Immunological evaluation of a synthetic clostridium difficile oligosaccharide conjugate vaccine candidate and identification of a minimal epitope

Article abstract of DOI:10.1021/ja401410y

Clostridium difficile is the cause of emerging nosocomial infections that result in abundant morbidity and mortality worldwide. Thus, the development of a vaccine to kill the bacteria to prevent this disease is highly desirable. Several recently identified bacterial surface glycans, such as PS-I and PS-II, are promising vaccine candidates to preclude C. difficile infection. To circumvent difficulties with the generation of natural PS-I due to its low expression levels in bacterial cultures, improved chemical synthesis protocols for the pentasaccharide repeating unit of PS-I and oligosaccharide substructures were utilized to produce large quantities of well-defined PS-I related glycans. The analysis of stool and serum samples obtained from C. difficile patients using glycan microarrays of synthetic oligosaccharide epitopes revealed humoral immune responses to the PS-I related glycan epitopes. Two different vaccine candidates were evaluated in the mouse model. A synthetic PS-I repeating unit CRM₁₉₇ conjugate was immunogenic in mice and induced immunoglobulin class switching as well as affinity maturation. Microarray screening employing PS-I repeating unit substructures revealed the disaccharide Rha-(1→3)-Glc as a minimal epitope. A CRM₁₉₇-Rha-(1→3)-Glc disaccharide conjugate was able to elicit antibodies recognizing the C. difficile PS-I pentasaccharide. We herein demonstrate that glycan microarrays exposing defined oligosaccharide epitopes help to determine the minimal immunogenic epitopes of complex oligosaccharide antigens. The synthetic PS-I pentasaccharide repeating unit as well as the Rha-(1→3)-Glc disaccharide are promising novel vaccine candidates against C. difficile that are currently in preclinical evaluation.

Full text of DOI:10.1021/ja401410y

Article
pubs.acs.org/JACS
Immunological Evaluation of a Synthetic Clostridium difficile
Oligosaccharide Conjugate Vaccine Candidate and Identification of a
Minimal Epitope
Christopher E. Martin,†,∥ Felix Broecker,†,∥ Matthias A. Oberli,†,§ Julia Komor,‡ Jochen Mattner,‡
Chakkumkal Anish,*,† and Peter H. Seeberger*,†
†Max Planck Institute of Colloids and Interfaces, 14424 Potsdam, Germany and Freie Universitat Berlin, 14195 Berlin, Germany
̈
‡Mikrobiologisches Institut − Klinische Mikrobiologie, Immunologie und Hygiene, Universitatsklinikum Erlangen and
̈
Friedrich-Alexander Universitat Erlangen-Nurnberg, D-91054 Erlangen, Germany
̈
̈
S
* Supporting Information
ABSTRACT: Clostridium difficile is the cause of emerging nosocomial infections that result in abundant morbidity and mortality
worldwide. Thus, the development of a vaccine to kill the bacteria to prevent this disease is highly desirable. Several recently
identified bacterial surface glycans, such as PS-I and PS-II, are promising vaccine candidates to preclude C. difficile infection. To
circumvent difficulties with the generation of natural PS-I due to its low expression levels in bacterial cultures, improved chemical
synthesis protocols for the pentasaccharide repeating unit of PS-I and oligosaccharide substructures were utilized to produce large
quantities of well-defined PS-I related glycans. The analysis of stool and serum samples obtained from C. difficile patients using
glycan microarrays of synthetic oligosaccharide epitopes revealed humoral immune responses to the PS-I related glycan epitopes.
Two different vaccine candidates were evaluated in the mouse model. A synthetic PS-I repeating unit CRM197 conjugate was
immunogenic in mice and induced immunoglobulin class switching as well as affinity maturation. Microarray screening
employing PS-I repeating unit substructures revealed the disaccharide Rha-(1→3)-Glc as a minimal epitope. A CRM197-Rha-(1→
3)-Glc disaccharide conjugate was able to elicit antibodies recognizing the C. difficile PS-I pentasaccharide. We herein
demonstrate that glycan microarrays exposing defined oligosaccharide epitopes help to determine the minimal immunogenic
epitopes of complex oligosaccharide antigens. The synthetic PS-I pentasaccharide repeating unit as well as the Rha-(1→3)-Glc
disaccharide are promising novel vaccine candidates against C. difficile that are currently in preclinical evaluation.
preventing colonialization by vaccination against surface
antigens may more effectively limit recurrence than toxin-
neutralizing approaches. Bacterial surface glycans are promising
vaccine targets due to their well-documented role in adhesion
to host tissues.2,16−19 Several cell-surface glycans of C. difficile
have been characterized,20,21 and conjugate vaccines comprised
of isolated or synthetic polysaccharide PS-II are in preclinical
development.22−25
Cell-surface polysaccharide PS-I consisting of a pentasac-
charide phosphate repeating unit [→4)-α-Rhap-(1→3)-β-Glcp-
(1→4)-[α-Rhap-(1→3)]-α-Glcp-(1→2)-α-Glcp-(1→P] was
originally described on ribotype 02722 but has also been
identified on other C. difficile strains.25 Chemical synthesis is
currently the only feasible approach to obtain large amounts of
pure and well-defined PS-I related glycans due to low
expression levels under culture conditions.22,25 PS-I pentasac-
charide repeating unit 1 equipped with an aminopentyl linker at
the reducing end has been synthesized previously (Figure 1).26
Here, we describe the design and chemical synthesis of six
PS-I related oligosaccharides 1−6 (Figure 1). Diglucoside 2
contains the reducing end moieties A and B, while trisaccharide
3 contains one additional glucose moiety C. Rha-(1→3)-Glc
disaccharide 4 is found twice (BD and CD′) in 1, whereas
INTRODUCTION
■
Clostridium difficile is a Gram-positive, spore-forming bacterium
and the most common cause of nosocomial diarrhea world-
wide.1,2 The disruption of the intestinal flora by antibiotics
allows for the colonization with and/or the overgrowth by
drug-resistant, toxin-producing C. difficile spores commonly
found in healthcare facilities, causing C. difficile infection (CDI),
which in its most severe forms is life-threatening.3 In recent
years, infection and death rates have been increasing drastically.
In addition to the main risk group, the elderly, children, young
adults, and pregnant women are now infected, thereby
increasing the social and economic burden.4−6 The emergence
of new C. difficile strains, such as ribotype 027 (also designated
BI, NAP1, or toxin type III), with increased virulence, toxin
production, and antibiotics resistance is partially responsible for
this development.7,8 Ribotype 027 has quickly spread
throughout North America, Europe, Asia, and Oceania9
where it causes growing numbers of infections.5 Vaccination
as an alternative to antibiotic treatment is therefore highly
desirable. Toxin-neutralizing immunization can protect against
the lethal challenge with C. difficile in hamsters,10−12 but toxin-
based vaccines cannot inhibit bacterial colonization, which
precedes toxin production. Furthermore, recurrent CDIs are
serious clinical problems affecting up to 30% of patients after
cessation of therapy, either due to recolonialization by the same
or reinfection with a different C. difficile strain.13−15 Thus,
Received: February 7, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja401410y | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 1. Synthetic PS-I pentasaccharide repeating unit 1 and substructures 2−6.
Figure 2. Monosaccharide building blocks 7−10 employed in the synthesis of oligosaccharides 1−6. Ar = 2-methyl-5-tert-butylphenyl.
a
Scheme 1. Synthesis of Pentasaccharide 1 and Substructures 2 and 3
a
Reagents and conditions: (a) 8, NIS/TfOH, Et2O, −20 to −10 °C, 69%; (b) N2H4·H2O, AcOH/pyridine, DCM, 96%; (c) H2, 10% Pd/C, MeOH,
THF, H2O, AcOH, 99%; (d) 9, NIS/TfOH, DCM, −30 to −17 °C, 92%; (e) cat. Pd(OAc)2, (3,4-dimethoxyphenyl)boronic acid, TBABr, K3PO4,
EtOH; (f) DDQ, aq. NaHCO3, H2O, DCM; (g) TBAF·3H2O, AcOH, DMF, 50 °C, 62% over 3 steps; (h) NaOMe, THF/MeOH; (i) H2, 10% Pd/
C, MeOH, THF, H2O, AcOH, 49% over 2 steps; (j) 10, TMSOTf, DCM, 4 Å MS, −40 to −20 °C, 88%; (k) NaOMe, THF/MeOH; (l) H2, 10%
Pd/C, MeOH, THF, H2O, AcOH, 60% over 2 steps.
trisaccharide 5 covers the sequence BCD′ and 6 consists of a
rhamnose residue (D/D′).
Glycan microarrays containing the C. difficile PS-I related
glycans 1−6 (Figure 1) were used to screen patient samples for
PS-I specific antibodies. The synthetic oligosaccharides
represent natural epitopes that play a role in the immune
response during CDI. Based on the insights gained from glycan
microarray studies, a pentasaccharide 1-CRM197 vaccine
candidate was immunologically evaluated. The glycoconjugate
proved immunogenic in mice, and immunoglobulin class-
B
dx.doi.org/10.1021/ja401410y | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
Scheme 2. Synthesis of Substructures 4−6
a
Reagents and conditions: Synthesis of 4 (A): (a) TBAF·3H2O, AcOH, DMF, 35 °C, 92%; (b) 10, TMSOTf, DCM, 4 Å MS, −40 to −20 °C, 86%;
(c) HO(CH2)5NBnCbz, NIS/TfOH, DCM, −20 to 0 °C, 91%; (d) NaOMe, THF/MeOH; (e) H2, 10% Pd/C, MeOH, THF, H2O, AcOH, 76%
over 2 steps. Synthesis of 5 (B): (f) HO(CH2)5NBnCbz, NIS/TfOH, Et2O, −10 to 0 °C, 39%; (g) N2H4·H2O, AcOH/pyridine, DCM, 81%; (h) 17,
NIS/TfOH, DCM, −20 to 0 °C, 95%; (i) NaOMe, THF/MeOH; (j) H2, 10% Pd/C, MeOH, THF, H2O, AcOH, 77% over 2 steps. Synthesis of 6
(C): (k) HO(CH2)5NBnCbz, TMSOTf, DCM, 4 Å MS, −30 to −20 °C, 94%; (l) NaOMe, THF/MeOH; (m) H2, 10% Pd/C, MeOH, THF, H2O,
AcOH, 94% over 2 steps. Ar = 2-methyl-5-tert-butylphenyl.
switching and affinity maturation were observed. Glycan array-
assisted epitope mapping using synthetic structures 1−6
revealed Rha-(1→3)-Glc disaccharide 4 as minimal epitope of
pentasaccharide 1. Mice immunized with a disaccharide 4-
CRM197 glycoconjugate produced antibodies that recognized
the PS-I pentasaccharide. Pentasaccharide 1 as well as
disaccharide 4 are promising vaccine candidates to protect
from CDI.
compared to our previous approach, where Fmoc was used as
opposed to PBB.26 Following conversion of 13 to diol 14, bis-
glycosylation with rhamnosyl-imidate 10 gave the fully
protected pentasaccharide 15. Global deprotection finally
gave 1. Disaccharide 2 and trisaccharide 3 were obtained by
catalytic hydrogenation of protected 12 and 14, respectively
(Scheme 1).
Oligosaccharides 4 and 5, containing terminal rhamnose
residues, were synthesized relying on disaccharide 17 that in
turn was obtained from building blocks 9 and 10 (Scheme 2).
Rhamnoside 6 resulted from glycosylation of the linker with 10,
followed by deprotection.
RESULTS AND DISCUSSION
■
Synthesis of PS-I Pentasaccharide Repeating Unit 1
and Oligosaccharide Substructures. Pentasaccharide 1 and
substructures 2−6 (Figure 1) were assembled from mono-
saccharide building blocks 7−10 (Figure 2). Two building
blocks (7, 10) had proven reliable previously, whereas the
protecting group patterns of monosaccharides 8 and 9 were
adjusted to overcome challenges encountered during previous
synthetic efforts.26 The nonparticipating benzyl group at C-2 of
thioglucoside 8 allows for α-selective glycosylation to α-
glucoside 7, bearing the linker at the reducing end.26 The two
orthogonal temporary protecting groups para-bromobenzyl
(PBB) ether27,28 at C-3 and levulinic (Lev) ester at C-4 of 8
allow for installation of the branching point. Thioglucoside 9
was freed from the 4,6-O-benzylidene conformational re-
striction utilized previously to exhibit rather two benzyl ethers
in the C-4 and C-6 positions. The tert-butyldimetylsilyl ether
(TBS) at C-3 masks the hydroxyl group serving as attachment
point for rhamnoside 10.
Chemical synthesis of several PS-I related oligosaccharide
structures 1−6 proved to be an efficient approach to obtain
large amounts of pure and well-defined glycans for the
generation of glycan microarrays and glycoconjugates.
Screening of Patient Samples for Antibodies to
Oligosaccharides 1−6. Vaccine candidates and C. difficile
diagnostics have to rely on glycan epitopes that are actually
expressed by the bacterium. As the expression of PS-I is likely
phase variable and thus not yet reliably established under
culture conditions,20,25 we chose the analysis of the native
infection as an indirect way to investigate the antigenic
properties by PS-I through the induction of antibody responses.
We screened samples from C. difficile patients for antibodies to
glycans 1−6 to determine whether these glycans represent true
epitopes of the natural PS-I polysaccharide. Two CDI-negative
control groups were included in this screening: one comprised
of patients without diarrhea and one with patients suffering
from diarrhea not caused by CDI. Initially, we aimed to detect
secreted mucosal IgA antibodies to PS-I since IgA is the
immunoglobulin critical for the control of mucosal infections.
Accordingly, high levels of mucosal IgA to toxin A have been
reported to correlate with protection from CDI.29 While IgA
antibodies to C. difficile toxins act by neutralizing toxicity,30,31
Linear assembly of 1 proceeded from the reducing to the
nonreducing end. Union of thioglucoside 8 and glucoside 7
furnished disaccharide 11 in 69% yield. After cleavage of the
levulinic ester, coupling of thioglycoside 9 and disaccharide 12
proceeded in 92% yield to give trisaccharide 13. A simple
protecting group change more than doubled the yield when
C
dx.doi.org/10.1021/ja401410y | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
IgA to surface antigens is proposed to inhibit colonialization of
C. difficile.32 Anti-C. difficile surface glycan PS-II IgA has been
identified in human fecal samples.23 In order to assess the
importance of the mucosal immune response to the PS-I
pentasaccharide 1 antigen during CDI, we first screened fecal
samples of 36 hospitalized CDI patients that had tested positive
for toxin A/B and the growth of C. difficile in bacterial cultures
obtained from stool specimens by glycan array. Twenty-five of
the patients had severe (persisting watery diarrhea, >5 times per
day for >10 days and/or fever and or blood/pus in stool and/or
abdominal cramping and pain) and 11 less severe (self-limiting
disease, watery diarrhea of <3 times a day for <5 days associated
with mild abdominal cramping and tenderness) clinical
symptoms. Ten individuals not colonized by C. difficile served
as controls. The known PS-II hexasaccharide epitope was also
included in this analysis.23 We observed that mucosal IgA
antibody levels to both PS-I and PS-II antigens were higher in
patients with less severe disease compared to the controls, while
the patients with more severe clinical symptoms had lower IgA
levels, with statistical significance (p ≤ 0.05) compared to the
patients with low disease severity (Figure 3A,B). This indicates
that higher antibody levels to these antigens correlate with
milder forms of the disease. To further follow the development
of mucosal immune responses to PS-I, we screened fecal
samples of 19 CDI patients suffering from acute CDI (tested
positive for toxin A/B and the bacterial growth) for the
presence of IgA to synthetic PS-I related epitopes using glycan
arrays containing saccharides 1−6. Twenty-three age- and sex-
matched controls without CDI (tested negative for toxin A/B)
were included. IgA antibodies specific to glycans 1−6 were
present in most fecal samples of both patient and control
groups, with highly variable antibody levels (Figure 3C). An
increase in specific anti-PS-I IgA was in general observed
around five days after the toxin ELISA became positive in
patient samples that were followed as part of a longitudinal
study. The CDI group showed a trend toward lower IgA levels
to pentasaccharide 1 and rhamnose-containing substructures
4−6, albeit without statistical significance likely due to small
sample sizes (Figure 3C). These findings suggest that low levels
of IgA antibodies to PS-I epitopes are associated with a risk of
acquiring CDI. However, we cannot exclude the possibility that
the lower IgA levels in patients’ feces might be due to
sequestration of antibodies through binding to a larger number
of bacteria.
IgG antibodies to a variety of C. difficile antigens, including
toxins33−36 and surface antigens,36−39 have been detected in
sera of healthy and infected individuals. Elevated serum IgG to
toxin and nontoxin antigens has been associated with
protection from CDI,40−42 suggesting that systemic IgG
responses to C. difficile antigens correlate with clinical
outcomes.43 To assess whether patient blood contains IgG
antibodies to PS-I epitopes, human sera were screened using
glycan microarrays. Blood from 35 reconvalescent patients
(diagnosed with CDI and recovered) and 24 age- and sex-
matched controls without a history of CDI was screened for
IgG recognizing glycans 1−6 using glycan microarrays. IgG
specific to all glycan antigens was present in most sera of both
reconvalescent patients and controls, with a high degree of
variation in antibody levels (Figure 3D). In reconvalescent
patients, statistically significantly higher IgG levels specific for
pentasaccharide 1 as well as disaccharide 2 and trisaccharide 3
were detected (Figure 3D), correlating with results obtained
Figure 3. Detection of antibodies to pentasaccharide 1 and
substructures 2−6 in human samples. (A) Exemplary microarray
scans representing IgA antibodies to the PS-I pentasaccharide 1 and
the PS-II hexasaccharide23 epitopes found in fecal samples from two
patients diagnosed with CDI; one with low disease severity and one
with high disease severity. (B) Secreted IgA levels to both epitopes
from fecal samples CDI patients with either low or high disease
severity and control individuals without diarrhea, as indicated within
the graph, expressed as MFI inferred by glycan array. Horizontal lines
indicate mean values. (C) Secreted IgA levels in fecal samples of CDI
patients and control patients with acute diarrhea due to other reasons
than CDI. (D) Serum IgG levels of reconvalescent CDI patients and
healthy controls. *P ≤ 0.1, **P ≤ 0.05, two-sided student’s t test.
Presence of IgG to rhamnose in human sera has been reported
previously.45,46 Correspondent microarray scans for (B−D) are
depicted in Supporting Information, Figures 1−3. Note that data
points for those patients with undetectable antibody levels are not
shown in the graphs.
with toxin A and B that had also been found to increase upon
infection with C. difficile.44
Elevated IgG to PS-I related antigens might be the result of
exposure of the PS-I antigen to the immune system during
CDI, raising the possibility that PS-I plays a role in immunity
against C. difficile. Nonetheless, it needs to be defined whether
serum IgG antibodies do not only correlate with the
responsiveness of the immune system to the bacterial
polysaccharides but also with the permeability and integrity
of the intestinal epithelium and the immune status of the
respective individual.
In summary, the presence of mucosal secreted IgA and serum
IgG to the PS-I related pentasaccharide 1 and substructures in
D
dx.doi.org/10.1021/ja401410y | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
Scheme 3. Conjugation of Pentasaccharide 1 to CRM197
a
Reagents and conditions: (a) di-N-succinimidyl adipate, Et3N DMSO; (b) CRM197, 100 mM sodium phosphate, pH 7.4.
the majority of individuals indicates that these synthetic glycans
represent biologically relevant epitopes that are recognized by
the humoral immune system during CDI. Increased systemic
IgG after reconvalescence from CDI suggests that an antibody
response to PS-I epitopes is caused by an increased exposure of
the bacterium to the immune system during the infection
process. Furthermore, low mucosal IgA levels might be a risk
factor for acquiring CDI. A part of the population appears to
secrete insufficient amounts of antibodies to prevent the
disease. A vaccine increasing the naturally existing antibody
levels to C. difficile glycan antigens would especially benefit this
group by leading to protection. Thus, the synthetic PS-I related
glycans are promising target molecules for vaccination
approaches against C. difficile.
Preparation and Characterization of a PS-I Pentasac-
charide-CRM197 Conjugate. The observation that the
synthetic PS-I related oligosaccharides 1−6 represent natural
epitopes of C. difficile prompted the immunological evaluation
of a pentasaccharide 1 protein conjugate. Conjugation to
immunogenic carrier proteins improves the usually poor
immunogenicity of oligosaccharide antigens and induces a T-
cell dependent immune response.47 The detoxified diphtheria
toxin variant CRM197 served as carrier protein. This particular
protein is a constituent of licensed conjugate vaccines against
Neisseria meningitidis, Haemophilus influenzae type B, and
Streptococcus pneumoniae.48 To prepare the glycoconjugate,
the primary amine present in the linker group of 1 was reacted
with excess di-N-succinimidyl adipate in DMSO. The resulting
monoester was reacted with the amine side chains of CRM197
lysine residues to obtain glycoconjugate 23 (Scheme 3).
Conjugation was confirmed by SDS-PAGE, showing a shift
toward a higher mass of the glycoconjugate compared with
unconjugated CRM197 and no evidence of unreacted protein
(Supporting Information, Figure 4A). MALDI-TOF mass
spectrometry analysis was used to determine the oligosacchar-
ide-to-CRM197 molar ratio (Supporting Information, Figure
4B). Mass analysis of glycoconjugate 23 revealed that an
average of 9.6 molecules 1 were loaded onto one molecule of
Figure 4. Serum antibody response to pentasaccharide 1 in mice
immunized with glycoconjugate 23. (A) C57BL/6 mice were
immunized with 23 at weeks zero (‘priming’) and two (‘boosting’)
using 3 μg of conjugated pentasaccharide 1 per injection. Sera
obtained at the indicated time points were subjected to microarray
analysis. Serum IgG, IgM, and IgA titers to pentasaccharide 1 at
different time points expressed as MFI values of n = 6 SEM are
shown in (B−D), respectively. White circles, no adjuvant; gray circles,
Alum adjuvant; black circles, Freund’s adjuvant. Serum IgG, IgM, and
IgA titers to pentasaccharide 1 at week 18 are shown in (E−G),
respectively. Bars represent mean values + SEM *P ≤ 0.1, **P ≤ 0.05,
***P ≤ 0.01 (ANOVA with Bonferroni correction). Note that week
18, values of the Alum adjuvant group represent means of n = 4, as two
mice of this group were sacrificed before for monoclonal antibody
development. The complete microarray scans are depicted in
Supporting Information, Figures 5−13.
CRM197
.
Immunological Evaluation of the PS-I Pentasacchar-
ide-CRM197 Conjugate. To assess the ability of glycoconju-
gate 23 to elicit an antibody response to pentasaccharide 1,
three groups of C57BL/6 mice (each group, n = 6) were
immunized, either without adjuvant, with Alum (Alhydrogel)
adjuvant, or with Freund’s adjuvant, in a prime-boost regimen
(Figure 4A).
Alum was selected as adjuvant since it is approved for use in
human vaccines.49 Freund’s adjuvant is an effective adjuvant in
mice that has been successfully employed to raise antibodies to
a synthetic oligosaccharide antigen.23 Each injection comprised
an amount of glycoconjugate corresponding to 3 μg of
pentasaccharide 1. The serum antibody response to 1 was
monitored by glycan microarray analysis up to 18 weeks after
the first injection (Figure 4A). Microarrays also included
CRM197 and a BSA-GlcNAc dummy conjugate to assess
antibody responses to the carrier protein and the generic
spacer moiety (composed of pentyl and adipoyl groups),
respectively. IgG, IgM, and IgA isotype antibodies to
E
dx.doi.org/10.1021/ja401410y | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 5. Serum IgG response to oligosaccharides 1−6 in mice immunized with glycoconjugate 23. (A) C57BL/6 mice were immunized as
described in the caption of Figure 4. Pooled sera (diluted 1:50 in PBS) were subjected to microarray analysis. (B) Representative microarray scans of
week 5 indicating serum IgG to the PS-I derived synthetic epitopes. Note that Alum-adjuvanted 23 predominantly induced antibodies to
pentasaccharide 1, while Freund’s-adjuvanted 23 also elicited high antibody titers to disaccharide 4 and trisaccharide 5. The complete microarray
scan is depicted in Supporting Information, Figure 14. (C) Antibody levels to PS-I derived antigens expressed as MFI values SD of four microarray
spots. White circles, no adjuvant; gray circles, Alum adjuvant; black circles, Freund’s adjuvant. Oligoglucoses 2 and 3 were omitted since responses to
them were almost undetectable. (D) Binding and stability values of antibodies to 1 and 4 in pooled sera of the Freund’s and Alum groups inferred by
SPR are plotted, expressed as percent increase to day zero. Binding and stability correlate with antibody affinity. Weeks after initial immunization are
indicated within the plot. Representative sensorgrams are shown in Supporting Information, Figure 15.
pentasaccharide 1 were detected in all three groups of mice
(Figure 4B−D), demonstrating that immunoglobulin class
switching was induced. Freund’s adjuvant elicited higher
antibody levels than Alum. Nonadjuvanted glycoconjugate
was immunogenic as well but induced a weaker and shorter-
lived antibody response. At week 18, serum IgG levels of the
Freund’s group (serum was diluted 1:100 in PBS) expressed as
mean fluorescence intensity (MFI) were 30-fold higher than
those of the nonadjuvanted group (Figure 4E) with statistical
significance (P ≤ 0.01) and 4-fold higher compared with the
Alum-adjuvanted group (P ≤ 0.1). Serum IgM levels were 12-
fold higher in the Freund’s group compared with the
nonadjuvanted group (P ≤ 0.01) and 4-fold higher compared
with the Alum-adjuvanted group (P ≤ 0.05) (Figure 4F).
Serum IgA levels were 28-fold higher in the Freund’s group
compared with the nonadjuvanted group (P ≤ 0.05) and 4-fold
higher compared with the Alum-adjuvanted group (P ≤ 0.1)
(Figure 4G). The higher levels of serum IgG (8-fold), IgM (4-
fold), and IgA (6-fold) in the Alum-adjuvanted group
compared with the nonadjuvanted group were without
statistical significance (P > 0.1), likely due to the small size
of the Alum group (n = 4) in these comparisons, since two
mice of the Alum group were sacrificed for monoclonal
antibody development prior to week 18. Antibody levels to the
CRM197 carrier protein and the adipoyl spacer moiety were also
increased when the glycoconjugate was administered with
adjuvants (Supporting Information, Figures 6−14). Immuniza-
tion with glycoconjugate 23 was specific for pentasaccharide 1
and did not elicit antibodies against control oligosaccharides, C.
difficile PS-II hexasaccharide23 and Leishmania lipophosphogly-
can capping tetrasaccharide50,51 (Supporting Information,
Figures 5−13).
antigen 1, we subjected pooled sera (diluted 1:50 in PBS) of
mice immunized with glycoconjugate 23 to microarray analysis
with slides containing oligosaccharides 1−6 (Figure 5A).
Antibodies raised using glycoconjugate 23 and Freund’s
adjuvant not only recognized pentasaccharide 1 but also
trisaccharide 5 and, more pronounced, disaccharide 4 (Figures
5B and C). In this group of mice, at week five after initial
immunization, serum IgG levels in the pooled sera expressed as
MFI were highest to pentasaccharide 1 (54 876 5279, mean
SD of four microarray spots), followed by disaccharide 4 (36
484
8679), trisaccharide 5 (29 123
2586), and
monorhamnose 6 (5999
1269). Interestingly, while the
serum IgG levels to pentasaccharide 1, disaccharide 4, and
trisaccharide 5 increased from weeks three to five, those to
rhamnose 6 decreased, indicating that antibodies to mono-
rhamnose are only transiently elicited. Antibodies recognizing
the more complex rhamnose-containing glycan epitopes show a
more robust induction. There was no significant induction of
antibodies recognizing oligoglucose di- (2) and trisaccharide
(3) (Supporting Information, Figure 14), even though both
oligosaccharides appear to be immunodominant epitopes
during natural infection (Figure 3B,C). Thus, contributions of
other microbial infections that elevate responses to 2 and 3
cannot be ruled out. A similar epitope recognition pattern was
observed in mice immunized with nonadjuvanted glycoconju-
gate 23 but with overall lower antibody levels and a shorter-
lived immune response (Figure 5B,C). Here, serum IgG levels
to pentasaccharide 1, trisaccharide 5, disaccharide 4, and
rhamnose 6 peaked at week three and decreased in week five.
At week three, serum IgG levels were highest to pentasacchar-
ide 1, followed by disaccharide 4, trisaccharide 5, and rhamnose
6. These observations indicate that immunizations with
glycoconjugate 23 benefit from administration with an
Glycan Microarray-Assisted Epitope Mapping of PS-I
Pentasaccharide 1. To identify the minimum size epitope of
F
dx.doi.org/10.1021/ja401410y | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
adjuvant. Again, serum IgG to the oligoglucoses 2 and 3 was
not detectable.
disaccharide 4. Even in mice receiving Freund’s adjuvant that
elicited the highest IgG titers, titers against 6 increased only
transiently, indicating that disaccharide 4, and not rhamnose 6,
is the minimal epitope of 1. This minimal epitope, however,
was not accessible when glycoconjugate 23 was administered
with the human-approved Alum adjuvant. Nevertheless
disaccharide 4 is an interesting structure to be investigated as
vaccine candidate as the data suggest that it could induce cross-
reactive antibodies to the pentasaccharide 1. Moreover, it can
be procured with reduced synthetic efforts.
Preparation, Characterization and Immunological
Evaluation of a PS-I Disaccharide 4-CRM197 Conjugate.
Disaccharide 4 emerged as the major antigenic determinant of
pentasaccharide 1. Hence, we expected that immunization with
4 should not only induce antibodies recognizing disaccharide 4
but also pentasaccharide 1. To test this hypothesis, we
conjugated 4 to CRM197 for immunological evaluation.
Conjugation was confirmed by SDS-PAGE and MALDI-TOF
mass spectrometry indicated an average loading of 9.8
molecules disaccharide 4 per molecule CRM197 (Supporting
Information, Figure 16). C57BL/6 mice were immunized with
glycoconjugate 24 in the presence of Freund’s adjuvant in a
prime-boost regimen using 3 μg carbohydrate-based doses per
injection (Figure 6A). We chose the more potent Freund’s
The epitope recognition pattern in mice immunized with
glycoconjugate 23 in the presence of Alum adjuvant differed
from the other two groups. While antibody levels specific for 1
were comparable to mice immunized with Freund’s adjuvant,
relatively low levels of antibodies to epitopes 5 and 4 were
observed. Interestingly, while the serum IgG level to
pentasaccharide 1 increased from week three to five, at the
same time the antibody response to trisaccharide 5,
disaccharide 4, and monorhamnose 6 decreased (Figure 5C).
This result indicates that a robust immune response was
exclusively elicited to the complete pentasaccharide epitope 1.
To gain further insights into the nature of the humoral
immune response to the PS-I related epitopes and the
differences observed in mice immunized with Freund’s and
Alum adjuvants, binding of serum antibodies to epitopes 1 and
4 was analyzed by surface plasmon resonance (SPR). Pooled
sera of these two groups of mice (1:50 dilution in PBS) were
flown over a CM5 sensor chip carrying immobilized epitopes 1
and 4. Antibodies to both and 1 and 4 in the sera of the
Freund’s group showed increasing binding and stability values
over time from week two over three to five, indicating affinity
maturation to both epitopes during the course of the immune
response (Figure 5D). Similarly, antibodies in the sera of the
Alum group elicited increasing affinity to epitope 1 over time.
In contrast, the affinity of serum antibodies to 4 was overall
lower and decreased over time, with a minimum at week five.
These results verify the epitope recognition to 1 and 4 observed
in the glycan array studies (Figure 5B,C).
The differences in the recognition of oligosaccharides 1−6
between the Freund’s and Alum-adjuvanted groups of mice
might be attributed to the nature of the immune responses that
are predominantly elicited by the adjuvants. Freund’s adjuvant
has been shown to evoke primarily a Th1-specific response
profile including the generation of IgG2a, IgG2b, and IgG3
antibodies in mice.52,53 Alum induces primarily a Th2-type
response of IgG1 antibodies.53−55 Th1-type responses have
been linked to epitope spreading, the spread of the immune
response to other areas of the antigen including substructures,
or cryptic epitopes.56 Our data indicate that disaccharide 4 is a
cryptic epitope of pentasaccharide 1 that is accessible only via a
Th1-type response induced by Freund’s adjuvant but not by a
Th2-type response elicited by Alum, as shown by lower
antibody responses to substructure epitopes of 1 and
decreasing affinity to epitope 4 during the course of the
immune response (Figure 5B−D). Further investigations will
need to address these questions and determine the IgG
subclasses induced by glycoconjugates administered with
Freund’s or Alum adjuvant. While Alum as primarily Th2-
directing adjuvant is most commonly used in human-approved
vaccines,49 the oil-in water emulsion adjuvant MF59, approved
for seasonal flu vaccinations,57 has been shown to elicit both
Th1- and Th2-type responses58 and has been successfully
employed to mount an antibody response to oligosaccharide
antigens formulated as CRM197 conjugates in mice.24 MF59 is
thus an interesting human-approved adjuvant for glycoconju-
gate 23 to elicit an antibody response to the cryptic epitope 4.
The response to disaccharide 4 was highest compared to the
other substructures when glycoconjugate 23 was administered;
titers and robustness to 4 were similar to pentasaccharide 1 in
the Freund’s and nonadjuvanted groups of mice. In all three
groups, IgG levels to rhamnose 6 were lower than to
Figure 6. Serum IgG response to oligosaccharides 1−6 in mice
immunized with glycoconjugate 24. (A) Three C57BL/6 mice were
immunized with 24 at weeks zero (‘priming’), two (‘first boosting’)
and five (‘second boosting’) using 3 μg of conjugated disaccharide 4
per injection and Freund’s adjuvant. Two of these mice, 5943 and
5944, mounted an antibody response to disaccharide 4. (B) Serum IgG
levels of two mice to 1−6 expressed as MFI. Data points show mean
SD of n = 4 microarray spots. (C) The mice were boosted with 23 at
the indicated time point. (D) Serum IgG levels of the two mice
expressed as MFI to 1−6. Responses to 2 and 3 were omitted since
they were undetectable. The microarray scans are depicted in
Supporting Information, Figure 17.
adjuvant rather than Alum since we expected lower
immunogenicity of the smaller structure 4 compared with the
pentasaccharide 1. Indeed, glycoconjugate 24 proved less
immunogenic than 23, which is in accordance with the notion
that simple carbohydrates, such as disaccharides, are usually less
immunogenic than more complex ones.61−64 Two boosting
immunizations were required to evoke a detectable immune
response against 4 in the serum of two of three mice (Figure
G
dx.doi.org/10.1021/ja401410y | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
6B). The antibodies also recognized pentasaccharide 1 and
trisaccharide 5 as well as monorhamnose 6. Recognition of the
rhamnose-containing epitopes cannot be explained by
rhamnose-specific antibodies, since mouse 5944, prior to
immunization, had antibodies to monorhamnose 6 that did
neither recognize 1, 4, nor 5 (Figure 6B), and after the second
boosting, IgG titers to 1 and 4 were both higher than those to
6. This evidence lends further credence to the notion that
disaccharide 4 is the minimal epitope of pentasaccharide 1.
These findings underscore that disaccharide 4 is a minimal
size epitope of pentasaccharide 1. Furthermore, the presence of
anti-disaccharide 4 antibodies in human sera and fecal samples
indicates that this disaccharide is a natural epitope with a
possible role during CDI (Figure 3). Thus, disaccharide 4 is a
valid target structure for vaccination approaches against C.
difficile. While oligosaccharide epitopes for conjugate vaccine
development should ideally be larger,59−63 immunogenic
disaccharide epitopes have conferred protection against
Streptococcus pneumoniae type 364 and Candida albicans65 in
preclinical settings. Thus, immunization with disaccharide 4
may protect against CDIs.
only feasible way to investigate whether antibodies raised with
synthetic oligosaccharides 1 and 4 recognize the natural PS-I
polysaccharide on the bacteria, since its expression has not yet
been reliably established under culture conditions.20,25 The
presence of IgA antibodies in the stool specimens of CDI
patients provides indirect evidence for the expression of PS-I
and PS-II by C. difficile in the intestinal tract during infection.
Cross-reactivity of naturally raised antibodies to the employed
oligosaccharides indicates that the oligosaccharides represent
natural epitopes of the PS-I polysaccharide. Challenge studies
will show whether antibodies raised with the synthetic
oligosaccharide epitopes recognize the natural PS-I poly-
saccharide. This will also be an opportunity to investigate
immunization regimes that specifically promote intestinal IgA
secretion like transcutaneous immunization with immunoadju-
vantive proteins, such as C. difficile toxin A,68 which would
complement vaccination with PS-I related oligosaccharides.
In a broader sense the findings reported here are of relevance
for the future design of synthetic carbohydrate antigens.
Despite advances in chemical synthesis,69−74 the assembly of
complex oligosaccharides still poses challenges, particularly on
large scale. Reducing carbohydrate antigens to less complex
substructures by a systematic reductionist approach, delivering
a molecule that still contains the desired antigenic properties is
an important step toward cost-efficient synthetic carbohydrate-
based vaccines. Monoclonal antibodies against pentasaccharide
1 and disaccharide 4 are being developed currently for passive
immunization and to serve as diagnostic tools. Challenge
studies with pentasaccharide 1 and disaccharide 4 antigens will
be performed to evaluate their effectiveness in preventing CDI.
CONCLUSION
■
This study highlights the utility of synthetic oligosaccharides to
identify and define glycan cell-surface epitopes as basis for
antibody and vaccine development. The present focus is placed
on C. difficile that causes serious infections and where a vaccine
is urgently needed.2 The chemical synthesis of PS-I
pentasaccharide repeating unit 1 and related substructures 2−
6 provided access to comparatively large amounts of pure
glycans that were not available by isolation from natural
sources. Glycan microarray analyses revealed that secreted IgA
and serum IgG antibodies to pentasaccharide 1 and its
substructures are found in the majority of the population,
indicating that the synthetic glycans represent naturally
occurring epitopes. Despite the observed interindividual
variations in antibody responses, low IgA titers correlated
with more severe acute disease among CDI patients. The
variations among control individuals and CDI patients might
reflect, for example, differences in the exposure status as
children have been reported to be frequently asymptomatically
colonized.66,67
ASSOCIATED CONTENT
* Supporting Information
■
S
Complete experimental procedures, NMR spectra of key
compounds, Scheme S3, Figures 1−17, and methods. This
information is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
Chakkumkal.Anish@mpikg.mpg.de; Peter.Seeberger@mpikg.
mpg.de
■
It has originally been proposed that PS-I is a stochastically
expressed antigen. However, the failure to reliably detect PS-I
in different C. difficile isolates might be due to phase-variable
expression rather than lack of biosynthetic machinery.25 In
contrast, our glycan array screens suggest that PS-I is a
common antigen expressed in vivo. These findings prompted us
to evaluate PS-I pentasaccharide 1 as vaccine candidate antigen.
Mice immunized with a pentasaccharide 1-CRM197 glyco-
conjugate mounted a robust antibody response to pentasac-
charide 1, including immunoglobulin class switching and
affinity maturation. Glycan microarray technology proved an
elegant means to deduce the minimal antigenic determinant of
complex oligosaccharide antigens. Here, we provide proof-of-
concept for glycan microarray-assisted antigen definition by
reducing the C. difficile PS-I pentasaccharide 1 antigen to
disaccharide 4. Our findings were supported by the observation
that antibodies raised against disaccharide 4-CRM197 glyco-
conjugate recognized pentasaccharide 1. Thus, disaccharide 4,
in addition to pentasaccharide 1, is a promising vaccine
candidate to be taken into challenge trials in animal models.
Recreating the native infection with C. difficile is currently the
Present Address
§Koch Institute for Integrative Cancer Research, Massachusetts
Institute of Technology, Cambridge, Massachusetts 02139,
United States
Author Contributions
∥These authors contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank the Max-Planck Society and the Korber Foundation
■
̈
for generous financial support of the vaccine work. This work
was supported by the German Federal Ministry of Education
and Research (03IS2201G and 0315447). The authors would
like to thank Dr. Bernd Lepenies for advice and support for
animal ethics, Dr. Claney L. Pereira for careful review of the
manuscript, Dr. James C. Klein and Anika Reinhardt for
stimulating and fruitful discussions, and Annette Wahlbrink and
Melanie Leddermann for excellent technical assistance. We
thank Drs. Elizabeth Araujo, Andreas Kremer, und Jochen
H
dx.doi.org/10.1021/ja401410y | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(30) Kim, K.; Pickering, L. K.; DuPont, H. L.; Sullivan, N.; Wilkins,
T. J. Infect. Dis. 1984, 150, 57.
(31) Wada, N.; Nishida, N.; Iwaki, S.; Ohi, H.; Miyawaki, T.;
Taniguchi, N.; Migita, S. Infect. Immun. 1980, 29, 545.
(32) Calabi, E.; Calabi, F.; Phillips, A. D.; Fairweather, N. F. Infect.
Immun. 2002, 70, 5770.
Wacker for obtaining serum samples and Judith Korner and Dr.
Schoerner for obtaining stool specimens. We also acknowledge
Pfenex, Inc. for providing CRM197 at a reduced price for
́
academic institutions.
̈
REFERENCES
(33) Kelly, C. P.; Pothoulakis, C.; Orellana, J.; LaMont, J. T.
Gastroenterology 1992, 102, 35.
■
(1) Knoop, F. C.; Owens, M.; Crocker, I. C. Clin. Microbiol. Rev.
1993, 6, 251.
(34) Nakamura, S.; Mikawa, M.; Nakashio, S.; Takabatake, M.;
Okado, I.; Yamakawa, K.; Serikawa, T.; Okumura, S.; Nishida, S.
Microbiol. Immunol. 1981, 25, 345.
(35) Viscidi, R.; Laughon, B. E.; Yolken, R.; Bo-Linn, P.; Moench, T.;
Ryder, R. W.; Bartlett, J. G. J. Infect. Dis. 1983, 148, 93.
(36) Sanchez-Hurtado, K.; Corretge, M.; Mutlu, E.; McIlhagger, R.;
Starr, J. M.; Poxton, I. R. J. Med. Microbiol. 2008, 57, 717.
(37) Drudy, D.; Calabi, E.; Kyne, L.; Sougioultzis, S.; Kelly, E.;
Fairweather, N.; Kelly, C. P. FEMS Immunol. Med. Microbiol. 2004, 41,
237.
(38) Pechine, S.; Janoir, C.; Collignon, A. J. Clin. Microbiol. 2005, 43,
5018.
(39) Wright, A.; Drudy, D.; Kyne, L.; Brown, K.; Fairweather, N. F. J.
Med. Microbiol. 2008, 57, 750.
(40) Kyne, L.; Warny, M.; Qamar, A.; Kelly, C. P. N. Engl. J. Med.
(2) Rebeaud, F.; Bachmann, M. F. Expert Rev. Vaccines 2012, 11, 469.
(3) Thomas, C.; Stevenson, M.; Riley, T. V. J. Antimicrob. Chemother.
2003, 51, 1339.
(4) Rouphael, N. G.; O’Donnell, J. A.; Bhatnagar, J.; Lewis, F.;
Polgreen, P. M.; Beekmann, S.; Guarner, J.; Killgore, G. E.; Coffman,
B.; Campbell, J.; Zaki, S. R.; McDonald, L. C. Am. J. Obstet. Gynecol.
2008, 198, 635 e1.
(5) Zilberberg, M. D.; Shorr, A. F.; Kollef, M. H. Emerging Infect. Dis.
2008, 14, 929.
(6) Nylund, C. M.; Goudie, A.; Garza, J. M.; Fairbrother, G.; Cohen,
M. B. Arch. Pediatr. Adolesc. Med. 2011, 165, 451.
(7) McDonald, L. C.; Killgore, G. E.; Thompson, A.; Owens, R. C.,
Jr.; Kazakova, S. V.; Sambol, S. P.; Johnson, S.; Gerding, D. N. N. Engl.
J. Med. 2005, 353, 2433.
(8) Warny, M.; Pepin, J.; Fang, A.; Killgore, G.; Thompson, A.;
Brazier, J.; Frost, E.; McDonald, L. C. Lancet 2005, 366, 1079.
(9) Freeman, J.; Bauer, M. P.; Baines, S. D.; Corver, J.; Fawley, W. N.;
Goorhuis, B.; Kuijper, E. J.; Wilcox, M. H. Clin. Microbiol. Rev. 2010,
23, 529.
2000, 342, 390.
(41) Kyne, L.; Warny, M.; Qamar, A.; Kelly, C. P. Lancet 2001, 357,
189.
(42) Aronsson, B.; Granstrom, M.; Mollby, R.; Nord, C. E. Infection
1985, 13, 97.
(43) Kelly, C. P.; Kyne, L. J. Med. Microbiol. 2011, 60, 1070.
(44) Wullt, M.; Noren, T.; Ljungh, A.; Akerlund, T. Clin. Vaccine
Immunol. 2012, 19, 1552.
(45) Oyelaran, O.; McShane, L. M.; Dodd, L.; Gildersleeve, J. C. J.
Proteome Res. 2009, 8, 4301.
(46) Huflejt, M. E.; Vuskovic, M.; Vasiliu, D.; Xu, H.; Obukhova, P.;
Shilova, N.; Tuzikov, A.; Galanina, O.; Arun, B.; Lu, K.; Bovin, N. Mol.
Immunol. 2009, 46, 3037.
(47) Avery, O. T.; Goebel, W. F. J. Exp. Med. 1931, 54, 437.
(48) Shinefield, H. R. Vaccine 2010, 28, 4335.
(10) Fernie, D. S.; Thomson, R. O.; Batty, I.; Walker, P. D. Dev Biol
Stand 1983, 53, 325.
(11) Giannasca, P. J.; Zhang, Z. X.; Lei, W. D.; Boden, J. A.; Giel, M.
A.; Monath, T. P.; Thomas, W. D., Jr. Infect. Immun. 1999, 67, 527.
(12) Kim, P. H.; Iaconis, J. P.; Rolfe, R. D. Infect. Immun. 1987, 55,
2984.
(13) Fekety, R.; McFarland, L. V.; Surawicz, C. M.; Greenberg, R. N.;
Elmer, G. W.; Mulligan, M. E. Clin. Infect. Dis. 1997, 24, 324.
(14) O’Neill, G. L.; Beaman, M. H.; Riley, T. V. Epidemiol. Infect.
1991, 107, 627.
(49) Clements, C. J.; Griffiths, E. Vaccine 2002, 20 (Suppl3), S24.
(50) Hewitt, M. C.; Seeberger, P. H. J. Org. Chem. 2001, 66, 4233.
(51) Hewitt, M. C.; Seeberger, P. H. Org. Lett. 2001, 3, 3699.
(52) Cribbs, D. H.; Ghochikyan, A.; Vasilevko, V.; Tran, M.;
Petrushina, I.; Sadzikava, N.; Babikyan, D.; Kesslak, P.; Kieber-
Emmons, T.; Cotman, C. W.; Agadjanyan, M. G. Int. Immunol. 2003,
15, 505.
(53) Germann, T.; Bongartz, M.; Dlugonska, H.; Hess, H.; Schmitt,
E.; Kolbe, L.; Kolsch, E.; Podlaski, F. J.; Gately, M. K.; Rude, E. Eur. J.
Immunol. 1995, 25, 823.
(54) Lefeber, D. J.; Benaissa-Trouw, B.; Vliegenthart, J. F.;
Kamerling, J. P.; Jansen, W. T.; Kraaijeveld, K.; Snippe, H. Infect.
Immun. 2003, 71, 6915.
(55) Wuorimaa, T.; Kayhty, H.; Leroy, O.; Eskola, J. Vaccine 2001,
19, 1863.
(15) Wilcox, M. H.; Spencer, R. C. J. Hosp. Infect. 1992, 22, 85.
(16) Cywes, C.; Wessels, M. R. Nature 2001, 414, 648.
(17) Soell, M.; Diab, M.; Haan-Archipoff, G.; Beretz, A.; Herbelin,
C.; Poutrel, B.; Klein, J. P. Infect. Immun. 1995, 63, 1380.
(18) Graveline, R.; Segura, M.; Radzioch, D.; Gottschalk, M. Int.
Immunol. 2007, 19, 375.
(19) Weidenmaier, C.; Kokai-Kun, J. F.; Kristian, S. A.; Chanturiya,
T.; Kalbacher, H.; Gross, M.; Nicholson, G.; Neumeister, B.; Mond, J.
J.; Peschel, A. Nat. Med. 2004, 10, 243.
(20) Ganeshapillai, J.; Vinogradov, E.; Rousseau, J.; Weese, J. S.;
Monteiro, M. A. Carbohydr. Res. 2008, 343, 703.
(21) Reid, C. W.; Vinogradov, E.; Li, J.; Jarrell, H. C.; Logan, S. M.;
Brisson, J.-R. Carbohydr. Res. 2012, 354, 65.
(22) Ganeshapillai, J.; Vinogradov, E.; Rousseau, J.; Weese, J. S.;
Monteiro, M. A. Carbohydr. Res. 2008, 343, 703.
(56) Sercarz, E. E. J. Autoimmun. 2000, 14, 275.
(57) Galli, G.; Medini, D.; Borgogni, E.; Zedda, L.; Bardelli, M.;
Malzone, C.; Nuti, S.; Tavarini, S.; Sammicheli, C.; Hilbert, A. K.;
Brauer, V.; Banzhoff, A.; Rappuoli, R.; Del Giudice, G.; Castellino, F.
Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 3877.
(23) Oberli, M. A.; Hecht, M.-L.; Bindschadler, P.; Adibekian, A.;
̈
Adam, T.; Seeberger, P. H. Chem. Biol. 2011, 18, 580.
(24) Adamo, R.; Romano, M. R.; Berti, F.; Leuzzi, R.; Tontini, M.;
Danieli, E.; Cappelletti, E.; Cakici, O. S.; Swennen, E.; Pinto, V.;
Brogioni, B.; Proietti, D.; Galeotti, C. L.; Lay, L.; Monteiro, M. A.;
Scarselli, M.; Costantino, P. ACS Chem. Biol. 2012, 7, 1420.
(25) Bertolo, L.; Boncheff, A. G.; Ma, Z.; Chen, Y. H.; Wakeford, T.;
Friendship, R. M.; Rosseau, J.; Weese, J. S.; Chu, M.; Mallozzi, M.;
Vedantam, G.; Monteiro, M. A. Carbohydr. Res. 2012, 354, 79.
(26) Martin, C. E.; Weishaupt, M. W.; Seeberger, P. H. Chem.
Commun. 2011, 47, 10260.
(27) Liu, X.; Seeberger, P. H. Chem. Commun. 2004, 7, 1708.
(28) Plante, O. J.; Buchwald, S. L.; Seeberger, P. H. J. Am. Chem. Soc.
2000, 122, 7148.
(29) Warny, M.; Vaerman, J. P.; Avesani, V.; Delmee, M. Infect.
Immun. 1994, 62, 384.
(58) O’Hagan, D. T.; Wack, A.; Podda, A. Clin. Pharmacol. Ther.
2007, 82, 740.
(59) Svenson, S. B.; Lindberg, A. A. Infect. Immun. 1981, 32, 490.
(60) Anderson, P. W.; Pichichero, M. E.; Insel, R. A.; Betts, R.; Eby,
R.; Smith, D. H. J. Immunol. 1986, 137, 1181.
(61) Pozsgay, V.; Chu, C.; Pannell, L.; Wolfe, J.; Robbins, J. B.;
Schneerson, R. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 5194.
(62) Phalipon, A.; Costachel, C.; Grandjean, C.; Thuizat, A.;
Guerreiro, C.; Tanguy, M.; Nato, F.; Vulliez-Le Normand, B.; Belot,
F.; Wright, K.; Marcel-Peyre, V.; Sansonetti, P. J.; Mulard, L. A. J.
Immunol. 2006, 176, 1686.
I
dx.doi.org/10.1021/ja401410y | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(63) Verez-Bencomo, V.; Fernandez-Santana, V.; Hardy, E.; Toledo,
M. E.; Rodriguez, M. C.; Heynngnezz, L.; Rodriguez, A.; Baly, A.;
Herrera, L.; Izquierdo, M.; Villar, A.; Valdes, Y.; Cosme, K.; Deler, M.
L.; Montane, M.; Garcia, E.; Ramos, A.; Aguilar, A.; Medina, E.;
Torano, G.; Sosa, I.; Hernandez, I.; Martinez, R.; Muzachio, A.;
Carmenates, A.; Costa, L.; Cardoso, F.; Campa, C.; Diaz, M.; Roy, R.
Science 2004, 305, 522.
(64) Benaissa-Trouw, B.; Lefeber, D. J.; Kamerling, J. P.;
Vliegenthart, J. F.; Kraaijeveld, K.; Snippe, H. Infect. Immun. 2001,
69, 4698.
(65) Bundle, D. R.; Nycholat, C.; Costello, C.; Rennie, R.; Lipinski,
T. ACS Chem. Biol. 2012, 7, 1754.
(66) Bolton, R. P.; Tait, S. K.; Dear, P. R.; Losowsky, M. S. Arch. Dis.
Child. 1984, 59, 466.
(67) Rousseau, C.; Poilane, I.; De Pontual, L.; Maherault, A.-C.; Le
Monnier, A.; Collignon, A. Clin. Infect. Dis. 2012, 55, 1209.
(68) Ghose, C.; Kalsy, A.; Sheikh, A.; Rollenhagen, J.; John, M.;
Young, J.; Rollins, S. M.; Qadri, F.; Calderwood, S. B.; Kelly, C. P.;
Ryan, E. T. Infect. Immun. 2007, 75, 2826.
(69) Plante, O. J.; Seeberger, P. H. Curr. Opin. Drug Discovery Dev.
2003, 6, 521.
(70) Seeberger, P. H.; Werz, D. B. Nature 2007, 446, 1046.
(71) Hecht, M. L.; Stallforth, P.; Silva, D. V.; Adibekian, A.;
Seeberger, P. H. Curr. Opin. Chem. Biol. 2009, 13, 354.
(72) Lepenies, B.; Yin, J.; Seeberger, P. H. Curr. Opin. Chem. Biol.
2010, 14, 404.
(73) Lepenies, B.; Seeberger, P. H. Immunopharmacol. Immunotoxicol.
2010, 32, 196.
(74) Yang, Y.; Martin, C. E.; Seeberger, P. H. Chem. Sci. 2012, 3, 896.
J
dx.doi.org/10.1021/ja401410y | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX