G Model

ARTICLE IN PRESS

JVAC 15111 17

Vaccine xxx (2014) xxxxxx

Contents lists available at ScienceDirect

Vaccine
journal homepage: www.elsevier.com/locate/vaccine

Development of a recombinant toxin fragment vaccine for Clostridium

difcile infection

Q1

4
5
6
7
8
9
10
11

Q2
Q3

Jerzy Karczewski a, , Julie Zorman a , Su Wang a , Matthew Miezeiewski d , Jinfu Xie a ,

Keri Soring d , Ioan Petrescu b , Irene Rogers b , David S. Thiriot c , James C. Cook a ,
Mihaela Chamberlin e , Rachel F. Xoconostle a , Debbie D. Nahas a , Joseph G. Joyce a ,
Jean-Luc Bodmer a,1 , Jon H. Heinrichs a , Susan Secore a
a

Merck Research Laboratories, Vaccine Basic Research, West Point, PA, United States
Merck Research Laboratories, Laboratory Animal Resources, West Point, PA, United States
c
Merck Research Laboratories, Vaccine Drug Product Development, West Point, PA, United States
d
Eurons Laboratories, Lancaster, PA, United States
e
Agile 1, Torrance, CA, United States
b

12

13
23

a r t i c l e

i n f o

a b s t r a c t

14
15
16

Article history:
Available online xxx

17

22

Keywords:
Clostridium difcile
Recombinant vaccine
TcdA
TcdB

24

1. Introduction

18
19
20
21

25
26

Clostridium difcile infection (CDI) is the major cause of antibiotic-associated diarrhea and pseudomembranous colitis, a disease associated with signicant morbidity and mortality. The disease is mostly of
nosocomial origin, with elderly patients undergoing anti-microbial therapy being particularly at risk. C.
difcile produces two large toxins: Toxin A (TcdA) and Toxin B (TcdB). The two toxins act synergistically to
damage and impair the colonic epithelium, and are primarily responsible for the pathogenesis associated
with CDI. The feasibility of toxin-based vaccination against C. difcile is being vigorously investigated. A
vaccine based on formaldehyde-inactivated Toxin A and Toxin B (toxoids) was reported to be safe and
immunogenic in healthy volunteers and is now undergoing evaluation in clinical efcacy trials. In order
to eliminate cytotoxic effects, a chemical inactivation step must be included in the manufacturing process of this toxin-based vaccine. In addition, the large-scale production of highly toxic antigens could be
a challenging and costly process.
Vaccines base on non-toxic fragments of genetically engineered versions of the toxins alleviate most
of these limitations. We have evaluated a vaccine assembled from two recombinant fragments of TcdB
and explored their potential as components of a novel experimental vaccine against CDI. Golden Syrian
hamsters vaccinated with recombinant fragments of TcdB combined with full length TcdA (Toxoid A)
developed high titer IgG responses and potent neutralizing antibody titers. We also show here that the
recombinant vaccine protected animals against lethal challenge with C. difcile spores, with efcacy
equivalent to the toxoid vaccine. The development of a two-segment recombinant vaccine could provide
several advantages over toxoid TcdA/TcdB such as improvements in manufacturability.
2014 Published by Elsevier Ltd.

Clostridium difcile is the most common cause of infectious

diarrhea in hospitalized patients in the developed world. There

Abbreviations: CDI, Clostridium difcile infection; CROP, clostridial repetitive

oligopeptide; ED, enzymatic domain.
Q4 Corresponding author at: 1634 Sylvan Drive, Blue Bell, PA 19422, United States.
Tel.: +1 215 652 3505; fax: +1 215 652 2142; mobile: +1 267 546 7339.
E-mail addresses: jerzy karczewski@merck.com, jerzykarczewski@comcast.net,
jerzy.karczewski@fhcmb.org (J. Karczewski).
1
Current address: Genocea Biosciences, Cambridge, MA, United States.

are an estimated 500,000 C. difcile infection (CDI) cases in the

United States each year [4,5]. Infection with C. difcile, an anaerobic,
gram-positive, spore-forming bacillus, usually occurs as a complication of antibiotic therapy due to the disruption of normal colonic
ora caused by the antibiotic treatment. The manifestations of
CDI range from asymptomatic colonization to symptomatic disease
and include antibiotic-associated diarrhea and pseudomembranous colitis. Furthermore, disease may progress to toxic megacolon,
sepsis, and death. Advanced age (65 years), antibiotic use,
immunosuppression, health care exposure, duration of hospitalization and serious underlying illness are consistently identied as
risk factors for CDI [1].

http://dx.doi.org/10.1016/j.vaccine.2014.02.026
0264-410X/ 2014 Published by Elsevier Ltd.

Please cite this article in press as: Karczewski J, et al. Development of a recombinant toxin fragment vaccine for Clostridium difcile
infection. Vaccine (2014), http://dx.doi.org/10.1016/j.vaccine.2014.02.026

27
28
29
30
31
32
33
34
35
36
37
38

G Model
JVAC 15111 17

J. Karczewski et al. / Vaccine xxx (2014) xxxxxx

2
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90

ARTICLE IN PRESS

The rst line of treatment for C. difcile infections (CDI) is

to discontinue ongoing antibiotic treatment and to administer
metronidazole or vancomycin [1,2]. While this treatment is effective in most cases, a signicant percentage of patients undergo
relapse. In addition, a newly identied, hypervirulent strain of
C. difcile called BI/NAP1/027 has been implicated in outbreaks
associated with increased morbidity and mortality since the early
2000s [38]. This epidemic strain is responsible for increased incidence of C. difcile-associated diarrhea related not only to antibiotic
exposure, but also associated with GI surgery, prolonged hospitalization and immune-compromising conditions [9]. The emergence
of this and potentially other hypervirulent strains (such as 078)
prompted the search for alternative methods to treat and control
infections leading to CDI. In addition to the development of novel
narrow-spectrum antibiotics such as daxomicin [10], two other
approaches are being pursued: therapeutic monoclonal antibodies
[11] and prophylactic vaccination [12].
C. difcile produces two potent exotoxins: Toxin A (TcdA) and
Toxin B (TcdB). These toxins induce a broad range of pathologies
locally (inammation, colonic epithelium damage) [13], and potentially could induce cardiotoxicity, as demonstrated in zebrash
embryos [14]. Both toxins are encoded on a 19 kb region of the
genome referred to as the pathogenicity locus or PaLOC [15] and
function through glucosylation of Rho-family GTP-ases leading to
a disruption of the gut mucosa [16].
TcdA and TcdB are large proteins (308 kDa and 270 kDa, respectively) and share a relatively high degree of amino acid sequence
homology (50%) and similar multi-domain structures. The toxins
are orthologous to the Large Clostridial Toxin family, which also
includes TcsL, TcsH, TcnA and TpeL [17,18]. The N-terminal, enzymatic domain (ED) of each protein contains all enzymatic activities
necessary for glucosylation of several GTPases such as RhoA, Rac
and Cdc42 [19,20]. The central hydrophobic core constitutes a pore
forming domain that mediates translocation of the ED to the host
cell cytoplasm, while the C-terminus of the protein consists of
a highly repetitive region referred to as the clostridial repetitive
oligopeptide (CROP) domain that binds cellular receptors.
Toxins enter cells by receptor-mediated endocytosis and
undergo an auto-catalytic cleavage in the cytosol which releases
the enzymatic domain. The ED subsequently catalyzes glucosylation of small GTPases, leading to disaggregation of the cytoskeleton
and cell death (for a review see [21]).
The aim of the current study was to evaluate the efcacy of a vaccine consisting of two separately expressed recombinant fragments
of TcdB combined with full-length, formaldehyde-inactivated TcdA
(Toxoid A). We demonstrate that hamsters immunized with this
vaccine developed high titer neutralizing antibodies and longlasting (>3 months) protective immunity against lethal challenge
with C. difcile. The inclusion of fragments encompassing the complete sequence of TcdB presents a novel and promising approach
for development of a safe and efcacious vaccine for prevention of
CDI as well as for potentially reducing costs of manufacturing.

91

2. Materials and methods

92

2.1. Cloning and expression of recombinant antigens

93
94
95
96
97
98
99
100

The genes encoding TcdB fragments, based on the source

sequence of C. difcile strain VPI 10463 source sequence were
codon optimized for expression in E. coli, synthesized by Genescript (Piscataway, NJ). The fragments were then cloned into
apET30a expression vector (EMD Millipore, Billerica, MA) with
C-terminal His-tag, and transformed into competent BL21 (DE3)
Escherichia coli cells (New England Biolabs, Ipswich, MA). One
liter cultures were grown in Luria-Bertani (LB) broth containing

30 g/mL kanamycin to an optical density (OD) at 600 nm of 0.6.

Expression was induced by the addition of 1 mM isopropyl -dthiogalactopyranoside (IPTG). Incubation continued for 4 h at 37 C.
Cells were harvested by centrifugation and cell pellets were stored
frozen at 70 C. TcdA fragment A-C1 was expressed using the Saccharomyces cerevisiae expression system previously described by
Jansen et al. [22].

2.2. Extraction and purication of recombinant antigens

TcdB fragments were puried from E. coli cell pellets using
immobilized metal afnity chromatography (IMAC) followed by
ion-exchange chromatography. Briey, each E. coli pellet was suspended in lysis buffer (50 mM Tris pH 8.0, 300 mM NaCl, 20 mM
imidazole) supplemented with the protease inhibitor cocktail
Complete (Roche Diagnostics, Mannheim, Germany) and 2500
units of nuclease (Benzonase , EMD Chemicals, Gibbstown, NJ) and
the cell suspension was passed through a microuidizer (Model
110Y, Microuidics Corp. Newton, MA). The lysate was claried
by centrifugation (20,000 g, 30 min), ltered through a 0.22
ExpressTM Plus ltration device (Millipore Corp., Billerica, MA)
and loaded onto a 5 mL HiTrap IMAC FF resin (GE Healthcare BioSciences Corp. Piscataway, NJ) equilibrated in 50 mM Tris pH 8.0,
300 mM NaCl, 20 mM imidazole and eluted with a 20250 mM
linear imidazole gradient in the same buffer. Fractions containing TcdB fragments were pooled and diluted 5-fold with 20 mM
Tris, pH 7.4 and subsequently loaded onto a 6 mL Resource Q
column (GE Healthcare Bio-Sciences Corp. Piscataway, NJ). TcdB
fragments were eluted with a NaCl gradient, dialyzed against
HEPES-buffered saline (10 mM Hepes, pH 7.4, 150 mM NaCl) and
ltered through a 0.22 Supor membrane lter (Pall Corporation,
Port Washington, NY) to make the nal product for immunization
studies.
The TcdA-C1 (amino acids 18322710) was puried to homogeneity from S. cerevisiae cell paste. Briey, the cell paste was
suspended in 0.2 M MOPS, pH 7.0 and ammonium sulfate added
to 50% saturation at 4 C with mixing, for 30 min. And the precipitate was harvested by centrifugation (10,000 g, 30 min, 4 C),
dissolved in 50 mM MOPS, pH = 7.0 containing 0.1% CHAPS and
diluted 50-fold with 50 mM MOPS pH 7.0. A-C1 fragment was
puried by sequential chromatography on the cation exchange
column containing POROS 50HS resin (Applied Biosystems,
Foster City, CA) followed by multimode chromatography on
hydroxyapatite HA resin (CHT-Type II 5 mL, BioRad, Hercules,
CA).

2.3. Biophysical characterization of recombinant fragment B-TC

The puried fragment B-TC was dialyzed into 10 mM phosphate
pH 7.0 and the circular dichroism (CD) spectrum was acquired
from 300 to 180 nm at 20 C using a Jasco J810 spectropolarimeter. Protein concentration was determined by amino acid analysis
and was used to convert the raw CD data to molar ellipticity.
Secondary structure estimation was performed using the CDPro
analytical software suite and the programs CDSSTR, SELCON3 and
CONTINLL [23,24,25,26,27,28]. Analytical ultracentrifugation was
performed using sedimentation equilibrium analysis on fragment
B-TC in 10 mM sodium phosphate, pH 7.0, 0.15 M NaCl at 20 C. The
analysis was carried out at rotor speeds of 6000 RPM, 8000 RPM, and
10,000 RPM for 16 h in a Beckman XL-I analytical ultracentrifuge
and the data was tted using HeteroAnalysis, ver. 1.1.33 software
(James Cole, Jeffrey Larry, Analytical Ultracentrifugation Facility,
Biotechnology and Bioservices Center, University of Connecticut,
Storrs, CT).

Please cite this article in press as: Karczewski J, et al. Development of a recombinant toxin fragment vaccine for Clostridium difcile
infection. Vaccine (2014), http://dx.doi.org/10.1016/j.vaccine.2014.02.026

101
102
103
104
105
106
107

108

109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143

144

145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160

G Model
JVAC 15111 17

ARTICLE IN PRESS
J. Karczewski et al. / Vaccine xxx (2014) xxxxxx

161

2.4. Toxin inactivation and formulation

174

TcdA and TcdB (puried from C. difcile strain VPI10463, List Biologicals, Campbell, CA) and selected recombinant fragments (B-ED
and B-TC) were reconstituted in 20 mM HEPES-buffered saline (pH
7.4) at 0.30.4 mg/mL. Formaldehyde was added to a nal concentration of 0.43% (v/v) and l-lysine was added to a nal concentration
of 4.2 mg/mL. Inactivation proceeded for 18 days 4 C and then the
protein solutions were dialyzed against 20 mM HEPES buffer (pH
7.4), containing 100 mM sodium chloride and sterilized by passage
through a 0.22 lter.
Protein antigens were co-formulated with aluminum hydroxyphosphate sulfate adjuvant (AHPS, Merck & Co, Inc.) and
ISCOMATRIXTM adjuvant (CSL Limited, Parkville, Australia) in saline
to prepare each component of the vaccine.

175

2.5. Hamster immunogenicity and protection studies

162
163
164
165
166
167
168
169
170
171
172
173

194

Golden Syrian hamsters, (male, 90120 g), were obtained from

Charles River Laboratories (Wilmington, MA), and were individually maintained in lter-lid cages. Hamsters were immunized
intramuscularly in the left quadriceps on days 0, 21, 42 and 63
with a 200 L vaccine dose containing 10 g of combined TcdB
fragments and10 g of TcdA toxoid. Control animals received adjuvant alone. Serum samples were collected via retro-orbital bleeding
immediately prior to each immunization and on day 77 and stored
frozen. For evaluation of vaccine efcacy against a lethal C. difcile spore challenge, animals were oro-gavaged with one dose
of 10 mg/kg clindamycin ve days prior to spore administration
and 16 days following the nal vaccine dose. C. difcile strain
VPI 10463spores 2ED90, 600 Cfu)were administered by orogavage and animals were monitored for symptoms of CDI or death
for the duration of the study (21 days post-challenge).The MantelCox test was applied to assess statistical differences in survival
rates between groups. All hamster experiments were performed
under Institutional Animal Care and Use Committee guidelines as
approved by Merck and Co., Inc.

195

2.6. Serological assays

176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193

196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212

213
214
215
216
217
218
219
220

2.6.1. ELISA
Thermo Reacti-Bind (Thermo Lab systems, Waltham, MA) 96well plates were coated with either TcdA or TcdB at 50 ng/well or
25 ng/well, respectively in PBS overnight at 4 C. Wells were washed
with PBS containing 0.05% Tween 20 (PBS-T) and blocked in PBS
containing 1% bovine serum albumin (Sigma, St. Louis, MO). Serial
dilutions (1:2) of hamster sera were incubated at room temperature
for 1.5 h. Plates were then washed with PBS-T and incubated with
goat anti-hamster IgG conjugated to horseradish peroxidase (HRP)
(Abcam, Cambridge, MA) for 1 h at room temperature. Following
an additional wash, 50 L of 3,3 ,5,5 -tetramethylbenzidine (TMB)
substrate (Thermo Labsystems, Rockford, IL) was added and after
15 min, the reaction was stopped with 50 L 0.4 N sulfuric acid and
the absorbance measured at 450 nm (Powerwave X Select, Bio-Tek
Instruments, Winooski, VT). Antibodytiters were calculated using
four parameter regression tting of the titration curve using Prism
5 for Windows software (GraphPad Software, Inc. La Jolla, CA).
2.6.2. Neutralization and cytotoxicity assay
Native TcdA or TcdB was pre-incubated with test sera for 1.5 h
at 4 C, applied to Vero cells grown in 384 well plates and incubated
in humidied CO2 incubators at 37 C for 48 h. The cells were xed,
permeabilized, and stained with Alexa Fluor 488 phalloidin and an
image of the cell monolayer was acquired using the ImageXpress
Velos Scanning Cytometer (Molecular Devices, LLC, Sunnyvale, CA).
The total cell surface area in each well was plotted against the

Table 1
Recombinant toxin fragments selected for in vivo evaluation, their molecular
weigths and purication yields.

.
Fragment

Amino acids

Mw (kDa)

Yield (mg/L)

B-ED
B-TC
B-C0
B-C1
B-C2
B-C3
B-C4
A-C1

1543
5452367
17862367
18362367
18362101
19492275
21022367
18322710

62.2
207
68
62.4
31.6
39.1
32
99.5

505
39
822
582
1836
686
800
20

dilution of sera (for neutralization titers) or against the toxin concentration (for cytotoxicity assessment) and titers were calculated
using four parameter regression tting using Prism 5 for Windows
software (GraphPad Software, Inc. La Jolla, CA).
3. Results
3.1. Design, preparation and formulation of recombinant Toxin B
fragments
We designed a vaccine composed of two recombinant fragments
covering an entire sequence of TcdB (Table 1). The rst fragment,
B-ED comprised the N-terminal region (amino acids 1543) and
contained the entire enzymatic domain of TcdB. To minimize the
possibility of cytotoxicity resulting from cellular uptake of this
fragment, four residues were mutated: W102A, D288A, E515Q
and S518A, which have been shown to be critical for the catalytic activity of the enzyme [2931]. The second component of
the vaccine (B-TC) encompassed the translocation domain through
the C-terminus (residues 5452367); therefore, when combined
together, fragments B-ED and B-TC contained the entire amino acid
sequence from full-length TcdB. The CROPS regions were identied
as being immunodominant and shown to induce potent neutralizing antibodies [32], we therefore prepared and evaluated several
shorter CROPS-derived fragments of TcdB (B-C0B-C4) and one Cterminal fragment of TcdA (Table 1).
TcdB fragments were expressed in E. coli and puried from soluble extracts in a two-step process including IMAC followed by
anion exchange chromatography. A similar strategy was initially
attempted to produce recombinant fragments of TcdA, including
the larger fragment A-TC and the shorter CROPS-derived fragment
A-C1; however, expression yields as well as solution stability did
not allow for production of sufcient quantities of proteins necessary to support in vivo studies. Expression of the A-C1 fragment
in S. cerevisiae followed by sequential chromatography on cation
exchange and hydroxyapatite chromatography allowed the desired
product to be puried to homogeneity (>95%). The overall purication yields from E. coli extracts ranged from 39 to 1836 mg/L of E. coli
culture, as shown in Table 1. The puried TcdA and TcdB fragments
migrated on SDS-PAGE at a molecular weight corresponding to the
monomeric form of each protein, as calculated from the amino acid
sequence (Fig. 1A). The purity of all antigens was greater than 90%
with very little evidence of degradation, as conrmed by Western
blot analysis (data not shown).

Please cite this article in press as: Karczewski J, et al. Development of a recombinant toxin fragment vaccine for Clostridium difcile
infection. Vaccine (2014), http://dx.doi.org/10.1016/j.vaccine.2014.02.026

221
222
223
224

225

226
227

228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261

G Model
JVAC 15111 17
4

ARTICLE IN PRESS
J. Karczewski et al. / Vaccine xxx (2014) xxxxxx
Table 2
Combinations of TcdA TcdB antigens selected for in vivo evaluation in clindamycininduced enterocolitis model.

Group 1
Group 2
Group 3
Group 4
Group 5
Group 6
Group 7
Group 8
Group 9

TcdA antigen

TcdB antigen

TcdA
TcdA
TcdA
TcdA
TcdA
TcdA
A-C1
A-C1
AAHS/IMX

TcdB
B-ED+B-TC(+F)
B-ED+B-TC
B-ED+B-C0
B-ED+B-C1
B-TC
B-ED+B-TC(-F)
TcdB
/IMX

et al. [33] demonstrated that a number of chimeric TcdB variants,

even when lacking the domain critical for translocation could retain
their full cytotoxic activity. For all recombinant antigens, including
the enzymatic domain B-ED as well as combinations of B-ED with
C-terminal fragments (B-TC, B-C0 and B-C1), we detected no cytotoxic activity even at micromolar concentrations (Fig. 1B) whereas
TcdA and TcdB exerted potent cytotoxic effects with an EC50 of
1.8 pM and 0.02 pM, respectively.
Table 2 presents the various combinations of recombinant fragments selected for direct comparison with a benchmark vaccine
consisting of full-length TcdA and TcdB puried from the C. difcile
bacterium and inactivated by treatment with formaldehyde (group
1). A two-segment vaccine containing recombinant fragments BED and B-TC (groups 2 and 3) was tested as formaldehyde treated
(group 2) and untreated combination. In addition, several additional combinations including B-TC, B-C0 and B-C1 (groups 35),
B-TC fragment alone (group 6), A-C1, B-ED, and B-TC (group 7), and
A-C1 and native TcdB (group 8) were tested.

3.2. Immunogenicity of the recombinant fragments

Fig. 1. (A) Analysis of recombinant toxin fragments by SDS/PAGE. 5 g of each protein was separated on NuPage 412% gel and stained with Coomassie Blue. (B)
Cytotoxic activity of native and recombinant toxin fragments. The potency of fragments or unmodied combinations was measured in a cell-based cytotoxicity assay.
The values for all recombinant fragments are plotted as the highest concentrations
tested (no cytotoxicity was observed at these concentrations).

Secondary structure estimation and solution-based size of puried fragment B-TC (207 kDa) was determined by circular dichroism
263
and analytical ultracentrifugation, respectively. CD indicated 22%
264
-helix, 29% -sheet, 21% turns, and 29% unordered compo265
nents suggesting that the recombinant antigen adopted a complex
266
secondary structure not dominated by any single structural com267
ponent. In comparison, previously published results have reported
268
that full length native TcdB is composed of 36% -helix, 21%
269
-sheet, 18% turns and 25% unordered components [33]. Sedimen270
tation equilibrium analysis of B-TC yielded a molecular weight of
271
726,745 Da, consistent with the formation of trimers in solution.
272
Recombinant fragments and combinations thereof were evalu273
Q5 ated in a cellular-based cytotoxicity assay. Studies by Genisyuerek
274
262

Hamsters were immunized four times with the antigen groups

listed in Table 2 and the post-dose four sera was evaluated by ELISA.
All immunized animals developed IgG antibodies specic for TcdA
as measured by ELISA (Fig. 2A) as well as antibodies to TcdB (Fig. 2B).
There were no statistically signicant differences between titers in
groups 18 as measured by MantelCox test. No C. difcile-specic
IgG response was detected in the adjuvant control group 9 (Fig. 2A
and B). We also evaluated fragments B-C2, B-C3 and B-C4 representing shorter fragments of CROPS, however these fragments failed to
induce potent IgG responses (data not shown) and therefore these
fragments were excluded from subsequent evaluation.

3.3. Serum neutralizationactivity

Neutralizing antibody titers were evaluated inpost-dose
4immune sera. All groups immunized with recombinant antigen combinations (Groups 28) developed neutralizing antibodies
against both TcdA and TcdB (Fig. 2C and D, respectively). We did not
observe statistically signicant differences in groups 18 in neutralization titers as measured against TcdA (MantelCox test). In
the case of TcdB, a statistically signicant increase in neutralizing
antibody titers was observed in groups 2 and 3 as compared to toxoid group 1. The remaining groups (48), did not reach statistical
signicance (p > 0.05) relative to group 1. Control animals receiving
adjuvant alone (group 9) did not produce measurable antibodies
that neutralized the cytotoxic effects of either TcdA or TcdB (Fig. 2C
and D). Similarly, preimmune sera obtained from all of the animals
was incapable of neutralizing toxicity at the lowest dilution tested
(data not shown).

Please cite this article in press as: Karczewski J, et al. Development of a recombinant toxin fragment vaccine for Clostridium difcile
infection. Vaccine (2014), http://dx.doi.org/10.1016/j.vaccine.2014.02.026

275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292

293

294
295
296
297
298
299
300
301
302
303
304

305

306
307
308
309
310
311
312
313
314
315
316
317
318
319
320

G Model
JVAC 15111 17

ARTICLE IN PRESS
J. Karczewski et al. / Vaccine xxx (2014) xxxxxx

Fig. 2. Systemic antibody response in the individual sera samples collected at day 77. (A and B) ELISA titers were measured against TcdA (Panel A) and TcdB (Panel B). (C and
D) Neutralizing titers for TcdA (Panel C) and TcdB (Panel D) were measured. The average titers (n = 9) SEM are shown.

321
322

323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352

3.4. In vivo efcacy of fragment based vaccine in hamster

challenge model
The clindamycin-induced lethal enterocolitis model in Golden
Syrian hamsters is commonly utilized for studying the mechanism of C. difcile infections as well as protection mediated by
vaccines [34] and antibodies [32] as well as small molecule antibiotics [35]. To evaluate the efcacy of our recombinant vaccines,
hamsters immunized with vaccine combinations (group 18) or
with adjuvant control (group 9) were treated with clindamycin, to
disrupt the gut microbiota and render the treated animals susceptible to subsequent challenge with C. difcile spores (Fig. 3A). Five
days following treatment, all animals were challenged with a lethal
dose of C. difcile spores VPI10463and monitored for development
of disease symptoms. We observed that all animals in the adjuvant control group developed diarrhea (wet tail) and died within
4280 h after challenge (group 9, Fig. 3B). The extent of protection in the vaccinated groups varied signicantly. We observed
near complete protection in the toxoid-vaccinated animals (group
1, Fig. 3B). In this group, eight out of nine hamsters survived for at
least 3 weeks post challenge. We observed similar complete protection in group 2, which received the recombinant two-segment
TcdB vaccine combined with toxoid A. Despite being protected from
death, signicant weight loss was evident in both groups 1 and 2.
All animals in these groups regained weight and resumed a normal diet within 23 weeks post challenge (Fig. 3D). No disease
relapse was observed up to two months post-challenge. In addition,
statistically signicant protection, compared to adjuvant control
was observed in groups 36, with 8090% survival at 1 week post
challenge and the overall survival between 40% and 70% recorded
at the end of the study. MantelCox analysis revealed no statistically signicant differences in the survival between groups 36
(P = 0.3605). Partial protective efcacy was observed in groups 7

and 8 which were immunized with the recombinant fragment of

TcdA (A-C1) combined with either the two-segment recombinant
TcdB (group 7) or with native TcdB (group 8) as shown in Fig. 3C.
Survival in both groups was improve das compared to the adjuvant control (P < 0.007), however these two vaccine combinations
performed substantially poorer than the full-length toxoid combination (group 1) with overall 20% and 40% survival (in groups 8 and
7, respectively) at the completion of the study.
4. Discussion
Prevention of nosocomial C. difcile infections represents a
major unmet medical need. Development of an efcacious vaccine is a major focus in multiple leading academic and industrial
research centers. An investigational vaccine comprised of formalininactivated full length toxoids TcdA and TcdB prepared from C.
difcile bacterial culture is currently undergoing clinical trials [36].
In order to produce a vaccine from the native organism, the cytotoxicity of puried toxins must be reduced by several orders of
magnitude, typically via chemical inactivation, such as formaldehyde treatment. A phase 1 study demonstrated the relative safety
of such a toxoid vaccine in a small group of healthy volunteers
[37]. Recombinant expression technology represents a convenient
platform for the high yield production of protein subunit vaccines
while offering the added advantage of enabling genetic alterations
via site-directed mutagenesis or truncations to obtain antigens of
desired properties. In this study we report the development of
a vaccine candidate shown to be ofcious in hamsters and comprising detoxied fragments of TcdB produced in a recombinant
expression system and combined with full length TcdA. The cytotoxicity of native TcdB was eliminated by expressing, purifying and
formulating two individual segments of the protein then combining
the formulations for immunization. We found that immunization of

Please cite this article in press as: Karczewski J, et al. Development of a recombinant toxin fragment vaccine for Clostridium difcile
infection. Vaccine (2014), http://dx.doi.org/10.1016/j.vaccine.2014.02.026

353
354
355
356
357
358
359
360

361

362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383

G Model
JVAC 15111 17
6

ARTICLE IN PRESS
J. Karczewski et al. / Vaccine xxx (2014) xxxxxx

Fig. 3. Evaluation of C. difcile vaccines in Golden Syrian hamster model of CDI.

(A) The outline of in vivo challenge model. (B) The groups of hamsters (n = 9) that
were immunized with recombinant TcdB fragments combined with Toxoid A were
challenged with C. difcile spores and the survival monitored. (C) Hamsters were
immunized with recombinant fragment of TcdA (A-C1) combined with Toxoid B,
challenged with C. difcile spores and the survival measured. (D) Surviving hamsters
in group 1 and 2 were monitored for weight changes for the duration of the study
and continued for an additional month.

384
385
386
387
388
389
390
391
392
393
394
395
396

hamsters with a formaldehyde treated two-segment TcdB component combined with a full-length, formaldehyde inactivated TcdA
component provided long-lasting (>3 months) immunity with100%
survival against lethal challenge with C. difcile. Moreover this
vaccine was as efcacious as a full-length toxoid vaccine. To our
knowledge this is the rst report to demonstrate equivalent efcacy
of a recombinant TcdB and toxoid antigen, where both molecules
were evaluated within the same experiment, at the same antigen
dose and formulation and using the same C. difcile challenge strain.
In contrast, we observed only partial efcacy in all other groups
(groups 38), immunized with one of several shorter fragments of
TcdB and/or one fragment of TcdA in combination with full-length
TcdB toxoid.

The efcacy of recombinant C. difcile vaccines has been

reported by others. A vaccine based on genetically fusing the Cterminal domains of TcdA and TcdB with a four amino acid linker
and formulated with aluminum hydroxide was recently reported
by Tian et al. [38]. This chimeric antigen was efcacious in a hamster model using C. difcile strain 630 as the challenge. We believe
that the current results, in which we challenged hamsters immunized with our vaccine constructs with the higher toxin-producing
strain VPI 10463 [39], represents a more rigorous demonstration of
efcacy.
A second chimeric-based vaccine in which the C-terminal CROPS
domain of TcdB variant carrying three detoxifying amino acid
mutations was replaced by the corresponding CROPS segment of
TcdA was recently described by Wang et al. [40]. Animals immunized with this construct displayed protective immunity up to 80 h
post challenge. These results are in good agreement with the survival outcome we report in the current study.
Interestingly, our current study demonstrated that formaldehyde treatment of a B-ED and B-TC fragment vaccine resulted in a
signicant improvement of efcacy relative to the untreated combination (Fig. 3, group 2 vs. group 3). The mechanism for this
enhanced efcacy is not entirely understood. Binding antibody
titers as measured by ELISA or neutralizing antibody titers as measured by the cytotoxicity assay were not increased in group 2,
suggesting that a mechanism not directly related to the neutralization of the cytotoxic effect may exist. Such effects may be related
to improvements in structural stability of the antigenic site by
treatment with formaldehyde. The biophysical characterizations
reported herein demonstrate that B-TC is a structurally complex
molecule containing multiple -helical and -sheet components
as determined by CD analysis. Analytical centrifugation using equilibrium sedimentation analysis shows that the B-TC fragment forms
trimeric complexes in solution. The formation of similar multimeric
structures has been reported previously for native toxins using
differential light scattering (DLS) analysis [33,41]. This multimerization is believed to be a physiologically relevant phenomenon and
could play a role during pore formation following endocytosis of
TcdB (reviewed in [42]). Formaldehyde-crosslinked TcdB was also
shown to exhibit an improved thermal stability as compared with
untreated TcdB [33]. It is likely that the B-TC fragment undergoes
similar modications following formaldehyde treatment which
may lead to improved stability and enhanced vaccine efcacy.
The mechanism of protection in the hamster lethal challenge
model may be complex in that factors other than anti-toxin antibodies may play a role. Our current study demonstrates that levels
of anti-toxin IgG and neutralizing antibody elicited by recombinant
fragment vaccine are comparable to those elicited by full length
inactivated toxoid, yet these titers do not correlate well with the
nal extent of protection measured three weeks post-challenge. It
appears plausible that neutralizing antibody titers measured in our
cytotoxicity assay reect a reduction in TcdB binding to host cells,
most likely mediated by antibodies directed against the CROPS
region. However, other regions of the toxin proteins may still play
a role in progression of the disease and it is possible that a broader
repertoire of antibodies to those regions is elicited by the toxoid
vaccine, A recent report by Marozsan et al. [43] demonstrated that
monoclonal antibodies against the enzymatic domain of TcdB combined with an anti TcdA antibody supported long-lasting protection
in a hamster model of CDI. In contrast, a different set of monoclonal antibodies, directed against the C-terminal domains of TcdA
and TcdB, only partially protected hamsters from lethal challenge
with C. difcile. This study supported the role of the enzymatic
domain in development of protective immunity. It would be of
interest to evaluate a combination of antibodies directed against
two distinct regions of TcdB to better understand the contribution
of multiple epitopes to protection [43] Additional animal efcacy

Please cite this article in press as: Karczewski J, et al. Development of a recombinant toxin fragment vaccine for Clostridium difcile
infection. Vaccine (2014), http://dx.doi.org/10.1016/j.vaccine.2014.02.026

397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462

G Model
JVAC 15111 17

ARTICLE IN PRESS
J. Karczewski et al. / Vaccine xxx (2014) xxxxxx

studies will be required to dene which regions of TcdA and TcdB

mediate protection from challenge and thus are critical for inclu465
sion in a nal vaccine composition. Similarly, additional work is
466
required to understand the effect of chemical and genetic inactiva467
tion on the structure, stability and antigenicity of both native and
468
recombinant toxins and fragments thereof. Finally, adjuvant and
469
formulation considerations will be important to produce vaccine
470
candidates that offer long-lived protective efcacy in humans.
471
We have attempted to conduct a similar study with TcdA, how472
ever our attempts to produce set of analogous fragments were not
473
successful. The preliminary results with A-C1 encourage further
474
research to establish an alternative expression platform allowing
475
production of sufcient quantities of antigens to support in vivo
476
studies.
477
In summary, our wok shows that a fragment-based recombi478
nant vaccine comprising two segments of TcdB, combined with full
479
length toxoid A was highly immunogenic in a hamster lethal chal480
lenge model of CDI. The vaccine stimulated development of robust
481
neutralizing antibody titers and demonstrated complete protection
482
which was equivalent to a full length TcdA/TcdB toxoid vaccine.
483
Q6 This approach presents an attractive and novel pathway for further
484
development of a low cost, efcacious and potentially safer vaccine
485
to prevent CDI infections.
463
464

486

Acknowledgements

489

We thank Mark Knower, Patricia Rebbeck and Noelle Dahl-Parisi

for their help in developing andperforming the hamster model of
CDI.

490

References

487
488

491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534

[1] Lo Vecchio A, Zacur GM. Clostridium difcile infection: an update on epidemiology, risk factors, and therapeutic options. Curr Opin Gastroenterol
2012;28(January (1)):19.
[2] Cohen SH, Gerding DN, Johnson S, Kelly CP, Loo VG, McDonald LC, et al. Clinical
practice guidelines for Clostridium difcile infection in adults: 2010 update by
the ociety for healthcare epidemiology of America (SHEA) and the infectious
diseases society of America (IDSA). Infect Control Hosp Epidemiol 2010;31(May
(5)):43155.
[3] Loo VG, Bourgault AM, Poirier L, Lamothe F, Michaud S, Turgeon N, et al. Host
and pathogen factors for Clostridium difcile infection and colonization. N Engl
J Med 2006;365(November (18)):1693703.
[4] McDonald LC, Killgore GE, Thompson A, Owens Jr RC, Kazakova SV, Sambol SP,
et al. An epidemic, toxin gene-variant strain of Clostridium difcile. N Engl J Med
2005;353(December (23)):243341.
[5] Muto CA. Why are antibiotic-resistant nosocomial infections spiraling out of
control? Infect Control Hosp Epidemiol 2005;26(January (1)):102.
[6] Pepin J, Valiquette L, Alary ME, Villemure P, Pelletier A, Forget K, et al. Clostridium difcile-associated diarrhea in a region of Quebec from 1991 to 2003: a
changing pattern of disease severity. CMAJ 2004;171(August (5)):46672.
[7] Zar FA, Bakkanagari SR, Moorthi KM, Davis MB. A comparison of vancomycin
and metronidazole for the treatment of Clostridium difcile-associated diarrhea,
stratied by disease severity. Clin Infect Dis 2007;45(August (3)):3027.
[8] Pituch H, Bakker D, Kuijper E, Obuch-Woszczatynski P, Wultanska D, Nurzynska
G, et al. First isolation of Clostridium difcile PCR-ribotype 027/toxinotype III in
Poland. Pol J Microbiol 2008;57(3):2678.
[9] Hookman P, Barkin JS. Clostridium difcile associated infection, diarrhea and
colitis. World J Gastroenterol 2009;15(April (13)):155480.
[10] Venugopal AA, Johnson S. Fidaxomicin: a novel macrocyclic antibiotic approved
for treatment of Clostridium difcile infection. Clin Infect Dis 2011;54(February
(4)):56874.
[11] Lowy I, Molrine DC, Leav BA, Blair BM, Baxter R, Gerding DN, et al. Treatment
with monoclonal antibodies against Clostridium difcile toxins. N Engl J Med
2010;362(3):197205.
[12] Kotloff KL, Wasserman SS, Losonsky GA, Thomas Jr W, Nichols R, Edelman R,
et al. Safety and immunogenicity of increasing doses of a Clostridium difcile
toxoid vaccine administered to healthy adults. Infect Immun 2001;69(February
(2)):98895.
[13] Lyerly DM, Krivan HC, Wilkins TD. Clostridium difcile: its disease and toxins.
Clin Microbiol Rev 1988;1(January (1)):118.
[14] Hamm EE, Voth DE, Ballard JD. Identication of Clostridium difcile toxin B cardiotoxicity using a zebrash embryo model of intoxication. Proc Natl Acad Sci
U S A 2006;103(September (38)):1417681.
[15] Cohen SH, Tang YJ, Silva Jr J. Analysis of the pathogenicity locus in Clostridium
difcile strains. J Infect Dis 2000;181(February (2)):65963.

[16] Riegler M, Sedivy R, Sogukoglu T, Castagliuolo I, Pothoulakis C, Cosentini E,

et al. Epidermal growth factor attenuates Clostridium difcile toxin A- and Binduced damage of human colonic mucosa. Am J Physiol 1997;273(November
(5 Pt 1)):G101422.
[17] Amimoto K, Noro T, Oishi E, Shimizu M. A novel toxin homologous to large
clostridial cytotoxins found in culture supernatant of Clostridium perfringens
type C. Microbiology (Reading, England) 2007;153(April (Pt 4)):1198206.
[18] von Eichel-Streiber C, Boquet P, Sauerborn M, Thelestam M. Large clostridial
cytotoxins a family of glycosyltransferases modifying small GTP-binding proteins. Trends Microbiol 1996;4(October (10)):37582.
[19] Spyres LM, QaDan M, Meader A, Tomasek JJ, Howard EW, Ballard JD. Cytosolic delivery and characterization of the TcdB glucosylating domain by using a
heterologous protein fusion. Infect Immun 2001;69(January (1)):599601.
[20] Jank T, Giesemann T, Aktories K. Rho-glucosylating Clostridium difcile toxins
A and B: new insights into structure and function. Glycobiology 2007;17(April
(4)):15R22R.
[21] Voth DE, Ballard JD. Clostridium difcile toxins: mechanism of action and role
in disease. Clin Microbiol Rev 2005;18(April (2)):24763.
[22] Jansen KU, Rosolowsky M, Schultz LD, Markus HZ, Cook JC, Donnelly JJ,
et al. Vaccination with yeast-expressed cottontail rabbit papillomavirus (CRPV)
virus-like particles protects rabbits from CRPV-induced papilloma formation.
Vaccine 1995;13(November (16)):150914.
[23] Johnson WC. Analyzing protein circular dichroism spectra for accurate secondary structures. Proteins 1999;35(May (3)):30712.
[24] Provencher SW, Glockner J. Estimation of globular protein secondary structure
from circular dichroism. Biochemistry 1981;20(January (1)):337.
[25] Sreerama N, Venyaminov SY, Woody RW. Estimation of the number of
alpha-helical and beta-strand segments in proteins using circular dichroism
spectroscopy. Protein Sci 1999;8(February (2)):37080.
[26] Sreerama N, Woody RW. A self-consistent method for the analysis of protein
secondary structure from circular dichroism. Anal Biochem 1993;209(February
(1)):3244.
[27] Sreerama N, Woody RW. Poly(pro)II helices in globular proteins: identication
and circular dichroic analysis. Biochemistry 1994;33(August (33)):100225.
[28] Sreerama N, Woody RW. Estimation of protein secondary structure from
circular dichroism spectra: comparison of CONTIN, SELCON, and CDSSTR
methods with an expanded reference set. Anal Biochem 2000;287(December
(2)):25260.
[29] Busch C, Hofmann F, Gerhard R, Aktories K. Involvement of a conserved
tryptophan residue in the UDP-glucose binding of large clostridial cytotoxin
glycosyltransferases. J Biol Chem 2000;275(May (18)):1322834.
[30] Jank T, Giesemann T, Aktories K. Clostridium difcile glucosyltransferase toxin
B-essential amino acids for substrate binding. J Biol Chem 2007;282(November
(48)):3522231.
[31] Reinert DJ, Jank T, Aktories K, Schulz GE. Structural basis for the function of
Clostridium difcile toxin B. J Mol Biol 2005;351(September (5)):97381.
[32] Babcock GJ, Broering TJ, Hernandez HJ, Mandell RB, Donahue K, Boatright
N, et al. Human monoclonal antibodies directed against toxins A and B
prevent Clostridium difcile-induced mortality in hamsters. Infect Immun
2006;74(November (11)):633947.
[33] Salnikova MS, Joshi SB, Rytting JH, Warny M, Middaugh CR. Physical characterization of Clostridium difcile toxins and toxoids: effect of the formaldehyde crosslinking on thermal stability. J Pharm Sci 2008;97(September
(9)):373552.
[34] Torres JF, Lyerly DM, Hill JE, Monath TP. Evaluation of formalin-inactivated
Clostridium difcile vaccines administered by parenteral and mucosal routes of
immunization in hamsters. Infect Immun 1995;63(December (12)):461927.
[35] Ochsner UA, Bell SJ, OLeary AL, Hoang T, Stone KC, Young CL, et al. Inhibitory
effect of REP3123 on toxin and spore formation in Clostridium difcile, and
in vivo efcacy in a hamster gastrointestinal infection model. J Antimicrob
Chemother 2009;63(May (5)):96471.
[36] Foglia G, Shah S, Luxemburger C, Pietrobon PJ. Clostridium difcile: development
of a novel candidate vaccine. Vaccine 2012;30(June (29)):43079.
[37] Greenberg RN, Marbury TC, Foglia G, Warny M. Phase I dose nding studies
of an adjuvanted Clostridium difcile toxoid vaccine. Vaccine 2012;30(March
(13)):22459.
[38] Tian JH, Fuhrmann SR, Kluepfel-Stahl S, Carman RJ, Ellingsworth L, Flyer DC.
A novel fusion protein containing the receptor binding domains of C. difcile
toxin A and toxin B elicits protective immunity against lethal toxin and spore
challenge in preclinical efcacy models. Vaccine 2012;30(June (28)):424958.
[39] Merrigan M, Venugopal A, Mallozzi M, Roxas B, Viswanathan VK, Johnson
S, et al. Human hypervirulent Clostridium difcile strains exhibit increased
sporulation as well as robust toxin production. J Bacteriol 2010;192(October
(19)):490411.
[40] Wang H, Sun X, Zhang Y, Li S, Chen K, Shi L, et al. A chimeric toxin vaccine protects against primary and recurrent Clostridium difcile infection. Infect Immun
2012;80(August (8)):267888.
[41] Palace GP, Lazari P, Norton K. Analysis of the physicochemical interactions between Clostridium difcile toxins and cholestyramine using liquid
chromatography with post-column derivatization. Biochim Biophys Acta
2001;1546(March (1)):17184.
[42] Pruitt RN, Lacy DB. Toward a structural understanding of Clostridium difcile
toxins A and B. Front Cell Infect Microbiol 2012;2(March):28.
[43] Marozsan AJ, Ma D, Nagashima KA, Kennedy BJ, Kang YK, Arrigale RR, et al. Protection against Clostridium difcile infection with broadly neutralizing antitoxin
monoclonal antibodies. J Infect Dis 2012;206(September (5)):70613.

Please cite this article in press as: Karczewski J, et al. Development of a recombinant toxin fragment vaccine for Clostridium difcile
infection. Vaccine (2014), http://dx.doi.org/10.1016/j.vaccine.2014.02.026

535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620