Snake phospholipase A2 neurotoxins enter neurons, bind specifically to mitochondria, and open their transition pores

Snake presynaptic neurotoxins with phospholipase A(2) activity are potent inducers of paralysis through inhibition of the neuromuscular junction. These neurotoxins were recently shown to induce exocytosis of synaptic vesicles following the production of lysophospholipids and fatty acids and a sustained influx of Ca(2+) from the medium. Here, we show that these toxins are able to penetrate spinal cord motor neurons and cerebellar granule neurons and selectively bind to mitochondria. As a result of this interaction, mitochondria depolarize and undergo a profound shape change from elongated and spaghetti-like to round and swollen. We show that snake presynaptic phospholipase A(2) neurotoxins facilitate opening of the mitochondrial permeability transition pore, an inner membrane high-conductance channel. The relative potency of the snake neurotoxins was similar for the permeability transition pore opening and for the phospholipid hydrolysis activities, suggesting a causal relationship, which is also supported by the effect of phospholipid hydrolysis products, lysophospholipids and fatty acids, on mitochondrial pore opening. These findings contribute to define the cellular events that lead to intoxication of nerve terminals by these snake neurotoxins and suggest that mitochondrial impairment is an important determinant of their toxicity.     

Membrane interactions of tetanus and botulinum neurotoxins: a photolabelling study with photoactivatable phospholipids

Tetanus and botulinum neurotoxins (TeNT and BoNT) bind strongly and specifically to the nervous tissue, as it can be inferred from their potency and from their effects restricted to the nervous system. The molecular basis of these properties are presently unknown. As a first approach, we have investigated the interaction of TeNT and BoNT with model membranes by photolabelling with phospholipid analogues carrying the photoreceptor group at different positions of the lipid molecule in order to probe different membrane regions. We found that at neutral pH TeNT and BoNTs (type A, B and E) adsorb onto the surface of negatively charged liposomes. Polysialogangliosides increase this interaction only slightly thus suggesting that they provide a minor contribution to toxin lipid binding. On this basis we propose that clostridial neurotoxins bind to lipids via both a predominant unspecific interaction with negatively charged lipids (including gangliosides) and a specific, but weaker, interaction with polysialogangliosides. At acidic pH values both chains of these neurotoxins are labelled strongly by photogroups located in the hydrophobic milieu of the membrane with a pH dependence that overlaps the range of pH values reached in the endosomal lumen. This result is consistent with their insertion into the lipid bilayer in agreement with the idea that clostridial neurotoxins may penetrate into cells via intracellular low pH compartments.     

Interaction between the two subdomains of the C-terminal part of the botulinum neurotoxin A is essential for the generation of protective antibodies

The botulinum neurotoxin A C-terminal fragment (Hc), which mediates the binding of the toxin to neuronal cell surface receptors, comprises two subdomains, Hc-N (amino acids 873-1095) and Hc-C (amino acids 1096-1296). In order to define the minimal fragment of Hc carrying protective antigenic properties, Hc, Hc-N and Hc-C have been produced as recombinant proteins in Escherichia coli, and have been tested for their antigenicity in mouse protection assays. Hc, Hc-N and Hc-C induced similar antibody levels as shown by ELISA. However, a single immunization with Hc (10 microg) fully protected mice challenged with 10(3) mouse lethal dose 50 of toxin, whereas Hc-N, Hc-C, or Hc-N plus Hc-C did not give any protection. Triple immunizations with Hc-N or Hc-C were necessary to induce a higher level of protection. Circular dichroism and fluorescence studies showed that the isolated subdomains were folded and stable. However, an intense near-UV dichroic signal was only observed in the Hc spectrum, revealing a highly structured interface between both subdomains. Taken together, the results show that the generation of protective antibodies requires the whole Hc domain and especially the native structure of the interfacial region between Hc-N and Hc-C.     

Disassembly of the reconstituted synaptic vesicle membrane fusion complex in vitro.

The interaction of the presynaptic membrane proteins SNAP-25 and syntaxin with the synaptic vesicle protein synaptobrevin (VAMP) plays a key role in the regulated exocytosis of neurotransmitters. Clostridial neurotoxins, which proteolyze these polypeptides, are potent inhibitors of neurotransmission. The cytoplasmic domains of the three membrane proteins join into a tight SDS-resistant complex (Hayashi et al., 1994). Here, we show that this reconstituted complex, as well as heterodimers composed of syntaxin and SNAP-25, can be disassembled by the concerted action of the N-ethylmaleimide-sensitive factor, NSF, and the soluble NSF attachment protein, alpha-SNAP. alpha-SNAP binds to predicted alpha-helical coiled-coil regions of syntaxin and SNAP-25, shown previously to be engaged in their direct interaction. Synaptobrevin, although incapable of binding alpha-SNAP individually, induced a third alpha-SNAP binding site when associated with syntaxin and SNAP-25 into heterotrimers. NSF released prebound alpha-SNAP from full-length syntaxin but not from a syntaxin derivative truncated at the N-terminus. Disassembly of complexes containing this syntaxin mutant was impaired, indicating a critical role for the N-terminal domain in the alpha-SNAP/NSF-mediated dissociation process. Complexes containing C-terminally deleted SNAP-25 derivatives, as generated by botulinal toxins type A and E, were dissociated more efficiently. In contrast, the N-terminal fragment generated from synaptobrevin by botulinal toxin type F produced an SDS-sensitive complex that was poorly dissociated.     

Interaction of botulinum and tetanus toxins with the lipid bilayer surface.

The interaction of botulinum neurotoxins serotypes A, B and E (from Clostridium botulinum) and of tetanus neurotoxin (from Clostridium tetani) with the surface of liposomes made of different lipid compositions was studied by photolabelling with a radioiodinated photoactive phosphatidylethanolamine analogue [125I-dipalmitoyl (3,4-azidosalicylamido)phosphatidylethanolamine]. When the vesicles were made of negatively charged lipids (asolectin), each of these neurotoxic proteins was radioiodinated, thus providing evidence for their attachment to the membrane surface. The presence of gangliosides on liposome membranes enhanced fixation of the neurotoxic proteins to the lipid vesicle surface. Both the heavy and light chains of the clostridial neurotoxins were involved in the attachment to the lipid bilayer surface. Each of the toxins tested here attached poorly to liposomes made of zwitterionic lipids (egg phosphatidylcholine), even when polysialogangliosides were present. The data suggest that the binding of botulinum and tetanus neurotoxins to their target neuronal cells involves negatively charged lipids and polysialogangliosides on the cell membrane.     

Tetanus, botulinum and snake presynaptic neurotoxins

Tetanus and botulinum neurotoxins, produced by anaerobic bacteria of the genus Clostridium, are the most toxic proteins known and are solely responsible for the pathogenesis of tetanus and botulism. They are metallo-proteases that enter nerve terminals and cleave proteins of the neuroexocytosis apparatus causing a persistent, but reversible, inhibition of neurotransmitter release. Botulinum neurotoxins are used in the therapy of many human syndromes caused by hyperactive nerve terminals. Snake presynaptic PLA2 neurotoxins block nerve terminals by binding to the nerve membrane and catalyzing phospholipid hydrolysis with production of lysophospholipids and fatty acids. These compounds change the membrane conformation causing enhanced fusion of synaptic vesicle via hemifusion intermediate with release of neurotransmitter and, at the same time, inhibition of vesicle fission and recycling. It is possible to envisage clinical applications of the lysophospholipid/fatty acid mixture to inhibit hyperactive superficial nerve terminals.

---

     

Flap loop of GluD2 binds to Cbln1 and induces presynaptic differentiation

Glutamate receptor δ2 (GluD2) is selectively expressed on the postsynaptic spines at parallel-fiber (PF)-Purkinje neuron (PN) synapses. GluD2 knockout mice show a reduced number of PF-PN synapses, suggesting that GluD2 is involved in synapse formation. Recent studies revealed that GluD2 induces presynaptic differentiation in a manner dependent on its N-terminal domain (NTD) through binding of Cbln1 secreted from cerebellar granule neurons. However, the underlying mechanism of the specific binding of the NTD to Cbln1 remains elusive. Here, we have identified the flap loop (Arg321-Trp339) in the NTD of GluD2 (GluD2-NTD) as a crucial region for the binding to Cbln1 and the induction of presynaptic differentiation. Both induction of presynaptic differentiation and binding of Cbln1 were abolished in the HEK cells expressing not wild-type GluD2 but GluD2 with mutations in the flap loop. Especially, single amino acid substitution of either Arg321 or Trp323 to alanine was sufficient to disable the GluD2 function. Finally, a homology model of GluD2-NTD suggested that the flap loop is located at the distal end, which appears consistent with an interaction with Cbln1 and a presynaptic varicosity.     

How do presynaptic PLA2 neurotoxins block nerve terminals?

Snake presynaptic neurotoxins with phospholipase A2 activity block nerve terminals in an unknown way. Here, we propose that they enter the lumen of synaptic vesicles following endocytosis and hydrolyse phospholipids of the inner leaflet of the membrane. The transmembrane pH gradient drives the translocation of fatty acids to the cytosolic monolayer, leaving lysophospholipids on the lumenal layer. Such vesicles are highly fusogenic and release neurotransmitter upon fusion with the presynaptic membrane, but cannot be retrieved because of the high local concentration of fatty acids and lysophospholipids, which prevents vesicle neck closure.     

Tetanus and botulinum neurotoxins: mechanism of action and therapeutic uses

The clostridial neurotoxins responsible for tetanus and botulism are proteins consisting of three domains endowed with different functions: neurospecific binding, membrane translocation and proteolysis for specific components of the neuroexocytosis apparatus. Tetanus neurotoxin (TeNT) binds to the presynaptic membrane of the neuromuscular junction, is internalized and transported retroaxonally to the spinal cord. The spastic paralysis induced by the toxin is due to the blockade of neurotransmitter release from spinal inhibitory interneurons. In contrast, the seven serotypes of botulinum neurotoxins (BoNTs) act at the periphery by inducing a flaccid paralysis due to the inhibition of acetylcholine release at the neuromuscular junction. TeNT and BoNT serotypes B, D, F and G cleave specifically at single but different peptide bonds, of the vesicle associated membrane protein (VAMP) synaptobrevin, a membrane protein of small synaptic vesicles (SSVs). BoNT types A, C and E cleave SNAP-25 at different sites located within the carboxyl-terminus, while BoNT type C additionally cleaves syntaxin. The remarkable specificity of BoNTs is exploited in the treatment of human diseases characterized by a hyperfunction of cholinergic terminals.     

Cell entry strategy of clostridial neurotoxins

Tetanus neurotoxin and botulinum neurotoxins are the causative agents of tetanus and botulism. They block the release of neurotransmitters from synaptic vesicles in susceptible animals and man and act in nanogram quantities because of their ability to specifically attack motoneurons. They developed an ingenious strategy to enter neurons. This involves a concentration step via complex polysialo gangliosides at the plasma membrane and the uptake and ride in recycling synaptic vesicles initiated by binding to a specific protein receptor. Finally, the neurotoxins shut down the synaptic vesicle cycle, which they had misused before to enter their target cells, via specific cleavage of protein core components of the cellular membrane fusion machinery. The uptake of four out of seven known botulinum neurotoxins into synaptic vesicles has been demonstrated to rely on binding to intravesicular segments of the synaptic vesicle proteins synaptotagmin or synaptic vesicle protein 2. This review summarizes the present knowledge about the cell receptor molecules and the mode of toxin-receptor interaction that enables the toxins' sophisticated access to their site of action.     

Functional duality and structural uniqueness of depressant insect-selective neurotoxins

Depressant insect-selective neurotoxins derived from scorpion venoms (a) induce in blowfly larvae a short, transient phase of contraction similar to that induced by excitatory neurotoxins followed by a prolonged flaccid paralysis and (b) displace excitatory toxins from their binding sites on insect neuronal membranes. The present study was undertaken in order to examine the basis of these similarities by comparing the primary structures and neuromuscular effects of depressant and excitatory toxins. A new depressant toxin (LqhIT2) was purified from the venom of the Israeli yellow scorpion. The effects of this toxin on a prepupal housefly neuromuscular preparation mimic the effects on the intact animal; i.e., a brief period of repetitive bursts of junction potentials is followed by suppression of their amplitude and finally by a block of neuromuscular transmission. Loose patch clamp recordings indicate that the repetitive activity has a presynaptic origin in the motor nerve and closely resembles the effect of the excitatory toxin AaIT. The final synaptic block is attributed to neuronal membrane depolarization, which results in an increase in spontaneous transmitter release; this effect is not induced by excitatory toxin. The amino acid sequences of three depressant toxins were determined by automatic Edman degradation. The depressant toxins comprise a well-defined family of polypeptides with a high degree of sequence conservation. This group differs considerably in primary structure from the excitatory toxin, with which it shares identical or related binding sites, and from the two groups of scorpion toxins that affect sodium conductance in mammals. The two opposing pharmacological effects of depressant toxins are discussed in light of the above data.     

In vitro binding of isolated synaptic vesicles to presynaptic plasma membranes: activation by Ca2+ and protein kinase C.

An in vitro model to study the molecular control of binding of highly purified synaptic vesicles to presynaptic plasma membranes has been developed. Presynaptic plasma membranes were immobilized by dotting onto nitrocellulose, and binding of iodinated synaptic vesicle membranes was studied under varying experimental conditions. Synaptic vesicles bind to presynaptic plasma membranes in the presence of Ca2+ and ATP. Binding is reduced in the presence of EGTA and abolished by the calmodulin antagonist trifluoperazine. Vesicle binding is stimulated 5-fold after incubation--prior to dotting--of presynaptic plasma membranes with ATP in the presence of the phorbol-ester 12-O-tetradecanoylphorbol-13-acetate (1 microM) and 2.5-fold after preincubation with Ca2+ (50 microM). Pretreatment of plasma membranes with alkaline phosphatase strongly reduces vesicle binding. Microsomes prepared from bovine liver did not bind to presynaptic plasma membranes. Our results suggest that activation of protein kinase C and Ca2+ stimulate binding of synaptic vesicles to the presynaptic membrane. In the intact nerve terminal this interaction may represent an initial step in synaptic vesicle exocytosis.     

Effect of cholinergic ligands on the lipids of acetylcholine receptor-rich membrane preparations from torpedo californica

Ion permeation, triggered by ligand-receptor interaction, is associated with the primary events of membrane depolarization at the neuromuscular junction and synaptic connections. To explore the possible sites of ion permeation, the long-lived fluorescent probe pyrene (fluorescence lifetime ～400 nsec) has been inserted into the lipid phase of acetylcholine receptor-rich membrane (AcChR-M) preparations from Torpedo californica. The pyrene probe is susceptible to both fluidity and permeability changes in the lipid bilayer. These changes are detected by variations in the rate of decay of the excited singlet state of pyrene after pulsation with a 10-nsec ruby laser flash. Variations of these lifetimes in the membrane preparations alone or in the presence of quenchers show that binding of cholinergic agonists and antagonists, neurotoxins, and local anesthetics to AcChR-M produces varying effects on the properties of the pyrene probe in the lipid phase. It is concluded that binding of cholinergic ligands to the receptor does not significantly alter the fluidity or permeability of the lipids in the bilayer in contact with pyrene. On the other hand, local anesthetics do affect these properties.     

Clostridial neurotoxins: new tools for dissecting exocytosis

Tetanus toxin and botulinal toxins are potent inhibitors of neuronal exocytosis. Within the past five years the protein sequences of all eight neurotoxins have been determined, their mode of action as metalloproteases has been established, and their intraneuronal targets have been identified. The toxins act by selectively proteolysing the synaptic vesicle protein synaptobrevin (VAMP) or the presynaptic membrane proteins syntaxin (HPC-1) and SNAP-25. These three proteins form the core of a complex that mediates fusion of carrier vesicles to target membranes. Tetanus and botulinal neurotoxins could serve in the future as tools to study membrane trafficking events, or even higher brain functions such as behaviour and learning.     

Role of key aromatic residues in the ligand-binding domain of alpha7 nicotinic receptors in the agonist action of beta-amyloid

Soluble β-amyloid (Aβ) resides in certain regions of the brain at or near picomolar concentration, rising in level during the prodromic stage of Alzheimer disease. Recently, we identified the homomeric α7 nicotinic acetylcholine receptor (α7-nAChR) as one possible functional target for picomolar Aβ. This study was aimed at addressing which residues in α7-nAChRs potentially interact with Aβ to regulate the presynaptic function of this receptor. Site-directed mutagenesis was carried out to study the key aromatic residues in the mouse α7-nAChR agonist-binding pocket. Mutations of tyrosine188 resulted in a decrease in activation of presynaptic α7-nAChRs by ACh and Aβ but with no change in response to nicotine, indicating the critical role of Tyr-188 in presynaptic regulation by Aβ. Coimmunoprecipitation additionally revealed direct binding of Aβ to α7-nAChRs and to the Tyr-188 mutant receptor. In contrast, mutations of Tyr-195 in α7-nAChR led to decreased activation by nicotine without apparent effects on ACh- or Aβ-induced responses. Agonist-induced responses of Tyr-93 mutant α7-nAChRs indicated possible interactions of nicotine and Aβ with its hydroxyl group, but there was no change in presynaptic responses after mutation of Trp-149. All of the mutants were shown to be expressed on the plasma membrane using cell surface labeling. Together, these results directly demonstrate an essential role for the aromatic residue Tyr-188 as a key component in the agonist binding domain for the activation of α7-nAChRs by Aβ.     

Leukotriene and prostaglandin production in rat brain synaptosomes treated with phospholipase A2 neurotoxins and enzymes

beta-Bungarotoxin (beta-BuTX) and notexin cause an irreversible blockade of neurotransmitter release through specific and potent effects at the presynaptic nerve terminal, however, the mechanism of action is uncertain. We examined the effects of beta-BuTX and notexin on LT and PG production in rat cerebrocortical synaptosomes in order to determine if eicosanoid production might mediate or regulate the pharmacological actions of these phospholipase A2 (PLA2) neurotoxins. The effects of the PLA2 enzymes isolated from Naja naja atra and Naja nigricollis snake venoms (which are not presynaptic selective) on LT and PG production were compared with the effects of beta-BuTX and notexin. N. n. atra PLA2, beta-BuTX, and notexin (all 50 nM) produced a time dependent rise in free fatty acids as measured in synaptic plasma membranes isolated from treated synaptosomes. Both the PLA2 neurotoxins and enzymes stimulated LTC4, LTB4, and PGE2 production, as measured by radioimmunoassay. In all cases, the PLA2 enzymes were more potent than the PLA2 neurotoxins. This observation correlates with their relative enzymatic potencies, as measured by free fatty acid generation. EDTA and BSA antagonized PLA2 induced LTB4 production and BSA also antagonized PLA2 induced PGE2 production. These results suggest that stimulation of eicosanoid production does not mediate the potent and specific presynaptic actions of beta-BuTX and notexin.     

Ammodytoxins, potent presynaptic neurotoxins, are also highly efficient phospholipase A2 enzymes

The enzymatic activity of ammodytoxins (Atxs), secreted phospholipases A(2) (sPLA(2)s) in snake venom, is essential for expression of their presynaptic neurotoxicity, but its exact role in the process is unknown. We have analyzed in detail the enzymatic properties of Atxs, their mutants, and homologues. The apparent rates of phospholipid hydrolysis by the sPLA(2)s tested vary by up to 4 orders of magnitude, and all enzymes display a strong preference for vesicles containing anionic phospholipids, phosphatidylglycerol or phosphatidylserine (PS), over those containing zwitterionic phosphatidylcholine (PC). Nevertheless, Atxs are quite efficient in hydrolyzing pure PC vesicles as well as PC-rich plasma membranes of intact HEK293 cells. The presence of anionic phospholipids in PC vesicles dramatically increases the interfacial binding affinity and catalytic activity of Atxs, but not of their nontoxic homologue ammodytin I(2), that displays unusually low binding affinity and enzymatic activity on PS-containing vesicles and HEK293 plasma membranes. Aromatic and hydrophobic residues on the interfacial binding surface of Atxs are important for productive binding to both zwitterionic and anionic vesicles, while basic and polar residues have a negative impact on binding to zwitterionic vesicles. When tightly bound to the membrane interface, Atxs can reach full enzymatic activity at low micromolar concentrations of Ca(2+). Although Atxs have evolved to function as potent neurotoxins that specifically target presynaptic nerve terminals, they display a high degree of phospholipolytic efficiency on various phospholipid membranes.     

Presynaptic enzymatic neurotoxins

Botulinum neurotoxins produced by anaerobic bacteria of the genus Clostridium are the most toxic proteins known, with mouse LD50 values in the 1-5 ng/kg range, and are solely responsible for the pathophysiology of botulism. These metalloproteinases enter peripheral cholinergic nerve terminals and cleave proteins of the neuroexocytosis apparatus, causing a persistent, but reversible, inhibition of neurotransmitter release. They are used in the therapy of many human syndromes caused by hyperactive nerve terminals. Snake presynaptic PLA2 neurotoxins block nerve terminals by binding to the nerve membrane and catalyzing phospholipid hydrolysis with production of lysophospholipids and fatty acids. These compounds change the membrane conformation, causing enhanced fusion of synaptic vesicle via hemifusion intermediate with release of neurotransmitter and, at the same time, inhibition of vesicle fission and recycling. It is possible to envisage clinical applications of the lysophospholipid/fatty acid mixture to inhibit hyperactive superficial nerve terminals.     

Mechanism of botulinum neurotoxin B and G entry into hippocampal neurons

Botulinum neurotoxins (BoNTs) target presynaptic nerve terminals by recognizing specific neuronal surface receptors. Two homologous synaptic vesicle membrane proteins, synaptotagmins (Syts) I and II, bind toxins BoNT/B and G. However, a direct demonstration that Syts I/II mediate toxin binding and entry into neurons is lacking. We report that BoNT/B and G fail to bind and enter hippocampal neurons cultured from Syt I knockout mice. Wild-type Syts I and II (but not Syt I loss-of-function toxin-binding domain mutants) restored binding and entry of BoNT/B and G in Syt I–null neurons, thus demonstrating that Syts I/II are protein receptors for BoNT/B and G. Furthermore, mice lacking complex gangliosides exhibit reduced sensitivity to BoNT/G, and binding and entry of BoNT/A, B, and G into hippocampal neurons lacking gangliosides is diminished. These data suggest that gangliosides are the shared coreceptor for BoNT/A, B, and G, supporting a double-receptor model for these three BoNTs for which the protein receptors are known.     

Tachykinin receptors in the rat brain bind alpha-bungarotoxin

alpha-Bungarotoxin was found to inhibit effectively the binding of 125I-labelled substance P and eledoisin to membrane and to solubilize preparations of the rat brain. Other postsynaptic neurotoxins exerted similar but less pronounced influence on the interaction of tachykinins with their receptors. The obtained results suggest that some alpha-bungarotoxin-binding polypeptides in brain are components of tachykinin receptors.