Glycan arrays containing synthetic Clostridium difficile lipoteichoic acid oligomers as tools toward a carbohydrate vaccine

Article abstract of DOI:10.1039/c3cc43545h

Clostridium difficile is a leading cause of severe nosocomial infections. Cell-surface carbohydrate antigens are promising vaccine candidates. Here we report the first total synthesis of oligomers of the lipoteichoic acid antigen repeating unit. Synthetic glycan microarrays revealed anti-glycan antibodies in the blood of patients that help to define epitopes for vaccine development.

Full text of DOI:10.1039/c3cc43545h

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Glycan arrays containing synthetic Clostridium difficile
lipoteichoic acid oligomers as tools toward a
carbohydrate vaccine†
Cite this: Chem. Commun., 2013,
49, 7159
Received 12th May 2013,
Accepted 17th June 2013
Christopher E. Martin,ab Felix Broecker,ab Steffen Eller,a Matthias A. Oberli,z
a
Chakkumkal Anish,a Claney L. Pereira*a and Peter H. Seeberger*ab
DOI: 10.1039/c3cc43545h
www.rsc.org/chemcomm
Clostridium difficile is a leading cause of severe nosocomial infections.
Cell-surface carbohydrate antigens are promising vaccine candidates.
Here we report the first total synthesis of oligomers of the lipoteichoic
acid antigen repeating unit. Synthetic glycan microarrays revealed
anti-glycan antibodies in the blood of patients that help to define
epitopes for vaccine development.
Clostridium difficile is a Gram-positive, spore-forming, anaerobic
bacterium colonizing the intestinal tract that can cause C. difficile
infections (CDIs). CDI is the most common cause of nosocomial
diarrhea worldwide and mostly attributed to antibiotic disruption
of the intestinal flora, allowing for overgrowth of drug-resistant,
toxin-producing C. difficile bacteria.1,2 The structures of three
glycans PS-I, PS-II and lipoteichoic acid (LTA) (also designated
PS-III) expressed on the bacterial cell surface are known.3–5
Due to low expression levels in the bacterial culture, chemical
synthesis is currently the only feasible way to access large
amounts of pure C. difficile glycans. Previous attention focused on
the synthetic repeating units of PS-I and PS-II,6–12 as glycoconjugates
containing synthetic PS-I and PS-II glycans were found to be
immunogenic. Vaccine candidates based on these glycoconjugates
are currently in preclinical development.7,9,10
Fig. 1 Retrosynthetic analysis of C. difficile LTA PS-III oligomers
1 and 2. CE:
2-cyanoethyl, Nap: 2-naphthylmethyl, All: allyl, Lev: levulinate, TCA: trichloroacetimidate.
Here we report the first synthesis of phosphodiester-bridged
oligomers of the LTA PS-III repeating unit. The recently elucidated
PS-III structure contains a glycolipid anchor -6)-b-D-Glcp-(1-6)-b-
D-Glcp-(1-6)-b-D-Glcp-(1-1)-Gro, where glycerol (Gro) is esterified synthesis of phosphorylated repeating unit monomer 1 and
with fatty acids.4 Through a phosphodiester bridge, the glyco- dimer 2 (Fig. 1). The synthetic PS-III oligomers were immobi-
lipid is connected to a polymer consisting of phosphodiester- lized on a microarray, together with synthetic PS-I and PS-II.
linked repeating units with the sequence [-6)-a-D-GlcpNAc- The glycan microarray containing all three known C. difficile
(1-3)-[-P-6]-a-D-GlcpNAc-(1-2)-D-GroA], GroA being glyceric cell-surface glycans was successfully used to detect anti-glycan
acid. The polymer repeating unit structure has so far only antibodies in the sera of CDI patients.§ This helps to define
been described for C. difficile.4 Therefore, we focused on the epitopes for vaccine development and diagnostic applications.
Deprotected oligomers 1 and 2 were obtained from fully
a Max Planck Institute of Colloids and Interfaces, 14424 Potsdam, Germany.
protected 3 and 4; benzyl protecting groups were chosen for
their removal under mild hydrogenolytic conditions (Fig. 1).
Protected monomer 3 was obtained by coupling the free alcohol
of sugar A of repeating unit 5 with phosphoramidite-linker 7.
This reaction proceeded via an intermediate phosphite that was
oxidized in situ to the corresponding phosphate. The vinyl ether
E-mail: ClaneyLebev.Pereira@mpikg.mpg.de, Peter.Seeberger@mpikg.mpg.de
b
¨
Freie Universitat Berlin, 14195 Berlin, Germany
† Electronic supplementary information (ESI) available: Experimental details and
full spectroscopic data for all new compounds. See DOI: 10.1039/c3cc43545h
‡ Current address: Koch Institute for Integrative Cancer Research, Massachusetts
Institute of Technology, Cambridge, MA 02139, USA.
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 7159--7161 7159

View Article Online
Communication
ChemComm
protecting group of sugar B was cleaved under these oxidative
conditions, yielding a free alcohol for further elongation. Key to
the efficient assembly of dimer 4 was repeating unit 6,
equipped with a phosphoramidite at ring A. Elongation at the
free alcohol of 3 with 6 was again followed by oxidative cleavage of
the vinyl ether, yielding 4 with a terminal alcohol. This one-pot
coupling, oxidation and deprotection procedure poses an advantage
over the previously described syntheses of teichoic acid molecules,
that rely on the dimethoxytrityl (DMT) protecting group, which
requires an additional acidic cleavage step.13–15 Repeating unit
5 was obtained by sequential glycosylation of glyceric acid 8 and
glucosamine trichloroacetimidate building blocks 9 and 10.
Non-participating azides were chosen as C-2 substituents in
order to ensure high a-selectivity.
Scheme 2 Synthesis of repeating units 5, 6 and 20. Reagents and conditions:
(a) 9, TMSOTf, DCM, Et2O, 4 Å MS, ꢁ20 1C to ꢁ10 1C, 81% (a : b = 9 : 1); (b) DDQ,
DCM, phosphate buffer, pH 7.2, 0 1C to rt, 80%; (c) 10, TMSOTf, DCM, Et2O,
4 Å MS, ꢁ20 1C to ꢁ10 1C, 69% (a : b = 8 : 1); (d) AcSH, pyridine, 67%;
(e) [Ir(COD)(PMePh2)2]+PF6ꢁ, THF; (f) N2H4ꢀH2O, AcOH, pyridine, DCM, 78% over
2 steps; (g) (iPr2N)2POCE, tetrazole, iPr2NH, 3 Å MS, DCM, MeCN, 82%; (h) I2, H2O,
THF, 89%; (i) H2, Pd/C, EtOH, H2O, AcOH, 78%.
Benzyl-protected (2R)-glyceric acid building block 8 was
synthesized in two steps from O-benzyl-D-serine, similar to
the procedures described for the (S) enantiomer (see ESI†).16,17
Glucosamine building blocks 9 and 10 were synthesized from
1118 and 14,18 respectively (Scheme 1).
The C-3 hydroxyl group of 11 was temporarily protected as a
2-naphthylmethyl (Nap) ether to give 12. After regioselective
opening of 4,6-O-benzylidene acetal, the C-6 alcohol was con-
verted to the levulinic ester. Removal of the anomeric TBS ether
with TBAF solution gave free lactol 13 that was converted to
trichloroacetimidate 9. Building block 10 was prepared in a
similar manner from 14; here the C-6 hydroxyl was converted
into an allyl ether, followed by TBS cleavage and formation of
an anomeric trichloroacetimidate.
Assembly of the a-GlcNAc-(1-3)-a-GlcpNAc-(1-2)-GroA
repeating unit proceeded in a linear manner from building
blocks 8–10 (Scheme 2). TMSOTf mediated glycosylation of
glycosyl-trichloroacetimidate 9 and glyceric acid building block
8 in a mixture of DCM and Et2O gave 16 in high yield and a : b
selectivity of 9 : 1. Removal of the temporary Nap protecting
group with DDQ resulted in 17. Glycosylation between 10 and
17 proceeded under identical conditions to those for the first
glycosylation in good yield and selectivity (a : b = 8 : 1) to give
pseudodisaccharide 18. Conversion of both azides to the
corresponding acetamides with thioacetic acid19 succeeded in
moderate yields to result in fully protected repeating unit 19.
Two step modification of the C-6 groups commenced with
iridium-catalyzed isomerization of the allyl double bond, to
form the corresponding vinyl ether, followed by cleavage of the
levulinic ester with hydrazine to obtain 5. Repeating unit 5 was
further converted to phosphoramidite
6 by employing
bis(diisopropylamino)cyanoethoxyphosphine in the presence
of diisopropylammonium tetrazolide.20–22 A sample of 5 was
subjected to oxidative vinyl ether cleavage with iodine, and the
resulting diol underwent palladium-catalyzed hydrogenation to
yield fully deprotected repeating unit 20. We were able to
confirm the structural assignment of the repeating unit by
comparing NMR spectra of 20 (see ESI†) with those reported
for the dephosphorylated, isolated repeating unit.4
Assembly of C. difficile PS-III oligomers proceeded by con-
densation of phosphoramidites with alcohols. Iodine oxidized
the phosphite intermediate to the corresponding phosphotrie-
ster in situ and simultaneously cleaved the C-6 vinyl ether
protecting group (Scheme 3). In order to achieve complete
conversion, three equivalents of linker phosphoramidite 7
(see ESI†) were reacted with repeating unit 5 in the presence
of 5-(ethylthio)tetrazole (ETT). Subsequent oxidation with
iodine gave phosphotriester 3 in nearly quantitative yield.
Elongation of 3 with 1.4 equivalents of phosphoramidite 6 gave
phosphotriester bridged dimer 4 in 78% yield.
Final deprotection of 3 and 4 proceeded in two steps. First
the 2-cyanoethyl group was cleaved from the phosphotriester to
give the phosphodiester, using triethylamine.23 Removal of the
benzyl groups was achieved by palladium-catalyzed hydrogeno-
lysis at elevated hydrogen pressure (4 bar). Comparison of NMR
spectra obtained for 1 and 2 (see ESI†) with those reported for
the isolated LTA polymer4 showed overall excellent agreement.
Microarrays were prepared by coupling 2 as well as synthetic PS-I8
and PS-II7 (Fig. 2) antigen repeating units on N-hydroxysuccinimide
activated glass slides. The glycan arrays were incubated with sera of
CDI patients.§ Serum IgG antibodies that specifically recognized the
synthetic glycans were visualized using a fluorescence-labelled sec-
ondary antibody. Serum IgG against the PS-I and PS-II repeating
units was detected in eleven and ten of twelve screened samples,
respectively. Anti-2 IgG was detected in half of the samples. Repre-
sentative samples A, B and C are shown in Fig. 2.
Scheme 1 Synthesis of 9 and 10. Reagents and conditions: 1a (a) NapBr, NaH,
DMF, 0 1C to rt, 87%; (b) BH3ꢀTHF, TMSOTf, DCM, 0 1C to 10 1C; (c) LevOH, EDC,
DMAP, DCM; (d) TBAF, AcOH, THF, 0 1C, 84% over 3 steps; (e) CCl3CN, K2CO3,
DCM, quant. 1b (a) NaOMe, MeOH; (b) AllBr, NaH, TBAI, THF, 0 1C to rt; (c) TBAF,
AcOH, THF, 0 1C, 75% over 3 steps; (d) CCl3CN, K2CO3, DCM, 82%.
c
7160 Chem. Commun., 2013, 49, 7159--7161
This journal is The Royal Society of Chemistry 2013

View Article Online
ChemComm
Communication
Scheme 3 Synthesis of C. difficile PS-III oligomers 1 and 2. Reagents and conditions: (a) 5-(ethylthio)tetrazole (ETT), 3 Å MS, MeCN; I2, H2O, THF (3: 98%; 4: 78%);
(b) Et3N; (c) H2 (4 bar), Pd/C, EtOH, H2O, AcOH (1: 63% over 2 steps; 2: 72%).
Notes and references
§ CDI patient sera were kindly provided by Prof. Jochen Mattner
¨
(Universitatsklinikum Erlangen).
1 F. C. Knoop, M. Owens and I. C. Crocker, Clin. Microbiol. Rev., 1993,
6, 251–265.
2 F. Rebeaud and M. F. Bachmann, Expert Rev. Vaccines, 2012, 11, 469–479.
3 J. Ganeshapillai, E. Vinogradov, J. Rousseau, J. S. Weese and
M. A. Monteiro, Carbohydr. Res., 2008, 343, 703–710.
4 C. W. Reid, E. Vinogradov, J. Li, H. C. Jarrell, S. M. Logan and
J. R. Brisson, Carbohydr. Res., 2012, 354, 65–73.
5 M. A. Monteiro, Z. Ma, L. Bertolo, Y. Jiao, L. Arroyo, D. Hodgins,
M. Mallozzi, G. Vedantam, M. Sagermann, J. Sundsmo and H. Chow,
Expert Rev. Vaccines, 2013, 12, 421–431.
Fig. 2 Detection of anti-LTA antibodies in the blood of CDI patients. Represen-
tative microarray scans (samples A, B and C) are shown. The intensity of red spots
corresponds to the abundance of the respective anti-glycan antibodies. Anti-2
antibodies were detected in samples A and B.
6 E. Danieli, L. Lay, D. Proietti, F. Berti, P. Costantino and R. Adamo,
Org. Lett., 2011, 13, 378–381.
¨
7 M. A. Oberli, M.-L. Hecht, P. Bindschadler, A. Adibekian, T. Adam
and P. H. Seeberger, Chem. Biol., 2011, 18, 580–588.
8 C. E. Martin, M. W. Weishaupt and P. H. Seeberger, Chem. Commun.,
2011, 47, 10260–10262.
Sample A contained high IgG levels against PS-I and lower
levels against PS-II and 2. Patient B had IgG primarily against 2
and to a lesser extent against PS-I, while anti-PS-II IgG was
almost undetectable. High levels of anti-PS-II and lower levels
of anti-PS-I IgG were detected for C, anti-2 antibodies were not
present.
Anti-PS-I and anti-PS-II antibodies in patient samples have
been investigated recently.7,12 The presence of anti-2 antibodies
in the blood of CDI patients suggests that PS-III may well be a
relevant C. difficile antigen. Further studies to elucidate the role
of PS-III in CDI are currently underway.
9 R. Adamo, M. R. Romano, F. Berti, R. Leuzzi, M. Tontini, E. Danieli,
E. Cappelletti, O. S. Cakici, E. Swennen, V. Pinto, B. Brogioni,
D. Proietti, C. L. Galeotti, L. Lay, M. A. Monteiro, M. Scarselli and
P. Costantino, ACS Chem. Biol., 2012, 7, 1420–1428.
10 L. Bertolo, A. G. Boncheff, Z. Ma, Y. H. Chen, T. Wakeford,
R. M. Friendship, J. Rosseau, J. S. Weese, M. Chu, M. Mallozzi,
G. Vedantam and M. A. Monteiro, Carbohydr. Res., 2012, 354, 79–86.
11 Y. Jiao, Z. Ma, D. Hodgins, B. Pequegnat, L. Bertolo, L. Arroyo and
M. A. Monteiro, Carbohydr. Res., 2013, DOI: 10.1016/j.carres.2013.03.018.
12 C. E. Martin, F. Broecker, M. A. Oberli, J. Komor, J. Mattner, C. Anish
and P. H. Seeberger, J. Am. Chem. Soc., 2013, 135, 9713–9722.
´
13 W. F. J. Hogendorf, L. J. van den Bos, H. S. Overkleeft, J. D. C. Codee
and G. A. van der Marel, Bioorg. Med. Chem., 2010, 18, 3668–3678.
14 W. F. J. Hogendorf, A. Kropec, D. V. Filippov, H. S. Overkleeft,
Due to low expression levels under culture conditions,
chemical synthesis is currently the only feasible method to
access sufficient amounts of pure PS-III related glycans.
A combination of carbohydrate and phosphoramidite chemistry
led to the first synthesis of PS-III oligomers and confirmation of
the structural assignment of the repeating unit. Antibodies
against synthetic PS-III were found in the blood of infected
patients. Based on these findings, immunological evaluation of
synthetic PS-III glycoconjugates is currently underway. Further-
more, the diagnostic potential of herein described synthetic
antigens will be addressed and compared with the conventional
diagnostic approach that relies on detection of C. difficile toxins
via enzyme-linked immunosorbent assay (ELISA).
´
J. Huebner, G. A. van der Marel and J. D. C. Codee, Carbohydr. Res.,
2012, 356, 142–151.
15 W. F. J. Hogendorf, L. N. Lameijer, T. J. M. Beenakker,
´
H. S. Overkleeft, D. V. Filippov, J. D. C. Codee and G. A. van der
Marel, Org. Lett., 2012, 14, 848–851.
16 C. L. Lencina, A. Dassonville-Klimpt and P. Sonnet, Tetrahedron:
Asymmetry, 2008, 19, 1689–1697.
17 D. X. Hu, M. O’Brien and S. V. Ley, Org. Lett., 2012, 14, 4246–4249.
18 H. A. Orgueira, A. Bartolozzi, P. Schell, R. E. J. N. Litjens,
E. R. Palmacci and P. H. Seeberger, Chem.–Eur. J., 2003, 9, 140–169.
19 N. Shangguan, S. Katukojvala, R. Greenberg and L. J. Williams,
J. Am. Chem. Soc., 2003, 125, 7754–7755.
20 A. D. Barone, J.-Y. Tang and M. H. Caruthers, Nucleic Acids Res.,
1984, 12, 4051–4061.
21 W. Bannwarth and A. Trzeciak, Helv. Chim. Acta, 1987, 70, 175–186.
22 C. M. Pedersen, I. Figueroa-Perez, B. Lindner, A. J. Ulmer,
¨
We acknowledge the Max-Planck Society, the Korber Foun-
¨
U. Zahringer and R. R. Schmidt, Angew. Chem., Int. Ed., 2010, 49,
dation and the German Federal Ministry of Education and
Research [03IS2201G] for generous financial support.
2585–2590.
23 H. M. Hsiung, Tetrahedron Lett., 1982, 23, 5119–5122.
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 7159--7161 7161