Unit: Anaerobe Bacteria and Toxins

Director: Popoff Michel R.

Regulation of the toxinogenesis in Clostridium botulinum and Clostridium tetani, investigation of the function of the regulatory proteins BotR and TetR, analysis of the two component systems of regulation. Study of large clostridial toxins (C. sordellii toxins) which glucosylate Rho and Ras GTPases and investigation of apoptosis and signaling pathways involved in the control of the actin cytoskeleton and intercellular junctions. Effects of C. perfringens epsilon toxin on cell barriers.

The National Reference Center for Anaerobes and Botulism is involved in the survey of botulism and identification of anaerobe isolates (see ).

Research on botulism

Study of the regulation of toxinogenesis in Clostridium botulinum

The botulinum neurotoxins (BoNT; 150 kDa) are associated non covalently with other non toxic proteins (associated non toxic proteins, ANTPs) to form large size botulinum complexes (300 to 900 kDa). The genes encoding BoNT and ANTPs are clustered on DNA fragment called botulinum locus. They are organized in two operons, one containing the genes of the botulinum neurotoxin and the non toxic non hemagglutinin protein (NTNH), and the other one contains the genes of the three hemagglutinins (HAs). Another gene encoding for a protein (BotR) which has the characteristics of DNA binding proteins, is localized in the 5' part of the botulinum locus in C. botulinum A and B. This gene is conserved in C. tetani (tetR). We have previously shown that BotR is a positive regulator of the expression of bont and antps genes by overexpressing botR gene in C. botulinum A or inhibiting it by an antisens mRNA construction. BotR/A from C. botulinum type A was produced as recombinant protein with a six His tag in E. coli and purified by affinity chromatography. Purified BotR/A and crude lysate from recombinant C. botulinum A which overproduces BotR/A were tested in gel shift assay with DNA fragments from the promoter regions of the two operons. The crude lysates from recombinant C. botulinum induced a gel shift, whereas purified BotR/A did not. However BotR/A in the presence of RNA polymerase core enzyme from E. coli caused a gel retardation of promoter DNAs. This indicates that BotR/A is probably an alternative sigma factor of the same family than TxeR from Clostridium difficile. The BotR/A-RNA polymerase core enzyme complex binds preferentially to the promoter DNAs from the two operons of the botulinum locus and weakly to the promoters of the other genes. This indicates that bont and antps genes are essentially transcribed as polycistronic genes.

Passage of the intestinal barrier by botulinum neurotoxin A

The intestinal cells CaCo-2 grown on filters form polarized cell monolayers with tight and adherent junctions. Purified BoNT/A or BoNT/A in complex form was inoculated on the apical side of CaCo-2 cells. The passage of the neurotoxin was monitored by measurement of the lethal activity on mice. The integrity of the cell monolayers checked by measurement of the electrical trans epithelial resistance and immunofluorescent staining of actin, apical junctions (ZO1), and adherens junctions (E-cadherin) was not modified by botulinum neurotoxin treatment. BoNT/A alone was able to cross CaCo-2 cell monolayers (yield of passage 1-5%), but the passage was increased when BoNT/A was associated to ANTPs. The BoNT/A passage was inhibited at 4°C. Moreover, the BoNT/A passage was saturable over a period of 20-30 min. This indicates that BoNT/A crosses CaCo-2 cell monolayers by a receptor mediated transcytosis mechanism.

Clostridium sordellii LT induces apoptosis in human myeloid HL60 cells

It is by synthetazing a lethal toxin (LT) that some strains of Clostridium sordellii are pathogenic and can induce a gaz gangrene. LT is a high molecular weight protein (250 kDa) which penetrates into cells. It contains a glucosyl-transferase activity towards several small GTPases and thus, leading to their inactivation. The targets of LT toxin from C. sordellii strain IP82 are majorly Rac, a small GTPase from the Ro family which acts on actin polymerization, and GTPases from the Ras family including Ras, Rap and Ral. LT from the strain 9048 has slighly different targets as it also inactivates Cdc42 another GTPase from the Rho family but has no effect on Ral GTPases.

In studying the effect of LT-IP82 on phospholipase D, an enzyme activity that can be regulated by Ral and Rho family GTPases, we observed that this toxin induced the death of human myeloid HL-60 cells. Death by apoptosis was demonstrated, as the cells presented a nucleus with chromatin condensation and nucleus fragmentation, their DNA was typically fragmented and non-permeable cells were labelled with Annexin-V FITC, which binds to phophatidylserine residues exposed on the outer leaflet of the plasma membrane in apoptotic cells. We have shown that LT-IP82 induces cell death via an intrinsic apoptotic pathway as it was found to disrupt mitochondrial homeostasis. This was characterized by a decrease in mitochondrial transmembrane potential and cardiolipin alterations. Time-course studies of caspase activation revealed that caspase-9 was activated prior to caspase-8. Protection of mitochondria by Bcl-2 overexpression prevented mitochondrial changes as well as apoptosis induction. Moreover, we showed that cell exposure to LT-IP82 leads to a co-localization of the toxin with a mitochondrial marker within 2 hours (see photo).

In an attempt to determined whether apoptosis is related to the effect of LT-IP82 on one of the small GTPases, we provided evidence that cortical actin disorganization is not the cause of cell death suggesting that Rac inactivation does not initiate the apoptotic process. Indeed other toxins such as iota toxin (Ia-Ib) from C. perfringens or cytochalasin D, which also disorganize the actin cytoskeleton do not induce apoptosis. Moreover, we also observed that LT-9048, when used at the same concentration as LT-IP82, is not able to induce HL-60 cell apoptosis. This lethal toxin, has slightly different targets from that of LT-IP82 as it inactivates Cdc42 another small GTPase from the Rho family which also play a role on cortical actin organization. In contrast it has no effect on Ral GTPases suggesting that Ral could have an essential role on HL-60 cell orientation towards surviving or death providing that the apoptotic effect of LT-IP82 is linked to the toxin glucosyl-transferase activity ant thus to its effect on one of its targets.

Clostridial toxins which modify the cell barriers

Some Clostridium produce potent toxins acting on cell barriers such as intestinal or kidney epithelium, which constitute the first line of defense of the organisms against bacterial invasion. A model of cell barrier is experimentally obtained by culture of cell monolayer on filter.

Epsilon toxin, produced by C. perfringens B and D, is responsible for severe enterotoxemia in animals, and in addition this toxin can cause lesions in humans. We have previously shown that epsilon toxin binds to a cell surface receptor on dog kidney cells (MDCK cell line) and forms highly stable and large size complexes, which correspond to toxin oligomerization and pore formation. Subsequently, the toxin induces the loss of intracellular K+, the cellular increase of Na+ and Cl-, and a rapid death of the cell monitored by the cell viability mitochondria-dependent test (MTT). Epsilon toxin also induces the cell entry of propidium iodide in parallel to the cell death, indicating that the toxin forms large pores. In artificial lipid bilayers, we have shown that epsilon toxin induces non selective pores permeable to hydrophilic solutes up to a molecular mass of at least1 kDa.

Epsilon toxin increases dramatically and rapidly the permeability of MDCK cell monolayers cultivated on filter. This activity was related to the pore formation on cell membranes. Epsilon toxin binds to either the apical or basolateral cell surface, but it has not been visualized in intracellular compartment (Fig. 2). Actin cytoskeleton, intercellular junctions, and paracellular permeability are not directly modified by epsilon toxin.

Large clostridial toxins such as C. difficile toxins A and B and C. sordellii LT, glucosylate small G proteins from the Rho and Ras family and therefore alter the actin cytoskeleton. The study of LT on the actin cytoskeleton and intercellular junctions on intestinal cell monolayers cultivated on filter is in progress.


1. Human promyelocytic HL-60 cells were treated with LT-IP82 (100ng/ml) for 2 and 16 hours, then fixed with 2% PFA. LT localization at the level of mitochondrion was studied by confocal microscopy using a polyclonal antibody against the N-term domain of the toxin and a monoclonal antibody (MAB 1273) against a mitochondrial protein.

2. Cellular localization of epsilon toxin on polarized MDCK cells grown on filter. Epsilon toxin was incubated either on the apical (A) or basolateral (B) cell side. Cells were fixed and immunostained with fluorescent anti-toxin antibodies and observed by confocal microscopy. Note that the immunostaining of epsilon toxin on the cell surface corresponding to that used for the toxin incubation, and the absence of intracellular compartment immunostaining.

Keywords: Clostridium, botulism, toxins, Rho-GTPases, apoptosis

Activity Reports 2003 - Institut Pasteur

Page Top research Institut Pasteur homepage

If you have problems with this Web page, please write to rescom@pasteur.fr