Synthesis of two linear PADRE conjugates bearing a deca- or pentadecasaccharide B epitope as potential synthetic vaccines against Shigella flexneri serotype 2a infection

DOI: 10.1002/chem.200400903

Source and publish data:

Chemistry - A European Journal p. 1625 - 1635 (2005)

Update date:2022-09-26

Topics:

Authors:

Belot, Frederic

Guerreiro, Catherine

Baleux, Francoise

Mulard, Laurence A.

Read Full Text PDF DownLoad Join now for total 90,000,000 free articles

Article abstract of DOI:10.1002/chem.200400903

The blockwise synthesis of the 2-aminoethyl glycosides of a deca- and a pentadecasaccharide made of two and three repeating units, respectively, of the Shigella flexneri serotype 2a specific polysaccharide is reported. The strategy relies on trifluoromethanesulfonic acid mediated glycosylation of a pentasaccharide building block acting as a glycosyl donor and a potential glycoside acceptor. Both targets were made available in amounts large enough for their subsequent conversion into glycoconjugates. Indeed, efficient elongation of the spacer through an acetylthioacetyl moiety and subsequent conjugation onto a Pan HLA DR-binding epitope (PADRE) T-cell-universal peptide resulted in two fully synthetic neoglycopeptides, which will be evaluated as potential vaccines against S. flexneri serotype 2a infections.

Full text of DOI:10.1002/chem.200400903

FULL PAPER

Synthesis of Two Linear PADRE Conjugates Bearing a Deca- or

Pentadecasaccharide B Epitope as Potential Synthetic Vaccines against

Shigella flexneri Serotype 2a Infection**

Frꢀdꢀric Bꢀlot, Catherine Guerreiro, FranÅoise Baleux, and Laurence A. Mulard*^[a]

Abstract: The blockwise synthesis of

the 2-aminoethyl glycosides of a deca-

and a pentadecasaccharide made of

two and three repeating units, respec-

tively, of the Shigella flexneri seroty-

pe 2a specific polysaccharide is report-

ed. The strategy relies on trifluorome-

thanesulfonic acid mediated glycosyla-

potential glycoside acceptor. Both tar-

gets were made available in amounts

large enough for their subsequent con-

version into glycoconjugates. Indeed,

efficient elongation of the spacer

through an acetylthioacetyl moiety and

subsequent conjugation onto Pan

HLA DR-binding epitope (PADRE)

T-cell-universal peptide resulted in two

fully synthetic neoglycopeptides, which

will be evaluated as potential vaccines

against S. flexneri serotype 2a infec-

tions.

Keywords:

antigens

carbo-

hydrates · glycoconjugates · glyco-

peptides · vaccines

tion of

a pentasaccharide building

block acting as a glycosyl donor and a

Introduction

69% of the episodes and 61% of all deaths attributable to

shigellosis involve children under five years of age, which is

of utmost concern. When it is considered that sanitary con-

ditions are not likely to improve rapidly in those areas at

risk and that the global impact of shigellosis cannot be ade-

quately controlled with the available means, a safe and ef-

fective vaccine against the most common serotypes of Shi-

gella would offer great potential to control the disease.

Indeed, the development of a vaccine against shigellosis is

of high priority, as stated by the World Health Organization

in its program against enteric diseases.^[2]Several options, re-

sulting in the development of various experimental vaccines,

have been undertaken to reach this goal.^[3–5]However, there

are as yet no licensed vaccines for shigellosis.

As for several other Gram-negative bacteria, S. flexneri 2a

lipopolysaccharide (LPS) is a major surface antigen. It is

both an essential virulence factor and a major target of the

infected hostꢀs immune response.^[6,7]Protection is serotype-

specific, a fact which points to the O-specific polysaccharide

(O-SP) moiety of the LPS as the target of the protective

immune response. In fact, the repeating unit of this polymer

of less than 30 kD defines the bacterium serotype. This

knowledge has been taken into account and has led to clini-

cal studies of at least three families of LPS-based candidate

Shigella vaccines, with two of them specifically involving the

O-SP moiety in the form of either the detoxified LPS or

synthetic fragments thereof. A critical point in the design of

such polysaccharide vaccines is that O-SPs are T-cell-inde-

Diarrheal diseases resulting from bacterial or viral infections

account for more than 3 million deaths annually. Available

field data point to shigellosis, or bacillary dysentery, as the

major form of infection leading to this poor prognostic.

Indeed, the number of annual episodes of shigellosis was re-

cently estimated to be 164.7 million, with some 1.1 million

deaths among the victims.^[1]Significantly, 163.2 million of

these episodes were reported in developing countries, partic-

ularly in areas where sanitary conditions are insufficient. Of

the four species of Shigella, Shigella flexneri is the major

one responsible for the endemic form of the disease in de-

veloping countries. Furthermore, field data indicate that

among the various S. flexneri strains known to be pathogenic

in humans, serotype 2a is the most prevalent. Infection,

which is spread by the feco-oral route, results from coloniza-

tion of the digestive tract by a number of bacteria as low as

a 100; this results in a high transmission rate. More than

[a] Dr. F. Bꢁlot, C. Guerreiro, Dr. F. Baleux, Dr. L. A. Mulard

Unitꢁ de Chimie Organique, URA CNRS 2128, Institut Pasteur

28 rue du Dr Roux, 75724 Paris Cedex 15 (France)

Fax : (+33)145-688-404

E-mail: lmulard@pasteur.fr

[**] PADRE=Pan HLA DR-binding epitope. Synthesis of ligands related

to the O-specific polysaccharides of Shigella flexneri serotype 2a and

Shigella flexneri serotype 5a, Part 13. For Part 12, see reference [22].

Chem. Eur. J. 2005, 11, 1625 – 1635

DOI: 10.1002/chem.200400903

ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1625

pendent antigens,^[8,9]which are not immunogenic by them-

selves. Nevertheless, benefiting from the successful conver-

sion of bacterial capsular polysaccharides from T-cell-inde-

pendent antigens to T-cell-dependent ones through their co-

valent coupling to a protein carrier, it was shown that O-SPs

could be turned into potent immunogens. Several polysac-

charide–protein conjugates, targeting either S. sonnei, S. dys-

enteriae 1, or S. flexneri 2a, were shown to be safe and im-

munogenic in adults^[7,10]and also in young children when

evaluated in the latter.^[11]In addition, a recent outbreak of

S. sonnei during a randomized, double-blind, field trial al-

lowed Robbins and co-workers to demonstrate that a candi-

date vaccine made of the corresponding detoxified LPS co-

valently linked with the nontoxic recombinant exoprotein A

of Pseudomonas aeruginosa conferred 74% protection in

military recruits.^[12]Even though encouraging data are avail-

able, detoxified-LPS–protein conjugate vaccines remain

complex constructs, especially when obtained from random-

ly activated polysaccharides. Their immunogenicity depends

on several parameters, among which are the length and

nature of the hapten as well as its loading on the protein. It

may be assumed that the use of well-defined synthetic oligo-

saccharides (OSs) suitable for single-point attachment on to

the carrier would allow better control and, consequently, the

optimization of the above-mentioned parameters. The fact

that low-molecular-weight OSs mimicking bacterial antigen-

ic determinants are immunogenic when conjugated onto a

protein carrier was demonstrated in the late 1930s^[13,14]and

has been exploited on several occasions since then.^[15]Sever-

al reports indicated that neoglycoproteins incorporating OSs

comprising one repeating unit or smaller fragments were im-

munogenic in mice.^[16–18]Furthermore, others demonstrated

that conjugates incorporating haptens mimicking a single re-

peating unit of the natural antigen could induce fully protec-

tive antibodies in mice,^[19,20]

tococcus pneumoniae 3.^[19]Analogous observations resulted

from extensive work in the field of Haemophilus influen-

zae b glycoconjugate vaccines.^[26]In the latter case, the suc-

cess of the approach in humans was recently demonstrat-

ed.^[27]In addition, another important point was addressed,

that is, the need for the development of alternatives to the

most commonly used protein carriers that are compatible

with administration in humans. Relying on synthetic T-

helper peptides derived from conventional protein carriers

such as tetanus toxoid^[28]or selected from the pathogenꢀs

own proteins^[29]may be an option. To overcome the limita-

tions associated with the extensive polymorphism of human

leucocyte antigen (HLA) molecules, recombinant carrier

proteins constituted by strings of several human T-helper

antigens from various pathogen origins have also been pro-

posed.^[30]In other strategies targeting an efficient T-helper

contribution in humans, nonnatural T-helper peptides, such

as the Pan HLA DR-binding epitope (PADRE), were engi-

neered based on their capacity to bind to a large number of

HLA class II molecules.^[31]PADRE was found to be efficient

when evaluated as a carrier for carbohydrate B antigens,

such as the lacto-N-fucopentose II and a dodecasaccharide

from the O-SP of Salmonella typhimurium,^[32]as well as

more recently when conjugated to various S. pneumoniae

polysaccharide antigens.^[33]

An extension of the latter approach to the design of neo-

glycopeptides as potential vaccines against S. flexneri 2a in-

fection is reported in the following section.

Results and Discussion

The O-SP of S. flexneri 2a is a branched heteropolysacchar-

ide defined by the pentasaccharide repeating unit I.^[34,35]It

whereas short OSs representing

only part of the natural antigen

repeating unit appeared to al-

ready contain epitopes capable

of inducing protection in rabbits.^[20]Along this line, we re-

cently reported the synthesis of three fully synthetic glyco-

conjugates as potential vaccines against S. flexneri 2a infec-

tion.^[21]These incorporated short OS haptens, representative

of either part of or the whole repeating unit of the O-SP of

S. flexneri 2a.

The thymus dependence of an OS–protein conjugate in-

creases with shortening of the OS length. However, it is

known that epitope size can vary between species.^[22]In par-

ticular, longer OSs may be required for an optimal immune

response in humans.^[23]Indeed, it is anticipated that the

better, in terms both of antigenicity and secondary structure,

the O-SP mimics used as haptens, the better the immunoge-

nicity of the resulting glycoconjugates. For various reasons,

it is theorized that haptens made of at least two contiguous

repeating units may be necessary for the corresponding OS

conjugates to induce antipolysaccharide antibodies efficient-

ly,^[24]as illustrated in the case of S. dysenteriae 1^[25]or Strep-

features a linear tetrasaccharide backbone, which is

common to all S. flexneri O-SPs and comprises an N-acetyl

glucosamine (D) and three rhamnose residues (A, B, C).

The specificity of the serotype is associated with the a-d-

glucopyranosyl residue linked to position 4 of rhamnose C.

We described recently the synthesis of the ECD, B(E)CD,

and AB(E)CD fragments functionalized with an aminoethyl

spacer at their reducing end, and we demonstrated that the

latter could serve as a suitable anchoring point,^[21]as illus-

trated by the synthesis of the corresponding PADRE neo-

glycopeptides. Parallel studies into the recognition of syn-

thetic fragments of the S. flexneri 2a O-SP by protective

monoclonal antibodies outlined the impact of chain elonga-

tion on the recognition process.^[36]Taking both sets of data

into account, we report herein on the synthesis of the 2-ami-

noethyl glycosides of a decasaccharide (1) and a pentadeca-

saccharide (2), which correspond to a dimer [AB(E)CD]₂

and a trimer [AB(E)CD]₃of the branched pentasaccharide

1626

ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

Chem. Eur. J. 2005, 11, 1625 – 1635

Potential Vaccines against Shigella Infections

FULL PAPER

Scheme 1. Retrosynthetic analysis of the aminoethyl glycoside haptens 1 and 2. a=d-alanine, Ac=acetyl, Bn=benzyl, Bz=benzoyl, Pfp=pentafluoro-

phenyl, TCA=trichloroacetimidate, X=cyclohexylalanine, Z=aminocaproic acid.

I, respectively (Scheme 1). The synthesis is based on a mod-

ular approach involving three partners. Basically, it relies on

1) the use of appropriate haptens (3 and 4, respectively)

functionalized at the anomeric position with an aminoethyl

spacer, 2) the incorporation of a thioacetyl acetamido linker

as a masked thiol functionality, and 3) the use of a PADRE

peptide derivatized by a maleimido group on a C-terminal

lysine residue (5).

In considering targets 3 and 4, a disconnection at the D–

A linkage would appear most appropriate. However, others

have shown that such a disconnection strategy was not suita-

ble even with di- or trisaccharide building blocks;^[37,38]thus,

this route was avoided. More recently, disconnections at the

A–B, B–C, and C–D linkages were evaluated in this labora-

tory when the methyl glycoside of the frame-shifted deca-

saccharide D’A’B’(E’)C’DAB(E)C was successfully synthe-

sized.^[39]It was demonstrated on that occasion that discon-

nection at the C–D linkage was indeed appropriate for the

construction of large fragments of the S. flexneri 2a O-SP.

Based on our experience in the field, we designed a block-

wise strategy to synthesize targets 3 and 4 that involved an

AB(E)C tetrasaccharide donor (6), a DAB(E)C potential

acceptor acting as a donor (7), and the recently disclosed ac-

ceptor 8,^[21]bearing a masked aminoethyl spacer, as a pre-

cursor to the reducing end D residue (Scheme 1). Although

permanent blocking of the OH groups at positions 4_Dand

6_Dwith an isopropylidene acetal may appear somewhat un-

usual, this choice was a key feature of the strategy. It was

based on former observations in the methyl glycoside series,

when it was demonstrated that its use could overcome some

of the known drawbacks of the corresponding benzylidene

acetal,^[40,41]including the poor solubility. In order to reduce

the number of synthetic steps, it was found appropriate to

access the AB(E)C donor and the DAB(E)C building block

from a common key AB(E)C tetrasaccharide intermediate 9

(see Scheme 2).^[39]Most of all, the design of the pentasac-

charide building block 7 was a key element to success.

Indeed, a leading concept of the overall strategy was to limit

the number of transformations at later stages in the synthe-

ses. With respect to the choice of 7, the readerꢀs attention is

drawn to 1) the permanent blocking of positions 4_Dand 6_D

with an isopropylidene acetal, 2) the introduction of a par-

ticipating benzoyl group, which is resistant to Zemplꢁn de-

acylation, at position 2_A, 3) the temporary protection of po-

sition 3_Das an orthogonal acetate, 4) the early introduction

of the required acetamido functionality at position 2_D, and

5) the activation of the anomeric position as a trichloroacet-

imidate (TCA). Indeed, the syntheses disclosed herein are

based on the use of the trichloroacetimidate chemistry,^[42]

and known building blocks were used whenever possible.

Synthesis of the tetrasaccharide building block

(Scheme 2): The fully protected tetrasaccharide 9 could be

obtained in high yield when the condensation was run on a

Chem. Eur. J. 2005, 11, 1625 – 1635

www.chemeurj.org

ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1627

L. A. Mulard et al.

ꢀ

Scheme 2. Synthesis of building blocks 6 and 7. a) 1. [Ir(cod){PCH₃(C₆H₅)₂}₂]⁺PF₆(cat.), THF, room temperature, 2 h; 2. I₂, THF/water, room tempera-

ture, 1 h; 3. CCl₃CN, DBU, CH₂Cl₂, 08C, 1 h; b) see ref. [21]; c) Bu₃SnH, AIBN (cat.), toluene, 1008C, 1 h; d) MeONa (cat.), MeOH, room temperature,

25 min; e) Me₂C(OMe)₂, PTSA (cat.), DMF, room temperature, overnight; f) Ac₂O, pyridine, room temperature, 2.5 h; g) 1. [Ir(cod){PCH₃(C₆H₅)₂}₂]⁺

ꢀ

PF₆(cat.), THF, room temperature, 2 h; 2. HgO, HgBr₂, acetone/water, room temperature, 1 h; h) CCl₃CN, DBU, CH₂Cl₂, 08C, 1 h. AIBN=2,2’-azobis-

isobutyronitrile, All=allyl, cod=cycloocta-1,5-diene, DBU=1,8-diazabicyclo [5.4.0]undec-7-ene, DMF=N,N-dimethylformamide, PTSA=p-toluenesul-

fonic acid, THF=tetrahydrofuran.

15 g scale. Preparation of 6 was conveniently achieved with

67% yield from this crucial intermediate according to a con-

ventional protocol, namely, selective removal of the anome-

ric allyl group and subsequent activation upon treatment of

the resulting hemiacetal with trichloroacetonitrile in the

presence of catalytic DBU.

groups were stable to such neutral conditions.^[44]Subsequent

treatment of 15 with trichloroacetonitrile in the presence of

catalytic DBU cleanly gave the key building block 7 (82%

from 14).

Synthesis of the aminoethyl decasaccharide 3 (Scheme 3):

Previous glycosidation attempts in the series indicated that,

when run at low temperature or at room temperature, reac-

tions with the D acceptor 8 occasionally resulted in a some-

what poor yield of the condensation product. This was tenta-

tively explained by the still rather low solubility of 8. When

using 1,2-DCE as the solvent, the condensation could be

performed at higher temperature, which proved rewarding.

Indeed, optimized coupling conditions of 7 and 8, used in

slight excess, relied on the concomitant use of a catalytic

amount of triflic acid in the presence of 4 ꢃ molecular

sieves as the promoter and 1,2-DCE as the solvent, while

the condensation was performed at 758C, according to a

known protocol^[45]that had recently been adapted to the use

of acceptor 8 in the S. flexneri series.^[21]The fully protected

hexasaccharide 16 was isolated in a satisfactory 76% yield.

The resistance of the two isopropylidene acetals to the harsh

acidic conditions of the glycosidation reaction is noteworthy.

The fact that the hemiacetal 15, resulting from the hydroly-

sis of the excess donor, can be recovered is a great advant-

age if one considers scaling up the process (not described).

Resistance of isolated benzoyl groups to Zemplꢁn transes-

terification has been reported.^[46–48]It was also observed pre-

viously in the series, upon attempted removal of a benzoyl

group located at position 2_C.^[39]Thus, as anticipated, selective

deacetylation of the OH group at position 3 of the nonre-

ducing residue gave the D’AB(E)CD acceptor 17 in a yield

of 97%; this result confirmed the orthogonality of the vari-

ous protecting groups in use at this stage. Condensation of

the latter and 6 was performed in 1,2-DCE with triflic acid

as the promoter. One may note that although the condensa-

Synthesis of the pentasaccharide building block

(Scheme 2): We recently described the synthesis of the

DAB(E)C building block 10 bearing a trichloroacetamido

function at position 2_Dby starting from the tetrasaccharide

9. Compound 10, now conveniently prepared on a 5–10 g

scale, was used successfully as the donor in the synthesis of

the D’A’B’(E’)C’DAB(E)C decasaccharide, once it had been

converted into the corresponding trichloroacetimidate.^[39]

However, for the present study, we reasoned that conversion

of the trichloroacetamide moiety into the required acet-

amide at an early stage in the synthesis was preferable. Re-

ductive free-radical dechlorination of 10 by using Bu₃SnH in

the presence of catalytic AIBN allowed the conversion of

the N-trichloroacetyl moiety into an N-acetyl group, to give

the known compound 11 (68%), previously obtained ac-

cording to an alternative and somewhat lower yielding strat-

egy.^[39]Controlled de-O-acetylation of 11 by using a catalytic

amount of methanolic sodium methoxide gave the triol 12,

which was next converted into the corresponding alcohol 13

upon reaction with 2,2-dimethoxypropane (81% from 11).

Conventional acetylation at position 3_Dthen gave the fully

protected intermediate 14 (94%), with the good overall

yield of this three-step conversion (11!14, 76%) outlining

its usefulness. The latter compound was transformed into

the hemiacetal 15 by following a two-step process, involving

iridium-complex-promoted isomerization of the allyl moiety

into the corresponding propen-1-yl group^[43]and hydrolysis

of the latter upon treatment with mercuric chloride, since it

was originally demonstrated that labile isopropylidene

1628

ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

Chem. Eur. J. 2005, 11, 1625 – 1635

Potential Vaccines against Shigella Infections

FULL PAPER

Scheme 3. Synthesis of the decasaccharide hapten 20 bearing a masked-thiol-type spacer. a) TfOH (cat.), MS, 1,2-DCE, 758C, 2.5 h; b) MeONa (cat.),

MeOH, room temperature, 3 h; c) TfOH (cat.), MS, 1,2-DCE, ꢀ358C!108C, 2.5 h; d) 50% aqueous TFA, CH₂Cl₂, 08C; 3 h; e) 1. MeONa (cat.),

MeOH, 558C, overnight; 2. 10% Pd/C, 1n aqueous HCl, EtOH/EtOAc, room temperature, 72 h; f) SAMA-Pfp, 0.1m phosphate buffer (pH 7.4), room

temperature, 1 h. 1,2-DCE=1,2-dichloroethane, MS=4 ꢃ molecular sieves, SAMA-Pfp=S-acetylthioglycolic acid pentafluorophenyl ester, Tf=triflate=

trifluoromethanesulfonyl, TFA=trifluoroacetic acid.

tion involves the construction of the C–D linkage, thus

somewhat resembling the preparation of hexasaccharide 16,

heating was not required and the glycosylation went

smoothly at 108C to give the fully protected decasaccharide

18 (82%). Acidic hydrolysis of the acetals gave the tetraol

19 (75%). Transesterification of the acyl groups was best

performed by overnight heating of 19 in methanolic sodium

methoxide. Final hydrogenolysis of the benzyl groups and

concomitant conversion of the azido group into the corre-

sponding amine gave the target 3 (71% from 19). As ob-

served previously,^[21,49]the latter transformation was best

performed under acidic conditions.

sistance of the isolated benzoyl groups at position 2_Cto

methanolic transesterification could be overcome by running

the transesterification step at high temperature. Lastly, con-

ventional hydrogenolysis of the benzyl groups and concomi-

tant reduction of the azide moiety allowed the smooth con-

version of the de-O-acylated intermediate into the pentade-

casaccharide hapten 4 (65% from 24). Interestingly, al-

though the number of synthetic steps involved may be some-

what challenging, the steps are generally high yielding and

large amounts of 4 are reachable.

Synthesis of the target conjugates 1 and 2: Chemoselective

ligation of the carbohydrate B and peptide Tepitopes was

achieved through coupling of the carbohydrate haptens pre-

functionalized with a thiol function and a maleimido group

correctly introduced at the C terminus of the T-helper pep-

tide, which allows specific and high-yielding modification of

the former in the presence of other nucleophiles.^[50]Based

on reported data on the immunogenicity of various male-

imide-derived coupling agents,^[51]4-(N-maleimido)-n-buta-

noyl was selected as the linker. It was covalently linked to

the side-chain amino group of a lysine residue added to the

C terminus of the PADRE sequence (PADRE-Lys) accord-

ing to an in-house process,^[21]which differs from that de-

scribed previously by others.^[32]Treatment of 3 and 4 with

SAMA-Pfp resulted in the site-selective elongation of their

aminoethyl spacers with a thioacetyl acetamido linker to

yield 20 and 25 in 61% and 63% yield, respectively

(Schemes 3 and 4).^[21]Derivatization could be monitored by

reversed-phase (RP) HPLC with detection at 215 nm, and

structure confirmation was based on MS and NMR spectro-

scopy analysis. Conjugation of the carbohydrate haptens to

the maleimido-activated PADRE-Lys (5) was performed in

phosphate buffer at pH 6.0 in the presence of hydroxyla-

mine^[52]and monitored by RP HPLC. Lastly, RP HPLC pu-

rification gave the target neoglycopeptides 1 and 2 as single

Synthesis of the aminoethyl pentadecasaccharide

(Scheme 4): The rather convenient access to building block

7 allowed larger sequences to be targeted. Thus, having the

hexasaccharide acceptor 17 in hand, we repeated the two-

step glycosylation/deacetylation process involving 7. Analo-

gously to the condensation step leading to the fully protect-

ed decasaccharide, condensation of 17 and the pentasacchar-

ide donor 7 in the presence of triflic acid was run at a tem-

perature below 58C. Under such conditions, the fully pro-

tected undecasaccharide 21 was isolated in an excellent

yield of 90%, a result that once more outlines the compati-

bility of the rather labile isopropylidene groups with the gly-

cosylation conditions in use. Zemplꢁn transesterification at

the nonreducing OH group at position 3_Dof 21, resulting in

the required acceptor 22 (91%), proved to be efficient. Con-

densation of this key intermediate with the tetrasaccharide

trichloroacetimidate donor 6 was performed according to

the same protocol, by using triflic acid as the promoter. The

fully protected pentadecasaccharide 23 was isolated in a sat-

isfactory yield of 82%. Conversion of 23 into the target 4

was performed by running the stepwise sequence described

for the preparation of 3. Acidic hydrolysis of the isopropyli-

dene groups afforded the hexaol 24 (83%). Again, the re-

Chem. Eur. J. 2005, 11, 1625 – 1635

www.chemeurj.org

ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1629

L. A. Mulard et al.

Scheme 4. Synthesis of the pentadecasaccharide hapten 25 bearing a masked-thiol-type of spacer. a) TfOH (cat.), MS, 1,2-DCE, ꢀ308C!58C, 2.5 h;

b) MeONa (cat.), MeOH, room temperature, 3 h; c) 50% aqueous TFA, CH₂Cl₂, 08C, 3 h; d) 1. MeONa (cat.), MeOH, 558C, overnight; 2. 10% Pd/C,

1n aqueous HCl, EtOH/EtOAc, room temperature, 48 h; e) SAMA-Pfp, 0.1m phosphate buffer (pH 7.4), room temperature, 2 h.

appropriately adjusted polarity. Detection was effected, when applicable,

with UV light and/or by charring with orcinol (35 mm) in 4n aqueous

H₂SO₄or EtOH/H₂SO₄(95/5). Preparative chromatography was per-

products, whose identity was assessed by MS analysis, in

yields of 44 and 67%, respectively.

formed by elution from columns of silica gel 60 (particle size: 60–43 mm).

RP HPLC (215 nm) was performed with a Kromasil 5 mm C18 100 ꢃ

4.6ꢄ250 mm analytical column (flow rate: 1 mLmin^ꢀ1). NMR spectra

Conclusion

were recorded at 208C on a Bruker Advance 400 spectrometer (400 MHz

for ¹H, 100 MHz for ¹³C). The external references were TMS (0.00 ppm

The synthesis of the O-SP of S. flexneri Y by way of poly-

condensation of a tritylated cyanoethylidene tetrasaccharide

has been reported by others.^[53]However, this is to our

knowledge is the first report of the total synthesis of fully

defined oligomeric repeating-unit glycosides mimicking the

branched bacterial O-SPs in the S. flexneri series. The strat-

egy disclosed herein gives access to extended fragments of

the O-SP of S. flexneri serotype 2a in a spacer-armed form

suitable for immunological studies. Indeed, by using so-

lution-phase methods, large enough amounts were made for

the synthesis of fully synthetic oligosaccharide conjugates as

potential vaccines targeting S. flexneri serotype 2a infection.

The preparation of such conjugates in the form of two linear

PADRE–oligosaccharide conjugates is exemplified. It is

noteworthy that the synthesis of the pentadecasaccharide 2

differs from that of the decasaccharide 1 by only two addi-

tional steps.

for both ¹H and ¹³C) for solutions in CDCl₃, and dioxane (67.4 ppm for

¹³C) and trimethylsilyl-3-propionic acid sodium salt (0.00 ppm for H) for

solutions in D₂O. Proton signal assignments were made by first-order

analysis of the spectra as well as analysis of two-dimensional ¹H–¹H cor-

relation maps (COSY). Of the two magnetically nonequivalent geminal

protons at C-6, the one resonating at lower field is denoted H-6a and the

one at higher field is denoted H-6b. Interchangeable assignments in the

¹H and ¹³C NMR spectra are marked with an asterisk. Sugar residues in

oligosaccharides are serially lettered according to the lettering of the re-

peating unit of the O-SP and are identified by a subscript in the listing of

signal assignments. Low-resolution mass spectra were obtained by fast-

atom-bombardment (FAB) MS in the positive-ion mode with dithioery-

thridol/dithio-l-threitol (4:1, MB) as the matrix, in the presence of NaI,

and with xenon as the gas. Anhydrous dichloromethane (DCM) and 1,2-

dichloroethane (1,2-DCE), sold on molecular sieves, were used as re-

ceived. 4 ꢃ powdered molecular sieves were kept at 1008C and activated

before use by heating at 2508C under vacuum. Solid-phase peptide syn-

thesis was performed by using standard 9-fluorenylmethoxycarbonyl

(Fmoc) chemistry protocols on a Pioneer peptide synthesizer (Applied

Biosystem). Fmoc-Lys(iv-Dde)-OH, Fmoc-Cha-OH, Fmoc-d-Ala-OH,

Fmoc-e-Ahx-OH and Boc-d-Ala-OH were purchased from NovaBio-

chem (iv-Dde=1-(4,4-dimethyl-2,6-dioxocyclohexlidene)-3-methylbutyl,

Cha (X)=cyclohexylalanine, e-Ahx (Z)=aminocaproic acid). All others

reagents and amino acids were purchased from Applied Biosystem.

Experimental Section

(2-O-Acetyl-3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!2)-(3,4-di-O-

benzyl-a-l-rhamnopyranosyl)-(1!3)-[2,3,4,6-tetra-O-benzyl-a-d-gluco-

pyranosyl-(1!4)]-2-O-benzoyl-a-l-rhamnopyranosyl trichloroacetimidate

(6): (1,5-Cyclooctadiene)bis(methyldiphenylphosphine)iridium(i) hexa-

General methods: Optical rotations were measured at 258C with a

Perkin–Elmer model 241 MC automatic polarimeter. TLC on precoated

slides of silica gel 60 F₂₅₄(Merck) was performed with solvent mixtures of

1630

ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

Chem. Eur. J. 2005, 11, 1625 – 1635

Potential Vaccines against Shigella Infections

FULL PAPER

fluorophosphate (17 mg) was dissolved in THF (10 mL), and the resulting

red solution was degassed in an argon stream. Hydrogen was then bub-

bled through the solution, causing the color to change to yellow. The so-

lution was then degassed again in an argon stream. A solution of 9^[39]

(550 mg, 0.37 mmol) in THF (3 mL) was degassed and added. The mix-

ture was stirred at room temperature for 2 h. A solution of I₂(188 mg,

0.74 mmol) in a mixture of THF (3.5 mL) and water (1 mL) was added.

The mixture was stirred at room temperature for 1 h and the volatiles

were evaporated. The residue was taken up in DCM and the organic

phase was washed successively with a solution of 5% aqueous NaHSO₃,

5% aqueous NaHCO₃, water, and brine. The organic phase was dried by

being passed through a phase separator and was concentrated. The resi-

due was eluted from a column of silica gel with cyclohexane/EtOAc (9:1)

to give the corresponding hemiacetal (489 mg, 91%). Trichloroacetoni-

trile (360 mL) and DBU (6 mL) were added to a solution of the residue

(479 mg) in anhydrous DCM (3.1 mL) at 08C. After 1 h, toluene was

added and the mixture was concentrated. The residue was eluted from a

column of silica gel with cyclohexane/EtOAc (90:10!85:15) and 0.2%

Et₃N to give 6 as a white foam (389 mg, 74%); [a]_D=+228 (c=1,

CHCl₃); ¹H NMR: d=8.72 (s, 1H, C=NH), 8.00–7.00 (m, 45H, Ph), 6.39

(d, 1H, J_1,2=2.5 Hz, H-1_C), 5.60 (dd, 1H, J_2,3=3.0 Hz, H-2_C), 5.58 (dd,

1H, J_1,2=1.7, J_2,3=3.0 Hz, H-2_A), 5.12 (d, 1H, J_1,2=3.2 Hz, H-1_E), 5.08

(m, 2H, H-1_A, 1_B), 5.00–4.00 (m, 16H, CH₂Ph), 4.20 (dd, 1H, H-3_C), 4.05

(dd, 1H, H-3_E), 4.00–3.35 (m, 14H, H-2_E, 4_E, 5_E, 6a_E, 6b_E, 4_C, 5_C, 2_B, 3_B,

4_B, 5_B, 3_A, 4_A, 5_A), 2.05 (s, 3H, OAc), 1.42, 1.36, 1.00 (3d, 9H, H-6_A, 6_B,

6_C) ppm; ¹³C NMR: d=170.3, 165.8 (2C, C=O), 138–127 (Ph), 99.9 (2C,

C-1_A, 1_B), 98.5 (C-1_E), 94.7 (C-1_C), 82.1, 81.2, 80.4, 80.0, 79.1, 78.1, 78.0,

75.2, 71.7, 71.2, 70.7, 69.5, 69.4, 68.7 (16C, C-2_A, 3_A, 4_A, 5_A, 2_B, 3_B, 4_B, 5_B,

2_C, 3_C, 4_C, 5_C, 2_E, 3_E, 4_E, 5_E), 76.0, 75.7, 75.5, 75.1, 74.3, 73.3, 72.2, 71.2

(8C, PhCH₂), 68.5 (C-6_E), 21.4 (OAc), 19.2, 18.5, 18.1 (C-6_A, 6_B, 6_C) ppm;

elemental analysis calcd (%) for C₉₁H₉₆Cl₃NO₂₀: C 67.05, H 5.94, N 0.86;

found: C 66.44, H 6.21, N 0.93.

(2.65 g, 1.47 mmol) was dissolved in MeOH (20 mL). MeONa was added

until a value of pH 10 was reached. The mixture was stirred for 25 min

then treated with IR 120 (H⁺) resin until a neutral pH value was

reached. The solution was filtered and concentrated. The residue was

eluted from a column of silica gel with DCM/MeOH (9:1) to give the ex-

pected triol 12 which was then treated overnight at room temperature

with 2,2-dimethoxypropane (11 mL, 0.1 mol) and PTSA (20 mg,

0.17 mmol) in DMF (20 mL). Et₃N was added and the solution was

evaporated. The residue was eluted from a column of silica gel with cy-

clohexane/EtOAc (1:1) and 0.2% Et₃N to give 13 as a white foam

(2.05 g, 81% from 11); [a]_D=+38 (c=1, CHCl₃); H NMR: d=6.98–8.00

(m, 45H, Ph), 6.17 (brs, 1H, NH_D), 5.82 (m, 1H, All), 5.30 (dd, 1H, J_1,2

1.0, J_2,3=3.0 Hz, H-2_C), 5.11–5.25 (m, 2H, All), 5.06 (brs, 1H, H-1_A), 4.92

(d, 1H, J_1,2=3.1 Hz, H-1_E), 4.88 (d, 1H, J_1,2=1.6 Hz, H-1_B), 4.84 (brs,

1H, H-1_C), 4.35 (d, 1H, H-1_D), 4.34 (dd, 1H, H-2_B), 4.20–4.80 (m, 16H,

CH₂Ph), 4.05 (dd, 1H, H-2_A), 3.36 (dd, 1H, H-2_E), 2.90–4.10 (m, 22H,

All, H-2_D, 3_A, 3_B, 3_C, 3_D, 3_E, 4_A, 4_B, 4_C, 4_D, 4_E, 5_A, 5_B, 5_C, 5_D, 5_E, 6a_D, 6b_D,

6a_E, 6b_E), 1.5 (s, 3H, NHAc), 1.2–0.9 (m, 15H, C(CH₃)₂, H-6_A, 6_B,

6_C) ppm; ¹³C NMR: d=172.7, 164.9 (2C, C=O), 137.7–116.7 (Ph, All),

102.3 (C-1_D), 100.2 (C-1_B), 100.0 (C-1_A), 98.9 (C(CH₃)₂), 97.2 (C-1_E), 95.1

(C-1_C), 82.1, 82.0, 81.8, 81.6, 80.6, 80.3, 79.0, 78.8, 78.3, 77.8, 77.6, 75.7,

75.6, 75.0, 74.3, 72.8, 71.8, 71.6, 70.8, 70.3, 69.0, 68.5, 67.8, 67.4, 61.9, 60.8,

60.5, 29.4 (C(CH₃)₂), 22.7 (NHAc), 19.0 (C(CH₃)₂), 18.9, 18.4, 18.2 (3C,

C-6_A, 6_B, 6_C) ppm; FAB MS for C₁₀₁H₁₁₅NO₂₄[M]⁺(1726.9): m/z: 1749.7

[M+Na]⁺; elemental analysis calcd (%) for C₁₀₁H₁₁₅NO₂₄·H₂O: C 69.52,

H 6.76, N 0.80; found: C 69.59, H 6.71, N 0.57.

Allyl (2-acetamido-3-O-acetyl-4,6-O-isopropylidene-2-deoxy-b-d-gluco-

pyranosyl)-(1!2)-(3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!2)-(3,4-di-

O-benzyl-a-l-rhamnopyranosyl)-(1!3)-[2,3,4,6-tetra-O-benzyl-a-d-glu-

copyranosyl-(1!4)]-2-O-benzoyl-a-l-rhamnopyranoside (14): A mixture

of 13 (2.05 g, 1.19 mmol) in pyridine (60 mL) was cooled to 08C. Acetic

anhydride (20 mL) was added and the solution was stirred for 2.5 h. The

solution was concentrated and coevaporated with toluene. The residue

was eluted from a column of silica gel with cyclohexane/EtOAc (2:1) and

0.2% Et₃N to give 14 as a white foam (1.99 g, 94%); [a]_D=+18 (c=1,

Allyl

(2-acetamido-3,4,6-tri-O-acetyl-2-deoxy-b-d-glucopyranosyl)-(1!

2)-(3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!2)-(3,4-di-O-benzyl-a-l-

rhamnopyranosyl)-(1!3)-[2,3,4,6-tetra-O-benzyl-a-d-glucopyranosyl-

(1!4)]-2-O-benzoyl-a-l-rhamnopyranoside (11):

mixture of 10^[39]

CHCl₃); H NMR: d=6.95–8.00 (m, 45H, Ph), 5.82 (m, 1H, All), 5.46 (d,

(3.14 g, 1.6 mmol), Bu₃SnH (2.5 mL, 9.3 mmol), and AIBN (240 mg) in

dry toluene (40 mL) was stirred for 30 min at room temperature under a

stream of dry argon and was then heated for 1 h at 1008C, cooled, and

concentrated. The residue was eluted from a column of silica gel with pe-

troleum ether/EtOAc (3:2) to give 11 as a white foam (2.0 g, 68%);

1H, J_2,NH=8.0 Hz, NH_D), 5.29 (dd, 1H, J_1,2=1.0, J_2,3=3.0 Hz, H-2_C),

5.11–5.25 (m, 2H, All), 5.00 (brs, 1H, H-1_A), 4.90 (d, 1H, J_1,2=3.1 Hz, H-

1_E), 4.85 (d, 1H, J_1,2=1.6 Hz, H-1_B), 4.83 (brs, 1H, H-1_C), 4.70 (dd, 1H,

_2,3=J_3,4=10.0 Hz, H-3_D), 4.44 (d, 1H, H-1_D), 4.34 (dd, 1H, H-2_B), 4.20–

4.80 (m, 16H, CH₂Ph), 4.02 (dd, 1H, H-2_A), 3.37 (dd, 1H, H-2_E), 2.90–

4.10 (m, 21H, All, H-2_D, 3_A, 3_B, 3_C, 3_E, 4_A, 4_B, 4_C, 4_D, 4_E, 5_A, 5_B, 5_C, 5_D, 5_E,

6a_D, 6b_D, 6a_E, 6b_E), 1.92 (s, 3H, OAc), 1.57 (s, 3H, NHAc), 1.27–0.90 (m,

15H, C(CH₃)₂, H-6_A, 6_B, 6_C) ppm; ¹³C NMR: d=171.3, 170.3, 166.2 (3C,

C=O), 138.7–117.9 (Ph, All), 103.9 (C-1_D), 101.5 (C-1_B), 101.4 (C-1_A),

99.9 (C(CH₃)₂), 98.5 (C-1_E), 96.3 (C-1_C), 82.1, 81.7, 81.6, 80.3, 80.1, 78.8,

78.1, 77.8, 76.0, 75.8, 75.3, 75.1, 74.7, 74.2, 73.6, 73.3, 72.7, 71.9, 71.4, 70.8,

69.0, 68.8, 68.7, 68.4, 68.1, 67.8, 62.1, 55.0 (C-2_D), 30.0 (C(CH₃)₂), 23.5

(NHAc), 21.6 (OAc), 19.2 (C(CH₃)₂), 19.0, 18.3, 18.2 (3C, C-6_A, 6_B,

6_C) ppm; FAB MS for C₁₀₃H₁₁₇NO₂₅[M]⁺(1769.0): m/z: 1791.9 [M+Na]⁺;

elemental analysis calcd (%) for C₁₀₃H₁₁₇NO₂₅: C 69.93, H 6.67, N 0.79;

found: C 69.77, H 6.84, N 0.72.

[a]_D=+38 (c=1, CHCl₃); H NMR: d=8.00–7.00 (m, 45H, Ph), 5.82 (m,

1H, All), 5.58 (d, 1H, J_2,NH=8.0 Hz, N-H_D), 5.35 (dd, 1H, J_1,2=1.0, J_2,3

2.3 Hz, H-2_C), 5.19 (m, 2H, All), 5.10 (d, 1H, J_1,2=1.0 Hz, H-1_A), 4.92

(dd, 1H, J_2,3=10.5, J_3,4=10.5 Hz, H-3_D), 4.92 (d, 1H, J_1,2=3.3 Hz, H-1_E),

4.90 (d, 1H, J_1,2=1.7 Hz, H-1_B), 4.89 (d, 1H, H-1_C), 4.88 (dd, 1H, J_4,5

10.0 Hz, H-4_D), 4.62 (d, 1H, J_1,2=8.5 Hz, H-1_D), 4.90–4.35 (m, 16H,

CH₂Ph), 4.40 (m, 1H, H-2_B), 4.10–4.00 (m, 2H, All), 4.08 (dd, 1H, J_2,3

2.4 Hz, H-2_A), 4.02 (dd, 1H, H-3_C), 3.91 (m, 1H, H-2_D), 3.90–3.70 (m,

11H, H-4_C, 5_C, 3_A, 5_A, 6a_D, 6b_D, 3_E, 4_E, 5_E, 6a_E, 6b_E), 3.61 (dd, 1H, J_3,4

9.5 Hz, H-3_B), 3.55 (m, 1H, H-5_B), 3.41–3.40 (m, 3H, H-4_A, 5_D, 2_E), 3.47

(m, 1H, J_4,5=9.5, J_5,6=6.1 Hz, H-5_B), 3.35–3.33 (m, 3H, H-4_A, 5_D, 2_E),

3.25 (dd, 1H, H-4_B), 1.95, 1.70 (3 s, 9H, OAc), 1.65 (s, 3H, NHAc), 1.32

(d, 3H, J_5,6=6.1 Hz, H-6_A), 1.30 (d, 3H, J_5,6=6.0 Hz, H-6_C), 0.97 (d, 3H,

(2-Acetamido-3-O-acetyl-4,6-O-isopropylidene-2-deoxy-b-d-glucopyrano-

syl)-(1!2)-(3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!2)-(3,4-di-O-

benzyl-a-l-rhamnopyranosyl)-(1!3)-[2,3,4,6-tetra-O-benzyl-a-d-gluco-

pyranosyl-(1!4)]-2-O-benzoyl-a-l-rhamnopyranosyl trichloroacetimidate

(7): 1,5-Cyclooctadiene-bis(methyldiphenylphosphine)iridium hexafluoro-

phosphate (50 mg, 58 mmol) was dissolved in THF (10 mL), and the re-

sulting red solution was degassed in an argon stream. Hydrogen was then

bubbled through the solution, causing the color to change to yellow. The

solution was then degassed again in an argon stream. A solution of 14

(1.8 g, 1.02 mmol) in THF (20 mL) was degassed and added. The mixture

was stirred at room temperature overnight, then concentrated to dryness.

The residue was dissolved in acetone (9 mL), then water (2 mL), mercu-

ric chloride (236 mg), and mercuric oxide (200 mg) were added succes-

sively. The mixture was protected from light and stirred at room tempera-

ture for 2 h, then the acetone was evaporated. The resulting suspension

was taken up in DCM, washed twice with 50% aqueous KI, water, and

saturated aqueous NaCl, dried, and concentrated. The residue was eluted

_5,6=6.0 Hz, H-6_B) ppm; ¹³C NMR: d=171.1, 170.8, 170.2, 169.6, 166.2

(5C, C=O), 138.2–118.5 (Ph, All), 103.1 (C-1_D), 101.4 (C-1_B), 101.2 (C-

1_A), 98.5 (C-1_E), 96.4 (C-1_C), 82.2 (C-3_E), 81.7 (C-2_E), 81.7 (C-4_A), 80.4

(C-4_B), 80.2 (C-3_C), 79.0 (C-3_A), 78.6 (C-3_B), 78.1 (C-2_A), 77.8 (C-4_C), 77.6

(C-4_E), 76.0, 75.8, 75.4, 74.7, 74.3, 74.2, 73.3, 70.5 (8C, CH₂Ph), 74.9 (C-

2_B), 72.7 (C-2_C), 72.6 (C-3_D), 71.9 (2C, C-5_E, 5_D), 69.1 (C-5_B), 68.9 (2C,

All, C-5_A), 68.3 (C-6_E), 67.8 (C-5_C), 62.3 (C-6_D), 54.6 (C-2_D), 23.5

(NHAc), 21.1, 21.0, 20.8 (3C, OAc), 19.0 (C-6_C), 18.4 (C-6_A), 18.2 (C-

6_B) ppm; FAB MS for C₁₀₄H₁₁₇NO₂₇[M]⁺(1913.1): m/z: 1936.2 [M+Na]⁺;

elemental analysis calcd (%) for C₁₀₄H₁₁₇NO₂₇: C 68.90, H 6.50, N 0.77;

found: C 68.64, H 6.66, N 1.05.

Allyl

(2-acetamido-4,6-O-isopropylidene-2-deoxy-b-d-glucopyranosyl)-

(1!2)-(3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!2)-(3,4-di-O-benzyl-a-

l-rhamnopyranosyl)-(1!3)-[2,3,4,6-tetra-O-benzyl-a-d-glucopyranosyl-

(1!4)]-2-O-benzoyl-a-l-rhamnopyranoside (13): The pentasaccharide 11

Chem. Eur. J. 2005, 11, 1625 – 1635

www.chemeurj.org

ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1631

L. A. Mulard et al.

from a column of silica gel with cyclohexane/EtOAc (3:2) and 0.2%

Et₃N to give the corresponding hemiacetal 15. Trichloroacetonitrile

(2.4 mL) and DBU (72 mL) were added to a solution of the residue in an-

hydrous DCM (24 mL) at 08C. After 1 h, the mixture was concentrated.

The residue was eluted from a column of silica gel with cyclohexane/

EtOAc (3:2) and 0.2% Et₃N to give 7 as a colorless oil (1.58 g, 82%

from 14); [a]_D=+28 (c=1, CHCl₃); ¹H NMR: d=8.62 (s, 1H, NH),

6.95–8.00 (m, 45H, Ph), 6.24 (d, 1H, J_1,2=2.6 Hz, H-1_C), 5.48 (dd, 1H,

5.10 (d, 1H, J_1,2=1.0 Hz, H-1_A), 4.99 (d, 1H, J_1,2=8.3 Hz, H-1_D), 4.96 (d,

1H, J_1,2=3.2 Hz, H-1_E), 4.90 (d, 1H, J_1,2=1.0 Hz, H-1_B), 4.86 (d, 1H,

_1,2=1.0 Hz, H-1_C), 4.52 (d, 1H, J_1,2=7.5 Hz, H-1_D’), 4.37 (dd, 1H, H-2_B),

4.22 (dd, 1H, H-3_D), 4.02 (dd, 1H, H-2_A), 4.80–4.00 (m, 16H, CH₂Ph),

4.00–2.95 (m, 30H, H-2_D, 4_D, 5_D, 6a_D, 6b_D, 2_E, 3_E, 4_E, 5_E, 6a_E, 6b_E, 3_C, 4_C,

5_C, 3_B, 4_B, 5_B, 3_A, 4_A, 5_A, 2_D’, 3_D’, 4_D’, 5_D’, 6a_D’, 6b_D’, OCH₂CH₂N₃), 2.00–

0.92 (6s and 3d, 27H, NHAc, C(CH₃)₂, H-6_A, 6_B, 6_C) ppm; ¹³C NMR

(partial): d=173.9, 172.1, 166.3 (3C, C=O), 140.0–125.0 (Ph), 103.6 (C-

1_D’), 101.7 (C-1_B), 101.2 (C-1_A), 100.2 (C(CH₃)₂), 100.2 (C-1_D), 99.9

(C(CH₃)₂), 98.2 (C-1_E), 97.8 (C-1_C), 51.1 (CH₂N₃), 29.4, 29.3, 23.9, 22.8,

19.6, 19.2, 18.9, 18.4, 18.2 (C-6_A, 6_B, 6_C, NHAc, C(CH₃)₂) ppm; FAB MS

for C₁₁₁H₁₃₁N₅O₂₉[M]⁺(1999.2): m/z: 2021.8 [M+Na]⁺; elemental analy-

sis calcd (%) for C₁₁₁H₁₃₁N₅O₂₉: C 66.68, H 6.60, N 3.50; found: C 66.63,

H 6.78, N 3.32.

_2,3=3.0 Hz, H-2_C), 5.41 (d, 1H, J_2,NH=8.4 Hz, NH_D), 4.99 (brs, 1H, H-

1_A), 4.92 (d, 1H, J_1,2=3.2 Hz, H-1_E), 4.88 (d, 1H, J_1,2=1.6 Hz, H-1_B), 4.69

(dd, 1H, J_2,3=J_3,4=10.0 Hz, H-3_D), 4.44 (d, 1H, H-1_D), 4.34 (dd, 1H, H-

2_B), 4.20–4.80 (m, 16H, CH₂Ph), 4.02 (dd, 1H, H-2_A), 3.38 (dd, 1H, H-

2_E), 2.90–4.10 (m, 19H, H-2_D, 3_A, 3_B, 3_C, 3_E, 4_A, 4_B, 4_C, 4_D, 4_E, 5_A, 5_B, 5_C, 5_D,

5_E, 6a_D, 6b_D, 6a_E, 6b_E), 1.95 (s, 3H, OAc), 1.55 (s, 3H, NHAc), 1.30–0.85

(m, 15H, C(CH₃)₂, H-6_A, 6_B, 6_C) ppm; ¹³C NMR: d=172.4, 171.4, 166.9

(3C, C=O), 140.2–128.9 (Ph), 104.2 (C-1_D), 101.4 (2C, C-1_A, 1_B), 101.1

(C(CH₃)₂), 98.0 (C-1_E), 94.8 (C-1_C), 92.4 (CCl₃), 82.1, 81.5, 80.2, 80.1,

78.6, 78.1, 77.8, 77.6, 76.0, 75.8, 75.5, 75.0, 74.3, 74.2, 73.5 (C-3_D), 73.4,

71.9, 71.4, 71.0, 70.5, 69.2, 68.8, 68.3, 68.1, 62.1, 54.9 (C-2_D), 29.3

(C(CH₃)₂), 23.4 (NHAc), 21.4 (OAc), 19.2 (C(CH₃)₂), 19.0, 18.2, 18.1 (3C,

C-6_A, 6_B, 6_C) ppm; FAB MS for C₁₀₂H₁₁₃Cl₃N₂O₂₅[M]⁺(1873.3): m/z:

1896.3 [M+Na]⁺; elemental analysis calcd (%) for C₁₀₂H₁₁₃Cl₃N₂O₂₅: C

65.40, H 6.08, N 1.50; found: C 65.26, H 6.02, N 1.31.

2-Azidoethyl (2-O-acetyl-3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!2)-

(3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!3)-[2,3,4,6-tetra-O-benzyl-a-

d-glucopyranosyl-(1!4)]-(2-O-benzoyl-a-l-rhamnopyranosyl)-(1!3)-(2-

acetamido-2-deoxy-4,6-O-isopropylidene-b-d-glucopyranosyl)-(1!2)-

(3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!2)-(3,4-di-O-benzyl-a-l-

rhamnopyranosyl)-(1!3)-[2,3,4,6-tetra-O-benzyl-a-d-glucopyranosyl-

(1!4)]-(2-O-benzoyl-a-l-rhamnopyranosyl)-(1!3)-2-acetamido-2-

deoxy-4,6-O-isopropylidene-b-l-glucopyranoside (18): A mixture of alco-

hol 17 (110 mg, 55 mmol), trichloroacetimidate 6 (179 mg, 110 mmol) and

4 ꢃ molecular sieves in anhydrous 1,2-DCE (2.5 mL) was stirred for 1 h

under dry argon. After cooling at ꢀ358C, triflic acid (5 mL, 50 mmol) was

added dropwise and the mixture was stirred for 2.5 h while it was allowed

to reach 108C. Et₃N (25 mL) was added, and the mixture was filtered and

concentrated. The residue was eluted from a column of silica gel with tol-

uene/EtOAc (4:1!3:1) and 0.2% Et₃N to give 18 as a white foam

(158 mg, 82%); [a]_D=+188 (c=1, CHCl₃); ¹H NMR: d=8.00–6.90

2-Azidoethyl (2-acetamido-3-O-acetyl-2-deoxy-4,6-O-isopropylidene-b-d-

glucopyranosyl)-(1!2)-(3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!2)-

(3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!3)-[2,3,4,6-tetra-O-benzyl-a-

d-glucopyranosyl-(1!4)]-(2-O-benzoyl-a-l-rhamnopyranosyl)-(1!3)-2-

acetamido-2-deoxy-4,6-O-isopropylidene-b-d-glucopyranoside (16):

mixture of donor

(745 mg, 0.4 mmol) and acceptor 8^[21](170 mg,

0.51 mmol), 4 ꢃ molecular sieves, and dry 1,2-DCE (12 mL) was stirred

for 1 h then cooled to 08C. Triflic acid (25 mL) was added. The stirred

mixture was allowed to reach room temperature over 10 min then stirred

again for 2.5 h at 758C. After cooling to room temperature, Et₃N

(100 mL) was added and the mixture was filtered. After evaporation, the

residue was eluted from a column of silica gel with cyclohexane/EtOAc

(1:2) and 0.2% Et₃N to give 16 as a white foam (615 mg, 76%); [a]_D=+

08 (c=1, CHCl₃); ¹H NMR: d=6.95–7.90 (m, 45H, Ph), 6.02 (d, 1H,

(90H, m, Ph), 5.90 (d, 1H, J_2,NH=7.0 Hz, NH_D), 5.58 (d, 1H, J_2,NH

7.5 Hz, NH_D’), 5.45, 5.22 (m, 2H, J_1,2=1.0, J_2,3=2.0 Hz, H-2_C, 2_C’), 5.12

(dd, 1H, H-2_A’), 5.11 (d, 1H, J_1,2=8.3 Hz, H-1_D), 5.05 (d, 1H, J_1,2

1.0 Hz, H-1_A), 5.01 (d, 1H, J_1,2=3.2 Hz, H-1_E), 4.96 (d, 1H, J_1,2=1.0 Hz,

H-1_C), 4.94 (m, 2H, H-1_E, 1_B), 4.86 (d, 1H, H-1_B), 4.82 (d, 1H, H-1_C),

4.72 (d, 1H, H-1_D’), 4.70 (d, 1H, H-1_A’), 4.90–4.20 (m, 36H, 16ꢄOCH₂Ph,

H-2_B, 2_B’, 3_D, 3_D’), 4.00–2.90 (m, 45H, H-2_D, 4_D, 5_D, 6a_D, 6b_D, 3_C, 4_C, 5_C, 2_E,

3_E, 4_E, 5_E, 6a_E, 6b_E, 3_B, 4_B, 5_B, 2_A, 3_A, 4_A, 5_A, 2_D’, 4_D’, 5_D’, 6a_D’, 6b_D’, 3_C’, 4_C’,

5_C’, 2_E’, 3_E’, 4_E’, 5_E’, 6a_E’, 6b_E’, 3_B’, 4_B’, 5_B’, 3_A’, 4_A’, 5_A’, OCH₂CH₂N₃), 2.00 (s,

3H, NHAc), 1.88 (s, 3H, OAc), 1.86 (s, 3H, NHAc), 1.40–0.82 (m, 30H,

H-6_A, 6_B, 6_C, 6_A’, 6_B’, 6_C’, C(CH₃)₂) ppm; ¹³C NMR (partial): d=172.1,

171.4, 170.2, 166.2, 165.9 (5C, C=O), 102.7 (C-1_D’), 101.6, 101.2 (2C, C-1_B,

1_B’), 101.1 (C-1_A), 99.8 (C-1_D), 99.7 (C-1_C), 98.2 (2C, C-1_E, 1_A’), 97.2 (2C,

C-1_C, 1_E), 63.3, 62.6 (2C, C-6_E, 6_E’), 60.0, 57.8 (2C, C-2_D, 2_D’), 51.0

(CH₂N₃), 29.5, 29.4 (2C, C(CH₃)₂), 24.0 (2C, NHAc), 21.3 (OAc), 19.6,

_2,NH=7.1 Hz, NH_D), 5.46 (d, 1H, J_2,NH=8.6 Hz, NH_D’), 5.20 (dd, 1H,

_1,2=1.0, J_2,3=3.0 Hz, H-2_C), 5.03 (d, 1H, J_1,2=8.1 Hz, H-1_D), 5.02 (brs,

1H, H-1_A), 4.92 (d, 1H, J_1,2=3.1 Hz, H-1_E), 4.85 (d, 1H, J_1,2=1.6 Hz, H-

1_B), 4.82 (brs, 1H, H-1_C), 4.70 (dd, 1H, H-3_D’), 4.44 (d, 1H, H-1_D’), 4.30

(dd, 1H, H-2_B), 4.20–4.80 (m, 16H, CH₂Ph), 3.99 (dd, 1H, H-2_A), 3.37

(dd, 1H, H-2_E), 2.90–3.95 (m, 29H, H-2_D, 2_D’, 3_A, 3_B, 3_C, 3_D, 3_E, 4_A, 4_B, 4_C,

4_D, 4_D’, 4_E, 5_A, 5_B, 5_C, 5_D, 5_D’, 5_E, 6a_D, 6b_D, 6a_D’, 6b_D’, 6a_E, 6b_E,

OCH₂CH₂N₃), 2.00 (s, 3H, NHAc), 1.92 (s, 3H, OAc), 1.57 (s, 3H,

NHAc), 1.27–0.90 (m, 21H, 2ꢄC(CH₃)₂, H-6_A, 6_B, 6_C) ppm; ¹³C NMR:

d=172.1, 171.5, 170.3, 166.2 (4C, C=O), 139.0–127.7 (Ph), 103.9 (C-1_D’),

101.7 (C-1_B), 101.2 (C-1_A), 100.0 (C-1_D), 99.9, 99.8 (2C, C(CH₃)₂), 98.3

(C-1_E), 97.8 (C-1_C), 82.0, 81.7, 81.5, 80.8, 80.2, 80.1, 78.9, 78.6, 78.0, 77.9,

76.0, 75.9, 75.8, 75.3, 74.8, 74.6, 74.2, 74.0, 73.6, 73.5, 73.4, 73.0, 71.9, 71.4,

70.8, 69.1, 69.0, 68.8, 68.6, 68.0, 67.7, 67.6, 62.6, 62.1, 60.8, 59.7 (C-2_D),

55.0 (C-2_D’), 51.1 (CH₂N₃), 29.5 (C(CH₃)₂), 29.3 (C(CH₃)₂), 23.9 (NHAc),

23.5 (NHAc), 21.3 (OAc), 19.7 (C(CH₃)₂), 19.2 (C(CH₃)₂), 18.8, 18.4, 18.2

(3C, C-6_A, 6_B, 6_C) ppm; FAB MS for C₁₁₃H₁₃₃N₅O₃₀[M]⁺(2041.3): m/z:

2064.2 [M+Na]⁺; elemental analysis calcd (%) for C₁₁₃H₁₃₃N₅O₃₀: C

66.49, H 6.57, N 3.43; found: C 65.93, H 6.57, N 2.61.

19.5 (2C, C(CH₃)₂), 19.1, 18.9, 18.8, 18.5, 18.2, 18.1 (6C, C-6_A, 6_B, 6_C, 6_A’

6_B’, 6_C’) ppm; FAB MS for C₂₀₀H₂₂₅N₅O₄₈[M]⁺(3446.9): m/z: 3489.5

[M+Na]⁺.

2-Azidoethyl (2-O-acetyl-3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!2)-

(3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!3)-[2,3,4,6-tetra-O-benzyl-a-

d-glucopyranosyl-(1!4)]-(2-O-benzoyl-a-l-rhamnopyranosyl)-(1!3)-(2-

acetamido-2-deoxy-b-d-glucopyranosyl)-(1!2)-(3,4-di-O-benzyl-a-l-

rhamnopyranosyl)-(1!2)-(3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!3)-

[2,3,4,6-tetra-O-benzyl-a-d-glucopyranosyl-(1!4)]-(2-O-benzoyl-a-l-

rhamnopyranosyl)-(1!3)-2-acetamido-2-deoxy-b-d-glucopyranoside (19):

A solution of TFA (2 mL) and water (2 mL) was added dropwise to a so-

lution of 18 (630 mg, 181 mmol) in DCM (12 mL) at 08C. The mixture

was stirred for 3 h at this temperature then concentrated by coevapora-

tion with water and then with toluene. The residue was eluted from a

column of silica gel with toluene/EtOAc (1:1) to give 19 as a white foam

(460 mg, 75%); [a]_D=+98 (c=1, CHCl₃); FAB MS for C₁₉₄H₂₁₇N₅O₄₈

[M]⁺(3386.8): m/z: 3409.2 [M+Na]⁺; elemental analysis calcd (%) for

2-Azidoethyl (2-acetamido-2-deoxy-4,6-O-isopropylidene-b-d-glucopyra-

nosyl)-(1!2)-(3,4-di-O-benzyl-a-l-rhamnopyranosyl)-(1!2)-(3,4-di-O-

benzyl-a-l-rhamnopyranosyl)-(1!3)-[2,3,4,6-tetra-O-benzyl-a-d-gluco-

pyranosyl-(1!4)]-(2-O-benzoyl-a-l-rhamnopyranosyl)-(1!3)-2-acetami-

do-2-deoxy-4,6-O-isopropylidene-b-d-glucopyranoside (17): The hexasac-

charide 16 (615 mg, 0.30 mmol) was dissolved in MeOH (8 mL). MeONa

was added until a value of pH 9 was reached. The mixture was stirred for

3 h then treated with IR 120 (H⁺) resin until a neutral pH value was

reached. The solution was filtered and concentrated. The residue was

eluted from a column of silica gel with DCM/MeOH (25:1) and 0.2%

Et₃N to give 17 as a white foam (590 mg, 97%); [a]_D=+18 (c=1,

CHCl₃); ¹H NMR: d=8.00–7.00 (m, 45H, Ph), 6.10 (d, 1H, NH_D’), 6.05

(d, 1H, J_2,NH=7.4 Hz, NH_D), 5.20 (dd, 1H, J_1,2=1.7, J_2,3=3.0 Hz, H-2_C),

₁₉₄H₂₁₇N₅O₄₈·H₂O: C 68.43, H 6.45, N 2.06; found: C 68.40, H 7.02, N

1.61.

2-Aminoethyl a-l-rhamnopyranosyl-(1!2)-a-l-rhamnopyranosyl-(1!3)-

[a-d-glucopyranosyl-(1!4)]-a-l-rhamnopyranosyl-(1!3)-2-acetamido-2-

deoxy-b-d-glucopyranosyl-(1!2)-a-l-rhamnopyranosyl-(1!2)-a-l-rham-

nopyranosyl-(1!3)-[a-d-glucopyranosyl-(1!4)]-a-l-rhamnopyranosyl-

1632

ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

Chem. Eur. J. 2005, 11, 1625 – 1635

Potential Vaccines against Shigella Infections

FULL PAPER

(1!3)-2-acetamido-2-deoxy-b-d-glucopyranoside (3): A mixture of 19

(130 mg, 38 mmol) in MeOH (4 mL) was treated with MeONa until a

value of pH 9 was reached. The mixture was stirred for 1 h at room tem-

perature then heated at 558C overnight. After the mixture was cooled to

room temperature, IR 120 (H⁺) resin was added until a neutral pH value

was reached, and the solution was filtered and concentrated. The residue

was eluted from a column of silica gel with DCM/MeOH (25:1!20:1) to

give an amorphous residue. A solution of this residue in EtOH (1.5 mL),

EtOAc (150 mL), 1m HCl (66 mL, 2 equiv) was hydrogenated in the pres-

ence of Pd/C (100 mg) for 72 h at room temperature. The mixture was fil-

tered and concentrated into a residue which was eluted from a column of

C-18 resin with water then lyophilized to afford amorphous 3 as a white

foam (41 mg, 71%); [a]_D=ꢀ78 (c=1, water); ¹H NMR (D₂O; partial):

d=4.90 (m, 2H, J_1,2=3.5 Hz, H-1_E, 1_E’), 4.82, 4.76, 4.72, 4.67, 4.52, 4.51

(6ꢄbrs, 6H, H-1_A, 1_B, 1_C, 1_A’, 1_B’, 1_C’), 4.41 (d, 1H, J_1,2=8.6 Hz, H-1_D*),

4.29 (d, 1H, J_1,2=8.6 Hz, H-1_D’*), 1.77 (s, 6H, NHAc), 1.15–0.96 (m,

18H, H-6_A, 6_B, 6_C, 6_A’, 6_B’, 6_C’) ppm; ¹³C NMR (D₂O; partial): d=174.8,

174.7 (2C, C=O), 102.6 (C-1_D*), 102.9, 101.8, 101.6, 101.4, 101.3 (6C, C-

1_A, 1_B, 1_C, 1_A’, 1_B’, 1_C’), 100.8 (C-1_D’*), 97.9 (2C, C-1_E, 1_E’), 56.0, 56.4 (2C,

2 C-6_D, 6_D’), 22.7, 22.6 (2C, NHAc), 18.2, 17.2, 17.0, 16.9 (6C, C-6_A, 6_B,

6_C, 6_A’, 6_B’, 6_C’) ppm; HRMS (MALDI) calcd (%) for C₆₆H₁₁₃N₅O₄₅Na

[M+Na]⁺: 1690.6544; found: 1690.6537.

0.27 mmol) was dissolved in MeOH (30 mL). MeONa was added until a

value of pH 9 was reached. The mixture was stirred for 3 h then treated

with IR 120 (H⁺) resin until a neutral pH value was reached. The so-

lution was filtered and concentrated. The residue was eluted from a

column of silica gel with toluene/EtOAc (1:1) and 0.2% Et₃N to give 22

as a white foam (900 mg, 91%); [a]_D=+158 (c=1, CHCl₃); ¹H NMR

(partial): d=6.95–8.00 (m, 90H, Ph), 6.19 (brs, 1H, NH_D*), 5.96 (d, 1H,

_2,NH=6.8 Hz, NH_D’*), 5.57 (d, 1H, J_2,NH=6.8 Hz, NH_D’’*), 5.22 (dd, 1H,

H-2_C*), 5.13 (dd, 1H, H-2_C’*), 5.10 (d, 1H, H-1_D), 5.07 (brs, 1H, H-1_A*),

5.04 (brs, 1H, H-1_A’*), 4.96 (d, 1H, H-1_E*), 4.94 (d, 1H, H-1_E’*), 4.85

(brs, 1H, H-1_B*), 4.84 (brs, 1H, H-1_B’*), 4.82 (brs, 1H, H-1_C*), 4.70 (d,

1H, H-1_C’*), 4.67 (d, 1H, H-1_D*), 4.44 (d, 1H, H-1_D’*), 4.20–4.80 (m,

16H, CH₂Ph), 2.00, 1.85, 1.58 (3ꢄs, 9H, NHAc), 1.37–0.80 (m, 36H,

C(CH₃)₂, H-6_A, 6_B, 6_C, 6_A’, 6_B’, 6_C’) ppm; ¹³C NMR (partial): d=172.8,

170.9, 170.3, 165.1, 164.7 (5C, C=O), 139.0–127.7 (Ph), 103.5, 103.1 (2C,

C-1_D, 1_D’), 101.5 (2C, C-1_B, 1_B’), 101.2, 101.1 (2C, C-1_A, 1_A’), 99.9 (C-1_D’’),

99.0, 98.8, 98.7 (3C, C(CH₃)₂), 98.3 (C-1_E*), 98.1 (2C, C-1_C*, 1_E’*), 97.8

(C-1_C’*), 82.1, 82.0, 81.9, 81.7, 81.6, 81.5, 80.6, 80.3, 80.2, 80.1, 79.7, 79.1,

78.9, 78.5, 77.9, 77.6, 75.7, 74.9, 74.6, 74.3, 73.3, 73.0, 72.7, 71.9, 71.8, 69.1,

68.9, 68.7, 68.5, 68.0, 67.8, 67.7, 67.6, 67.5, 62.6, 62.3, 61.9, 60.5, 59.9, 57.4,

55.0 (3C, C-2_D, 2_D’, 2_D’’), 51.0 (CH₂N₃), 29.5, 29.3 (3C, C(CH₃)₂), 24.0,

23.9, 22.7 (3C, NHAc), 19.7, 19.6, 19.3 (3C, C(CH₃)₂), 19.0, 18.9, 18.6,

18.5, 18.2, 18.1 (6C, C-6_A, 6_B, 6_C, 6_A’, 6_B’, 6_C’) ppm; FAB MS for