Covering the most recent advances in our understanding of toxins from venomous animals and microbes as well as that of their targets, this book expertly addresses the many intriguing and unsolved questions concerning; proteomics studies of the "toxinome", intimate modes of toxin actions, molecular basis of specificity, pleiotropic properties of toxins and structural biology of toxins.

Through twenty-seven chapters the authors discuss the role of structural genomics in toxinology, how toxins are subject to accelerated evolution, how toxins can be exploited as models for the design of new drugs, and what the future holds for the treatment of snake bites.

In order to address these challenging aspects, the authors have posed crystal-clear questions. Based on the most precise knowledge the attendant reasoning shows how toxinology has become an important area of biochemistry and is directly associated with advances in cellular microbiology, molecular pharmacology, molecular physiology, cell biology, protein engineering and many other disciplines.

Perspectives in Molecular Toxinology

---

John Wiley & Sons

Copyright © 2002 John Wiley & Sons, Ltd
All right reserved.
ISBN: 0-471-49503-4

---

Chapter One

Bacterial Toxins with Metalloprotease Activity

ORNELLA ROSSETTO and CESARE MONTECUCCO

1.1 INTRODUCTION

Metalloproteases are hydrolytic enzymes characterised by an active site containing a metal atom, usually zinc. They include amino-peptidases, carboxy-peptidases and endopeptidases depending on whether they remove an N- or a C-terminal residue or cleave internal peptide bonds of the protein substrate. Zinc-dependent endopeptidases are characterised by the presence of a zinc-binding motif consisting of His-Glu-X-X-His. Hundreds of different metalloproteases are produced by microorganisms and are involved in their metabolic activities or are released outside to cleave substrates not elaborated by the microorganism itself. The present chapter will only deal with bacterial metalloproteases which act on a specific cellular or tissue target of the mammalian bacterial host.

The recent determination of their primary sequence has led to the discovery of the metalloproteolytic activity of the bacterial toxins responsible for tetanus, botulism and anthrax. The protease domain of these toxins enters into the cytosol where it displays a zinc-dependent endopeptidase activity of remarkable specificity. Tetanus (TeNT) and botulinum toxins (BoNTs) cleave three protein components of the neuroexocytosis machinery leading to the blockade of neurotransmitter release and consequent paralysis. BoNTs are increasingly used in medicine for the treatment of human diseases characterised by hyperfunction of cholinergic terminals.

The lethal factor of Bacillus anthracis is specific for the MAP kinase-kinases that are cleaved within their amino-terminus. In this case, however, such a specific biochemical lesion has not yet been correlated with the pathogenesis of anthrax.

Fragilysin (BFT) is produced by toxigenic strains of the intestinal pathogen Bacteroides fragilis and attacks E-cadherin.

1.2 TETANUS AND BOTULINUM NEUROTOXINS 1.2.1 INTRODUCTION

Tetanus neurotoxin (TeNT) and botulinum neurotoxin (BoNT) were identified as the sole cause of tetanus and botulism, respectively, a little over a century ago, after the discovery of the anaerobic and spore-forming bacteria of the genus Clostridium. There are seven types of BoNT (indicated with letters from A to G) which differ in antigenicity and biochemical activity. BoNTs bind to and enter peripheral cholinergic terminals, causing a sustained block of acetylcholine (ACh) release, with ensuing flaccid paralysis and autonomic symptoms. Tetanus neurotoxin (TeNT) acts on the CNS and blocks neurotransmitter release at the inhibitory interneurons of the spinal cord, resulting in a frequently lethal spastic paralysis. Despite the opposite clinical symptoms of tetanus and botulism, the neurotoxins affect the same neuronal function: neuroexocytosis.

1.2.2 GENETICS AND STRUCTURE

In C. tetani and in C. botulinum G, the neurotoxin genes are contained within large plasmids, whereas in C. botulinum A, B, E and F the neurotoxin genes have a chromosomal localisation and in C. botulinum C and D toxins are encoded by bacteriophages. Usually one bacterium harbours one toxin gene, but several cases of multiple toxin genes have been reported. These genes do not contain a secretion signal sequence and the protein neurotoxins are released by bacterial autolysis as single polypeptide chains of 150 kDa, which are later activated by a specific proteolytic cleavage within a loop subtended by a highly conserved disulphide bridge. The heavy chain (H, 100 kDa) and the light chain (L, 50 kDa) remain associated via noncovalent interactions and by the conserved interchain S-S bond, whose integrity is essential for neurotoxicity, but which has to be reduced to allow the display of the metalloproteolytic activity of the L chain in the cytosol (Figure 1.1A).

The crystallographic structures of BoNT/A and BoNT/B and of the C-terminal part of the heavy chain of TeNT revealed that the 50 kDa receptor binding domain, termed Hc, consists of two sub-domains. The N-terminal part of He (HcN) consists of sixteen R-strands and four at-helices arranged in a jelly roll motif, closely similar to that of carbohydrate-binding proteins of the legume lectin family. The amino acid sequence of this sub-domain is highly conserved among BoNTs and TeNT, suggesting a closely similar three-dimensional structure. In contrast, the sequence of the C-terminal part of He (HcC) is poorly conserved, but folds similarly to proteins of the trypsin inhibitor family. On the basis of experiments performed with TeNT, it was suggested that HcC plays a major role in neurospecific binding.

The N-terminal part of the heavy chain (HN) features two ~ 100 Å-long antiparallel [alpha]-helices, similar to those of the membrane interacting proteins colicins and influenza hemagglutinin. The HNS of the CNTs are highly homologous and their predicted secondary structures are also highly similar, in agreement with their proposed role in transmembrane translocation of the L chain.

The L chain is a metalloprotease with little protein-protein interaction with the adjacent translocation domain (HN), which is in turn linked to the receptor- binding domain. At the centre of the long cleft-shaped active site there is a zinc atom coordinated via the two histidines and the glutamic residues of the zinc-binding motif, and by Glu262 in BoNT/A and Glu268 in BoNT/B, a residue conserved among clostridial neurotoxins which corresponds to Glu271 of TeNT (Figure 1.1 A). The Glu residue of the motif is particularly important because it coordinates the water molecule which actually performs the hydrolytic reaction of proteolysis. Its mutation leads to complete inactivation of these neurotoxins. The critical role of Glu271 of TeNT and Glu262 in BoNT/A has been shown to be that of providing a negatively charged carboxylate moiety and in preparation). This active site architecture is similar to that of thermolysin and identifies a primary sphere of residues essential to the catalytic function, which coincides with the zinc coordinating residues. In addition, it appears that a secondary layer of residues, less close to the zinc centre, is present at the active site of clostridial neurotoxins (Figure 1.1 B). Among these residues, Arg363 and Tyr366 in BoNT/A could play a role in the catalytic activity of this family of metalloproteases. In particular, Tyr366 in BoNT/A (corresponding to Tyr373 in BoNT/B and to Tyr375 in TeNT) point its phenolic ring inside the cleft-shaped active site of the toxin. The mutation of Tyr375 with an alanine inactivates the TeNT L chain, clearly indicating that this residue plays a critical role in the hydrolysis of the substrate. It has been proposed that Tyr373 of BoNT/B assists the hydrolysis reaction by donating a proton to the amide nitrogen of VAMP Phe77 which, together with bound water molecules, stabilises the leaving group.

The active site of the L chain faces the H chain in the unreduced toxin, accounting for its lack of proteolytic activity, and becomes accessible to the substrate upon reduction of the interchain disulphide bridge. Their proteolytic activity is zinc-dependent and heavy metal chelators such as ortho-phenantroline, which remove bound zinc, generate inactive apo-neurotoxins, but the active site metal atom can be reacquired upon incubation of apo-toxin in zinc-containing buffers. The biochemical and structural properties of clostridial neurotoxins define them as a distinct group of metalloproteases, whose origin cannot at present be traced to any of the known families of metalloprotease.

Such structural organisation of the CNTs has been shaped by evolution to fulfil the requirements of their mode of neuron intoxication which consists of four steps: 1) binding, 2) internalisation, 3) membrane translocation, and 4) proteolytic cleavage of their substrates (Figure 1.2).

1.2.3 NEURONAL INTOXICATION

From the site of production or adsorption (intestine or wounds), BoNTs and TeNT diffuse in the body fluids, up to the presynaptic membrane of cholinergic terminals where they bind very specifically. The He domain plays a major role in neurospecific binding [22], but additional regions may be involved in determining the remarkable specificity for cholinergic terminals of CNTs.

Identification of the presynaptic receptor(s) of CNTs has been attempted by several investigators. Polysialogangliosides are certainly involved together with as yet unidentified proteins of the presynaptic membrane. The presence of both lectin-like and protein binding sub-domains in the He domain supports the suggestion that CNTs bind strongly and specifically to the presynaptic membrane because they display multiple interactions with sugar- and protein-binding sites. Recently, BoNT/B was shown to bind strongly to the synaptic vesicle protein synaptotagmin II only in the presence of polysialogangliosides, but its role in vivo remains to be established. Identification of the receptors for the various CNTs will constitute a major advance in the understanding of the mechanism of neuron intoxication and help to improve current therapeutic protocols employing BoNT to treat human syndromes of hyperfunction of cholinergic terminals and excessive muscle contraction.

As depicted in Figure 1.2, the L chains of CNTs block neuroexocytosis by acting in the cytosol and they reach this cell compartment following endocytosis and membrane translocation. They are internalised inside acidic cellular compartments via a temperature- and energy-dependent process. Nerve stimulation facilitates intoxication by CNTs and a close link exists between stimulus-contraction coupling and endocytosis at nerve terminals. Hence endocytosis and other factors such as nerve stimulation-dependent proteolytic activity in the cytosol may partly account for this effect, which is potentially very relevant for the development of novel protocols of therapy employing BoNT. The protein receptor of TeNT would be responsible for its inclusion in an endocytic vesicle that moves in a retrograde direction all along and inside the axon, whereas BoNTs' protein receptors would guide them inside vesicles that acidify within the NMJ. The TeNT-carrying vesicles reach the cell body in the spinal cord and then move to dendritic terminals to release the toxin in the intersynaptic space. TeNT equilibrates between pre- and post-synaptic membranes and then binds and enters the inhibitory interneurons of the spinal cord via synaptic vesicle endocytosis.

To reach the cytosol the L chain has to cross the hydrophobic barrier of the vesicle membrane (Figure 1.2) and the acidity of the lumen is essential for such a movement. CNTs have to be exposed to a low pH step for nerve intoxication to occur. Acidic pH does not induce a direct activation of the toxin via a structural change, but is required in the process of transmembrane translocation of the L chain itself. CNTs undergo a low pH-driven conformational change from a water-soluble 'neutral' structure to an 'acid' structure characterised by the surface exposure of hydrophobic patches, which lead the H and L chains in the hydrocarbon core of the lipid bilayer. Following the low pH-induced membrane insertion, BoNTs and TeNT form transmembrane ion channels in planar lipid bilayers of low conductance. There is a general consensus that the toxin channels participate in the process of transmembrane translocation of the L domain, from the vesicle membrane to the nerve terminal cytosol, but there is no agreement on how this process may take place. The BoNT translocation domain is different from those of other pore-forming toxins since the long pair of [alpha]-helices, with their triple helix bundle at either end, resemble more some coiled coil viral proteins which do not translocate through pores but change structure at low pH and insert into membranes. It has been proposed that the L chain translocates across the vesicle membrane within a channel opened laterally to lipids, rather than inside a wholly proteinaceous pore, accounting for the fact that the L chain does contact the fatty acid chains of lipids during translocation. The H chain is suggested to form a transmembrane hydrophilic cleft that nests in the passage of the partially unfolded L chain with its hydrophobic segments facing the lipids. Facing the cytosolic neutral pH, the L chain refolds and regains its water-soluble neutral conformation. This model is also supported by the finding that the protein-translocating channel of the endoplasmic reticulum has been shown to be open laterally to lipids.

Once in the cytosol, CNTs exploit their catalytic activity. BoNTs and TeNT are remarkably specific proteases that recognise and cleave only three proteins, the so-called SNARE proteins, which form the core of the neuroexocytosis machinery. TeNT, BoNT/B, /D, /F and /G cleave VAMP, at different single peptide bonds; BoNT/A and /E cleave SNAP-25 at different sites within the COOH-terminus whereas BoNT/C cleaves both syntaxin and SNAP-25. Strikingly, TeNT and BoNT/B cleave VAMP at the same peptide bond (Gln76-Phe77) and yet, when injected into the animal, they cause the opposite symptoms of tetanus and botulism, respectively, conclusively demonstrating that the different symptoms of the two diseases derive from different sites of intoxication rather than from a different molecular mechanism of action.

VAMP, SNAP-25 and syntaxin form a heterotrimeric coil- coiled complex, termed the SNARE complex, which induces the juxtaposition of vesicle to the target membrane and is involved in their fusion. VAMP is a family of vesicular SNAREs with a short C-terminal tail facing the vesicle lumen, a single transmembrane domain and the remaining N-terminal part exposed to the cytosol. Different VAMP isoforms are located on different cell vesicles and contribute to address each vesicle to its appropriate target membrane with which it will fuse. VAMP-1 and -2 are the isoforms mainly involved in the binding and fusion of neurotransmitter-containing synaptic vesicles with the presynaptic membrane (neuroexocytosis). Syntaxin is anchored to target membranes via a C-terminal hydrophobic tail. Of the many syntaxin isoforms presently known, syntaxin IA, 1B and 2 are the isoforms mainly involved in neuroexocytosis. SNAP-25 (few isoforms) are 25 kDa SNARE proteins bound to the target membrane via fatty acids covalently linked to cysteine residues present in the middle of the polypeptide chain.

The proteolysis of one SNARE protein prevents the formation of the complex and consequently the release of the neurotransmitter. The SNARE complex is insensitive to CNT proteolysis, as expected on the basis of the fact that proteases are known to attack predominantly unstructured exposed loops.

Continues...

---

Excerpted from Perspectives in Molecular Toxinology Copyright © 2002 by John Wiley & Sons, Ltd . Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

---