Synthesis of the heparin-based anticoagulant drug fondaparinux

Article abstract of DOI:10.1002/anie.201404154

Fondaparinux, a synthetic pentasaccharide based on the heparin antithrombin-binding domain, is an approved clinical anticoagulant. Although it is a better and safer alternative to pharmaceutical heparins in many cases, its high cost, which results from th

Full text of DOI:10.1002/anie.201404154

Angewandte
Chemie
DOI: 10.1002/anie.201404154
Carbohydrates
Synthesis of the Heparin-Based Anticoagulant Drug Fondaparinux**
Cheng-Hsiu Chang, Larry S. Lico, Teng-Yi Huang, Shu-Yi Lin, Chi-Liang Chang,
Susan D. Arco, and Shang-Cheng Hung*
Abstract: Fondaparinux, a synthetic pentasaccharide based on
the heparin antithrombin-binding domain, is an approved
clinical anticoagulant. Although it is a better and safer
alternative to pharmaceutical heparins in many cases, its high
cost, which results from the difficult and tedious synthesis, is
a deterrent for its widespread use. The chemical synthesis of
fondaparinux was achieved in an efficient and concise manner
from commercially available d-glucosamine, diacetone a-d-
glucose, and penta-O-acetyl-d-glucose. The method involves
suitably functionalized building blocks that are readily acces-
sible and employs shared intermediates and a series of one-pot
reactions that considerably reduce the synthetic effort and
improve the yield.
low-molecular-weight variants.[6] However, nature-sourced
heparins are preferred over fondaparinux because of their
lower price.
Fondaparinux synthesis is made very demanding by the
regio- and stereochemical aspects of the assembly and the
strategic placement of multiple sulfonate groups. The estab-
lished route involves approximately 55 steps, which drastically
decrease the overall yield.[5,7] a-Glucosaminylation, in partic-
ular, mainly relies on the anomeric effect, which is insufficient
in curbing production of the unwanted b isomer. The 1,2-trans
glycosylation involving the uronic acid precursors, although
feasible through neighboring group participation, is con-
founded by the different C2 functionalizations in the final
product. Hence, as often is the case in general heparin
syntheses,[8] protecting-group selection is critical. Recent
efforts to improve fondaparinux synthesis have only showed
limited success.[9]
To provide an efficient and concise access to fondapar-
inux, we conceived a route through the building blocks 5 or 6,
7, and 8 (Scheme 1), all readily attainable from commercially
available starting materials. The approach followed the
formation of the tetrasaccharide donor 3 and its subsequent
coupling with the reducing end acceptor 4 to generate the
fully protected precursor (2) of the target compound 1.
Shared intermediates and one-pot reactions are used through-
out the process to reduce the steps and minimize wasteful
purification stages. The synthetic design exploits our estab-
lished methods, especially regioselective one-pot multipro-
tection,[10] highly stereoselective a-glucosaminylation influ-
enced by the orthogonal protecting groups in 8 (N3, PBB, 2-
NAP, and TBDPS),[11] and late-stage oxidation[11a,c,12] to avoid
the low reactivity and base sensitivity associated with uronate
esters. The glucose derivatives 5 and 6 can be prepared in one
pot from the tetrasilylated compound 10, which can be
obtained from penta-O-acetyl-d-glucose in two steps.[10] The
anhydro-l-idose 7 can be acquired from diacetone d-glucose
(10) in five steps with 37% overall yield.[12a,13] Lev and Bz
groups were selected to facilitate neighboring-group partic-
ipation and allow the introduction of the crucial sulfate
groups at O2 and O3 of the respective IdoA and GlcN units,
while ensuring that the O2 position of GlcAwould be free. Ac
groups were used to protect the primary alcohols destined for
oxidation.
H
eparin is a polysulfated polysaccharide with alternating d-
glucosamine (GlcN) and either d-glucuronic acid (GlcA) or
l-iduronic acid (IdoA) units. Within its microheterogeneous
chain is a particular pentasaccharide motif that binds and
activates antithrombin, an inhibitor of the blood coagulation
cascade.[1] This property forms the basis for the clinical
potency of heparin in the treatment of thromboembolic
disorders.[2] Nevertheless, active monitoring is needed during
heparin therapy because serious complications such as
heparin-induced thrombocytopenia, uncontrolled bleeding,
and osteoporosis may occur.[3] Moreover, the safety and
quality of the heparin supply chain was called into question
when a recent contamination with oversulfated chondroitin
sulfate turned fatal.[4] The quest for precisely targeted anti-
coagulant activity that avoids the problems associated with
nature-sourced heparins led to the development of fondapar-
inux (1), a synthetic pentasaccharide derived from the
aforementioned antithrombin-binding sequence.[5] Fondapar-
inux is safer and displays comparable to superior efficacy and
pharmacological properties to unfractionated heparin and its
[*] Dr. C.-H. Chang, L. S. Lico, Dr. T.-Y. Huang, Dr. S.-Y. Lin,
Dr. C.-L. Chang, Prof. Dr. S.-C. Hung
Genomics Research Center, Academia Sinica
No. 128 Academia Road, Section 2, Taipei 115 (Taiwan)
E-mail: schung@gate.sinica.edu.tw
Dr. C.-H. Chang
Department of Chemistry, National Tsing Hua University
No. 101, Section 2, Kuang-Fu Road, Hsinchu 300 (Taiwan)
L. S. Lico, Prof. Dr. S. D. Arco
Institute of Chemistry, University of the Philippines
Diliman, Quezon City 1101 (Philippines)
The GlcN building blocks 4 and 8 were prepared from d-
glucosamine·HCl (11; Scheme 2). Conversion of the amino
group into an azide, followed by peracetylation yielded
compound 12. Next, a BF3-assisted one-pot thioglycosylation
and deacetylation supplied the triol 13. A new extended one-
pot procedure starting from 13 that includes per-O-silylation
with HMDS,[14] 4,6-O-napthylmethylidene formation, and
regioselective ring opening installed the 2-NAP group at
[**] This work was supported by the Ministry of Science and Technology
(NSC 100-2113-M-001-019-MY3 and NSC 101-2628-M-001-006-
MY3), National Health Research Institutes (NHRI-EX101-10146NI),
and Academia Sinica.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201403154.
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
with methanol and 2-NAP cleavage with DDQ in one pot. For
this reaction, the b product was also produced in 15% yield.
Evidently, tetrasaccharide 3 could be produced in a [2 +
2]-manner in a reaction requiring a dilevulinylated disacchar-
ide acceptor (18) and a benzoylated disaccharide donor (21).
The acceptor can be prepared from monosaccharides 7 and 8
through a known diol intermediate (17), which we have
previously accessed[11a] by using a palladium- and tin-cata-
lyzed reaction. This time, we decided to avoid these metal
catalysts by following a simpler approach (Scheme 3). When
Scheme 3. Synthesis of the disaccharide acceptor 18. Reagents and
conditions: a) NIS, TfOH, CH2Cl2, À60 to À408C, 3 h, 70%;
b) NaOMe, MeOH, 92%; c) LevOH, EDC, DMAP, CH2Cl2, 08C, 16 h;
DDQ, H2O, RT, 4 h, 70%. EDC=1-ethyl-3-(3-dimethylaminopropyl)car-
bodiimide.
Scheme 1. A retrosynthetic analysis of fondaparinux. 2-NAP=2-naph-
thylmethyl, Ac=acetyl, Bn=benzyl, Bz=benzoyl, Lev=levulinyl,
PBB=p-bromobenzyl, TBDPS=tert-butyldiphenylsilyl, TMS=trimethyl-
silyl, Tol=4-tolyl.
using the thioglycoside 15 instead of 8, the glycosylation of
idose 7 (1.5 equiv) led to the a disaccharide 16 (J1’,2’ = 3.5 Hz)
in 70% yield, together with its easily separable b isomer
(10%). Self-condensation of the donor was not observed.
Zemplꢀn conditions lead to the removal of the Bz group to
afford the diol 17, which underwent one-pot levulinylation
and denaphthylmethylation to furnish compound 18.
The construction of compound 3 was initially planned to
include the glucosyl acceptor 5, but this proved troublesome
upon the condensation of 18 with donor 21[11c] (Scheme 4).
Glycosylation efficiency decreased, presumably as a result of
the electron-withdrawing esters in both compounds. To
circumvent this problem, we temporally used Bn in place of
Ac in the glucosyl building block (20[10a]). Fortunately, the
coupling of 22 and 18 effectively produced the tetrasaccharide
23. Further exposure to Ac2O and TMSOTf led to the
concurrent primary Bn-to-Ac exchange[15] and anhydro-ring
acetolysis. Although the triacetate 24 was ultimately
obtained, the [2+2]-coupling required 3 equiv of the donor
22 to make up for the low reactivity of the acceptor 18. The
selective conversion of the 6-O-Bn group also required a large
excess of TMSOTf (10 equiv) in a reaction that must be
maintained at À408C to avoid side products. These issues
prompted us to look for an alternative approach to construct
24 and eventually reach 3.
Scheme 2. Preparation of the glucosamine building blocks. Reagents
and conditions: a) 1) TfN3, CuSO4, Et3N, MeCN, MeOH, 08C to RT,
2 d; 2) Ac2O, DMAP, 08C to RT, 16 h, 84%; b) p-thiocresol, BF3·Et2O,
RT, 8 h; MeOH, reflux, 16 h, 78% (a/b=3.5/1); c) HMDS, TMSOTf,
CH2Cl2, 08C, 1 h; 2-NAPCHO, TMSOTf, 08C, 2 h; BH3·THF, TMSOTf,
08C, 8 h; TBAF, RT, 16 h, 88%; d) TBDPSCl, imidazole, DMAP, CH2Cl2,
91%; e) NaH, PBBBr, DMF, 95%; f) MeOH, NIS, TfOH, Et2O, À60 to
À408C, 3 h; DDQ, MeCN, RT, 2 d, 57%. DDQ=2,3-dichloro-5,6-
dicyano-1,4-benzoquinone, DMAP=4-N,N-dimethylaminopyridine,
DMF=dimethylformamide, HMDS=hexamethyldisilazane, NIS=N-
iodosuccinimide, TBAF=tetrabutylammonium fluoride, Tf=trifluoro-
methanesulfonyl.
This time the assembly of tetrasaccharide 3 was carried
out through sequential monosaccharide addition from the
reducing to the nonreducing end (Scheme 5). With the
glucosyl donor 6 (1.3 equiv), the stereoselective extension of
18 worked successfully. The addition of aqueous TFA to the
same reaction vessel resulted in the hydrolysis of the
O4.[10] The sequential protection of the diol 14 with TBDPS
and PBB groups gave the donor 8. Finally, the a-methyl
glycoside 4 (57%, J1,2 = 3.2 Hz) was prepared by glycosylation
2
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
[11a]
building blocks). Treatment with TMSSTol and ZnI2
directly to the desired thioglycoside 3. NMR data showing
1’’’,2’’’ = 3.8 Hz and the W-coupling between 1-H and 3-H[11c]
led
J
(see the Supporting Information) confirmed the stereochem-
istry of the newly formed glycosidic bonds. The coupling of
this donor with the acceptor 4 (1.2 equiv) furnished the
pentasaccharide 2, the structure of which was supported by
NMR analysis as above.
Functional-group transformations were then carried out.
The Ac groups in 2 were selectively removed by using
Mg(OMe)2 to get the diol 27. TEMPO oxidation of the freed
alcohol functionalities, followed by hydrazine-mediated
cleavage of the Lev groups furnished the diol 28. Further
methyl ester formation and desilylation delivered the pentaol
29. The exposure of 29 to SO3·Et3N led to the corresponding
pentasulfate 30, which was saponified to afford compound 31.
Hydrogenolysis cleaved all of the remaining protecting
groups and simultaneously reduced the azide to the amine.
Final N-sulfonation supplied the crude material, which was
passed through size-exclusion and ion-exchange columns to
isolate fondaparinux (1) as a sodium salt. NMR and mass
spectrometry analyses and HPLC comparison with the
commercial product confirmed the structure of 1 (see the
Supporting Information).
Scheme 4. Synthesis of the tetrasaccharide 24. Reagents and condi-
tions: a) 1) NIS, acetone, H2O, 08C, 2 h, 81%; 2) K2CO3, CCl3CN,
CH2Cl2, RT, 16 h, 95%; b) AgOTf, CH2Cl2, À40 to À208C, 4 h, 21: 70%,
22: 72%; c) NIS, TfOH, CH2Cl2, À40 to À208C, 4 h, 71% (when using
22); d) TMSOTf, Ac2O, À408C, 4 h, 61%.
benzylidene ring to produce the diol 25 (J1’’,2’’ = 8.5 Hz).
Regioselective acetylation at the primary alcohol delivered
the trisaccharide acceptor 26. The stereoselective a-glucosa-
minylation of 26 with thioglycoside 8 (2 equiv) and the
acetolysis of the anhydro ring were accomplished in one pot
to furnish compound 24. This new route towards 24 is more
straightforward and gave better overall yield compared to our
initial procedure (13.9% versus 10.8% from the component
In summary, we synthesized fondaparinux in 32 collective
steps from commercially available starting materials and in
0.63% overall yield over 22 linear steps from d-glucos-
amine·HCl. The approach benefitted from carefully selected
orthogonal protecting groups and readily accessible mono-
saccharide building blocks. The use of common intermediates
and one-pot reactions effectively reduced the synthetic and
purification steps. This method represents an improved route
Scheme 5. Synthesis of fondaparinux. Reagents and conditions: a) NIS, TfOH, CH2Cl2, À60 to À408C, 3 h; TFA, H2O, RT, 1 h, 58%; b) Ac2O,
Et3N, CH2Cl2, 08C to RT, 16 h, 82%; c) NIS, TfOH, CH2Cl2, À788C to À208C, 4 h; TMSOTf, Ac2O, À208C, 8 h, 65%; d) TMSSTol, ZnI2, CH2Cl2, RT,
1 h, 85%; e) NIS, TfOH, CH2Cl2, À208C to 08C, 4 h, 61%; f) Mg(OMe)2, MeOH, CH2Cl2, RT, 16 h, 83%; g) 1) TEMPO, BAIB, CH2Cl2, H2O, RT,
24 h; 2) NH2NH2, Pyr, AcOH, 08C, 2 h; h) 1) CH2N2, Et2O, CH2Cl2, RT, 16 h; 2) HF·Pyr, Pyr, RT, 3 d, 60% from 27; i) SO3·Et3N, DMF, 608C, 73%;
j) LiOH, H2O2, THF, 378C, 3 d, 71%; k) 1) Pd(OH)2, H2, phosphate buffer (pH 7), MeOH, RT, 2 d, 80%; 2) SO3·Pyr, NaOH, H2O, RT, 2 d, 81%.
BAIB=bis(acetoxy)iodobenzene, TEMPO=2,2,6,6-tetramethyl-1-piperidinyloxyl free radical, Pyr=pyridine.
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
Venot, J. E. Turnbull, G.-J. Boons, J. Am. Chem. Soc. 2009, 131,
17394 – 17405; d) Z. Wang, Y. Xu, B. Yang, G. Tiruchinapally, B.
Sun, R. Liu, S. Dulaney, J. Liu, X. Huang, Chem. Eur. J. 2010, 16,
8365 – 8375; e) S. U. Hansen, G. J. Miller, C. Cole, G. Rushton, E.
Avizienyte, G. C. Jayson, J. M. Gardiner, Nat. Commun. 2013, 4,
2016; f) M. M. L. Zulueta, S.-Y. Lin, S.-C. Hung, Trends Glyco-
sci. Glycotechnol. 2013, 25, 141 – 158; g) M. M. L. Zulueta, S.-Y.
Lin, Y.-P. Hu, S.-C. Hung, Curr. Opin. Chem. Biol. 2013, 17,
1023 – 1029.
that may prove a useful basis for producing fondaparinux for
pharmaceutical use.
Received: April 10, 2014
Published online: && &&, &&&&
Keywords: anticoagulant · carbohydrates · heparin ·
.
one-pot reactions · total synthesis
[9] a) Y. Xu, S. Masuko, M. Takieddin, H. Xu, R. Liu, J. Jing, S. A.
Mousa, R. J. Linhardt, J. Liu, Science 2011, 334, 498 – 501; b) F.
Lin, G. Lian, Y. Zhou, Carbohydr. Res. 2013, 371, 32 – 39; c) R.
Kovi, A. Naik, B. Patel, M. Madala (Apicore), WO 2013/003001
A1, 2013; d) P. P. Patel, C. Ma, K. K. Ohrr, S. Nadji (Reliable
Pharmaceutical), WO 2013/115817 A1, 2013; e) T. Li, H. Ye, X.
Cao, J. Wang, Y. Liu, L. Zhou, Q. Liu, W. Wang, J. Shen, W.
Zhao, P. Wang, ChemMedChem 2014, 9, 1071 – 1080.
[10] a) C.-C. Wang, J.-C. Lee, S.-Y. Luo, S. S. Kulkarni, Y.-W. Huang,
C.-C. Lee, K.-L. Chang, S.-C. Hung, Nature 2007, 446, 896 – 899;
b) C.-C. Wang, S. S. Kulkarni, J.-C. Lee, S.-Y. Luo, S.-C. Hung,
Nat. Protoc. 2008, 3, 97 – 113; c) T.-Y. Huang, M. M. L. Zulueta,
S.-C. Hung, Org. Lett. 2011, 13, 1506 – 1509; d) T.-Y. Huang,
M. M. L. Zulueta, S.-C. Hung, Org. Biomol. Chem. 2014, 12,
376 – 382.
[11] a) Y.-P. Hu, S.-Y. Lin, C.-Y. Huang, M. M. L. Zulueta, J.-Y. Liu,
W. Chang, S.-C. Hung, Nat. Chem. 2011, 3, 557 – 563; b) C.-C.
Chung, M. M. L. Zulueta, L. T. Padiyar, S.-C. Hung, Org. Lett.
2011, 13, 5496 – 5499; c) M. M. L. Zulueta, S.-Y. Lin, Y.-T. Lin,
C.-J. Huang, C.-C. Wang, C.-C. Ku, Z. Shi, C.-L. Chyan, D. Irene,
L.-H. Lim, T.-I. Tsai, Y.-P. Hu, S. D. Arco, C.-H. Wong, S.-C.
Hung, J. Am. Chem. Soc. 2012, 134, 8988 – 8995.
[12] a) S.-C. Hung, X.-A. Lu, J.-C. Lee, M. D.-T. Chang, S.-l. Fang, T.-
c. Fan, M. M. L. Zulueta, Y.-Q. Zhong, Org. Biomol. Chem.
2012, 10, 760 – 772; b) Y.-P. Hu, Y.-Q. Zhong, Z.-G. Chen, C.-Y.
Chen, Z. Shi, M. M. L. Zulueta, C.-C. Ku, P.-Y. Lee, C.-C. Wang,
S.-C. Hung, J. Am. Chem. Soc. 2012, 134, 20722 – 20727.
[13] L.-D. Lu, C.-R. Shie, S. S. Kulkarni, G.-R. Pan, X.-A. Lu, S.-C.
Hung, Org. Lett. 2006, 8, 5995 – 5998.
[1] a) I. Capila, R. J. Linhardt, Angew. Chem. 2002, 114, 426 – 450;
Angew. Chem. Int. Ed. 2002, 41, 390 – 412; b) R. J. Linhardt, J.
Med. Chem. 2003, 46, 2551 – 2564.
[2] a) C. A. A. van Boeckel, M. Petitou, Angew. Chem. 1993, 105,
1741 – 1761; Angew. Chem. Int. Ed. Engl. 1993, 32, 1671 – 1690;
b) M. Petitou, B. Casu, U. Lindahl, Biochimie 2003, 85, 83 – 89.
[3] J. I. Weitz, L.-A. Linkins, Expert Opin. Invest. Drugs 2007, 16,
271 – 282.
[4] D. B. Blossom, A. J. Kallen, P. R. Patel, A. Elward, L. Robinson,
G. Gao, R. Langer, K. M. Perkins, J. L. Jaeger, K. M. Kurkjian,
M. Jones, S. F. Schillie, N. Shehab, D. Ketterer, G. Venkatara-
man, T. K. Kishimoto, Z. Shriver, A. W. McMahon, K. F. Austen,
S. Kozlowski, A. Srinivasan, G. Turabelidze, C. V. Gould, M. J.
Arduino, R. Sasisekharan, N. Engl. J. Med. 2008, 359, 2674 –
2684.
[5] M. Petitou, C. A. A. van Boeckel, Angew. Chem. 2004, 116,
3180 – 3196; Angew. Chem. Int. Ed. 2004, 43, 3118 – 3133.
[6] a) J. M. Walenga, W. P. Jeske, F. X. Frapaise, R. L. Bick, J.
Fareed, M. M. Samama, Expert Opin. Invest. Drugs 2002, 11,
397 – 407; b) S. Middeldorp, Thromb. Res. 2008, 122, 753 – 762;
c) M. Nagler, M. Haslauer, W. A. Wuillemin, Thromb. Res. 2012,
129, 407 – 417.
[7] a) C. A. A. van Boeckel, T. Beetz, J. N. Vos, A. J. M. de Jong,
S. F. van Aelst, R. H. van den Bosch, J. M. R. Mertens, F. A.
van der Vlugt, J. Carbohydr. Chem. 1985, 4, 293 – 321; b) M.
Petitou, J.-C. Jacquinet, P. Sinay, J. Choay, J.-C. Lorraeau, M.
Nassr, US Patent 4,818,816, 1989.
[8] a) C. Noti, J. L. de Paz, L. Polito, P. H. Seeberger, Chem. Eur. J.
2006, 12, 8664 – 8686; b) F. Baleux, L. Loureiro-Morais, Y.
Hersant, P. Clayette, F. Arenzana-Seisdedos, D. Bonnaffꢀ, H.
Lortat-Jacob, Nat. Chem. Biol. 2009, 5, 743 – 748; c) S. Arungun-
dram, K. Al-Mafraji, J. Asong, F. E. Leach III, I. J. Amster, A.
[14] A. A. Joseph, V. P. Verma, X.-Y. Liu, C.-H. Wu, V. M. Dhur-
andhare, C.-C. Wang, Eur. J. Org. Chem. 2012, 744 – 753.
[15] P. Angibeaud, J.-P. Utille, Synthesis 1991, 737 – 738.
4
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Carbohydrates
Working against the clot: The synthetic
anticoagulant fondaparinux, a pentasac-
charide based on the antithrombin-bind-
ing domain of heparin, was prepared in
a concise and efficient manner in the
shortest route reported to date. The
application of one-pot strategies, the use
of common intermediates, and the effi-
cient preparation of monosaccharide
building blocks from commercial sources
are key features of this approach.
C.-H. Chang, L. S. Lico, T.-Y. Huang,
S.-Y. Lin, C.-L. Chang, S. D. Arco,
S.-C. Hung*
&&&& — &&&&
Synthesis of the Heparin-Based
Anticoagulant Drug Fondaparinux
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!