Bacterial glycosyltransferase toxins - Wiley Online Library

Cellular Microbiology (2015) 17(12), 1752–1765

doi:10.1111/cmi.12533 First published online 4 November 2015

Microreview Bacterial glycosyltransferase toxins Thomas Jank,1 Yury Belyi2,3 and Klaus Aktories1,3* 1 Institute for Experimental and Clinical Pharmacology and Toxicology, Albert-Ludwigs University of Freiburg, Freiburg, Germany. 2 Gamaleya Research Institute, Moscow, 123098, Russia. 3 Freiburg Institute for Advanced Studies (FRIAS), AlbertLudwigs University of Freiburg, Freiburg, Germany. Summary Mono-glycosylation of host proteins is a common mechanism by which bacterial protein toxins manipulate cellular functions of eukaryotic target host cells. Prototypic for this group of glycosyltransferase toxins are Clostridium difficile toxins A and B, which modify guanine nucleotide-binding proteins of the Rho family. However, toxin-induced glycosylation is not restricted to the Clostridia. Various types of bacterial pathogens including Escherichia coli, Yersinia, Photorhabdus and Legionella species produce glycosyltransferase toxins. Recent studies discovered novel unexpected variations in host protein targets and amino acid acceptors of toxin-catalysed glycosylation. These findings open new perspectives in toxin as well as in carbohydrate research.

Introduction Bacterial protein toxins are major virulence factors, which in many cases are crucial for host–pathogen interaction and infection. Recent studies have shown that the spectrum of mechanisms by which bacterial protein toxins attack host cells is amazingly broad (Just et al., 1995a; Cui et al., 2010; Aktories, 2011; Goody et al., 2012; Chen et al., 2013; Jank et al., 2013; Li et al., 2013; Pearson et al., 2013; Young et al., 2014). While membrane-damaging or pore-forming toxins can directly reach their target (Parker and Feil, 2005), many bacterial protein toxins have their main targets inside host cells. Therefore, sophisticated delivery systems were Received 30 July, 2015; revised 5 October, 2015; accepted 6 October, 2015. *For correspondence. E-mail klaus.aktories@ pharmakol.uni-freiburg.de; Tel. +49-761-2035301; Fax +49-7613035311.

developed, allowing membrane translocation of protein toxins into target cells (Sandvig et al., 2004; Trujillo et al., 2006). These delivery systems are integrated parts of the toxin molecule (e.g. AB-toxins) or complex cell puncturing devices, which in general depend on the direct contact of the bacteria with host cells (e.g. type III, type IV and type VI secretion systems) (Barison et al., 2013; Galan et al., 2014; Sarris et al., 2014; Chandran and Waksman, 2015). In the latter cases, the translocated toxic proteins are often assigned as ‘effectors’ as compared with exotoxins, which are released from the bacteria into the environment. Many bacterial protein toxins and ‘effectors’ (here, termed ‘toxins’) exhibit enzyme activities to manipulate host cell functions by diverse post-translational modifications (Aktories, 2011). ADP-ribosylation was one of the first mechanisms identified to be used by bacterial toxins (Honjo et al., 1968; Deng and Barbieri, 2008; Simon et al., 2014). Other common mechanisms are N-glycosidation (Endo et al., 1988), adenylylation (AMPylation) (Worby et al., 2009; Yarbrough et al., 2009), proteolytic cleavage (Schiavo et al., 2000), deamidation (Flatau et al., 1997; Schmidt et al., 1997; Orth et al., 2009) and glycosylation (Just et al., 1995a; Belyi et al., 2006; Jank et al., 2013; Li et al., 2013). Especially exciting is the rapidly growing field of glycosyltransferase toxins. Clostridium difficile toxins A (TcdA) and B (TcdB), which modify Rho family proteins by glucosylation, are the prototypes of this family (Just et al., 1995a). Related glycosyltransferase toxins have been identified in Legionella, Escherichia coli, Photorhabdus and Yersinia species. In this review, we will discuss the several types of glycosyltransferase toxins. The main focus will be on the molecular mechanisms and the functional consequences of glycosylation of host cell proteins by bacterial protein toxins. Clostridial glycosylating toxins TcdA and TcdB, which are prototypic for clostridial glycosylating toxins, are the major virulence factors of C. difficile. This pathogen causes antibiotic-associated diarrhoea and pseudomembranous enterocolitis (Just and Gerhard, 2004; Voth and Ballard, 2005). C. difficile infection is often the consequence of antibiotic treatment that alters the intestinal microbiome, allowing colonization of the pathogen and production of toxins. The toxins (TcdA and TcdB) cause inflammation and damage of the gut mucosa mainly by glucosylation of Rho proteins, resulting

© 2015 John Wiley & Sons Ltd

cellular microbiology

Glycosyltransferase toxins in diarrhoea and/or colitis (Voth and Ballard, 2005). After years of discussion whether TcdA or TcdB is more important for induction of diseases (Lyras et al., 2009; Kuehne et al., 2010), recent experimental evidence points to a most prominent role of TcdB as a virulence factor of C. difficile. This view is supported by studies of isogenic C. difficile toxin mutants of an epidemic clinical isolate, showing that the C. difficile mutant, which produces only TcdB, causes comparable symptoms as the wild type strain and more severe disease than the solely TcdA-producing mutant (Carter et al., 2015). However, the precise pathogenetic pathways involved in disease are still not completely understood (see succeeding text). Besides the epidemic isolate, several other clinical relevant C. difficile strains produce toxins that differ in sugar donor and protein substrate specificities and induce different types of diseases (Just and Gerhard, 2004). The ABCD model of the toxins and their up-take TcdA and TcdB are constructed according the ABCD model (Fig. 1A) (Jank and Aktories, 2008). The N-terminal

1753

region A (A for activity) harbours the glucosyltransferase domain (GTD) (Hofmann et al., 1997). The C-terminal region B (B for binding) is suggested to be involved in receptor binding (Von Eichel-Streiber et al., 1992). Region C (C for cutting) represents the protease domain (Egerer et al., 2007), which is involved in auto-catalytic cleavage and processing of the toxins. Finally, region D (D for delivery) is most likely involved in translocation of the toxin into the cytosol (Genisyuerek et al., 2011). This part might be also responsible for toxin binding to target cells (Genisyuerek et al., 2011; Olling et al., 2011; Yuan et al., 2015). Several steps are involved in toxin up-take (Fig. 1B). The first step is the binding of TcdA and TcdB to target cells involving the C-terminal B domain, which adopts a beta-solenoid structure suggested to bind carbohydrates (Greco et al., 2006). Recent data indicate that not only the C-terminus but also part of the delivery domain may be involved in toxin membrane binding (Olling et al., 2011; Schorch et al., 2014). For TcdB, it was reported that chondroitin sulfate proteoglycan-4 (Yuan et al., 2015)

Fig. 1. Domain arrangement of glycosyltransferase toxins and up-take of C. difficile toxin B. A. Architecture of C. difficile Toxin B (TcdB), the prototype of clostridial glycosylating toxins, according the ABCD-domain organization (Activity, Binding, Cutting and Delivery) in comparison with the glycosylating toxins PaTox (Photorhabdus asymbiotica Toxin), Afp18 (Antifeeding prophage 18 effector from Yersinia ruckeri), Lgt1 and SetA (Legionella pneumophila glycosyltransferases), NleB (EPEC effector). B. Schematic of the cellular uptake of toxin A and B (TcdA, TcdB). The toxins released by the bacteria bind to their host cell surface receptors and are taken up by clathrin-mediated endocytosis. Upon acidification of endosomes, the toxins alter their structural conformation and insert into the endosomal membrane, which results in pore formation and translocation of the N-terminal part of the toxin. Inositol hexakisphosphate (InsP6) triggers the autocatalytic cleavage of the toxins by activation of the protease domain and releases the glycosyltransferase domain (GTD) into the cytoplasm. Here, the GTD is targeted to the inner leaflet of the plasma membrane by the N-terminal four helix bundle where the substrate Rho is located. By using host UDP-glucose as sugar donor, the GTD glycosylates threonine-35/37 of Rho GTPases. This results in impaired downstream signalling and, finally, in the pathogenic effects seen in C. difficile infections. GTD, glycosyltransferase domain; APD, autoprocessing domain; CROPs, combined repetitive oligopeptides; D, deamidase; PI(3)P-binding, phosphatidylinositol-3-phosphate binding domain; DXD, aspartate-X-aspartate-motif. © 2015 John Wiley & Sons Ltd, Cellular Microbiology, 17, 1752–1765

1754

T. Jank, Y. Belyi and K. Aktories

and/or poliovirus receptor-like 3 protein (LaFrance et al., 2015) might interact with TcdB in a region other than the C-terminal combined repetitive oligopeptides domain. Toxin binding is followed by clathrin-dependent endocytosis and traffic to acidic early endosomes (Barth et al., 2001; Papatheodorou et al., 2010). Here, the low pH may induce structural rearrangements allowing the insertion of hydrophobic areas of the toxins into endosomal membranes and translocation of at least parts of the toxins (domains A and C) into the cytosol (Barth et al., 2001). This translocation is not well understood but may involve pore formation. The autoprocessing protease domain is activated by inositol hexakisphosphate and then cleaves and releases the GTD into the cytoplasm of the host cell (Egerer et al., 2007; Reineke et al., 2007). Expression of a single-domain heavy chain antibody, which inhibits autoprocessing, in target cells, indicates that autoprocessing occurs in the cytosol and not in endosomal compartments (Li et al., 2015). Eventually, the GTD localizes to the cell membrane, where Rho proteins are glucosylated. Inactivation of Rho proteins by glucosylation Rho proteins comprise a group of ~20 highly related GTP-binding proteins, which act as molecular switches and belong to the superfamily of Ras proteins (Burridge and Wennerberg, 2004; Jaffe and Hall, 2005). They are master regulators of the actin cytoskeleton and control motile functions, cell morphology and polarity (Ridley and Hall, 1992; Hall, 1998; Nobes and Hall, 1999). They are involved in regulation of proliferation, cell cycle progression and cell division. Important for host–pathogen interactions are their crucial roles in regulation of epithelial barrier functions, phagocytosis, superoxide

anion production, migration, secretion and immune cell signalling (Caron and Hall, 1998; Aktories, 2011; Lemichez and Aktories, 2013; Popoff, 2014). The activity of Rho proteins is regulated by a GTPase cycle. In the GDP-bound form, the proteins are inactive. GTP/GDP exchange, induced by guanine nucleotide exchange factors (GEFs), activates Rho proteins. Then, they interact with a large spectrum of effector proteins like protein kinases, phospholipases or adaptor proteins. Hydrolysis of GTP, which is facilitated by GTPaseactivating proteins (GAPs), terminates the active state (Jaffe and Hall, 2005). Rho proteins are isoprenylated, which favours membrane binding of the GTP-bound form of Rho proteins. On the other hand, inactive GDP-bound Rho is associated with GDI proteins (guanine nucleotide dissociation inhibitors) and localized in the cytosol. Several members of the Rho protein family are glucosylated by TcdA and TcdB (Just et al., 1995b; Just et al., 1995a). Well known is the modification of RhoA, B, C, Rac1, 2 and Cdc42. Other substrates are given in Table 1. All these Rho proteins are modified at a highly conserved threonine (threonine-37 in RhoA, B and C and threonine-35 in other Rho GTPases) (Just and Gerhard, 2004). This threonine residue is essential for nucleotide binding and located in the effector region (switch-1 region) of Rho (Jaffe and Hall, 2005). Glucosylation of Rho proteins inhibits interaction with effectors and thereby blocks Rho-dependent signalling (Sehr et al., 1998). Glucosylation also inhibits the actions of GAPs and GEFs on Rho. Moreover, glucosylated Rho sticks to cell membranes, because the interaction to GDIs and the extraction from the cell membranes is impaired (Genth et al., 1999). Inhibition of Rho–effector interaction by glucosylation is probably most significant for the toxin actions on cells, which is characterized by arborization

Table 1. Glycosyltransferase toxins. Toxin

Organism

Donor substrate

Targets

TcdA (toxin A) TcdB (toxin B) TcsL (lethal toxin) TcsH (haemorrhagic toxin) TpeL

Clostridium difficilea C. difficilea C. sordelliib C. sordelliib

UDP-Glc UDP-Glc UDP-Glc UDP-Glc

C. perfringensc

TcnA (alpha toxin) PaTox Afp18 NleB Lgt1, 2, 3 SetA

C. novyid Photorhabdus asymbiotica Yersinia ruckeri EPEC Legionella pneumophila L. pneumophila

UDP-GlcNAc/ (UDP-Glc)e UDP-GlcNAc UDP-GlcNAc UDP-GlcNAc UDP-GlcNAc UDP-Glc UDP-Glc

RhoA/B/C, Rac1, Cdc42, (TC10, Rap1, Rap2A, Ral, Ras, RhoG)e RhoA/B/C, Rac1, Cdc42, TC10, RhoG, (Rap1/2, Ral, Ras)e Ras, Rac, RhoG, TC10, Rap, Ral, (Cdc42)e RhoA/B/C, Rac1, Cdc42, TC10, TCL, H-Ras, N-Ras, K-Ras, Rap2A, RalC H-, K-, N-, R-Ras, Rap1B, Rap2A, Ral, (Rac1/2/3)e Rho, Rac, Cdc42 RhoA/B/C, Rac1/2/3, Cdc42 RhoA Death domain containing proteins: TRADD, FADD, RIPK1 eEF1A Unknown

a.

Strain VPI 10463. Strain VPI 9048. c. Strain JGS1495. d. Strain 19402. e. Minor substrate. b.

© 2015 John Wiley & Sons Ltd, Cellular Microbiology, 17, 1752–1765

Glycosyltransferase toxins and rounding up of cells with loss of cell–cell contacts (Fiorentini and Thelestam, 1991). Because Rho proteins are involved in many signalling processes, the inhibitory effects of the toxins are manyfold (Just and Gerhard, 2004; Voth and Ballard, 2005). Role of glycosylation in pathogenesis of C. difficile infection Clostridium difficile infection is characterized by diarrhoea and, in more severe cases, by pseudomembranous enterocolitis, which may be lethal. Symptoms of C. difficile infections depend on the presence of the glycosyltransferase toxins TcdA and TcdB. How can glycosylation of Rho proteins induce disease? Two major pathogenetic mechanisms are important (Fig. 2). First, as discussed previously, glucosylation of Rho proteins inhibits their functions as master regulators of the actin cytoskeleton. Therefore, multiple F-actin-dependent processes are disturbed. These include damage of tight and adherence junctions and loss of cell–cell contacts, resulting in increased epithelial permeability, which is probably the cause of diarrhoea. Moreover, reduced cell adherence causes apoptosis and cell loss. However, epithelial cell renewal is limited, and cell proliferation is inhibited, because glucosylated Rho blocks cell cycle progression and actin-dependent cytokinesis. The second major pathogenetic mechanism is the activation of the inflammasome by glycosylated RhoA, which is probably the cause of inflammation and colitis induced by C. difficile (Ng et al., 2010). While glycosylated RhoA exhibits ‘loss of function’ in regulation of the actin cytoskeleton, it exhibits ‘gain of function’ as an activator of the inflammasome. Feng Shao and coworkers identified Pyrin encoded by the mediterranean fever gene MEFV as a sensor for glycosylated RhoA (but also for RhoA modified in the switch-1 region by ADP-ribosylation, AMPylation or deamidation) (Xu et al., 2014). Pyrin interacts with ASC (apoptosisassociated speck-like protein), which is an adaptor protein that recruits and activates pro-caspase 1 (Lu and Wu, 2015). Caspase-1 is a central regulator of the innate immune defence and activates IL-1β and IL-18. These cytokines cause (among other effects) release of IL-8 and interferon-γ (IFN-γ) respectively. Increase in IL-8 and INF-γ induced by C. difficile toxins has been frequently reported (Linevsky et al., 1997; Warny et al., 2000; Ishida et al., 2004; Jafari et al., 2013). IL-8 is one of the most potent attractants for neutrophils, which explains the strong neutrophil invasion into colon mucosa in the course of C. difficile infection and is probably responsible for mucosal damage but also for control of infection. Less well understood are toxin effects, which depend on the presence of the glycosyltransferase domain of TcdA and TcdB but not on its transferase activity (Farrow et al., 2013; Wohlan et al., 2014). Reactive oxygen production induced © 2015 John Wiley & Sons Ltd, Cellular Microbiology, 17, 1752–1765

1755

by the toxins (independent of its glycosyltransferase activity) has been suggested (Farrow et al., 2013). These effects appear to be Rac dependent (Farrow et al., 2013; Wohlan et al., 2014). Other clostridial glycosylating toxins Various clostridial species produce glycosyltransferase toxins related to TcdA and TcdB. Clostridium sordellii, which causes myonecrosis (Aldape et al., 2006) and toxic shock syndrome (Fischer et al., 2005), produces lethal (TcsL) (Just et al., 1996) and haemorrhagic toxin (TcsH) (Genth et al., 2014). Clostridium novyi, which induces myonecrosis (Samlaska and Maggio, 1996) and soft tissue infections of injecting drug users (Brett et al., 2005), produces α-toxin (TcnA) (Selzer et al., 1996). The latest described member of the toxin group is Clostridium perfringens toxin TpeL (Amimoto et al., 2007; Nagahama et al., 2010), which does not contain the C-terminal binding domain (B or combined repetitive oligopeptide domain). All these toxins share Rho/Ras proteins as eukaryotic targets but differ slightly in substrate specificity (Table 1). However, in any case, modification of Ras/Rho proteins occurs at the same amino acid, namely, threonine-35/37. Also the sugar donor differs between the toxins with uridine diphosphate (UDP)-glucose (UDPGlc) for TcdA, TcdB, TcsL, TcsH and UDP-Nacetylglucosamine (UDP-GlcNAc) for TcnA and TpeL (Table 1) (Selzer et al., 1996; Nagahama et al., 2010; Guttenberg et al., 2012). The latter accepts also UDPglucose to minor extent. In general, these toxins follow the typical ABCD architecture (Fig. 1A). Structural aspects of clostridial glycosylating toxins The three-dimensional structure of the GTD of TcdB (Fig. 3A) (Reinert et al., 2005) reveals a GT-A-like glycosyltransferase listed as glucosyltransferase family 44 (CAZY, http://www.cazy.org/). This family comprises more than 300 genes. The enzyme domain (543 amino acids) of TcdB consists of 11 β-strands and 21 α-helices. The catalytic core is formed by 234 amino acids, resembling a Rossmann-like fold consisting of an αβα sandwich of a seven-stranded β-sheet. The additional 309 resides mainly form α-helices. Functions of the subdomains are not clear with the exception of the 4helical bundle at the N-terminus, which is involved in intracellular membrane binding (Fig. 3A). This subdomain is conserved in various bacterial proteins and protein toxins (e.g. the C1 domain in Pasteurella multocida toxin PMT) unrelated to clostridial glycosyltransferases (Kitadokoro et al., 2007; Geissler et al., 2010). The hallmark of GT-A type of glycosyltransferases is the DxD motif, in the case of TcdB D286-x-D288 (Fig. 3D). The DxD motif is involved in UDP binding and Mn2+

1756


Fig. 2. Pathophysiological consequences of glycosylation of Rho proteins. Rho, Rac and Cdc42 proteins are glycosylated by C. difficile toxins A (TcdA) and B (TcdB). Thereby, the Rho proteins are inhibited in their functions to regulate the actin cytoskeleton with inhibition of motile functions of cells, blockade of phagocytosis and secretion. Cell–cell contacts and cell adherence is disturbed resulting in increase in epithelial permeability, cells loss and apoptosis. Cell renewal is blocked by inhibition of proliferation, cytokinesis and cell cycle progression. Glycosylated RhoA mediates activation of the inflammasome via Pyrin and adaptor ASC. How glycosylated RhoA is sensed by Pyrin is not clear. Changes in the actin cytoskeleton may play a role (Xu et al., 2014). ASC activates caspase-1, which activates IL-1β and IL-18 and may cause pyroptosis. IL-1β releases IL-8 (and other cytokines, e.g. IL-6), which is a strong attractor of neutrophils and IL-18 causes release of IFN-γ.

coordination. D286 binds the 3′hydroxyl group of ribose of UDP and the 3′hydroxyl group of glucose. In addition, it is involved in coordination of Mn2+ via a water molecule. The other crucial asparagine (D288) directly coordinates the divalent cation. Tryptophan-102 (WU) stabilizes the uracil ring by aromatic stacking. Furthermore, essential for the glucosylation reaction is tryptophan-520 (Wflex). Determination of the crystal structures of clostridial glycosyltransferases in the apo form (without a sugar donor) and with hydrolysed or uncleaved UDP-glucose (Ziegler et al., 2008) revealed that tryptophan-520 undergoes a dramatic conformational change, depending on the presence of UDP-glucose (Fig. 3D). Thus, an open and a closed state of the enzyme is defined by the positioning of W520. The catalytic core of the transferase with typical essential residues is highly conserved among clostridial glycosylating toxins (Ziegler et al., 2008; Pruitt et al., 2012). However, these pivotal residues are also found in other glycosylating toxins (Fig. 3B and C) (Hurtado-Guerrero et al., 2010; Lu et al., 2010; Jank et al., 2013). Thus, sequence comparison of the catalytic core of TcdB with other bacterial effectors and toxins allowed identification of new glycosylating toxins from the genus Photorhabdus, Yersinia and Legionella.

Tyrosine glycosylation of Rho proteins by Photorhabdus asymbiotica and Yersinia ruckeri toxins PaTox, a GlcNAc transferase from Photorhabdus asymbiotica Bacteria from the genus Photorhabdus are symbionts in the intestinal tract of entomopathogenic nematodes (Forst et al., 1997; ffrench-Constant and Bowen, 2000). The nematodes invade larvae and kill the insects by toxins produced by Photorhabdus bacteria, which are released from the nematodes. Killed insects are then an optimal environment and food source for proliferation of nematodes and bacteria. Recently, it was shown that PaTox, a toxin from P. asymbiotica (name reflects earlier suggestion of a non-symbiosis life style), is lethal for Galleria mellonella larvae and causes destruction of the actin cytoskeleton in cell culture (Jank et al., 2013). The toxin has a mass of ~335 kDa and possesses a glycosyltransferase domain between residues 2115 and 2449, which shares sequence similarity with clostridial glycosylating toxins (Fig. 1A). PaTox also modifies Rho proteins by glycosylation (Jank et al., 2013). The glucosyltransferase uses UDP-GlcNAc as a sugar donor and modifies Rho, Rac and Cdc42. Mass © 2015 John Wiley & Sons Ltd, Cellular Microbiology, 17, 1752–1765

Glycosyltransferase toxins

1757

Fig. 3. Structural comparison of bacterial glycosyltransferase toxins. The glycosyltransferase domains of the following are shown: A. C. difficile toxin B (TcdB) (PDB 2BVL). B. P. asymbiotica PaTox (PDB 4MIX). C. Legionella glycosyltransferase Lgt1 (PDB 3JSZ). The central Rossman domains defining the glycosyltransferase type A (GT-A)-fold are shown in blue. Protruding domains involved in plasma membrane interaction or substrate recognition are depicted in red. D. Structural alignment of the catalytic core of TcdB, PaTox and Lgt1 highlights the spatial conservation of amino acid residues. The DxD-motif and 2+ U the sugar binding GT-A triad (amino acids are labelled in magenta) interact with the divalent cation (M ) and the sugar respectively. Tryptophane U flex flex (W ) stacks to the uracil ring of UDP and Tryptophane (W ), located on the flexible loop, interacts with the β-phosphate and flips outwards during product release. UDP-glucose is shown from Lgt1. E. Schematic of the effects of P. asymbiotica PaTox and Y. ruckeri Afp18. PaTox and Afp18 modify Rho at switch-1 tyrosine-34 by GlcNAcylation and thereby prevent the interaction to GEFs, GAPs and downstream effectors. PaTox modifies RhoA, Rac and Cdc42, whereas Afp18 seems only to modify RhoA. The consequences of glycosylation by PaTox is the blockade of actin-driven processes as phagocytosis and cytoskeleton stabilization, which results in melanization and death of insect larvae. Glycosylation of RhoA by Afp18 prevents actin-mediated cytokinesis and gastrulation and blocks zebrafish embryo development.

spectrometric analyses reveal that Rho proteins are modified at tyrosine-32/34 but not at threonine-35/37 as known for clostridial glucosylating toxins (Fig. 3E) (Jank et al., 2013). GlcNAcylation of RhoA in tyrosine-34 is highly specific for this amino acid. Change of tyrosine-34 to threonine prevents modification by PaTox. RhoA modified by TcdB at threonine-37 is still glycosylatable by PaTox at tyrosine-34. Modification of a tyrosine residue is unique within the families of glycosyltransferases with the exception of glycogenin, which starts glycogen synthesis with an auto-glycosylation at a tyrosine residue (Smythe et al., 1988). RhoA tyrosine-34 is located in the switch-1 region, and GlcNAcylation of this residue results in inhibition of Rho protein interaction with effectors, GAPs and GEFs, while GDP- or GTP-binding per se is not affected. © 2015 John Wiley & Sons Ltd, Cellular Microbiology, 17, 1752–1765

The crystal structure of the GTD of PaTox exhibits a very small GT-A family enzyme with a Rossman-like fold, consisting of a central six-stranded β-sheet interposed by α-helices on both sides (Fig. 3B) (Jank et al., 2013). Typical for GT-A transferases, GTD of PaTox contains a DxD motif, involved in Mn2+ coordination, which is extended to a DxDD motif. An exchange of these residues to asparagine (D2276N, D2278N or D2279N) strongly inhibits transferase activity. Residue 2170 is the conserved tryptophan (WU) that is involved in aromatic stacking of the uracil moiety of UDP-GlcNAc. Further conserved residues involved in sugar donor binding are D2260 and R2263. Equivalent residues are present in clostridial glycosyltransferases (Fig. 3D, labelled in magenta). Nuclear magnetic resonance spectroscopy of

1758


GlcNAcylated RhoA shows that GTD of PaTox (like clostridial glycosylating toxins) is a retaining glycosyltransferase, which attaches GlcNAc onto Rho in an α-anomeric configuration (Jank et al., 2013). Again similar to clostridial glycosylating toxins, the GTD of PaTox localizes to the negatively charged inner leaflet of the cell membrane. This localization is guided by a polybasic region, consisting of R2134, K2135, R2139, K2140 in helix-1 at the N-terminus of the GTD (Jank et al., 2015). In silico comparison of the distances of the GTD of PaTox and the GTD of TcdB from the cell membrane surface to the catalytic sites of the toxins reveals an amazing similarity, suggesting the same geometry for the interaction with Rho proteins. In contrast to TcdB and TcdA, which prefer GDP-bound RhoA, PaTox needs the active GTP-bound form of RhoA for efficient glycosylation (Just et al., 1995a). It was a surprising finding that RhoA can be activated by a deamidase domain of PaTox, which is located downstream of the glycosyltransferase domain. However, activation of RhoA was only observed when the GTD (Rho inactivating!) was inhibited by mutation of the DxDD motif (Jank et al., 2013). PaTox-induced Rho activation is indirect and follows the activation of heterotrimeric Gq/i proteins caused by the deamidase domain. This domain shares structural similarity with Salmonella enterica SPI-2 effector SseI (Bhaskaran and Stebbins, 2012) and Pasteurella multocida toxin PMT, a well-known deamidase that also indirectly activates Rho via activation of heterotrimeric G proteins (Orth et al., 2009). Thus, it appears that activities for the activation and the inhibition of Rho proteins are localized in the same protein toxin. The pathophysiological role of PaTox is not clear. Especially, the timing of activation and inactivation of Rho proteins by the toxin remains to be clarified. PaTox kills insect larvae (Jank et al., 2013), which might be explained by the essential role of Rho proteins. However, one has to consider that Photorhabdus species produce a large array of different types of toxins. Rho proteins are not only inactivated but also activated by Photorhabdus toxins. For example, Photorhabdus luminescens produces the Tc toxin PTC5. This tripartite toxin contains the TccC5 component, which activates Rho proteins by ADPribosylation (Lang et al., 2010). Insects possess highly effective innate immunity, and Rho proteins are essentially involved in defence mechanisms like phagocytosis by hemocytes, melanization and encapsulation (Bidla et al., 2007; Fauvarque and Williams, 2011). Afp18, a glycosyltransferase from Yersinia ruckeri Yersinia ruckeri is a fish pathogen that causes enteric redmouth disease (Rucker, 1966), an infection mainly of salmonid fish species (e.g. rainbow trout), which is accompanied by severe economic losses in fish farm industries. A major virulence factors appear to be the

antifeeding prophage (Afp) complex, which is related to phage tail-derived type VI secretion systems (Coulthurst, 2013). Particularly high similarity exists between Y. ruckeri Afp complex and the Afp complex of Serratia entomophila (Heymann et al., 2013). The structures of Afps, which comprise 18 proteins, are similar to constituents of prophages with a contractile sheath, an inner tube, baseplate components and tail fibres, but miss a phage head component. The toxic effector of the syringe-like injection machine, which is delivered into target cells, is apparently Afp18. Afp18 is a glycosyltransferase, which GlcNAcylates specifically RhoA. Similar to PaTox, Afp18 modifies RhoA at tyrosine-34, resulting in inhibition of RhoA signalling. Crystal structure analyses of RhoA, which was GlcNAcylated by Afp18 at tyrosine-34, revealed a specific conformation of the switch regions dissimilar to typical GDP or GTP-bound forms of the GTPase. As known for PaTox and also for clostridial glucosylating toxins, modification of Rho by Afp18 results in attachment of the sugar (GlcNAc or glucose) in an αanomeric configuration, which is important for the stability of the modification (see succeeding text). Afp18 has been studied in developing zebrafish embryos, where the bacterial effector causes inhibition of cytokinesis, actindependent motility and cell blebbing, eventually, resulting in blockade of gastrulation (Fig. 3E). All these effects are mainly caused by a blockade of RhoA function, while Rac and Cdc42 play no or only a minor role. Glucosylating Legionella pneumophila toxins Various Legionella species contain toxins/effectors, which modify eukaryotic target protein by glycosylation (Belyi et al., 2003; Belyi et al., 2006; Jank et al., 2012; Belyi et al., 2013). Such an enzyme is Lgt1 that causes glucosylation of eukaryotic elongation factor 1A. While the natural host of Legionella pneumophila is amoeba, the Gram-negative pathogen can infect mammalian host cells (predominantly lung epithelial cells, alveolar macrophages and monocytes) to cause Legionnaires’ disease, a severe type of pneumonia (Swanson and Hammer, 2000; Fields et al., 2002). Legionella depends on intracellular proliferation in a specific Legionella-containing vacuole (Horwitz and Silverstein, 1980; Bitar et al., 2004). It is suggested that Legionella produces more than 300 effectors, which are injected into the cytosol of target cells by the type IV secretion system Icm/Dot (Ensminger and Isberg, 2009; Isberg et al., 2009). Among these are glucosyltransferases Lgt 1, 2 and 3 (Belyi et al., 2008). Lgt1 has a molecular mass of ~60 kDa and consists of 525 amino acids (Belyi et al., 2003) (Fig. 1A). While the primary amino acid sequence shares little similarity with other proteins, a small central region around a DxD motif (D246-X-D248 of Lgt1) is characterized by conserved © 2015 John Wiley & Sons Ltd, Cellular Microbiology, 17, 1752–1765

Glycosyltransferase toxins amino acids also found in typical glycosyltransferase toxins (Fig. 3D). The crystal structure of Lgt1 reveals three domains (Fig. 3C). The N-terminal domain consists of seven α-helices and is unrelated to any glycosyltransferase or toxin (Hurtado-Guerrero et al., 2010; Lu et al., 2010). The middle domain represents the glycosyltransferase core with the conserved DxD motif and a sugar recognition (GT-A) triad consisting of aspartate-230, arginine-233 and aspartate-246 (Fig. 3D, labelled in magenta). This sugar recognition triad is also conserved in other glycosyltransferase toxins. Similar as for clostridial glycosyltransferases, the uracil of UDPglucose stacks against the indole moiety of tryptophan-139 (WU). And again, as in clostridial glycosylating toxins, a loop with a tryptophan residue (W520, Wflex) covers and closes the catalytic cleft. The third domain, which mainly consists of α-helices, is inserted at the end of the glycosyltransferase domain and is formed by residues 323–444. Modification of elongation factor 1A Lgt1 modifies elongation factor 1A (eEF1A) by using UDPglucose as sugar donor (Belyi et al., 2006). eEF1A is a highly abundant, conserved protein that is essential for protein syntheses (Ramakrishnan, 2002). Thus, the functional consequence of Lgt1-catalysed glucosylation is protein synthesis inhibition (Fig. 4A) (Belyi et al., 2006; Belyi et al., 2012). eEF1A is a GTP-binding protein and responsible for the delivery of aminoacyl tRNA (aa-tRNA) to the A-site of mRNA-charged ribosomes (Fig. 4A). Here, eEF1A is released from aa-tRNA on hydrolysis of GTP. The cycle can be repeated after GTP/GDP exchange catalysed by the nucleotide exchange factor elongation factor 1Bα (Achenbach and Nierhaus, 2013). eEF1A consists of three domains (Fig. 4A, right insert). Domain 1 is a Ras-like GTP-binding domain (G-domain). Domains 2 and 3 consist of β-strands and comprise beta barrel structures (Andersen et al., 2001). The complete switch-1-region of the G domain undergoes major conformational changes upon GTP or GDP-binding, and the spatial relationships of the three domains change strongly with consequences for binding of interacting partners (Crepin et al., 2014). L. pneumophila Lgt1 mono-O-glucosylates serine-53 of eEF1A by forming an α-anomeric sugar–protein linkage (Belyi et al., 2006; Belyi et al., 2009). Serine-53 is located in the G domain of the GTP-binding protein within a loop region that connects helices A* and A′ of eEF1A (Fig. 4A). This region is not present in prokaryotic elongation factor Tu, which is the bacterial ortholog of eEF1A; therefore, Legionella EF-Tu is not a substrate for Lgt1. Isolated eEF1A is rather a poor substrate for Lgt1. The preferred substrate of Lgt1 is the ternary complex of GTPbound eEF1A with aminoacylated tRNA, suggesting that a specific conformation of eEF1A is essential for the ability to serve as substrate for Lgt1 (Tzivelekidis et al., 2011). The functional consequence of Lgt1-catalysed glucosylation of © 2015 John Wiley & Sons Ltd, Cellular Microbiology, 17, 1752–1765

1759

eEF1A is protein synthesis inhibition (Belyi et al., 2006; Belyi et al., 2012). However, it is still not clear which step in ribosome translation is blocked by glucosylation. In addition to Lgt1, most (if not all) strains of L. pneumophila produce also Lgt3 (Belyi et al., 2006). This enzyme has a molecular mass of about 100 kDa and is characterized by a conserved C-terminal repetitive region of unknown function. The N-terminal part is homologous to Lgt1 but slightly elongated. Lgt3 seems to be 10-times more efficient than Lgt1. Some strains of L. pneumophila produce Lgt2, a 70 kDa protein with considerable similarity to Lgt1. All these enzymes glucosylate serine-53 of eEF1A. Because synthesis of Lgt3 is maximal at early stages of bacterial growth, while Lgt1 and Lgt2 culminate at stationary phase, it is suggested that a temporal regulation of gene expression is important for infection of target cells by Legionella (Belyi et al., 2008). Legionella effector SetA Recently, another glycosyltransferase effector from L. pneumophila was discovered, named subversion of eukaryotic vesicle trafficking A (SetA). SetA is injected via the Dot/Icm system from the bacteria into the cytoplasm of the eukaryotic host cell (Franco et al., 2009; Huang et al., 2011). SetA interferes with the secretory pathway in host cells, blocking the transportation from the endoplasmic reticulum (ER) through the Golgi to lysosomes (Heidtman et al., 2009). The protein is composed of an N-terminal glycosyltransferase domain and a C-terminal localization domain (Jank et al., 2012). The localization domain specifically interacts with the phospholipid phosphatidylinositol-3-phosphate [PtdIns(3)P], which is a specific constituent of early endosomes. This domain might target SetA to the surface of vesicles (Jank et al., 2012). The N-terminal GTD of SetA resembles a GT-A type enzyme with the typical DxD-motif (D134x-D136), and the sugar binding triad consisting of residues D118, R121 and D134. In in vitro reactions, SetA hydrolyses specifically UDP-glucose, and it transfers the glucose moiety to threonine and serine residues on itself (auto-glucosylation) and also to histones H3.1 and H4, which serve as artificial substrates. Mutations of the DxD-motif blocks SetA-mediated UDP-glucose hydrolysis, histone glucosylation, mammalian cell death and endocytic secretory trafficking, demonstrating that the glucosyltransferase activity is crucial for the functionality of SetA (Heidtman et al., 2009; Jank et al., 2012). However, the eukaryotic target, which is glucosylated by SetA, remains still unidentified. Enteropathogenic E. coli NleB GlcNAcylates death domains Enteropathogenic E. coli (EPEC) is a gastrointestinal pathogen that is the most common cause of infantile

1760


Fig. 4. Actions of Legionella glucosyltransferases Lgt1 and E. coli (EPEC)-effector NleB. A. Legionella glucosyltransferases block protein translation. Lgts catalyse the glycosylation of eukaryotic elongation factor (eEF) 1A, which is important for efficient delivery of aminoacyl-tRNA (aa-tRNA) to the translating ribosome. Glucosylation of eEF1A at serine-53 prevents the ribosomal function and halts elongation of the nascent peptide chain. Right panel, crystal structure of yeast elongation factor eEF1A-GDP (adapted from PDB 1IJF) with Lgt1 glycosylation site serine-53 marked. B. Schematic of the function of enteropathogenic E. coli (EPEC)-effector NleB on cell death and inflammasome pathways.

diarrhoea (Clarke et al., 2003). Pathogenesis depends on the induction of typical so-called attaching and effacing lesions, which are characterized by redistribution of the actin cytoskeleton and formation of an actin-rich cup-like pedestal at the contact site of bacteria. However, host–pathogen interaction of EPEC is much more complicated than only inducing morphological changes at the contact site. It is well known that EPEC is able to inhibit cytokine release and inflammatory responses of target cells (Wong et al., 2011). The type-III secretion effector of EPEC NleB inhibits NFκB signalling. However, this is only observed when NFκB is stimulated via TNF receptors (Fig. 4B) (Newton et al., 2010). Recently, it was discovered that NleB is a glycosyltransferase, sharing structural similarity with the GT-8 family of glycosyltransferases (Li et al., 2013; Pearson et al., 2013). This GT-A family enzyme also possesses a DxD motif and depends on Mn2+ for glycosylation. NleB GlcNAcylates the death domains of TRADD (TNFR1-associated death domain), FADD (FAS-associated death domain) and RIPK1 (receptor-interacting serine/threonine-protein kinase 1)

(Fig. 4B). Surprisingly, modification by NleB occurs at an arginine residue (R235 of TRADD and R117 of FADD), which is unique for glycosyltransferases. Glycosylation of the death domains prevents protein interaction and inhibits formation of the caspase-8-activating death-inducing signalling complex, which contains TRADD, RIPK1 and FADD. Thus, NleB also attacks immune mechanisms and pathways like large clostridial glycosylating toxins, Photorhabdus glycosyltransferase PaTox and Yersinia Afp18; however, in this case, a completely different protein target is modified, which is not a GTP-binding protein like Rho proteins or eEF1A. Conclusions Glycosyltransferase toxins, which modify eukaryotic target proteins, are important virulence factors of different types of microbial pathogens, including Gram-negative as well as Gram-positive bacteria. Under normal physiological conditions, many eukaryotic cytosolic and nuclear proteins © 2015 John Wiley & Sons Ltd, Cellular Microbiology, 17, 1752–1765

Glycosyltransferase toxins (most likely > 1000) are modified by endogenous mono-OGlcNAcylation (Hart, 1997b; Zachara and Hart, 2002; Kamemura and Hart, 2003; Hart et al., 2011). All these modifications are catalysed by only one highly conserved glycosyltransferase, namely, O-GlcNAc transferase (OGT) (Haltiwanger et al., 1990). OGT is an inverting glycosyltransferase and forms a β-anomeric sugar–protein bond from UDP-GlcNAc. OGT-catalysed mono-O-GlcNAcylation of cytosolic and nuclear proteins is rapidly reversed by glycosidase O-GlcNAcase (OGA), which exclusively cleaves β-anomeric sugar–protein bonds (Hart, 1997a; Vocadlo, 2012). Thus, endogenous mono-O-GlcNAcylation is often compared with phosphorylation/dephosphorylation cycles involved in signalling (Hart et al., 2011). In contrast, the aforementioned toxins are retaining glycosyltransferases and form α-anomeric sugar–protein linkages, which are not cleaved by OGA. Thereby, the stability of the sugar–protein bond is high and the toxin effect is long-lasting. Endogenous GlcNAcylation may cause activation or inactivation of substrate proteins, depending on the site of modification and the type of protein substrate (Hart, 1997b; Zachara and Hart, 2002; Kamemura and Hart, 2003; Hart et al., 2011). In general, glycosyltransferase toxins inhibit target proteins and block their functions by direct steric hindrance of protein–protein interaction or by preventing the active conformation of the modified protein. However, in case of toxin-catalysed RhoA glycosylation, it is different. While this modification blocks the interaction of RhoA with all effectors studied so far, it also results in an activation signal for the inflammasome (Xu et al., 2014). Thus, it might be that an inhibitory signal is regulated by RhoA, which is relieved after glycosylation. Up to date studies indicate that sugar donors of glycosyltransferase toxins are only UDP-glucose and UDP-GlcNAc. The protein acceptor diversity appears to be higher than the diversity of sugar donors, at least including many small GTPases of the Ras/Rho families, eEF1A and death domains of proteins involved in TNF signalling. Surprising is the variability of the type of the acceptor amino acid, which is glycosylated by the toxins. While threonine and serine are also typical amino acid acceptors for mono-O-glycosylation in eukaryotic cells, glycosylation of the amino acid tyrosine is unique for bacterial toxins (exception auto-glycosylation of glycogenin). Also mono-Nglycosylation of arginine is so far only observed with bacterial toxins. This means, we can learn a lot about glycosylation reactions from bacterial protein toxins. Therefore, studies on glycosyltransferase toxins are not only essential for understanding their role as microbial virulence factors in host–pathogen interactions but may be also instrumental to answer emerging questions of carbohydrate research concerning protein substrate recognition, acceptor specificity or molecular mechanisms of retaining glycosylation reactions. © 2015 John Wiley & Sons Ltd, Cellular Microbiology, 17, 1752–1765

1761

Acknowledgements The research leading to these results has received funding from the People Programme (Marie Curie Actions) of the European Union’s Seventh Framework Programme (FP7/2007-2013) under REA grant agreement no. 609305.

References Achenbach, J., and Nierhaus, K.H. (2013) Translocation at work. Nat Struct Mol Biol 20: 1019–1022. Aktories, K. (2011) Bacterial protein toxins that modify host regulatory GTPases. Nat Rev Microbiol 9: 487–498. Aldape, M.J., Bryant, A.E., and Stevens, D.L. (2006) Clostridium sordellii infection: epidemiology, clinical findings, and current perspectives on diagnosis and treatment. Clin Infect Dis 43: 1436–1446. Amimoto, K., Noro, T., Oishi, E., and Shimizu, M. (2007) A novel toxin homologous to large clostridial cytotoxins found in culture supernatant of Clostridium perfringens type C. Microbiology 153: 1198–1206. Andersen, G.R., Valente, L., Pedersen, L., Kinzy, T.G., and Nyborg, J. (2001) Crystal structures of nucleotide exchange intermediates in the eEF1A-eEF1Balpha complex. Nat Struct Biol 8: 531–534. Barison, N., Gupta, R., and Kolbe, M. (2013) A sophisticated multi-step secretion mechanism: how the type 3 secretion system is regulated. Cell Microbiol 15: 1809–1817. Barth, H., Pfeifer, G., Hofmann, F., Maier, E., Benz, R., and Aktories, K. (2001) Low pH-induced formation of ion channels by Clostridium difficile toxin B in target cells. J Biol Chem 276: 10670–10676. Belyi, I., Popoff, M.R., and Cianciotto, N.P. (2003) Purification and characterization of a UDP-glucosyltransferase produced by Legionella pneumophila. Infect Immun 71: 181–186. Belyi, Y., Jank, T., and Aktories, K. (2013) Cytotoxic glucosyltransferases of Legionella pneumophila. Curr Top Microbiol Immunol 376: 211–226. Belyi, Y., Niggeweg, R., Opitz, B., Vogelsgesang, M., Hippenstiel, S., Wilm, M., and Aktories, K. (2006) Legionella pneumophila glucosyltransferase inhibits host elongation factor 1A. Proc Natl Acad Sci U S A 103: 16953–16958. Belyi, Y., Stahl, M., Sovkova, I., Kaden, P., Luy, B., and Aktories, K. (2009) Region of elongation factor 1A1 involved in substrate recognition by Legionella pneumophila glucosyltransferase LGT1-identification of LGT1 as a retaining glucosyltransferase. J Biol Chem 284: 20167–20174. Belyi, Y., Tabakova, I., Stahl, M., and Aktories, K. (2008) Lgt: a family of cytotoxic glucosyltransferases produced by Legionella pneumophila. J Bacteriol 190: 3026–3035. Belyi, Y., Tartakovskaya, D., Tais, A., Fitzke, E., Tzivelekidis, T., Jank, T., et al. (2012) Elongation factor 1A is the target of growth inhibition in yeast caused by Legionella pneumophila glucosyltransferase Lgt1. J Biol Chem 287: 26029–26037. Bhaskaran, S.S., and Stebbins, C.E. (2012) Structure of the catalytic domain of the Salmonella virulence factor SseI. Acta Crystallogr D Biol Crystallogr 68: 1613–1621. Bidla, G., Dushay, M.S., and Theopold, U. (2007) Crystal cell rupture after injury in Drosophila requires the JNK pathway, small GTPases and the TNF homolog Eiger. J Cell Sci 120: 1209–1215.

1762


Bitar, D.M., Molmeret, M., and Abu, K.Y. (2004) Molecular and cell biology of Legionella pneumophila. Int J Med Microbiol 293: 519–527. Brett, M.M., Hood, J., Brazier, J.S., Duerden, B.I., and Hahne, S.J. (2005) Soft tissue infections caused by spore-forming bacteria in injecting drug users in the United Kingdom. Epidemiol Infect 133: 575–582. Burridge, K., and Wennerberg, K. (2004) Rho and Rac take center stage. Cell 116: 167–179. Caron, E., and Hall, A. (1998) Identification of two distinct mechanisms of phagocytosis controlled by different Rho GTPases. Science 282: 1717–1721. Carter, G.P., Chakravorty, A., Pham Nguyen, T.A., Mileto, S., Schreiber, F., Li, L., et al. (2015) Defining the roles of TcdA and TcdB in localized gastrointestinal disease, systemic organ damage, and the host response during Clostridium difficile infections. MBio 6: e00551. Chandran, D.V., and Waksman, G. (2015) Structural biology of bacterial type IV secretion systems. Annu Rev Biochem 84: 603–629. Chen, Y., Tascon, I., Neunuebel, M.R., Pallara, C., Brady, J., Kinch, L.N., et al. (2013) Structural basis for Rab1 deAMPylation by the Legionella pneumophila effector SidD. PLoS Pathog 9: e1003382. Clarke, S.C., Haigh, R.D., Freestone, P.P., and Williams, P.H. (2003) Virulence of enteropathogenic Escherichia coli, a global pathogen. Clin Microbiol Rev 16: 365–378. Coulthurst, S.J. (2013) The Type VI secretion system – a widespread and versatile cell targeting system. Res Microbiol 164: 640–654. Crepin, T., Shalak, V.F., Yaremchuk, A.D., Vlasenko, D.O., McCarthy, A., Negrutskii, B.S., et al. (2014) Mammalian translation elongation factor eEF1A2: X-ray structure and new features of GDP/GTP exchange mechanism in higher eukaryotes. Nucleic Acids Res 42: 12939–12948. Cui, J., Yao, Q., Li, S., Ding, X., Lu, Q., Mao, H., et al. (2010) Glutamine deamidation and dysfunction of ubiquitin/NEDD8 induced by a bacterial effector family. Science 329: 1215–1218. Deng, Q., and Barbieri, J.T. (2008) Molecular mechanisms of the cytotoxicity of ADP-ribosylating toxins. Annu Rev Microbiol 62: 271–288. Egerer, M., Giesemann, T., Jank, T., Satchell, K.J., and Aktories, K. (2007) Auto-catalytic cleavage of Clostridium difficile toxins A and B depends on a cysteine protease activity. J Biol Chem 282: 25314–25321. Endo, Y., Tsurugi, K., Yutsudo, T., Takeda, Y., Ogasawara, T., and Igarashi, K. (1988) Site of action of a Vero toxin (VT2) from Escherichia coli O157:H7 and of Shiga toxin on eukaryotic ribosomes. RNA N-glycosidase activity of the toxins. Eur J Biochem 171: 45–50. Ensminger, A.W., and Isberg, R.R. (2009) Legionella pneumophila Dot/Icm translocated substrates: a sum of parts. Curr Opin Microbiol 12: 67–73. Farrow, M.A., Chumbler, N.M., Lapierre, L.A., Franklin, J.L., Rutherford, S.A., Goldenring, J.R., and Lacy, D.B. (2013) Clostridium difficile toxin B-induced necrosis is mediated by the host epithelial cell NADPH oxidase complex. Proc Natl Acad Sci U S A 110: 18674–18679. Fauvarque, M.O., and Williams, M.J. (2011) Drosophila cellular immunity: a story of migration and adhesion. J Cell Sci 124: 1373–1382.

ffrench-Constant, R.H., and Bowen, D.J. (2000) Novel insecticidal toxins from nematode-symbiotic bacteria. Cell Mol Life Sci 57: 828–833. Fields, B.S., Benson, R.F., and Besser, R.E. (2002) Legionella and Legionnaires’ disease: 25 years of investigation. Clin Microbiol Rev 15: 506–526. Fiorentini, C., and Thelestam, M. (1991) Clostridium difficile toxin A and its effects on cells. Toxicon 29: 543–567. Fischer, M., Bhatnagar, J., Guarner, J., Reagan, S., Hacker, J.K., Van Meter, S.H., et al. (2005) Fatal toxic shock syndrome associated with Clostridium sordellii after medical abortion. N Engl J Med 353: 2352–2360. Flatau, G., Lemichez, E., Gauthier, M., Chardin, P., Paris, S., Fiorentini, C., and Boquet, P. (1997) Toxin-induced activation of the G protein p21 Rho by deamidation of glutamine. Nature 387: 729–733. Forst, S., Dowds, B., Boemare, N., and Stackebrandt, E. (1997) Xenorhabdus and Photorhabdus spp.: bugs that kill bugs. Annu Rev Microbiol 51: 47–72. Franco, I.S., Shuman, H.A., and Charpentier, X. (2009) The perplexing functions and surprising origins of Legionella pneumophila type IV secretion effectors. Cell Microbiol 11: 1435–1443. Galan, J.E., Lara-Tejero, M., Marlovits, T.C., and Wagner, S. (2014) Bacterial type III secretion systems: specialized nanomachines for protein delivery into target cells. Annu Rev Microbiol 68: 415–438. Geissler, B., Tungekar, R., and Satchell, K.J. (2010) Identification of a conserved membrane localization domain within numerous large bacterial protein toxins. Proc Natl Acad Sci U S A 107: 5581–5586. Genisyuerek, S., Papatheodorou, P., Guttenberg, G., Schubert, R., Benz, R., and Aktories, K. (2011) Structural determinants for membrane insertion, pore formation and translocation of Clostridium difficile toxin B. Mol Microbiol 79: 1643–1654. Genth, H., Aktories, K., and Just, I. (1999) Monoglucosylation of RhoA at threonine 37 blocks cytosol-membrane cycling. J Biol Chem 274: 29050–29056. Genth, H., Pauillac, S., Schelle, I., Bouvet, P., Bouchier, C., Varela-Chavez, C., et al. (2014) Haemorrhagic toxin and lethal toxin from Clostridium sordellii strain vpi9048: molecular characterization and comparative analysis of substrate specificity of the large clostridial glucosylating toxins. Cell Microbiol 16: 1706–1721. Goody, P.R., Heller, K., Oesterlin, L.K., Muller, M.P., Itzen, A., and Goody, R.S. (2012) Reversible phosphocholination of Rab proteins by Legionella pneumophila effector proteins. EMBO J 31: 1774–1784. Greco, A., Ho, J.G., Lin, S.J., Palcic, M.M., Rupnik, M., and Ng, K.K. (2006) Carbohydrate recognition by Clostridium difficile toxin A. Nat Struct Mol Biol 13: 460–461. Guttenberg, G., Hornei, S., Jank, T., Schwan, C., Lu, W., Einsle, O., et al. (2012) Molecular characteristics of Clostridium perfringens TpeL toxin and consequences of mono-O-GlcNAcylation of Ras in living cells. J Biol Chem 287: 24929–24940. Hall, A. (1998) Rho GTPases and the actin cytoskeleton. Science 279: 509–514. Haltiwanger, R.S., Holt, G.D., and Hart, G.W. (1990) Enzymatic addition of O-GlcNAc to nuclear and cytoplasmic proteins. identification of a uridine diphospho-N-acetylglucosamine: © 2015 John Wiley & Sons Ltd, Cellular Microbiology, 17, 1752–1765

Glycosyltransferase toxins peptide beta-N-acetylglucosaminyltransferase. J Biol Chem 265: 2563–2568. Hart, G.W. (1997a) Dynamic O-GlcNAcylation of nuclear and cytoskeletal proteins. Annu Rev Biochem 66: 315–335. Hart, G.W. (1997b) Dynamic O-linked glycosylation of nuclear and cytoskeletal proteins. Annu Rev Biochem 66: 315–335. Hart, G.W., Slawson, C., Ramirez-Correa, G., and Lagerlof, O. (2011) Cross talk between O-GlcNAcylation and phosphorylation: roles in signaling, transcription, and chronic disease. Annu Rev Biochem 80: 825–858. Heidtman, M., Chen, E.J., Moy, M.Y., and Isberg, R.R. (2009) Large-scale identification of Legionella pneumophila Dot/Icm substrates that modulate host cell vesicle trafficking pathways. Cell Microbiol 11: 230–248. Heymann, J.B., Bartho, J.D., Rybakova, D., Venugopal, H.P., Winkler, D.C., Sen, A., et al. (2013) Three-dimensional structure of the toxin-delivery particle antifeeding prophage of Serratia entomophila. J Biol Chem 288: 25276–25284. Hofmann, F., Busch, C., Prepens, U., Just, I., and Aktories, K. (1997) Localization of the glucosyltransferase activity of Clostridium difficile toxin B to the N-terminal part of the holotoxin. J Biol Chem 272: 11074–11078. Honjo, T., Nishizuka, Y., and Hayaishi, O. (1968) Diphtheria toxin-dependent adenosine diphosphate ribosylation of aminoacryl transferase II and inhibition of protein synthesis. J Biol Chem 243: 3553–3555. Horwitz, M.A., and Silverstein, S.C. (1980) Legionnaires’ disease bacterium (Legionella pneumophila) multiples intracellularly in human monocytes. J Clin Invest 66: 441–450. Huang, L., Boyd, D., Amyot, W.M., Hempstead, A.D., Luo, Z. Q., O’Connor, T.J., et al. (2011) The E Block motif is associated with Legionella pneumophila translocated substrates. Cell Microbiol 13: 227–245. Hurtado-Guerrero, R., Zusman, T., Pathak, S., Ibrahim, A.F., Shepherd, S., Prescott, A., et al. (2010) Molecular mechanism of elongation factor 1A inhibition by a Legionella pneumophila glycosyltransferase. Biochem J 426: 281–292. Isberg, R.R., O’Connor, T.J., and Heidtman, M. (2009) The Legionella pneumophila replication vacuole: making a cosy niche inside host cells. Nat Rev Microbiol 7: 13–24. Ishida, Y., Maegawa, T., Kondo, T., Kimura, A., Iwakura, Y., Nakamura, S., and Mukaida, N. (2004) Essential involvement of IFN-gamma in Clostridium difficile toxin A-induced enteritis. J Immunol 172: 3018–3025. Jafari, N.V., Kuehne, S.A., Bryant, C.E., Elawad, M., Wren, B. W., Minton, N.P., et al. (2013) Clostridium difficile modulates host innate immunity via toxin-independent and dependent mechanism(s). PLoS One 8: e69846. Jaffe, A.B., and Hall, A. (2005) Rho GTPases: biochemistry and biology. Annu Rev Cell Dev Biol 21: 247–269. Jank, T., and Aktories, K. (2008) Structure and mode of action of clostridial glucosylating toxins: the ABCD model. Trends Microbiol 16: 222–229. Jank, T., Bogdanovic, X., Wirth, C., Haaf, E., Spoerner, M., Bohmer, K.E., et al. (2013) A bacterial toxin catalyzing tyrosine glycosylation of Rho and deamidation of Gq and Gi proteins. Nat Struct Mol Biol 20: 1273–1280. Jank, T., Bohmer, K.E., Tzivelekidis, T., Schwan, C., Belyi, Y., and Aktories, K. (2012) Domain organization of Legionella effector SetA. Cell Microbiol 14: 852–868. © 2015 John Wiley & Sons Ltd, Cellular Microbiology, 17, 1752–1765

1763

Jank, T., Trillhaase, C., Brozda, N., Steinemann, M., Schwan, C., Suss, R., and Aktories, K. (2015) Intracellular plasma membrane guidance of Photorhabdus asymbiotica toxin is crucial for cell toxicity. FASEB J 29: 2789–2802. Just, I., and Gerhard, R. (2004) Large clostridial cytotoxins. Rev Physiol Biochem Pharmacol 152: 23–47. Just, I., Selzer, J., Hofmann, F., Green, G.A., and Aktories, K. (1996) Inactivation of Ras by Clostridium sordellii lethal toxincatalyzed glucosylation. J Biol Chem 271: 10149–10153. Just, I., Selzer, J., Wilm, M., Von Eichel-Streiber, C., Mann, M., and Aktories, K. (1995a) Glucosylation of Rho proteins by Clostridium difficile toxin B. Nature 375: 500–503. Just, I., Wilm, M., Selzer, J., Rex, G., Von Eichel-Streiber, C., Mann, M., and Aktories, K. (1995b) The enterotoxin from Clostridium difficile (ToxA) monoglucosylates the Rho proteins. J Biol Chem 270: 13932–13936. Kamemura, K., and Hart, G.W. (2003) Dynamic interplay between O-glycosylation and O-phosphorylation of nucleocytoplasmic proteins: a new paradigm for metabolic control of signal transduction and transcription. Prog Nucleic Acid Res Mol Biol 73: 107–136. Kitadokoro, K., Kamitani, S., Miyazawa, M., HanajimaOzawa, M., Fukui, A., Miyake, M., and Horiguchi, Y. (2007) Crystal structures reveal a thiol protease-like catalytic triad in the C-terminal region of Pasteurella multocida toxin. Proc Natl Acad Sci U S A 104: 5139–5144. Kuehne, S.A., Cartman, S.T., Heap, J.T., Kelly, M.L., Cockayne, A., and Minton, N.P. (2010) The role of toxin A and toxin B in Clostridium difficile infection. Nature 467: 711–713. LaFrance, M.E., Farrow, M.A., Chandrasekaran, R., Sheng, J., Rubin, D.H., and Lacy, D.B. (2015) Identification of an epithelial cell receptor responsible for Clostridium difficile TcdB-induced cytotoxicity. Proc Natl Acad Sci U S A 112: 7073–7078. Lang, A.E., Schmidt, G., Schlosser, A., Hey, T.D., Larrinua, I. M., Sheets, J.J., et al. (2010) Photorhabdus luminescens toxins ADP-ribosylate actin and RhoA to force actin clustering. Science 327: 1139–1142. Lemichez, E., and Aktories, K. (2013) Hijacking of Rho GTPases during bacterial infection. Exp Cell Res 319: 2329–2336. Li, S., Shi, L., Yang, Z., Zhang, Y., Perez-Cordon, G., Huang, T., et al. (2015) Critical roles of Clostridium difficile toxin B enzymatic activities in pathogenesis. Infect Immun 83: 502–513. Li, S., Zhang, L., Yao, Q., Li, L., Dong, N., Rong, J., et al. (2013) Pathogen blocks host death receptor signalling by arginine GlcNAcylation of death domains. Nature 501: 242–246. Linevsky, J.K., Pothoulakis, C., Keates, S., Warny, M., Keates, A.C., LaMont, J.T., and Kelly, C.P. (1997) IL-8 release and neutrophil activation by Clostridium difficile toxin-exposed human monocytes. Am J Physiol 273: G1333–G1340. Lu, A., and Wu, H. (2015) Structural mechanisms of inflammasome assembly. FEBS J 282: 435–444. Lu, W., Du, J., Stahl, M., Tzivelekidis, T., Belyi, Y., Gerhardt, S., et al. (2010) Structural basis of the action of glucosyltransferase Lgt1 from Legionella pneumophila. J Mol Biol 396: 321–331.

1764


Lyras, D., O’Connor, J.R., Howarth, P.M., Sambol, S.P., Carter, G.P., Phumoonna, T., et al. (2009) Toxin B is essential for virulence of Clostridium difficile. Nature 458: 1176–1179. Nagahama, M., Ohkubo, A., Oda, M., Kobayashi, K., Amimoto, K., Miyamoto, K., and Sakurai, J. (2010) Clostridium perfringens TpeL glycosylates the Rac and Ras subfamily proteins. Infect Immun 79: 905–910. Newton, H.J., Pearson, J.S., Badea, L., Kelly, M., Lucas, M., Holloway, G., et al. (2010) The type III effectors NleE and NleB from enteropathogenic E. coli and OspZ from Shigella block nuclear translocation of NF-kappaB p65. PLoS Pathog 6: e1000898. Ng, J., Hirota, S.A., Gross, O., Li, Y., Ulke-Lemee, A., Potentier, M.S., et al. (2010) Clostridium difficile toxininduced inflammation and intestinal injury are mediated by the inflammasome. Gastroenterology 139: 542–552, 552. Nobes, C.D., and Hall, A. (1999) Rho GTPases control polarity, protrusion, and adhesion during cell movement. J Cell Biol 144: 1235–1244. Olling, A., Goy, S., Hoffmann, F., Tatge, H., Just, I., and Gerhard, R. (2011) The repetitive oligopeptide sequences modulate cytopathic potency but are not crucial for cellular uptake of Clostridium difficile toxin A. PLoS One 6 e17623. Orth, J.H., Preuss, I., Fester, I., Schlosser, A., Wilson, B.A., and Aktories, K. (2009) Pasteurella multocida toxin activation of heterotrimeric G proteins by deamidation. Proc Natl Acad Sci U S A 106: 7179–7184. Papatheodorou, P., Zamboglou, C., Genisyuerek, S., Guttenberg, G., and Aktories, K. (2010) Clostridial glucosylating toxins enter cells via clathrin-mediated endocytosis. PLoS One 5: e10673. Parker, M.W., and Feil, S.C. (2005) Pore-forming protein toxins: from structure to function. Prog Biophys Mol Biol 88: 91–142. Pearson, J.S., Giogha, C., Ong, S.Y., Kennedy, C.L., Kelly, M., Robinson, K.S., et al. (2013) A type III effector antagonizes death receptor signalling during bacterial gut infection. Nature 501: 247–251. Popoff, M.R. (2014) Bacterial factors exploit eukaryotic Rho GTPase signaling cascades to promote invasion and proliferation within their host. Small GTPases 5 pii:e28209. Pruitt, R.N., Chumbler, N.M., Rutherford, S.A., Farrow, M.A., Friedman, D.B., Spiller, B., and Lacy, D.B. (2012) Structural determinants of Clostridium difficile toxin A glucosyltransferase activity. J Biol Chem 287: 8013–8020. Ramakrishnan, V. (2002) Ribosome structure and the mechanism of translation. Cell 108: 557–572. Reineke, J., Tenzer, S., Rupnik, M., Koschinski, A., Hasselmayer, O., Schrattenholz, A., et al. (2007) Autocatalytic cleavage of Clostridium difficile toxin B. Nature 446: 415–419. Reinert, D.J., Jank, T., Aktories, K., and Schulz, G.E. (2005) Structural basis for the function of Clostridium difficile toxin B. J Mol Biol 351: 973–981. Ridley, A.J., and Hall, A. (1992) The small GTP-binding protein rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors. Cell 70: 389–399. Rucker, R.R. (1966) Redmouth disease of rainbow trout (Salmo gairdneri). Bull Off Int Epizoot 65: 825–830.

Samlaska, C.P., and Maggio, K.L. (1996) Subcutaneous emphysema. Adv Dermatol 11: 117–151. Sandvig, K., Spilsberg, B., Lauvrak, S.U., Torgersen, M.L., Iversen, T.G., and van Deurs, B. (2004) Pathways followed by protein toxins into cells. Int J Med Microbiol 293: 483–490. Sarris, P.F., Ladoukakis, E.D., Panopoulos, N.J., and Scoulica, E.V. (2014) A phage tail-derived element with wide distribution among both prokaryotic domains: a comparative genomic and phylogenetic study. Genome Biol Evol 6: 1739–1747. Schiavo, G., Matteoli, M., and Montecucco, C. (2000) Neurotoxins affecting neuroexocytosis. Physiol Rev 80: 717–766. Schmidt, G., Sehr, P., Wilm, M., Selzer, J., Mann, M., and Aktories, K. (1997) Gln 63 of Rho is deamidated by Escherichia coli cytotoxic necrotizing factor-1. Nature 387: 725–729. Schorch, B., Song, S., van Diemen, F.R., Bock, H.H., May, P., Herz, J., et al. (2014) LRP1 is a receptor for Clostridium perfringens TpeL toxin indicating a two-receptor model of clostridial glycosylating toxins. Proc Natl Acad Sci U S A 111: 6431–6436. Sehr, P., Joseph, G., Genth, H., Just, I., Pick, E., and Aktories, K. (1998) Glucosylation and ADP-ribosylation of Rho proteins – effects on nucleotide binding, GTPase activity, and effector-coupling. Biochemistry 37: 5296–5304. Selzer, J., Hofmann, F., Rex, G., Wilm, M., Mann, M., Just, I., and Aktories, K. (1996) Clostridium novyi α-toxin-catalyzed incorporation of GlcNAc into Rho subfamily proteins. J Biol Chem 271: 25173–25177. Simon, N.C., Aktories, K., and Barbieri, J.T. (2014) Novel bacterial ADP-ribosylating toxins: structure and function. Nat Rev Microbiol 12: 599–611. Smythe, C., Caudwell, F.B., Ferguson, M., and Cohen, P. (1988) Isolation and structural analysis of a peptide containing the novel tyrosyl-glucose linkage in glycogenin. EMBO J 7: 2681–2686. Swanson, M.S., and Hammer, B.K. (2000) Legionella pneumophila pathogenesis: a fateful journey from amoebae to macrophages. Annu Rev Microbiol 54: 567–613. Trujillo, C., Ratts, R., Tamayo, A., Harrison, R., and Murphy, J.R. (2006) Trojan horse or proton force: finding the right partner(s) for toxin translocation. Neurotox Res 9: 63–71. Tzivelekidis, T., Jank, T., Pohl, C., Schlosser, A., Rospert, S., Knudsen, C.R., et al. (2011) Aminoacyl-tRNA-charged eukaryotic elongation factor 1A is the bona fide substrate for Legionella pneumophila effector glucosyltransferases. PLoS One 6: e29525. Vocadlo, D.J. (2012) O-GlcNAc processing enzymes: catalytic mechanisms, substrate specificity, and enzyme regulation. Curr Opin Chem Biol 16: 488–497. Von Eichel-Streiber, C., Sauerborn, M., and Kuramitsu, H.K. (1992) Evidence for a modular structure of the homologous repetitive C-terminal carbohydrate-binding sites of Clostridium difficile toxins and Streptcoccus mutans glucosyltransferases. J Bacteriol 174: 6707–6710. Voth, D.E., and Ballard, J.D. (2005) Clostridium difficile toxins: mechanism of action and role in disease. Clin Microbiol Rev 18: 247–263. © 2015 John Wiley & Sons Ltd, Cellular Microbiology, 17, 1752–1765

Glycosyltransferase toxins Warny, M., Keates, A.C., Keates, S., Castagliuolo, I., Zacks, J. K., Aboudola, S., et al. (2000) p38 MAP kinase activation by Clostridium difficile toxin A mediates monocyte necrosis, IL-8 production, and enteritis. J Clin Invest 105: 1147–1156. Wohlan, K., Goy, S., Olling, A., Srivaratharajan, S., Tatge, H., Genth, H., and Gerhard, R. (2014) Pyknotic cell death induced by Clostridium difficile TcdB: chromatin condensation and nuclear blister are induced independently of the glucosyltransferase activity. Cell Microbiol 16: 1678–1692. Wong, A.R., Pearson, J.S., Bright, M.D., Munera, D., Robinson, K.S., Lee, S.F., et al. (2011) Enteropathogenic and enterohaemorrhagic Escherichia coli: even more subversive elements. Mol Microbiol 80: 1420–1438. Worby, C.A., Mattoo, S., Kruger, R.P., Corbeil, L.B., Koller, A., Mendez, J.C., et al. (2009) The fic domain: regulation of cell signaling by adenylylation. Mol Cell 34: 93–103. Xu, H., Yang, J., Gao, W., Li, L., Li, P., Zhang, L., et al. (2014) Innate immune sensing of bacterial modifications of Rho

© 2015 John Wiley & Sons Ltd, Cellular Microbiology, 17, 1752–1765

1765

GTPases by the Pyrin inflammasome. Nature 513: 237–241. Yarbrough, M.L., Li, Y., Kinch, L.N., Grishin, N.V., Ball, H.L., and Orth, K. (2009) AMPylation of Rho GTPases by Vibrio VopS disrupts effector binding and downstream signaling. Science 323: 269–272. Young, J.C., Clements, A., Lang, A.E., Garnett, J.A., Munera, D., Arbeloa, A., et al. (2014) The Escherichia coli effector EspJ blocks Src kinase activity via amidation and ADP ribosylation. Nat Commun 5: 5887. Yuan, P., Zhang, H., Cai, C., Zhu, S., Zhou, Y., Yang, X., et al. (2015) Chondroitin sulfate proteoglycan 4 functions as the cellular receptor for Clostridium difficile toxin B. Cell Res 25: 157–168. Zachara, N.E., and Hart, G.W. (2002) The emerging significance of O-GlcNAc in cellular regulation. Chem Rev 102: 431–438. Ziegler, M.O., Jank, T., Aktories, K., and Schulz, G.E. (2008) Conformational changes and reaction of clostridial glycosylating toxins. J Mol Biol 377: 1346–1356.