Posttranslational Regulation of Botulinum Neurotoxin Production in Clostridium botulinum Hall A-hyper

ABSTRACT Botulinum neurotoxins (BoNTs) are the most toxic substances known to humankind and are the causative agents of the neuroparalytic disease botulism. Despite the overall importance of BoNTs in public health and safety, as a bioterrorism concern, and in pharmaceutical development, little is known about the molecular mechanisms mediating BoNT stability and degradation in various environments. Previous studies using Clostridium botulinum strain ATCC 3502 revealed that high levels of arginine (20 g/liter) repressed BoNT production approximately 1,000-fold. In the present study, the mechanisms of toxin reduction in arginine-enriched cultures of C. botulinum strain Hall A-hyper, which we have previously genetically manipulated using ClosTron technology, were explored. Cultures were grown in toxin production medium (TPM) and TPM enriched with arginine. Cultures were analyzed for growth (optical density at 600 nm [OD600]), changes in pH, and BoNT formation and stability. Our data indicate that arginine enrichment of C. botulinum strain Hall A-hyper cultures results in a pH shift that induces pH-dependent posttranslational control mechanisms. We further show that independent of arginine, maintenance of an acidic culture pH during growth of C. botulinum strain Hall A-hyper plays a central role in toxin stability and that an extracellular metalloprotease produced by the culture results in BoNT degradation at pH levels between ⁓6.5 and 8.0. IMPORTANCE Botulinum neurotoxin (BoNT) is a public health and bioterrorism concern as well as an important and widely used pharmaceutical, yet the regulation of its synthesis by BoNT-producing clostridia is not well understood. This paper highlights the role of environmentally controlled posttranslational regulatory mechanisms influencing processing and stability of biologically active BoNTs produced by C. botulinum. The results of this work will help enhance public health and safety measures and our ability to evaluate safety risks of novel BoNTs and improve production and quality of BoNTs for pharmaceutical use.

citrulline, ornithine, carbamoyl phosphate, ATP, and ammonia (37,38,41). Arginine deiminase pathways, consisting of a cluster of genes termed arc genes, have also been demonstrated in numerous other organisms, including clostridial species (37,(41)(42)(43)(44). In this study, homologous genes comprising an arginine catabolism pathway were identified in the genome of C. botulinum strain Hall A-hyper. Results from mutational analyses of several of these genes and from nutritional analyses indicated that the reduction of toxicity was not a direct effect of arginine or one of its metabolites but rather appeared to be related to a metabolically driven increase in culture pH during growth in excess arginine. Further studies showed pH-dependent proteolytic degradation of BoNT in culture media, indicating a strong pH-dependent, posttranslational control mechanism for toxin production by C. botulinum strain Hall A-hyper, which in turn is controlled by metabolic changes during growth.

RESULTS
Clostridium botulinum strain Hall A-hyper contains all genes necessary for a complete arginine deaminase pathway. The genome sequence of C. botulinum strain Hall A-hyper (45) was examined for the presence of homologs to genes that are known to be involved in arginine catabolism pathways in other organisms including several clostridia (37,(41)(42)(43)(44). The annotated genome of C. botulinum strain Hall A-hyper includes several arc genes central to the arginine deaminase pathway (37,(41)(42)(43), as well as an ornithine catabolism gene (ornithine cyclodeaminase, CLC2447), and two genes (argG, CLC2544, and argH, CLC2543) controlling arginine formation from citrulline. Catabolism of arginine through the ADI pathway results in the production of ammonia, citrulline, ornithine, carbamoyl phosphate, CO 2 , and ATP, in addition to other metabolites produced from these products (35,37,(41)(42)(43). Based on these data together with the identified putative involved genes in strain Hall A-hyper, an arginine catabolism pathway for this strain was constructed (Fig. 1).
Botulinum neurotoxin production by C. botulinum strain Hall A-hyper is strongly suppressed by catabolism of arginine or citrulline during growth. Previously, we had observed an ;1,000-fold suppression of active BoNT production in a C. botulinum type A strain, ATCC 3502, by 2% (0.115 M) arginine, a concentration present in arginine-rich foods such as some seeds, nuts, and meats, added to complex medium (toxin production medium [TPM]) (39). Similar suppression was observed in C. botulinum strain Hall A-hyper (Fig. 2). In order to investigate the role of arginine catabolism in toxin suppression in C. botulinum strain Hall A-hyper, the genes encoding two key enzymes in the arginine catabolism pathway, ArcB (catalyzing catabolism of citrulline to ornithine and carbamoyl phosphate) and ArgG (catalyzing conversion of citrulline to arginine through a feedback loop), were inactivated using ClosTron technology, yielding strains Hall A-hyper-argG-and -arcB- (46,47). Attempts to inactivate arcC (catalyzing conversion of carbamoyl phosphate to NH 3 , CO 2 , and ATP) using two different ClosTron constructs resulted in lack of survival of colonies, indicating that arcC is an essential gene in C. botulinum Hall A-hyper. C. botulinum strain Hall A-hyper and the two derivatives argGand arcBwere examined in complex TPM and TPM supplemented with 2% arginine. Mouse bioassay of culture supernatants and Western blot FIG 2 Repression of BoNT production in C. botulinum and derivatives by arginine and metabolites. C. botulinum strains Hall A-hyper (wt), Hall A-hyper/ArcB-, and Hall A-hyper/ArgG-were grown in toxin production medium (TPM) supplemented with arginine, citrulline, or ornithine. (A) Toxicity of trypsinized culture supernatants was determined by the mouse bioassay and is shown as mouse LD 50 /ml. Values represent the mean from four mice and standard deviation. Asterisks represent a statistically significant difference (P , 0.05). (B) Western blot assays analyzing whole-culture samples (including cell pellets and supernatant) using anti-BoNT/A specific antibody. Control is 50 ng of purified BoNT/A1. Marker is SeeBlue Plus2 prestained protein standard (Life Technologies), and the molecular weight of the marker bands is indicated on the left in kDa. NR indicates a nonreduced sample, and R indicates a sample reduced with 100 mM dithiothreitol (DTT). Numbers at left are molecular masses in kilodaltons.
analysis of whole-culture samples at day 6 of incubation showed that neither mutation affected BoNT production or growth characteristics in TPM ( Fig. 2A and B). However, supplementation of the medium with arginine or citrulline resulted in strong suppression of BoNT production in the wild type (wt) and the argGstrain but not in the arcBstrain (Fig. 2B). In addition, a strong increase in cell density for the wt and argGstrains was observed in the presence of 2% arginine and a limited increase was observed in the presence of 2% citrulline, as determined by an ;2.5and 1.5-fold increase in optical density at 600 nm (OD 600 ), respectively (see Fig. S1 in the supplemental material). Supplementation of the medium with an equal concentration of the (downstream) catabolite ornithine, on the other hand, did not affect BoNT production in either the wt or the mutant strains and did not increase cell density ( Fig. 2B and Fig. S1).
Taken together, these data indicate that arginine and citrulline must be catabolized at least past the step involving the arcB gene to result in suppression of BoNT production and that the presence of 2% arginine or citrulline in a strain lacking a functional arcB gene, which is required to further metabolize citrulline, does not result in BoNT suppression. This indicates that neither arginine nor citrulline is likely to play a direct role in BoNT suppression, while their catabolism causes the suppression. Since ornithine, a catabolite of arginine or citrulline, had no effect on toxin production, the data further suggest that ornithine or its catabolites are also not directly responsible for BoNT suppression by arginine and citrulline. Thus, other catabolites in the ADI pathway, such as carbamoyl phosphate, ATP, or ammonia may be involved in BoNT regulation (Fig. 1). Since attempts to disrupt the arcC gene were unsuccessful, as it likely is an essential gene, this hypothesis could not be tested directly.
In order to determine the minimal concentration of arginine required for BoNT repression in C. botulinum strain Hall A-hyper, triplicate cultures were grown in TPM supplemented with either 0%, 0.1%, 0.3%, 1.0%, or 2% arginine. Analysis of day 6 culture samples by Western blotting showed a significant decrease in BoNT production in medium enriched with 1% and 2% arginine compared to cultures grown in medium with arginine concentrations below #0.3% (P , 0.003) (Fig. 3B). Similarly, absorbance at OD 600 showed a divergence for peak growth and lysis patterns (Fig. 3A) as previously observed for other cultures that resulted in BoNT suppression (Fig. S1). Cultures enriched with arginine at 1% and 2% grew to a greater density with OD 600 values of 7.4 and 8.8, respectively, and did not reach full lysis until after day 6. Cultures enriched with #0.3% arginine grew as the nonenriched culture did to densities of about 2.5 to 3 OD 600 and lysed rapidly within 24 h post-stationary phase (Fig. 3A). These data indicate that the suppression of BoNT production and enhanced cell density by arginine requires at least 1% arginine with no apparent gradual dose response.
BoNT suppression by arginine in C. botulinum Hall A-hyper occurs posttranscriptionally. Since BoNT production by C. botulinum involves a series of intricate and only partially understood steps of transcription, processing, complex assembly, and proteolytic activation, we next sought to determine whether arginine catabolism affects BoNT production at the transcriptional level. Northern blot analyses of total RNA of C. botulinum strain Hall A-hyper grown in TPM or TPM with 2% arginine for up to 24 h revealed distinct ;7.5-kb (corresponding to bicistronic ntnh1bont mRNA) and ;3.9-kb (corresponding to monocistronic bont) bands with a specific bont/a mRNA probe (Fig. 4A). Peak mRNA levels were observed at about 8 to 12 h, consistent with previously reported results showing detectable bont mRNA during late log and early stationary phase only (34). The patterns for relative expression of the NTNH-BoNT-linked genes were similar between the two culture treatments, with quantitative data (including bicistronic and monocistronic mRNA) of triplicate plots showing no significant difference in overall mRNA levels but a slightly earlier peak in medium containing arginine, which corresponds to an earlier peak in cell density (OD 600 values not shown) (Fig. 4B).
Western blot analysis of cell lysates of the same cultures, on the other hand, showed significantly stronger BoNT formation for the culture grown in TPM than for the culture grown in TPM with arginine, where BoNT bands were barely detectable (Fig. 4C). In addition, while BoNT levels in the culture grown in TPM appear to increase over 24 h, the BoNT levels in the culture grown in TPM with arginine decreased after log phase. These results indicate that arginine does not directly influence transcriptional control of BoNT production.
Suppression of BoNT production by C. botulinum Hall A-hyper in medium containing 2% arginine is correlated with increased culture pH. BoNT produced by C. botulinum is predominantly released into the culture supernatant during the lysis stage of cell growth. Since cell lysis is delayed in C. botulinum strain Hall A-hyper grown in TPM with 2% arginine, BoNT levels in culture supernatant and cell pellets of cultures grown in TPM or TPM with 2% arginine for up to 4 days were examined by Western blotting. In cultures grown in TPM, increasing amounts of BoNT in culture supernatants were observed over the 4-day time period, and amounts in cell pellets increased earlier up to a peak at 24 h, followed by a slow decrease ( Fig. 5B and C). The increase of toxin in culture supernatant with correlating decrease in cell pellets after 24 h is in agreement with BoNT release from the cells during the lysis phase, although a surprisingly large amount of toxin remained associated with the cell pellet even after 96 h. In the cultures grown in TPM with 2% arginine, there was very little detectable toxin in the culture supernatant as expected. In addition, significantly less toxin was associated with the cell pellet compared to that of TPM cultures (P , 0.001), with slowly increasing amounts throughout the 4-day study period ( Fig. 5B and C). This indicates that the decrease of BoNT in culture supernatant is not due to lack of release of the toxin from the cells due to slower lysis but that BoNT formation in cells is also decreased.
To further investigate the mechanism underlying the suppression of BoNT production by arginine, culture pH was examined during growth of these cultures. While the pH of the culture grown in TPM steadily decreased during log phase from the starting pH of about 7 to an acidic pH of about 6, there consistently was a shift in pH to alkaline values (;7.5 to 7.8) at 8 h with 2% arginine supplementation (Fig. 5A). The same divergence of pH values was also observed for cultures grown in medium enriched with 1% and 2% arginine compared to cultures enriched with arginine at or below 0.3% and in the cultures used to prepare RNA samples for the Northern blot analyses (Fig. S2). Upon additional review, a correlation of BoNT suppression and increased culture pH was also observed for cultures of C. botulinum strain Hall A-hyper wt, Hall A-hyper argG-, and Hall A-hyper arcBin TPM and TPM supplemented with arginine, citrulline, or ornithine (Table S1). Together, these data strongly suggest a correlation of increased culture pH with alkaline levels and BoNT suppression.
Suppression of BoNT production by C. botulinum Hall A-hyper in medium containing 2% arginine is due to a shift to alkaline culture pH. Since arginine catabolism produces an excess of free ammonia (Fig. 1), we hypothesized that the increase in ammonia leads to an increase in culture pH, which in turn might lead to BoNT Regulation of BoNT Production instability and degradation. To test this hypothesis, BoNT production was monitored in cultures of C. botulinum grown in TPM or TPM with 2% arginine for 82 h while manually reversing the pH pattern normally observed for cultures grown in TPM or TPM with 2% arginine. Since addition of various buffers to the culture medium was not sufficient to prevent a change in pH during growth, medium was adjusted manually by repeated addition of HCl or NaOH after the log phase of growth, at which time the shift in pH naturally occurs in these cultures ( Fig. 5A and Fig. S2). C. botulinum strain Hall A-hyper wt cultures grown in TPM were adjusted to maintain an average pH of~7.4, while cultures grown in TPM with arginine were adjusted to an average pH of~6.2 (Fig. 6A). Western blot analysis of 82-h nonreduced culture samples clearly indicated that suppression of BoNT/A was correlated with pH and not with the presence of arginine in medium ( Fig. 6B and C).
This hypothesis was further examined by repeating the same assay reversing the pH in parallel with cultures of C. botulinum strains Hall A-hyper, Hall A-hyper arcB-, and Hall A-hyper argGgrown in TPM with 2% arginine. Cultures of strain Hall A-hyper arcBdropped to an acidic pH of ;5.7 to 6.0 at around 8 h of growth as expected, and a second set of the same cultures was manually adjusted to an alkaline pH of ;7.5 to 7.8. Cultures of Hall A-hyper wt or Hall A-hyper argG-, on the other hand, increased in pH at around 8 h to an alkaline pH of about 7.8, and a second set of cultures for each strain was manually adjusted to an acidic pH of 5.7 to 6.0. Western blot analysis of cultures at 96 h indicated that the previously observed BoNT/A suppression in strain Hall A-hyper arcBand wt cultures supplemented with 2% arginine (Fig. 2B) was reversed by adjusting culture pH to an alkaline value (;7.5) (Fig. 6D and E). Conversely, BoNT/A stability was induced for strain Hall A-hyper argGin TPM cultures with 2% arginine by maintaining an acidic pH below 6.0. These data confirm that BoNT stability in TPM with 2% arginine is dependent on culture pH and is independent of the presence of arginine, suggesting an important role of metabolically induced culture pH in BoNT stability.
BoNT is proteolytically degraded in culture supernatants by a metalloprotease in a pH-dependent manner. To further explore the mechanism by which pH influences the stability of BoNT, in vitro experiments were conducted that examined BoNT stability in spent culture supernatants from C. botulinum strain Hall A-hyper grown in TPM for 48 h. The culture supernatants were pH adjusted to a range of values, sterile filtered, and incubated at 37°C for up to 4 days, the time period for which the C. botulinum production culture is usually grown before toxin harvest. BoNT complex added to fresh TPM or TPM with 2% arginine, both at pH 7.4, was stable over the 4-day time period with no apparent degradation (control, not shown). Comparison of nonreduced BoNT levels from the very acidic pH group (pH 5.6) group with that of the mildly acidic pH group (pH 6.4) revealed a progressive decrease in BoNT levels until day 4, accompanied by an increase in smaller bands indicative of breakdown products ( Fig. 7A and B). At about neutral pH (pH 7.2), an even stronger and faster decrease in BoNT levels was observed, concomitant with a greater increase in apparent breakdown products. Interestingly, at an alkaline pH of 8.0, the decrease in BoNT was less than at pH 7.2 and mirrored that at pH 6.4. While there are two predominant bands in the breakdown products that are similar in size to the 100-kDa HC and the 50-kDa LC of BoNTs, these bands were observed in nonreduced samples which would not allow for separation of HC and LC. In addition, detailed Western blot analyses with HC-and LCspecific antibodies indicated that the 100-kDa band is larger than the HC and contains the LC and a portion of the HC, whereas the apparent 50-kDa band is smaller than the LC and contains an HC fragment (data not shown). These data indicate proteolytic degradation of BoNT in a pH-dependent manner.
In order to investigate proteolytic involvement in toxin instability at pH 7.2, the same assay using spent 48-h culture supernatants of C. botulinum strain Hall A-hyper grown in TPM was performed at pH 7.2, and either a broad-spectrum protease inhibitor cocktail (PIC) or EDTA (metalloprotease inhibitor) was added before incubation. The PIC contained inhibitors of serine proteases, cysteine proteases, aspartic acid proteases, and aminopeptidases that are typically present in cell lysates. Quantitative Western blot analysis showed that the BoNT in nonreduced supernatant samples not treated with protease inhibitors decreased almost 2-fold by day 1 and 19-fold by day 4. Addition of PIC had no effect on this degradation, but addition of EDTA completely prevented protein degradation ( Fig. 7C and D). These data indicate that a metalloprotease is responsible for pH-dependent protein degradation of BoNT/A in C. botulinum strain Hall A-hyper culture supernatant.

DISCUSSION
The molecular mechanisms governing BoNT formation during growth of C. botulinum in various environments are not well understood, and the intricate steps involved have not been characterized. While several transcriptional control mechanisms have been identified as relatively minor regulators of BoNT production (14,29,30,(32)(33)(34), metabolic and environmental factors have been shown to strongly affect BoNT production (35)(36)(37)(38)(39). The data presented here for the first time suggest pH-dependent posttranslational proteolytic processing via a metalloprotease as a major mechanism of controlling BoNT production. The culture pH of C. botulinum cultures changes during growth in rich medium (34,48) and is affected by environmental and nutritional factors that can alter the metabolism of the bacteria (39,48). An acidic pH has been associated with greater toxin yield (48,49), but the underlying mechanisms are unknown.
In a previous study, we had observed ;1,000-fold suppression in active BoNT formation by C. botulinum if 2% arginine was added to complex medium during growth. In an attempt to understand the underlying mechanisms, the arginine catabolism pathway of C. botulinum strain Hall A-hyper was delineated using bioinformatics (Fig. 1). Second, select enzymes in the pathway were inactivated by ClosTron mutagenesis. Results from investigations into BoNT formation by the Hall A-hyper strain and these mutants in complex medium supplemented with either arginine or one of its metabolites indicated an indirect effect of arginine on BoNT formation ( Fig. 2 and 3). Further, analysis of mRNA by Northern blotting indicated that the addition of 2% arginine into culture medium did not significantly alter transcription of the bont gene (Fig. 4).
During these studies, it became apparent that a drop in culture pH to acidic levels during growth correlated with greater toxin production, while an increase to alkaline pH levels correlated with decreased BoNT/A levels (see Table S1 and Fig. S1 in the supplemental material). Further analysis confirmed that C. botulinum strain Hall A-hyper cultures grown in TPM underwent a gradual decrease in pH during log phase, reaching a pH of ;6 to 6.3, which stabilized at stationary phase and then increased very slightly with continued incubation (Fig. 5). Cultures grown in TPM supplemented with 2% arginine, on the other hand, underwent an increase in pH during log phase to alkaline levels of 7.5 to 7.8, which stabilized during stationary phase (Fig. 5). In addition, analysis of BoNT/A production during growth indicated an increase of intracellular BoNT/A until 24 h (late stationary phase) followed by a slight decrease of cell pellet-associated BoNT/A, which correlated with a gradual increase of BoNT/A in culture supernatants. This pattern is in agreement with toxin production during log phase and gradual release or secretion from the cells during the late log and lysis phases. In the cultures grown in TPM supplemented with arginine, on the other hand, almost no BoNT/A was detected in supernatant at any point, whereas cell pellet-associated BoNT/A increased throughout the 4-day study period. However, even cell pellet-associated BoNT/A levels were at least 5-fold reduced compared to those of cultures grown in TPM. Taken together, these data suggest posttranscriptional control of BoNT/A production in TPM with 2% arginine and indicate a potential correlation with culture pH.
To further examine the role of culture pH in BoNT/A production by C. botulinum strain Hall A-hyper, the culture pH during growth in TPM or TPM with 2% arginine was manually adjusted to alkaline levels instead of the normally acidic TPM culture, and to acidic levels for the normally alkaline TPM-2%-arginine cultures. This reversal in culture pH led to a reversal in BoNT/A production as well, with the now acidified TPM-2%-arginine culture producing normal levels of BoNT/A and the alkaline TPM culture resulting in barely detectable and largely degraded BoNT/A (Fig. 6). Similarly, pH reversal in in cultures of C. botulinum strain Hall A-hyper argGor arcBgrown in TPM with 2% arginine confirmed correlation of decreased BoNT levels with alkaline pH. These data excluded a direct effect of arginine in controlling BoNT/A production and instead implicated culture pH as a crucial factor. Analyses of the effect of culture pH on the stability of BoNT/A in 48-h spent culture medium of C. botulinum strain Hall A-hyper grown in TPM further showed gradual degradation of BoNT/A at alkaline pH yet high stability at 37°C over 4 days at acidic pH (Fig. 7). Fresh TPM spiked with isolated BoNT/A complex resulted in no degradation at alkaline pH (data not shown), indicating the likely presence of a C. botulinum-produced protease in spent culture medium. In fact, addition of EDTA prevented BoNT/A degradation in spent culture medium (Fig. 7), further implicating a metalloprotease in toxin degradation.
These data suggest that BoNT/A production in proteolytic C. botulinum strain Hall A-hyper is strongly regulated on a posttranslational level by a pH-dependent metalloprotease that is produced during growth. The lower cell-associated BoNT/A levels at alkaline pH indicate that this proteolysis occurs inside and outside the bacterial cell. It has previously been described that Gram-positive bacteria adjust their intracellular pH based on extracellular pH changes (50). The metalloprotease involved in BoNT/A degradation is currently uncharacterized, and there are several possible mechanisms by which pH may play a role. First, most biologic enzymes are pH dependent and often show peak activity at neutral pH. In addition, pH-dependent conformational changes of the NTNH-BoNT/A1 complex causing instability at neutral or alkaline pH have been demonstrated (51), which could result in structural changes making the BoNT/A1 more susceptible to proteolytic attack by the metalloprotease.
Interestingly, the LC is a catalytically active Zn 2 -dependent metalloprotease, and the possibility that the metalloprotease(s) that is essential for production of biologically active BoNT in proteolytic C. botulinum is also involved in BoNT degradation cannot be ruled out (52,53). BoNTs are produced as single 150-kDa polypeptides and are proteolytically converted to the more active dichain form of the 100-kDa HC and 50-kDa LC linked by a disulfide bond (8). The data show that BoNT/A in culture supernatants is mostly present in the "nicked" dichain form, while BoNT/A extracted from the cell pellet was only partially converted to the 100-and 50-kDa HC and LC, respectively (Fig. 5B). This indicates that both proteolytic processing and degradation of BoNT/A predominantly occur in the extracellular environment and raises the possibility that the same protease(s) may be responsible for both processes under different environmental conditions. In fact, a previous study has shown that BoNT/A and BoNT/E are degraded at 44°C by a protease with similar properties as the protease that activates BoNTs at lower temperatures, suggesting the possibility that it may be one protease with various functions at different temperatures (54). The data presented here are consistent with this study and suggest pH as an additional environmental or metabolically determined factor in controlling protease activity.
In conclusion, this study shows that the strong suppression of BoNT/A production in proteolytic C. botulinum by arginine is an indirect consequence of the catabolism of arginine by raising the pH to alkaline levels, most likely due to excess free ammonia production. Under normal (laboratory) growth conditions, proteolytic C. botulinum cultures acidify during growth to a pH of 5.7 to 6.2. The main effect of the alkaline pH appears to be the posttranslational proteolytic degradation of BoNT/A complex by a metalloprotease, although an additional effect on pretranslational regulation cannot be entirely excluded based on the results of this study. Both the protease, or proteases, involved in degradation and the molecular mechanisms underlying the pH dependence remain unknown, and future studies are required to resolve these outstanding questions. The findings from this study enhance our understanding of the mechanism(s) involved in BoNT formation by C. botulinum under various environmental or metabolic conditions, which is relevant for environmental and food safety, clinical efficacy, and production methods of pharmaceutical BoNTs. For ClosTron mutagenesis experiments, reinforced clostridial medium (RCM) with 1% agar (both from Difco) and 2 mg/ml resazurin (as an indicator of anaerobic conditions) was used. Escherichia coli strain CA434 (56) was used as the donor strain for conjugation of the ClosTron plasmids into C. botulinum Hall A-hyper. E. coli cultures were grown aerobically at 37°C in a shaker at 225 rpm, in either Luria broth (LB) or SOC broth (57,58). Antibiotics were used at the following concentrations: 20 mg/ml thiamphenicol, 12.5 mg/ml chloramphenicol, 25 mg/ml erythromycin, and 250 mg/ml cycloserine.

MATERIALS AND METHODS
Mouse bioassay. An intraperitoneal (i.p.) time to death method was used to estimate toxicity of the cultures as previously described (39). In short, culture supernatant samples were treated with 5 mg/ml of trypsin (Worthington Biochemical Corporation) and incubated at 37°C for 1 h to proteolytically activate the BoNT. The treated culture supernatants were sterile filtered and mixed at a 1:1 ratio with GelPhos buffer (0.2% gelatin and 30 mM sodium phosphate, pH 6.3), and 0.5 ml per mouse was injected i.p. into groups of four female ICR mice (Harlan Laboratories). Survival time was recorded, and 50% lethal dose (LD 50 )/ml was estimated from the curve fit formula: y = 8334208.65 Â 10 20.025x (39). Statistical analysis was performed with a paired Student t test (equal variance, two-tailed), and values were considered different with a P value of ,0.05.
ClosTron mutagenesis. All ClosTron plasmids were designed using the Petruka algorithm on clostron.com (46,59) and ordered from DNA2.0 (Atum.bio, Newark, CA). ClosTron mutagenesis was performed as previously described (60). Briefly, ClosTron plasmid DNA was transformed into E. coli strain CA434 and transferred by conjugation into C. botulinum strain Hall A-hyper, as previously described (46,47,60). The mating mixture of CA434 and Hall A-hyper was then plated onto RCM agar supplemented with cycloserine (to select for Hall A-hyper) and thiamphenicol (to select for acquisition of the ClosTron plasmid). Thiamphenicol-resistant colonies were restreaked on cycloserine-and erythromycin-supplemented media to select for Hall A-hyper mutants with the integrated intron and activated erythromycin resistance gene. Selected colonies were then serially restreaked onto erythromycin-supplemented medium plates to ensure purity and to cure the mutants of the ClosTron plasmid. Curing of the plasmid was verified by screening for loss of thiamphenicol resistance. Clones were analyzed by PCR using primers flanking the intron integration site ( Table 1), such that mutants would produce a PCR product 1.8 kbp larger than wild type, due to integration of the intron. PCRs were performed using Phusion Master h to a PhosphorImager screen, and the capture and quantification of the hybridization signals were performed using a Typhoon FLA 9000 imaging scanner and ImageQuant software (GE Healthcare Bio-Sciences Corp., Piscataway, NJ). Relative ntnh-bont bicistronic and monocistronic bont mRNA levels were measured together. The mRNA expression levels were normalized to the levels of the TPM cultures grown for 12 h, which were defined as 100%. Average and standard deviations from triplicate samples were determined. Hybridization probes. PCR primers (IDT Technologies) were used to amplify three different regions of the BoNT/A1 gene (Table 1). C. botulinum strain Hall A-hyper chromosomal DNA was used as a template. PCRs were performed using Phusion high-fidelity PCR master mix with high-fidelity buffer (New England Biolabs Inc., Ipswich, MA) in a ProFlex PCR cycler (Applied Biosystems by Life Technologies, Bedford, MA). The PCR fragments were purified from a 1% agarose gel using the Monarch DNA gel extraction kit (New England Biolabs Inc., Ipswich, MA) following manufacturer's instructions. Fragments were inserted into pJET1.2 vector using the CloneJET PCR cloning kit (ThermoFisher Scientific, Waltham, MA), and DNA sequences were confirmed by DNA sequencing (University of Wisconsin-Madison Biotechnology Center). Clones of BoNT/A1 gene fragments inserted downstream of the T7 promoter in the reverse orientation of the BoNT gene were used for generation of antisense RNA probes. The plasmids were linearized with restriction enzyme XbaI, the enzyme was inactivated for 20 min at 65°C, and linearized plasmids were then treated with DNA polymerase I (Klenow fragment) (New England Biolabs Inc., Ipswich, MA) in the presence of deoxynucleoside triphosphates (dNTPs) to generate blunt ends of the DNA molecules. The linearized DNA templates were purified (Monarch PCR and DNA cleanup kit; New England Biolabs Inc., Ipswich, MA) following manufacturer's instructions. 32 P-labeled antisense RNA hybridization probes were generated from the three BoNT/A1 gene fragments using the HiScribe T7 high-yield RNA synthesis kit (New England Biolabs Inc., Ipswich, MA) and [ 32 P]UTP (3,000 Ci/mmol: PerkinElmer, Inc.) following manufacturer's instructions. Unincorporated nucleotides were removed from the synthesized probes by spin column chromatography (NucAway columns; Invitrogen) according to manufacturer's instructions. All three probes were combined and used at 1.5 Â 10 6 cpm/ml in hybridization with RNA membranes. Oligonucleotide primers used for amplification of BoNT gene fragments are shown in Table 1.
Statistics. All values were measured in at least triplicate values, and values are expressed as average 6 standard deviation (SD). Differences between samples were assessed using paired Student's t test where statistical significance is assumed for P , 0.05.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.