Anatomy Of A Neurotoxin
By: Carol A. Rouzer, VICB Communications
Published: April 9, 2013
Three-dimensional analysis of transmission electron micrographs provides insight into the structure of botulinum neurotoxin progenitor complexes.
Botulinum neurotoxin (BoNT), the most potent known toxin, is produced by several species of bacteria belonging to the genus Clostridium. The seven serotypes (A through G) of BoNT share a similar structure, comprising a 50 kD light chain and a 100 kD heavy chain. Targeting the neuromuscular junction, the heavy chain of BoNT binds to receptors on the membrane of the presynaptic neuron, allowing entry of the light chain into the cell. The light chain is a zinc metalloprotease that cleaves and destroys protein components of the neuron’s synaptic membrane fusion complex. As a result, the neuron is unable to release neurotransmitter to stimulate the postsynaptic muscle cell, leading to flaccid paralysis of the intoxicated organism. Bacteria release BoNTs in the form of a progenitor complex (PC) containing up to four additional neurotoxin-associated proteins (NAPs). The oral potency of the PC, as measured by the median lethal dose, is higher than that of BoNT alone, suggesting that the NAPs play an important role in BoNT-mediated toxicity. Yet, the exact nature of their role remains a mystery. To solve that mystery, VICB member Borden Lacy and her collaborator Melanie Ohi have elucidated the three-dimensional structure of the BoNT PC [D. A. Benefield et al. (2013) Proc. Natl. Acad. Sci. U.S.A., published online March 18, DOI:10.1073/pnas.1222139110.].
Depending on the serotype, BoNTs form PCs of varying sizes, designated 12S (~300 kDa), 16S (~500 kDa), and 19S (~900 kDa), on the basis of their behavior during ultracentrifugation. The 12S PC contains BoNT plus one additional protein, designated nontoxic nonhemagglutinin (NTNH), which is similar in structure to the BoNT. The larger PCs also contain three hemagglutinin proteins (HA1, HA2, and HA3), named for their ability to aggregate red blood cells. Desirée Benefield, a graduate student in the Lacy lab began the investigation of PC structure by obtaining images of the BoNT/A1, BoNT/B, and BoNT/E serotype PCs using transmission electron microscopy (TEM). The images (Figure 1) suggested that the 16S BoNT/A1 and BoNT/B PCs comprised an ovoid body attached to three flexible arms, while the 12S BoNT/E PC exhibited only the ovoid body. The flexible arms in the case of the BoNT/A1 and BoNT/B PCs appeared to have “pincher-like” structures at the ends (Figure 1).
Figure 1. TEM images of the BoNT PC serotypes A1 (left), B (center), and E (right). BoNT/A1 and BoNT/B appear to have an ovoid body attached at one end to three flexible arms. At the end of the arms is a “pincher-like” structure. BoNT/E comprises the ovoid body only. Reproduced with permission from D. A. Benefield et al. (2013) Proc. Natl. Acad. Sci. U.S.A., published online March 18, DOI:10.1073/pnas.1222139110. Copyright 2013, Benefield et al.
To obtain a three-dimensional model of the BoNT/A1 PC, the investigators obtained two sets of TEM images of over 15,000 complexes, one with the sample grid placed perpendicular to the electron beam and the other with the sample tilted at -55o. Analysis of the combined images using the random conical tilt method initially provided a set of averaged two-dimensional views of the complex. Two distinct orientations emerged. The “flat” orientation showed the ovoid body with two arms, while the “prong” orientation revealed the body with just one arm (Figure 2). Despite the appearance of three arms in the original images, these were not seen in the averaged and refined results. However, when the flat and prong orientations were used to elucidate the three-dimensional structure of the complex, two models, one with two arms and one with a single arm emerged. Combining the two models produced a structure comprising an ovoid body capped with a triangular plate bearing three flexible arms with pincher-like ends (Figure 3).
Figure 2. The results of image averaging yielded two orientations for the BoNT/A1 particles. The “flat” orientation (left) exhibited an ovoid body with two arms. The “prong” orientation (right) suggested an ovoid body with only one arm. Reproduced with permission from D. A. Benefield et al. (2013) Proc. Natl. Acad. Sci. U.S.A., published online March 18, DOI:10.1073/pnas.1222139110. Copyright 2013, Benefield et al.
Figure 3. 3D reconstruction of images in the flat and prong orientations yielded structures showing two arms (A) or one arm (B), respectively, attached to an ovoid body. Combining the two structures yielded a complete structure with three arms (C). Reproduced with permission from D. A. Benefield et al. (2013) Proc. Natl. Acad. Sci. U.S.A., published online March 18, DOI:10.1073/pnas.1222139110. Copyright 2013, Benefield et al.
Previously published data provided crystal structures of portions of PCs isolated from BoNTs of different serotypes. These included a BoNT-NTNH dimer from the A1 serotype, an (HA1)2-HA2 trimer from the D serotype, and an HA3 trimer from the C serotype. Comparison of these structures with the model obtained from the TEM data revealed that the BoNT-NTNH dimer could easily be superimposed onto the ovoid body portion of the complex, while the HA3 trimer overlaid nicely into the triangular plate, and the (HA1)2-HA2 trimer fit into the pincher-like arms (Figure 4). The orientation of the BoNT-NTNH dimer at first was uncertain, since either the BoNT or the NTNH light chain region could be placed next to the triangular plate. However, use of an antibody directed against the BoNT light chain indicated that it was NTNH that came in direct contact with the triangular plate, while BoNT assumed the most distal position of the ovoid body. A similar approach revealed that the BoNT/B 16S PC assumed an overall structure very similar to that of the BoNT/A1 PC, while the BoNT/E 12S PC structure consisted of only the ovoid body.
Figure 4. Crystal structures of the BoNT/A1-NTNH complex (left), the BoNT/D (HA1)2-HA2 trimer (center), and the BoNT/C HA3 trimer (right) were superimposed into the 3D structure of the BoNT/A1 PC (below). In this case, the 3D model obtained from the flat orientation is shown. Reproduced with permission from D. A. Benefield et al. (2013) Proc. Natl. Acad. Sci. U.S.A., published online March 18, DOI:10.1073/pnas.1222139110. Copyright 2013, Benefield et al.
These results suggest that the 16S PC comprises one molecule each of BoNT and NTNH, three molecules each of HA2 and HA3, and six molecules of HA1. The resulting molecular mass would be ~760 kDa, somewhat larger than the 500 kDa predicted from ultracentrifugation. It is interesting to note that the BoNT molecule comes into direct contact only with NTNH, suggesting that NTNH may play a role in protecting BoNT from acidic or proteolytic degradation. This is particularly important because BoNT’s primary route of access into the body is through the digestive system. Indeed, substantial prior evidence supports a protective role for NTNH. In contrast, the finding that the HA proteins do not contact BoNT indicates that they likely play a different function. Some reports have indicated that the PC facilitates BoNT’s attachment to and transport across the epithelium of the digestive tract. The HA proteins bear a number of sugar binding sites that may be important for attachment to mucins or cell surface glycoproteins. The Lacy and Ohi team proposed that the presence of these binding sites on the three flexible arms of the PC might promote exploration and sampling of the epithelial surface for appropriate sites of binding and entry of the toxin (Figure 5). Together, the data provide a strong foundation upon which to build a better understanding of how BoNT gains access to the body and reaches its target at the neuromuscular junction.
Figure 5. Two views of the proposed model of the BoNT/A1 PC showing the (HA1)2
-HA2 trimer (light and dark blue), the HA3 trimer (rose and pink), NTNH (green), and BoNT (gold). Bright red highlights show the locations of sugar binding sites. Reproduced with permission from D. A. Benefield et al. (2013) Proc. Natl. Acad. Sci. U.S.A., published online March 18, DOI:10.1073/pnas.1222139110. Copyright 2013, Benefield et al.