Search tips
Search criteria 


Logo of iaiPermissionsJournals.ASM.orgJournalIAI ArticleJournal InfoAuthorsReviewers
Infect Immun. 2006 October; 74(10): 5617–5624.
PMCID: PMC1594926

Comparison of Extracellular and Intracellular Potency of Botulinum Neurotoxins


Levels of botulinum neurotoxin (BoNT) proteolytic activity were compared using a cell-free assay and living neurons to measure extracellular and intracellular enzymatic activity. Within the cell-free reaction model, BoNT serotypes A and E (BoNT/A and BoNT/E, respectively) were reversibly inhibited by chelating Zn2+ with N,N,N′,N′-tetrakis (2-pyridylmethyl) ethylenediamine (TPEN). BoNT/E required relatively long incubation with TPEN to achieve total inhibition, whereas BoNT/A was inhibited immediately upon mixing. When naïve Zn2+-containing BoNTs were applied to cultured neurons, the cellular action of each BoNT was rapidly inhibited by subsequent addition of TPEN, which is membrane permeable. Excess Zn2+ added to the culture medium several hours after poisoning fully restored intracellular toxin activity. Unlike TPEN, EDTA irreversibly inhibited both BoNT/A and -E within the cell-free in vitro reaction. Excess Zn2+ did not reactivate the EDTA-treated toxins. However, application of EDTA-treated BoNT/A or -E to cultured neurons demonstrated normal toxin action in terms of both blocking neurotransmission and SNAP-25 proteolysis. Different concentrations of EDTA produced toxin preparations with incrementally reduced in vitro proteolytic activities, which, when applied to living neurons showed undiminished cellular potency. This suggests that EDTA renders the BoNT proteolytic domain conformationally inactive when tested with the cell-free reaction, but this change is corrected during entry into neurons. The effect of EDTA is unrelated to Zn2+ because TPEN could be applied to living cells before or after poisoning to produce rapid and reversible inhibition of both BoNTs. Therefore, bound Zn2+ is not required for toxin entry into neurons, and removal of Zn2+ from cytosolic BoNTs does not irreversibly alter toxin structure or function. We conclude that EDTA directly alters both BoNTs in a manner that is independent of Zn2+.

Botulinum neurotoxins (BoNTs) cause muscle paralysis by cleaving key proteins in nerve terminals. Combined incidents of botulism caused by BoNT serotypes A and E (BoNT/A and BoNT/E, respectively) constitute 50 to 60% of all botulism outbreaks in the United States (41, 49). BoNT poisoning frequently entails critical care medical treatment throughout the course of the disease because no cure exists to reverse the consequent paralysis. The duration and severity of symptoms may last for many months. After a botulism case is confirmed, strong emphasis is placed on identifying BoNTs in foods or environmental samples in order to document the source of the poisoning, prevent further spread of the disease, and help understand the molecular basis by which these toxins are transmitted.

The light-chain (LC) domain of each BoNT is a Zn-dependent metalloprotease (33). Cellular toxicity from the LC is imparted by two other regions of the BoNT protein that mediate toxin binding to nerve terminals and LC translocation into the neuronal cytoplasm (28). Within the nerve terminal, BoNT/A and -E LCs cut the synaptic protein SNAP-25, resulting in a blockade of nerve-muscle communication (18). Each LC is relatively large, about 50 kDa, and must enter the nerve terminal by unfolding within an acidic endosomal compartment and threading through a narrow toxin-induced channel (17, 23). Within the neuron, the LC refolds, where it may be phosphorylated by neuronal kinases (5, 11, 13) prior to enzymatically cleaving SNAP-25.

Comparison of molecular and biochemical characteristics between BoNT/A and -E is of interest because both cleave the same neuronal protein, SNAP-25, yet paralysis caused by each persists for different durations. Paralysis from BoNT/A can endure for months, and paralysis from BoNT/E lasts for several days (1, 22, 40). Recently, we demonstrated that BoNT/A translocates into neurons more slowly than BoNT/E, although cell binding kinetics were similar for both BoNTs (20). Previously, we showed that neurotransmission is inhibited differently by BoNT/A or -E because of the distinct SNAP-25 sites cut by each toxin (21). In the present study, we have compared extra- and intracellular proteolytic activities of BoNT/A and -E. Extracellular proteolysis of SNAP-25 never happens in vivo; however, this type of reaction is commonly used to study the toxin enzymatic properties and substrate-toxin binding interactions and to screen small drug inhibitors to the LC domain (26, 35-39, 42, 48). In the present study, extracellular BoNT activity was measured by a method established by Schiavo and Montecucco (32) using recombinant SNAP-25. In addition to this assay, cultured neurons were used to assess toxin entry into living cells to compare the biological action of each toxin with its in vitro activity (3, 24, 44, 50).

Although the Zn protease character of BoNTs is very well established, our results indicate that under cell-free conditions BoNT/A and BoNT/E protease activity can be masked to various degrees using EDTA, and this masking is independent of Zn2+. Paradoxically, the loss of this extracellular activity does not reflect cellular potency: both toxins fully regain intracellular activity when applied to living cells. In contrast to the EDTA effect, the Zn2+-specific chelator N,N,N′,N′-tetrakis (2-pyridylmethyl) ethylenediamine (TPEN) reversibly inhibits the BoNTs in both assays. The results presented here rectify an apparent contradiction between previous reports in which separate studies examined either EDTA or TPEN, leading to opposite results and different interpretations (15, 27, 33, 46). The current data allow the coexistence of the competing observations and suggest that the discrepancies were due to different model systems and different reaction conditions employed in each study. A hypothesis is discussed to account for the combined observations. The findings provide valuable information for BoNT detection assays and for small-molecule inhibitor screening assays used to develop BoNT antagonists.



Polyclonal anti-mouse and anti-rabbit antibodies conjugated to alkaline phosphatase, mouse anti-HPC-1, rabbit anti-SNAP-25, phenylmethylsulfonyl fluoride (PMSF), TPEN, and ultrapure dimethyl sulfoxide (DMSO) were obtained from Sigma Chemical Co. (St. Louis, MO). Monoclonal mouse anti-SNAP-25 was purchased from Sternberger Monoclonal Antibodies (Baltimore, MD). TPEN was prepared as a 10 or 20 mM stock solution in DMSO and diluted into BoNT-containing solutions such that the final DMSO concentration was 0.1% or less. DMSO at these levels had no effect on BoNT/A or -E. All electrophoresis reagents were from Bio-Rad (Hercules, CA). Polyvinylidene difluoride (PVDF) microporous (0.45 μm) membrane was obtained from Millipore Corporation (Millerica, MA).

Cell cultures.

Timed pregnant C57BL/6J mice were obtained from the Frederick Cancer Research and Development Center, Frederick, MD. On gestation day 13, animals were rendered unconscious in a CO2 atmosphere. Embryos were surgically removed from the amniotic sac, and the spinal cords were isolated. Neuronal cultures were prepared as described previously (30, 50). Briefly, spinal cords were removed from approximately 20 fetal mice, dissociated with trypsin, and plated on Vitrogen-coated dishes (Collagen Corp., Palo Alto, CA) at a density of 0.5 × 106 to 1 × 106 cells/dish. Cultures were maintained for 3 weeks at 37°C in a humidified atmosphere of 90% air-10% CO2 before addition of toxins. Cultures were grown in Eagle's minimum essential medium (formula 82-0234AJ; Invitrogen, Carlsbad, CA) supplemented with 5% heat-inactivated horse serum and a mixture of complex factors (tissue culture medium). Protein extracts were prepared with ice-cold lysis buffer (0.6% Triton X-100, 0.6% Tween 20, 25 mM HEPES, pH 7.2, containing 1 mM PMSF). After 1 h of incubation on ice, the contents were collected and boiled for 4 min and then dialyzed in 20 mM HEPES, pH 7.2, for 20 h.

In vitro cleavage of SNAP-25 assay. The catalytic activity of BoNT/A or -E was analyzed using recombinant SNAP-25 (32). Full-length SNAP-25 cDNA was a gift from G. A. Oyler, and the protein was expressed and purified as described previously. Reactions of 50 mM NaCl and 20 mM HEPES, pH 7.2, were initiated after reducing BoNT/A (20 nM) or -E (4 nM) with dithiothreitol (5 mM) for 60 min at room temperature. In some experiments, TPEN (100 μM) was added to the reaction mixture or to the toxin before initiating the reactions. Samples were collected at recorded time intervals and mixed with two volumes of sodium dodecyl sulfate (SDS) reducing buffer (2% SDS, 5% β-mercaptoethanol, 55 mM Tris-Cl, pH 6.8, 10% glycerol, 0.05% bromophenol blue) and then heated at 95°C for 5 min prior to separation by SDS-polyacrylamide gel electrophoresis.

Gel electrophoresis and Western blotting.

Equal volumes of lysate were loaded onto 11% or 16.5% acrylamide gels prepared by the method of Laemmli (25). Proteins were separated using 0.1 M Tris-Tricine as a run buffer (pH 8.3) (31) and then transferred to PVDF membrane using 192 mM glycine, 25 mM Tris, pH 8.3, and 7% methanol (22).

The PVDF membrane was treated with Tris-buffered saline (TBS; 0.5 M NaCl, 25 mM Tris, pH 7.5) containing 4% nonfat dry milk for 1 h. Primary antibodies (anti-SNAP-25 and antisyntaxin) in blocking buffer were incubated with the membrane at room temperature using gentle agitation for 60 min followed by two sequential 10-min washes with TTBS (TBS containing 0.05% Triton X-100) at room temperature. A secondary antibody conjugated to alkaline phosphatase was similarly applied to the membrane, followed by two sequential washes with TTBS, one wash with TBS, and a 1-min rinse with deionized water before visualizing the bound alkaline phosphatase with alkaline phosphatase conjugate substrate kit (Bio-Rad).

Botulinum neurotoxin preparation.

Preparations of BoNT/A and -E toxin complex were from Wako Chemicals (Richmond, VA) with reported activities of 2.0 × 107 and 1.0 × 107 50% lethal doses per mg, respectively. Clostridium botulinum proteolytic strains release BoNT/A as a biologically active dichain toxin (8, 29). BoNT/E, in contrast, is generated within nonproteolytic C. botulinum strains as an inactive, single-chain protoxin (6). BoNT/E is routinely activated by limited proteolysis in vitro, as first described in 1956 (29). In this study, BoNT/E (1 mg/ml) was activated by incubation for 30 min at 37°C with 0.3 mg/ml trypsin (type XI, bovine pancreas) in 30 mM HEPES, pH 6.75 (10, 47). Soybean trypsin inhibitor type I-S (0.5 mg/ml) was added in the same buffer and incubated at room temperature for 5 min to inhibit further trypsinization. Toxin was aliquoted and stored at −20°C. Stock solutions of BoNT/A (1 mg/ml) were aliquoted and stored at −20°C without further treatment. Each experiment utilized a new aliquot of toxin to ensure uniform activity. Cell-free reactions were carried out using toxin concentrations of either 20 nM BoNT/A or 4 nM BoNT/E, assuming molecular masses of 500 kDa and 300 kDa for BoNT/A and -E, respectively, as indicated by the supplier.

Data analysis.

Scanned images of Western blots were produced and stored utilizing CorelDraw 11 (Corel Corp., Ottawa, Canada) and Photoshop 7.0 (Adobe Systems, San Jose, CA). In some cases, the bands of Western blots were quantified by densitometry (Kodak Image Station 440CF; Kodak, Rochester, NY). The data are presented as the mean ± standard error. Student's t test was used to analyze the variance between control and treated group. The significance level was P < 0.05.


TPEN effect on BoNT/A and -E in cell-free, in vitro reactions.

Incubation of soluble SNAP-25 with BoNT/A led to proteolysis of SNAP-25 over 60 min (Fig. (Fig.1,1, lanes 1 to 4). Zn2+ was removed from BoNT/A by incubating the toxin with the Zn2+-specific chelator TPEN (0.1 mM) for 2 min. Following this, Zn-depleted toxin was diluted into reaction mixtures containing TPEN alone or TPEN and Zn2+ (0.2 mM). The results show that TPEN caused a complete blockade of BoNT/A extracellular activity, and this activity was restored readily with excess Zn2+. In contrast, incubation of BoNT/E with 0.1 mM TPEN for several minutes did not reduce BoNT/E activity. To achieve blockade of BoNT/E, the toxin was incubated with TPEN for up to 60 min prior to initiating the reaction with SNAP-25 (Fig. (Fig.2).2). TPEN concentration was constant throughout the experiment. BoNT/E catalytic activity was decreased incrementally over the course of 60 min (Fig. (Fig.2A,2A, lanes 13 to 16). Dilution of the apo-BoNT/E into a Zn2+-containing reaction mixture immediately restored BoNT/E activity (Fig. (Fig.2A,2A, lanes 17 to 20). These findings indicate that Zn2+ is removed from BoNT/A more rapidly than from BoNT/E, but addition of excess Zn2+ readily restores extracellular proteolytic activity to both BoNTs.

FIG. 1.
TPEN effect on BoNT/A. Intact SNAP-25 is in lane 1; BoNT/A-cleaved SNAP-25 appears as a lower band in panels 2 to 4 and 8 to 10. The proteolysis of SNAP-25 by BoNT/A (lanes 2 to 4) was inhibited totally by prior incubation with 0.1 mM TPEN (lanes 5 to ...
FIG. 2.
TPEN effect on BoNT/E. (A) Reaction conditions were similar to Fig. Fig.1,1, except that BoNT/E was present at 4 nM. The SNAP-25 Western blot shows the incremental inhibition of BoNT/E incubated with 0.1 mM TPEN for various times. BoNT incubation ...

TPEN effect on BoNT/A and -E in cultured neurons.

We next did two experiments to examine if intoxication by BoNT/A and -E into living neurons could be altered by Zn2+ chelation. Because toxin entry into neurons is dependent upon synaptic activity, we first tested if TPEN would inhibit BoNT catalytic activity after the Zn-containing toxins had entered cells. This was done by exposing cultures to 500 pM native BoNT for 5 min at 37°C and then washing the cultures to remove unbound, extracellular BoNT (Fig. (Fig.3A).3A). Fresh medium containing TPEN (10 μM) was applied to some cultures after cells were washed to remove free toxin. All cultures were kept at 37°C for 2.5 h to allow the relatively slow translocation process to proceed to completion. After this time, excess Zn2+ (15 μM) was added to some cultures. After a total of 3.5 h, cultures were dissolved in SDS sample buffer and SNAP-25 proteolysis was examined by Western blot analysis. TPEN applied immediately after toxin exposure was able to inhibit cytosolic BoNT/A and -E. Intracellular toxin activity was restored with the addition of extracellular Zn2+ (Fig. (Fig.3A).3A). Therefore, we conclude that Zn2+ can be rapidly and reversibly removed from intracellular BoNTs.

FIG. 3.
TPEN reversibly inhibits BoNT in living neurons. (A) Neuronal cell cultures were exposed to BoNT/A (lanes 2 to 4) or BoNT/E (lanes 5 to 7) for 5 min, and then cultures were washed with buffer containing TPEN. Untreated cells were maintained in fresh medium ...

We then tested if Zn2+ chelation before addition to cell cultures would influence BoNT entry into neurons. This was important since previous studies had concluded that Zn2+ chelation destroys BoNT/A biological function (15). Each BoNT (0.5 μM) was treated with TPEN for 90 min at room temperature prior to being diluted into culture medium (250 pM final toxin, 50 nM 10 μM TPEN, 1.5 μM Zn2+). Some toxin samples were not treated with TPEN but underwent identical preparation prior to being applied to cultures. Neither BoNT/A nor BoNT/E was inactivated by prior treatment with TPEN as determined by SNAP-25 Western blot results, and they exhibited biological function that was approximately equivalent to that of native BoNTs (Fig. (Fig.3B).3B). In a similar procedure, each toxin was treated with TPEN and then diluted into culture medium supplemented with various concentrations of Zn2+ and TPEN. If the amount of Zn2+ was less than that of TPEN, neither toxin exhibited activity within neurons. Yet, when the amount of Zn2+ exceeded that of TPEN, normal toxin function in terms of reduced Ca2+-dependent neurotransmission and proteolysis of SNAP-25 was observed within the neurons (Fig. (Fig.3C3C).

EDTA effect in cell-free assay and in cultured neurons.

Treatment of BoNT/A or -E with EDTA produced different results from TPEN treatment. EDTA required at least 60 min of incubation with each toxin to fully inhibit in vitro catalytic activity (Fig. (Fig.4A).4A). Although EDTA inhibited both BoNTs, addition of excess Zn2+ did not restore extracellular toxin catalytic activity when the cell-free assay was used. The extracellular proteolytic activities of both BoNT/A and -E were irreversibly inhibited by EDTA treatment. When EDTA-treated toxins were applied to living neurons, however, intracellular catalytic activities of BoNT/A and -E were readily detected by blockade of neurotransmission and by cleavage of SNAP-25 (Fig. 4B and C). To further test this, BoNT/A and -E (2 μM) were incubated with different concentrations of EDTA (0.05, 0.2, and 0.5 mM) for 60 min and then diluted into reaction mixtures so that the final EDTA concentration was less than 5 μM. This treatment caused incremental reduction of in vitro proteolytic activity (Fig. (Fig.5,5, top frames). In contrast, application of the EDTA-treated toxin preparations to living neurons produced uniform intracellular toxin activity in all conditions (Fig. (Fig.5,5, bottom frames).

FIG. 4.
EDTA irreversibly inactivates BoNT/A and -E catalytic activity in extracellular proteolytic assays, but activity is restored within living neurons. (A) Each BoNT was incubated with EDTA for 90 min at room temperature prior to dilution into reaction mixtures ...
FIG. 5.
Comparison of extracellular and intracellular BoNT activities after treatment with various concentrations of EDTA. BoNT/A and -E were incubated with the indicated EDTA concentrations for 90 min prior to being diluted into reaction mixtures. For the extracellular ...

Time-dependent return of cellular BoNT protease activity.

In a final set of experiments, TPEN or EDTA was used to test if Zn2+ depletion altered or slowed the membrane translocation process of the LC into the cytosol by monitoring the rate at which cellular BoNT activity resumed upon addition of Zn2+. Cultures were exposed to either of the Zn-containing BoNTs, BoNT/A or BoNT/E. Proteolysis of SNAP-25 and blockade of neurotransmission were assessed at various times up to 3.5 h after applying toxins. SNAP-25 proteolysis and a reduction of [3H]neurotrasmitter release proceeded in a time-dependent manner when native BoNTs were tested (Fig. (Fig.6).6). EDTA treatment of either toxin had no effect on the onset or extent of toxin cellular activity. When TPEN was used, BoNT/A activity was inhibited throughout the experiment, except where Zn2+ was added 2.5 h later. In this case, the reactivated toxin cleaved SNAP-25 rapidly, and in 0.5 h the extent of SNAP-25 proteolysis was equivalent to that of native BoNT/A (no TPEN), which had reacted within the cytosol for 3 h to produce a gradual accumulation of cleaved SNAP-25 and the consequential blockade of neurotransmission (Fig. (Fig.6).6). Similar results were obtained for BoNT/E (data not shown). This implies that Zn2+-depleted BoNT/A and Zn2+-depleted BoNT/E LCs can enter neurons unimpeded and each LC resides in the neuronal cytosol as an inactive enzyme until excess extracellular Zn2+ is added. The rapid accumulation of cleaved SNAP-25 after addition of excess Zn2+ to the medium indicates the inactive LC molecules reside within the local proximity of SNAP-25.

FIG. 6.
Time course of restoration of intracellular BoNT/A activity following inhibition with TPEN. (A) Neurotransmitter release was measured at various times after applying either BoNT alone or BoNT and TPEN to cells. Incremental reduction in neurotransmission ...


BoNTs are 150-kDa proteins expressed by Clostridium botulinum, Clostridium baratii, and Clostridium butyricum, and are Zn2+ metalloproteases (16, 33, 51). The toxins have confined substrate specificity. BoNT/A and -E are known to only cleave the 206-amino-acid protein SNAP-25 at Q197-R198 and R180-I181, respectively (4, 34). These serotypes were studied because they represent the extremes of BoNT duration in vivo. We report that removing Zn2+ from the catalytic LC domain reversibly inhibits each. Bound Zn2+ is unnecessary for toxin binding and translocation into living cells. Intracellular catalytic activity of each apo-BoNT is readily reconstituted upon entering neurons if Zn2+ is added subsequently. In spite of the apparent unimportance of Zn2+ to cellular entry, cell-free proteolysis reactions yielded fundamentally conflicting results.

In the case of TPEN, chelation of toxin-bound Zn2+ inhibited BoNT/A and -E proteolytic activity in vitro at different rates. BoNT/A readily parted with its active-site Zn2+ atom. When excess Zn2+ was added, the toxin quickly reconstituted its catalytic activity, consistent with the BoNT/A LC having a flexible protein structure (7). In contrast, the BoNT/E LC required longer incubation with TPEN—at least 60 min to achieve total inhibition. Therefore, extracellular BoNT/E seemed to have a less flexible structure than BoNT/A (5). Both BoNTs bind Zn2+ with similar dissociation values (33), and neither BoNT parts with Zn2+ during exhaustive dialysis in Zn2+-depleted buffer (15, 33). Resin-bound chelators do not inhibit BoNT/A, leading to the interpretation that TPEN physically extracts Zn2+ from BoNT/A (46). Extending this interpretation, TPEN does not readily extract Zn2+ from the BoNT/E LC.

EDTA has previously been shown to physically denature the tertiary structure of BoNT/A, which was attributed to removal of Zn2+ (15, 27). When 65Zn2+ was used, the binding was shown to be reversible but the conformational changes to the LC were not corrected upon reacquisition of 65Zn2+. In those studies, both the 150-kDa BoNT/A molecule and the purified 50-kDa A LC were treated with EDTA. Partial (30%) enzymatic activity of the purified LC could be retrieved with excess Zn2+. However, when EDTA-treated holo-BoNT/A was applied to detergent-permeabilized cells, only naïve toxin was capable of cleaving SNAP-25; EDTA-treated toxin was deemed irreversibly devoid of in vivo activity (15). Toxin entry into permeabilized cells, however, bypassed the membrane translocation step, which is required for BoNT entry into cells. In essence, permeabilized cells represent a type of cell-free assay where BoNT activity is a measure of toxin diffusing to the substrate and cleaving it. The results using permeabilized cells are identical to our cell-free results: EDTA irretrievably inactivated the BoNTs. Our experiments using cultured neurons, which retain all the steps of the natural entry pathway showed that EDTA-treated BoNT/A and -E generated normal cellular biological function, identical to those found in many other studies that had used intact nerve-muscle preparations or living animals (9, 33, 43, 45, 46).

Previous work has shown that chelation of Zn2+ does not compromise overall BoNT/A potency, but under some circumstances can delay the onset of paralysis (19, 43, 46). From these studies, it is clear that the effect of EDTA or TPEN is temporary. Yet, these in vivo and in situ experiments provide little insight into the BoNT LC structure. We confirm that EDTA reduces extracellular enzymatic activity for BoNT/A, and we extend this finding to BoNT/E. Although we do not directly assess the toxin structure, the results are consistent with EDTA denaturing the LCs of both BoNTs as had been described in previous studies (15, 27). The loss of extracellular enzymatic activity can be total or partial, depending on the EDTA concentration, and extracellular catalytic activity towards SNAP-25 in a cell-free reaction cannot be readily restored with excess Zn2+.

Given the thorough investigation into the structural effects of EDTA on BoNT/A (15, 27), we suggest that the same structural alterations were generated in the present work whenever EDTA was used to remove Zn2+ from BoNT/A or -E. The functional observations reported here for the in vitro assay are in agreement with in vitro data published along with the structural analysis. However, the structural changes imparted to the LC of BoNT/A by EDTA (15), and now presumed to similarly affect BoNT/E (current work), do not alter toxin potency towards living neurons and may be unrelated to Zn2+. Toxin binding and uptake into neurons proceed unimpeded regardless of whether Zn2+ is bound to the LC domain. Upon translocating into the cell cytosol, the LC of each BoNT produces normal enzymatic activity. Unrecognized by previous studies, this finding suggests that if the LCs are indeed physically denatured by EDTA, the changes are reversed most likely during the translocation step where a favorable situation exists for the proper refolding of LC as it enters the cytosol.

To directly test the effect of Zn2+ on BoNT/E in cultured neurons, we performed a two-part experiment. The first part tested if TPEN would inhibit intracellular toxin activity. This was done by applying normal Zn2+-containing BoNT/A or -E to cultures for several minutes followed by washout of free toxin (20). Immediate addition of TPEN, which is membrane permeable, rapidly inhibited each BoNT. Addition of free Zn2+ restored BoNT activity. Therefore, intracellular BoNTs were reversibly inhibited by TPEN. Intracellular BoNT/E was instantly inhibited by TPEN, which is in contrast to the 60-min inactivation period required in the cell-free assay; therefore, the BonT/E LC must become more flexible upon entering nerve terminals (5), suggesting that the LC domain unfolds during translocation. Removal of Zn2+ during this process appears to be inconsequential to the toxin translocation mechanism.

We next tested if Zn2+ altered toxin entry into neurons. Both BoNTs were incubated with TPEN for 90 min before addition to cultures. When Zn2+ was added to the cultures several hours later, cytosolic toxicity from both BoNTs was fully and immediately restored. The combined results indicate that the TPEN effect on the LC does not affect toxin entry into neurons. Both apo-BoNTs retain the fundamental ability to enter neurons: binding, internalization, and translocation are independent of Zn2+. These results support earlier observations and conclusions for BoNT/A in muscle preparations (45, 46) and show BoNT/E is similar to BoNT/A in this respect. This is particularly noteworthy since cytosolic BoNT/E is naturally much less stable than BoNT/A (14, 22). Temporary removal of Zn2+ from cytoplasmic BoNT/E did not cause a general loss of intracellular BoNT/E activity, and upon addition of excess Zn2+, the toxin regained the same level of activity as the native toxin (Fig. (Fig.33 and and6).6). Therefore, removing Zn2+ did not denature or destabilize BoNT/E in the neurons.

The combined data imply that translocation is critical for restoring intracellular proteolytic activity of BoNT/A and -E after treatment with EDTA. Extracellular BoNT/A and -E are affected similarly by EDTA, and both toxins regain activity upon entering neurons where physiological Zn2+ supports robust BoNT catalytic activity. We show that apo-BoNTs translocate into neurons, reside near SNAP-25, and remain catalytically inactive until free Zn2+ becomes available.

It is important to note that many previous zinc studies presented contradictions that have persisted without resolution. In 1989, it was found that a variety of phenanthroline and hydroxyquinoline-based chelators irreversibly inhibited BoNT and tetanus neurotoxin lethality in vivo but EDTA had no effect on the biological function (2). In 1992, three studies showed that the toxins contained a homologous zinc-binding domain within the toxin LC and that Zn2+ was reversibly removed by treatment with EDTA (16, 33, 51). The physiological effect of Zn2+ on toxin activity was investigated further after the membrane-permeable Zn-chelator TPEN was found to delay muscle paralysis in the isolated mouse hemidiaphragm tissue (9, 43, 45, 46). Without exception, these studies have revealed that Zn2+ contributes directly to the paralytic step of BoNT intoxication but Zn2+ does not influence toxin binding to nerve terminals or the LC translocation step. In light of these results, however, physical biochemical analysis indicated that Zn2+ provided structural stability to the LC if EDTA was used to chelate the Zn2+ atom (15, 27). Although the structural analysis was rigorous, the results led to the erroneous conclusion that Zn2+ itself rather than the EDTA interaction with the toxin destroyed BoNT in vivo potency. The experimental approach that led to this conclusion utilized a 5-min BoNT/A reaction to measure initial rate proteolysis of SNAP-25. Subsequently, Simpson et al. (46) reexamined the effect of Zn2+ and concluded that BoNT potency in vitro was reversibly inhibited by EDTA by using an extended BoNT reaction that proceeded for 4 h using substantially large amounts of BoNT/A.

From our findings, we suggest that BoNTs can exhibit various extracellular activities while retaining undiminished in vivo toxicity. This characteristic should be considered for drug development, since current screening methods employ in vitro, cell-free reaction assays (37-39). Furthermore, although the reduction in catalytic activity is observed following treatment with EDTA, similar effects were reported for several organic metal chelators (2), and it would not be surprising if other chemicals or environmental factors can exert similar effects on BoNT LC under cell-free conditions. Because the TPEN effect was reversible upon addition of Zn2+, the EDTA effect must be due to some action other than removing Zn2+ from the LC active site. One possibility is the hypothesized role of toxin-bound Ca2+ which has been identified in BoNT/B (12). However, unlike BoNT/A, the structure of BoNT/B is not altered by EDTA and treatment with a Ca2+-specific divalent chelator did not affect the in vitro catalytic activity of BoNT/A or -B (12). Another possibility may entail the greater water solubility of EDTA compared to TPEN. Both chelators have approximately the same binding affinity for Zn2+, which implies that the substantial hydrophilic character of EDTA may cause nonspecific structural changes to the LC as EDTA diffuses to the Zn-binding site. The relatively long time frame required for EDTA Zn2+ chelation suggests the toxin LC must gradually unfold to allow EDTA access to the Zn2+ atom. Therefore, the EDTA effect on BoNT/A and -E remains unclear but is clearly unrelated to a simple interaction with toxin-bound Zn2+.

In terms of diagnostic assays, it is possible that within the variety of conditions encountered by BoNTs in nature, such as type of food, pH, ionic strength, storage temperatures, etc., BoNTs may have altered LC structures and reduced enzymatic properties in the extracellular milieu when tested with cell-free catalytic assays, while in vivo lethality remains undiminished. Based on these findings, the amount of LC activity determined by in vitro measurement is most likely underrepresentative of the amount and biological potency of BoNTs.


We thank E. A. Neale (NICHD, NIH, Bethesda, MD) for providing the neuronal cultures used in this study. We greatly appreciate Robert E. Sheridan (1950 to 2004) (USAMRICD, Edgewood, MD) for sharing his valuable insight and encouragement regarding several aspects of this work.

This research was supported by NIAID research grant IAG Y2-A1-3744-02.


Editor: J. T. Barbieri


1. Adler, M., J. E. Keller, R. E. Sheridan, and S. S. Deshpande. 2001. Persistence of botulinum neurotoxin A demonstrated by sequential administration of serotypes A and E in rat EDL muscle. Toxicon 39:233-243. [PubMed]
2. Bhattacharyya, S. D., and H. Sugiyama. 1989. Inactivation of botulinum and tetanus toxins by chelators. Infect. Immun. 57:3053-3057. [PMC free article] [PubMed]
3. Bigalke, H., F. Dreyer, and G. Bergey. 1985. Botulinum A neurotoxin inhibits non-cholinergic synaptic transmission in mouse spinal cord neurons in culture. Brain Res. 360:318-324. [PubMed]
4. Binz, T., J. Blasi, S. Yamasaki, A. Baumeister, E. Link, T. C. Sudhof, R. Jahn, and H. Niemann. 1994. Proteolysis of SNAP-25 by types E and A botulinal neurotoxins. J. Biol. Chem. 269:1617-1620. [PubMed]
5. Blanes-Mira, C., C. Ibanez, G. Fernandez-Ballester, R. Planells-Cases, E. Perez-Paya, and A. Ferrer-Montiel. 2001. Thermal stabilization of the catalytic domain of botulinum neurotoxin E by phosphorylation of a single tyrosine residue. Biochemistry 40:2234-2242. [PubMed]
6. Bonventre, P. F., and L. L. Kempe. 1960. Physiology of toxin production by Clostridium botulinum types A and B. IV. Activation of the toxin. J. Bacteriol. 79:24-32. [PMC free article] [PubMed]
7. Cai, S., H. K. Sarkar, and B. R. Singh. 1999. Enhancement of the endopeptidase activity of botulinum neurotoxin by its associated proteins and dithiothreitol. Biochemistry 38:6903-6910. [PubMed]
8. Das Gupta, B. R., and H. Sugiyama. 1972. Role of a protease in natural activation of Clostridium botulinum neurotoxin. Infect. Immun. 6:587-590. [PMC free article] [PubMed]
9. Deshpande, S. S., R. E. Sheridan, and M. Adler. 1995. A study of zinc-dependent metalloendopeptidase inhibitors as pharmacological antagonists in botulinum neurotoxin poisoning. Toxicon 33:551-557. [PubMed]
10. Duff, J. T., G. G. Wright, and A. Yarinsky. 1956. Activation of Clostridium botulinum type E toxin by trypsin. J. Bacteriol. 72:455-460. [PMC free article] [PubMed]
11. Encinar, J. A., A. Fernandez, J. A. Ferragut, J. M. Gonzalez-Ros, B. R. DasGupta, M. Montal, and A. Ferrer-Montiel. 1998. Structural stabilization of botulinum neurotoxins by tyrosine phosphorylation. FEBS Lett. 429:78-82. [PubMed]
12. Eswaramoorthy, S., D. Kumaran, J. Keller, and S. Swaminathan. 2004. Role of metals in the biological activity of Clostridium botulinum neurotoxins. Biochemistry 43:2209-2216. [PubMed]
13. Ferrer-Montiel, A. V., J. M. Canaves, B. R. DasGupta, M. C. Wilson, and M. Montal. 1996. Tyrosine phosphorylation modulates the activity of clostridial neurotoxins. J. Biol. Chem. 271:18322-18325. [PubMed]
14. Foran, P. G., N. Mohammed, G. O. Lisk, S. Nagwaney, G. W. Lawrence, E. Johnson, L. Smith, K. R. Aoki, and J. O. Dolly. 2003. Evaluation of the therapeutic usefulness of botulinum neurotoxin B, C1, E, and F compared with the long lasting type A. Basis for distinct durations of inhibition of exocytosis in central neurons. J. Biol. Chem. 278:1363-1371. [PubMed]
15. Fu, F. N., R. B. Lomneth, S. Cai, and B. R. Singh. 1998. Role of zinc in the structure and toxic activity of botulinum neurotoxin. Biochemistry 37:5267-5278. [PubMed]
16. Fujii, N., K. Kimura, N. Yokosawa, K. Tsuzuki, and K. Oguma. 1992. A zinc-protease specific domain in botulinum and tetanus neurotoxins. Toxicon 30:1486-1488. [PubMed]
17. Hoch, D. H., M. Romero-Mira, B. E. Ehrlich, A. Finkelstein, B. R. DasGupta, and L. L. Simpson. 1985. Channels formed by botulinum, tetanus, and diphtheria toxins in planar lipid bilayers: relevance to translocation of proteins across membranes. Proc. Natl. Acad. Sci. USA 82:1692-1696. [PubMed]
18. Jahn, R., and H. Niemann. 1994. Molecular mechanisms of clostridial neurotoxins. Ann. N. Y. Acad. Sci. 733:245-255. [PubMed]
19. Kalandakanond, S., and J. A. Coffield. 2001. Cleavage of SNAP-25 by botulinum toxin type A requires receptor-mediated endocytosis, pH-dependent translocation, and zinc. J. Pharmacol. Exp. Ther. 296:980-986. [PubMed]
20. Keller, J. E., F. Cai, and E. A. Neale. 2004. Uptake of botulinum neurotoxin into cultured neurons. Biochemistry 43:526-532. [PubMed]
21. Keller, J. E., and E. A. Neale. 2001. The role of the synaptic protein snap-25 in the potency of botulinum neurotoxin type A. J. Biol. Chem. 276:13476-13482. [PubMed]
22. Keller, J. E., E. A. Neale, G. Oyler, and M. Adler. 1999. Persistence of botulinum neurotoxin action in cultured spinal cord cells. FEBS Lett. 456:137-142. [PubMed]
23. Koriazova, L. K., and M. Montal. 2003. Translocation of botulinum neurotoxin light chain protease through the heavy chain channel. Nat. Struct. Biol. 10:13-18. [PubMed]
24. Kurokawa, Y., K. Oguma, N. Yokosawa, B. Syuto, R. Fukatsu, and I. Yamashita. 1987. Binding and cytotoxic effects of Clostridium botulinum type A, C1 and E toxins in primary neuron cultures from foetal mouse brains. J. Gen. Microbiol. 133:2647-2657. [PubMed]
25. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685. [PubMed]
26. Li, L., T. Binz, H. Niemann, and B. R. Singh. 2000. Probing the mechanistic role of glutamate residue in the zinc-binding motif of type A botulinum neurotoxin light chain. Biochemistry 39:2399-2405. [PubMed]
27. Li, L., and B. R. Singh. 2000. Role of zinc binding in type A botulinum neurotoxin light chain's toxic structure. Biochemistry 39:10581-10586. [PubMed]
28. Montecucco, C., and G. Schiavo. 1995. Structure and function of tetanus and botulinum neurotoxins. Q. Rev. Biophys. 28:423-472. [PubMed]
29. Ohishi, I., and G. Sakaguchi. 1977. Activation of botulinum toxins in the absence of nicking. Infect. Immun. 17:402-407. [PMC free article] [PubMed]
30. Ransom, B. R., E. Neale, M. Henkart, P. N. Bullock, and P. G. Nelson. 1977. Mouse spinal cord in cell culture. I. Morphology and intrinsic neuronal electrophysiologic properties. J. Neurophysiol. 40:1132-1150. [PubMed]
31. Schagger, H., and G. von Jagow. 1987. Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166:368-379. [PubMed]
32. Schiavo, G., and C. Montecucco. 1995. Tetanus and botulism neurotoxins: isolation and assay. Methods Enzymol. 248:643-652. [PubMed]
33. Schiavo, G., O. Rossetto, A. Santucci, B. R. DasGupta, and C. Montecucco. 1992. Botulinum neurotoxins are zinc proteins. J. Biol. Chem. 267:23479-23483. [PubMed]
34. Schiavo, G., A. Santucci, B. R. Dasgupta, P. P. Mehta, J. Jontes, F. Benfenati, M. C. Wilson, and C. Montecucco. 1993. Botulinum neurotoxins serotypes A and E cleave SNAP-25 at distinct COOH-terminal peptide bonds. FEBS Lett. 335:99-103. [PubMed]
35. Schmidt, J. J., and K. A. Bostian. 1997. Endoproteinase activity of type A botulinum neurotoxin: substrate requirements and activation by serum albumin. J. Protein Chem. 16:19-26. [PubMed]
36. Schmidt, J. J., and K. A. Bostian. 1995. Proteolysis of synthetic peptides by type A botulinum neurotoxin. J. Protein Chem. 14:703-708. [PubMed]
37. Schmidt, J. J., and R. G. Stafford. 2002. A high-affinity competitive inhibitor of type A botulinum neurotoxin protease activity. FEBS Lett. 532:423-426. [PubMed]
38. Schmidt, J. J., R. G. Stafford, and K. A. Bostian. 1998. Type A botulinum neurotoxin proteolytic activity: development of competitive inhibitors and implications for substrate specificity at the S1′ binding subsite. FEBS Lett. 435:61-64. [PubMed]
39. Schmidt, J. J., R. G. Stafford, and C. B. Millard. 2001. High-throughput assays for botulinum neurotoxin proteolytic activity: serotypes A, B, D, and F. Anal. Biochem. 296:130-137. [PubMed]
40. Sellin, L. C., J. A. Kauffman, and B. R. Dasgupta. 1983. Comparison of the effects of botulinum neurotoxin types A and E at the rat neuromuscular junction. Med. Biol. 61:120-125. [PubMed]
41. Shapiro, R. L., C. Hatheway, and D. L. Swerdlow. 1998. Botulism in the United States: a clinical and epidemiologic review. Ann. Intern. Med. 129:221-228. [PubMed]
42. Sharma, S. K., and B. R. Singh. 2004. Enhancement of the endopeptidase activity of purified botulinum neurotoxins A and E by an isolated component of the native neurotoxin associated proteins. Biochemistry 43:4791-4798. [PubMed]
43. Sheridan, R. E., and S. S. Deshpande. 1995. Interactions between heavy metal chelators and botulinum neurotoxins at the mouse neuromuscular junction. Toxicon 33:539-549. [PubMed]
44. Sheridan, R. E., T. J. Smith, and M. Adler. 2005. Primary cell culture for evaluation of botulinum neurotoxin antagonists. Toxicon 45:377-382. [PubMed]
45. Simpson, L. L., J. A. Coffield, and N. Bakry. 1993. Chelation of zinc antagonizes the neuromuscular blocking properties of the seven serotypes of botulinum neurotoxin as well as tetanus toxin. J. Pharmacol. Exp. Ther. 267:720-727. [PubMed]
46. Simpson, L. L., A. B. Maksymowych, and S. Hao. 2001. The role of zinc binding in the biological activity of botulinum toxin. J. Biol. Chem. 276:27034-27041. [PubMed]
47. Sugiyama, H., B. R. DasGupta, and K. H. Yang. 1974. Toxicity of purified botulinal toxin fed to mice. Proc. Soc. Exp. Biol. Med. 147:589-591. [PubMed]
48. Swaminathan, S., and S. Eswaramoorthy. 2000. Structural analysis of the catalytic and binding sites of Clostridium botulinum neurotoxin B. Nat. Struct. Biol. 7:693-699. [PubMed]
49. Tucker, N., J. Sobel, S. E. Maslanka, P. M. Griffin, and R. V. Tauxe. 2002. Surveillance for botulism: summary of 2001 data. Centers for Disease Control and Prevention, Atlanta, Ga.
50. Williamson, L. C., S. C. Fitzgerald, and E. A. Neale. 1992. Differential effects of tetanus toxin on inhibitory and excitatory neurotransmitter release from mammalian spinal cord cells in culture. J. Neurochem. 59:2148-2157. [PubMed]
51. Wright, J. F., M. Pernollet, A. Reboul, C. Aude, and M. G. Colomb. 1992. Identification and partial characterization of a low affinity metal-binding site in the light chain of tetanus toxin. J. Biol. Chem. 267:9053-9058. [PubMed]

Articles from Infection and Immunity are provided here courtesy of American Society for Microbiology (ASM)