Phosphorylation of the synthetic hexasaccharide repeating unit is essential for the induction of antibodies to clostridium difficile PSII cell wall polysaccharide

Article abstract of DOI:10.1021/cb300221f

Clostridium difficile is emerging worldwide as a major cause of nosocomial infections. The negatively charged PSII polysaccharide has been found in different strains of C. difficile and, thereby, represents an important target molecule for a possible carb

Full text of DOI:10.1021/cb300221f

Articles
pubs.acs.org/acschemicalbiology
Phosphorylation of the Synthetic Hexasaccharide Repeating Unit Is
Essential for the Induction of Antibodies to Clostridium difficile PSII
Cell Wall Polysaccharide
Roberto Adamo,*,†,⊥ Maria R. Romano,†,⊥ Francesco Berti,† Rosanna Leuzzi,† Marta Tontini,†
Elisa Danieli,† Emilia Cappelletti,† Osman S. Cakici,† Erwin Swennen,† Vittoria Pinto,† Barbara Brogioni,†
Daniela Proietti,† Cesira L. Galeotti,† Luigi Lay,‡ Mario A. Monteiro,§ Maria Scarselli,†
and Paolo Costantino†
†Research Center, Novartis Vaccines and Diagnostics, Via Fiorentina 1, 53100 Siena, Italy
‡Department of Organic and Industrial Chemistry, University of Milan, Via G. Venezian 21, 20133 Milan, Italy
§Department of Chemistry, University of Guelph, 50 Stone Road East, Guelph, ON, Canada N1G 2W1
S
* Supporting Information
ABSTRACT: Clostridium difficile is emerging worldwide as a major
cause of nosocomial infections. The negatively charged PSII
polysaccharide has been found in different strains of C. difficile
and, thereby, represents an important target molecule for a possible
carbohydrate-based vaccine. In order to identify a synthetic
fragment that after conjugation to a protein carrier could be able
to induce anti-PSII antibodies, we exploited a combination of
chemical synthesis with immunochemistry, confocal immunofluor-
escence microscopy, and solid state NMR. We demonstrate that the
phosphate group is crucial in synthetic glycans to mimic the native
PSII polysaccharide; both native PSII and a phosphorylated
synthetic hexasaccharide repeating unit conjugated to CRM197 elicit
comparable immunogenic responses in mice. This finding can aid
design and selection of carbohydrate antigens to be explored as vaccine candidates.
Clostridium difficile is the most important identifiable cause of
healthcare-acquired diarrhea.1,2 The incidence and mortality
associated with this Gram-positive spore-forming anaerobic
pathogen have dramatically increased in the recent years, with a
reported attack rate of 22.5 cases per 1000 hospital admissions
in Canada and a significantly high mortality rate of 6.9%.3
Outbreaks of severe and recurrent C. difficile infections in the
U.S. and Canada are associated with the spread of the
hypervirulent ribotype 027 strain or North American pulsotype
1 (NAP1).4 This strain has subsequently been registered also in
many European countries,5 although a prevalence of ribotypes
other than 027 has been reported.6
The incidence of this microorganism is increasing in
hospitals worldwide as a consequence of the widespread use
of broad-spectrum antibiotics.7 Treatment failures and
recurrences with antibiotics such as vancomycin and
metronidazole are emphasizing the need for the discovery of
new therapeutic and preventative agents.8 The major virulence
factors of C. difficile are the two large exotoxins A (TcdA) and
B (TcdB).9 In addition, a limited number of isolates also
produce a binary toxin (CDT), whose patho-physiological role
in C. difficile infection (CDI) remains poorly understood.10
Infusion with human mAbs against toxins A and B appears to
be protective against secondary CDI episodes.11 However, the
anticipated high costs associated with this passive immuno-
therapy render a vaccine more attractive in terms of efficacy,
availability, and affordability.1
Recently the structural analysis of the cell wall poly-
saccharides of C. difficile ribotype 027 and two additional
strains, MOH900 and MOH718, identified two different
structures, named PSI and PSII. Only PSII was found to be
expressed by all the three analyzed strains.12 On the basis of the
hypothesis that this polysaccharide might be a conserved
surface antigen, in the perspective of a vaccine that could be
effective in inducing bacterial clearance following infection, we
considered the use of PSII for the development of a possible
carbohydrate-based anti-C. difficile vaccine. The immunoge-
nicity of weak T-independent antigens, such as carbohydrates,
can be increased in quantity and quality by conjugation to
protein carriers, which induce T-cell help.13 This strategy has
already found application in the development of a number of
carbohydrate-based vaccines against Haemophilus influenzae
Received: February 24, 2012
Accepted: May 23, 2012
Published: May 23, 2012
© 2012 American Chemical Society
1420
dx.doi.org/10.1021/cb300221f | ACS Chem. Biol. 2012, 7, 1420−1428

ACS Chemical Biology
Articles
Figure 1. PSII repeating unit and retrosynthetic strategies A20 and B for the preparation of its synthetic fragments.
type b, Neisseria meningitidis, Streptococcus pneumoniae, and
Group B Streptococcus.14
polyanionic polysaccharide antigens.19 Therefore, in conjunc-
tion with the question whether the single repeating unit,
conjugated to a carrier protein, is sufficient to elicit anti PSII
antibodies in mice, we thought it important to also address the
question whther phosphorylation of the synthetic oligosacchar-
ide is necessary.
We previously disclosed the first synthesis of the
phosphorylated hexasaccharide PSII repeating unit 1 bearing
a linker for conjugation to the carrier protein, and its
nonphosphorylated counterpart 2 (Figure 1).20 Synthetic
glycans 1 and 2 are suitable candidates to determine the
contribution of phosphorylation at the nonreducing end
terminal residue of PSII in the polysaccharide immunogenicity.
According to our retrosynthetic scheme (Figure 1), the two
structures 1 and 2 can be prepared from a common
tetrasaccharide AB(D)C intermediate 4, which we considered
per se an interesting fragment to be investigated. Since the β-D-
Glcp-(1−3)-β-D-GalpNAc-(1−4)- (CB) arm is repeated twice
in the hexasaccharide repeating unit, we were interested in
understanding whether this polysaccharide sub unit could still
be immunogenic.
Glycoconjugate vaccines are among the safest and most
efficacious vaccines developed so far. A breakthrough for this
type of vaccine was made in 2004 by a Cuban and Canadian
team15 who reported the large-scale synthesis and the
introduction of a synthetic oligosaccharide vaccine against H.
influenzae type b in humans. Advantages of synthetic
carbohydrate-based vaccines include their well-defined chemical
structures and lack of the impurities present in polysaccharides
obtained from bacterial isolation. In the past decade, a variety of
oligosaccharides resembling the surface carbohydrates of
pathogens have been synthesized, coupled to carrier proteins,
and proved to elicit protective antibodies in animal models.16
Synthetic oligosaccharides, besides being possible vaccine
candidates, can help to elucidate the structural moieties of
the biological polysaccharide that are essential to induce
antibodies.17 This step is crucial for the design of a new
generation of improved and safer vaccines obtained either from
chemical synthesis or bacterial source.
Accordingly we decided to investigate synthetic oligosac-
charides as candidates for a C. difficile conjugate vaccine. PSII is
a polyanionic polysaccharide composed of a hexaglycosyl
phosphate repeating unit: [-6)-β-D-Glcp-(1−3)-β-D-GalpNAc-
(1−4)-α-D-Glcp-(1−4)-[β-D-Glcp-(1-]-β-D-GalpNAc-(1−3)-α-
D-Manp-(1-P].12 Besides C. difficile, other pathogenic bacteria
display negatively charged polysaccharides (carboxylated or
phosphorylated) on their surface. The presence of phospho-
diester-linked repeating units is a structural motif widely
occurring in many pathogens, such as Haemophilus influenzae
type b, serogroup A meningococcus, and serotypes 6 (A, B and
C) and 19 (A and F) of Streptococcus pneumonia.18
The tetrasaccharide AB(D)C 3 would also make it possible
to examine more thoroughly the branching point of the
hexaglycosyl unit, as it has been reported that branches can play
a crucial role determining or deleting the immunogenicity of
short oligosaccharides.17,21 Accordingly, we have developed the
synthesis of the tetrasaccharide subunit 3, via the AB(D)C
intermediate 5.
The preparation of a nonphosporylated hexasaccharide and
its conjugation to the nontoxic mutant of diphteria toxin
CRM197,
22 a carrier protein widely used in commercial vaccines,
was recently reported by Oberli et al.,23 who demonstrated the
immunogenicity of the glycan by the presence of its own
specific IgG antibodies in the serum of immunized mice.
Moreover, specific IgA antibodies in the stool of hospital
patients infected with C. difficile recognized the synthetic PSII
hexasaccharide hapten. However, no proof about the
Although a variety of studies have been conducted to identify
the critical structural parameters that are important for
polysaccharide and conjugate vaccines antigenicity and
immunogenicity, to our knowledge little is known about the
immunodominance of the negatively charged groups in
1421
dx.doi.org/10.1021/cb300221f | ACS Chem. Biol. 2012, 7, 1420−1428

ACS Chemical Biology
Articles
a
Scheme 1. Synthesis of PSII Tetrasaccharide Fragment
a
Reagents and conditions: (i) p-methoxyphenol, BF3·Et2O, CH2Cl2, 93%; (ii) NaOMe, MeOH; PhCH(OMe)2, p-TsOH, CH3CN, 83% (over 2
steps); (iii) NIS, TfOH, CH2Cl2, 98%; (iv) Me3N·BH3, BF3·Et2O, CH3CN, 87%; (v) NIS, TfOH, toluene−dioxane, 82%; (vi) Me3N·BH3, BF3·Et2O,
CH3CN, 99%; (vii) ClAc2O, pyridine−CH2Cl2, 91%; (viii) CAN, CH3CN−H2O, 62%; (ix) Ac2O, pyridine; PhSH, BF3·Et2O, CH2Cl2, 77%; (x)
NaOMe, MeOH; PhCH(OMe)2, p-TsOH, CH3CN, 50% (over 2 steps); (xi) CH3C(OCH2CH3)3, p-TsOH, CH3CN; AcOH−H2O, 67% (over 2
steps); (xii) 21, NIS, TfOH, CH2Cl2, 65%; (xiii) 3 M NaOH, THF; Ac2O, MeOH, 81% (over 2 steps); (xiv) H2, 10% Pd−C, AcOH, EtOH, 74%.
immunogenicity of the native PSII polysaccharide and whether
synthetic fragments can be used in place of it has been hitherto
reported.24
confirming the PSII expression in all the analyzed strains, with a
minor extent in 630 strain.
R20291 strain, also known as Stoke Mandeville (ribotype
027) was selected as the best producer and used for subsequent
fermentation development. The polysaccharide was released by
acidolysis, followed by filtration for inactivation of any spores
(see Supporting Information, Paragraph 1.2). The purification
process was mainly based on fractionation by precipitation with
CaCl2/EtOH and enzymatic treatments to remove protein,
nucleic acid, and peptidoglycan contaminants, followed by
different chromatographic steps to remove low molecular
weight oligosaccharides and contaminants (see Supporting
Information, Figure S1). As a first step 1% CaCl2 and 20%
EtOH were added to precipitate the protein contaminants that
were removed by centrifugation; subsequent increase of the
EtOH concentration to 80% led to PSII precipitation. PSII was
then treated with DNase and RNase for nucleic acid removal,
and mutanolysin, a muralytic enzyme that cleaves β-N-
acetylmuramyl-(1,4)-N-acetylglucosamine linkages, for pepti-
doglycan removal (>1%, see Supporting Information, Para-
graph 1.2).
In the present paper we investigate the role of the phosphate
group in the hexaglycosyl repeating unit 1 in order to
antigenically represent the PSII polysaccharide, exploiting a
combination of chemical synthesis with physicochemical and
immunological techniques that were applied to compare native
PSII and related chemically synthesized structures 1, 2, and 3.
The PSII polysaccharide was isolated and purified from
bacterial growth in view of the coupling to the carrier protein.
The native polymer and the three synthetic glycans 1, 2 and 3
were conjugated to CRM197, and the immunogenicity of the
resulting glycoconjugates was tested in mice. Polyclonal
antibodies raised from immunization with the CRM197
conjugates of the synthetic oligosaccharides 1, 2, 3 and native
PSII were then utilized to verify by confocal microscopy
imaging the accessibility of surface epitopes in C. difficile cell
samples. Analyzed strains were selected on the basis of a fast
screening through solid state NMR analysis of a wide range of
C. difficile bacterial samples collected from clinical settings.
At this stage, after fractionation by precipitation with CaCl2/
EtOH to remove the enzymes, an anionic exchange
chromatography was carried out to separate the low molecular
weight (avDP 2−6) from the higher molecular weight species
(DP> 7). The material, analyzed by the phenol−sulphuric acid
test for saccharide content and by NMR to determine the
averaged degree of polymerization (avDP), was concentrated
by tangential flow filtration (TFF) through a 5 kDa cutoff
membrane and diafiltered in order to reach the appropriate
volume and buffer composition for the next step. The
RESULTS AND DISCUSSION
■
PSII Production, Purification, and Characterization.
Five strains (M68, M120, R20291, 630, Nt2023) were tested to
check PSII polysaccharide expression. The strains were grown,
and after inactivation and extraction, supernatants were
1
analyzed by H NMR. The spectra obtained were consistent
with the published structure,12 and anomeric protons of unit A
at 5.44 ppm, D at 4.95 ppm, B at 4.76 ppm, and B′ at 4.60 ppm
(Man, α-Glc and GalNAc, respectively) were clearly detectable,
1422
dx.doi.org/10.1021/cb300221f | ACS Chem. Biol. 2012, 7, 1420−1428

ACS Chemical Biology
Articles
a
Scheme 2. Conjugation of Synthetic Glycans
a
Reagents and conditions: (i) di-N-hydroxy-succinimidyl adipate ester, Et3N DMSO; (ii) CRM197, 100 mM NaPi, pH 7.
polyanionic PSII was then loaded onto a cationic exchange
chromatography to remove positively charged contaminants.
Finally, size exclusion chromatography was performed to
remove the impurities of low molecular weight.
Protein contamination was determined by MicroBCA assay,
Bradford assay, and aminoacid (AA) analysis; comparing the
results obtained by the different assays, we found that AA
analysis and the Bradford assay are more suitable to assess the
purity of this polysaccharide (see Supporting Information,
Table S1 and Figure S2).
Tetrasaccharide 4 is prepared in turn from the four
monosaccharide building blocks 6, 7, 8, and 9 by first
assembling the linear trisaccharide ABC and then inserting
the α-Glc D unit. Synthon 6 can present a benzyl group at C-6
position for the preparation of compound 2, whereas this
position can be differentiated with an acetate ester for the
selective introduction of the phosphate group in hexasaccharide
1.
Besides the development of strategy A, which has been
extensively discussed in our previous report,20 we explored an
alternative approach leading to the tetrasaccharide AB(D)C 5
starting from building blocks 9, 10, 11, and 12 (Figure 1,
strategy B). In this case the branched trisaccharide B(D)C 17 is
initially assembled to be later coupled to the monosaccharide
acceptor 12 (Scheme 1). For the synthesis of the disaccharide
unit CB we envisioned the use of a p-methoxyphenyl (MP)
moiety for temporary protection of the anomeric position of 14
to be later activated as trichloroacetimidate donor.25
The structure of PSII released by acidolysis and purified was
1
1
verified by H and 31P NMR analysis. H−31P HMBC NMR
indicated that the phosphate-C-1-α-mannose linkage was less
stable in comparison with the C-6-β-glucose-phosphate, and
thus the phosphate group remained linked to the C-6 position
of Glc C′ unit (see Supporting Information, Paragraph 1.3 and
Figure S3).
The 31P NMR profile of the polysaccharide showed two spin
systems, which were assigned to the phosphodiester (Pde) and
phosphomonoester (Pme) signals at 0.5 and −1.5 ppm,
respectively. avDP was calculated by the integral ratios of
selected proton signals (avDP = [H1A/(H1A-α + H1A-β) + 1])
and phosphorus signals (avDP = [(Pde/Pme) + 1]).
For this purpose, glycosylation with 1326 of p-methoxyphe-
nol in the presence of BF3·Et2O afforded exclusively 14, with
unexpected α configuration. The subsequent deacetylation
under Zemplen conditions, followed by 4,6-O-benzylidene
formation, provided the galactosamine acceptor 11. Glyco-
sylation of 11 with donor 10,27 promoted by NIS-TfOH,
afforded disaccharide 15 in 98% yield, which was subjected to
regioselective opening of the benzylidene acetal to furnish
acceptor 16 (87% yield). The glycosylation of 16 with
ethylthioglycoside 928 permitted the stereoselective introduc-
tion of the α-linkage and provided trisaccharide 17 in high yield
(82%).20 After regioselective acetal ring opening of compound
17 and chloroacetylation of the resulting alcohol 18, the p-
methoxyphenyl group was removed with ammonium cerium-
(IV) nitrate (CAN) to give 20.25 Acetylation of the anomeric
position and conversion to phenyl thioglycoside gave the
trisaccharide donor 21.
Design and Synthesis of PSII Related Oligosacchar-
ides. The observation that the phosphate group was bound to
the C-6 of Glc C′ unit suggested the design of the synthesis of
the phosphorylated hexasaccharide repeating unit 1 in order to
mimic the nonreducing end terminal repeating unit of native
PSII and its nonphosphorylated counterpart 2 to determine the
effect of the charged group. As shown in Figure 1 (strategy A),
the two hexasaccharides 1 and 2 can be assembled by a (4 + 2)
convergent approach from a common tetrasaccharide inter-
mediate AB(D)C 4. By this strategy the challenging insertion of
the 1,2-cis glycosidic linkage between residues D and B is
accomplished in an early stage of our multistep synthesis.
1423
dx.doi.org/10.1021/cb300221f | ACS Chem. Biol. 2012, 7, 1420−1428

ACS Chemical Biology
Articles
The mannosyl acceptor 12, bearing the tether for
conjugation, was prepared by methanolysis of 2220 followed
by 4,6-O-benzylidene protection to give diol 23. Regioselective
acetylation of the axial 2-hydroxy group of 23 gave the
mannosyl acceptor 12, which was glycosylated with the
trisaccharide donor 21 to afford the target tetrasaccharide 5
in 65% yield. Eventually, tetrasaccharide 5 was fully deprotected
in three steps, including Troc cleavage with simultaneous esters
saponification, N-acetylation (81% yield over two steps), and
hydrogenolysis (74% yield), providing target 3, which was used
to explore its immunogenicity in comparison with the other
PSII fragments.
Conjugation to CRM197. The conjugation of the
synthesized PSII fragments 1, 2, and 3 was achieved by
reaction of the linker-equipped oligosaccharide with an excess
of di-N-hydroxysuccinimidyl adipate in DMSO (Scheme 2), in
order to obtain the formation of the half ester exclusively,
which was condensed with the amino groups of the protein in
phosphate buffer (pH 7).21 Glycoconjugates using different
molar carbohydrate/protein ratios were prepared, and the
corresponding saccharide loadings determined by MALDI-
TOF mass spectrometry analysis are reported in Table 1.
ranged from 0.2 to 0.3 (w/w), which correspond to 1−2 mols
of PSII (considering a molecular weight of 9−24 kDa) coupled
to the protein.
Immunological Evaluation of Glycoconjugates. PSII-
CRM197 conjugate and neoglycoconjugates from synthetic
glycans were evaluated for their ability to elicit anti PSII
antibodies. With this aim, groups of 8 Balb/C mice were
immunized with 2.5 μg carbohydrate based doses of Tetra-
(26a), Hexa- (28b), HexaP- (30), and PSII-CRM197 (33b)
conjugates, respectively. The biomolecules were formulated
with the adjuvant MF59, an oil in water emulsion frequently
used for seasonal flu vaccination.30 MF59 has been proven to
induce effective long-term protection in the elderly, which is the
same target population estimated for a C. difficile vaccine.31,32
Adjuvant alone in phosphate buffer (PBS) was used as a
negative control.
Sera were analyzed by ELISA for their content of anti PSII
IgG. PSII-HSA, prepared with the same method described for
the CRM197 conjugate, was used for the coating of the plates.
Only CRM197 conjugates obtained from the native poly-
saccharide and the phosphorylated hexasaccharide 1 were able
to induce IgG antibodies which bind to PSII (Figure 2).
In addition, PSII and phosphorylated glycan 1 conjugates
were also the only compounds in the set of analyzed
glycoconjugates that elicited low levels of anti-PSII IgM
antibodies (see Supporting Information, Paragraph 4 and
Figure S6). The tetrasaccharide 3 and the nonphosphorylated
hexasaccharide 2, albeit eliciting self-specific antibodies, showed
neither IgG nor IgM anti-PSII titers, demonstrating that the
presence of the charged phosphate is crucial to induce
antibodies recognizing the polysaccharide.
Table 1. Characteristics of the Synthetic Glycoconjugates
saccharide loading
(mol/mol)
conjugation efficiency
a
compd stoichiometry
(%)
26a
26b
28a
28b
30
40
30
40
30
40
15.5
9.8
39
33
34
33
26
13.6
9.8
10.4
Specificity of the raised IgG toward PSII polysaccharide was
assessed by competitive ELISA using PSII as inhibitor (see
Supporting Information, Figure S7). As it has been reported
that the nonphosphorylated hapten 2 was recognized by IgA
antibodies isolated from patients infected with C. difficile,23 sera
from mice immunized with the PSII-CRM197 conjugate were
used to check their capability to bind the synthetic PSII
fragments conjugated to a non CRM197 protein (for dot-blot
analysis, see Supporting Information Figure S8).
Unlike tetrasaccharide 3, which showed no binding, sera
containing anti-PSII antibodies were found to bind to both 1
and its nonphosphorylated counterpart 2. Antibodies directed
against the PSII polysaccharide can thereby detect the presence
of both the phosphorylated and the nonphosporylated hapten,
but only the synthetic hexasaccharide 1 induces antibodies that
recognize the native PSII. The truncation of the C′B′ arm from
the repeating unit results in loss of immunogenicity, and sera of
mice immunized with PSII conjugates do not contain anti-
tetrasaccharide antibodies. Therefore, the phosphorylated
hexasaccharide 1 represents the minimal structure able to
mimic the native PSII, and the phosphate group (in the form of
either mono or diester) is likely part of the minimal
conformational epitope.
a
Moles saccharide/moles protein used for conjugation, on the basis of
spectrophotometric determination of active ester in the oligosacchar-
ide sample.
The number of sugar chains linked to the carrier proteins
were in agreement with SDS-PAGE profiles (see Supporting
Information, Paragraph 3.1 and Figure S4) and showed that the
shorter and uncharged glycan 3 could be covalently attached
more efficiently to the protein than the longer and charged
hexasaccharide 1, according to a trend that has been previously
observed for similar synthetic oligosaccharides.29
Considering that the reducing Man end was not phosphory-
lated, PSII was coupled to CRM197 by a selective modification
of this sugar to generate aldehyde groups. Reduction of PSII
with NaBH4, followed by a controlled periodate oxidation, gave
the aldehyde groups for subsequent coupling to the lysine
residues of CRM197 via reductive amination (Scheme 3).
PSII oligomers with different avDPs (8, 11, and 21) were
used for glycoconjugates preparation (see Supporting Informa-
tion, Paragraph 3.2). The complete disappearance of
unglycosylated CRM197 (see Supporting Information, para-
graph 3.3 and Figure S4) confirmed that the reaction
conditions enable the efficient conjugation of the protein.
The purification of the conjugates was performed by size
exclusion chromatography that permitted efficient separation
from the unbound saccharide. Total and free saccharide and
protein contents of glycoconjugates are shown in Table 2. After
purification, no signals of either residual aldehyde or hydrated
aldehyde were detected in the 1H NMR spectrum of the
glycoconjugates, indicating that no further quenching of these
highly reactive groups was needed. The degrees of glycosylation
Confocal Immunofluorescence Microscopy and Solid
State NMR on Clinical Isolates. In order to investigate
whether antibodies induced by PSII and the synthetic glycans 1,
2, and 3 conjugated to CRM197 could recognize the native PSII
directly on the C. difficile surface, fixed bacteria were analyzed
by immunofluorescence. The R20291 strain was chosen for this
analysis, as it is the source of PSII purification and the main
polysaccharide producer among the strains analyzed. We found
that bacteria were labeled with antibodies against the purified
1424
dx.doi.org/10.1021/cb300221f | ACS Chem. Biol. 2012, 7, 1420−1428

ACS Chemical Biology
Articles
a
Scheme 3. Conjugation of PSII
a
Reagents and conditions. i. NaBH4, H2O; ii. NaIO4, H2O; iii. NaCNBH3, pH 6.5.
isolates belonging to 6 ribotypes (see Supporting Information),
in addition to R20291 and 630 strains, using the purified PSII
polysaccharide NMR solution spectrum collected as a positive
Table 2. Characteristics of PSII-CRM197 Conjugates
compd DP saccharide/protein (w/w) saccharide loading (mol/mol)
33a
33b
33c
8
21
11
0.31
0.24
0.26
2.07
0.61
1.31
1
control (Figure 4). The anomeric region of the H HR-MAS
PSII and the phosphorylated hexasaccharide 1 CRM197
conjugates (Figure 3a,b); on the contrary, an extremely weak
labeling was observed with antibodies against the tetrasacchar-
ide 3 and the nonphosphorylated hexasaccharide 2 CRM197
conjugates (Figure 3c,d).
Since the bacteria were not uniformly labeled, it can be
hypothesized that the heterogeneity of the surface structures,
containing different carbohydrate antigens in addition to PSII18
could result in bacterial subpopulations with variable staining.
However, in agreement with the immunogenicity data, this
evidence strongly suggests that the purified PSII and the
phosphorylated hexasaccharide 1 CRM197 conjugates induce
antibodies recognizing accessible epitopes on the bacterial
polysaccharide surface.
As the presence of PSII polysaccharide has been demon-
strated in only three strains,12 we considered its expression
within a larger collection of strains an important point to be
evaluated in view of a possible vaccine application. In this
regard we analyzed by 1H HR-MAS NMR a panel of 40 clinical
Figure 2. IgG levels detected in individual post 3 sera (sera collected
two weeks after the third immunization) of Balc/C mice against PSII-
HSA conjugate coating; each dot represents single mouse sera;
horizontal dark bars indicate geometric mean titers of each group with
95% statistical confidence intervals as colored bars.
1425
dx.doi.org/10.1021/cb300221f | ACS Chem. Biol. 2012, 7, 1420−1428

ACS Chemical Biology
Articles
Figure 3. Confocal immunofluorescence microscopy of the R20291 strain using the following sera: α-PSII-CRM197 (a), α-Hexa-P-CRM197 (b), α-
Tetra-CRM197 (c), and α-Hexa-CRM197 (d). Labeling was visualized using secondary anti-IgG antibodies conjugated to red-fluorescent Alexa-
Fluor568. Positive and negative controls were sera against whole fixed bacteria (e) and an unrelated CRM197 conjugate (f), respectively. The
fluorescence images (left panels) were merged to light microscopy images to show overlaps (right panels).
and 126) was selected for confocal immunofluorescence
studies. Sera against PSII-CRM197 and HexaP-CRM197
conjugates were able to stain polysaccharide structures on the
surface of strains belonging to ribotypes 001, 018, 078 and 126,
whereas no staining was detected on strains belonging to 012
(630 strain and an additional clinical isolate) and 014 ribotypes.
However we could confirm the presence of PSII in the 630
strain by GC−MS analysis.
The reasons for this different behavior revealed by confocal
microscopy remain to be clarified. Inspection of publicly
available genome sequences shows that the putative capsule
biosynthesis locus (CD2768−CD2780) is well conserved
across strains belonging to ribotypes 012, 027, 001, and 078
1
(strains 630, R202291, ATCC43255, and QCD-23m63
respectively),33−35 leading to excluding that such differences
can be ascribed to the absence of specific genes.
We could hypothesize that PSII in the 012 and 014 ribotypes
was not efficiently accessible to antibodies, due to the presence
of other surface structures. Alternatively, the inability to detect
PSII could be exclusively due to the specific experimental
conditions, which do not reflect the in vivo regulation
mechanisms in these strains.
Conclusion. It is becoming more and more evident that C.
difficile is the cause for many nosocomial infections, especially
in the U.S. and Canada. In the past decade, the organism has
evolved to become more virulent and resistant to antibiotics.
Figure 4. Comparison of the H HR-MAS NMR (400 MHz, 90%
H2O + 10% D2O, 298 K) spectra collected from the two different
1
strains R20291 (a) and B1 (b) and the PSII H NMR (400 MHz,
D2O, 298 K) spectrum of R20291 measured in solution (c).
NMR spectra (signals at 5.40, 4.95, 4.70, 4.60 ppm) was used as
fingerprint to identify PSII in the heterogeneous phase of the
analyzed samples from inactivated cells, except for strains 630
(012 ribotype) and 014 where anomeric signals were not clearly
detectable.
To support the NMR data and evaluate the presence of the
surface polysaccharide in the context of intact bacteria, a
representative strain for each ribotype (001, 012, 014, 018, 078
1426
dx.doi.org/10.1021/cb300221f | ACS Chem. Biol. 2012, 7, 1420−1428

ACS Chemical Biology
Articles
We investigated the possibility of using the PSII polysaccharide,
isolated from the bacterial surface of the hypervirulent epidemic
strain BI/NAP1/027, as a potential antigen for a glycoconju-
gate vaccine. To achieve a better understanding of the relevant
epitopes in the polysaccharide and highlight possible synthetic
candidates, we applied a combination of chemical synthesis
with physicochemical and immunochemical techniques. In
particular we focused on elucidating the role of the phosphate
group in PSII repeating unit.
A procedure for PSII purification was established, and native
polysaccharide and synthetic glycans 1, 2 and 3 were
conjugated to CRM197. Immunological evaluation of the
glycoconjugates in mice highlighted the on/off role of the
phosphate group to mimic PSII, as only the conjugated
hexasaccharide 1 and the native polysaccharide were able to
elicit comparable IgG titers against PSII.
AUTHOR INFORMATION
Corresponding Author
*E-mail: roberto.adamo@novartis.com.
■
Author Contributions
⊥These authors contributed equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We gratefully acknowledge N. Norais and M. Biagini for their
kind contribution in discussion on C. difficile glycoproteins.
■
REFERENCES
■
(1) Koo, H. L., Garey, K. W., and DuPont, H. L. (2010) Future novel
therapeutic agents for Clostridium difficile infection. Expert Opin. Invest.
Drugs, 825−836.
By immunofluorescence microscopy we assessed that surface
PSII expressed by R20291 strain is accessible in vitro by sera
against PSII-CRM197 and HexaP-CRM197 conjugates, thus
confirming the ELISA data. Concurrent utilization of sera
from mice immunized with conjugated native polysaccharide
and synthetic glycans proved that the staining in confocal
microscopy can be correlated only to the presence of PSII on
the bacterial cell wall, thereby excluding any possible surface
contaminant.
(2) Knoop, F. C., Owens, M., and Crocker, I. C. (1993) Clostridium
difficile: clinical disease and diagnosis. Clin. Microbiol. Rev 6, 251−265.
(3) Loo, V. G., Poirier, L., Miller, M. A., Oughton, M., Libman, M. D.,
Michaud, S., Bourgault, A.-M., Nguyen, T., Frenette, C., Kelly, M.,
Vibien, A., Brassard, P., Fenn, S., Dewar, K., Hudson, T. J., Horn, R.,
́
Rene, P., Monczak, Y., and Dascal, A. (2005) A predominantly clonal
culti-institutional outbreak of Clostridium difficile-associated diarrhea
with high morbidity and mortality. N. Engl. J. Med. 353, 2442−2449.
(4) Warny, M., Pepin, J., Fang, A., Killgore, G., Thompson, A.,
Brazier, J., Frost, E., and Clifford McDonald, L. (2005) Toxin
production by an emerging strain of Clostridium difficile associated
with outbreaks of severe disease in North America and Europe. Lancet
366 (9491), 1079−1084.
PSII expression was finally analyzed on a panel of clinical
1
isolates by H HR-MAS NMR and confocal immunofluor-
escence microscopy. The presence of the surface PSII
polysaccharide was revealed for ribotypes 027, which is
endemic in North America,4 and in strains belonging to
ribotypes 001, 078 and 126, known to have significant
incidence in causing severe nosocomial disease in Europe.6
Therefore, by a combination of chemical synthesis,
immuchemistry, solid state NMR, and confocal microscopy
we can identify the native PSII and the phosphorylated
synthetic hexasaccharide repeating unit 1 as two possible
candidates for either a glycoconjugate or a multiantigen vaccine
against C. difficile. Our finding highlights for the first time, by
well-defined synthetic structures, that phosphorylation of the
terminal residue of short length fragments of phosphodiester-
linked glycopolymers is an essential determinant for their
immunogenicity. Noteworthy, a single synthetic repeating unit
phosphorylated at the terminal residue was in our case
sufficient to mimic the bacterial native phosphopolymer.
Since phosphodiester linkages occur in the surface-exposed
polysaccharides of many pathogenic bacteria, we expect that
these results will open new perspectives in the design and
selection of carbohydrate antigens to be explored as vaccine
candidates.
(5) Kuijper, E. J., Barbut, F., Brazier, J. S., Kleinkauf, N., Eckmanns,
T., Lambert, M. L., Drudy, D., Fitzpatrick, F., Wiuff, C., Brown, D. J.,
Coia, J. E., Pituch, H., Reichert, P., Even, J., Mossong, J., Widmer, A.
F., Olsen, K. E., Allerberger, F., Notermans, D. W., Delmee
́
, M.,
Coignard, B., Wilcox, M., Patel, B., Frei, R., Nagy, E., Bouza, E., Marin,
M., Akerlund, T., Virolainen-Julkunen, A., Lyytikainen, O., Kotila, S.,
̈
Ingebretsen, A., Smyth, B., Rooney, P., Poxton, I., Monnet., R., and
Kuijper, D. L. (2008) Update of Clostridium difficile infection due to
PCR ribotype 027 in Europe. Eurosurveillance 13 (31), 18942−18949.
(6) Bauer, M. P., Notermans, D. W., van Benthem, B. H., Brazier, J.
S., Wilcox, M. H., Rupnik, M., Monnet, D. L., van Dissel, J. T., and
Kuijper, E. J. (2011) Clostridium difficile infection in Europe: a
hospital-based survey. Lancet 377 (9759), 63−73.
(7) Kyne, L., Hamel, M. B., Polavaram, R., and Kelly, C. P. (2002)
Health care costs and mortality associated with nosocomial diarrhea
due to Clostridium difficile. Clin. Infect. Dis. 34, 346.
(8) Gerding, D. N., Muto, C. A., and Owens, R. C. (2008) Treatment
of Clostridium difficile infection. Clin. Infect. Dis. 46, S32−S42.
(9) Pruitt, R. N., Chambers, M. G., Ng, K. K.-S., Ohib, M. D., and
Lacy, D. B. (2010) Structural organization of the functional domains of
Clostridium difficile toxins A and B. Proc. Natl. Acad. Sci. U.S.A. 107,
13467−13472.
(10) Sun, X., Savidge, T., and Feng, H. (2010) The Enterotoxicity of
Clostridium difficile Toxins. Toxins 2, 1848−1880.
(11) Parks, T., Gkrania-Klotasas, E., and Nicholl, C. (2010)
Treatment with monoclonal antibodies against Clostridium difficile
toxins. N. Engl. J. Med. 362, 1444−1445.
(12) Ganeshapillai, J., Vinogradov, E., Rousseau, J., Scott Weese, J.,
and Monteiro, M. A. (2008) Clostridium difficile cell-surface
polysaccharides composed of pentaglycosyl and hexaglycosyl phos-
phate repeating units. Carbohydr. Res. 343, 703−710.
(13) Broker, M., Dull, P. M., Rappuoli, R., and Costantino, P. (2009)
Chemistry of a new investigational quadrivalent meningococcal
conjugate vaccine that is immunogenic at all ages. Vaccine 27,
5574−5580.
ASSOCIATED CONTENT
■
S
* Supporting Information
Materials and methods for PSII isolation, purification, and
characterization; experimental procedures and characterization
of new compounds, NMR spectra of compounds 14, 11, 15, 17,
19, 23, 12, and 3; procedures for preparation of glycoconju-
gates and their characterization; immunochemical analysis;
methods for confocal immunofluorescence microscopy and
HR-MAS NMR. This material is available free of charge via the
Internet at http://pubs.acs.org.
(14) Costantino, P., Rappuoli, R., and Berti, F. (2011) The design of
semi-synthetic and synthetic glycoconjugate vaccines. Expert Opin.
Drug Discovery 6 (10), 1045−1067.
1427
dx.doi.org/10.1021/cb300221f | ACS Chem. Biol. 2012, 7, 1420−1428

ACS Chemical Biology
Articles
(15) Verez-Bencomo, V., Fernandez-Santana, V., Hardy, E., Toledo,
M. E., Rodrıguez, 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., Martınez, R., Muzachio, A.,
Carmenates, A., Costa, L., Cardoso, F., Campa, C., Diaz, M., and Roy,
R. (2004) A synthetic conjugate polysaccharide vaccine against
Haemophilus influenzae type b. Science 305, 522−525.
(16) Morelli, L., Poletti, L., and Lay, L. (2011) Carbohydrates and
immunology: synthetic oligosaccharide antigens for vaccine formula-
tion. Eur. J. Org. Chem., 5723−5777.
(17) Safari, D., Dekker, H. A. T., Joosten, J. A. F., Michalik, D.,
Carvalho de Souza, A., Adamo, R., Lahmann, M., Sundgren, A.,
Oscarson, S., Kamerling, J. P., and Snippe, H. (2008) Identification of
the smallest structure capable of evoking opsonophagocytic antibodies
against Streptococcus pneumoniae Type 14. Infect. Immun. 76 (10),
4615−4623.
(18) Nikolaev, A. V., Botvinko, I. V., and Ross, A. J. (2007) Natural
phosphoglycans containing glycosyl phosphate units: structural
diversity and chemical synthesis. Carbohydr. Res. 342, 297−344.
(19) Anderson, P. W., Pichichero, M. E., Stein, E. C., Porcelli, S.,
Betts, R. F., Connuck, D. M., Korones, D., Insel, R. A., Zahradnik, J.
M., and Eby, R. (1989) Effect of oligosaccharide chain length, exposed
terminal group, and hapten loading on the antibody response of
human adults and infants to vaccines consisting of Haemophilus
influenzae type b capsular antigen uniterminally coupled to the
diphteriae protein CRM197. J. Immunol. 142 (7), 2464−2468.
(20) Danieli, E., Lay, L., Proietti, D., Berti, F., Costantino, P., and
Adamo, R. (2011) First synthesis of C. difficile PS-II Cell wall
polysaccharide repeating unit. Org. Lett. 13 (3), 378−381.
(31) O’Hagan, D. T., Rappuoli, R., De Gregorio, E., Tsai, T., and Del
Giudice, G. (2011) MF59 adjuvant: the best insurance against
influenza strain diversity. Expert Rev. Vaccines 10 (4), 447−462.
(32) van den Brand, J. M. A., Kreijtz, J. H. C. M., Bodewes, R.,
Stittelaar, K. J., van Amerongen, G., Kuiken, T., Simon, J., Fouchier, R.
A. M., Del Giudice, G., Rappuoli, R., Rimmelzwaan, G. F., and
Osterhaus, A. D. M. E. (2011) Efficacy of vaccination with different
combinations of MF59-adjuvanted and nonadjuvanted seasonal and
pandemic influenza vaccines against pandemic H1N1 (2009) influenza
virus infection in ferrets. J. Virol. 85 (6), 2851−2858.
(33) Sebaihia, M., Wrem, B. W., Mullany, P., Fairwheather, N. F.,
Minton, N., Stabler, R., Thomson, N. R., Roberts, A. P., Cerdeno-
̃
́
Tarraga, A. M., Wang, H., Holden, M. T., Wright, A., Churcher, C.,
Quail, M. A., Baker, S., Bason, N., Brooks, K., Chillingtworth, T.,
Cronini, A., Davis, P., Dowd, L., Fraser, A., Feltwel, T., Hance, Z.,
Holroyd, S., KìJagels, K., Moule, S., Mungall, K., Price, C.,
Rabbinowitsch, E., Sharp, S., Simmonds, M., Steven, K., Unwin, L.,
Whithead, S., Dupuy, B., Dougan, G., Barrell, B., and Parkhill, J.
(2006) The multidrug-resistant human pathogen Clostridium difficile
has a highly mobile, mosaic genome. Nat. Genet. 38 (7), 779−786.
(34) He, M., Sebaihia, M., Lawley, T. D., Stabler, R. A., Dawson, L.
F., Martin, M. J., Holt, K. E., Seth-Smith, H. M., Quail, M. A., Rance,
R., Brooks, K., Churcher, C., Harris, D., Bentley, S. D., Burrows, C.,
Clark, L., Corton, C., Murray, V., Rose, G., Thurston, S., van Tonder,
A., Walker, D., Wren, B. W., Dougan, G., and Parkhill, J. (2010)
Evolutionary dynamics of Clostridium difficile over short and long time
scales. Proc. Natl. Acad. Sci. U.S.A. 107 (6), 7527−7532.
(35) Forgetta, V., Oughton, M. T., Marquis, P., Brukner, I.,
Blanchette, R., Haub, K., Magrini, V., Gerding, D. L., Loo, V. G.,
Mile, M. A., Mulvey, M. R., Rupnik, M., Dascal, A., and Dewar, K.
(2011) Fourteen-genome comparison identifies DNA markers for
severe-disease-associated strains of Clostridium difficile. J. Clin.
Microbiol. 49 (6), 2230−2238.
(21) Adamo, R., Tontini, M., Brogioni, G., Romano, M. R.,
Costantini, G., Danieli, E., Proietti, D., Berti, F., and Costantino, P.
(2011) Synthesis of laminarin fragments and evaluation of a β-(1,3)
glucan hexasaccaride-CRM197 conjugate as vaccine candidate against
Candida albicans. J. Carbohydr. Chem. 30, 249−280.
(22) Giannini, G., and Rappuoli, R. (1984) The amino acid sequence
of the two non-toxic mutants of diphtheria toxin: CRM45 and CRM197
.
Nucleic Acids Res. 25 (12), 4063−4069.
(23) Oberli, M. A., Hecht, M.-L., Bindschadler, P., Adibekian, A.,
Adam, T., and Seeberger, P. H. (2011) A possible oligosaccharide-
conjugate vaccine candidate for Clostridium difficile is antigenic and
immunogenic. Chem. Biol. 18, 580−588.
(24) Preliminary results on immunogenicity of PSII-CRM197 and
HexaP-CRM197 were presented by Danieli, E., Tontini, M., Romano,
M. R., Berti, F., Costantino, P., and Adamo, R. at 16th Eurocarb, 3−7
July 2011, Sorrento (Italy).
(25) Mori, M., Ito, Y., and Ogawa, T. (1989) A highly stereoselective
and practical synthesis of cyclomannohexaose, cyclo (→4)-[α-D-Man]
5- α-D-Man-(1→). Carbohydr. Res. 192, 131−146.
(26) Higashi, K., and Susaki, H. (1992) A novel glycosydation
promoted by trimethylsilyl halide and zinc triflate. Chem. Pharm. Bull.
40, 2019−2022.
́ ́
(27) Chenede, A., Perrin, E., Rekaï, E. D., and Sinay, P. (1994)
̈
Synthesis of C-disaccharides by a samaium diiodie induced intra-
molecular coupling reaction. Synlett, 420−422.
(28) Mandal, M., Dukdin, V. Y., Geng, X., and Danishefsky, S. J.
(2004) In pursuit of carbohydrate-based HIV vaccines, part 1: the total
synthesis of hybrid-type gp120 fragments. Angew. Chem., Int. Ed. 43,
2557−2561.
(29) Lefeber, D. J., Kamerling, J. P., and Vliegenthart, J. F. G. (2001)
Synthesis of Streptococcus pneumoniae type 3 neoglycoproteins varying
in oligosaccharide chain length, loading and carrier protein. Chem.
Eur. J. 7 (20), 4411−4421.
(30) 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., and Castellino,
F. (2009) Adjuvanted H5N1 vaccine induces early CD4+ T cell
response that predicts long-term persistence of protective antibody
levels. Proc. Natl. Acad. Sci. U.S.A. 106, 3877−3882.
1428
dx.doi.org/10.1021/cb300221f | ACS Chem. Biol. 2012, 7, 1420−1428