary. Aquat. Bot. 25, 139. doi: 10.1016/0304-3770(86)90050-1
Hielm, S., Hyytia, E., Andersin, A. B., and Korkeala, H. (1998). A high prevalence of Clostridium botulinum type E in Finnish freshwater and Baltic Sea sediment samples. J. Appl. Microbiol. 84, 133. doi: 10.1046/j.1365-2672.1997.00331.x
Hill, K. K., and Smith, T. J. (2013). Genetic diversity within Clostridium botulinum serotypes, botulinum neurotoxin gene clusters and toxin subtypes. Curr. Top. Microbiol. Immunol. 364, 1. doi: 10.1007/978-3-642-33570-9_1
Hill, K. K., Xie, G., Foley, B. T., Smith, T. J., Munk, A. C., Bruce, D., et al. (2009). 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. BMC Biol. 7:66. doi: 10.1186/1741-7007-7-66
Hinderink, K., Lindstrom, M., and Korkeala, H. (2009). Group I Clostridium botulinum strains show significant variation in growth at low and high temperatures. J. Food Prot. 72, 375. doi: 10.4315/0362-028X.JFP-11-187
Hubalék, Z., and Halouzka, J. (1991). Persistence of Clostridium botulinum type-C in blow fly (Calliphoridae) larvae as a possible cause of avian botulism in spring. J. Wildl. Dis. 27, 1, 81. doi: 10.7589/0090-3558-27.1.81
Huss, H. H. (1980). Distribution of Clostridium botulinum. Appl. Environ. Microbiol. 39, 764.
Johannsen, A. (1963). Clostridium botulinum in Sweden and adjacent waters. J. Appl. Bacteriol. 26, 43. doi: 10.1111/j.1365-2672.1963.tb01153.x
Karasawa, T., Wang, X., Maegawa, T., Nakamura, S., Hang’ombe, B. M., and Isogai, E. (2000). Demonstration of botulinum toxins of types B and D in soil samples from Zambia. Ann. Trop. Med. Parasitol. 94, 409.
Koepke, R., Sobel, J., and Arnon, S. S. (2008). Global occurrence of infant botulism, 1976. Pediatrics 122, e73. doi: 10.1542/peds.2007-1827
Kouguchi, H., Suzuki, T., Hasegawa, K., Mutoh, S., Watanabe, T., Niwa, K., et al. (2006). Quantitative detection of gene expression and toxin complex produced by Clostridium botulinum serotype D strain 4947. J. Microbiol. Methods 67, 416. doi: 10.1016/j.mimet.2006.04.013
Lafrancois, B. M., Riley, S. C., Blehert, D. S., and Ballmann, A. E. (2011). Links between type E botulism outbreaks, lake levels, and surface water temperatures in Lake Michigan, 1963. J. Great Lakes Res. 37, 86. doi: 10.1016/j.jglr.2010.10.003
Lalitha, K. V., and Gopakumar, K. (2000). Distribution and ecology of Clostridium botulinum in fish and aquatic environments of a tropical region. Food Microbiol. 17, 535. doi: 10.1006/fmic.2000.0346
Lalitha, K. V., and Surendran, P. K. (2002). Occurrence of Clostridium botulinum in fresh and cured fish in retail trade in Cochin (India). Int. J. Food Microbiol. 72, 169. doi: 10.1016/S0168-1605(01)00632-8
Leclair, D., Farber, J. M., Doidge, B., Blanchfield, B., Suppa, S., Pagotto, F., et al. (2013a). Distribution of Clostridium botulinum type E strains in Nunavik, Northern Quebec, Canada. Appl. Environ. Microbiol. 79, 646. doi: 10.1128/AEM.05999-11
Leclair, D., Fung, J., Isaac-Renton, J. L., Proulx, J. F., May-Hadford, J., Ellis, A., et al. (2013b). Foodborne botulism in Canada, 1985. Emerg. Infect. Dis. 19, 961. doi: 10.3201/eid1906.120873
Long, S. C., and Tauscher, T. (2006). Watershed issues associated with Clostridium botulinum: a literature review. J. Water Health 4, 277.
Miller, L. G. (1975). Observations on distribution and ecology of Clostridium botulinum type E in Alaska. Can. J. Microbiol. 21, 920. doi: 10.1139/m75-136
Murrell, W. G., and Stewart, B. J. (1983). Botulism in New South-Wales, 1980. Med. J. Aust. 1, 13.
Nightingale, K. W., and Ayim, E. N. (1980). Outbreak of botulism in Kenya after ingestion of white ants. Br. Med. J. 281, 1682. doi: 10.1136/bmj.281.6256.1682-a
Nol, P., Rocke, T. E., Gross, K., and Yuill, T. M. (2004). Prevalence of neurotoxic Clostridium botulinum type C in the gastrointestinal tracts of tilapia (Oreochromis mossambicus) in the Salton Sea. J. Wildl. Dis. 40, 414. doi: 10.7589/0090-3558-40.3.414
Peck, M. W. (2009). Biology and genomic analysis of Clostridium botulinum. Adv. Microb. Physiol. 55, 183. doi: 10.1016/S0065-2911(09)05503-9
Perez-Fuentetaja, A., Clapsadl, M. D., Einhouse, D., Bowser, P. R., Getchell, R. G., and Lee, W. T. (2006). Influence of limnological conditions on Clostridium botulinum type E presence in eastern Lake Erie sediments (Great Lakes, USA). Hydrobiologia 563, 189. doi: 10.1007/s10750-005-0011-1
Perez-Fuentetaja, A., Clapsadl, M. D., Getchell, R. G., Bowser, P. R., and Lee, W. T. (2011). Clostridium botulinum type E in Lake Erie: Inter-annual differences and role of benthic invertebrates. J. Great Lakes Res. 37, 238. doi: 10.1016/j.jglr.2011.03.013
Quortrup, E. R., and Holt, A. L. (1941). Detection of potential botulinus-toxin-producing areas in Western Duck Marshes with suggestions for control. J. Bacteriol. 41, 363.
Riley, S. C., Munkittrick, K. R., Evans, A. N., and Krueger, C. C. (2008). Understanding the ecology of disease in Great Lakes fish populations. Aquat. Ecosyst. Health Manag. 11, 321. doi: 10.1080/14634980802301638
Rocke, T. E., Euliss, N. H., and Samuel, M. D. (1999). Environmental characteristics associated with the occurrence of avian botulism in wetlands of a northern California refuge. J. Wildl. Manag. 63, 358. doi: 10.2307/3802520
Rocke, T. E., and Samuel, M. D. (1999). Water and sediment characteristics associated with avian botulism outbreaks in wetlands. J. Wildl. Manag. 63, 1249. doi: 10.2307/3802842
Sandler, R. J., Rocke, T. E., Samuel, M. D., and Yuill, T. M. (1993). Seasonal prevalence of Clostridium-botulinum type-C in sediments of a Northern California wetland. J. Wildl. Dis. 29, 533. doi: 10.7589/0090-3558-29.4.533
Sandler, R. J., Rocke, T. E., and Yuill, T. M. (1998). The inhibition of Clostridium botulinum type C by other bacteria in wetland sediments. J. Wildl. Dis. 34, 4, 830. doi: 10.7589/0090-3558-34.4.830
Segner, W. P., Schmidt, C. F., and Boltz, J. K. (1971). Enrichment, isolation, and cultural characteristics of marine strains of Clostridium botulinum type C. Appl. Microbiol. 22, 1017.
Shapiro, R. L., Hatheway, C., and Swerdlow, D. L. (1998). Botulism in the United States: a clinical and epidemiologic review. Ann. Intern. Med. 129, 221. doi: 10.7326/0003-4819-129-3-199808010-00011
Shin, N. R., Byun, S. H., Chun, J. H., Shin, J. H., Kim, Y. J., Kim, J. H., et al. (2010). An outbreak of type C botulism in waterbirds: Incheon, Korea. J. Wildl. Dis. 46, 912. doi: 10.7589/0090-3558-46.3.912
Skulberg, A., and Holt, G. (1987). 𠇊vian botulism in scandinavia,” in Avian Botulism, an International Perspective, eds M. W. Eklund and V. R. Dowell, Jr. (Springfield, IL: Charles C Thomas), 107.
Smith, G. R., Milligan, R. A., and Moryson, C. J. (1978). Clostridium botulinum in aquatic environments in Great Britain and Ireland. J. Hyg. 80, 431. doi: 10.1017/S0022172400024906
Smith, L. D. (1975). Inhibition of Clostridium botulinum by strains of Clostridium perfringens isolated from soil. Appl. Microbiol. 30, 319.
Stringer, S. C., Carter, A. T., Webb, M. D., Wachnicka, E., Crossman, L. C., Sebaihia, M., et al. (2013). Genomic and physiological variability within group II (non-proteolytic) Clostridium botulinum. BMC Genomics 14:333. doi: 10.1186/1471-2164-14-333
Stringer, S. C., Webb, M. D., and Peck, M. W. (2011). Lag time variability in individual spores of Clostridium botulinum. Food Microbiol. 28, 228. doi: 10.1016/j.fm.2010.03.003
Suen, J. C., Hatheway, C. L., Steigerwalt, A. G., and Brenner, D. J. (1988). Clostridium argentinense sp. nov.: a genetically homogeneous group composed of all strains of Clostridium botulinum toxin type G and some nontoxigenic strains previously identified as Clostridium subterminale or Clostridium hastiforme. Int. J. Syst. Bacteriol. 38, 375. doi: 10.1099/00207713-38-4-375
Suhadi, F., Thayib, S. S., and Sumarsono, N. (1981). Distribution of Clostridium botulinum around fishing areas of the western part of Indonesian waters. Appl. Environ. Microbiol. 41, 1468.
Takeda, M., Tsukamoto, K., Kohda, T., Matsui, M., Mukamoto, M., and Kozaki, S. (2005). Characterization of the neurotoxin produced by isolates associated with avian botulism. Avian Dis. 49, 376. doi: 10.1637/7347-022305R1.1
Umeda, K., Wada, T., Kohda, T., and Kozaki, S. (2013). Multi-locus variable number tandem repeat analysis for Clostridium botulinum type B isolates in Japan: comparison with other isolates and genotyping methods. Infect. Gen. Evol. 16, 298. doi: 10.1016/j.meegid.2013.02.022
Vidal, D., Anza, I., Taggart, M. A., Perez-Ramrez, E., Crespo, E., Hofle, U., et al. (2013). Environmental factors influencing the prevalence of a Clostridium botulinum type C/D mosaic strain in nonpermanent Mediterranean wetlands. Appl. Environ. Microbiol. 79, 4264. doi: 10.1128/AEM.01191-13
Webb, M. D., Pin, C., Peck, M. W., and Stringer, S. C. (2007). Historical and contemporary NaCl concentrations affect the duration and distribution of lag times from individual spores of nonproteolytic Clostridium botulinum. Appl. Environ. Microbiol. 73, 2118. doi: 10.1128/AEM.01744-06
Wetzel, R. G. (2001). Limnology. Lake and River Ecosystems, Third Edn. San Diego, CA: Academic Press.
Woudstra, C., Skarin, H., Anniballi, F., Fenicia, L., Bano, L., Drigo, I., et al. (2012). Neurotoxin gene profiling of Clostridium botulinum types C and D native to different countries within Europe. Appl. Environ. Microbiol. 78, 3120. doi: 10.1128/AEM.07568-11
Yamakawa, K., Kamiya, S., Nishida, S., Yoshimura, K., Yu, H., Lu, D. Y., et al. (1988). Distribution of Clostridium botulinum in Japan and in Shinkiang district of China. Microbiol. Immunol. 32, 579. doi: 10.1111/j.1348-0421.1988.tb01419.x
Yamakawa, K., Kamiya, S., Yoshimura, K., Nakamura, S., and Ezaki, T. (1990). Clostridium botulinum in the soil of Kenya. Ann. Trop. Med. Parasitol. 84, 201.
Yamakawa, K., and Nakamura, S. (1992). Prevalence of Clostridium botulinum type E and coexistence of C. botulinum nonproteolytic type B in the river soil of Japan. Microbiol. Immunol. 36, 583. doi: 10.1111/j.1348-0421.1992.tb02058.x
Yule, A. M., Barker, I. K., Austin, J. W., and Moccia, R. D. (2006). Toxicity of Clostridium botulinum type E neurotoxin to Great Lakes fish: implications for avian botulism. J. Wildl. Dis. 42, 479. doi: 10.7589/0090-3558-42.3.479
Zechmeister, T. C., Kirschner, A. K. T., Fuchsberger, M., Gruberl, S. G., Suss, B., Rosengarten, R., et al. (2005). Prevalence of botulinum neurotoxin C1 and its corresponding gene in environmental samples from low and high risk avian botulism areas. ALTEX 22, 185.
Zulkifly, S., Hanshew, A., Young, E. B., Lee, P., Graham, M. E., Graham, M. E., et al. (2012). The epiphytic microbiota of the globally widespread macroalga Cladophora glomerata (Chlorophyta, Cladophorales). Am. J. Bot. 99, 1541. doi: 10.3732/ajb.1200161
Keywords : Clostridium botulinum, botulism, serotype, spore, anaerobe, lakes, wetlands, soil
Citation: Espelund M and Klaveness D (2014) Botulism outbreaks in natural environments – an update. Front. Microbiol. 5:287. doi: 10.3389/fmicb.2014.00287
Received: 28 February 2014 Accepted: 24 May 2014
Published online: 11 June 2014.
Marie Archambault, University of Montreal, Canada
John W. Austin, Bureau of Microbial Hazards – Health Canada, Canada
Copyright © 2014 Espelund and Klaveness. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Mari Espelund, Protection and Societal Security Division, Norwegian Defence Research Establishment, P.O. Box 25, N-2027 Kjeller, Norway e-mail: [email protected]
† Mari Espelund and Dag Klaveness have contributed equally to this work.