crossmark
Analysis of TcdB Proteins within the Hypervirulent Clade 2 Reveals
an Impact of RhoA Glucosylation on Clostridium difficile
Proinflammatory Activities
Carlos Quesada-Gómez,a Diana López-Ureña,a Nicole Chumbler,b Heather K. Kroh,b Carolina Castro-Peña,a César Rodríguez,a
Josué Orozco-Aguilar,c,d Sara González-Camacho,d Alexandra Rucavado,e Caterina Guzmán-Verri,f Trevor D. Lawley,g
D. Borden Lacy,b,h Esteban Chaves-Olartea
Facultad de Microbiología and Centro de Investigación en Enfermedades Tropicales, Universidad de Costa Rica, San José, Costa Ricaa; Department of Pathology,
Microbiology, and Immunology, Vanderbilt University School of Medicine, Nashville, Tennessee, USAb; Facultad de Farmacia, Universidad de Costa Rica, San José, Costa
Ricac; Laboratorio de Ensayos Biológicos, Escuela de Medicina, Universidad de Costa Rica, San José, Costa Ricad; Instituto Clodomiro Picado, Facultad de Microbiología,
Universidad de Costa Rica, San José, Costa Ricae; Programa de Investigación en Enfermedades Tropicales, Escuela de Medicina Veterinaria, Universidad Nacional, Heredia,
Costa Ricaf; Host-Microbiota Interactions Laboratory, Wellcome Trust Sanger Institute, Hinxton, United Kingdomg; The Veterans Affairs Tennessee Valley Healthcare
System, Nashville, Tennessee, USAh
Clostridium difficile strains within the hypervirulent clade 2 are responsible for nosocomial outbreaks worldwide. The increased
pathogenic potential of these strains has been attributed to several factors but is still poorly understood. During a C. difficile
outbreak, a strain from this clade was found to induce a variant cytopathic effect (CPE), different from the canonical arborizing
CPE. This strain (NAP1V) belongs to the NAP1 genotype but to a ribotype different from the epidemic NAP1/RT027 strain.
NAP1V and NAP1 share some properties, including the overproduction of toxins, the binary toxin, and mutations in tcdC.
NAP1V is not resistant to fluoroquinolones, however. A comparative analysis of TcdB proteins from NAP1/RT027 and NAP1V
strains indicated that both target Rac, Cdc42, Rap, and R-Ras but only the former glucosylates RhoA. Thus, TcdB from hyper-
virulent clade 2 strains possesses an extended substrate profile, and RhoA is crucial for the type of CPE induced. Sequence com-
parison and structural modeling revealed that TcdBNAP1 and TcdBNAP1V share the receptor-binding and autoprocessing activi-
ties but vary in the glucosyltransferase domain, consistent with the different substrate profile. Whereas the two toxins displayed
identical cytotoxic potencies, TcdBNAP1 induced a stronger proinflammatory response than TcdBNAP1V as determined in ex vivo
experiments and animal models. Since immune activation at the level of intestinal mucosa is a hallmark of C. difficile-induced
infections, we propose that the panel of substrates targeted by TcdB is a determining factor in the pathogenesis of this pathogen
and in the differential virulence potential seen among C. difficile strains.
Clostridium difficile, a Gram-positive spore-forming anaerobe, host cell receptor(s) (11). The middle part of the toxin representsis the leading cause of antibiotic-associated diarrhea in hospi- the translocation region with autoprocessing activity mediated by
talized patients (1). Antibiotic treatment modifies the balance of an autoprotease domain (9). The GTD located in the N-terminal
commensal microbiota, allowing C. difficile to extensively colo- region is composed of a catalytic and a substrate recognition sub-
nize the gut. The resulting C. difficile infection (CDI) leads to a domain; this region is responsible for the cytopathic activity in the
variety of clinical outcomes that range from mild diarrhea to po- host cell cytosol (12).
tentially fatal pseudomembranous colitis (2). C. difficile strains producing a variant TcdB have been previ-
The main virulence factors associated with CDI are two large ously reported, mainly in TcdA-negative strains (13–15). In cul-
exotoxins, TcdA and TcdB. The toxins are encoded by the tcdA tured cells, these TcdB variants induce a CPE characterized by the
and tcdB genes, respectively, which are located in a 19.6-kb patho-
genicity locus (PaLoc) together with the tcdE (holin-like), tcdC
(putative negative regulator), and tcdR (sigma factor) genes (3, 4).
Received 14 October 2015 Returned for modification 23 November 2015
The toxins glucosylate small GTPases (5), and their combined Accepted 4 January 2016
action results in colonic tissue inflammation and massive colonic Accepted manuscript posted online 11 January 2016
fluid secretion (2). In cell cultures treated with C. difficile toxins, Citation Quesada-Gómez C, López-Ureña D, Chumbler N, Kroh HK, Castro-Peña C,
monoglucosylation of RhoA, Rac1, and Cdc42 disrupts the actin Rodríguez C, Orozco-Aguilar J, González-Camacho S, Rucavado A, Guzmán-Verri C,
cytoskeleton and causes an arborizing cytopathic effect (CPE) (5). Lawley TD, Lacy DB, Chaves-Olarte E. 2016. Analysis of TcdB proteins within the
TcdB is a 270-kDa cytotoxin, and its mechanism of action in- hypervirulent clade 2 reveals an impact of RhoA glucosylation on Clostridium
difficile proinflammatory activities. Infect Immun 84:856 – 865.
volves host cell receptor binding (6), uptake by endocytosis (7), doi:10.1128/IAI.01291-15.
pH-dependent pore formation (8), translocation across the endo- Editor: V. B. Young
somal membrane (9), host factor-dependent autoprocessing (10), Address correspondence to Esteban Chaves-Olarte, esteban.chaves@ucr.ac.cr.
and release of the glucosyltransferase domain (GTD) into the host C.Q.-G. and D.L.-U. contributed equally to the work.
cell (11). The C-terminal domain of the holotoxin contains a Copyright © 2016 Quesada-Gómez et al. This is an open-access article distributed
number of short, homologous regions with combined repetitive under the terms of the Creative Commons Attribution 4.0 International license.
oligopeptides (CROPs) and is thought to be important for binding
856 iai.asm.org Infection and Immunity March 2016 Volume 84 Number 3
Downloaded from http://iai.asm.org/ on July 12, 2017 by UNIV DE COSTA RICA SISTEMA DE
Variant TcdB in C. difficile Hypervirulent Clade 2
collapse of the actin cytoskeleton with complete rounding of the ST37). For automatic annotation, we used Prokka (36) and custom C.
cell body and detachment from the surface in contrast to the clas- difficile databases. For core genome multialignment, variant calling, and
sic arborizing effect (13). This variant CPE is due to a different core genome phylogeny, we used the Harvest suite (37) and FigTree (http:
pattern of glucosylated GTPases since classic TcdB modifies //tree.bio.ed.ac.uk/software/figtree/). For multilocus sequence typing
RhoA, Rac, and Cdc42 whereas variant TcdB targets Rac, Cdc42, (MLST), we used the MLST 1.7 tool maintained by the Center for
Rap, Ral, and R-Ras (5, 13, 14, 16). Furthermore, variations based Genomic Epidemiology at the Danish Technical University (38). PaLoc
and TcdB sequences were extracted manually and aligned with MAFFT
on the PaLoc sequence have classified these groups of strains in
(39) or MUSCLE (40). For these sequences, phylogenetic tree estimation
separate toxinotypes (17). through maximum likelihood was done using Fasttree (41). TcdB recom-
The epidemic NAP1/RT027 C. difficile strains have rapidly bination was detected using DualBrothers (42).
spread and have been responsible for epidemic outbreaks world- Quantitation of secreted toxins. The NAP1V and NAP1 strains were
wide (18, 19). Among the factors that have been proposed to con- grown in TYT broth (3% Bacto tryptose, 2% yeast extract, and 0.1%
tribute to the increased virulence of these strains are resistance to thioglycolate, pH 6.8) for 24 h, as described previously (29). Decimal
fluoroquinolones, higher sporulation capacity, and increased pro- dilutions of these supernatants were added to HeLa cell monolayers. The
duction of toxins (20–22). It has been demonstrated that TcdB cells were monitored for appearance of CPE. Specific TcdB antiserum
from epidemic NAP1/RT027 strains possesses an increased cyto- (TechLab) was used to neutralize the effect of the toxin. Nontoxigenic C.
toxic capacity on different cell types due to a more efficient auto- difficile ATCC 700057 was used as a negative control. Cytotoxicity was
processing activity, which would result in a more rapid release of expressed as the inverse of the dilution of the supernatants that caused
the enzymatic domain into the cytosol (23). These results indicate 50% cell rounding in the monolayers (CPE50). The amount of toxins was
quantified by Western blotting, for which the proteins from bacteri-
that altered TcdB activity could be an additional important factor
um-free supernatants at 24 h were concentrated by methanol-chloro-
for the increased pathogenesis of NAP1 strains. form precipitation (43). Proteins were separated in SDS-PAGE gels
In this work, we describe a C. difficile NAP1 strain from the and electrotransferred to polyvinylidene difluoride (PVDF) mem-
hypervirulent clade 2 carrying a variant TcdB (TcdBNAP1V). In branes. The membranes were probed with monoclonal anti-TcdA
contrast to TcdB from the classic NAP1/RT027 strain, TcdBNAP1V (TTC8) or anti-TcdB (2CV) antibody (tgcBIOMICS) (43). Chemilumi-
does not glucosylate Rho and partially targets Cdc42. Whereas the nescent signals emitted by a goat anti-mouse IgG-horseradish peroxidase
cytopathic potency of this TcdBNAP1V is similar to that of TcdB conjugate (Invitrogen) in the presence of the Lumi-Light Plus Western
purified from classic NAP1 strains, it induces a significantly lower blotting substrate (Roche) were recorded with a ChemiDoc XRS docu-
quantity of proinflammatory mediators in the ligated loop model, mentation system (Bio-Rad). Transcripts of tcdA and tcdB were quantified
suggesting that the panel of glucosylated small GTPases deter- by real-time quantitative PCR (qRT-PCR) as described previously (44).
mines the biological outcome induced by C. difficile toxins. The amplification followed conditions previously reported (44). The rel-
ative expression of genes was calculated by the threshold cycle (CT)
MATERIALS AND METHODS method using rpoA transcript as the endogenous control (45).
Isolation and characterization of C. difficile strains and fluoroquin- Toxin purification. TcdB proteins were obtained from supernatants
olone resistance. The NAP1 strains were isolated from stool samples ac- of NAP1 strains grown in a dialysis system culture and purified as de-
cording to the protocols previously described (24). Fragments of tcdA, scribed previously (46). The purity of the toxins was determined by SDS-
tcdB, cdtB, and tcdC were amplified by PCR using primers and conditions PAGE and mass spectrometry which indicated the presence of peptides
previously reported (25, 26). MICs of ciprofloxacin, moxifloxacin, and derived from TcdB only and not TcdA (data not shown).
levofloxacin were determined using agar dilution according to guidelines Cytopathic effect produced by NAP1 toxins. Confluent 3T3 fibro-
of the Clinical and Laboratory Standards Institute (CLSI; M11-A7). Re- blasts, Vero cells, and HeLa cells grown in 12-mm glass slides were intox-
sistance breakpoints were 4 g · ml1. Mutations in the fluoroquin- icated with 0.2 nM TcdBNAP1 and TcdBNAP1V. The cells were immobilized
olone resistance-determining region of gyrA and gyrB and in the tcdC and fixed according to previously described protocols (47). The CPE was
genes were identified using Artemis (27) and BLAST tools. evaluated by phase-contrast, fluorescence, and scanning electron micros-
PFGE typing. The pulsed-field gel electrophoresis (PFGE) procedure copy as indicated in the figure legends.
was derived from published protocols (28, 29). Bacteria from 6- to 8-h In vitro glycosyltransferase activity. The TcdB ability to glycosylate
cultures in brain heart infusion (BHI) were disrupted in lysis buffer. Aga- different monomeric GTPases was examined through a radioactivity as-
rose plugs were prepared by mixing equal volumes of bacterial suspen- say, as previously described (48, 49), and Western blot assays. Briefly, for
the radioactive test, UDP-[14sions and SeaKem Gold agarose (Lonza) in Tris-EDTA (TE) buffer con- C]glucose (250 mCi/mmol; PerkinElmer),
taining SDS. After overnight digestion with SmaI (Roche), DNA GTD, each recombinant GTPase– glutathione S-transferase (GST), and
fragments were separated on 1% agarose (Bio-Rad) gels. Images were ana- each TcdB were mixed in a reaction buffer. After 1 h of incubation at 37°C,
lyzed with the BioNumerics software, v5.1 (Applied Maths), and the patterns the proteins were separated by SDS-PAGE. Glycosylation of GTPase was
were compared to those deposited in the database of the National Microbiol- analyzed by phosphorimaging. For graphical representation, band density
ogy Laboratory, Public Health Agency of Canada (Michael R. Mulvey). was measured with ImageQuant TL. For the Western blotting, the same
PCR-restriction fragment length polymorphism (RFLP) analysis reactions and conditions were used and assays were performed using
and ribotyping. For toxinotyping, C. difficile VPI 10463 was used as a UDP-glucose (Sigma). After protein separation by SDS-PAGE, the pro-
control according to the published protocols (30). For ribotyping, primer teins were transferred to PVDF membranes. The GTPase was detected
sequences and reaction conditions were taken from the work of Bidet et al. with monoclonal anti-RhoA antibody (Abcam; ab54835) by Western
(31). blotting. The control RhoA-GST proteins were stained with Coomassie
Whole-genome sequencing, MLST, and PaLoc/TcdB comparison. blue.
Whole-genome sequences were obtained using multiplexed paired-end RhoA, Rac1, and Cdc42 GTPase activation assays. The TcdB ability
libraries and the sequencing-by-synthesis HiSeq platform (Illumina). to inactivate GTPases was determined on confluent 3T3 fibroblasts grown
Reads were assembled using Velvet (32), contigs of 300 bp were scaf- in Dulbecco modified Eagle medium (DMEM) supplemented with 5%
folded with SSPACE (33), and gaps were filled using GapFiller (34). The fetal bovine serum (FBS) (Sigma). For the pulldown steps, GTP-RhoA was
resulting scaffolds were ordered using Mauve (35) and the genomes of precipitated with GST-tagged Rho binding domain (RBD) and GTP-Rac1
reference strain R20291 (NAP1/RT027/ST01) or M68 (NAP9/RT017/ and GTP-Cdc42 were precipitated with GST-p21 binding domain (PBD).
March 2016 Volume 84 Number 3 Infection and Immunity iai.asm.org 857
Downloaded from http://iai.asm.org/ on July 12, 2017 by UNIV DE COSTA RICA SISTEMA DE
Quesada-Gómez et al.
Confluent 3T3 fibroblasts cultured in 6-well plates were intoxicated with
0.2 nM TcdB of NAP1V and NAP1 strains under the conditions indicated
in the figure legends. After the intoxication, the cells were treated as pre-
viously described (47). Briefly, cells were washed with phosphate-buffered
saline (PBS) and lysed with precipitation buffer. Lysates were centrifuged
and incubated with Rho binding domain (RBD) of the human Rhotekin
protein, which had been expressed as a GST fusion protein (RBD-GST),
or Rac/Cdc42 (p21) binding domain (PBD) of the human p21-activated
kinase 1 protein, which had been expressed as a GST fusion protein (PBD-
GST). Active proteins were pulled down by centrifugation, resolved by
SDS-PAGE, and transferred to PVDF membranes. GTPases were detected
using anti-RhoA (Abcam; ab54835), anti-Rac1 (Abcam; ab33186), or an-
ti-Cdc42 (Abcam; ab41429) antibody by Western blotting. For detection
of RhoA glucosylation using the monoclonal antibody, HeLa cells, Vero
cells, and 3T3 fibroblasts were intoxicated with 0.2 nM TcdB from NAP1
or NAP1V strains for 6 and 24 h. After intoxication, cells were lysed in 2%
SDS and 20 g of each lysate was separated by 10% SDS-PAGE, electro-
transferred to PVDF membranes, and probed with the anti-RhoA mono-
clonal antibody.
Structural analysis and TcdB GTD modeling. The homology models
were made using Modeler and Chimera bioinformatics tools, as described
previously (49), based on the tcdB sequences of NAP1, NAP1V, and VPI
10463 strains. Adjustments to the multiple-sequence alignment con-
structed by using ClustalW (50) were made based on the structure-based
alignment performed by superimposing the structures of TcdB proteins.
Kinetics of CPE induced by toxins. Confluent HeLa cells were intoxi-
cated with 10 pM TcdBNAP1 and TcdB . The percentage of round cells in
FIG 1 PFGE and core genome-based analysis of the phylogenetic relatedness
NAP1V
each well was evaluated every hour for a period of 12 h and then at 24 h. of NAP1 strains analyzed. (A) Two different SmaI macrorestriction patterns
were detected, 001 and 279. The variant NAP1 strain was allocated to the latter
Determination of TNF- induction. Confluent RAW 264.7 cells were group, and thus, we named it NAP1V. NAP4 (TcdA
 TcdB) and NAP9
intoxicated with 0.5 nM TcdBNAP1 and TcdBNAP1V for 6 h. The concen- (TcdA TcdB) strains were included in the dendrogram for comparative
tration of tumor necrosis factor alpha (TNF-) in the supernatants was de- purposes. (B) A phylogenetic reconstruction based on core SNPs revealed that
termined by commercial enzyme-linked immunosorbent assay (ELISA) ac- the NAP1V strain was more closely related to NAP1 reference strains and
cording to the instructions of the manufacturer (R&D Systems). clinical isolates than to contemporary NAP2, NAP4, NAP6, and NAP9 clinical
Murine ileal loop model. Animal experimental procedures were ap- isolates (four-digit identifiers after the PFGE pattern) and to CD_630 and VPI
proved (CICUA-38-14) by the University of Costa Rica Animal Care and 10463 strains. The genomes of reference NAP1/RT027 (CD196, R20291, B1-
Use Committee. Male Swiss mice of 20 to 25 g were subjected to fasting 12, and BI17), NAP9/RT017 (M68), CD_630, and VPI 10463 strains were
overnight and anesthetized with ketamine (60 mg/kg of body weight) and included in the analysis to validate the results of the PFGE typing method.
xylazine (5 mg/kg) (Eremer Pharma). Through a midline laparotomy, an
ileal loop was ligated, and 10 g of each toxin or the corresponding con-
trol solution was injected. Mice were sacrificed 4 h after inoculation, and
the length and weight of the intestinal loops were recorded (51). The netic reconstruction based on core single nucleotide polymor-
neutrophil accumulation in homogenized ileal tissue was evaluated phisms (SNPs) revealed that NAP1V is more closely related to
through determination of myeloperoxidase (MPO) activity with the o- historical and epidemic NAP1/RT027/ST01 strains than to TcdA-
dianisidine dihydrochloride (Sigma) and H2O2 assay (52). The concen- negative NAP9/RT017/ST37 strains with genes encoding variant
trations of the proinflammatory cytokines interleukin-1 (IL-1), IL-6, TcdB (53) (Fig. 1B). This relationship to the clade 2 of hyperviru-
and TNF- in ileal tissue homogenates were determined by commercial lent lineages postulated by Griffiths et al. (54) was confirmed
ELISA according to the instructions of the manufacturer (R&D Systems). through ribotyping and MLST, as both the NAP1V (RT019/ST67)
Nucleotide sequence accession numbers. All reads were deposited at and the NAP1 (RT027/ST01) strains belong to this clade (54). In
the European Nucleotide Archive in study PRJEB5034 under the run ac- agreement with this finding, the NAP1 strain carries the tcdA,
cession numbers ERR467598 (LIBA-5784), ERR467603 (LIBA-6277), VtcdB, and cdtB genes and presents an 18-bp deletion and a single-
ERR467599 (LIBA-5785), ERR467582 (LIBA-5757), and ERR467583
(LIBA-5758). base-pair deletion at position 117 in tcdC, characteristic of NAP1/
RT027 strains. On the other hand, NAP1V was not resistant to
RESULTS fluoroquinolones and did not present the amino acid transition
A NAP1 strain inducing a variant cytopathic effect. In a prelim- from Thr82 (ACT) to Ile (ATT) in GyrA as observed in classic
inary study performed on a collection of clinical isolates from NAP1/RT027 strains (55).
tertiary care hospitals, the presence of the NAP1 genotype was The levels of secreted TcdA and TcdB and the expression of
reported (24). Among 33 NAP1 isolates analyzed, we found a tcdA and tcdB transcripts were measured to determine whether the
strain whose supernatant induced a cytopathic effect (CPE) dif- NAP1V strain produces increased amounts of toxin relative to the
ferent from the classic arborizing CPE observed for the other classical epidemic NAP1/RT027 strain. Titration of toxin activity
NAP1 strains (data not shown). PFGE analysis indicated that the in bacterium-free supernatants indicated that the NAP1 and
SmaI macrorestriction pattern of this particular variant strain was NAP1V strains induced similar CPE50 titers (Fig. 2A), and in
279, while the pattern of the other NAP1 strains was 001 (Fig. 1A). agreement, the levels of secreted toxins were similar for the two
This strain, here designated NAP1V, was analyzed by whole-ge- strains (Fig. 2B). tcdA and tcdB mRNAs were quantified by real-
nome sequencing and comparative genome analyses. A phyloge- time quantitative PCR. The level of both transcripts was signifi-
858 iai.asm.org Infection and Immunity March 2016 Volume 84 Number 3
Downloaded from http://iai.asm.org/ on July 12, 2017 by UNIV DE COSTA RICA SISTEMA DE
Variant TcdB in C. difficile Hypervirulent Clade 2
FIG 2 The NAP1V strain produces increased amounts of TcdA and TcdB. (A)
Twenty-four-hour bacterium-free supernatants were titrated by 10-fold dilu-
tions on HeLa cell monolayers. Twenty-four hours after inoculation with the
indicated supernatant, the dilution inducing a cytopathic effect (CPE) in 50%
of the cells was calculated by visual examination under the microscope. Each
bar represents the means 	 standard deviations of CPE50 from three replicates.
*, P 
 0.05 (one-way Kruskal-Wallis test followed by Mann-Whitney U test).
(B) Proteins from bacterium-free supernatants were precipitated and sepa- FIG 3 The NAP1V strain induces a variant CPE. HeLa cells, Vero cells, and
rated by 7.5% SDS-PAGE. Proteins were electrotransferred to PVDF mem- 3T3 fibroblasts were treated with TcdBNAP1 and TcdBNAP1V. Cells were treated
branes and probed with monoclonal antibodies to TcdA and TcdB. (C) Total until a CPE was achieved in 100% of the cells. Images obtained by phase-
RNA was prepared from the indicated strains at 5, 8, and 24 h during the contrast microscopy show rounding as well as detachment of the cells caused
growth curve. RNA was retrotranscribed, and cDNA was quantified by RT- only by TcdBNAP1V. In order to see actin cytoskeleton modifications, fibro-
PCR using primers specific for tcdA and tcdB. Results displayed represent the blasts were stained with fluorescein isothiocyanate-phalloidin. TcdBNAP1-
means 	 standard deviations from three independent experiments. *, P 
 0.05 treated cells show a classical arborizing effect. TcdBNAP1V-treated cells that had
compared to NAP4; **, P 
 0.05 compared to NAP1 (one-way analysis of not been detached were fixed and show cell rounding without an arborizing
variance with Bonferroni’s correction). effect. Cells were visualized with a Nikon Eclipse 80i fluorescence microscope.
Effects on cells induced by the toxins were visualized by scanning electron
microscopy (SEM) using an S-3700N (Hitachi) electron microscope.
cantly higher in both the NAP1 and NAP1V strains than in control
strains at all times tested (Fig. 2C). Interestingly, the NAP1V strain
produces even more toxin transcripts than does the NAP1 coun- vitro assay. TcdBNAP1 modified a panel of substrates characteristic
terpart, a detail that should be considered in future experiments of classic TcdB proteins, with RhoA, Rac1, and Cdc42 being read-
dealing with the regulation of these genes. Altogether, these results ily glucosylated (Fig. 4A). Interestingly, we observed modification
demonstrate that the NAP1V strain is closely related to the epi- to a lesser extent of Rap1, Rap2, and R-Ras, which has not been
demic NAP1/RT027 strains but displays distinctive genotypic and reported previously for classic TcdB proteins inducing arborizing
phenotypic characteristics associated with TcdB that we further CPE. On the other hand, TcdBNAP1V glucosylated Rac1, but the
explored. glucosylation of RhoA and Cdc42 was significantly diminished
TcdBNAP1V induces a variant CPE related to a distinct (Fig. 4B). As with TcdBNAP1, TcdBNAP1V was able to glucosylate
GTPase glucosylation pattern. To analyze the cytopathic char- Rap1, Rap2, and R-Ras at low levels.
acteristics of toxin B from NAP1V (TcdBNAP1V) and compare the To confirm the panel of substrates modified by the toxins, we
toxin with those from a classic NAP1 strain (TcdBNAP1), both monitored the ex vivo glucosylation of RhoA, Rac1, and Cdc42
toxins were purified. After intoxication of HeLa cells, Vero cells, after intoxication of cultured cells by pulldown assays. When 3T3
and 3T3 fibroblasts with TcdBNAP1, the classical arborizing CPE cells were incubated with TcdBNAP1, Rho-GTP was undetectable
was observed (Fig. 3). In contrast, TcdBNAP1V induced cell round- at 6 h (Fig. 4C). In contrast, Rho-GTP was detected in cells intox-
ing and detachment but no protrusions or arborizing effects (Fig. icated with TcdBNAP1V for up to 24 h, confirming the lack of mod-
3). Hence, TcdBNAP1V was responsible for the variant CPE pro- ification of this small GTPase in the in vitro assay (Fig. 4C). Both
duced by NAP1V supernatants. TcdB proteins inactivated Rac1 after 6 and 24 h of treatment, again
Next, we determined the glucosylation pattern of TcdBNAP1 confirming the results of the in vitro glucosylation assay (Fig. 4C).
and TcdBNAP1V using a panel of small GTPases and a radioactive in Additionally, the level of Cdc42-GTP exhibited a significant and
March 2016 Volume 84 Number 3 Infection and Immunity iai.asm.org 859
Downloaded from http://iai.asm.org/ on July 12, 2017 by UNIV DE COSTA RICA SISTEMA DE
Quesada-Gómez et al.
FIG 5 A monoclonal antibody to RhoA detects modification of this small
GTPase by TcdBNAP1 but not by TcdBNAP1V. (A) Recombinant purified RhoA
was incubated with TcdBNAP1 and TcdBNAP1V in the presence of UDP-glucose.
The preparations were separated by SDS-PAGE and detected by Coomassie
blue staining. Parallel samples were transferred to PVDF membranes and de-
veloped by Western blotting (WB) using the monoclonal antibody to RhoA.
(B) HeLa cells, 3T3 fibroblasts, and Vero cells were treated for the indicated
times with TcdBNAP1 or TcdBNAP1V. Cell lysates of treated cells and nontreated
control cells were separated by SDS-PAGE, transferred to PVDF membranes, and
FIG 4 The NAP1 strain does not glucosylate RhoA. (A) TcdB and revealed with the monoclonal antibody to RhoA by Western blotting. As a loadingV NAP1
TcdB were tested for their ability to glycosylate a panel of recombinant control, membranes were also revealed with a monoclonal antibody to Cdc42.NAP1V
GTPases using UDP-[14C]glucose as a cosubstrate. Labeled bands were de-
tected by phosphorimaging analysis. (B) The band intensities of the GTPase
glycosylation were quantified by densitometry. Each experiment was normal- able tool to monitor Rho modification by large clostridial cytotox-
ized to Rac1 signal. Means 	 standard deviations from three independent
experiments are shown. (C) Effect of TcdB and TcdB on the activa- ins. To further explore this concept, different cell lines (HeLa cells,NAP1 NAP1V
tion state of small GTPases. 3T3 fibroblasts were intoxicated with TcdBNAP1 3T3 fibroblasts, and Vero cells) were intoxicated for 6 and 24 h
and TcdBNAP1V for the indicated times. After treatment, cells were lysed. One with either TcdBNAP1 or TcdBNAP1V. RhoA was detected only in
part of the lysates was used as a control for total amount of GTPases, and the lysates prepared from cells intoxicated with TcdB (Fig. 5B),
other one was incubated with PBD-GST or RBD-GST-Sepharose beads. Active NAP1V
proteins were pulled down and analyzed by Western blotting. GTPases were confirming that the ability to target this small GTPase is the main
detected using anti-RhoA, anti-Rac, and anti-Cdc42, respectively. Cells treated difference at the level of substrates between the two toxins.
with TcdBNAP1 show inactivation of RhoA, whereas cells intoxicated with TcdBNAP1V sequence combines the enzymatic domain of
TcdBNAP1V do not. Cytotoxic necrotizing factor 1 (CNF) from Escherichia coli variant toxins with the receptor-binding domain of TcdB from
was used as a positive control for GTPase activation. Negative-control cells the hypervirulent clade 2. Since TcdB
were left untreated. NAP1V
clearly presents dif-
ferent phenotypic behavior than TcdBNAP1, we focused on differ-
ences at the sequence level. Phylogenetic analysis indicates that the
PaLoc of the NAP1V strain is more closely related to that of classic
consistent decrease at 6 h after intoxication with TcdBNAP1 and NAP1/RT027 strains than to TcdA-negative strains carrying vari-
Cdc42-GTP completely disappeared after 24 h of treatment (Fig. ant TcdB proteins (Fig. 6). Next, we determined the toxinotype of
4C). However, TcdBNAP1V was able to decrease the level of Cdc42- the NAP1V strain. The tcdB polymorphisms of the B1 fragment
GTP only after 24 h of intoxication, indicating that this small (containing the coding region for the TcdB glucosyltransferase
GTPase is not a preferred substrate of this variant TcdB (Fig. 4C). domain) (Fig. 7A) from NAP1V were identical to those of TcdA-
Interestingly, no signal was detected in the control for total Rho negative strain NAP9/RT017 and different from that of NAP1/
(loading control) in cells treated with TcdBNAP1, indicating either RT027 (Fig. 7B). The restriction patterns for tcdA were the same
that the protein is degraded after glucosylation or that, alterna- for the two strains (Fig. 7C). Thus, the toxinotype of the NAP1V
tively, the antibody to Rho that we used in this work does not strain (toxinotype XXIII) is not the classic one found in NAP1
recognize the glucosylated isoform (Fig. 4C). To distinguish be- strains (toxinotype III) and rather coincides with the toxinotype
tween these two possibilities, recombinant Rho was incubated present in TcdA-positive strains carrying variant TcdB proteins.
with either TcdBNAP1 or TcdBNAP1V in the presence of UDP-glu- A detailed analysis and comparison of the sequences from the
cose. Rho was detected by Coomassie blue staining after treatment different TcdB proteins indicates that the primary sequence of the
with both toxins but was not detected by Western blotting after glucosyltransferase domain of TcdBNAP1V is more closely related
treatment with TcdBNAP1 (Fig. 5A). This result indicates that the to the corresponding region of TcdB proteins inducing a variant
monoclonal antibody used does not interact with glucosylated CPE than to that of TcdB proteins inducing a classic arborizing
Rho and confirms the fact that TcdBNAP1V does not modify this CPE (Fig. 8A). Indeed, the identity in the first 546 amino acid
small GTPase. Thus, this monoclonal antibody represents a valu- residues between TcdBNAP1V and TcdBNAP9 is 100%, whereas that
860 iai.asm.org Infection and Immunity March 2016 Volume 84 Number 3
Downloaded from http://iai.asm.org/ on July 12, 2017 by UNIV DE COSTA RICA SISTEMA DE
Variant TcdB in C. difficile Hypervirulent Clade 2
FIG 6 Phylogenetic relationship of the pathogenicity locus (PaLoc) of NAP1V
and NAP1 strains to that of reference strains (VPI 10463, CD_630, NAP7/
RT078_M120, NAP9/RT017_M68, and NAP1/RT027_R20291) and clinical
isolates (NAP4, NAP6, and NAP9). The PaLoc sequence of NAP1V clustered
together with those of clinical (NAP1-001_5768) and reference NAP1/027
(R20291) strains rather than with sequences from TcdA TcdB variant
strains (NAP9/017).
between TcdBNAP1V and TcdBNAP1 is 80%. Furthermore, the
identity in the substrate specificity domain (amino acids 365 to
516) between TcdBNAP1V and TcdBNAP1 is only 62%. The glu-
cosyltransferase domain (GTD) sequences of TcdB FIG 7 Toxinotyping of NAP1 strains. The polymorphisms obtained from theNAP1V and B1 and A3 regions of the tcdA and tcdB genes were analyzed by digestion with
TcdBNAP1 were analyzed in the context of the VPI 10463 refer- AccI (A) and HindIII (H) restriction enzymes. (A) Representation of the am-
ence strain (identical to the 630 reference strain). Although the plified regions. (B) The restriction polymorphisms of the tcdB fragments from
core residues of the GTDs are conserved between TcdB the NAP1V (toxinotype XXII) and the tcdA-negative tcdB NAP9 (toxinotypeNAP1V
and TcdB VIII) strains are indistinguishable and different from the corresponding pat-VPI10463, the surface residues are divergent (Fig. 8B). tern from the NAP1 strain. (C) The NAP1 and NAP1 (toxinotype III) strains
These divergent residues are predicted to be involved in the sub- Vhave the same restriction pattern of the tcdA fragment.
strate affinity of the GTD. In contrast, the GTDs of TcdBNAP1 and
TcdBVPI10463 are very similar (Fig. 8B). The CROPs domain of
TcdBNAP1V is highly similar to the corresponding region of TcdB of myeloperoxidase (MPO) activity as an indicator of tissue neu-
proteins from classic NAP1 strains (Fig. 8A). The identity in this trophil infiltration and the levels of IL-1 and IL-6 to indicate
region (amino acids 1645 to 2366) between TcdBNAP1V and immune activation at the ileal tissue level. TcdBNAP1 caused a sta-
TcdBNAP1 is 99%. tistically significant increase in MPO activity, whereas TcdBNAP1V
These data indicate the possibility of a recombination event elicited a reaction undistinguishable from that of the control (Fig.
that led to the sequence encoding TcdBNAP1V. To explore this, we 9C). The levels of IL-1 and IL-6 were significantly increased by
applied a Bayesian approach to infer changes in tree topologies TcdBNAP1, and again, TcdBNAP1V was unable to induce any reac-
and evolutionary rates using TcdB sequences from strains NAP1V, tion (Fig. 9C).
R20291 (NAP1/RT027), M120 (NAP7/RT078), and NAP9/
RT017. Figure 8C shows the dominant tree topology within each DISCUSSION
partition of the alignment, with NAP1V and NAP9/RT017 being The increase in rate and severity of CDI has been linked to the
the closest neighbors with 100% probability in the first 635 resi- emergence and spread of the epidemic NAP1 strain (19, 56). This
dues of the alignment and NAP1V and R20291 (NAP1/RT027) genotype has acquired several genetic determinants that contrib-
showing the same result from residue 1465 onward. These topol- ute to its increased virulence; among these, the overproduction of
ogies imply a possible recombination event between TcdB pro- toxins (linked to mutations in the tcdC gene), the presence of a
teins from NAP1/RT027 and NAP9/RT017 strains. binary toxin, and resistance to fluoroquinolones have been con-
TcdBNAP1V and TcdBNAP1 have similar cytopathic potencies sidered to play an important role (21, 22, 56). It has been shown
but exert different biological effects. To assess the cytopathic po- that TcdB is essential for C. difficile virulence and that its glucosyl-
tency of TcdBNAP1V and TcdBNAP1, both toxins were titrated on transferase activity is required for activity in an ileal loop model
HeLa cells. The two toxins elicited similar kinetic profiles of cell (57, 58). In this context, previous studies have concluded that
intoxication, indicating that their cytopathic potencies are similar variations in the sequence of this toxin lead to an augmented cy-
(Fig. 9A). TcdBNAP1V and TcdBNAP1 were also tested for their abil- topathic potency and play a crucial role in the increased virulence
ity to induce TNF- production by RAW murine macrophages. of NAP1 strains. This increased cytotoxicity is due to a more effi-
Despite the similar cytopathic potencies, the release of this cyto- cient delivery of the enzymatic domain to the cytosol (23, 59).
kine was statistically higher in cells treated with TcdBNAP1 (Fig. However, the role of the substrate pattern of TcdB within the
9B). The pathogenic potential of both toxins was assayed in the hypervirulent clade 2 has not been analyzed. In the present study,
murine ligated ileal loop model. We measured the concentration we describe a NAP1 isolate from an outbreak setting that produces
March 2016 Volume 84 Number 3 Infection and Immunity iai.asm.org 861
Downloaded from http://iai.asm.org/ on July 12, 2017 by UNIV DE COSTA RICA SISTEMA DE
Quesada-Gómez et al.
FIG 8 TcdBNAP1V shares with TcdBNAP1 the receptor-binding domain but not the enzymatic domain. (A) Sequence alignment of (i) variant toxins B from
NAP9_M68, CD_1470, and CD_8864 reference strains inducing a variant (V) CPE; (ii) TcdBNAP1 from a clinical isolate and a reference epidemic NAP1/RT027
strain (R20291) inducing a classic (C) CPE; and (iii) TcdBNAP1V. Black lines represent disagreements in the sequence of TcdBVPI10463, which was selected as a
reference for the alignment. The blue box highlights a distinct glucosyltransferase region shared between NAP1V and other variant strains. The green box shows
sequence stretches in the repetitive CROPs domains shared between TcdBNAP1, TcdBR20291, and TcdBNAP1V. (B) Comparison of the TcdBNAP1 and TcdBNAP1V
sequences in the context of the TcdB GTD structure (PDB 2BVM, VPI 10463 sequence). Sequence conservation on the putative GTPase-binding face compared
to the GTD from C. difficile VPI 10463 with NAP1 and NAP1V TcdB GTD structures is shown (red, conserved; blue, not conserved). UDP-glucose is depicted in
white in the GTD active site. (C) Resulting recombination detection graphs using TcdB sequences from strains NAP9 (M68, RT017 reference strain), R20291
(epidemic NAP1/RT027 reference strain), NAP7 (epidemic M120, RT078 reference strain), and NAP1V. Signs of possible recombination events are represented
as changes in the topology graphs (first row) that appear at the most probable topology between the segments. The cross of the topological lines (green and red
lines) indicates recombination breakpoints. Resulting trees are compatible with a scenario in which TcdBNAP1V emerged through recombination of tcdB
sequences from NAP9 and NAP1 strains.
a variant TcdB. Our goals were to understand the differences in strain belonging to the NAP1/RT027 genotype, the NAP1V/RT019
the CPEs induced by this toxin and the potential role of the panel macrorestriction pattern differs from that of the classical NAP1/
of modified substrates in the biological effects induced by TcdB RT027 strains and its toxinotype differs due to variations within
proteins secreted by strains from the hypervirulent clade 2 and to the tcdB-encoded N-terminal region. Indeed, the digestion pat-
elucidate the emergence of this NAP1 variant strain. tern of the amplified B1 fragment coding for the catalytic region of
Due to the PFGE classification and the particular CPE induced TcdBNAP1V was indistinguishable from the corresponding one
by bacterium-free supernatants derived from the NAP1V strain, presented in TcdA-negative strains. Interestingly, the NAP1V
we reasoned that this isolate would have phenotypic characteris- strain belongs to the toxinotype XXII, which has been described in
tics associated with important differences at the level of TcdB. The isolates harboring variant TcdB proteins (17, 61).
NAP1V strain is, in fact, a toxin-overproducing isolate, harboring The CPE induced by TcdBNAP1V, characterized by rounding
deletions in tcdC and carrying the binary toxin gene. Nonetheless, and detachment of intoxicated cells, resembles the effect induced
it is not resistant to fluoroquinolones, since it does not harbor the by Clostridium sordellii lethal toxin (TcsL) and C. difficile variant
typical mutation in gyrA found in NAP1 strains (60). Despite the TcdB proteins (13, 62).This effect, referred to as variant CPE, was
862 iai.asm.org Infection and Immunity March 2016 Volume 84 Number 3
Downloaded from http://iai.asm.org/ on July 12, 2017 by UNIV DE COSTA RICA SISTEMA DE
Variant TcdB in C. difficile Hypervirulent Clade 2
variant TcdB proteins, but in the case of TcdBNAP1V, there is par-
tial modification of this small GTPase. However, a complete glu-
cosylation of Cdc42, as determined by pulldown assays, was de-
tected only after 24 h of intoxication, indicating that glucosylation
of this protein is not involved in induction of the variant CPE,
which appears in the first few hours after addition of the toxin.
There is a clear correlation between the small GTPases modi-
fied and the type of CPE induced by TcdBNAP1V and variant TcdB
proteins from TcdA-negative strains. This concordance is in
agreement with the primary sequence of the toxins since the GTD
of TcdBNAP1V has a high identity to the corresponding domain
found in variant TcdB proteins. On the other hand, the autopro-
cessing domain and the carboxyl-terminal region of TcdBNAP1V
are almost identical to the corresponding regions from TcdBNAP1.
These results, along with sequence comparison, reveal that
TcdBNAP1V is a toxin of the classical NAP1/RT027 genotype but
with modifications within the enzymatic domain.
Recently, a strain belonging to clade 2, RT244/ST41, was re-
ported to display an increased virulence (62). As a NAP1 strain,
RT244/ST41 harbors a binary toxin; however, it does not produce
increased amounts of toxins and is fluoroquinolone susceptible.
These data indicate that the NAP1V/RT019 strain shares more
FIG 9 TcdBNAP1V has the same cytotoxic potency as TcdB but induces features with the classic NAP1/RT027 than RT244/ST41 andNAP1
fewer proinflammatory reactions. (A) HeLa cells were treated with equal con- would then be more closely related to the epidemic strain. Inter-
centrations (10 pM) of purified TcdBNAP1 and TcdBNAP1V. The percentage of estingly, the strain RT244/ST41 genome also seems to encode a
cells showing a toxin-induced CPE was calculated at the indicated times. (B) variant TcdB. A comparative and detailed assessment of the viru-
RAW cells were incubated for 6 h with equal concentrations (0.5 nM) of pu-
rified TcdB and TcdB . After incubation, the amount of TNF- re- lence potential of these three strains would allow one to determineNAP1 NAP1V
leased in the supernatant was determined by ELISA. Each bar represents the the relative contribution of factors such as the presence of binary
mean 	 standard deviation from three independent experiments. (C) Mouse toxin, overproduction of toxins, fluoroquinolone resistance, and
ligated ileal loops were inoculated with 10 g of purified TcdBNAP1 and type of toxin produced to the increased virulence displayed by
TcdBNAP1V for 4 h. After treatment, MPO activity and inflammatory cytokine
(IL-1 and IL-6) levels were determined. Means 	 standard deviations, n  5. members of this clade.
*, P 
 0.05, compared to the groups without asterisk (one-way analysis of Nosocomial outbreaks caused by TcdA-negative strains have
variance with Bonferroni’s correction). increased in the last decade (53, 64, 65). Interestingly, all these
strains have been reported to harbor variant TcdB proteins. This
might be an indication that the biological effects induced by classic
first reported in C. difficile strains that do not produce TcdA. TcdB proteins differ from those induced by variant TcdB proteins
While it has now been described in a wider range of strains (13, 17, and that TcdA-negative strains compensate for the lack of this
47), it had not been previously described within the NAP1/RT027 toxin by using a TcdB with a different panel of substrates. When
genotype; in all these cases, the variant CPE has been attributed to we compared the responses to TcdB on the ligated loop model, we
TcdB (16). The induction of a variant CPE by variant TcdB pro- could indeed find a significant biological difference between the
teins correlates with substrate profiles that differ from the panel two toxins. Whereas TcdBNAP1 induced an immune activation,
targeted by TcdB from reference strain VPI 10463. Whereas the TcdBNAP1V induced a much milder and almost undetectable re-
latter modifies Rho, Rac, and Cdc42, variant TcdB proteins also sponse. Secretion of TNF- by macrophages has been associated
modify R-Ras, Rap, and Ral (5, 47). A more detailed analysis of the with the glucosyltransferase activity of C. difficile toxins (66), and
consequences of small GTPase modification indicated that R-Ras since TcdBNAP1V and TcdBNAP1 have a high degree of identity in
glucosylation and transient RhoA activation determine the ap- the regions determining receptor binding and entrance to the cell
pearance of a variant CPE since R-Ras glucosylation leads to in- and the two toxins have similar cytopathic potencies, we assume
tegrin inactivation, and as a consequence, focal adhesions disas- that the biological differences detected in our assays are due to a
semble, causing detachment (47). In contrast to TcdB from strain differential panel of substrates glucosylated. Since the main differ-
VPI 10463, TcdBNAP1, which induces a classic arborizing CPE, ence in substrates is at the level of RhoA modification, we hypoth-
seems to have an extended substrate pattern since it was able to esize that glucosylation of this small GTPase enhances the proin-
modify Rap and R-Ras. These additional targeted substrates might flammatory response induced by C. difficile toxins. The use of a
have a role in the increased biological effects induced by TcdBNAP1 panel of purified toxins with differing substrate panels like the
(57, 59). Interestingly, the main difference at the level of substrates ones indicated in this article in a wide range of experimental mod-
between TcdBNAP1V and TcdBNAP1 is the modification of RhoA. els would allow the dissection of the relevance of the modification
Thus, the ability to target this small GTPase also seems to play an of different small GTPases in the outcome of CDI.
important role in defining the type of CPE that will be induced,
since RhoA-modifying toxins induce an arborizing CPE and non- ACKNOWLEDGMENTS
RhoA-modifying toxins induce a variant CPE (63). In addition, We thank Michael R. Mulvey, George Golding, and Tim Du (National
previous studies have shown that Cdc42 is not glycosylated by Microbiology Laboratory, Winnipeg, MB, Canada) for access to the NAP
March 2016 Volume 84 Number 3 Infection and Immunity iai.asm.org 863
Downloaded from http://iai.asm.org/ on July 12, 2017 by UNIV DE COSTA RICA SISTEMA DE
Quesada-Gómez et al.
type database and technical assistance with the PFGE analyses and Ed- 2001. Variant toxin B and a functional toxin A produced by Clostridium
gardo Moreno for critical reading of the manuscript. difficile C34. FEMS Microbiol Lett 198:171–176. http://dx.doi.org/10.1111
We declare no conflicts of interest. /j.1574-6968.2001.tb10638.x.
The funders had no role in study design, data collection and interpre- 15. Rupnik M, Kato N, Grabnar M, Kato H. 2003. New types of toxin
tation, or the decision to submit the work for publication. A-negative, toxin B-positive strains among Clostridium difficile isolates
from Asia. J Clin Microbiol 41:1118 –1125. http://dx.doi.org/10.1128
FUNDING INFORMATION /JCM.41.3.1118-1125.2003.16. Huelsenbeck J, Dreger S, Gerhard R, Barth H, Just I, Genth H. 2007.
The National Rectors Council, Costa Rica (CONARE) provided funding Difference in the cytotoxic effects of toxin B from Clostridium difficile
to Carlos Quesada-Gómez under grant numbers 803-B1-654 and 803-B4- strain VPI 10463 and toxin B from variant Clostridium difficile strain 1470.
652. The Postgraduate Studies System and Vice-rectory for Research of Infect Immun 75:801– 809. http://dx.doi.org/10.1128/IAI.01705-06.
the University of Costa Rica provided funding to Carlos Quesada-Gómez 17. Rupnik M. 2008. Heterogeneity of large clostridial toxins: importance of
under grant numbers 803-B5-107 and 803-B5-108. The National Council Clostridium difficile toxinotypes. FEMS Microbiol Rev 32:541–555. http:
of Science and Technology (CONICIT) provided funding to Esteban //dx.doi.org/10.1111/j.1574-6976.2008.00110.x.
Chaves-Olarte under grant number FORINVES FV-0004-13. HHS | Na- 18. Loo VG, Poirier L, Miller MA, Oughton M, Libman MD, Michaud S,
Bourgault A-M, Nguyen T, Frenette C, Kelly M, Vibien A, Brassard P,
tional Institutes of Health (NIH) provided funding to Borden Lacy under
Fenn S, Dewar K, Hudson TJ, Horn R, René P, Monczak Y, Dascal A.
grant number R01AI095755. Wellcome Trust provided funding to Trevor 2005. A predominantly clonal multi-institutional outbreak of Clostridium
D Lawley under grant number 098051. difficile-associated diarrhea with high morbidity and mortality. N Engl J
Med 353:2442–2449. http://dx.doi.org/10.1056/NEJMoa051639.
REFERENCES 19. He M, Miyajima F, Roberts P, Ellison L, Pickard DJ, Martin MJ,
1. Kelly CP, LaMont JT. 2008. Clostridium difficile—more difficult than Connor TR, Harris SR, Fairley D, Bamford KB, D’Arc S, Brazier J,
ever. N Engl J Med 359:1932–1940. http://dx.doi.org/10.1056 Brown D, Coia JE, Douce G, Gerding D, Kim HJ, Koh TH, Kato H,
/NEJMra0707500. Senoh M, Louie T, Michell S, Butt E, Peacock SJ, Brown NM, Riley T,
2. Voth D, Ballard J. 2005. Clostridium difficile toxins: mechanism of action Songer G, Wilcox M, Pirmohamed M, Kuijper E, Hawkey P, Wren BW,
and role in disease. Clin Microbiol Rev 18:247–263. http://dx.doi.org/10 Dougan G, Parkhill J, Lawley TD. 2013. Emergence and global spread of
.1128/CMR.18.2.247-263.2005. epidemic healthcare-associated Clostridium difficile. Nat Genet 45:109 –
3. Hundsberger T, Braun V, Weidmann M, Leukel P, Sauerborn M, von 113. http://dx.doi.org/10.1038/ng.2478.
Eichel-Streiber C. 1997. Transcription analysis of the genes tcdA-E of the 20. Drudy D, Quinn T, O’Mahony R, Kyne L, O’Gaora P, Fanning S. 2006.
pathogenicity locus of Clostridium difficile. Eur J Biochem 244:735–742. High-level resistance to moxifloxacin and gatifloxacin associated with a
http://dx.doi.org/10.1111/j.1432-1033.1997.t01-1-00735.x. novel mutation in gyrB in toxin-A-negative, toxin-B-positive Clostridium
4. Spigaglia P, Mastrantonio P. 2002. Molecular analysis of the patho- difficile. J Antimicrob Chemother 58:1264 –1267. http://dx.doi.org/10
genicity locus and polymorphism in the putative negative regulator of .1093/jac/dkl398.
toxin production (TcdC) among Clostridium difficile clinical isolates. J 21. Warny M, Pepin J, Fang A, Killgore G, Thompson A, Brazier J, Frost
Clin Microbiol 40:3470 –3475. http://dx.doi.org/10.1128/JCM.40.9.3470 E, McDonald LC. 2005. Toxin production by an emerging strain of
-3475.2002. Clostridium difficile associated with outbreaks of severe disease in North
5. Just I, Selzer J, Wilm M, von Eichel-Streiber C, Mann M, Aktories K. America and Europe. Lancet 366:1079 –1084. http://dx.doi.org/10.1016
1995. Glucosylation of Rho proteins by Clostridium difficile toxin B. Na- /S0140-6736(05)67420-X.
ture 375:500 –503. http://dx.doi.org/10.1038/375500a0. 22. Merrigan M, Venugopal A, Mallozzi M, Roxas B, Viswanathan VK,
6. von Eichel-Streiber C, Sauerborn M, Kuramitsu HK. 1992. Evidence for Johnson S, Gerding DN, Vedantam G. 2010. Human hypervirulent
a modular structure of the homologous repetitive C-terminal carbohy- Clostridium difficile strains exhibit increased sporulation as well as robust
drate-binding sites of Clostridium difficile toxins and Streptococcus mutans toxin production. J Bacteriol 192:4904 – 4911. http://dx.doi.org/10.1128
glucosyltransferases. J Bacteriol 174:6707– 6710. /JB.00445-10.
7. Papatheodorou P, Zamboglou C, Genisyuerek S, Guttenberg G, Akto- 23. Lanis JM, Barua S, Ballard JD. 2010. Variations in TcdB activity and the
ries K. 2010. Clostridial glucosylating toxins enter cells via clathrin- hypervirulence of emerging strains of Clostridium difficile. PLoS Pathog
mediated endocytosis. PLoS One 5:e10673. http://dx.doi.org/10.1371 6:e1001061. http://dx.doi.org/10.1371/journal.ppat.1001061.
/journal.pone.0010673. 24. Quesada-Gómez C, Rodríguez C, Gamboa-Coronado MDM, Rodrí-
8. Barth H, Pfeifer G, Hofmann F, Maier E, Benz R, Aktories K. 2001. Low guez-Cavallini E, Du T, Mulvey MR, Villalobos-Zúñiga M, Boza-
pH-induced formation of ion channels by Clostridium difficile toxin B in Cordero R. 2010. Emergence of Clostridium difficile NAP1 in Latin
target cells. J Biol Chem 276:10670 –10676. http://dx.doi.org/10.1074/jbc America. J Clin Microbiol 48:669 – 670. http://dx.doi.org/10.1128/JCM
.M009445200. .02196-09.
9. Egerer M, Giesemann T, Jank T, Satchell KJF, Aktories K. 2007. Auto- 25. Kato N, Ou CY, Kato H, Bartley SL, Brown VK, Dowell VRJ, Ueno K.
catalytic cleavage of Clostridium difficile toxins A and B depends on cys- 1991. Identification of toxigenic Clostridium difficile by the polymerase
teine protease activity. J Biol Chem 282:25314 –25321. http://dx.doi.org chain reaction. J Clin Microbiol 29:33–37.
/10.1074/jbc.M703062200. 26. Lemee L, Dhalluin A, Testelin S, Mattrat M-A, Maillard K, Lemeland
10. Reineke J, Tenzer S, Rupnik M, Koschinski A, Hasselmayer O, Schrat- J-F, Pons J-L. 2004. Multiplex PCR targeting tpi (triose phosphate
tenholz A, Schild H, von Eichel-Streiber C. 2007. Autocatalytic cleavage isomerase), tcdA (toxin A), and tcdB (toxin B) genes for toxigenic culture
of Clostridium difficile toxin B. Nature 446:415– 419. http://dx.doi.org/10 of Clostridium difficile. J Clin Microbiol 42:5710 –5714. http://dx.doi.org
.1038/nature05622. /10.1128/JCM.42.12.5710-5714.2004.
11. Pfeifer G, Schirmer J, Leemhuis J, Busch C, Meyer DK, Aktories K, 27. Rutherford K, Parkhill J, Crook J, Horsnell T, Rice P, Rajandream
Barth H. 2003. Cellular uptake of Clostridium difficile toxin B. Transloca- M-A, Barrell B. 2000. Artemis: sequence visualization and annotation.
tion of the N-terminal catalytic domain into the cytosol of eukaryotic cells. Bioinformatics 16:944 –945. http://dx.doi.org/10.1093/bioinformatics
J Biol Chem 278:44535– 44541. /16.10.944.
12. Li S, Shi L, Yang Z, Zhang Y, Perez-Cordon G, Huang T, Ramsey J, 28. Killgore G, Thompson A, Johnson S, Brazier J, Kuijper E, Pepin J, Frost
Oezguen N, Savidge TC, Feng H. 2015. Critical roles of Clostridium EH, Savelkoul P, Nicholson B, van den Berg RJ, Kato H, Sambol SP,
difficile toxin B enzymatic activities in pathogenesis. Infect Immun 83: Zukowski W, Woods C, Limbago B, Gerding DN, McDonald LC. 2008.
502–513. http://dx.doi.org/10.1128/IAI.02316-14. Comparison of seven techniques for typing international epidemic strains
13. Chaves-Olarte E, Löw P, Freer E, Norlin T, Weidmann M, von Eichel- of Clostridium difficile: restriction endonuclease analysis, pulsed-field gel
Streiber C, Thelestam M. 1999. A novel cytotoxin from Clostridium electrophoresis, PCR-ribotyping, multilocus sequence typing, multilocus
difficile serogroup F is a functional hybrid between two other large clos- variable-number tandem-repeat analysis, amplified fragment length poly-
tridial cytotoxins. J Biol Chem 274:11046 –11052. http://dx.doi.org/10 morphism, and surface layer protein A gene sequence typing. J Clin Mi-
.1074/jbc.274.16.11046. crobiol 46:431– 437. http://dx.doi.org/10.1128/JCM.01484-07.
14. Mehlig M, Moos M, Braun V, Kalt B, Mahony DE, Eichel-Streiber C. 29. Quesada-Gómez C, López-Ureña D, Acuña-Amador L, Villalobos-
864 iai.asm.org Infection and Immunity March 2016 Volume 84 Number 3
Downloaded from http://iai.asm.org/ on July 12, 2017 by UNIV DE COSTA RICA SISTEMA DE
Variant TcdB in C. difficile Hypervirulent Clade 2
Zúñiga M, Du T, Freire R, Guzmán-Verri C, del Mar Gamboa- toxin A glucosyltransferase activity. J Biol Chem 287:8013– 8020. http://dx
Coronado M, Lawley TD, Moreno E, Mulvey MR, de Castro Brito GA, .doi.org/10.1074/jbc.M111.298414.
Rodríguez-Cavallini E, Rodríguez C, Chaves-Olarte E. 2015. Emergence 50. Li K-B. 2003. ClustalW-MPI: ClustalW analysis using distributed and
of an outbreak-associated Clostridium difficile variant with increased vir- parallel computing. Bioinformatics 19:1585–1586. http://dx.doi.org/10
ulence. J Clin Microbiol 53:1216 –1226. http://dx.doi.org/10.1128/JCM .1093/bioinformatics/btg192.
.03058-14. 51. Pawlowski SW, Calabrese G, Kolling GL, Platts-Mills J, Freire R, Al-
30. Rupnik M, Avesani V, Janc M, von Eichel-Streiber C, Delmée M. 1998. cantaraWarren C, Liu B, Sartor RB, Guerrant RL. 2010. Murine model
A novel toxinotyping scheme and correlation of toxinotypes with sero- of Clostridium difficile infection with aged gnotobiotic C57BL/6 mice and
groups of Clostridium difficile isolates. J Clin Microbiol 36:2240 –2247. a BI/NAP1 strain. J Infect Dis 202:1708 –1712. http://dx.doi.org/10.1086
31. Bidet P, Lalande V, Salauze B, Burghoffer B, Avesani V, Delmee M, /657086.
Rossier A, Barbut F, Petit J-C. 2000. Comparison of PCR-ribotyping, 52. Pulli B, Ali M, Forghani R, Schob S, Hsieh KLC, Wojtkiewicz G,
arbitrarily primed PCR, and pulsed-field gel electrophoresis for typing Linnoila JJ, Chen JW. 2013. Measuring myeloperoxidase activity in bio-
Clostridium difficile. J Clin Microbiol 38:2484 –2487. logical samples. PLoS One 8:e67976. http://dx.doi.org/10.1371/journal
32. Zerbino DR, Birney E. 2008. Velvet: algorithms for de novo short read .pone.0067976.
assembly using de Bruijn graphs. Genome Res 18:821– 829. http://dx.doi 53. Du P, Cao B, Wang J, Li W, Jia H, Zhang W, Lu J, Li Z, Yu H, Chen C,
.org/10.1101/gr.074492.107. Cheng Y. 2014. Sequence variation in tcdA and tcdB of Clostridium difficile:
33. Boetzer M, Henkel CV, Jansen HJ, Butler D, Pirovano W. 2011. ST37 with truncated tcdA is a potential epidemic strain in China. J Clin
Scaffolding pre-assembled contigs using SSPACE. Bioinformatics 27:578 – Microbiol 52:3264 –3270. http://dx.doi.org/10.1128/JCM.03487-13.
579. http://dx.doi.org/10.1093/bioinformatics/btq683. 54. Griffiths D, Fawley W, Kachrimanidou M, Bowden R, Crook DW, Fung
34. Boetzer M, Pirovano W. 2012. Toward almost closed genomes with Gap- R, Golubchik T, Harding RM, Jeffery KJM, Jolley KA, Kirton R, Peto
Filler. Genome Biol 13:R56. http://dx.doi.org/10.1186/gb-2012-13-6-r56. TE, Rees G, Stoesser N, Vaughan A, Walker AS, Young BC, Wilcox M,
35. Rissman AI, Mau B, Biehl BS, Darling AE, Glasner JD, Perna NT. 2009. Dingle KE. 2010. Multilocus sequence typing of Clostridium difficile. J
Reordering contigs of draft genomes using the Mauve aligner. Bioinfor- Clin Microbiol 48:770 –778. http://dx.doi.org/10.1128/JCM.01796-09.
matics 25:2071–2073. http://dx.doi.org/10.1093/bioinformatics/btp356. 55. Walkty A, Boyd DA, Gravel D, Hutchinson J, McGeer A, Moore D,
36. Seemann T. 2014. Prokka: rapid prokaryotic genome annotation. Bioinfor- Simor A, Suh K, Taylor G, Miller M, Mulvey MR. 2010. Molecular
matics 30:2068–2069. http://dx.doi.org/10.1093/bioinformatics/btu153. characterization of moxifloxacin resistance from Canadian Clostridium
37. Treangen TJ, Ondov BD, Koren S, Phillippy AM. 2014. The Harvest difficile clinical isolates. Diagn Microbiol Infect Dis 66:419 – 424. http://dx
suite for rapid core-genome alignment and visualization of thousands of .doi.org/10.1016/j.diagmicrobio.2009.12.002.
intraspecific microbial genomes. Genome Biol 15:524. http://dx.doi.org 56. O’Connor JR, Johnson S, Gerding DN. 2009. Clostridium difficile infec-
/10.1186/s13059-014-0524-x. tion caused by the epidemic BI/NAP1/027 strain. Gastroenterology 136:
38. Larsen MV, Cosentino S, Rasmussen S, Friis C, Hasman H, Marvig RL, 1913–1924. http://dx.doi.org/10.1053/j.gastro.2009.02.073.
Jelsbak L, Sicheritz-Pontén T, Ussery DW, Aarestrup FM, Lund O. 57. Carter GP, Chakravorty A, Pham Nguyen TA, Mileto S, Schreiber F, Li
2012. Multilocus sequence typing of total-genome-sequenced bacteria. J L, Howarth P, Clare S, Cunningham B, Sambol SP, Cheknis A, Figueroa
Clin Microbiol 50:1355–1361. http://dx.doi.org/10.1128/JCM.06094-11. I, Johnson S, Gerding D, Rood JI, Dougan G, Lawley TD, Lyras D. 2015.
39. Katoh K, Standley DM. 2013. MAFFT multiple sequence alignment soft- Defining the roles of TcdA and TcdB in localized gastrointestinal disease,
ware version 7: improvements in performance and usability. Mol Biol Evol systemic organ damage, and the host response during Clostridium difficile
30:772–780. http://dx.doi.org/10.1093/molbev/mst010. infections. mBio 6:e00551-15. http://dx.doi.org/10.1128/mBio.00551-15.
40. Edgar RC. 2004. MUSCLE: multiple sequence alignment with high accu- 58. Yang Z, Zhang Y, Huang T, Feng H. 2015. Glucosyltransferase activity of
racy and high throughput. Nucleic Acids Res 32:1792–1797. http://dx.doi Clostridium difficile toxin B is essential for disease pathogenesis. Gut Mi-
.org/10.1093/nar/gkh340. crobes 6:221–224. http://dx.doi.org/10.1080/19490976.2015.1062965.
41. Guindon S, Dufayard J-F, Lefort V, Anisimova M, Hordijk W, Gascuel 59. Lanis JM, Heinlen LD, James JA, Ballard JD. 2013. Clostridium difficile
O. 2010. New algorithms and methods to estimate maximum-likelihood 027/BI/NAP1 encodes a hypertoxic and antigenically variable form of
phylogenies: assessing the performance of PhyML 3.0. Syst Biol 59:307– TcdB. PLoS Pathog 9:e1003523. http://dx.doi.org/10.1371/journal.ppat
321. http://dx.doi.org/10.1093/sysbio/syq010. .1003523.
42. Chan CX, Beiko RG, Ragan MA. 2006. Detecting recombination in 60. Drudy D, Kyne L, O’Mahony R, Fanning S. 2007. gyrA mutations in
evolving nucleotide sequences. BMC Bioinformatics 7:412. http://dx.doi fluoroquinolone-resistant Clostridium difficile PCR-027. Emerg Infect Dis
.org/10.1186/1471-2105-7-412. 13:504 –505. http://dx.doi.org/10.3201/eid1303.060771.
43. Kuehne SA, Cartman ST, Heap JT, Kelly ML, Cockayne A, Minton NP. 61. Geric B, Rupnik M, Gerding DN, Grabnar M, Johnson S. 2004. Distri-
2010. The role of toxin A and toxin B in Clostridium difficile infection. bution of Clostridium difficile variant toxinotypes and strains with binary
Nature 467:711–713. http://dx.doi.org/10.1038/nature09397. toxin genes among clinical isolates in an American hospital. J Med Micro-
44. Lyras D, O’Connor JR, Howarth PM, Sambol SP, Carter GP, Phu- biol 53:887– 894. http://dx.doi.org/10.1099/jmm.0.45610-0.
moonna T, Poon R, Adams V, Vedantam G, Johnson S, Gerding DN, 62. Lim SK, Stuart RL, Mackin KE, Carter GP, Kotsanas D, Francis MJ,
Rood JI. 2009. Toxin B is essential for virulence of Clostridium difficile. Easton M, Dimovski K, Elliott B, Riley TV, Hogg G, Paul E, Korman
Nature 458:1176 –1179. http://dx.doi.org/10.1038/nature07822. TM, Seemann T, Stinear TP, Lyras D, Jenkin GA. 2014. Emergence of a
45. Livak KJ, Schmittgen TD. 2001. Analysis of relative gene expression data ribotype 244 strain of Clostridium difficile associated with severe disease
using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. and related to the epidemic ribotype 027 strain. Clin Infect Dis 58:1723–
Methods 25:402– 408. http://dx.doi.org/10.1006/meth.2001.1262. 1730. http://dx.doi.org/10.1093/cid/ciu203.
46. Pruitt RN, Chambers MG, Ng KK-S, Ohi MD, Lacy DB. 2010. Structural 63. Jank T, Giesemann T, Aktories K. 2007. Rho-glucosylating Clostridium
organization of the functional domains of Clostridium difficile toxins A difficile toxins A and B: new insights into structure and function. Glyco-
and B. Proc Natl Acad Sci U S A 107:13467–13472. http://dx.doi.org/10 biology 17:15R–22R. http://dx.doi.org/10.1093/glycob/cwm004.
.1073/pnas.1002199107. 64. Drudy D, Harnedy N, Fanning S, Hannan M, Kyne L. 2007. Emergence
47. Chaves-Olarte E, Freer E, Parra A, Guzmán-Verri C, Moreno E, The- and control of fluoroquinolone-resistant, toxin A-negative, toxin B-posi-
lestam M. 2003. R-Ras glucosylation and transient RhoA activation de- tive Clostridium difficile. Infect Control Hosp Epidemiol 28:932–940. http:
termine the cytopathic effect produced by toxin B variants from toxin //dx.doi.org/10.1086/519181.
A-negative strains of Clostridium difficile. J Biol Chem 278:7956 –7963. 65. Elliott B, Squire MM, Thean S, Chang BJ, Brazier JS, Rupnik M, Riley
http://dx.doi.org/10.1074/jbc.M209244200. TV. 2011. New types of toxin A-negative, toxin B-positive strains among
48. Chaves-Olarte E, Weidmann M, Eichel-Streiber C, Thelestam M. 1997. clinical isolates of Clostridium difficile in Australia. J Med Microbiol 60:
Toxins A and B from Clostridium difficile differ with respect to enzymatic 1108 –1111. http://dx.doi.org/10.1099/jmm.0.031062-0.
potencies, cellular substrate specificities, and surface binding to cultured 66. Sun X, He X, Tzipori S, Gerhard R, Feng H. 2009. Essential role of the
cells. J Clin Invest 100:1734 –1741. http://dx.doi.org/10.1172/JCI119698. glucosyltransferase activity in Clostridium difficile toxin-induced secretion
49. Pruitt RN, Chumbler NM, Rutherford SA, Farrow MA, Friedman DB, of TNF-alpha by macrophages. Microb Pathog 46:298 –305. http://dx.doi
Spiller B, Lacy DB. 2012. Structural determinants of Clostridium difficile .org/10.1016/j.micpath.2009.03.002.
March 2016 Volume 84 Number 3 Infection and Immunity iai.asm.org 865
Downloaded from http://iai.asm.org/ on July 12, 2017 by UNIV DE COSTA RICA SISTEMA DE