Discovery of Oligosaccharide Antigens for Semi-Synthetic Glycoconjugate Vaccine Leads against Streptococcus suis Serotypes 2, 3, 9 and 14**

Article abstract of DOI:10.1002/anie.202103990

Streptococcus suis bacteria are one of the most serious health problems for pigs and an emerging zoonotic agent in humans working in the swine industry. S. suis bacteria express capsular polysaccharides (CPS) a major bacterial virulence factor that define the serotypes. Oligosaccharides resembling the CPS of S. suis serotypes 2, 3, 9, and 14 have been synthesized, glycans related to serotypes 2 and 9 were placed on glycan array surfaces to screen blood from infected pigs. Lead antigens for the development of semi-synthetic S. suis serotypes 2 and 9 glycoconjugate veterinary vaccines were identified in this way.

Full text of DOI:10.1002/anie.202103990

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Discovery of oligosaccharide antigens for semi-synthetic
glycoconjugate vaccine leads against Streptococcus suis
serotypes 2, 3, 9 and 14
Authors: Shuo Zhang, Mauro Sella, Julinton Sianturi, Patricia Priegue,
Dacheng Shen, and Peter H. Seeberger
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202103990
Link to VoR: https://doi.org/10.1002/anie.202103990

10.1002/anie.202103990
Angewandte Chemie International Edition
Discovery of oligosaccharide antigens for semi-synthetic glycoconjugate vaccine leads
against Streptococcus suis serotypes 2, 3, 9 and 14
Shuo Zhang[a][b]§, Mauro Sella[a]§, Julinton Sianturi[a]§, Patricia Priegue[a][b], Dacheng Shen[a][c]
and Peter H. Seeberger[a][b]
*
[a] Department of Biomolecular Systems, Max Planck Institute of Colloids and Interfaces,
Am Mühlenberg 1, 14476 Potsdam, Germany
[b] Institute of Chemistry and Biochemistry, Freie Universität Berlin, Arnimallee 22, 14195
Berlin, Germany
[c]
Present address: Department of Chemistry and Chemical Biology, Harvard University, 12
Oxford Street, Cambridge, MA 02138, USA
§ These authors contributed equally
* Corresponding author
Abstract
Streptococcus suis bacteria are one of the most serious health problems for pigs and an
emerging zoonotic agent in humans working in the swine industry. S. suis bacteria express
capsular polysaccharides (CPS) a major bacterial virulence factor that define the serotypes.
Oligosaccharides resembling the CPS of S. suis serotypes 2, 3, 9, and 14 have been synthesized,
glycans related to serotypes 2 and 9 were placed on glycan array surfaces to screen blood from
infected pigs. Lead antigens for the development of semi-synthetic S. suis serotype 2 and 9
glycoconjugate veterinary vaccines were identified in this way.
Introduction
Streptococcus suis causes bacterial infections in farm pigs globally,[1] but is also a commensal
bacterium that commonly inhabits the upper respiratory, digestive and reproductive systems of
pigs.[2,3] Virulent strains can infect the bloodstream and eventually result in septic shock and
meningitis in pigs, but can also cause septicemia and meningitis in humans.
S. suis is surrounded by a layer of polysaccharides forming the bacterial capsules that play a
fundamental role for pathogen survival,[4] protect the bacterium and are important virulence
factors. CPSs are able to trigger an adaptive immune response resulting in the production of
specific antibodies rendering polysaccharides attractive targets for antibacterial vaccine
development.[5] S. suis serotypes are distinguished based on the chemical composition of the
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
capsules.[6,7] Serotypes 1/2, 2, 3, 7, 9 are most frequently isolated from infected animals and
differ in geographical prevalence. Serotype 2 is particularly frequent in Europe and Asia, while
serotypes 3 and 9 were mostly found in North America.
Vetenary vaccines are an effective strategy to limit disease in farm animals and reduce the
spread of pathogens between animals and transmission to humans. Vaccinations help to reduce
antibiotic consumption and slow the development of antimicrobial resistance.[8] All currently
used antibacterial vetenary vaccines are prepared from live attenuated or inactivated bacteria
that suffer shortcomings in terms of safety, stability and in some cases limited
immunogenicity.[8–10] While glycoconjugate vaccines in humans are very successful, veterinary
glycoconjugate vaccines remain a largely unexplored opportunity.[11]
The CPS structures of four major S. suis serotypes (2, 3, 9, and 14) have been elucidated (Figure
1).[12–15] The CPSs include rare sugars, a variety of glycosidic linkages, anionic charges and
modifications such as acetyl esters and phosphodiesters. Serotypes 1, 2, and 14 are structurally
very similar.
Figure 1. Structures of the most common S. suis CPS repeating units. Highlighted structures were the focus of
the studies disclosed here.
S. suis vaccine development using isolated polysaccharides has focused exclusively on serotype
2.[16] To date, no single antigen has been shown to be more efficacious than a suspension of
killed or weakened bacteria,[16–18] or to be protective against S. suis serotype 9. Exact
carbohydrate epitopes responsible for inducing protective antibodies are still unknown but are
the basis for establishing structure-immunogenicity relationships for the design of carbohydrate
antigens for vaccine studies. Investigations with isolated native CPSs produced inconclusive
results.[19–21] Synthetic oligosaccharides related to CPS can help to determine antibody epitopes.
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Here, we describe the first synthesis of well-defined oligosaccharides resembling the S. suis
serotypes 2, 3, 9 and 14 CPSs. Glycans related to serotypes 2 and 9 were employed on the
surface of glycan arrays to identify lead structures for the development of semi-synthetic
glycoconjugate vaccines against S. suis.
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Synthesis of oligosaccharides related to S. suis serotype 2 CPS
The S. suis serotype 2 CPS[14] consists of a branched heptasaccharide repeating unit ([→4)[-
Neu5Ac(2→6)--D-Gal(1→4)--D-GlcNAc(1→3)]--D-Gal(1→4)-[-D-Gal(1→3)]--L-
Rha(1→4)--D-Glc(1→]) (Figure 2). The CPS from S. suis serotype 2 is essential for its
virulence as it prevents phagocytosis when the bacterium infiltrates the bloodstream.[22]
The CPS is the most promising antigen[23,24] as it can induce protective IgM antibodies[25,26]
despite being poorly immunogenic ─ low levels of anti-CPS antibodies were seen in pigs after
infection[18] or immunization.[23] An anti-serotype 2 glycoconjugate vaccine made from
capsular polysaccharides isolated from fermented bacteria was evaluated in immunization
experiments in animal models.[16] The poor immunogenicity of CPS can be overcome and
protection against S. suis can be achieved by active immunization with a glycoconjugate.
Five oligosaccharides (1-5) resembling the repeating unit of S. suis serotype 2 CPS were
designed, to obtain detailed structural information of antigenic epitopes of antibodies from S.
suis-infected pigs (Figure 2). Three shorter fragments were included: trisaccharide 1 resembles
the backbone while 2 and 3 represent the side-chain. Pentasaccharide 4 and hexasaccharide 5
were synthesized to cover almost the entire length of a repeating unit as well as branched
sequences, and to understand whether the terminal N-acetyl neuraminic acid is directly engaged
in antibody binding. All synthetic oligosaccharides contain an aminopentyl spacer at the
reducing end sugar for creating microarrays and protein conjugates.
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Figure 2. Structure of S. suis serotype 2 CPS and related oligosaccharides to be prepared from building blocks
6–12.
Seven orthogonally protected monosaccharide building blocks 6–12 were identified to create
1,2-cis glycosidic bonds, a branching point on L-rhamnose and the -sialyl linkage (Figure 2).
The synthesis of 1 via a linear approach used three monosaccharide building blocks (6-8)
(Scheme 1) and started with the introduction of the spacer at the reducing end monosaccharide
with a glycosylation between N-protected aminopentanol 13 and glucose thioglycoside 6,
followed by cleavage of the fluorenylmethoxycarbonyl (Fmoc) protective group to obtain 14.
The reducing end glucose was then glycosylated with rhamnose thioglycoside 7.[27] To assist
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
the formation of the -rhamnosidic linkage and prepare for the subsequent introduction of the
-galactose, the C-3 hydroxyl group is masked by a 2-pyridinecarbonyl ester (picoloyl ester –
Pico) that ensures a H-bond mediated stereodirecting effect and is orthogonal to the benzyl
ethers. The non-reducing end galactose was introduced with  configuration using known
galactosyl thioglycoside 8,[28] equipped with C-4 and C-6 acetyl esters to assist in the formation
of the 1,2-cis glycosidic bond. The glycosylation of disaccharide 15 with thioglycoside 8 was
carried out in a DCM/Et2O mixture to increase the -selectivity through solvent effects. Only
the -linked product 16 was detected on TLC and isolated. Finally, protected trisaccharide 16
was fully deprotected by ester hydrolysis using sodium methoxide in methanol followed by
catalytic hydrogenation, obtaining trisaccharide 1.
The linear synthesis of trisaccharide 2 used commercially available galactose building block 9
as starting point for both monosaccharides 17 and 20 (Supporting information). Spacer 13 was
regioselectively glycosylated with 17, before the resulting galactose was glycosylated with
commercially available glucosamine 10, furnishing disaccharide 19a after Fmoc removal
(Scheme 1). The disaccharide was finally glycosylated with galactose 20 to obtain fully
protected trisaccharide 21. Deacylation with sodium methoxide in methanol at 35 °C was
accompanied by partial hydrolysis of the trichloroacetamide due to the large excess of base
such that N-acetylation became necessary. Finally, catalytic hydrogenation removed all ethers
and afforded deprotected trisaccharide 2.
A convergent 2+2 glycosylation strategy was followed to assemble tetrasaccharide 3, by first
coupling acceptor 19a and disaccharide 23 that contains a preinstalled  sialyl glycosydic bond.
Disaccharide 23 was obtained by glycosylating galactose acceptor 22 with known sialyl
glycosyl imidate 12[29] at 60 °C in a DCM/CH3CN mixture. Isolation of the pure
diastereoisomers by careful silica column chromatography gave pure α-sialylated galactoside
23 and the corresponding isomer in a 6/1  ratio. The configuration was unequivocally
determined by measuring the long-range JC-1,H-3ax (Supporting information).
Disaccharide 23 was employed to glycosylate acceptor 19a, promoted by NIS and triflic acid.
These conditions proved ineffective for tetrasaccharide formation. To improve the reactivity of
the glycosylating agent, the thioglycoside was converted to the more reactive glycosyl
phosphate 24. The glycosylation of acceptor 19a using disaccharide 24 proceeded poorly as no
product was isolated from a complex mixture. Likely, insufficient acceptor nucleophilicity was
responsible for the failure of this reaction as both glycosylating agents 23 and 24 hydrolysed
but acceptor 22 was recovered. The C-4 hydroxyl group in ester protected glucosamine
acceptors is a poor nucleophile[30] and variations in the protecting group pattern can lead to
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
improved couplings. When disaccharide 19b, containing an ether group instead of an ester at
C-3 of the glucosamine unit was glycosylated using disaccharide phosphate 24, the desired
tetrasaccharide 25 was obtained in 67% yield. Removal of all protecting groups by ester
hydrolysis under basic conditions followed by catalytic hydrogenation gave unprotected
tetrasaccharide 3.
Scheme 1. Synthesis of linear oligosaccharides 1, 2 and 3
Oligosaccharides 4 and 5 were assembled using convergent syntheses from common
trisaccharide acceptor 29 (Scheme 2). Trisaccharide 29 contains a challenging -rhamnosidic
linkage on a doubly substituted terminal rhamnose residue. Building block 7 was not suitable
for the synthesis and fully orthogonal rhamnose building block 26 was prepared instead in seven
steps from commercially available rhamnose (Supporting information, Scheme S2). A
glycosylation of the aminopentanol linker using rhamnose tioglycoside 26 and NIS/TfOH as
activator resulted in low stereoselectivity (2.4/1, / as judged by NMR. Instead, bromine
activation[31] of thioglycoside 26 afforded spacer-linked rhamnose in a slow reaction with
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
higher stereoselectivity (10/1, ). Rhamnose 27 was then glycosylated with galactose 8 in a
DCM/Et2O mixture and no appreciable amounts of -linked galactose were isolated. To
perform a second glycosylation and install a second galactose residue, the silyl ether was
removed by TBAF to afford disaccharide acceptor 28 that was glycosylated with 9 to obtain
desired trisaccharide 29 after Fmoc removal. The C-4 hydroxyl of rhamnose acted as a good
nucleophile, despite its proximity to the  galactose unit.
With trisaccharide acceptor 29 in hand, the assembly of 4 continued with the preparation of
disaccharide imidate 30 (Scheme 2 and Supporting information, Scheme S3). Disaccharide 30
and acceptor 29 were coupled to obtain the protected pentasaccharide in 67% yield. Removal
of all protective groups yielded pure pentasaccharide 4.
Sialylated hexasaccharide 5 required the preparation of trisaccharide glycosyl imidate 32
(Scheme 2 and Supporting information, Scheme S4). A 3+3 glycosylation of trisaccharide
imidate 32 and trisaccharide 29 did not proceed below 20 °C as 32 degraded. Increasing
amounts of acid activator (up to 0.5 eq) and higher temperatures still resulted in highly complex
mixtures as only 11% of hexasaccharide 31 was obtained. An identical acceptor was
successfully employed in the synthesis of oligosaccharide 4. Therefore, likely trisaccharide
glycosylating agent 32 is responsible for the low yields although the reasons remained unclear.
Subsequently, ester cleavage and catalytic hydrogenation produced hexasaccharide 5.
Scheme 2. Synthesis of branched oligosaccharides 4 and 5
8
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
S. suis serotype 2 pentasaccharide is a potential vaccine candidate
Glycan arrays enable the screening of multiple serum samples to identify protective
glycotopes.[32] Synthetic oligosaccharides 1-5, the isolated native CPS, and other structurally
related glycans were immobilized on glass slides to identify antibody binding patterns in serum
samples from infected pigs and from rabbits immunized with native CPS (Figure 3). In addition,
human reference serum 007sp was screened as control.[33]
IgG antibodies from pig sera bound specifically to oligosaccharide 1 and, to a lesser extent, to
4. A five-fold stronger binding to oligosaccharides 4 and 5 was observed with anti-serotype 2
rabbit serum (Figure 3). Co-infections with other S. suis serotypes or bacteria in pigs may be
responsible for these differences as higher antibody levels are elicited following immunizations
compared to a natural infection.[34]
Trisaccharide 1 was recognized not only by pig sera, but also by S. pneumoniae-vaccinated
human serum (007sp). This finding is likely the result of immune cross-reaction due to
structural similarities in the CPS from different streptococcal species.[35,36]
Pentasaccharide 4, a structure that covers most of the native repeating unit, appears to be the
minimum glycotope useful to elicit an immune response, as both pig and rabbit antibodies
recognized it strongly. It is worth noting that almost identical fluorescence signals were
measured for compounds 4 and 5. The terminal sialic acid unit on hexasaccharide 5 neither
increased nor impaired binding to these antibodies. Therefore, sialic acid is not a fundamental
part of the minimal epitope of rabbit antibodies.
Additionally, class switch from IgM to IgG antibodies occurred as IgG titres were found to be
much higher than the IgM response (Supporting information, Figure S2), likely as result of B
cell differentiation to eliminate the pathogen following immunization or infection.[16]
Immunization resulted in higher antibody titres than natural infections,[34] suggesting that
vaccination with glycoconjugates is a promising means to elicit an even stronger immune
response. In conclusion, pentasaccharide 4 is an attractive lead for the development of a
glycoconjugate vaccine to protect from S. suis serotype 2.
9
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Figure 3. Glycan array analysis of S. suis serotype 2 oligosaccharides and native CPS. A) Printing pattern of
microarray and binding pattern of rabbit serum to immobilized glycans. B) IgG antibody binding to glycans. A
serum dilution of 1:100 was used. MFI, mean fluorescence in intensity (mean ± standard deviation); PB, printing
buffer; CWPS, cell wall polysaccharide; CPS and synthetic glycans, see Figure S1.
Synthesis of oligosaccharides related to S. suis serotype 3 CPS
The S. suis serotype 3 CPS disaccharide repeating unit [→4)--D-GlcpA-(1→3)--D-
QuipNAc4NAc-(1→] contains the rare diamino sugar di-N-acetyl-D-bacillosamine
(QuipNAc4NAc) and D-glucuronic acid (GlcA) (Figure 4).[12] QuipNAc4NAc is found in
various bacterial capsular polysaccharides, including N-linked glycoproteins on Campylobacter
jejuni, O-linked glycoproteins on Neisseria gonorrhoeae and O-antigens from many strains of
10
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Gram-negative bacteria.[37–39] The S. suis disaccharide repeating unit is unique and we report
the first synthesis of a series of oligosaccharides related to the native S. suis serotype 3 CPS as
basis for immunological studies.
Figure 4. Structure of S. suis serotype 3 CPS and synthetic oligosaccharide antigens resembling the S. suis serotype
3 CPS derived from building blocks 13, 41, and 42.
The main synthetic challenges are access to the rare diamino bacillosamine building block 42
and complex functional interconversions to be performed at the tri- and tetrasaccharide level.
Formation of glycosidic bonds to construct the oligosaccharides was challenging as several
electron-withdrawing groups reduce the nucleophilicity of glycosyl acceptors.[40,41] Since
electron-withdrawing carboxyl groups at C-6 may further reduce the reactivity of the glucose
donors,[42] they were introduced after the glycosylation. Finally, the densely nitrogen-
functionalized target molecules necessitated additional deprotection steps. Stereoselective -
glycosidic bond formations were ensured by neighboring participation with benzoyl groups
(Bz) on the glucose and a trichoroacetamide (TCA) group on bacillosamine.
Synthesis of the bacillosamine-containing reducing end commenced with the stereoselective
glycosylation between selenoglycoside 42 (Supporting information, Scheme S5) and protected
amino linker 13 using TMSOTf and NIS as activators to yield exclusively the -linked product
43 in 91% yield. Conversion of the azide group to the acetamide and contemporary reduction
of the TCA group was carried out using zinc powder in a THF/Ac2O/AcOH mixture. Removal
of the remaining benzyl ethers and the benzyloxycarbamate group was achieved by
11
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
hydrogenolysis using Pd/C in EtOAc/t-BuOH/H2O and gave linker-equipped bacillosamine 40
in 49% yield over two steps.
Cleavage of the 2-naphthylmethyl (Nap) protecting group on 43 with DDQ afforded alcohol 44
in 97% yield and subsequent uion of 44 and 41[43] yielded 67% -linked disaccharide 45. Even
when excess donor 41 (2.0 eq) and prolonged reaction times were employed, unreacted acceptor
44 was always observed. Acid hydrolysis of the benzylidene acetal on 45, followed by selective
oxidation of the C-6 hydroxyl group with TEMPO/BAIB and protection of the carboxyl moiety
as benzyl ester gave disaccharide 47 in 77% yield over two steps. The C-4 hydroxyl was
glycosylated with another bacillosamine unit to furnish -linked trisaccharide 48 in 65% yield.
Azide and TCA groups present in 47 and 48 were converted into acetamides by employing the
same conditions used for compound 43. Subsequent hydrolysis under basic conditions and
hydrogenolysis smoothly afforded oligosaccharides 38 and 36 in acceptable yield.
Linker-equipped glucuronic acid was prepared by first coupling protected glucose building
block 41 and 13 under NIS/TMSOTf promotion to furnish -linked product 49 in 71% yield.
The benzylidene acetal on 49 was then hydrolyzed, before regioselective oxidation with
TEMPO/BAIB and esterification furnished glucuronic acid benzyl ester 51 in 72% yield over
two steps. Ester groups on monosaccharide 51 were hydrolyzed under basic conditions, and
hydrogenolysis afforded deprotected 39 in 70% yield over two steps.
Further elaboration of 51 continued with a glycosylation using building block 42 to furnish
protected disaccharide 52 in 93% yield. The Nap ether was removed as preparation for another
glycosylation with building block 41 to yield protected trisaccharide 54. Removal of the
benzylidene acetal on 54 proved challenging. Treatment with (+)-camphor-10-sulfonic acid as
a catalyst and ethanethiol as acetal exchange reagent[44,45] gave trisaccharide diol 55 in low
yield. Instead, aqueous acetic acid (80%) at 60 °C smoothly delivered 55 in 87% yield
(Supporting information). The primary C-6 alcohol of diol 55 was then oxidized and converted
into the corresponding benzyl ester 56. Coupling of bacillosamine donor 42 with acceptor 56
produced tetrasaccharide 57 in 65% yield. Finally, oligosaccharides 52, 56 and 57 were globally
deprotected via Zn-mediated reduction, basic ester hydrolysis and hydrogenolysis, to obtain
oligosaccharides 37 (51% over three steps), 35 (50% over three steps) and 34 (40% over three
steps), respectively.
12
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Scheme 3. Synthesis of oligosaccharides 34-40 related to S. suis serotype 3 CPS
13
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Synthesis of oligosaccharides related to S. suis serotype 9 CPS
The S. suis serotype 9 CPS repeating unit consists of a branched tetrasaccharide with a
phosphorylated D-glucitol residue (Figure 5).[15] The presence of the labile C-4-keto sugar and
the phosphodiester group pose considerable synthetic challenges. Synthetic C-4-keto sugars are
difficult to handle during a total synthesis but the reduced form is more stable, while
maintaining the ability to induce an immune response against the CPS.[46] This isosteric
chemical modification facilitates subsequent scale-up. We focused on the synthesis of the C-4-
axial reduced form of S. suis serotype 9 CPS repeating unit and its analogues (Figure 5).
14
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Figure 5. Structure of S. suis serotype 9 CPS and related oligosaccharides 58-65 to be prepared from building
blocks 66-73.
Five differentially protected monosaccharide building blocks (66-70) and three H-
phosphonates (71, 72, 73) are needed to assemble the target oligosaccharides (Figure 5). Benzyl
ethers serve as permanent protecting groups while TCA and levulinoyl ester (Lev) ensure
anchimeric assistance as C-2 protective groups. The phosphorylated D-glucitol will be
introduced after assembly of the tetrasaccharide core.
Assembly of tetrasaccharide 79 commenced with the union of L-rhamnose 66 and D-fucosamine
67 in the presence of NIS/TfOH as promoter to afford disaccharide 74 in good stereoselectivity
15
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
(α/β = 1/6) (Scheme 4). Cleavage of the Nap ether with DDQ and β-pinene gave disaccharide
75 (61% yield)[47] ready for glycosylation with 68 to produce trisaccharide 76 with complete β-
selectivity. Cleavage of the levulinoyl ester with hydrazine acetate gave trisaccharide 77 in 91%
yield. Control over stereoselectivity via solvent effects yielded consistant results. The
NIS/TfOH-mediated glycosylation between thioglycoside 70 and 77 in a mixture of DCM and
Et2O produced 82% tetrasaccharide 78. Acetyl ester cleavage with sodium methoxide furnished
79. Global deprotection of oligosaccharides 74, 76 and 79 gave the corresponding non-
phosphorylated compounds 65, 62 and 59. H-phosphonates 71, 72 and 73 were coupled with
the hydroxyl group on 79 in the presence of pivaloyl chloride to give intermediate H-
phosphonate diesters.[48] Subsequent oxidation with iodine and water in the same pot furnished
the phosphates 80, 81 and 82 as triethylammonium salts in 69-86% yield over two steps. Global
deprotection of the phosphodiesters by Pd/C-catalyzed hydrogenolysis, followed by ion
exchange column chromatography furnished 58, 60 and 61 as sodium salts (Supporting
information).
Scheme 4. Synthesis of oligosaccharide fragments of S. suis serotype 9 CPS 58-65.
16
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
For the synthesis of trisaccharide phosphates, glycosylation between 75 and 69 yielded
trisaccharide 83 (73%) with complete β selectivity (Scheme 4). Removal of Nap provided 84
with a free C-3 hydroxyl group to be coupled with H-phosphonates 72 and 71. The levulinoyl
esters were cleaved to afford phosphate diesters 85 (73% yield) and 86 (66% yield). Target
phosphates 63 and 64 were obtained after global deprotection.
The antibody-binding activity of a synthetic oligosaccharide antigen depends on the structure
and configuration of the antigen.[49] In order to investigate the importance of the β-linkage in
the S. suis serotype 9 CPS, oligosaccharides 87-93 bearing an α-glycosidic linkage in place of
the native β-glycosidic bond between D-fucosamine and L-rhamnose based on the S. suis
serotype 9 CPS were prepared (Figure 6).
Figure 6. α Oligosaccharide antigens resembling S. suis serotype 9 CPS.
To install the α-linkage, building block 94 was prepared (Supporting information). Following
the assembly sequence used for the β-linked oligosaccharides, union of 66 and 94 at 0 °C
produced disaccharide 95 with high α-selectivity (Scheme 5). Conversion of the benzoyl ester
to the benzyl ether over two steps followed by the cleavage of the Nap ether afforded
disaccharide acceptor 98. Coupling of thioglycoside 68 with 98 produced exclusively β-product
99 in 87% yield. After levulinoyl ester cleavage, NIS/TfOH-mediated glycosylation of 100 with
70 produced tetrasaccharide 101. The azide moiety of 101 was reduced by zinc and the resulting
amine was acetylated to give 102 in 76% yield. Next, tetrasaccharide 103 with a free hydroxyl
was obtained after saponification of the acetyl ester and following hydrogenolysis gave 88.
Tetrasaccharide 103 was smoothly coupled with three different H-phosphonates 71, 72 and 73
17
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
to produce corresponding phosphates 104, 105 and 106. Subsequent hydrogenolysis and
sodium ion exchange chromatography produed pure 87, 89 and 90.
Scheme 5. Assembly of α oligosaccharides 87-93.
The presence of an acetyl ester in 100 rendered benzylation under basic conditions not feasible.
Treatment of 100 with freshly prepared silver oxide and benzyl bromide[50] produced 107 in
47% yield. The azide was converted to the corresponding acetamide using zinc and acetic
anhydride to yield 108 (69%). The synthesis of 91 was achieved after removal of the acetyl
group and hydrogenolysis of the benzyl ethers on 108 in 46% yield over two steps. To set the
stage for phosphorylation, deacetylation of 108 yielded the requisite free hydroxyl group that
was coupled with 72 and 71 to furnish phosphates 109 and 110 as the triethylammonium salts.
Global deprotection of the phosphates afforded the corresponding final glycosides 92 and 93.
Glycan array screening identifies a trisaccharide as S. suis serotype 9 glycoconjugate
vaccine lead
18
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Sera from pigs infected with S. suis serotype 9, from immunized rabbits and human reference
serum 007sp[51] were screened for antibodies binding to synthetic oligosaccharides and isolated
S. suis serotype 9 CPS using glycan microarrays (Figure 7). IgG antibodies present in pig sera
bound weakly to oligosaccharides revealing a complex binding pattern. Rabbit sera on the other
hand showed a clearer picture. Rabbit IgGs recognized strongly phosphorylated trisaccharides
63 and 64, disaccharide 65, and similarly the native CPS. Trisaccharide 63 includes both the
sugar sequence of 65 and a terminal phosphate monoester. This functional group appears to be
essential since the absence on phosphorylation on 62 strongly reduced binding. On the other
hand, the glucitol chain on 64 did not show a significant effect. Branched oligosaccharides 58-
61 were not bound specifically. These observations suggest that the minimal glycotope contains
L-rhamnose, D-fucosamine and a phosphate moiety, indicating that trisaccharide 63 is the
minimum glycotope useful to elicit an immune response.
Antibodies from 007sp serum bound to the longer oligosaccharides due to cross-reactivity with
CPS S. pneumoniae, possibly because of the branched α-galactose.[35,52] It had been shown
previously, that other S. suis serotypes cross-react with some serotypes of S. pneumoniae.[35,36]
The IgG response was higher than that of IgM, as IgM antibodies showed low or no binding
for most cases (Supporting information, Figure S4) indicating that isotype switching after
infection in pigs or immunization in rabbits and humans was induced.
Oligosaccharides with an α- in place of a β-linkage between D-fucosamine and L-rhamnose (87-
93) were found to be bound much weaker (Supporting information, Figure S5). This finding
suggests that the β-linkage, present in the native CPS,[15] has an important role for the
recognition by antibodies.[49]
In conclusion, trisaccharide 63 is an attractive lead for the development of a glycoconjugate
vaccine against S. suis serotype 9.
19
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Figure 7. Glycan array analysis of S. suis serotype 9 oligosaccharides and native CPS. A) Printing pattern of
microarray and binding of rabbit serum to immobilized glycans B) IgG antibody binding to glycans. A serum
dilution of 1:100 was used. MFI, mean fluorescence intensity (mean ± standard deviation); PB, printing buffer;
CWPS, cell wall polysaccharide; CPS and synthetic glycans, see Figure S3.
20
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Synthesis of oligosaccharides related to S. suis serotype 14 CPS
S. suis serotype 14 is responsible for pig and human infections mainly in Asia[6,53] and has been
less studied than the more prevalent serotypes. Expression of the CPS is fundamental to inhibit
phagocytosis in vitro and non-encapsulated bacteria are significantly less virulent in mouse
models.[54] The serotype 14 CPS[13] (Figure 8) consists of a hexasaccharide repeating unit
([→6)[-Neu5Ac(2→6)--D-Gal(1→4)--D-GlcNAc(1→3)]--D-Gal(1→3)--D-Gal(1→4)-
-D-Glc(1→]) composed of a trisaccharide backbone and a sialylated lactosamine side chain.
Compared to the S. suis serotype 2 the -rhamnose in the backbone is missing and the linkage
between the glucose and galactose units is 1→6), instead of 1→4).
Recent studies aimed at elucidating glycotopes responsible for the production of protective
antibodies[19] but more detailed information on the structure of carbohydrate epitopes is needed.
Moreover, the antigenic properties of the CPS, either alone or as part of a glycoconjugate, have
not been evaluated.
Several substructures related to the repeating unit of serotype 14 CPS were designed (Figure
8), including three oligosaccharides carrying an aminopentyl spacer at the reducing end. To
identify whether antibody binding involves mostly the backbone residues,[19] hexasaccharide
111 was synthesized. Pentasaccharide 112 was prepared to evaluate whether antibody epitopes
include the entire repeating unit. Hexasaccharide 113, representing its sialylated analogue, was
included to address the role of sialylation in S. suis serotype 14 CPS.
21
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Figure 8. Structure of S. suis serotype 14 CPS and retrosynthetic analysis of related oligosaccharides. (CMP=
cytidine monophosphate)
Automated glycan assembly (AGA) was used to assemble all oligosaccharides from four
building blocks and Merrifield resin functionalized with a photolabile linker (Scheme 6).
Considering the challenges encountered during sialylation of S. suis serotype 2
oligosaccharides, a chemoenzymatic approach was adopted to obtain hexasaccharide 113, via
one single step from pentasaccharide 112 using a sialyltransferase.
Hexasaccharide 111 consists of two repetitions of a trisaccharide. Glycosyl phosphates 114,
115 and 116 were employed in AGA to obtain protected hexasaccharide 120, using
glycosylation conditions previously optimized for glycosyl phosphates. The target compound
was obtained after resin cleavage and HPLC purification in 20% overall yield. Deprotection by
hydrogenolysis and basic ester hydrolysis produced compound 111 in 52% yield over two steps.
Pentasaccharide 112 contains the trisaccharide repeating unit in 111, plus a glucosamine and a
galactose. Just three building blocks are needed: one glucose, one galactose and galactosamine
117. Employing the conditions used previously to synthesize the hexasaccharide, AGA
proceeded smoothly and protected pentasaccharide 121 was obtained after resin cleavage and
HPLC purification in 56% overall yield. Deprotection was carried out by hydrogenolysis
followed by deacylation with sodium methoxide, to afford fully deprotected 112 in 53% yield.
Enzymatic sialylation of pentasaccharide 112 using a sialyltransferase from the marine
bacterium Photobacterium damsela[55] (Pd2,6ST) produced the 2→6) linkage in
hexasaccharide 113. To limit double sialylation on the oligosaccharide chain observed during
22
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
intial experiments, optimal reaction conditions (1.5 eq of CMP-Neu5Ac and 7 h reaction time)
were used to produce sialylated hexasaccharide 113 in 42% yield after purification.
Scheme 6. Synthesis of S. suis serotype 14 related oligosaccharides. Reagents and conditions: a) Building block
(5 eq), TMSOTf, ̶ 30 °C (5 min)→ ̶ 10 °C (40 min); b) Ac2O, MsOH, DCM ; c) Piperidine, DMF; d) hν; overall
yield (based on resin loading): 20% for 120; 56% for 121; e) H2, Pd/C, THF/MeOH/AcOH; f) NaOMe, MeOH;
40 °C; 52% over two steps for 111; 53% over two steps for 112.
23
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Conclusions
We describe the synthesis of a collection of 30 novel oligosaccharides resembling the capsular
polysaccharides related to four major serotypes (2, 3, 9, 14) of the bacterium Streptococcus
suis. The syntheses tackled challenges associated with complex glycan targets, such as
sialylation, the introduction of amino-rich sugars and labile phosphodiesters. The synthetic,
conjugation-ready glycans were printed onto array surface to give rise to glycan microarrays.
The glycan microarrays were used to screen the sera of pigs infected with different S. suis
serotypes as well as sera from immunized rabbits. With the help of glycan array studies, the
glycan epitopes of lead antigens for the development of semi-synthetic glycoconjugate vaccines
to protect from S. suis serotypes 2 and 9 were identified. Currently, vaccination of pigs for
challenge studies is being prepared in an effort to develop efficacious vaccines to protect pigs
as well as people working in the swine industry while reducing the use of antibiotics.
24
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
References
[1] K. VanderWaal, J. Deen, Proc. Natl. Acad. Sci. U. S. A. 2018, 115, 11495–11500.
[2] D. M. Williams, G. H. K. Lawson, A. C. Rowland, Res. Vet. Sci. 1973, 15, 352–362.
[3] K. Murase, T. Watanabe, S. Arai, H. Kim, M. Tohya, K. Ishida-Kuroki, T. H. Võ, T. P.
B. Nguyễn, I. Nakagawa, R. Osawa, N. H. Nguyễn, T. Sekizaki, PLoS One 2019, 14,
e0215983.
[4] I. S. Roberts, Annu. Rev. Microbiol. 1996, 50, 285–315.
[5] R. D. Astronomo, D. R. Burton, Nat. Rev. Drug Discov. 2010, 9, 308–324.
[6] G. Goyette-Desjardins, J.-P. Auger, J. Xu, M. Segura, M. Gottschalk, Emerg. Microbes
Infect. 2014, 3, 1–20.
[7] B. Haas, D. Grenier, Med. Mal. Infect. 2018, 48, 159–166.
[8] K. Hoelzer, L. Bielke, D. P. Blake, E. Cox, S. M. Cutting, B. Devriendt, E. Erlacher-
Vindel, E. Goossens, K. Karaca, S. Lemiere, M. Metzner, M. Raicek, M. Collell
Suriñach, N. M. Wong, C. Gay, F. Van Immerseel, Vet. Res. 2018, 49, 64.
[9] E. N. T. Meeusen, J. Walker, A. Peters, P.-P. Pastoret, G. Jungersen, Clin. Microbiol.
Rev. 2007, 20, 489–510.
[10] V. Vetter, G. Denizer, L. R. Friedland, J. Krishnan, M. Shapiro, Ann. Med. 2018, 50,
110–120.
[11] V. Gerdts, G. Mutwiri, J. Richards, S. van D. L. van den Hurk, A. A. Potter, Vaccine
2013, 31, 596–602.
[12] G. Goyette-Desjardins, E. Vinogradov, M. Okura, D. Takamatsu, M. Gottschalk, M.
Segura, Carbohydr. Res. 2018, 466, 18–29.
[13] M.-R. Van Calsteren, F. Gagnon, C. Calzas, G. Goyette-Desjardins, M. Okura, D.
Takamatsu, M. Gottschalk, M. Segura, Biochem. Cell Biol. 2012, 91, 49–58.
[14] M.-R. Van Calsteren, F. Gagnon, S. Lacouture, N. Fittipaldi, M. Gottschalk, Biochem.
Cell Biol. 2010, 88, 513–525.
[15] E. Vinogradov, G. Goyette-Desjardins, M. Okura, D. Takamatsu, M. Gottschalk, M.
Segura, Carbohydr. Res. 2016, 433, 25–30.
[16] G. Goyette-Desjardins, C. Calzas, T. C. Shiao, A. Neubauer, J. Kempker, R. Roy, M.
Gottschalk, M. Segura, Infect. Immun. 2016, 84, 2059–2075.
[17] C. G. Baums, C. Kock, A. Beineke, K. Bennecke, R. Goethe, C. Schröder, K.-H.
Waldmann, P. Valentin-Weigand, mSphere 2009, 16, 200–208.
[18] C. Calzas, P. Lemire, G. Auray, V. Gerdts, M. Gottschalk, M. Segura, Infect. Immun.
2015, 83, 441 LP – 453.
[19] M.-R. Van Calsteren, G. Goyette-Desjardins, F. Gagnon, M. Okura, D. Takamatsu, R.
Roy, M. Gottschalk, M. Segura, J. Biol. Chem. 2016, 291, 8387–8398.
[20] M. P. Lecours, N. Fittipaldi, D. Takamatsu, M. Okura, M. Segura, G. Goyette-
Desjardins, M. R. Van Calsteren, M. Gottschalk, Microbes Infect. 2012, 14, 941–950.
[21] G. Goyette-Desjardins, S. Lacouture, J. P. Auger, R. Roy, M. Gottschalk, M. Segura,
Pathogens 2019, 8, 1–20.
[22] N. Fittipaldi, M. Segura, D. Grenier, M. Gottschalk, Future Microbiol. 2012, 7, 259–
279.
[23] C. Calzas, M. Taillardet, I. S. Fourati, D. Roy, M. Gottschalk, H. Soudeyns, T.
Defrance, M. Segura, Pathog. (Basel, Switzerland) 2017, 6, 16.
[24] M. Segura, Expert Rev. Vaccines 2015, 14, 1587–1608.
[25] S. D. Elliott, F. Clifton-Hadley, J. Tai, J. Hyg. (Lond). 1980, 85, 275–285.
[26] N. Charland, M. Jacques, S. Lacouture, M. Gottschalk, Microbiology 1997, 143, 3607–
3614.
[27] M. Emmadi, N. Khan, L. Lykke, K. Reppe, S. G. Parameswarappa, M. P. Lisboa, S.-M.
Wienhold, M. Witzenrath, C. L. Pereira, P. H. Seeberger, J. Am. Chem. Soc. 2017, 139,
25
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
14783–14791.
[28] H. S. Hahm, M. Hurevich, P. H. Seeberger, Nat. Commun. 2016, 7, 12482.
[29] S. Cai, B. Yu, Org. Lett. 2003, 5, 3827–3830.
[30] D. Crich, V. Dudkin, J. Am. Chem. Soc. 2001, 123, 6819–6825.
[31] S. Kaeothip, J. P. Yasomanee, A. V Demchenko, J. Org. Chem. 2012, 77, 291–299.
[32] F. Broecker, P. H. Seeberger, Methods Mol. Biol. 2017, 1518, 227–240.
[33] D. Goldblatt, C. Y. Tan, P. Burbidge, S. McElhiney, L. McLaughlin, R. Tucker, M.
Rauh, M. Sidhu, P. C. Giardina, Clin. Vaccine Immunol. 2015, 22, 1154–1159.
[34] C.-A. Siegrist, in Plotkin’s Vaccines, Elsevier, 2018, pp. 16-34.e7.
[35] M. Gottschalk, J. Kolberg, N. Charland, M. Jacques, J. Clin. Microbiol. 1995, 33,
2492–2495.
[36] G. Goyette-Desjardins, E. Vinogradov, M. Okura, D. Takamatsu, M. Gottschalk, M.
Segura, Carbohydr. Res. 2019, 473, 36–45.
[37] N. M. Young, J. R. Brisson, J. Kelly, D. C. Watson, L. Tessier, P. H. Lanthier, H. C.
Jarrell, N. Cadotte, F. St. Michael, E. Aberg, C. M. Szymanski, J. Biol. Chem. 2002,
277, 42530–42539.
[38] E. STIMSON, M. VIRJI, S. BARKER, M. PANICO, I. BLENCH, J. SAUNDERS, G.
PAYNE, E. Richard MOXON, A. DELL, H. R. MORRIS, Biochem. J. 1996, 316, 29–
33.
[39] M. D. Hartley, M. J. Morrison, F. E. Aas, B. Børud, M. Koomey, B. Imperiali,
Biochemistry 2011, 50, 4936–4948.
[40] M. Heuckendorff, L. T. Poulsen, H. H. Jensen, J. Org. Chem. 2016, 81, 4988–5006.
[41] J. G. Taylor, X. Li, M. Oberthür, W. Zhu, D. E. Kahne, J. Am. Chem. Soc. 2006, 128,
15084–15085.
[42] W. Yang, K. Yoshida, B. Yang, X. Huang, Carbohydr. Res. 2016, 435, 180–194.
[43] E. Bedini, D. Esposito, M. Parrilli, Synlett 2006, 825–830.
[44] Y. Liu, J. Zeng, J. Sun, L. Cai, Y. Zhao, J. Fang, B. Hu, P. Shu, L. Meng, Q. Wan, Org.
Chem. Front. 2018, 5, 2427–2431.
[45] X. Xiao, Y. Zhao, P. Shu, X. Zhao, Y. Liu, J. Sun, Q. Zhang, J. Zeng, Q. Wan, J. Am.
Chem. Soc. 2016, 138, 13402–13407.
[46] M. P. Lisboa, N. Khan, C. Martin, F. F. Xu, K. Reppe, A. Geissner, S. Govindan, M.
Witzenrath, C. L. Pereira, P. H. Seeberger, Proc. Natl. Acad. Sci. U. S. A. 2017, 114,
11063–11068.
[47] D. Lloyd, M. Bylsma, D. K. Bright, X. Chen, C. S. Bennett, J. Org. Chem. 2017, 82,
3926–3934.
[48] S. Boonyarattanakalin, X. Liu, M. Michieletti, B. Lepenies, P. H. Seeberger, J. Am.
Chem. Soc. 2008, 130, 16791–16799.
[49] C. E. Martin, F. Broecker, S. Eller, M. A. Oberli, C. Anish, C. L. Pereira, P. H.
Seeberger, Chem. Commun. 2013, 49, 7159–7161.
[50] L. Van Hijfte, R. D. Little, J. Org. Chem. 1985, 50, 3940–3942.
[51] D. Goldblatt, B. D. Plikaytis, M. Akkoyunlu, J. Antonello, L. Ashton, M. Blake, R.
Burton, R. Care, N. Durant, I. Feavers, P. Fernsten, F. Fievet, P. Giardina, K. Jansen,
L. Katz, L. Kierstead, L. Lee, J. Lin, J. Maisonneuve, M. H. Nahm, J. Raab, S.
Romero-Steiner, C. Rose, D. Schmidt, J. Stapleton, G. M. Carlone, Clin. Vaccine
Immunol. 2011, 18, 1728–1736.
[52] S. van Selm, L. M. van Cann, M. A. B. Kolkman, B. A. M. van der Zeijst, J. P. M. van
Putten, Infect. Immun. 2003, 71, 6192 LP – 6198.
[53] A. Kerdsin, K. Oishi, S. Sripakdee, N. Boonkerd, P. Polwichai, S. Nakamura, R.
Uchida, P. Sawanpanyalert, S. Dejsirilert, J. Med. Microbiol. 2009, 58, 1508–1513.
[54] D. Roy, J.-P. Auger, M. Segura, N. Fittipaldi, D. Takamatsu, M. Okura, M. Gottschalk,
Can. J. Vet. Res. 2015, 79, 141–146.
26
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
[55] Y. Kajihara, T. Yamamoto, H. Nagae, M. Nakashizuka, T. Sakakibara, I. Terada, J.
Org. Chem. 1996, 61, 8632–8635.
27
This article is protected by copyright. All rights reserved.

10.1002/anie.202103990
Angewandte Chemie International Edition
Table-of-Contents:
30 novel oligosaccharides resembling the capsular polysaccharides related to four major
serotypes (2, 3, 9, 14) of Streptococcus suis (S. suis) were synthesized. The conjugation-ready
glycans were used to screen the sera of pigs and rabbits. With glycan array study, glycan
epitopes of lead antigens for the development of semi-synthetic glycoconjugate vaccines to
protect from S. suis serotypes 2 and 9 were identified.
Keywords: Carbohydrates; Glycans; Immunology; Oligosaccharides; Total synthesis
28
This article is protected by copyright. All rights reserved.