From polymer to size-defined oligomers: An expeditious route for the preparation of chondroitin oligosaccharides

Article abstract of DOI:10.1002/anie.200503551

(Chemical Equation Presented) A synthetic short cut: Controlled acid hydrolysis of polymeric chondroitin sulfate affords a tailor-made disaccharidic fragment in good yield. This compound acts as a versatile starting material for the efficient construction

Full text of DOI:10.1002/anie.200503551

Communications
Synthesis Design
DOI: 10.1002/anie.200503551
From Polymer to Size-Defined Oligomers: An
Expeditious Route for the Preparation of
Chondroitin Oligosaccharides**
Chrystel Lopin and Jean-Claude Jacquinet*
Chondroitin sulfates are structurally complex, polyanionic,
microheterogeneous, linear polysaccharides belonging to the
glycosaminoglycan family. They are ubiquitous components
of the extracellular matrix of all connective tissues, but are
also found on mammalian cell surfaces and in neural tissues.[1]
They are copolymers made of dimeric units composed of d-
glucuronic acid (GlcA) and 2-acetamido-2-deoxy-d-galactose
(GalNAc) arranged in the sequence [!4)-b-d-GlcA-(1!3)-
b-d-GalNAc-(1!]n, and contain on average one sulfate group
per disaccharide unit, but other types with sulfate(s) at
various positions are also known. These sulfation patterns
give rise to numerous biologically important functions, such as
cell–cell recognition,[2] brain development and regeneration,[3]
or binding to selectins,[4] and many other that are not
completely deciphered at the molecular level. Unfortunately,
structurally defined oligosaccharide fragments of chondroitin
and its sulfo forms are not easily prepared, even though they
are in great demand. Thus, rapid and efficient methods for the
preparation of chondroitin sulfate oligomers are critical to
advance our understanding of this important class of biomol-
ecules. Synthetic approaches should ideally provide ready
access and involve a small number of steps. Several syntheses
of chondroitin sulfate oligosaccharides have recently been
reported,[5] all starting from monomeric units. Since d-
GalNAc is a rare and somewhat expensive sugar, its
derivatives have been generally prepared by long routes. In
a continued effort to try to reduce the number of trans-
formations in a multistep procedure, we turned our attention
towards the possible chemical hydrolysis of chondroitin
sulfate polymers. These compounds are abundant and readily
available, and were formerly mainly obtained from bovine
cartilage. However, the discovery of bovine spongiform
encephalopathy has meant that they cannot be used anymore,
at least for medical purposes. Commercial products are now
obtained from cartilage of marine origin (shark, skate), and
[*] Dr. C. Lopin, Dr. J.-C. Jacquinet
Institut de Chimie Organique et Analytique—UMR CNRS 6005
FacultØ des Sciences
UniversitØ d’OrlØans, BP 6759
45067 OrlØans Cedex (France)
Fax: (+33)2-3841-7281
E-mail: jean-claude.jacquinet@univ-orleans.fr
[**] This research was supportedby the “Programme National de
Recherche sur les Maladies OstØoarticulaires” (INSERM Pro-A
0401) andby the Action ConcertØe Incitative “MØcanismes
Fondamentaux des Maladies” (BCMS 0152).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
2574
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 2574 –2578

Angewandte
Chemie
they should constitute an attractive starting material for
organic chemists.
Since the pioneering works of Levene[6] and of Davidson
and Meyer[7] on the structure of chondroitin sulfates, chon-
drosine, the basic repeating unit of chondroitin sulfates, can
be obtained by acid hydrolysis of the polymer, a process that
also results in desulfation and N-deacetylation of the resulting
disaccharide. This product contains a number of attractive
features. Besides its backbone, which still possesses d-GlcA
and d-GalNH2 arranged in the correct manner, there is an
interesting possibility to substitute the free amine group with
a powerful stereocontrolling auxiliary such as a trichloroace-
tyl group,[8] a process that was used successfully for the
synthesis of chondroitin sulfate oligosaccharides.[9] We thus
reinvestigated and improved the protocol devised by David-
son and Meyer, and the known[7] methyl ester 1 (Scheme 1)
Scheme 2. Synthesis of glycosides 7–9 and 11. Reagents andcondi-
tions: a) Ac2O, pyridine, 08C, 8 h; b) PhCHO, TFA, 24 h; then Ac2O,
pyridine, room temperature, 16 h, 72%; c) 75% TFA, CH2Cl2, 08C, 4 h;
then Ac2O, pyridine, room temperature, 6 h, 78%; d) hydrazine ace-
tate, DMF, 30 min; then Cl3CCN, DBU, CH2Cl2, 30 min, 70%; e) alco-
hol, TMSOTf, CH2Cl2, 30 min, 80–90%. Bn=benzyl, Pent=pentenyl,
NAP=2-naphthylmethyl, MP=4-methoxyphenyl, TFA=trifluoroacetic
acid, DMF=N,N-dimethylformamide, DBU=1,8-diazabicyclo-
[5,4,0]undec-7-ene, TMSOTf=trimethylsilyl trifluoromethanesulfonate.
Scheme 1. Preparation of building block 2. Reagents andconditions:
a) IR-120 [H+] resin, H2O; then 0.5m H2SO4, 1008C, 6 h; b) AcCl,
MeOH, 08C, 72 h; c) Cl3CCOCl, pyridine, 08C, 4 h; then CH2Cl2/
MeOH/pyridine, room temperature, 3 h. TCA=trichloroacetyl.
acetylation gave exclusively 3 in good yield. Removal of the
4,6-benzylidene acetal on 5 is a prerequisite since it has been
demonstrated[11] that derivatives in the d-galacto series
locked in an activated 4,6-acetal arrangement, despite the
presence of a participating group at C-2, led to mixtures of
glycosides in which the 1,2-cis species were predominant. The
peracetate 3 was then transformed in two steps into the
crystalline a-trichloroacetimidate 6. The glycosylating power
of 6 was then tested towards a set of alcohols. The coupling of
imidate 6 with benzyl alcohol, 4-pentenol, 2-naphthylmetha-
nol, and the poorly reactive 10[12] in the presence of a
trimethylsilyl triflate catalyst gave exclusively the correspond-
ing 1,2-trans-glycosides 7–9 and 11 in excellent yields.
A block approach was designed for the construction of
oligomers in which crucial coupling reactions were achieved
by using Schmidtꢀs trichloroacetimidate glycosylation proce-
dure.[13] The 2-naphthylmethyl (NAP)-substituted glycoside 9
was selected as a starting material since the NAP group[14]
should be retained after deprotection of the other hydroxy
groups and serve as a fluorophore in biological assays or
easily removed to afford reducing species—both useful tools
was obtained in good yield. Attempted chemoselective N-
trichloroacetylation of 1 failed, but trichloroacetamide 2 was
readily obtained through O,N-pertrichloroacetylation fol-
lowed by selective solvolysis of the trichloroacetate esters.
Unfortunately, acetylation of
2
was troublesome
(Scheme 2), and variation of the conditions (acetylating
reagent, solvent, temperature) led systematically to mixtures
of pyranose 3 and furanose 4 in an approximately 3/1 ratio.
This result is not a surprising[10] outcome in the d-galacto
series as a consequence of the unfavorable interaction of the
4-axial hydroxy group in the pyranoid derivative. To avoid
this ring equilibrium, 2 was first treated with benzaldehyde
and trifluoroacetic acid under thermodynamic control to lock
the galactosamine unit in the pyranose conformation through
formation of a 4,6-benzylidene acetal, and then was perace-
tylated in situ to give 5. Mild acid hydrolysis of 5 followed by
Angew. Chem. Int. Ed. 2006, 45, 2574 –2578ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2575

Communications
for the preparation of microarrays. Transesterification of 9
(Scheme 3) followed by acetalation with 2-methoxypropene
and CSA under kinetic control gave the crystalline bisacetal
12 together with a small amount of its 3’,4’ isomer, which was
easily recycled by acid hydrolysis. Levulinoylation of 12 gave
13, which was submitted to mild acid hydrolysis and
acetylation to afford the crystalline key intermediate 14.
This was then transformed into either acceptor 15 through
selective cleavage of the levulinoyl group or crystalline donor
16 by treatment with DDQ[15] followed by imidoylation.
Coupling of alcohol 15 with donor 16 in the presence of a
TMSOTf catalyst gave the crystalline tetrasaccharide deriv-
ative 17, which was readily transformed into acceptor 18 in
good yield. Further coupling of 18 with imidate 6 or 16 under
the same conditions gave hexasaccharide derivatives 19 and
20, respectively, in 65% yield. Deprotection of 19 was
achieved through transformation of the N-trichloroacetyl
groups into their N-acetyl congeners by radical reduction with
tributylstannane[8] followed by a two-step saponification with
lithium hydroperoxide[16] and sodium hydroxide, to give the
target hexasaccharide glycoside 21 in good yield. Thus, an
efficient route is now open for the preparation of size-defined
chondroitin oligomers containing a fluorescent glycoside as
useful probes for the study of the biosynthesis as well as the
polymerization of chondroitin sulfate chains.
The possibility was next examined of differentiating each
sugar unit so as to prepare the way for the synthesis of
chondroitin sulfo forms with important biological functions.
Another route was explored in which the more economical 4-
methoxyphenyl group was used as a temporary protection at
the anomeric reducing center. Surprisingly, the attempted
coupling of 3 or 6 with 4-methoxyphenol under various
conditions led to anomeric mixtures of glycosides (b ꢀ a
ꢀ 2 : 1). Since anomerization of the b-glycoside was not
observed in the reactions, this lack of stereoselectivity was
attributed to the 2-deoxy-2-trichloroacetamido group partic-
ipating too slowly, possibly as a result of the presence of the
bulky group at O-3 which prevents a rapid change in the
conformation of the d-galacto unit. To overcome this prob-
lem, the trichloromethyl oxazoline 22 (Scheme 4) was pre-
pared directly from 3 by using a protocol reported for the d-
gluco series.[8] Indeed, reaction of 22 with 4-methoxyphenol in
the presence of TMSOTf at 08C gave exclusively the b-
glycoside 23 in excellent yield. It is to be noted that the NAP-
substituted glycoside 9 could be obtained from 3 in 80%
overall yield via 22. Glycoside 9 was finally selected as a
starting material on the basis of its accessability and because
of the greater solubility of its derivatives compared to those of
its 4-methoxyphenyl analogues. Transesterification of 9 fol-
lowed by 4,6-benzylidene formation, 2’,3’-isopropylidenation,
and 4’-levulinoylation, as reported previously, gave the
intermediate 24 in 56% overall yield. Interestingly, the 2’,3’-
isopropylidene acetal on 24 could be selectively removed in
the presence of the 4,6-benzylidene group through mild acid
hydrolysis to give the crystalline diol 25 in 79% yield.
Variants in which the d-GalNAc moiety was sulfated were
planned to be synthesized later on from the common
precursor 31, in which benzoate esters were selected as
permanent protection for hydroxy groups that were not to be
Scheme 3. Synthesis of chondrosine hexasaccharide glycoside 21.
Reagents andconditions: a) NaOMe, MeOH; b) 2-methoxypropene,
CSA, DMF, 90 min, 67% from 9; c) LevOH, DCC, DMAP, CH2Cl2, 3 h,
90%; d) 60% AcOH, 1008C, 1 h; then Ac2O, pyridine, 16 h, 71%;
e) DDQ, CH2Cl2/MeOH, 24 h; then Cl3CCN, DBU, CH2Cl2, 30 min,
66%; f) hydrazine acetate, pyridine, 8 min, 89%; g) TMSOTf, CH2Cl2,
30 min, 57% for 17, 65% for 19 and 20; h) Bu3SnH, AIBN, DMAC,
808C, 2 h, 78%; then LiOH/H2O2, THF, ꢁ108C to room temperature,
16 h; then 4m NaOH, MeOH, room temperature, 4 h, 75%. CSA=
camphorsulfonic acid, LevOH=4-oxopentanoic acid, DCC=N,N-dicy-
clohexylcarbodiimide, DMAP=4-dimethylaminopyridine, DDQ=2,3-
dichloro-5,6-dicyanobenzoquinone, AIBN=2,2’-azobis(2-methylpropio-
nitrile), DMAC=N,N-dimethylacetamide.
A
2576
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 2574 –2578

Angewandte
Chemie
O-sulfonated. To achieve this goal, the intermediate 25
was submitted to 2’,3’-dibenzoylation, mild acid hydrol-
ysis of the 4,6-benzylidene acetal, and chloroacetylation
of the resulting 4,6-diol to give 26 in 76% overall yield.
This latter compound was either transformed into
acceptor 27, through selective cleavage of the levulinoyl
group, or donor 28, through oxidative removal of the
NAP glycoside with DDQ followed by imidoylation.
Coupling of alcohol 27 with imidate 28 in the presence of
a TMSOTf catalyst gave the tetrasaccharide derivative
29 in 71% yield. This intermediate should be ready for
further elongation at the nonreducing end after selective
cleavage of the O-lev group. Removal of the chloroace-
tate esters with thiourea gave the tetrol 30, the N-
trichloroacetyl groups of which were reduced by a
radical method to give the key intermediate 31, a
precursor of the 4,6-disulfated species (chondroitin
sulfate E). Selective benzoylation at O-6 of 31 with
benzoyl cyanide gave 32, a precursor of the 4-sulfated
species (chondroitin sulfate A), whereas controlled
regioselective sulfation at C-6[17] with the sulfur triox-
ide/trimethylamine complex followed by O-acetylation
at C-4 (to assess the positions of sulfation) afforded 33, a
direct precursor of the 6-sulfated species (chondroitin
sulfate C).
Differentiation between O-2’ and O-3’ was next
studied (Scheme 5), and was achieved starting from diol
25. Treatment of 25 with benzoyl cyanide gave exclu-
sively the crystalline 3’-benzoate 34 in 80% yield. This
was then transformed in good yield into the key
intermediate 35 through chloroacetylation at O-2’,
hydrolysis of the 4,6-benzylidene acetal, selective chlor-
oacetylation at O-6 at ꢁ208C, and benzoylation at O-4,
thus opening the way for the preparation of the 6,2’-sulfo
form (chondroitin sulfate D). Tin-mediated[18] benzoyla-
tion of 25 gave the 2’-benzoate 36 in 68% yield, along
with its 3’ isomer 34 (16% yield). It should now be
possible to transform the 2’-benzoate 36 by the same
strategy into precursors of the less common sulfo forms
such as 4,3’-disulfated (chondroitin sulfate K), 6,3’-dis-
ulfated (chondroitin sulfate L), and 4,6,3’-trisulfated
species (chondroitin sulfate M).
Scheme 4. Synthesis of precursors of chondroitin sulfates E (31),
A (32), andC ( 33). Reagents andconditions: a) BF 3·OEt2,
TMSBr, 2,4,6-collidine, Bu4NBr, CH2Cl2, 48 h, 89%; b) 2-naphthyl-
methanol, TMSOTf, CH2Cl2, 30 min, 90%; c) 4-methoxyphenol,
TMSOTf, CH2Cl2, 08C, 30 min, 90%; d) MeONa, MeOH; then
PhCHO, TFA, 4 h, 75%; e) 2-methoxypropene, CSA, DMF,
90 min, 83%; f) LevOH, DCC, DMAP, CH2Cl2, 1 h, 92%; g) 80%
AcOH, CH2Cl2, 20 h, 79%; h) PhCOCl, pyridine, 08C, 1 h, 95%;
i) 75% AcOH, 1008C, 30 min; then (ClAc)2O, pyridine, CH2Cl2,
08C, 1 h, 80%; j) hydrazine acetate, pyridine, room temperature,
8 min, 89%; k) DDQ, CH2Cl2/MeOH, room temperature, 24 h;
then Cl3CCN, DBU, CH2Cl2, 20 min, 68%; l) TMSOTf, CH2Cl2,
room temperature, 1 h, 71%; m) thiourea, pyridine/EtOH, 808C,
2 h, 90%; n) Bu3SnH, AIBN, DMAC, 808C, 2 h, 89%;
o) PhCOCN, pyridine, room temperature, 6 h, 88%;
p) Me3N·SO3, DMF, 408C, 2 h 30; then Ac2O, pyridine, room
temperature, 16 h; then SP-C25 [Na+] resin, 81%. Bz=benzoyl.
Angew. Chem. Int. Ed. 2006, 45, 2574 –2578ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2577

Communications
[7] E. A. Davidson, K. Meyer, J. Am. Chem. Soc. 1954, 76, 5686 –
5689.
[8] G. Blatter, J. M. Beau, J. C. Jacquinet, Carbohydr. Res. 1994, 260,
189 – 202.
[9] a) C. Coutant, J. C. Jacquinet, J. Chem. Soc. Perkin Trans. 1 1995,
1573; b) F. Belot, J.-C. Jacquinet, Carbohydr. Res. 2000, 326, 88 –
97.
[10] N. Karst, J.-C. Jacquinet, J. Chem. Soc. Perkin Trans. 1 2000,
2709 – 2717.
[11] a) D. Qiu, S. S. Gandhi, R. R. Koganty, Tetrahedron Lett. 1996,
37, 595 – 598; b) F. Belot, J. C. Jacquinet, Carbohydr. Res. 2000,
325, 93 – 106.
[12] F. Belot, J. C. Jacquinet, Carbohydr. Res. 1996, 290, 79 – 86.
[13] R. R. Schmidt, Angew. Chem. 1986, 98, 213 – 236; Angew. Chem.
Int. Ed. Engl. 1986, 25, 212 – 235.
[14] M. J. Gaunt, J. Yu, J. B. Spencer, J. Org. Chem. 1998, 63, 4172 –
4173.
[15] J. Xia, S. A. Abbas, R. D. Locke, C. F. Piskorz, J. L. Alderfer,
K. L. Matta, Tetrahedron Lett. 2000, 41, 169 – 173.
[16] H. Lucas, J. E. M. Basten, T. G. van Dinther, D. G. Meuleman,
S. F. van Aelst, C. A.A. van Boeckel, Tetrahedron 1990, 46,
8207 – 8228.
[17] J.-C. Jacquinet, L. Rochepeau-Jobron, J.-P. Combal, Carbohydr.
Res. 1998, 314, 283 – 288.
[18] S. David, S. Hanessian, Tetrahedron 1985, 41, 643 – 663.
Scheme 5. Access to precursors of CS-D (35) andchondroitin sulfates
K, L, andM ( 36). Reagents and conditions: a) PhCOCN, pyridine,
room temperature, 6 h, 80%; b) (ClAc)2O, pyridine, 08C, 1 h, 95%;
c) 75% AcOH, 1008C, 20 min; then (ClAc)2O, pyridine, CH2Cl2,
ꢁ208C, 30 min, 72%; d) PhCOCl, pyridine, 08C, 1 h, 91%; e) Bu2SnO,
dioxane/benzene, reflux, 7 h; then PhCOCl, room temperature, 8 h,
68%.
In conclusion, 1 can be readily prepared on a multigram
scale and its versatility has been demonstrated. This short-
ened sequence renders this approach very attractive and
competitive for the preparation of size-defined chondroitin
oligomers. Also relevant is the rapid access to precursors for
the preparation of various sulfo forms starting from the
common intermediate 25. This route can also be applied to
other biologically important members of the glycosaminogly-
can family such as dermatan sulfate and hyaluronic acid. The
synthesis of a collection of chondroitin oligomers and their
various sulfo forms based on this approach is underway, and
will be reported elsewhere in due course.
Received: October 6, 2005
Revised: November 12, 2005
Published online: March 13, 2006
Keywords: carbohydrates · glycosaminoglycans ·
.
oligosaccharides · synthesis design
[1] a) L. Kjell›n, U. Lindahl, Annu. Rev. Biochem. 1991, 60, 443 –
475; b) R. V. Iozzo, Annu. Rev. Biochem. 1998, 67, 609 – 652.
[2] M. Jalkanen, M. Jalkanen, J. Cell Biol. 1992, 116, 817 – 825.
[3] N. B. Schwartz, N. Domowicz, Glycoconjugate J. 2004, 21, 329 –
341.
[4] H. Kawashima, K. Atarashi, M. Hirose, S. Yamada, K. Sugahara,
M. Misayaka, J. Biol. Chem. 2002, 277, 12921 – 12930.
[5] a) N. Karst, J. C. Jacquinet, Eur. J. Org. Chem. 2002, 815 – 825;
b) N. A. Karst, R. J. Linhardt, Curr. Med. Chem. 2003, 10, 1993 –
2031; c) J. I. Tamura, M. Tokuyoshi, Biosci. Biotechnol. Bio-
chem. 2004, 68, 2436 – 2443; d) S. E. Tully, R. Mabon, C. I. Gama,
S. M. Tsai, X. Liu, L. C. Hsieh-Wilson, J. Am. Chem. Soc. 2004,
126, 7736 – 7737.
[6] P. A. Levene, J. Biol. Chem. 1941, 140, 267 – 277.
2578
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 2574 –2578