BoNT/G is the least-studied and the most recently reported of the seven serotypes produced by C. botulinum. Although BoNT/G is associated with a distinct species and metabolic group, the toxin shares multiple characteristics with the other six progenitor toxins. The seven serotypes have similar biochemical and molecular mechanisms of cell entry and membrane translocation. They cause disease by inhibiting synaptic transmission as a result of the enzymatic cleavage of the SNARE protein complex. In the present work, we detail the in silico comparison of BoNT/G progenitor toxin proteins to the other six serotypes of C. botulinum, as well as methods for the digestion, detection, and relative quantification of BoNT/G and its NAPs.
The comparison of the BoNT/G progenitor toxin with the other six serotypes was completed to determine/G's phenotypic relationship with the other BoNTs. In general, past analyses [7
] have included a comparison at the gene level; this study focuses solely on protein level. While comparisons of toxin and NTNH proteins to select serotypes have been previously described [23
], a complete comparison of all/G complex proteins (toxin, NTNH, HA70, and HA17) with the other six serotypes has not been previously reported. Phenetic analysis confirmed that the BoNT/G complex of proteins shared the most similarity with the/B serotype (Figure ), as previously reported [10
To determine the extent of/G's homology to the/B toxin serotype, we completed an in-depth comparison of six/B subtypes, 22 different accession numbers (Figure , additional files 2
). The comparison of individual domains--translocation domain, binding domain NT, binding domain CT, and peptidase--revealed the area of the toxin in which/G shares the greatest (translocation domain) and least (binding domain CT) similarity. Overall, each domain compared, between the two toxins, is greater than 50% similar. This comparison helped to determine which substrate peptide would be optimal to test the activity of/G. Although there are no direct indications that sequence similarity would imply overall identical functionality, similar sequences would allow similar crystal structures to form, suggesting similar functionality [24
]. It is currently known that both BoNT/B and/G cleave the Synaptobrevin protein;/B cleaves a Gln76
bond and/G an Ala81
bond five amino acids downstream (Table ). Because the cleavage sites of both toxins are relatively near one another--thus the similarity of their binding domain sequences and therefore structures--the same peptide substrate currently used to test/B activity was used to test/G activity [19
In order to confirm that the commercial BoNT/G complex was active and therefore could be considered analogous to the toxin complex found in clinical samples, various dilutions of the commercial toxin were tested using the Endopep-MS method previously described (Figure ) [19
]. In addition to confirming the toxin's activity, the Endopep-MS experiments indicated a new optimum temperature for/G activity. When reactions were pulsed at 47°C for 10 min, followed by incubation at 42°C for at least eight hours--as opposed to 37°C for a minimum of 17 hr--an increase in activity and in the quality of mass spectra produced was observed. Other serotypes of BoNT (/C and/D) are often associated with botulism in animals, avians, equines, and bovines, whose body temperatures are higher than those of humans. BoNT/G has yet to be associated with botulism in a particular organism; however, it is possible that/G would be more effective at causing disease in an organism with a higher body temperature than that of humans, similar to BoNT/C and/D.
Figure 6 Endopep-MS method confirmation of commercial BoNT/G activity. This is a representative spectrum indicating BoNT/G activity on a specific substrate peptide. 1Intact substrate, 2C-Terminus product mass 1762.9, and 3N-Terminus product mass 2281.8. The sequences (more ...)
Proteomic strategies and analyses used in this study were important to help define the characteristics of proteins associated with the BoNT/G complex. The 1D-SDS PAGE analysis confirmed the presence of the four expected complex proteins (BoNT, NTNH, HA70, and HA17), with relatively high sequence coverage for in gel digestion (Figure ). As expected the proteins, P21 and HA33, were not identified. P21, a positive regulator of gene expression, lies just upstream of NTNH on the toxin plasmid (Figure ) [10
]. The purpose of P21, in complex development, is not completely understood and previous reports have not identified it as part of the/G complex [11
]. HA33, a hemagglutinin component, is not found on the/G plasmid. The lack of evidence of the protein's presence further endorsed the theory that, unlike the other serotypes, HA33 is not associated with the/G complex [10
]. Two gel slices (Figure ; #6 and 11) out of 17 visually had protein but did not return any identifiable peptides when digested and analyzed. This could be due to a number of factors: the protein was relatively difficult to digest, there was not a sufficient amount of protein to digest, the sequence was not present in the database used, or post-translational modifications (PTMs) altered the protein sequence and did not allow for identification. The SDS-Page gel and in gel digestions confirmed visually and analytically which proteins are present in the commercial toxin complex and allowed us to continue to in solution digestions with some prior knowledge of which proteins should be identified.
As anticipated, the same proteins that were identified with the in gel digestions were also identified in the analysis of the in solution digestions. The four main complex components-- BoNT, NTNH, HA70, and HA17--were all identified with high confidence, and returned a large number of peptides. Hines et al. reported the use of a reduction and alkylation overnight digestion method that produced sequence coverages of 16% for BoNT, 10% for NTNH, 38% for HA70, and 49% for HA17 [18
]. The method used in our study allowed the recovery of more than four times the sequence coverage for BoNT at 66%, more than five times for NTNH at 57%, and more than double for both HA70 and HA17 at 91% and 99%, respectively.
BoNT complexes are difficult to digest in solution [18
]. This rapid high-temperature digestion method does not involve reduction and alkylation, unlike classical methods; instead, it uses an acid labile surfactant to solubilize the hydrophobic proteins. The increased solubility allows a denatured protein to be more susceptible to tryptic digestion, thereby increasing the rate of digestion and the number of tryptic peptides produced [25
]. It has also been previously reported that the use of high temperature for a short period of time is the best condition for the enzymatic activity of trypsin [26
This BoNT complex digestion method, in addition to analysis of the samples on two different electrospray (ESI) MS instruments using data-dependent (DDA) and data-independent MSE analysis, allowed for the detection of a greater number of peptides for each protein, leading to a greater overall sequence coverage than had previously been reported. This sequence coverage lends insight into the complex proteins being studied. A high percentage of sequence coverage indicates that there are few PTMs associated with the proteins, as well as no truncation. The presence of PTMs has been known to compromise protein identification, and truncated proteins do not function as expected.
In addition to providing enhanced sequence coverage, the use of data-independent MSE
analysis and label-free quantification software allowed us to relatively quantify the amount of each protein present in the BoNT/G complex (Table ). This quantification method has the advantage of being able to provide accurate estimates of relative protein abundance (often within 30% of the known values on most identified proteins in a mixture, without the much more rigorous requirements of targeted protein quantification methods. A percentage of abundance (by weight and molecules, separately) of each protein within the complex was determined, as well as an overall weight ratio of BoNT:NAPs and a molecular ratio of BoNT:NTNH:HA70:HA17. Analysis of the individual proteins within the complex illustrated that the weight of the toxin (30.4%) is almost equivalent to that of HA70 (27.8%) and about eight percent less than that of NTNH (38%); whereas HA17 makes up only a minute portion of the overall weight at just 3.7%. Conversely, analysis using molecular amounts indicated that the complex contains an equivalent amount of the toxin, NTNH, and HA17, whereas HA70 is almost twice as abundant. The nanogram and femtomole on column data sets signify a likely overall ratio of 1:3 BoNT:NAPs weight ratio and a 1:1:2:1 BoNT:NTNH:HA70:HA17 molar ratio. As stated earlier, the function of the NAPs has been shown to protect the neurotoxin in harsh environments [12
]. Due to this protective ability, in theory, a larger ratio of NAPs:BoNT, ie the greater the number of molecules of NAPs to BoNT, would protect more effectively the toxin from the acidic environment of the stomach. This potentially would increase the toxin's effectiveness at penetrating the mucosa of the intestine and entering the blood stream, increasing the toxin's chances of entering the synaptic cell and causing disease. Knowledge of the stoichiometry of proteins within the BoNT complexes would be useful to further understanding of NAPs significance and toxin potency.