John L. Middlebrook

Identification of the site at which phospholipase A2 neurotoxins localize to produce their neuromuscular blocking effects

Experiments were conducted on mouse hemidiaphragm preparations using five phospholipase A2 neurotoxins of differing chain structures and antigenicities [notexin (one chain); crotoxin (two chains not covalently bound), beta-bungarotoxin (two chains covalently bound); taipoxin (three chains), and textilotoxin (five chains; one copy each of three chains and two copies of a fourth chain)]. Three clostridial neurotoxins (botulinum neurotoxin types A and B, and tetanus toxin) were used in comparison experiments. Phospholipase A2 neurotoxins produced concentration-dependent blockade of neuromuscular transmission. There was no obvious relationship between chain structure and potency, but there was an indication of a relationship between chain structure and binding. The binding of notexin was substantially reversible, the binding of crotoxin was slightly reversible, and the binding of beta-bungarotoxin, taipoxin and textilotoxin was poorly reversible. Experiments with neutralizing antibodies indicated that phospholipase A2 neurotoxins became associated with binding sites on or near the cell surface. This binding did not produce neuromuscular blockade. When exposed to physiological temperatures and nerve stimulation, bound toxin disappeared from accessibility to neutralizing antibody. This finding suggests that there was some form of molecular rearrangement. The two most likely possibilities are: (1) there was a change in the conformation of the toxin molecule, or (2) there was a change in the relationship between the toxin and the membrane. The molecular rearrangement step did not produce neuromuscular blockade. At a later time there was onset of paralysis; the amount of time necessary for onset of blockade was a function of toxin concentration. Phospholipase A2 neurotoxins were not antagonized by drugs that inhibit receptor-mediated endocytosis. In addition, phospholipase A2 neurotoxins did not display the pH-induced conformational changes that are typical of other endocytosed proteins, such as clostridial neurotoxins. However, phospholipase A2 neurotoxins were antagonized by strontium, and this antagonism was expressed against toxins that were free in solution and toxins that were bound to the cell surface. Limited antagonism was expressed after toxins had undergone molecular rearrangement, and no antagonism was expressed after toxin-induced neuromuscular blockade. The cumulative data suggest that phospholipase A2 neurotoxins are not internalized to produce their poisoning effects. These toxins appear to act on the plasma membrane, and this is the site at which they initiate the events that culminate in neuromuscular blockade.

Primary sequence determination of the most basic myonecrotic phospholipase A2 from the venom of Vipera russelli

The most basic phospholipase A2 (PLA2), VRV-PL-VIIIa, was purified from (Sri Lankan) Vipera russelli venom. It is a major component of the venom, contributing over 40% to the whole venom PLA2 activity. The purity of VRV-PL-VIIIa was ascertained by electrophoresis and by reverse phase high-pressure liquid-chromatography (RP-HPLC). VRV-PL-VIIIa had an apparent mol. wt of 13,000 and was a single polypeptide. The protein was reduced, pyridylethylated and subjected to sequence analysis. The N-terminal amino acid sequence was established up to the 39th residue. Pyridylethylated VRV-PL-VIIIa was digested with endoprotease Glu-C, and several peptides were purified by RP-HPLC; six purified peptides were sequenced. The sequence of the C-terminal was established by sequencing a CNBr-produced peptide purified by RP-HPLC. Several peptides were also generated by digestion with endoprotease Asp-N. Two peptides were sequenced to obtain overlapping regions. The complete structure was deduced from sequences of overlapping peptides and through homology with other group II PLA2 sequences. Sequence homology was greatest with ammodytoxin A: 99 amino acid residues out of 121 occurred in identical positions. Myotoxin III of Bothrops asper showed 73% homology, 89 out of 121 residues. In agreement with the sequence data, polyclonal antiserum against VRV-PL-VIIIa cross-reacted in ELISA with ammodytoxin A and, to a lesser extent, with caudoxin.

Purification, sequencing and characterization of pseudexin phospholipases A2 from Pseudechis porphyriacus (Australian red-bellied black snake)

Pseudexin is the name given to a mixture of toxic phospholipase A2 isoenzymes isolated from the venom of the Australian red-bellied black snake, Pseudechis porphyriacus. We found that this mixture consists of three components: pseudexins A, B and C, which we individually purified by reverse phase chromatography or by hydrophobic interaction chromatography. Pseudexins A and B had relatively low specific toxicities in mice (i.p. LD50 of 1300 and 750 micrograms/kg, respectively), while C was non-toxic. All three had similar phospholipase A2 activities (43-53 muequiv H+ released/min/mg protein). The complete amino acid sequences of pseudexins A and B were determined. Amino acids were identical at 91 of the 117 residues. The first 28 residues of pseudexin C were determined, sufficient to show that C is structurally similar to A and B, but not identical with either. As judged by reactions with antisera against several other snake phospholipase A2 toxins, pseudexins A, B and C have very similar antigenic structures. We noted extensive homology with other phospholipases.

Mosquito inoculation: An alternative bioassay for toxins

Mosquitoes were evaluated as a bioassay host for several classes of biological toxins. Mosquitoes were sensitive to snake toxic or neurotoxic phospholipase A2 enzymes (but not to nontoxic phospholipase A2 enzymes), cobrotoxin, saxitoxin, microcystin and the scorpion insect sodium channel toxin. Mosquitoes were not sensitive to ricin, diphtheria toxin, anthrax toxin, botulinum toxin, tetanus toxin, conotoxin G or a scorpion sodium channel toxin toxic to mammals. Specific antisera neutralization tests with mosquitoes gave comparable results to those of a mouse assay. The mosquito is a suitable bioassay animal for many, but not all biological toxins, and offers a safer, more efficient and economical assay than mice.

[12] Isolation and characterization of the Botulinum neurotoxins

Publisher Summary This chapter describes the isolation and characterization of Clostridium botulinum neurotoxins. Clostridium botulinum is an anaerobic, spore-forming organism that is relatively ubiquitous in its distribution. The bacterium is known for its ability to produce two remarkably potent toxins. The first of these is a neurotoxin that acts preferentially on cholinergic nerve endings to block the release of acetylcholine. The second is a more diffusely acting toxin that, among other things, appears to promote the movement of fluid across membranes. Botulinum neurotoxin is isolated in seven immunologically distinct forms designated A, B, C, D, E, F, and G. Although the various neurotoxins are immunologically distinct—for example, there is little cross-neutralization, they do share certain structural and functional properties. In addition, pharmacological experiments suggest that the various serotypes may proceed through the same general sequence of events in producing blockade of transmitter release. The tendency to form complexes with certain other proteins is a characteristic of most of the botulinum neurotoxins .

Immunological relationships of phospholipase A2 neurotoxins from snake venoms

Polyclonal rabbit antisera were raised against ten snake phospholipase A2 neurotoxins and one snake phospholipase A2 cytotoxin. Immunological cross-reactivities between these toxins, two other snake phospholipase A2 enzymes and pancreatic phospholipase A2 were studied using ELISA technology. All snake phospholipase A2 neurotoxins fell into two main antigenic classes. One antigenic class was composed of all the elapid toxins tested (textilotoxin, taipoxin, notexin, pseudexin and beta-bungarotoxin), the cytotoxic phospholipase A2 from Naja naja atra and pancreatic phospholipase A2. beta-Bungarotoxin seemed to be in an immunological subclass of its own compared to the rest of the elapid toxins. The second antigenic class was comprised of crotalid and viperid phospholipase A2 neurotoxins (crotoxin, concolor toxin, Mojave toxin, vegrandis toxin, ammodytoxin and caudoxin). Our data indicated that the viperid toxins, caudoxin and ammodytoxin, were an immunological subclass apart from the crotalid toxins.

Cross-reactivity and neutralization by rabbit antisera raised against crotoxin, its subunits and two related toxins.

Antisera were raised against intact crotoxin (Crotalus durissus terrificus), Mojave toxin (Crotalus scutulatus scutulatus) and concolor toxin (Crotalus viridis concolor), as well as the subunits of crotoxin. Double immunodiffusion and enzyme-linked immunosorbent assays (ELISA) demonstrated antigenic similarity between these three purified toxins and their subunits. Additionally, when crotoxin antisera were pre-incubated with each of the three toxins before injection, the lethal activity of all were neutralized equally well. Antiserum was considerably more effective in neutralizing crotoxin in vivo when the toxin was injected i.m. than when injected i.v. Antisera against both intact crotoxin and its basic subunit were an order of magnitude more effective than crotoxin acidic subunit antiserum in crotoxin neutralization. Purified phospholipase A2 from Crotalus adamanteus and Crotalus atrox showed weak cross-reactivity with antisera raised against intact crotoxin and its subunits in the ELISA. Our results suggest that crotalid neurotoxins can be detected and neutralized by polyclonal antibodies raised against any intact toxin or basic subunit in this group of homologous toxins.

Effects of beta-bungarotoxin and Naja naja atra snake venom phospholipase A2 on acetylcholine release and choline uptake in synaptosomes

Beta-bungarotoxin is a potent presynaptically acting snake venom toxin that exhibits phospholipase A2 activity. We compared the effects of beta-bungarotoxin and a less toxic snake venom phospholipase A2 on synaptosomal 3H-acetylcholine release and 3H-choline uptake. The purpose of these experiments was to study the mode by which beta-bungarotoxin inhibits 3H-acetylcholine release in this preparation. Under non-depolarizing conditions, both beta-bungarotoxin and Naja naja atra phospholipase A2 stimulated 3H-acetylcholine release from a synaptosomal fraction preloaded with 3H-choline. Beta-bungarotoxin was more potent, but less efficacious, than N. naja atra phospholipase A2. In contrast, both toxins inhibited 3H-acetylcholine release from the synaptosomal fraction incubated with 3H-choline after toxin exposure. In agreement with the results obtained by monitoring acetylcholine release, beta-bungarotoxin and N. naja atra phospholipase A2 appeared to block 3H-choline uptake into the synaptosomal fraction non-competitively. Although the toxins may cause the release of unlabeled choline from synaptosomes, the block of labeled choline uptake could not be explained by decreased specific activity of 3H-choline in the bathing medium. Therefore, beta-bungarotoxin and N. naja atra phospholipase A2 block 3H-acetylcholine release from synaptosomes indirectly by inhibiting the uptake of 3H-choline necessary for 3H-acetylcholine synthesis. In comparing these results using 3H-choline to those in the literature obtained with deuterated choline, there appears to be a difference in apparent toxin action that relates to the type of label (3H or 2H) attached to choline.

Preparation of a crotoxin neutralizing monoclonal antibody

Crotoxin is a heterodimeric protein composed of an acidic and basic subunit from the venom of Crotalus durissus terrificus and is representative of a number of presynaptically acting neurotoxins found in the venom of rattlesnakes. Four different monoclonal antibodies, typed as IgG1 subclass, were raised against the basic subunit of this toxin. One was a potent neutralizing antibody of intact crotoxin, which could neutralize approximately 1.6 moles of purified crotoxin per mole of antibody. The monoclonal antibody enhanced the neutralizing ability of commercial polyvalent crotalid antivenom against the lethality of crude C. d. terrificus venom four-fold. Paradoxically, this monoclonal antibody by itself was ineffective against the lethality of crude C. d. terrificus venom. Using an enzyme-linked immunosorbent assay, we tested various proteins for competitive inhibition of binding of biotinylated-crotoxin to plates coated with the four individual monoclonal antibodies. Concolor toxin, vegrandis toxin, intact crotoxin, Mojave toxin, and the basic subunit of crotoxin showed increasing effectiveness as displacers of crotoxin from the neutralizing monoclonal antibody. None of the monoclonal antibodies reacted with purified phospholipase A2 enzymes from Crotalus atrox or Crotalus adamanteus, nor any of the components present in the crude venoms from four different elapids known to contain presynaptically acting neurotoxins, which show some sequence identity to crotoxin.

Effects of myonecrotic snake venom phospholipase A2 toxins on cultured muscle cells

We have attempted to establish a cell culture model suitable for molecular mechanism of action studies of necrotic phospholipases A2 (PLA2). Three myonecrotic PLA2 were purified, one basic PLA2 from Naja nigricollis venom and two basic PLA2 (VRV-PL-V and VRV-PL-VIIIa) from Vipera russelli venom. The effects of these PLA2 on several established muscle cell lines were evaluated. As judged by light microscopy, some, but not all, cell lines detached from the culture plate in a time- and concentration-related fashion. Naja nigricollis PLA2 was the most potent at eliciting this effect, followed by VRV-PL-V and VRV-PL-VIIIa. The two most sensitive cell lines, 1447 and 1456, were chosen for further study using N. nigricollis PLA2. Cellular protein and nucleic acid syntheses were inhibited by the toxin in a time- and dose-related manner. However, it appeared that most, if not all, of the inhibition was due to toxin-induced reduction of precursor uptake, suggesting effects at the plasma membrane level. The putative membrane effects were specific, in that uptake of calcium, choline or glucose was not inhibited by the toxin. Moreover, treating the cells with toxin failed to significantly increase lactate dehydrogenase release into the medium. Polyclonal antiserum prepared against N. nigricollis basic PLA2 neutralized the toxicity completely with 1456 cells, but only partially with the 1447 cell line. Both the 1447 and 1456 lines appear to be suitable as cell culture models for necrotizing PLA2 molecular mechanism of action studies.