Synthesis of a S-linked heparan sulfate trisaccharide as the substrate mimic of heparanase

Article abstract of DOI:10.1016/j.tetlet.2005.04.088

An approach to the construction of the β-(1→4)-S-linkage between a glucuronic and a glucosamine unit, and then to the synthesis of a heparan sulfate trisaccharide containing such a linkage (1) as a nonhydrolyzable substrate mimic of heparanase was developed.

Full text of DOI:10.1016/j.tetlet.2005.04.088

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 4337–4340
Synthesis of a S-linked heparan sulfate trisaccharide as
the substrate mimic of heparanase
Hongzhi Cao and Biao Yu*
State Key Laboratory of Bio-organic and Natural Products Chemistry, Shanghai Institute of Organic Chemistry,
Chinese Academy of Sciences, Shanghai 200032, China
Received 7 March 2005; revised 22 April 2005; accepted 25 April 2005
Abstract—An approach to the construction of the b-(1!4)-S-linkage between a glucuronic and a glucosamine unit, and then to the
synthesis of a heparan sulfate trisaccharide containing such a linkage (1) as a nonhydrolyzable substrate mimic of heparanase was
developed.
Ó 2005 Elsevier Ltd. All rights reserved.
Heparan sulfate (HS) proteoglycans, occurring at cell
surface and in the extracellular matrix (ECM), have
been attributed significant biological functions that are
generally mediated via interactions between the constit-
uent HS fragments and various proteins.1 Heparanase,
an endo-b-D-glucuronidase responsible for cleavage of
HS polysaccharides, thus contributes to the integrity
and functional state of cell surface and ECM, and plays
important roles in the control of normal and patholog-
ical processes, such as morphogenesis, tissue repair, neu-
rite outgrowth, inflammation, autoimmunity, and tumor
metastasis and angiogenesis.2 However, only recently,
the substrate specificity of heparanase has been convinc-
ingly characterized by employing purified recombinant
human heparanase and a series of structurally defined
HS/heparin oligosaccharides.3,4 The minimum recogni-
tion HS backbone was deduced to be the trisaccharide:
GlcN(NS,6S)-a-(1!4)-GlcUA-b-(1!4)-GlcN(NS,6S)
(where GlcN, GlcUA, NS, 6S represent D-glucosamine,
D-glucuronic acid, 2-N-sulfate, and 6-O-sulfate, respec-
tively).3 Replacing the native O-glycosidic linkage
between the GlcUA and the GlcN unit with a S-linkage
in the HS trisaccharide, leading to trisaccharide 1, might
provide a nonhydrolyzable substrate mimic, thus a
competitive inhibitor of heparanase to facilitate the
study of the function of this important enzyme. Such a
O!S strategy has been successfully applied in the study
of the interaction of oligosaccharides with proteins.5
And effective approaches to the synthesis of the S-linked
oligosaccharides have been developed.5 However, con-
struction of the S-linked oligosaccharide involving a glu-
curonic acid residue and the synthesis of a S-linked HS/
heparin oligosaccharide have not been reported. Here
we describe an effective approach to the synthesis of
the S-linked HS trisaccharide 1.
Chemical synthesis of HS/heparin oligosaccharides has
been a topic of intensive research.6 Nevertheless, in
planning the synthesis of a S-analog, that is, trisaccha-
ride 1, the feasibility of two quite effective tactics for
HS/heparin oligosaccharide synthesis is questionable:
(1) The sulfated hydroxyl groups are used to be pro-
tected with the persistent benzyl group (Bn), which
could be released cleanly via hydrogenolysis for sulfa-
tion at a later stage; while in the S-analog synthesis,
the sulfur (or the contaminant of sulfur containing re-
agents) might deteriorate the catalyst for hydrogenoly-
sis. (2) The carboxyl function on the GlcUA residue is
preferably introduced after assembly of the oligosaccha-
ride backbone via oxidation of the corresponding 6-OH
groups; such an oxidation might affect the S-glycosidic
linkage. To avoid these possible problems, we planed
to use p-methoxybenzyl group (PMB) for protection of
the hydroxyl groups requiring sulfation. Protecting the
remaining hydroxyl groups with acyl groups and surro-
gating the 2-amino function of the GlcN residues with
N3, that are common practice in HS/heparin synthesis,6
lead to the orthogonally protected trisaccharide 2 as the
precursor to the synthesis of 1 (Scheme 1). The carboxyl
function on the GlcUA residue should be introduced
Keywords: Heparan sulfate; S-Linked oligosaccharide; Glucuronic
acid; Glucosamine; Glycosylation.
*
Corresponding author. Tel.: +86 21 5492 5131; fax: +86 21 6416
6128; e-mail: byu@mail.sioc.ac.cn
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.04.088

4338
H. Cao, B. Yu / Tetrahedron Letters 46 (2005) 4337–4340
enyl bromide under similar conditions gave the corre-
OSO3Na
sponding 1-S-alkylglucuronate in 30% yield.9 Although
glycosidic coupling with glycosyl trichloroacetimidates
as donors has been successful in the synthesis of the
S-glycosidic linkages,10 glycosylation of the GlcN
4-thiol derivative 8 with GlcUA trichloroacetimidate
O
O
HO
HO
OSO3Na
O
ONa
O
NaSO3HN
O
S
HO
NaSO3HN
HO
OH
1
OMe
7
11 under the promotion of TMSOTf or BF3OEt2 failed
OPMB
O
to provide the S-linked disaccharide 6. Fortunately, the
SN2 displacement of the GlcUA bromide 9 with GlcN
4-thiol 8 proceeded smoothly in the presence of
Cs2CO3,12 affording the desired disaccharide 6 in a
satisfactory 56% yield. The b-linkage was confirmed by
RO
RO
OPMB
O
OMe
O
O
N
3O
S
RO
RO
N
OR
3OMe
the J1 ,2 value of 11.4 Hz (4.68 ppm, H-10) and the
0
0
2: R = Acyl groups
upfielded C-10 signal at 82.4 ppm.
OPMB
O
OMe
O
O
Taking advantage of the higher nucleophilicity of the
thiolate over alkyloxide, we expected to prepare the
S-linked disaccharide bearing a free 40-OH (i.e., 3a)
via direct condensation of a 4-OH-GlcUA 1-bromide
derivative (i.e., 11) with the 4-SH-GlcN derivative 8
(Scheme 3). Thus, treatment of methyl 1,2,3-tri-O-benz-
oyl-glucuronate 1013 with 33% HBr/HOAc at 0 °C pro-
vided the desired GlcUA bromide 11 in 71% yield.
Meanwhile, the corresponding 4-O-acetylated byprod-
uct (methyl 4-O-acetyl-2,3-di-O-benzoyl-glucuronate 1-
bromide) was isolated in 12% yield. Condensation of
the GlcUA 1-a-bromide 11, bearing a free 4-OH, with
GlcN 4-SH 8 in the presence of Cs2CO3 in DMF affor-
ded the desired b-(1!4)-S-linked disaccharide 3a in
good yield (67%). Glycosylation of the hindered 40-OH
in 3a with 2-azido-glucopyranosyl trichloroacetimidate
12 was achieved in the presence of tert-butyldimethylsi-
lyl triflate (TBSOTf, 0.25 equiv), providing the desired
trisaccharide 13 in a satisfactory 74% yield. The final
elaboration of the HS trisaccharide S-mimic 1 from
the fully protected 13 was succeeded expectedly. The
6,600-di-O-PMB group on trisaccharide 13 was removed
cleanly with dichlorodicyanobenzoquinone (DDQ)14
without affecting the S-glycosidic linkage. Subsequent
reduction of the 2,200-azido group was realized with
1,3-dithiopropanol in the presence of triethylamine in
a mixture solvent of pyridine and water,15 furnishing tri-
saccharide 14 in quantitative yield (for two steps).
Simultaneous sulfation of the resulting 6,600-di-OH and
HO
RO
S
RO
N
OR
3OMe
3: R = Acyl groups
Scheme 1.
before assembly of the S-linkage, therefore construction
of the S-glycosidic linkage between a GlcUA and a
GlcN residue (e.g., 3) is required. Two major strategies
have been employed in assembly of the (1!4)-S-glyco-
sidic linkage:5 (1) SN2-type displacement of a 4-O-leav-
ing group (mostly a triflate) of a glycosyl acceptor
with a sugar 1-thiolate or of a glycosyl halide with a su-
gar 4-thiolate; (2) glycosylation-type condensation be-
tween a glycosyl donor and a sugar 4-thiol. Both
strategies might be problematic when applied to the syn-
thesis involving a GlcUA residue, considering the viola-
bility of the GlcUA derivatives toward elimination
under SN2-type conditions7 and the inertness of the
GlcUA donors in glycosylation.8 Therefore we tried a
variety of conditions for the construction of the S-link-
age between a GlcUA and a GlcN residue (Scheme 2).
Condensation of GlcUA 1-b-thiol 49 with GalN 4-O-tri-
flate 5 in the presence of NaH in DMF afforded the
desired b-(1!4)-S-linked disaccharide 6 in a poor 22%
yield. Comparably, coupling of thiol 4 with undec-10-
OMe
O
OPMB
O
O
OMe
O
OTf
OPMB
O
O
a
AcO
AcO
AcO
S
+
BzO
SH
BzO
AcO
22%
OAc
N
3OMe
OAc
N3
OMe
6
4
5
OPMB
OMe
O
O
b
O
HS
AcO
6 (not detected)
O
CCl3
+
BzO
AcO
N
OAc
3OMe
NH
8
7
OMe
O
O
c
AcO
AcO
6
8
+
56%
AcO
Br
9
Scheme 2. Reagents and conditions: (a) NaH, DMF, rt; (b) TMSOTf or BF3OEt2, CH2Cl2, ꢀ78 °C to rt; (c) Cs2CO3, DMF, rt.

H. Cao, B. Yu / Tetrahedron Letters 46 (2005) 4337–4340
4339
OMe
O
OAc
O
O
OMe
O
O
a,b,c
6 steps
a
8, b
HO
BzO
AcO
AcO
HO
HO
D-Glc
OBz
OTBS
BzO
20% yield
BzO
N3
OBz
Br
OPMB
19
11
10
OPMB
O
p-MeOPh
O
O
OMe
O
d
OPMB
HO
BzO
O
O
OTBS
O
O
OTBS
BzO
AcO
CCl3
NH
O
S
N3
+
BzO
BzO
BzO
N3
21
20
N3
OBz
N
3 OMe
12
OPMB
O
3a
f
e
AcO
OPMB
O
OTBS
BzO
AcO
BzO
N3
g
OMe
O
OPMB
O
22
O
c
N
OPMB
O
3 O
S
BzO
BzO
AcO
BzO
12
OH
OBz
N
3 OMe
OH
O
13
N3
23
AcO
BzO
OMe
O
OH
O
O
Scheme 5. Reagents and conditions: (a) NaOMe/MeOH, rt; (b)
p-MeOPhCH(OMe)2, p-TsOH, DMF, 40 °C; (c) BzCl, DMAP,
pyridine, rt, 85% (for three steps); (d) NaBH3CN, TFA, 4 A MS,
H2N
f, g
d, e
O
S
1
BzO
BzO
˚
OBz
H2N
OMe
DMF, 0 °C, 97%; (e) Ac2O/pyridine, DMAP, rt, 100%; (f) 1 M TBAF
in THF, THF, rt, 87%; (g) Cl3CCN, DBU, CH2Cl2, 0 °C, 96%.
14
Scheme 3. Reagents and conditions: (a) 33% HBr/HOAc, 0 °C, 71%;
˚
(b) Cs2CO3, DMF, rt, 67%; (c) TBSOTf, 4A MS, toluene, ꢀ20 °C,
74%; (d) DDQ, CH2Cl2–H2O (10:1, v/v), rt, 99%; (e) HS(CH2)3SH,
Et3N, pyridine/H2O, rt, 100%; (f) SO3Æpyridine, Et3N, CH2Cl2, rt; (g)
NaOH (2 N), rt; followed by neutralization with IR-120 [H+], ion
exchange with IR-120 [Na+], and then desalting with G10, and
lyophilization, 23%.
and in 27% overall yield employing a literature proce-
dure.16 Blocking the 4,6-di-OH in 15 with p-methoxy-
benzylidene led to the isolation of 16 and its b-anomer,
in 33% and 48% yield, respectively. Protection of the
3-OH in 16 with a benzoyl group, followed by reductive
opening of the 4,6-O-p-methoxybenzylidene acetal with
a modified protocol using NaBH3CN–TFA in DMF17
led to the desired 6-O-PMB-galactopyranoside 17 in
excellent regioselectivity (97% yield). The released axial
4-OH was then transformed into a 4-O-triflate (Tf) by
treatment with trifluoromethanesulfonic anhydride
(Tf2O) in the presence of pyridine at 0 °C, providing 5,
which was directly used in the next substitution reaction.
Thus treatment of 5 with potassium thioacetate in DMF
afforded 4-S-acetyl-glucopyranoside 18 (in 87% yield).18
Selective removal of the S-acetate, in the presence of the
3-O-benzoate, was achieved with NaOMe in methanol,
providing 4-SH-glucopyranoside 8, which was utilized
in the subsequent transformation without purification.
2,200-di-NH2 groups in 14 with SO3Æpyridine in the pres-
ence of triethylamine afforded the corresponding tetra-
sulfated trisaccharide, which was directly subjected to
saponification to remove the Ac and Bz protective
groups. The final product 1 was successfully obtained
after ion exchange with IR-120 [Na+] resin, desalting
with G10 gel, and lyophilization.
The hither required glycosamine derivatives 5, 8, and 12
were conveniently prepared as depicted in Schemes 4
and 5. Methyl 2-azido-2-deoxy-galactopyranoside 15
was readily synthesized from D-galactose in six steps
OH
OH
3,4,6-Tri-O-Ac-2-azido-1-O-TBS-b-D-glucopyranoside 19
was readily prepared from glucose in six steps and
20% overall yield following a literature procedure.19
Removal of the 3,4,6-tri-O-acetyl groups in 19, followed
by subsequent blocking of the resulting 4,6-OH with a
p-methoxybenzylidene acetal and the 3-OH with a benzo-
ate, provided 20 conveniently in 85% yield (three steps).
Regioselective opening of the glucopyranose 4,6-O-p-
methoxybenzylidene acetal with NaBH3CN/TFA affor-
ded the desired 6-O-PMB derivative 21 in an excellent
97% yield. The resulting 4-OH was then protected with
a acetate, leading to 22. Cleavage of the 1-O-TBS group
with tetrabutylammonium fluoride (TBAF) in the pres-
ence of acetic acid furnished 23, which was converted
readily into the desired glycosyl trichloroacetimidate 12.
a
6 steps
O
D-Gal
OMe
27% yield
HO
N3
15
PhOMe-p
O
OH OPMB
O
O
d
b,c
O
5
BzO
HO
N3
OMe
17
N3
OMe
16
OPMB
O
f
e
AcS
8
BzO
N3
OMe
18
Scheme 4. Reagents and conditions: (a) p-MeOPhCH(OMe)2,
p-TsOH, DMF, 40 °C, 33% for 16 and 48% for the b isomer; (b)
BzCl, DMAP, pyridine, rt, 99%; (c) NaBH3CN, TFA, 4 A MS, DMF,
˚
In conclusion, the b-(1!4)-S-linkage between the
GlcUA and the GlcN unit was successfully constructed
via substitution of a GlcUA 1-a-bromide with a GlcN
0 °C, 82%; (d) Tf2O, DMAP, pyridine, 0 °C; (e) KSAc, DMF, 0 °C,
87%; (f) NaOMe, MeOH, rt.

4340
H. Cao, B. Yu / Tetrahedron Letters 46 (2005) 4337–4340
˚
Thunberg, L.; Ba¨ckstro¨m, G.; Wasteson, A.; Robinson,
4-thiolate. The synthesis of a HS trisaccharide contain-
ing such a S-linkage (i.e., 1), as a nonhydrolyzable sub-
strate mimic of heparanase, was achieved. The present
synthesis represents the first synthesis of a HS/heparin
oligosaccharide with a S-glycosidic linkage, and shall
be informative for the preparation of the diverse cong-
eners of the HS/heparin oligosaccharide S-mimics. Syn-
thesis of longer congeners (of 1) and measurement of
their heparanase inhibition activities are undergoing
and will be reported in due course.
¨
H. C.; Ogren, S.; Lindahl, U. J. Biol. Chem. 1982, 257,
10278–10282.
5. (a) Fairweather, J. K.; Driguez, H. In Carbohydrates
in Chemistry and Biology; Ernst, B., Hart, G. W., Sinay,
¨
P., Eds.; Wiley-VCH: Weinheim, 2000; Vol. 1, pp 531–
564; (b) Driguez, H. Top. Curr. Chem. 1997, 187, 86–
116.
6. For reviews on HS/heparin synthesis: (a) Petitou, M.; van
Boeckel, C. A. A. Angew. Chem., Int. Ed. 2004, 43, 3118–
3133; (b) Poletti, L.; Lay, L. Eur. J. Org. Chem. 2003, 16,
2999–3024; (c) Karst, N. A.; Linhardt, R. J. Curr. Med.
Chem. 2003, 10, 1993–2031; (d) Tamura, J. Trends
Glycosci. Glycotechnol. 2001, 13, 65–88.
Acknowledgements
7. For an example, see: Florio, P.; Thomson, R. J.; von
Itzstein, M. Carbohydr. Res. 2000, 328, 445–448.
8. Orgueira, H. A.; Bartolozzi, A.; Schell, P.; Litjens, R.;
Palmacci, E. R.; Seeberger, P. H. Chem. Eur. J 2003, 9,
140–169, and references cited therein.
Financial support from the Committee of Science and
Technology of Shanghai (04DZ19213, 02QMA1401,
and 03DZ19203) is gratefully acknowledged.
´ ´
9. Leydet, A.; Jeantet-Segonds, C.; Barthelemy, P.; Boyer,
B.; Roque, J. P. Recl. Trav. Chim. Pays-Bas 1996, 115,
Supplementary data
421–426.
10. For examples, see: (a) Eisele, T.; Windmuller, R.; Schmidt,
¨
R. R. Carbohydr. Res. 1998, 306, 81–91; (b) Lu, P. P.;
Hindsgaul, O.; Li, H.; Palcic, M. Can. J. Chem. 1997, 75,
790–800; (c) Andrew, J. S.; Pinto, B. M. Carbohydr. Res.
1995, 270, 51–62.
Experimental procedures and spectroscopic data for
new compounds. Supplementary data associated with
this article can be found, in the online version, at
doi:10.1016/j.tetlet.2005.04.088.
11. Fischer, B.; Nudelman, A.; Ruse, M.; Herzig, J.; Gottlieb,
H. E.; Keinan, E. J. Org. Chem. 1984, 49, 4988–
4993.
References and notes
12. Nilsson, U.; Johansson, R.; Magnusson, G. Chem. Eur. J.
1996, 2, 295–302.
1. (a) Capila, I.; Linhardt, R. J. Angew. Chem., Int. Ed. 2002,
41, 390–412; (b) Conrad, H. E. Heparin-Binding Proteins;
Academic: San Diego, 1998.
13. Lin, F.; Peng, W. J.; Xu, W.; Han, X. W.; Yu, B.
Carbohydr. Res. 2004, 339, 1219–1223.
14. Horita, K.; Yoshioka, T.; Tanaka, T.; Oikawa, Y.;
Yonemitsu, O. Tetrahedron 1986, 42, 3021–3028.
15. (a) Greilich, U.; Brescello, R.; Jung, K.-H.; Schmidt, R. R.
Liebigs Ann. Chem. 1996, 663–672; (b) Baylay, H.;
Standring, D. N.; Knowles, J. R. Tetrahedron Lett. 1978,
19, 3633–3634.
2. For reviews, see: (a) Vlodavsky, I.; Goldshmidt, O.;
Zcharia, E.; Atzmon, R.; Rangini-Guatta, Z.; Elkin, M.;
Peretz, T.; Friedmann, Y. Seminars in Cancer Biology
2002, 12, 121–129; (b) Parish, C. R.; Freeman, C.; Hulett,
M. D. Biochim. Biophys. Acta 2001, 1471, M99–M108; (c)
Bame, K. J. Glycobiology 2001, 11, 91R–98R; (d) Demp-
sey, L. A.; Brunn, G. J.; Platt, J. L. Trends Biochem. Sci.
2000, 25, 349–351.
16. (a) Lemieux, R. U.; Ratcliffe, R. M. Can. J. Chem. 1979,
57, 1244–1251; (b) Jacquinet, J. C.; Sinay, P. Carbohydr.
Res. 1987, 159, 229–253.
¨
3. Okada, Y.; Yamada, S.; Toyoshima, M.; Dong, J.;
Nakajima, M.; Sugahara, K. J. Biol. Chem. 2002, 277,
42488–42495.
4. For earlier research on the substrate specificity of hepa-
ranase, see: (a) Pikas, D. S.; Li, J.-P.; Vlodavsky, I.;
Lindahl, U. J. Biol. Chem. 1998, 273, 18770–18777; (b)
17. Kloosterman, M.; Kuyl-Yeheskiely, E.; van Boom, J. H.
Recl. Trav. Chim. Pays-Bas 1985, 104, 291–295.
18. Brossmer, R.; Mack, H. Tetrahedron Lett. 1981, 22, 933–
936.
19. Kinzy, W.; Schmidt, R. R. Liebigs Ann. Chem. 1985,
1537–1545.