Scalable synthesis of L-lduronic acid derivatives via stereocontrolled cyanohydrin reaction for synthesis of heparin-related disaccharides

Article abstract of DOI:10.1021/ol901723m

L-ldo cyanohydrln 3 was prepared from dlacetone-D-glucose In four steps and 76% overall yield and 90% de via cyanohydrin reaction of aldehyde 2. This process can be scaled to provide >1 mol of pure L-Ido cyanohydrin 3. Cyanohydrin 3 was elaborated to 1,2-Isopropylldineprotected L-ldo nitrile (8), lduronic amide 9, and known carboxy ester 10. Coupling of 8 and 9 with glucosamine donors leads to new types (6-cyano and 6-carboxamlde) of heparin-related disaccharides.

Full text of DOI:10.1021/ol901723m

ORGANIC
LETTERS
2009
Vol. 11, No. 20
4528-4531
Scalable Synthesis of L-Iduronic Acid
Derivatives via Stereocontrolled
Cyanohydrin Reaction for Synthesis of
Heparin-Related Disaccharides
Steen Uldall Hansen,*,† Marek Bara´th,‡,† Bader A. B. Salameh,§,†
Robin G. Pritchard,† William T. Stimpson,† John M. Gardiner,*,† and
Gordon C. Jayson|
School of Chemistry and Manchester Interdisciplinary Biocentre, The UniVersity of
Manchester, 131 Princess Street, Manchester M1 7DN, U.K., and Cancer Research
U.K. and UniVersity of Manchester Department of Medical Oncology, Christie
Hospital, Wilmslow Road, Manchester M20 4BX, United Kingdom
gardiner@manchester.ac.uk.
Received July 27, 2009
ABSTRACT
L-Ido cyanohydrin 3 was prepared from diacetone-D-glucose in four steps and 76% overall yield and 90% de via cyanohydrin reaction of
aldehyde 2. This process can be scaled to provide >1 mol of pure L-ido cyanohydrin 3. Cyanohydrin 3 was elaborated to 1,2-isopropylidine-
protected L-ido nitrile (8), iduronic amide 9, and known carboxy ester 10. Coupling of 8 and 9 with glucosamine donors leads to new types
(6-cyano and 6-carboxamide) of heparin-related disaccharides.
Heparin/heparan sulfates (HS) are linear glycosaminogly-
can (GAG) polysaccharides, consisting of alternating
N-glucosamino and hexuronic acid residues. They play
critical roles through binding to a wide range of enzymes,
growth factors, and other proteins regulating diverse
biological effects, with sulfation at several possible sites
being central to biological recognition and effect.1 These
include fibroblast growth factors (FGFs). The complex
heterogeneity of natural HS oligosaccharide sources means
that defining details of the roles and effects of specific
HS sequences in relation to specific FGFs2 (of which more
than 20 are known) needs to be addressed by specific
synthesis of a diversity of structurally defined HS se-
quences. Access to pure natural sequences and mimetics
also offers prospects for providing designed sequences for
specific therapeutic applications.3 In natural HS sequences,
† The University of Manchester.
(1) (a) Bishop, J.; Schuksz, M; Esko, J. D Nature 2007, 446, 1030–
1037. (b) Tumova, S.; Woods, A.; Couchman, J. R. Int. J. Biochem. Cell
Biol. 2000, 32, 269–288. (c) Sasisekharan, R.; Shriver, Z.; Venkataraman,
G.; Narayanasami, U. Nat. ReV. Cancer 2002, 2, 521–528.
‡ Current address: Institute of Chemistry, Centre of Glycomics, Slovak
Academy of Sciences, Dubravska Cesta 9, 845 38 Bratislava, Slovak
Republic.
§ Current address: Chemistry Department, The Hashemite University,
P.O. Box 150459, Zarqa 13115, Jordan.
(2) (a) Pellegrini, L. Curr. Opin. Struct. Biol. 2001, 629–634. (b) Hung,
K. W.; Kurnar, T. K. S.; Kathir, K. M.; Xu, P.; Ni, F.; Ji, H. H. Biochemistry
2005, 44, 15787–15798. (c) Olsen, S. K.; Ibrahimi, O. A.; Raucci, A.; Zhang,
F. Proc. Natl. Acad. Sci. 2004, 101, 935–941.
| Cancer Research U.K. and University of Manchester Department of
Medical Oncology.
10.1021/ol901723m CCC: $40.75
Published on Web 09/18/2009
 2009 American Chemical Society

the uronic acid component is either D-glucuronic or
L-iduronic acid, which differ only in their C5-configura-
tion. Various biological studies indicate that sequences
containing domains of L-iduronic components modulate
effects on a number of FGFs. Synthetic access to disac-
charide components of such oligosaccharides is thus of
considerable interest, as they are key precursors to the
synthesis of longer H/HS-like sequences.
Of the uronic acid components, L-iduronic acid provides
a key challenge to effective synthesis, as this sugar is not
readily available. Thus, efficient synthesis of L-iduronic acid
derivatives is critical to provision of structurally defined
saccharides for evaluation, and scalability is important for
any future therapeutic potential.
A number of previous syntheses of L-iduronic acid and
derivatives have been described,4 involving obtention of the
C5 L-configuration by diastereomer separation4a-c or inver-
sion of the C5 stereochemistry of D-sugar precursors,4d-i in
some cases with subsequent oxidation of intermediate L-idose
derivatives.4j-l Low temperature addition of tris(phenylthi-
o)methyllithium (a carboxylate surrogate)5 to 2 has also
provided stereoselective introduction of the C5 stereocenter.6
Oxidation of L-iditose components in oligosaccharides has
also been employed.7
tivity in most cases, though its utility has been applied to
synthesis of isotopically labeled sugars.8b
The diastereoselectivity of the cyanohydrin reaction of 2
has previously been reported to be poor (typically a 1:1
mixture).9 While addition of appropriate dialkylhydrazone
reagents has provided a stereoselective entry to the D-gluco
cyanohydrin (4), in two steps from 2,10 there is no prior
stereoselective access to 3. We thus undertook an reinves-
tigation of the conditions for this cyanohydrin reaction of
aldehyde 2, with a view to developing a more diastereose-
lective and scalable process.
Scheme 1. Stereocontrolled Cyanohydrin Synthesis
We report here our approach to stereocontrolled introduc-
tion of the key C5 L-ido strereochemistry, through reaction
at ambient temperature, under air and without the need for
anhydrous conditions, using inexpensive reagents. The key
stereochemical step involves purification by crystallization
and is thus designed to enable large-scale synthesis, facilitat-
ing access to larger amounts of various L-iduronate targets.
Scalability underpins the potential for viable development
of L-ido-containing heparan-related saccharides for clinical
evaluations.
We found the cyanohydrin reaction to be fast (complete
in <30 min) with or without additives, but that diastereo-
control, as previously reported,9 is very low at short reaction
times. However, following the reaction over longer periods
showed that, in reactions with added magnesium chloride,
the de increases11 and after 5 days affords 90% de in favor
of the L-ido configuration (Scheme 1 and Table 1a). During
this reaction increasing precipitation was observed. This
outcome is rationalized as a combination of cyanohydrin
epimer equilibration (slightly basic reaction conditions) and
the preferential crystallization of the L-ido configuration
diastereomer 3 from the reaction solution,12 driving equi-
librium toward the L-ido product. Pure 3 is then isolated in
high yields after recrystallization.
The cyanohydrin reaction has been used in carbohydrate
homologation chemistry8 (originally reported by Kiliani8a)
and, while high yielding, proceeds with low diastereoselec-
(3) Cole, C.; Jayson, G. C. Expert Opin. Biol. Ther. 2008, 8, 351–362.
(4) (a) Rochepeau-Jobron, L; Jacquinet, J.-C. Carbohydr. Res. 1997,
303, 395–406. (b) Hinou, H.; Kuorsawa, H.; Matsuoka, K.; Terunuma, D.;
Kuzuhara, H. Tetrahedron Lett. 1999, 40, 1501–1504. (c) Schell, P.;
Orgueira, H. A.; Roehrig, S.; Seeberger, P. H. Tetrahedron Lett. 2001, 42,
3811–3814. (d) Orgueira, H. A.; Bartolozzi, A.; Schell, P.; Litjens,
R. E. J. N.; Palmacci, E. R.; Seeberger, P. H. Chem.sEur. J. 2003, 9, 140–
169. (e) Ke, W.; Whitfield, D. M.; Gill, M.; Larocque, S.; Yu, S. H.
Tetrahedron Lett. 2003, 44, 7767–7770. (f) Ojeda, R.; de Paz, J. L.; Mart´ın-
Lomas, M.; Lassaletta, J. M. Synlett 1999, 1316–1318. (g) Jacquinet, J-C;
Petitoi, M; Duchaussoy, P.; Lederman, I.; Choay, J.; Torri, G.; Sinay, P.
Carbohydr. Res. 1984, 130, 221–241. (h) Yu, H. N.; Furukawa, J.-i.; Ikeda,
T.; Wong, C. H. Orglett 2004, 6, 723–726. Via C6 bromo: (i) Blanc-
Muesser, M.; Defaye, J.; Horton, D.; Tsai, J.-H. Methods Carbohydr. Chem.
1980, 8, 177–183. (j) Tatai, J.; Osztrovszky, G.; Kajta´r-Peredy, M.; Fu¨gedi,
P. Carbohydr. Res. 2007, 343, 596–606. (k) Codee, J. D. C.; Stubba, B.;
Schiattarella, M.; Overkleeft, H. S.; van Boeckel, C. A.; van Boom, J. H.;
van der Marel, G. A. J. Am. Chem. Soc. 2005, 127, 3767–3773. (l) Davis,
N. J.; Flitsch, S. L. Tetrahedron Lett. 1993, 34, 1181–1184.
Analysis of the effect of substrate concentration on the
diastereoselectivity (using 1.1 equiv of MgCl2 and KCN, 16 h
time point, where optimum de ) 82%), illustrates that the
de is also related to a critical concentration limit, below
which the de drops markedly (Table 1b).
(9) (a) Wolfrom, M. L.; Thomas, G. H. S. Methods Carbohydr.
Chem. 1963, 2, 32–4. (b) Shafizadeh, F.; Wolfrom, M. L. J. Am. Chem.
Soc. 1955, 77, 2568–2569. (c) Hudson, C. S. J. Am. Chem. Soc. 1951, 73,
4498–4499. Idogluco system gave 1:1 ratio under all conditions: (d) Rauter,
A.; Figueiredo, J.; Ismael, I.; Pais, M. J. Carbohydr. Chem. 1987, 6, 259–
272.
(5) Seebach, D. Angew. Chem., Int. Ed. Engl. 1967, 6, 442–443.
(6) (a) Lohman, G. J. S.; Hunt, D. K.; Hogermeier, J. A; Seeberger, P.
J. Org. Chem. 2003, 68, 7559–7561. (b) Gavard, O.; Hersant, Y. I.; Alais,
J.; Duverger, V.; Dilhas, A.; Bascou, A.; Bonnaffe´, D. Eur. J. Org. Chem.
2003, 3603–3620. (c) Lubineau, A.; Gavard, O.; Alais, J.; Bonnaffe´, D.
Tetrahedron Lett. 2000, 41, 307–311.
(10) (a) Ferna´ndez, R.; Mart´ın-Zamora, E.; Pareja, C.; Lassaletta, J. M.
J. Org. Chem. 2001, 66, 5201–5207. (b) Ferna´ndez, R.; Mart´ın-Zamora,
E.; Pareja, C.; Lassaletta, J. M. Tetrahedron Lett. 1996, 375787-
5790.
(7) Polat, T.; Wong, C.-H. J. Am. Chem. Soc. 2007, 127, 12795–12800.
(8) (a) Kiliani, H. Ber. 1887, 20, 339. (b) Synder, J. R.; Serianni, A. S.
Carbohydr. Res. 1987, 163, 169–188. (c) Synder, J. R.; Serianni, A. S. J.
Org. Chem. 1986, 51, 2694–2702.
(11) A graphical representation is available in the Supporting Informa-
tion.
(12) Solubilities in 1:1 MeOH/H2O: 3, 0.9 mg/mL, and 4, 1.9 mg/
mL.
Org. Lett., Vol. 11, No. 20, 2009
4529

Table 1. Effects of (a) Reaction Time and (b) Substrate
Concentration on Diastereoselectivity of Cyanohydrin Reaction
of 2a
(a) reaction timea
(b) substrate concentrationb,c
entry time/h yield/% de
C/M of 2
yield/%
de
1
2
3
4
5
6
7
0.3
1
2
4
16
97
94
94
96
94
94
93
9
28
69
74
82
87
90
0.5
93
94
91
96
83
82
75
36
0.27
0.17
0.06
48
120
a All reactions using 1.1 equiv of KCN and MgCl2. b 0.27 M of 2. c Time
) 16 h.
Figure 1. X-ray structure of L-ido configuration cyanohy-
drin 3.13
To determine the scope of additives in synthesis of 3,
evaluation of other metal salts and alternative magnesium
salts (MgSO4 and MgBr2) was undertaken (Table 2). Several
magnesium salts gave high diastereoselectivity, but magne-
sium chloride had the advantage of a superior overall yield,
with the optimum combination of isolated yield and de
obtained using 1.1 equiv of MgCl2.
converted into the 1,2-isopropylidine-protected pyranose 8
using a modified version of methodology for the analogous
carboxy ester system.6a Thus, removal of the isopropylidene
from 3 by treatment with aqueous TFA afforded deprotected
crystalline cyano sugar 5 in 84% yield (Scheme 2). The 1,2-
isopropylidene was then reformed using excess 2-methox-
ypropene and CSA, conditions leading preferentially to the
1,2-isoproylidine-protected pyranoside ring system. The
primary product was in fact 7, which contains an open acetal
attached to the 4-position. Product 7 was isolated in 62%
yield, along with 18% of the desired O-4-deprotected
pyranoside 8. This reaction proceeds via tris-acetal-protected
intermediate 6,14 which could be isolated when the reaction
was run for 2 h at 0 °C.
Table 2. Effects of Metal Salt on Diastereoselectivity of
Cyanohydrin Reaction of 2a
metal salt
equiv
yield (%)
de
none
n/a
0.55
1.1
1.1
1.1
64
80
94
77
76
33
82
87
82
83
MgCl2
MgCl2
MgBr2
MgSO4
Scheme 2
.
Conversion of L-Ido Cyanohydrin 3 into L-Ido
Cyanopyranoside 8a
a Reaction time ) 48 h.
The L-ido configuration of 3 was unambiguously estab-
lished by an X-ray structure determination (Figure 1).
The overall yield from diacetone-D-glucose (1a) to pure
L-ido cyanohydrin 3 was 76% (four steps). Moreover, the
residue in solution after recrystallization (a mixture of
L-ido and D-gluco cyanohydrins (3 and 4)) could be re-
equilibrated by stirring in 1:1 methanol-water with 0.1
equiv of potassium cyanide and magnesium chloride for
5 days, re-establishing 90% de of L-ido over D-gluco. The
mild conditions allied with purification of 3 by crystal-
lization enabled scale up to >1 mol scale. This offers
significant process cost advantages over prior homologa-
tions of 2.
a Note: O-4 deprotection of 7 also leads to 12% of starting furanoside
3.
L-Ido cyanohydrin 3 was elaborated to several different
L-iduronic acid derivatives, as acceptors for glycosidic
couplings. The 5-membered ring cyanohydrin 3 was first
The O-4 acetal was selectively removed to convert 7 into
8, using 0.1 equiv of p-TSA as under these conditions 8 is
(13) Data deposited with the Cambridge Crystallographic Database and
allocated the deposition no. CCDC 736690.
(14) NMR data indicate 6 is 4C1: JH2-H3 ) 9.8 Hz for 6, while H-4 in
7, for example, shows only J < 5 Hz.
4530
Org. Lett., Vol. 11, No. 20, 2009

only slowly converted back to the furanose 3, maximizing
overall conversion of 5 to 8.15
heparin disaccharides 13 and 15-20. Yields for glyco-
sylations of 8-10 using 4,6-O-benzylidine protected donor
12 [R ) R1 ) CHPh] were very similar to those for donors
having preinstalled differentiations of O-4/O-6. However,
anomeric selectivities proved lower, with anomeric ratios
(R/ꢀ) for 16 (7:2), 18 (3:2), and 20 (5:1).
The nitrile 8 was converted into the (crystalline) L-ido
amide 9 in excellent yield using hydrogen peroxide and
potassium carbonate (Scheme 3). This primary amide was
treated with dimethylformamide dimethyl acetal in dry
methanol16 to facilitate the direct conversion to the known
methyl ester 10 in 81% yield (with concomitant formation
of 9% of the furanoside 11, separable from 10.)
Scheme 4. L-Ido Disaccharide Syntheses
Scheme 3
.
Conversion of Idonitrile 8 to Iduronamide and
Iduronic Ester Derivatives
The ester 10 is known to be an excellent acceptor for
the synthesis of heparin-like disaccharides.4d,k,17 In our
hands, acceptor 10 was thus employed for synthesis of
heparin-related disaccharide units 13-16, providing good
yields and selectivities against variations in nature of the O-4
and O-6 protecting groups of donor 12,4d,18 thereby affording
useful intermediates with different alternatives to diverge to
disaccharides for further homologation (Scheme 4).
The utility of 6-cyano- or 6-carboxamide-bearing ac-
ceptors (in general) is essentially unexplored in oligosac-
charide synthesis. (Both cyano and amido L-ido interme-
diates 8 and 9 are novel.) We thus decided to evaluate
coupling of these acceptors with the 2-azido gluco
thioglycoside donors 12, which had, as anticipated,
performed well with the 6-carbomethoxy ido acceptor 10.
Thioglycoside donors 12 coupled with nitrile 8 to provide
a high yield of the R-linked disaccharides 17 and 18. The
amide 9 proved to be a poorer acceptor, affording amide-
containing R-linked disaccharides 19 and 20 but isolated
in lower yields.19 Notwithstanding lower yields using
amide 9, this establishes a route to novel 6-modified
In conclusion, a new scalable approach to the estab-
lished valuable L-iduronic acid building block (10) for
heparan sulfate syntheses has been developed. The route
to the key precursor L-ido cyanohydrin 3 is highly
stereoselective and efficient and offers significant advan-
tages on a larger scale as it involves no column chroma-
tography, proceeds at ambient temperature, has no re-
quirement for anhydrous conditions, and utilizes very
inexpensive reagents affording a highly crystalline product
in a single step. In addition to conversion of cyanohydrin
3 to a known L-iduronic ester acceptor derivative 10, novel
nitrile and primary amide L-iduronic intermediates 8 and
9 were also prepared and evaluated as acceptors in
glycoside couplings, yielding novel 6-cyano and 6-car-
boxamide L-ido disaccharide analogues. Further elabora-
tions of 3 into various heparan-sulfate related oligosac-
charides, and the uses of novel disaccharides, will be
reported elsewhere.
(15) Increasing acid concentration in the first step (to effect in situ
removal of the O-4 acetal of 7) was ineffective.
Acknowledgment. The Holt Foundation, CRUK (C2075/
A9106 and C2075/A5048), and MRC (G0601746) are thanked
for project funding, and EPSRC is thanked for NMR instru-
mentation (GR/L52246).
(16) Anelli, P.; Brocchetta, M.; Palano, D.; Visigalli, M. Tetrahedron
Lett. 1997, 38, 2367–2368.
(17) (a) Lohman, G.; Seeberger, P. J. Org. Chem. 2004, 69, 4081–4093.
(b) Orgueira, H.; Bartolozzi, A.; Schell, P.; Seeberger, P. Angew. Chem.,
Int. Ed. 2002, 41, 2128–2131
.
(18) 12 (R ) PMB, R1 ) Ac): (a) Karst, N.; Islam, T.; Avci, F.; Linhardt,
R. Tetrahedron Lett. 2004, 45, 6433–6437 12 (R ) R1 ) CHPh): (b) Martin-
Lomas, M.; Flores-Mosquera, M.; Chiara, J. Eur. J. Org. Chem. 2000, 154,
7–1562. Codee, J. D. C.; Litjens, R.; den Heeten, R.; Overkleeft, H.; van
Boom, J. H.; van der Marel, G. A Orglett 2003, 1519–1522 12 (R ) PMB,
R1 ) Bn): (c) Lindberg, J.; Ohberg, L.; Garegg, P. J.; Konradsson, P.
Supporting Information Available: Experimental and
copies of spectral data for compounds 3, 7-10, 13, and
15-20. This material is available free of charge via the
Internet at http://pubs.acs.org.
Tetrahedron Lett. 2002, 58, 1387–1398
(19) The same reactivity trends were evident using corresponding
trichloroacetimidate 2-azido gluco donors.
.
OL901723M
Org. Lett., Vol. 11, No. 20, 2009
4531