A modular strategy toward the synthesis of heparin-like oligosaccharides using monomeric building blocks in a sequential glycosylation strategy

Article abstract of DOI:10.1021/ja045613g

A novel flexible assembly strategy is described for the modular synthesis of heparin and heparan sulfates. The reported strategy uses monomeric building bocks to construct the oligosaccharide chain to attain a maximum degree of flexibility. In the assembly, 1-hydroxyl glucosazido- and 1-thio uronic acid donors are combined in a sequential glycosylation protocol using sulfonium triflate activator systems. The key 1-thio uronic acids were obtained in an efficient manner from diacetone glucose employing a chemo- and regioselective oxidation of partially protected glucose and idose thioglycosides.

Full text of DOI:10.1021/ja045613g

Published on Web 02/25/2005
A Modular Strategy Toward the Synthesis of Heparin-like
Oligosaccharides Using Monomeric Building Blocks in a
Sequential Glycosylation Strategy
Jeroen D. C. Code´e,† Bas Stubba,† Marialuisa Schiattarella,†
Herman S. Overkleeft,† Constant A. A. van Boeckel,‡ Jacques H. van Boom,†,§ and
Gijsbert A. van der Marel*,†
Contribution from the Leiden Institute of Chemistry, Gorlaeus Laboratories, Leiden UniVersity,
P.O. Box 9502, 2300 RA Leiden, The Netherlands, and Organon N. V., P.O. Box 20,
5340 BH Oss, The Netherlands
Received July 21, 2004; E-mail: g.marel@chem.leidenuniv.nl
Abstract: A novel flexible assembly strategy is described for the modular synthesis of heparin and heparan
sulfates. The reported strategy uses monomeric building bocks to construct the oligosaccharide chain to
attain a maximum degree of flexibility. In the assembly, 1-hydroxyl glucosazido- and 1-thio uronic acid
donors are combined in a sequential glycosylation protocol using sulfonium triflate activator systems. The
key 1-thio uronic acids were obtained in an efficient manner from diacetone glucose employing a chemo-
and regioselective oxidation of partially protected glucose and idose thioglycosides.
Introduction
targets, well-defined H and HS sequences are required. One of
the most powerful tools to obtain these complex oligosaccharides
Heparin and heparan sulfate (H and HS) are present in many
living organisms and are the most complex members of the
glycosaminoglycan (GAG) superfamily.1 They are composed
of a linear backbone built up from 20 to 200 R-1,4-linked
disaccharide moieties, which in turn are constructed from an
uronic acid (either D-glucuronic or L-iduronic) and a glu-
cosamine residue (Figure 1). H and HS are characterized by
the high degree of sulfation at both the hydroxyl and the amino
groups of the polymer. Sulfate groups have been found on the
3- and/or 6-position of the glucosamine residue and on the
2-position of the uronic acids. In addition, the amino groups of
the polysaccharide can be sulfated, acetylated, or unsubstituted.
The different substitution patterns of the disaccharide units,
together with the variant alternation of the GlcA-GlcN and
IdoA-GlcN moieties, give rise to innumerable H and HS
structures. This microheterogeneity is at the basis of the plethora
of regulatory functions H and HS fulfill, and it is now well-
established that H and HS play a pivotal role in processes such
as blood coagulation, inflammation, homeostasis, cell adhesion,
and cell growth.1,2 However, the microheterogeneity is also the
origin of the lack of detailed knowledge of the mode of H and
HS action at the molecular level. To elucidate the selective mode
of binding of specific H and HS fragments with their protein
is evidently organic synthesis.3 This is nowhere better illustrated
than by the anticoagulant heparin pentasaccharide. Since the
first heroic total syntheses4 two decades ago, an analogue to
this highly functionalized pentamer has been developed into a
novel antithrombotic drug (Arixtra),5 which very selectively
targets the anticoagulation factor Xa through activation of the
protease inhibitor antithrombin III (AT III). Although these total
syntheses represent a milestone in heparin synthesis, they were
highly target-orientated.6 To grant access to a wide range of
differentially substituted H and HS fragments, a general modular
synthetic strategy is required.7-9 Such a strategy uses a set of
(3) For a comprehensive review on the synthesis on GAGs, see: Yeung, B.
K. S.; Chong, P. Y. C.; Petillo, P. A. J. Carbohydr. Chem. 2002, 21, 799-
2002.
(4) (a) Sinay¨, P.; Jacquinet, J.-C. Carbohydr. Res. 1984, 132, C5-C9. (b)
Jaquinet, J.-C.; Petitou, M.; Duchoussoy, P.; Lederman, I.; Choay, J.; Torri,
G.; Sinay¨, P. Carbohydr. Res. 1984, 130, 221-241. (c) Van Boeckel, C.
A. A.; Beetz, T.; Vos, J. N.; De Jong, A. J. M.; Van Aelst, S. F.; Van den
Bosch, R. H.; Mertens, J. M. R.; Van der Vlugt, F. A. Carbohydr. Res.
1985, 4, 293-321.
(5) (a) Herbert, J.-M.; Petitou, M.; Lormeau, J. C.; Cariou, R.; Necciari, J.;
Magnani, H. N.; Zandberg, P.; Van Amsterdam, R. G. M.; Van Boeckel,
C. A. A. CardioVasc. Drug ReV. 1997, 15, 1-26. (b) Turpie, A. G. G.;
Eriksson, B. I.; Larssen, M. R.; Bauer, K. A. Curr. Opin. Hematol. 2003,
10, 327-332.
(6) For selected syntheses of analogous heparin pentasaccharides, see: (a) Van
Boeckel, C. A. A.; Petitou, M. Angew. Chem., Int. Ed. Engl. 1993, 32,
1671-1690. (b) Petitou, M.; Van Boeckel, C. A. A. Angew. Chem., Int.
Ed. 2004, 43, 3118-3133. (c) Jaurand, G.; Basten, J.; Lederman, I.; Van
Boeckel, C. A. A.; Petitou, M. Bioorg. Med. Chem. Lett. 1992, 2, 897-
900. (d) Basten, J.; Jaurand, G.; Olde-Hanter, B.; Petitou, M.; Van Boeckel,
C. A. A. Bioorg. Med. Chem. Lett. 1992, 2, 901-904. (e) Basten, J.;
Jaurand, G.; Olde-Hanter, B.; Duchaussoy, P.; Petitou, M.; Van Boeckel,
C. A. A. Bioorg. Med. Chem. Lett. 1992, 2, 905-910. (f) Westerduin, P.;
Van Boeckel, C. A. A.; Basten, J. E. M.; Broekhoven, M. A.; Lucas, H.;
Rood, A.; Van der Heijden, H.; Van Amsterdam, R. G. M.; Van Dinther,
T. G.; Meuleman, D. G.; Visser, A.; Vogel, G. M. T.; Damm, J. B. L.;
Overklift, G. T. Bioorg. Med. Chem. 1994, 2, 1267-1280. (g) Code´e, J.
D. C.; Van der Marel, G. A.; Van Boeckel, C. A. A.; Van Boom, J. H.
Eur. J. Org. Chem. 2002, 3954-3965.
† Leiden Institute of Chemistry.
‡ Organon N. V.
§ Deceased 7/31/2004.
(1) (a) Capila, I.; Lindhardt, R. J. Angew. Chem., Int. Ed. 2002, 41, 390-412.
(b) Turnbull, J.; Powell, A.; Guimond, S. Trends Cell Biol. 2001, 11, 75-
82. (c) Bernfield, M.; Go¨tte, M.; Park, P. W.; Reizes, O.; Fitzgerald, M.
L.; Lincecum, J.; Zako, M. Annu. ReV. Biochem. 1999, 68, 729-777. (d)
Rabenstein, D. A. Nat. Prod. Rep. 2002, 19, 312-331.
(2) (a) Lindahl, U.; Kusche-Gullberg, M.; Kjelle´n, L. J. Biol. Chem. 1998,
273, 24979-24982. (b) Esko, J. D.; Lindahl, U. J. Clin. InVest. 2001, 108,
169-173.
9
10.1021/ja045613g CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 3767-3773
3767

A R T I C L E S
Code´e et al.
Figure 1. Synthetic strategy.
closely related building blocks, which should allow for stan-
dardized coupling conditions and provide flexibility and variety
in use. To attain a maximum degree of flexibility in the assembly
of H and HS fragments, we present here the first modular
strategy that employs monomeric building blocks. The use of
monomeric modules will open up possibilities for the incorpora-
tion of a diverse set of easily available building blocks both
“natural” and “unnatural” (for example, conformationally
restricted, flexible, epimeric, and “non-GAG”) synthons, and it
avoids the use of far-advanced and therefore precious dimer or
oligomer motives. Efficiency in the monomeric assembly
strategy of the H and HS fragments is warranted by the
application of an iterative sequential glycosylation strategy,
which precludes excessive synthetic steps at the oligosaccharide
level.
has to enable the installation of sulfate groups on selected
hydroxyl and amino functions, the stereoselective construction
of the glucosamine-uronic acid backbone forms a major
obstacle. The inherent low reactivity of the uronic acid building
blocks (as compared to their non-oxidized glucose and idose
counterparts)11 requires a glycosylation methodology that at the
same time is robust enough to effectuate an efficient glycosy-
lation and watches over the stereocontrol in the formation of
the glycosidic linkages. In this respect, the R-glucosamine-
glucuronic acid linkage is of special interest, since the stereo-
selective construction of this linkage, in contrast to its R-glu-
cosamine-iduronic acid counterpart, has received very little
attention of the scientific community.12 Finally, an efficient route
of synthesis should be devised for the necessary building blocks.
The latter holds especially true for the iduronic acid synthons,
since idose or iduronic acid are not readily available from natural
or commercial sources.13,14
Synthetic Strategy
Any H and HS synthesis has to overcome a range of synthetic
hurdles imposed by the complex structure of the H and HS
saccharides. Besides the elaborate protecting group strategy that
The synthetic approach presented here is based on a sequential
glycosylation strategy using monomeric building blocks and
(10) (a) Sakairi, N.; Basten, J. E. M.; Van der Marel, G. A.; Van Boeckel, C.
A. A.; Van Boom, J. H. Chem.-Eur. J. 1996, 2, 1007-1012. (b) Das, S.
K.; Mallet, J.-M.; Esnault, J.; Driguez, P.-A.; Duchaussoy, P.; Sizun, P.;
He´rault, J.-P.; Herbert, J.-M.; Petitou, M.; Sinay¨, P. Angew. Chem., Int.
Ed. 2001, 40, 1670-1673. (c) Das, S. K.; Mallet, J.-M.; Esnault, J.; Driguez,
P.-A.; Duchaussoy, P.; Sizun, P.; He´rault, J.-P.; Herbert, J.-M.; Petitou,
M.; Sinay¨, P. Chem.-Eur. J. 2001, 7, 4821-4834. (d) Kovensky, J.; Mallet,
J.-M.; Esnault, J.; Driguez, P.-A.; Sizun, P.; He´rault, J.-P.; Herbert, J.-M.;
Petitou, M.; Sinay¨, P. Eur. J. Org. Chem. 2002, 3595-3603.
(7) (a) Poletti, L.; Lay, L. Eur. J. Org. Chem. 2003, 2999-3024. (b) Orgueira,
H. A.; Bartolozzi, A.; Schell, P.; Litjens, R. E. J. N.; Palmacci, E. R.;
Seeberger, P. H. Chem.-Eur. J. 2003, 9, 140-169. (c) Haller, M.; Boons,
G.-J. J. Chem. Soc., Perkin Trans. 1 2001, 814-822. (d) Kovensky, J.;
Duchaussoy, P.; Bono, F.; Salmivirta, M.; Sizun, P.; Herbert, J.-M.; Petitou,
M.; Sinay¨, P. Bioorg. Med. Chem. 1999, 7, 1567-1580.
(8) For modular approaches of [glucosamine-iduronic acid] oligomers, see:
(a) De Paz, J.-L.; Angulo, J.; Lassaletta, J.-M.; Nieto, P. M.; Redondo-
Horjaco, M.; Lozano, R. M.; Gime´nez-Gallego, G.; Mart´ın-Lomas, M.
ChemBioChem 2001, 2, 673-685. (b) Tabeur, C.; Mallet, J.-M.; Bono, F.;
Herbert, J.-M.; Petitou, M.; Sinay¨, P. Bioorg. Med. Chem. 1999, 7, 2003-
2012. (c) Poletti, L.; Fleischer, M.; Vogel, C.; Guerrini, M.; Torri, G.; Lay,
L. Eur. J. Org. Chem. 2001, 2727-2734. (d) Lee, J.-C.; Lu, X.-A.; Kulkarni,
S. S.; Wen, Y.-S.; Hung, S.-C. J. Am. Chem. Soc. 2004, 126, 476-477.
(9) For a modular approach toward a [glucosamine-iduronic acid] oligomer
on a soluble solid support, see: Ojeda, R.; De Paz, J.-L.; Mart´ın-Lomas,
M. Chem. Commun. 2003, 2486-2487.
(11) (a) Stachulski, A. V.; Jenkins, G. N. Nat. Prod. Rep. 1998, 172-186. (b)
Garegg, P. J.; Olsson, L.; Oscarson, S. J. Org. Chem. 1995, 60, 2200-
2204. (c) Krog-Jensen, C.; Oscarson, S. Carbohydr. Res. 1998, 308, 287-
296.
(12) Seeberger and co-workers employed a glycosylation strategy in which the
glucuronic acid acceptor building blocks were locked in the 1C4 conforma-
tion to control the stereochemical outcome of these condensations: (a)
Orguira, H. A.; Bartolozzi, A.; Schell, P.; Seeberger, P. H. Angew. Chem.,
Int. Ed. 2002, 41, 2128-2131. (b) Also see ref 7b.
9
3768 J. AM. CHEM. SOC. VOL. 127, NO. 11, 2005

Synthesis of Heparin-like Oligosaccharides
A R T I C L E S
capitalizes on the use of 1-thio uronic acid synthons and the
recently introduced sulfonium activator systems, as is depicted
in Figure 1.15 The 1-thio uronic acid building blocks represent
a highly disarmed class of donor glycosides and require a very
potent activator system to promote their condensation.13a It was
reasoned that the thiophilic sulfonium activator systems diphenyl-
sulfoxide/triflic anhydride (Ph2SO/Tf2O)16,17 and 1-benzene-
sulfinyl piperidine/triflic anhydride (BSP/Tf2O)18 should be
potent enough to activate these highly disarmed 1-thio uronic
acids,13a facilitating their condensation with the appropriate
glucosazide building blocks and therefore allowing their effec-
tive use in the assembly of H and HS fragments. To introduce
the R-glucosamine linkages, we selected 1-hydroxyl glucosazide
donors, which can be activated by the same sulfonium activator
systems.19 It is well-known that the stereochemical outcome of
any condensation greatly depends on the reactivity of the
coupling partners involved and that glycosyl donors and
acceptors of low reactivity generally benefit the formation of
the more thermodynamically favored product. Indeed, a very
high degree of R-selectivity was recurrently observed in the
condensation of 1-hydroxyl donors bearing a nonparticipating
C2-hydroxyl protecting group with a variety of rather unreactive
acceptors.15,20 Accordingly, the glycosylations of the unreactive
C4-hydroxyl of glucuronic acid acceptors with their designated
1-hydroxyl glucosazide donors should also predominantly
provide the R-linked products. Furthermore, the condensation
of L-iduronic acid acceptors and their D-glucosazide coupling
partners has been shown to proceed in a highly stereoselective
fashion to provide the desired R-glucosamine linkage guided
by double stereodifferentiation (matched pair) in the transition
state leading to the interglycosidic bond.21
and suitably protected 1-thio uronic acceptor 3, having a free
4-OH function. The resulting disaccharide can then immediately
be used in the next glycosylation event, in which the thio uronic
acid is condensed with glucosamine building block 4. Unmask-
ing of the anomeric hydroxyl function then paves the way for
a second glycosylation sequence in which the trisaccharide is
elongated with the second uronic acid building block 5 and the
third glucosamine 6 to furnish the pentasaccharide 1. Throughout
this synthetic approach we adopted a well-established protecting
group strategy:4-9 hydroxyls meant to be sulfated are protected
with acyl (acetyl and benzoyl) groups, benzyl groups are
installed on the remaining hydroxyls, and the amino function-
alities are masked as azides.
Synthesis of the Monomeric Building Blocks. The 1-thio
uronic acid synthons take up crucial positions in the strategy
outlined above. It is therefore of paramount importance to have
an efficient route of synthesis for these building blocks.
Although great improvements in the construction of 1-hydroxyl
glucuronic and iduronic acid building blocks have recently been
disclosed,13 no productive syntheses for their 1-thio counterparts
have appeared.22 Evidently, the presence of both the anomeric
thio function and the C5 carboxylic acid function within the
same glycoside significantly complicates their construction. To
this end, a new synthetic route to access these crucial building
blocks was devised. As is depicted in Scheme 1, we took full
advantage of the 2,2,6,6-tetramethyl piperidinyloxy free radical
(TEMPO)-[bis(acetoxy)iodo] benzene (BAIB) reagent combina-
tion,23 which we recently introduced as an efficient means for
the chemo- and regioselective oxidation of thioglycosides into
their corresponding 1-thio uronic acids.24,25 Partially protected
1-thio glucuronic acid precursor 10 was obtained in a straight-
forward manner from 1,2:5,6-di-O-isopropylidene-3-O-benzyl-
glucofuranose 7. Acidic hydrolysis of both acetonide functions
in 7 and ensuing acetylation provided the â-tetraacetate 8,26
which was transformed into thioglycoside 9. Saponification of
the acetate esters in 9 and formation of the 4,6-O-benzylidene
acetal was followed by protection of the C2-hydroxyl with a
benzoyl function to provide anchimeric assistance in the
forthcoming glycosylation reactions. Subsequent acidolysis of
the cyclic acetal furnished the phenyl thioglucoside 10.27 The
crucial oxidation step, in which the primary C6-alcohol was
selectively oxidized in the presence of both the anomeric
thiophenyl moiety and the secondary C4-alcohol function, was
accomplished by treatment of 10 with a catalytic amount of
TEMPO and a slight excess of BAIB as a co-oxidant in a
biphasic dichloromethane/water solvent system to provide the
glucuronic acid 11 in a yield of 83%. Transformation of the
carboxylic acid into the corresponding methyl ester completed
To probe the above-described assembly strategy, fully
protected pentasaccharide 1 (Figure 1) was selected as a model
saccharide. Thus, the construction of target compound 1 will
start with the condensation of 1-hydroxyl glucosazide donor 2
(13) For recent syntheses of iduronic acid building blocks, see: (a) Tabeur, C.;
Machetto, F.; Mallet, J.-M.; Duchaussoy, P.; Petitou, M.; Sinay¨, P.
Carbohydr. Res. 1996, 281, 253-275. (b) Lohman, G. J. S.; Hunt, D. K.;
Ho¨germaier, J. A.; Seeberger, P. H. J. Org. Chem. 2003, 68, 7559-7561.
(c) Ke, W.; Whitfield, D. M.; Gill, M.; Laroque, S.; Yu, S.-H. Tetrahedron
Lett. 2003, 44, 7767-7770. (d) Gavard, O.; Hersant, Y.; Alais, J.; Duverger,
V.; Dilhas, A.; Bascou, A.; Bonnaffe´, D. Eur. J. Org. Chem. 2003, 3603-
3620. (e) Lubineau, A.; Gavard, O.; Alais, J.; Bonnaffe´, D. Tertrahedron
Lett. 2000, 41, 307-311. (f) Dilhas, A.; Bonnaffe´, D. Carbohydr. Res.
2003, 338, 681-686. (g) Adinolfi, A.; Barone, G.; DeLorenzo, F.; Iadonisi,
A. Synlett 1999, 1316-1318. (h) Hung, S.-C.; Thopate, S. R.; Chi, F.-C.;
Chang, S.-W.; Lee, J.-C.; Wang, C.-C.; Wen, Y.-S. J. Am. Chem. Soc.
2001, 123, 3153-3154.
(14) For selected syntheses of dimer building blocks, see: (a) De Paz, J.-L.;
Ojeda, R.; Reichardt, N.; Mart´ın-Lomas, M. Eur. J. Org. Chem. 2003,
3308-3324. (b) Prabhu, A.; Venot, A.; Boons, G.-J. Org. Lett. 2003, 5,
4975-4978. (c) Haller, M.; Boons, G.-J. Eur. J. Org. Chem. 2002, 2033-
2038. (d) Also see ref 13e.
(15) Code´e, J. D. C.; Van den Bos, L. J.; Litjens, R. E. J. N.; Overkleeft, H. S.;
Van Boom, J. H.; Van der Marel, G. A. Org. Lett. 2003, 5, 1947-1950.
(16) (a) Code´e, J. D. C.; Litjens, R. E. J. N.; Den Heeten, R.; Overkleeft, H. S.;
Van Boom, J. H.; Van der Marel, G. A. Org. Lett. 2003, 5, 1519-1522.
(b) Code´e, J. D. C.; Van den Bos, L. J.; Litjens, R. E. J. N.; Overkleeft, H.
S.; Van Boeckel, C. A. A.; Van Boom, J. H.; Van der Marel, G. A.
Tetrahedron 2004, 60, 1057-1064.
(22) Although the synthesis of 1-thio iduronic acid building blocks has been
reported, it suffered from low yields and the thioglycosides could not be
exploited in productive glycosylations towards heparin disaccharides; see
ref 13a.
(23) (a) De Mico, A.; Margarita, R.; Parlanti, L.; Vescovi, A.; Piancatelli, G. J.
Org. Chem. 1997, 62, 6974-6977. (b) Epp, J. B.; Widlanski, T. S. J. Org.
Chem. 1999, 64, 293-295.
(17) Gin and co-workers originally developed the Ph2SO/Tf2O reagent system
for the activation of 1-hydroxyl donors. See ref 19.
(18) (a) Crich, D.; Smith, M. J. Am. Chem. Soc. 2001, 123, 9015-9020. (b)
Crich, D.; Li, H. J. Org. Chem. 2002, 67, 4640-4646. (c) Crich, D.; De la
Mora, M.; Vinod, A. U. J. Org. Chem. 2003, 68, 9523-9532.
(19) (a) Garcia, B. A.; Poole, J. L.; Gin, D. Y. J. Am. Chem. Soc. 1997, 119,
7597-7598. (b) Garcia, B. A.; Gin, D. Y. J. Am. Chem. Soc. 2000, 122,
4269-4279. (c) Nguyen, H. M.; Poole, J. L.; Gin, D. Y. Angew. Chem.,
Int. Ed. 2001, 40, 414-417.
(20) Van den Bos, L. J.; Code´e, J. D. C.; Van Boom, J. H.; Overkleeft, H. S.;
Van der Marel, G. A. Org. Biomol. Chem. 2003, 1, 4160-4165.
(21) Spijker, N. M.; Van Boeckel, C. A. A. Angew. Chew., Int. Ed. Engl. 1991,
30, 180-183.
(24) Van den Bos, L. J.; Code´e, J. D. C.; Van der Toorn, J. C.; Boltje, T. J.;
Overkleeft, H. S.; Van der Marel, G. A. Org. Lett. 2004, 6, 2165-2168.
(25) For alternative oxidation methods, see: (a) Allanson, N. M.; Liu, D.; Chi,
F.; Jain, R. K.; Chen, A.; Ghosh, M.; Hong, L.; Sofia, M. J. Tetrahedron
Lett. 1998, 39, 1889-1892. (b) Magaud, D.; Grandjean, C.; Doutheau, A.;
Anket, D.; Shevchik, V.; Cotte-Patta, N.; Robert-Beaudouy, J. Tetrahedron
Lett. 1997, 38, 241-244. (c) Also see ref 11b.
(26) Takeo, K.; Kitamura, S.; Murata, Y. Carbohydr. Res. 1992, 224, 111-
122.
(27) For the synthesis of the analogous thioethyl compound, see: Ziegler, T.;
Eckhardt, E.; Birault, V. J. Org. Chem. 1993, 58, 1090-1099.
9
J. AM. CHEM. SOC. VOL. 127, NO. 11, 2005 3769

A R T I C L E S
Code´e et al.
Scheme 1. Synthesis of the 1-Thio Uronic Acid Building Blocksa
a Reagents and conditions: (a) PhSH, BF3‚OEt2, DCM, 0 °C to room temperature, 9: 85%, 15: 91% (4:1 R/â); (b) i. KOtBu, MeOH/DCM; ii. PhCH(OMe)2,
p-TsOH, ACN/DMF, 50 °C; iii. BzCl, pyridine, 0 °C to room temperature; iv. 60% TFA in H2O, DCM, 10: 80%; (c) TEMPO, BAIB, DCM, H2O, 11: 83%,
18: 83% 20: 81%; (d) CH2N2, DMF, 3: 91%, 19: 96%, 5: 83%; (e) EtSH, BF3.OEt2, DCM, 0 °C to room temperature, 79% (1.5:1 R/â); (f) i. KOtBu,
MeOH/DCM; ii. DMP, p-TsOH, DMF; iii. BzCl, pyridine, 0 °C to room temperature; iv. 80% HOAc in H2O, 60 °C, 16: 79%; 17: 81% (anomers partially
separated).
Scheme 2. Synthesis of the Glucosazide Building Blocksa
a Reagents and conditions: (a) i. TBDPSCl, pyridine; ii. Ac2O, pyridine, 96%; (b) BnBr, NaH, DMF, 89%; (c) TiCl4, DCM, 85%; (d) Lev2O, dioxane,
pyridine, 100%; (e) 10% TFA in Ac2O, 0 °C to room temperature; (f) 6% piperidine in THF, 27: 87% (two steps), 2: 94% (two steps), 30: 91% (two steps);
(g) 30, Ph2SO, Tf2O, TTBP,38 DCM, -40 °C, then 2-propanol, 82%, (2:3 R/â); (h) H2NNH2‚H2O, HOAc, pyridine, 86%.
the synthesis of the glucuronic acid building block 3. This highly
efficient strategy was also applied to the construction of the
1-thio iduronic acids 5 and 19. Glucose 7 was converted into
its C5-epimer, as reported in the literature,4c and transformed
into the idopyranose 13. The anomeric thio function was
installed to provide idopyranoses 14 and 15, which were
subjected to a similar sequence of reactions as their glucose
congener, to give the 1-thio idose derivatives 16 and 17. BAIB-
mediated TEMPO oxidation of C6 in 16 and 17 proceeded
uneventfully to afford the desired iduronic acids 18 and 20,
which were treated with diazomethane to afford the uronic acid
ester building blocks 19 and 5, respectively.
liberated29 to provide 24, which will function as a masked
equivalent of compound 4 in the assembly of the target
pentasaccharide 1.30 The orthogonally protected 25 was obtained
after protection of the 4-OH in 24 with a levulinoyl group.
Acidolysis of the 1,6-anhydro bridge in 22, 23, and 25 and
selective unmasking of the anomeric hydroxyl functions gave
the corresponding glucosazide building blocks 27, 2, and 30.
Capping of the reducing end of glucosazide 30 with an
2-propanol moiety provided, after delevulinoylation, the terminal
building block 6 for the oligosaccharide assembly.
Glycosylations and Assembly of the Oligosaccharides
The construction of the glucosazide building blocks used in
this study is depicted in Scheme 2 and followed straightforward
synthetic methodologies. Thus, regioselective introduction of
the tert-butyldiphenylsilyl group28 in 21 and acetylation of the
remaining hydroxyl furnished the first fully protected glu-
cosazide 22. Benzylation of the two hydroxyls in 21 gave
compound 23, of which the C4-hydroxyl was selectively
With all the required building blocks in hand, the formation
of the crucial R-glucosamine linkages was examined. It is now
well-established that the condensation of a D-glucopyranoside
donor, bearing a C2-nonparticipating group and a L-idose or
L-iduronic acid acceptor, proceeds highly stereoselectively to
(29) Hori, H.; Nishida, Y.; Ohrui, H.; Meguro, H. J. Org. Chem. 1989, 54, 1346-
1353.
(30) Glucosazide 24 presents a practical glycosyl acceptor, combining orthogonal
protection with efficient reactivity. Furthermore, the locked anomeric
configuration of 24 simplifies the characterization of reaction products as
compared to the analogous glucosazide 4 (R ) Ac), being a mixture of
anomers.
(28) When the tert-butyldimethylsilyl group was employed for the protection
of compound 21 both the OH-3 and OH-4 silylated products were
obtained: Izumi, M.; Tsuruta, O.; Harayama, S.; Hashimoto, H. J. Org.
Chem. 1997, 62, 992-998.
9
3770 J. AM. CHEM. SOC. VOL. 127, NO. 11, 2005

Synthesis of Heparin-like Oligosaccharides
A R T I C L E S
Table 1. Dehydrative Glycosylations Using 1-Hydroxyl Donors and Uronic Acid Acceptorsa
a
Ph2SO, Tf2O, TTBP, DCM, -40 °C to room temperature. b anomeric ratio determined from anomeric mixture by NMR. c BSP, Tf2O, TTBP, DCM,
-40 °C to room temperature.
form the R-linked products,4,6,7b,10,12 which is further guided by
favorable steric effects in the transition state leading to the
glycosidic bond. Analogous condensations employing D-glucose
or D-glucuronic acid acceptors generally proceed less stereo-
selectively to provide R/â-mixtures. From a stereochemical
viewpoint, the formation of the R-D-glucosazide-(1 f 4)-L-
iduronic acid linkage thus poses no problems, whereas the
formation of the corresponding R-D-glucosazide-D-glucuronic
acid glycosidic bond presents a challenge, particularly in
“armed” glycosylation systems. Consequently, we were eager
to find out how glucuronic acid acceptors behave in the
dehydrative glycosylations with 1-hydroxyl glucosazide donors,
and therefore a series of model glycosylations were conducted,
the outcome of which are summarized in Table 1. As can be
seen from entries 1-4, the model glucuronic acids 33 and 34
performed well in the Ph2SO/Tf2O-mediated condensations,
which all proceeded in good yield. In addition, the stereochem-
ical outcome of the condensations proved to be very satisfac-
tory: in all cases the condensation proceeded with a high degree
of R-selectivity. In line with the expectations (vide supra), the
increased reactivity of the coupling partners in entry 4 led to
the formation of some â-linked disaccharide. When the reactivity
of the dibenzylated donor 2 was tempered by employing the
less reactive BSP/Tf2O activator system,31 (entry 4b) a scarcely
better R/â-ratio was obtained at the expense of a slightly lower
yield. Iduronic acid acceptor 357b and donor 27 also performed
well under the dehydrative conditions to furnish the R-linked
disaccharide 40 with complete stereoselectivity (entry 5).
Encouraged by these results, we returned to the 1-thio uronic
acid acceptors 3, 5, and 19 (Scheme 3). Glucosazide 2 was
activated using Ph2SO in combination with Tf2O and subse-
quently condensed with glucuronic acid 3 to provide the
R-linked target disaccharide 42 as the sole product in a very
rewarding 91% yield. Unfortunately, the condensation of
(31) The glycosylating species that arises from reaction with the BSP/Tf2O
system (i.e., glucosazide 52) is more stable and therefore less reactive than
the corresponding glucosazide that is generated using the Ph2SO/Tf2O
activator (i.e., 53).
9
J. AM. CHEM. SOC. VOL. 127, NO. 11, 2005 3771

A R T I C L E S
Code´e et al.
Scheme 3. Glycosylations with 1-Thio Uronic Acid Acceptors 3, 19, and 5a
a Reagents and conditions: (a) Ph2SO, Tf2O, TTBP, DCM, -40 °C to room temperature, 42: 91%; 43: 31% (2.5:1 R/â), 44: 19%, 45: 24%; 46: 43%.
glucosazide 30 and iduronic acid acceptor 19 was less successful
and led to a mixture of products. Three major products were
isolated: in addition to the desired disaccharide 43 (31% yield),
the R-thioglycoside 44 and the 1-â-O-benzoyl-L-iduronic acid
45 were formed in 24 and 19% yield, respectively. Apparently
the OH-4 in 19 is not nucleophilic enough compared to the
ethylthio function at the relatively reactive anomeric center in
19, and therefore a substantial amount of aglycon transfer takes
place.32 Unwanted intermolecular aglycon transfer reactions have
previously been circumvented by changing the protecting groups
in the acceptor glycosides.32b Such a solution could not be
applied here since the protecting group strategy prescribes a
benzyl group33 and a participating acyl function at the C3- and
C2-hydroxyls of the iduronic acid synthon, respectively. There-
fore, it was decided to tune the reactivity of the anomeric thio
function in the acceptor iduronic acid building block, and the
R/â-thioethyl function in 19 was changed to an R-thiophenyl
group in 5.34,35 This subtle modification indeed proved effective
in avoiding the undesired aglycon transfer and led to a more
productive glycosylation, as shown in Scheme 3.
(i.e., 24 and 50). Gratifyingly, the BSP/Tf2O system, developed
in the Crich laboratory,16 also proved to be potent enough for
the activation of 1-thio glucuronic acid 42, and this time addition
of glucosazide 24 to the activated thioglycoside led to the desired
condensation reaction to stereoselectively afford trisaccharide
47 in 76% yield.36 The BSP/Tf2O activator could also be
employed to smoothly activate 1-thio iduronic acid 43 (Scheme
4). We were pleased to establish that glucosamine acceptor 50,37
which is sterically rather encumbered and bears a carbamate
function, was compatible with the developed condensation
conditions and could be coupled with the activated iduronic acid
43 to provide trisaccharide 51, having two orthogonally
protected amino functionalities, in 55% yield.
Having explored both the acceptor and donor properties of
the 1-thio iduronic acid building blocks, the assembly of the
target pentasaccharide 1 was continued. To this end, trimer 47
was transformed into a donor amenable for the second glyco-
sylation sequence. Thus, the anomeric hydroxyl function of
trisaccharide 47 was unmasked by opening of the anhydro bridge
and subsequent deacetylation of the anomeric center. The chain
elongation of compound 48 commenced by extending the
trisaccharide with the phenyl 1-thio iduronic acid building block
5. The dehydrative condensation proceeded with excellent
stereoselectivity to yield tetrasaccharide 49 in 51% yield. No
aglycon transfer was observed during the glycosylation. The
tetrasaccharide 49 could directly be used in the second
condensation event of this glycosylation sequence. Activation
of the anomeric thiophenyl group in 49 with the BSP/Tf2O
reagent system again proceeded efficiently, and ensuing addition
of the terminal glucosamine building block 6 completed the
assembly of target pentasaccharide 1.
Now the stage was set to probe the glycosylating properties
of the thio uronic disaccharides (Scheme 4). At first, the highly
reactive Ph2SO/Tf2O thiophilic promotor system17 was applied
for the activation of the disarmed thio disaccharides. It was
found that this reagent system very effectively activated the
anomeric center of glucuronic acid 42 at low temperature (-60
°C), as judged by TLC analysis. Surprisingly, no ensuing
condensation took place with a choice of acceptor glucosazides
(32) (a) Belot, F.; Jacquinet, J.-C. Carbohydr. Res. 1996, 290, 79-86. (b) Zhu,
T.; Boons, G.-J. Carbohydr. Res. 2000, 329, 709-715.
(33) Naturally ocurring H and HS contains only uronic acid residues which bear
unsulfated hydroxyl functions at the C-3 position. Therefore, a benzyl group
is required during the assembly of the oligosaccharides. For the synthesis
of “non-natural” H and HS sequences bearing sulfates at the OH-3 of the
uronic acids, a benzoyl group can be introduced in the building blocks.
(34) Thiophenyl glycosides are slightly less reactive than their thioethyl
counterparts: (a) Fu¨gedi, P.; Garegg, P. J. Carbohydr. Res. 1986, 149, C9-
C12. (b) Zuurmond, H.; Van der Laan, S. C.; Van der Marel, G. A.; Van
Boom, J. H. Carbohydr. Res. 1991, 215, C1-C3. (c) Lahmann, M.;
Oscarson, S. Can. J. Chem. 2002, 80, 889-893.
Conclusion
In conclusion, we have developed the first modular assembly
strategy for the synthesis of heparin and heparan sulfates that
(36) Premixing of the donor and acceptor glycoside led to the formation of the
acceptor BSP-OTf adduct 54. Similar results were obtained previously,
as reported in ref 16b.
(37) Beetz, T.; Van Boeckel, C. A. A. Tetrahedron Lett. 1986, 27, 5889-5892.
(38) Crich, D.; Smith, M.; Yao, Q.; Picione, J. Synthesis 2001, 323-326.
(35) The anomeric configuration of the glycosides has an effect on their
reactivity, with the â-glycoside being the more reactive donor of the two
anomers: Lemieux, R. U.; Hendricks, K. B.; Stick, R. V.; James, K. J.
Am. Chem. Soc. 1975, 97, 4056-4062.
9
3772 J. AM. CHEM. SOC. VOL. 127, NO. 11, 2005

Synthesis of Heparin-like Oligosaccharides
A R T I C L E S
Scheme 4. Assembly of the Oligosaccharidesa
a Reagents and conditions: (a) BSP, Tf2O, DCM, -60 °C to room temperature, 47: 76%, 1: 53%, 51: 55%; (b) i. 10% TFA in Ac2O, 0 °C to RT; ii.
6% piperidine in THF, 92%; (c) Ph2SO, Tf2O, TTBP, DCM, -40 °C to room temperature, 51%.
employs monomeric building blocks. The development of this
strategy has been made possible by the identification of the
1-thio glucuronic and iduronic acids as powerful synthons for
H and HS assembly. First, a very efficient route of synthesis
was disclosed for these building blocks. Next they were
combined with 1-hydroxyl glucosazides in a sequential glyco-
sylation protocol, in which both donor types were transformed
into efficient glycosylating agents by the application of the
powerful electrophilic sulfonium activator systems Ph2SO/Tf2O
and BSP/Tf2O, which are introduced for the first time in H and
HS assembly. In addition, application of these sulfonium
activator systems for the condensation of the 1-hydroxyl donors
and their uronic acid acceptors led to highly stereoselective
glycosylations, required for the construction of the oligosac-
charide backbone. The reported strategy thus presents all
requirements for the modular assembly of H and HS fragments
as well as other members of the glycosaminoglycan superfamily,
such as the chrondoitin and dermatan sulfates: efficient
synthesis of building blocks, rapid and stereoselective assembly
of the oligosaccharide chain, and a high degree of flexibility,
offered by the monomeric synthons. It should be noted that the
intermediate oligomeric building blocks (e.g., 42, 48, and 49)
can also find potential in a more convergent block strategy,
which uses, for example, tri- and disaccharides. However, a
linear strategy that iteratively uses closely related (monomeric)
building blocks benefits from standardized condensation pro-
cedures and the ability to interchange the building blocks in a
very straightforward manner and hence streamlines the oligomer
assembly process.
Acknowledgment. This work was financially supported by
The Netherlands Technology Foundation (STW). We are
grateful to F. Lefeber and C. Erkelens for recording NMR
spectra and B. Hofte for high resolution mass measurements.
Supporting Information Available: Full experimental and
characterization details for all new compounds. This material
is available free of charge via the Internet at http://pubs.acs.org.
JA045613G
9
J. AM. CHEM. SOC. VOL. 127, NO. 11, 2005 3773