Dissection des mécanismes synaptiques de la libération des neurotransmetteurs par l’utilisation de toxines bactériennes

Bacterial toxins as tools to decipher exocytotic mechanisms in neurones

¹ Laboratoire de Neurobiologie Cellulaire, U PR 9009 du CNRS, 5, rue Biaise Pascal, 67084 Strasbourg Cédex.
² Dipartimento di Farmacologia e Fisiologia Umana, Facoltà di Medicina e Chirurgia, Università di Bari, Italie.
³ Interaction Bactérie Cellule, Institut Pasteur, 28, rue du Docteur Roux, 75 724 Paris Cédex.
⁴ A qui la correspondance doit être adressée.

Résumé

L’identification des mécanismes moléculaires mis en jeu lors de l’exocytose des neurotransmetteurs nécessite l’utilisation d’approches expérimentales qui permettent d’affecter spécifiquement certaines des protéines synaptiques impliquées. Le mécanisme d’action et les protéines cibles de plusieurs toxines bactériennes étant connus, ces dernières peuvent être utilisées comme autant d’outils moléculaires pour modifier sélectivement la fonctionnalité de certaines des protéines impliquées dans la libération des neurotransmetteurs ou pour révéler leur implication dans cette fonction. Après une revue rapide du mécanisme d’action et des conditions d’utilisation en neurobiologie de trois groupes de toxines bactériennes, nous illustrons notre propos par des exemples tirés de nos expériences. Nous abordons ainsi successivement les neurotoxines clostridiales qui clivent des protéines de la machinerie de fusion, des toxines (ADP-ribosyl-transférase, glucosyl-transférase ou désamidase) qui inactivent ou activent différentes petites protéines G- monomériques (Rho, Rac, CDC42, Ras, Ral, Rap, Rab) de la superfamille Ras, et des toxines qui ADP- ribosylent le monomère d’actine.

Abstract

Several bacterial toxins are powerful and highly specific tools for studying basic mechanisms involved in cell biology. Whereas the clostridial neurotoxins are widely used by neurobiologists, many other toxins (i.e. toxins acting on small G-proteins or actin) are still overlooked.

Botulinum neurotoxins (BoNT, serotypes A-G) and tetanus neurotoxin (TeNT), known under the generic term of clostridial neurotoxins, are characterized by their unique ability to selectively block neurotransmitter release. These proteins are formed of a light (Mr ~ 50) and a heavy (Mr ~ 100) chain which are disulfide linked. The cellular action of BoNT and TeNT involves several steps: heavy chain-mediated binding to the nerve ending membrane, endocytosis, and translocation of the light chain (their catalytic moiety) into the cytosol. The light chains each cleaves one of three, highly conserved, proteins (VAMP/synaptobrevin, syn- taxin, and SNAP-25 also termed SNAREs) implicated in fusion of synaptic vesicles with plasma membrane at the release site. Hence, when these neurotoxins areapplied extracellularly, they can be used as specific tools to inhibit evoked and spontaneous transmitter release from certain neurones whereas, when the membrane limiting steps are bypassed by the mean of intracellular applications, BoNTs orTeNT can be used to affect regulated secretion in various cell types.

Several members of the Rho GTPase family have been involved in intracellular trafficking of synaptic vesicles and secretory organelles. As they are natural targets for several bacterial exoenzymes or cytotoxins, their role in neurotransmitter release can be probed by examining the action of these toxins on neurotransmission. Such toxins include: /') the non permeant C3 exoenzymes from C. botulinum or C. limo- sum which ADP-ribosylate and thereby inactivate Rho, ii) exoenzyme S from Pseudomonas aeruginosa which ADP-ribosylates different members of the Ras, Rab, Ral and Rap families, iii) toxin B from C. difficile which glucosylates Rho, Rac and CDC42, iv) lethal toxin from C. sordellii which glucosylates Rac, Ras and to a lesser extent, Rap and Ral, but noton Rho or CDC42, and v) CNF deamidases secreted by pathogenic strains of E. coli which activate Rho and, to a lesser extent, CDC42. Since these toxins or exoenzymes have no or little ability to enter into the neurones, they must be applied intraneuronally to bypass the membrane limiting steps. Injection of several of these toxins into Aplysia neurones allowed us to reveal a new role for Rac in the control of exocytosis.

ADP-ribosylating enzymes, which specifically act on monomeric actin (C2 binary toxin from C. botulinum and iota toxin from C. perfringens), are potential tools to probe the role of actin filaments during secretion.