Automated solid phase synthesis of teichoic acids

Article abstract of DOI:10.1039/c1cc13132j

This communication describes the first automated solid phase synthesis of teichoic acids (TAs) and the preparation by this method of a number of well-defined TA structures, which were probed for their antigenicity. An opsonophagocytic killing assay reveal

Full text of DOI:10.1039/c1cc13132j

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 8961–8963
www.rsc.org/chemcomm
COMMUNICATION
Automated solid phase synthesis of teichoic acidswz
Wouter F. J. Hogendorf,a Nico Meeuwenoord,a Herman S. Overkleeft,a Dmitri V. Filippov,a
Diana Laverde,b Andrea Kropec,b Johannes Huebner,b Gijsbert A. Van der Marel*a and
a
´
Jeroen D. C. Codee*
Received 27th May 2011, Accepted 27th June 2011
DOI: 10.1039/c1cc13132j
This communication describes the first automated solid phase
synthesis of teichoic acids (TAs) and the preparation by this
method of a number of well-defined TA structures, which were
probed for their antigenicity. An opsonophagocytic killing assay
revealed a clear TA-length–activity relationship and indicated a
promising candidate for future vaccine development.
methodology for the synthesis of TA structures, its use in the
generation of a small TA-library and the initial assessment of
the antigenic properties of the synthetic TA fragments.
Our first focus was the generation of TA structures
consisting of repeating glycerol phosphate residues, as present
on the cell wall of Gram-positive bacteria like Staphylococcus
aureus and the commensal bacterium Enterococcus faecalis.1
Whereas the health threat of S. aureus is widely recognized,
the latter species has been regarded to be relatively harmless.
Recently, however it has become clear that E. faecalis can
present a serious risk, especially to immunocompromized
individuals, causing bacterimea, urinary tract infections,
peritonitis and endocarditis.6 The presence of multiple
antibiotic resistant determinants in most clinically relevant
enterococci isolates urges the development of alternative treatment
and prevention strategies.7 In this context, the determination
of specific antigenic epitopes can form the basis for future
vaccine development.
Based on the state-of-the-art in nucleic acid synthesis8 we
decided to employ cyanoethyl protected phosphoramidite
building blocks9 of which the alcohol function to be elongated
was protected with the dimethoxytrityl (DMT) group to allow
mildly acidic, yet fast and efficient cleavage and simultaneous
online monitoring of coupling efficiency. The automated
synthesis of the TA oligomers was executed on controlled
pore glass (CPG), with a base labile succinyl linker using a
Teichoic acids (TAs) are polyanionic glycopolymers built up
from repeating alditol (e.g. glycerol, ribitol or mannitol)
phosphate units, which are seemingly randomly decorated
with D-alanine esters and carbohydrate substituents. They
are prominent constituents of the cell wall of Gram-positive
bacteria,1 and are important for cell wall integrity, nutrient
uptake, cation homeostasis, neutralizing antibiotics. They
mediate various extracellular interactions and since they
protrude from the bacterial cell wall they interact with our
immune system. In this regard they have been implied as
antigenic structures and as immunostimulatory molecules that
act on the innate immune system through interaction with
Toll-like receptors.2 Because of the microheterogeneity of
TAs, the establishment of structure–activity relationships has
been a challenge and the biological activity of TA preparations
has been the subject of debate.3 We therefore set out to develop
synthetic strategies to produce well-defined structures.4,5 The
repetitive nature of the TA structures and the fact that the
repeating units are connected through phosphodiester linkages
make them attractive synthetic targets for an automated solid
phase approach. Besides the potential to rapidly access a
library of compounds, automated solid phase synthesis is also
very attractive for the assembly of longer oligomers, for which
solution phase synthesis is too time and labour intensive.
We here describe the development of automated solid phase
¨
commercially available AKTAt oligopilott synthesizer. To
investigate the use of C-2 functionalized glycerol phos-
phoramidite building blocks we explored the incorporation
of a-glucosyl substituted glycerol phosphates in our synthetic
scheme, because a-glucopyranose is a common substituent of
S. aureus TA. With the aim to provide TA fragments that can
be functionalized, an aminohexanol spacer was incorporated.
The following building blocks were required: glycerol phos-
phoramidite 1a and a-glucosyl glycerolphosphoramidite 2a, to
build up the glycerol phosphate chain, glycerol succinate 1b
and 2b to functionalize aminopropyl-CPG, and aminohexanol
phosphoramidite 3 to terminate the TA chains. We previously
reported5 the synthesis of glycerol 1 and spacer 3 and the
synthesis of the glucosyl building block 2 is depicted in
Scheme 1.z
a Leiden Institute of Chemistry, Leiden University, P.O. Box 9502,
2300 RA Leiden, The Netherlands.
E-mail: jcodee@chem.leidenuniv.nl, marel_g@chem.leidenuniv.nl;
Fax: +31-5274307; Tel: +31-5274280
b Division of Infectious Diseases, Department of Medicine,
University Medical Center Freiburg, Hugstetter Strasse 55,
79106 Freiburg, Germany
w This article is part of the ChemComm ‘Glycochemistry and glycobiology’
web themed issue.
z Electronic supplementary information (ESI) available: Experimental
procedures and characterization data for all new compounds. See
DOI: 10.1039/c1cc13132j
A key step in the construction of building block 2a/b is the
stereoselective introduction of the glycosidic bond, which was
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 8961–8963 8961

View Article Online
5-benzylthiotetrazole (5-BTT, 22.5 equiv.) as an activator12
in acetonitrile for 5 minutes. Oxidation of the intermediate
phosphites was achieved with I2 in pyridine/H2O (1 min), after
which a capping step (N-methylimidazole/acetic anhydride
(Ac2O)/2,6-lutidine) was introduced to cap any unreacted
alcohol functionalities. To complete the elongation cycle the
DMT-groups were removed with 3% DCA in toluene. The
coupling efficiency using this protocol was generally 498% as
judged from the automatic DMT-count. In the 6th cycle of
the hexamer assembly, the glycerol phosphoramidite 1a was
replaced by benzyloxycarbonyl-aminohexanol phosphoramidite
3 (2 x 5 equiv.) to terminate the sequence.y Cleavage of
the hexamer from the CPG using aqueous ammonia, and
concomitant cyanoethyl removal liberated the partially protected
TA-hexamer 10. We examined two different purification
protocols for the purification of the partially protected
oligomers, RP-HPLC purification and anion exchange
chromatography followed by desalination. In the case of
hexamer 10, the former protocol provided the desired oligomer
in the highest yield (18% overall from aminopropyl CPG). At
the onset of our studies we realized that the phosphodiesters in
our synthetic targets could be potentially labile to base treatment,
in analogy to the lability of RNA fragments.9 To investigate
the base lability of the partially protected TA oligomer, we
subjected hexamer 10 to a treatment with 25% aqueous
ammonia for a prolonged period of time at room temperature
and at 40 1C. No degradation of the hexamer could be
detected by LCMS analysis after 24 h at 40 1C indicating that
transesterification of the potentially labile phosphodiesters is
not a significant risk under the reaction conditions used. Using
the conditions described above, the following TA-oligomers
were assembled: unsubstituted 10-mer 11, unsubstituted 14-mer
12, unsubstituted 20-mer 13 and monoglucosyl substituted
6-mers 14 and 15. TA-fragment 15 was assembled on the
CPG resin functionalized with C-2 glycosyl glycerol succinate
2b (loading: 100 mmol gꢀ1). The results of the syntheses are
summarized in Table 1, from which it becomes clear that the
syntheses of the longer and substituted fragments proceeded
with even greater efficiency than the assembly of hexamer 10.
With increasing size of the oligomers, purification by ion
exchange chromatography became more efficient than RP-HPLC
purification. For TA 20-mer 13 and the glucosyl substituted
hexamers 14 and 15 LCMS analysis of the crude reaction
products showed significant peak broadening and therefore
these compounds were solely purified using the ion-exchange
protocol. From the excellent yields of glucosyl substituted
hexamers 14 and 15 it becomes clear that the glycerol C-2
glucose substitution had no adverse effect on the coupling
Scheme 1 Building blocks for the automated solid phase TA
synthesis.
accomplished by employing benzylidene S-phenyl glucosyl
donor 4 in a diphenylsulfoxide-triflic anhydride mediated
glycosylation of glycerol alcohol 5.10,11 The resulting glucosyl-
glycerol 6 was transformed into 2a/b as follows: isomerization of
the allyl ether and subsequent cleavage of the resulting enol
ether using iodine under basic conditions liberated the primary
alcohol (7), which was capped with a DMT group to give
compound 8. Next the silyl group was removed to provide
alcohol 9, which was either phosphitylated to provide the
phosphoramidite building block 2a, or treated with succinic
anhydride to give the glucosylglycerol linker 2b.
With the required building blocks in hand we set out to
explore the automated solid phase TA oligomer synthesis
(Scheme 2). As a first objective we attempted the assembly
of a spacer containing TA-hexamer 10. To this end aminopropyl
CPG was functionalized with succinyl glycerol 1b. Cleavage of
the DMT group with 3% dichloroacetic acid (DCA) in toluene
and concomitant determination of the loading led to a resin
with a loading of 100 mmol gꢀ1. For the coupling step we
treated the resin with 5 equiv. of phosphoramidite 1a and
Table 1 Results from the automated solid phase synthesis and
deprotection
Yield
Yield
Entry Compound (RP-HPLC) (ion exchange) (yield)
Deprotection
1
2
3
4
5
6
10
11
12
13
14
15
18%
29%
8%
10%
16%
21%
24%
32%
36%
16 (65%)
17 (78%)
18 (84%)
19 (95%)
20 (68%)
21 (86%)
Scheme 2 Automated solid phase TA synthesis.
8962 Chem. Commun., 2011, 47, 8961–8963
c
This journal is The Royal Society of Chemistry 2011

View Article Online
glycerolphosphate building blocks could also be used in the
synthesis to allow the assembly of substituted TA fragments.
Employing the solid phase methodology, a small library of TA
fragments was generated which was tested for activity in
an opsonophagocytic inhibition assay, revealing a clear
TA-length–activity relationship. The assay also revealed
glucosyl substituted TA-hexamers 20 and 21 as promising lead
candidates for future vaccine development. We are currently
exploring the full scope of the method in the generation of a
larger library of TAs, bearing diverse substitution patterns to
unravel more detailed structure–activity relationships.
Notes and references
Fig. 1 Opsonophagocytic inhibition assay. a-LTA 12030 represents
the killing of the sera at 1 : 200 dilution. LTA is used as a positive
y The use of the commercially available N-MMT-6-aminohexanol
phosphoramidite led to the formation of N-cyanoethyl side products
in the cleavage/deprotection step. Although this could be prevented by
the use of 1,4-dithiothreitol (DTT) in the cleavage cocktail, we prefer
the use of the stable benzyl carbamate 3.
control and is used at a concentration of 100 mg mLꢀ1
.
efficiency of the building blocks used. Notably, the benzylidene
functionality on the glucosyl moiety in 15 proved to be stable
to the repeated detritylation steps. Global deprotection of the
partially protected TA-fragments was achieved by hydrogenolysis
of the benzyl ethers, CBz-group and benzylidene functionality
over palladium black followed by desalination (Scheme 2),
uneventfully leading to TA-target structures 16–21, as
summarized in Table 1. The structure of the final products
was confirmed by NMR spectroscopy and high resolution
mass spectrometry. In the NMR spectra the relative ratio of
the integrals of the peaks belonging to the spacer CH2 groups,
the spacer CH2–N group and the terminal glycerol CH2–O
moiety with respect to the bulk signal originating from all of
the other glycerol protons provided proof for the integrity of
the structures.
1 (a) W. Fischer, Adv. Microb. Physiol., 1988, 29, 233;
(b) F. C. Neuhaus and J. Baddiley, Microbiol. Mol. Biol. Rev.,
2003, 67, 686; (c) C. Weidenmaier and A. Peschel, Nat. Rev.
Microbiol., 2008, 6, 276; (d) I. B. Naumova, A. S. Shashkov,
E. M. Tul’skaya, G. M. Streshinskaya, Y. I. Kozlova,
N. V. Potekhina, L. I. Evtusshenko and E. Stackebrandt, FEMS
Microbiol. Rev., 2001, 25, 269.
2 (a) S. Morath, A. Geyer and T. Hartung, J. Exp. Med., 2001,
193, 393; (b) S. Morath, A. Geyer, I. Spreitzer, C. Hermann and
T. Hartung, Infect. Immun., 2002, 70, 938; (c) S. Morath, S. von
Aulock and T. Hartung, J. Endotoxin Res., 2005, 11, 348.
3 (a) S. von Aulock, T. Hartung and C. Hermann, J. Immunol., 2007,
178, 2610; (b) M. Hashimoto, M. Furuyashiki and Y. Suda,
J. Immunol., 2007, 178, 2610; (c) M. Hashimoto,
K. Tawaratsumida, H. Kariya, A. Kiyohara, Y. Suda, F. Krikae,
T. Kirikae and F. Gotz, J. Immunol., 2006, 177, 3162.
4 See for example: (a) A. Stadelmaier, S. Morath, T. Hartung and
R. R. Schmidt, Angew. Chem., Int. Ed., 2003, 42, 916;
(b) I. Figueroa-Perez, A. Stadelmaier, S. Deininger, S. von Aulock,
T. Hartung and R. R. Schmidt, Carbohydr. Res., 2006, 341, 2901;
(c) I. Figueroa-Perez, A. Stadelmaier, S. Morath, T. Hartung and
R. R. Schmidt, Tetrahedron: Asymmetry, 2005, 16, 493.
5 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.
6 I. G. Sava, E. Heikens and J. Huebner, Clin. Microbiol. Infect.,
2010, 16, 533.
7 (a) S. Koch, M. Hufnagel and J. Huebner, Expert Opin. Biol. Ther.,
2004, 4, 1519; (b) C. Theilacker, W. A. Krueger, A. Kropec and
J. Huebner, Vaccine, 2004, 22, S31.
8 S. L. Beaucage and M. H. Caruthers, Tetrahedron Lett., 1981,
22, 1859.
9 (a) C. B. Reese, Org. Biomol. Chem., 2005, 3, 3851;
(b) S. L. Beaucage, Curr. Opin. Drug Discovery Dev., 2008,
11, 203; (c) S. L. Beaucage and R. P. Iyer, Tetrahedron, 1993,
49, 6123.
10 (a) D. Crich and W. Cai, J. Org. Chem., 1999, 64, 4976;
(b) D. Crich and O. Vinogradova, J. Org. Chem., 2006, 71, 8473;
(c) D. Crich, Acc. Chem. Res., 2010, 43, 1144.
The prepared TA fragments were tested in an opsono-
phagocytic inhibition assay to establish their antigenic activity.13
In this assay, rabbit sera raised against purified E. faecalis
LTA are used to kill E. faecalis bacteria. Blocking the opsonic
antibodies in the sera with an inhibitor will lead to reduced
killing.z The inhibitory potential of the synthesized TA fragments
at a concentration of 100 mg mLꢀ1 is displayed in Fig. 1, which
shows that the smallest fragment tested, hexamer 16, is capable
of inhibiting the opsonophagocytic killing. For the 10-, 14-
and 20-mer a length-dependence is observed for the inhibitory
potential. The longer the fragments are the better the binding
is to the opsonic antibodies, resulting in reduced killing.
Interestingly, the glucose substituted TA fragments 20 and
21 were found to be very potent inhibitors. This is striking
since the a-glucosyl substituent is found in S. aureus TA, but
has not been found in E. faecalis TA. Nonetheless, compound
21 proved to be the most active compound of the series,
making this a promising candidate for the future development
of a vaccine comprising 21 or a close analogue as a synthetic
TA-antigen.
11 (a) J. D. C. Code
H. S. Overkleeft, J. H. van Boom and G. A. van der Marel, Org.
Lett., 2003, 5, 1947; (b) J. D. C. Codee, T. J. Boltje and G. A. van
´
e, L. J. van den Bos, R. E. J. N. Litjens,
´
der Marel, Carbohydrate Chemistry: Proven Methods, 2011, vol. 1,
ch 6, p. 67.
We have described the development of automated solid phase
methodology to synthesize glycerol phosphate teichoic acid
fragments. Tailor made glycerol phosphoramidite building
blocks were used in combination with a commercially available
synthesizer, to produce partially protected TA fragments.
With a full coupling cycle, taking approximately 15 minutes,
a TA 20-mer was produced in 5 hours. Functionalized
12 R. Weltz and S. Muller, Tetrahedron Lett., 2002, 43, 795.
13 (a) C. Theilacker, Z. Kaczynski, A. Kropec, F. Fabretti, T. Sange,
O. Holst and J. Huebner, Infect. Immun., 2006, 74, 5703;
(b) C. Theilacker, A. Kropec, O. Holst and J. Huebner, Int. J.
Med. Microbiol., 2006, 296, 97; (c) I. Toma, C. Theilacker, I. Sava,
A. Kropec, F. Hammer and J. Huebner, Int. J. Antimicrob. Agents,
2007, 29, S512.
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 8961–8963 8963