Structural determinants for membrane insertion, pore formation and translocation of Clostridium difficile toxin B

Clostridium difficile toxins A and B bind to eukaryotic target cells, are endocytosed and then deliver their N-terminal glucosyltransferase domain after processing into the cytosol. Whereas glucosyltransferase, autoprocessing and cell-binding domains are well defined, structural features involved in toxin delivery are unknown. Here, we studied structural determinants that define membrane insertion, pore formation and translocation of toxin B. Deletion analyses revealed that a large region, covering amino acids 1501-1753 of toxin B, is dispensable for cytotoxicity in Vero cells. Accordingly, a chimeric toxin, consisting of amino acids 1-1550 and the receptor-binding domain of diphtheria toxin, caused cytotoxic effects. A large N-terminal part of toxin B (amino acids 1-829) was not essential for pore formation (measured by (86) Rb(+) release in mammalian cells). Studies using C-terminal truncation fragments of toxin B showed that amino acid residues 1-990 were still capable of inducing fluorescence dye release from large lipid vesicles and led to increased electrical conductance in black lipid membranes. Thereby, we define the minimal pore-forming region of toxin B within amino acid residues 830 and 990. Moreover, we identify within this region a crucial role of the amino acid pair glutamate-970 and glutamate-976 in pore formation of toxin B.     

Cholesterol-dependent pore formation of Clostridium difficile toxin A

The large clostridial cytotoxins toxin A and toxin B from Clostridium difficile are major virulence factors known to cause antibiotic-associated diarrhea and pseudomembranous colitis. Both toxins mono-glucosylate and thereby inactivate small GTPases of the Rho family. Recently, it was reported that toxin B, but not toxin A, induces pore formation in membranes of target cells under acidic conditions. Here, we reassessed data on pore formation of toxin A in cells derived from human colon carcinoma. Treatment of 86Rb+-loaded cells with native or recombinant toxin A resulted in an increased efflux of radioactive cations induced by an acidic pulse. The efficacy of pore formation was dependent on membrane cholesterol, since cholesterol depletion of membranes with methyl-beta-cyclodextrin inhibited 86Rb+ efflux, and cholesterol repletion reconstituted pore-forming activity of toxin A. Similar results were obtained with toxin B. Consistently, methyl-beta-cyclodextrin treatment delayed intoxication of cells in a concentration-dependent manner. In black lipid membranes, toxin A induced ion-permeable pores only in cholesterol containing bilayers and at low pH. In contrast, release of glycosylphosphatidylinositol-anchored structures by phosphatidylinositol specific phospholipase C treatment did not reduce cell sensitivity toward toxins A and B. These data indicate that in colonic cells toxin A induces pore formation in an acidic environment (e.g. endosomes) similar to that reported for toxin B and suggest that pore formation by clostridial glucosylating toxins depends on the presence of cholesterol.     

Structural determinants of Clostridium difficile toxin A glucosyltransferase activity

The principle virulence factors in Clostridium difficile pathogenesis are TcdA and TcdB, homologous glucosyltransferases capable of inactivating small GTPases within the host cell. We present crystal structures of the TcdA glucosyltransferase domain in the presence and absence of the co-substrate UDP-glucose. Although the enzymatic core is similar to that of TcdB, the proposed GTPase-binding surface differs significantly. We show that TcdA is comparable with TcdB in its modification of Rho family substrates and that, unlike TcdB, TcdA is also capable of modifying Rap family GTPases both in vitro and in cells. The glucosyltransferase activities of both toxins are reduced in the context of the holotoxin but can be restored with autoproteolytic activation and glucosyltransferase domain release. These studies highlight the importance of cellular activation in determining the array of substrates available to the toxins once delivered into the cell.     

Structural basis for the function of Clostridium difficile toxin B

Toxin B is a member of the family of large clostridial cytotoxins which are of great medical importance. Its catalytic fragment was crystallized in the presence of UDP-glucose and Mn2+. The structure was determined at 2.2 A resolution, showing that toxin B belongs to the glycosyltransferase type A family. However, toxin B contains as many as 309 residues in addition to the common chainfold, which most likely contribute to the target specificity. A superposition with other glycosyltransferases shows the expected positions of the acceptor oxygen atom during glucosyl transfer and indicates further that the reaction proceeds probably along a single-displacement pathway. The C1' donor carbon atom position is defined by the bound UDP and glucose. It assigns the surface area of toxin B that forms the interface to the target protein during the modifying reaction. A docking attempt brought the known acceptor atom, Thr37 O(gamma1) of the switch I region of the RhoA:GDP target structure, near the expected position. The relative orientation of the two proteins was consistent with both being attached to a membrane. Sequence comparisons between toxin B variants revealed that the highest exchange rate occurs around the active center at the putative docking interface, presumably due to a continuous hit-and-evasion struggle between Clostridia and their eukaryotic hosts.     

Mechanism of membrane translocation by anthrax toxin: insertion and pore formation by protective antigen

Proteolytic activation of receptor-bound protective antigen (PA) at the cell surface removes PA₂₀, allowing PA₆₃ to oligomerize and form a ring-shaped heptameric prepore. The prepore binds edema factor (EF) and lethal factor (LF) and, after endocytosis and trafficking of the complex to an acidic, vesicular compartment, it undergoes membrane insertion and mediates translocation of EF/LF to the cytosol. Data from membrane conductance experiments support a model of membrane insertion in which the 2β2–2β3 loop of PA, which is disordered in native PA and the prepore, forms a 14-stranded transmembrane β-barrel. Recent studies on the process of prepore-to-pore conversion and our current understanding of the mechanism of pH-dependent translocation will be described.     

Amino acid residues involved in membrane insertion and pore formation of Clostridium botulinum C2 toxin

The actin-ADP-ribosylating Clostridium botulinum C2 toxin consists of the enzymatic component C2I and the binding component C2II. C2II forms heptameric channels involved in translocation of the enzymatic component into the target cell. On the basis of the heptameric toxin channel, we studied functional consequences of mutagenesis of amino acid residues probably lining the lumen of the toxin channel. Substitution of glutamate-399 of C2II with alanine blocked channel formation and cytotoxicity of the holotoxin. Although cytotoxicity and rounding up of cells by C2I were completely blocked by exchange of phenylalanine-428 with alanine, the mutation increased potassium conductance caused by C2II in artificial membranes by about 2-3-fold over that of wild-type toxin. In contrast to its effects on single-channel potassium conductance in artificial membranes, the F428A mutation delayed the kinetics of pore formation in lipid vesicles and inhibited the activity of C2II in promoting (86)Rb (+) release from preloaded intact cells after pH shift of the medium. Moreover, F428A C2II exhibited delayed and diminished formation of C2II aggregates at low pH, indicating major changes of the biophysical properties of the toxin. The data indicate that phenylalanine-428 of C2II plays a major role in conformational changes occurring during pore formation of the binding component of C2II.     

Adaptation of Clostridium difficile toxin A for use as a protein translocation system

A cellular delivery system is a useful biotechnology tool, with many possible applications. Two derivatives of Clostridium difficile toxin A (TcdA) have been constructed (GFP-TcdA and Luc-TcdA), by fusing reporter genes to functional domains of TcdA, and evaluated for their ability to translocate their cargo into mammalian cells. The cysteine protease and receptor binding domains of TcdA have been examined and found to be functional when expressed in the chimeric construct. Whereas GFP failed to internalize in the context of the TcdA fusion, significant cellular luciferase activity was detected in vero cell lysates after treatment with Luc-TcdA. Treatment with bafilomycin A1, which inhibits endosomal acidification, traps the luciferase activity within endosomes. To further understand these results, clarified lysates were subjected to molecular weight sieving, demonstrating that active luciferase was released from Luc-TcdA after translocation and internal processing.     

Simultaneous detection of toxin A and toxin B genetic determinants of Clostridium difficile using the multiplex polymerase chain reaction.

A multiplex polymerase chain reaction was developed to simultaneously detect the presence of toxin A and toxin B genes of Clostridium difficile. A 1050-bp fragment of the toxin B gene and a 1217-bp fragment of the toxin A gene were amplified from 42 toxic strains of C. difficile; however, from 10 nontoxic strains the toxin gene fragments were not amplified; these data demonstrate that this multiplex polymerase chain reaction procedure can be used to differentiate between toxic and nontoxic strains. This sensitive and specific multiplex polymerase chain reaction for C. difficile toxins may prove to be a valuable diagnostic procedure.     

Bacteriocins: mechanism of membrane insertion and pore formation

Lactic acid bacteria produce several types of pore forming peptides. Class I bacteriocins are lantibiotics that contain (methyl)lanthionine residues that may form intramolecular thioether rings. These peptides generally have a broad spectrum of activity and form unstable pores. Class II bacteriocins are small, heat stable peptides mostly with a narrow spectrum of activity. Most bacteriocins interact with anionic lipids that are abundantly present in the membranes of Gram-positive bacteria.'Docking molecules' may enhance the conductivity and stability of lantibiotic pores, while'receptors' in the target membrane may determine specificity of class II bacteriocins. Insertion into the membrane of many bacteriocins is proton motive force driven. Lantibiotics may form pores according to a'wedge-like' model, while class II bacteriocins may enhance membrane permeability either by the formation of a'barrel stave' pore or by a'carpet' mechanism.     

Cellular stability of Rho-GTPases glucosylated by Clostridium difficile toxin B

Mono-glucosylation of Rho, Rac, and Cdc42 by Clostridium difficile toxin B (TcdB) induces changes of actin dynamics and apoptosis. When fibroblasts were treated with TcdB, an apparent decrease of the cellular Rac1 level was observed when applying anti-Rac1(Mab 102). This decrease was not based on degradation as inhibition of the proteasome by lactacystin did not stabilise cellular Rac1 levels. The application of anti-Rac1 (Mab 23A8) showed that the cellular Rac1 level slightly increased in TcdB-treated fibroblasts; thus, the apparent loss of cellular Rac1 was not due to degradation but due to impaired recognition of glucosylated Rac1 by anti-Rac1 (Mab 102). In contrast, recognition of RhoA by anti-RhoA (Mab 26C4) and Cdc42 by anti-Cdc42 (Mab 44) was not altered by glucosylation; a transient decrease of cellular RhoA and Cdc42 in TcdB-treated fibroblasts was indeed due to proteasomal degradation, as inhibition of the proteasome by lactacystin stabilised both cellular RhoA and Cdc42 levels. The finding that the apparent decrease of Rac1 reflects Rac1 glucosylation offers a valuable tool to determine Rac1 glucosylation.     

The role of toxin A and toxin B in the virulence of Clostridium difficile

During the past decade, there has been a striking increase in Clostridium difficile nosocomial infections worldwide predominantly due to the emergence of epidemic or hypervirulent isolates, leading to an increased research focus on this bacterium. Particular interest has surrounded the two large clostridial toxins encoded by most virulent isolates, known as toxin A and toxin B. Toxin A was thought to be the major virulence factor for many years; however, it is becoming increasingly evident that toxin B plays a much more important role than anticipated. It is clear that further experiments are required to accurately determine the relative roles of each toxin in disease, especially in more clinically relevant current epidemic isolates.     

Clostridium difficile toxin B induces autophagic cell death in colonocytes

Toxin B (TcdB) is a major pathogenic factor of Clostridum difficile. However, the mechanism by which TcdB exerts its cytotoxic action in host cells is still not completely known. Herein, we report for the first time that TcdB induced autophagic cell death in cultured human colonocytes. The induction of autophagy was demonstrated by the increased levels of LC3‐II, formation of LC3⁺ autophagosomes, accumulation of acidic vesicular organelles and reduced levels of the autophagic substrate p62/SQSTM1. TcdB‐induced autophagy was also accompanied by the repression of phosphoinositide 3‐kinase (PI3K)/Akt/mechanistic target of rapamycin (mTOR) complex 1 activity. Functionally, pharmacological inhibition of autophagy by wortmannin or chloroquine or knockdown of autophagy‐related genes Beclin 1, Atg5 and Atg7 attenuated TcdB‐induced cell death in colonocytes. Genetic ablation of Atg5, a gene required for autophagosome formation, also mitigated the cytotoxic effect of TcdB. In conclusion, our study demonstrated that autophagy serves as a pro‐death mechanism mediating the cytotoxic action of TcdB in colonocytes. This discovery suggested that blockade of autophagy might be a novel therapeutic strategy for C. difficile infection.     

Calcium and calmodulin in cellular intoxication with Clostridium difficile toxin B

In cultured human lung fibroblasts treated with Clostridium difficile toxin B, the development of the cytopathogenic effect was inhibited by the proton ionophore monensin but was not affected by some other ionophores. The calcium channel blockers verapamil and LaCl3 protected the cells against intoxication, as did the calmodulin antagonists trifluoperazine, amitriptyline, R 24571, and dansylcadaverine. Since these agents could not prevent intoxication when added after the toxin internalization was completed, we suggest that calmodulin and uptake of extracellular calcium are needed for the internalization but not for the cytosolic action of the toxin.     

Polyphosphate-mediated protection from cellular intoxication with Clostridium difficile toxin B

The influence of polyphosphorylated compounds on intoxication of human lung fibroblasts with Clostridium difficile toxin B was studied. ATP, as well as other nucleoside di-, tri-, and tetraphosphates, inorganic polyphosphates and polyphosphorylated sugars, caused a dose-dependent (1–5 mM range) delay in the appearance of the cytopathogenic effect. With a longer phosphate chain, the delay was more pronounced, although the cytopathogenic effect always developed finally, reaching the level of the control within 20 h. Toxin preparations contained one fraction of molecules able to bind ATP, besides one non-binding fraction. The protective effect of ATP did not depend on its energy producing ability. Neither was the protective effect due to an inactivation of the toxin per se, or to an interference with binding of the toxin to the cells. ATP was protective even upon addition 10 min after the toxin binding step. In the presence of ATP, the toxin remained accessible to neutralization with antitoxin. In analogy with the P-site on diphtheria toxin, we postulate that C. difficile toxin B contains a polyphosphate-binding site. This site is separate from the receptor-binding site, but involved in the interaction of toxin B with the cell surface shortly after the binding step.     

Clostridium difficile toxins A and B: Receptors,pores, and translocation into cells

The most potent toxins secreted by pathogenic bacteria contain enzymatic moieties that must reach the cytosol of target cells to exert their full toxicity. Toxins such as anthrax, diphtheria, and botulinum toxin all use three well-defined functional domains to intoxicate cells: a receptor-binding moiety that triggers endocytosis into acidified vesicles by binding to a specific host-cell receptor, a translocation domain that forms pores across the endosomal membrane in response to acidic pH, and an enzyme that translocates through these pores to catalytically inactivate an essential host cytosolic substrate. The homologous toxins A (TcdA) and Toxin B (TcdB) secreted by Clostridium difficile are large enzyme-containing toxins that for many years have eluded characterization. The cell-surface receptors for these toxins, the non-classical nature of the pores that they form in membranes, and mechanism of translocation have remained undefined, exacerbated, in part, by the lack of any structural information for the central ～1000 amino acid translocation domain. Recent advances in the identification of receptors for TcdB, high-resolution structural information for the translocation domain, and a model for the pore have begun to shed light on the mode-of-action of these toxins. Here, we will review TcdA/TcdB uptake and entry into mammalian cells, with focus on receptor binding, endocytosis, pore formation, and translocation. We will highlight how these toxins diverge from classical models of translocating toxins, and offer our perspective on key unanswered questions for TcdA/TcdB binding and entry into mammalian cells.     

Production, purification and characterization of Clostridium difficile toxic proteins different from toxin A and from toxin B

The purification and characterization of three new proteins called C1, C2, and C3 from Clostridium difficile are described. Their estimated molecular mass were about 350 (C1), 270 (C2) and 140 (C3) kDa, consisting of subunits of 39 (C1), 43 (C2) and 41 (C3) kDa, respectively. Immunodiffusion revealed that the three proteins contained similar but not identical antigenic determinants to toxin A. Each protein induced a cytotonic effect on hamster ovaric cells; the combined proteins, had a specific activity on cells 5-times higher than that of toxin A. In rat intestinal loops, they induced a clear fluid secretion, while toxin A elicited a haemorrhagic fluid response. The cytotonic activities of all three proteins were abolished by antiserum against toxin A, while antiserum against toxin B inhibited only the activity of the 270 kDa protein. In contrast to toxin A, the cytotoxicity of the three proteins was inactivated by trypsin. Thus, the chemical, antigenic and biological properties of these proteins differed from those of toxin A and toxin B.     

Phosphorylation of cellular proteins in response to treatment with Clostridium difficile toxin B and Clostridium sordellii toxin L.

Toxin B from Clostridium difficile induces typical morphological changes of cultured cells consisting of rounding up and arborization, which are associated with a dramatic disruption of microfilaments. In this study, we show that toxin L, a cytotoxin produced by bacterial strain Clostridium sordellii, has similar effects on cultured cells including the redistribution of F-actin and of the adhesion plaque protein vinculin. It has been assumed that the mechanisms involved in cytopathic effects of toxin B are related to the function of an unidentified component that regulates the organization of the actin cytoskeleton. We demonstrate that the treatment of cultured astrocytes with toxin B or toxin L alters the incorporation of inorganic phosphate into several proteins. Immunoblot analysis revealed that one of these proteins is tropomyosin. Since tropomyosin stabilizes microfilaments and inhibits the severing activity of gelsolin, the toxin-induced phosphorylation may counteract this inhibition resulting in severing of microfilaments and capping of short filaments. A decrease in the radioactivity associated with intermediate filament protein vimentin was also detected using a monoclonal antibody which specifically recognizes a phosphorylated epitope of vimentin. Since vimentin is an in vivo substrate for various protein kinases, these data are in favor of broad effects of these toxins. Direct measurement of protein kinase C in cells exposed to toxin B or to toxin L did not reveal a significant change in protein kinase C activity. Furthermore, treatments with toxins do not increase cAMP levels, suggesting that toxins do not activate protein kinase A. Although further studies are required to determine the primary target site for the clostridial cytotoxin B and L, our results show that they provoke the alteration in the phosphorylation of cellular proteins.     

Anthrax toxin receptor 2 determinants that dictate the pH threshold of toxin pore formation

The anthrax toxin receptors, ANTXR1 and ANTXR2, act as molecular clamps to prevent the protective antigen (PA) toxin subunit from forming pores until exposure to low pH. PA forms pores at pH approximately 6.0 or below when it is bound to ANTXR1, but only at pH approximately 5.0 or below when it is bound to ANTXR2. Here, structure-based mutagenesis was used to identify non-conserved ANTXR2 residues responsible for this striking 1.0 pH unit difference in pH threshold. Residues conserved between ANTXR2 and ANTXR1 that influence the ANTXR2-associated pH threshold of pore formation were also identified. All of these residues contact either PA domain 2 or the neighboring edge of PA domain 4. These results provide genetic evidence for receptor release of these regions of PA as being necessary for the protein rearrangements that accompany anthrax toxin pore formation.     

Purification and characterisation of two forms of toxin B produced by Clostridium difficile

Toxin B from Clostridium difficile was purified to homogeneity by gel filtration and high resolution ion exchange chromatography. Two forms of toxin B were found. Form 1 which seemed to consist of two identical subunits of 220-300 kDa; femtogram amounts of this toxin induced rounding of fibroblast cells. Form 2 contained subunits of 43 kDa and 105 kDa; the stoichiometric ratio probably being 4:1; picogram amounts were needed to induce rounding of fibroblast cells. Immunological studies suggested that both subunit types were antigenic and had epitopes which were identical with those of form 1.