Dialkylphosphates as stereodirecting protecting groups in oligosaccharide synthesis

Article abstract of DOI:10.1002/anie.200602699

(Chemical Equation Presented) Double duty: A phosphoric ester at the C2 position in glycosyl donors directs the glycosylation to occur selectively at the anomeric carbon atom with complete 1,2-trans selectivity. Since the phosphoric ester can be removed to give the hydroxy function, this group serves as a stereodirecting protecting group in oligosaccharide synthesis.

Full text of DOI:10.1002/anie.200602699

Angewandte
Chemie
Glycosylation
DOI: 10.1002/anie.200602699
Dialkylphosphates as Stereodirecting Protecting
Groups in Oligosaccharide Synthesis**
Scheme 1. Dialkylphosphates as stereodirecting protecting groups.
Takeshi Yamada, Kazunobu Takemura,
Jun-ichi Yoshida, and Shigeru Yamago*
report on the use of dialkylphosphates for the protection of
alcohols. Furthermore, while there are a few examples of the
use of diphenylphosphinamides as stereodirecting protecting
groups in glucosamine synthesis,[10] their synthetic potential
has not been well recognized.
Thioglycoside 1a, which has a 2,2-dimethyltrimethylene
(DMTM) phosphate group at the C2 position, was activated
by treatment with benzenesulfenylpiperidine (BSP) and triflic
anhydride (Tf2O) in the presence of 2,6-di-tert-butyl-4-
methylpyridine (DTBMP) at À608C,[11] and cyclohexanol
was added to this reaction mixture. The desired 1,2-trans-
glycoside b-2a formed as the sole product in 76% yield
(Scheme 2). The formation of a-2a and the corresponding
The stereoselective formation of the glycosidic linkage is one
of the most important tasks in synthetic carbohydrate
chemistry, because the glycoside structure plays a crucial
role in many important biological processes involving oligo-
saccharides.[1] To date, 1,2-trans-glycosides have been pre-
pared through the intramolecular participation of the acyl
protecting groups at the neighboring C2 position.[2] However,
this method has serious drawbacks owing to the formation of
orthoester side products.[3] The formation cannot be com-
pletely surpressed even when sterically and electronically
protected pivalates are used.[4] While the orthoester can be
isomerized to give the 1,2-trans-glycoside, the isomerization
[3f,j,k,5]
frequently results in lowefficiency
or formation of the
undesired stereoisomer.[3i,6] During our studies to develop an
iterative glycosylation,[7,8] we faced the same problem, which
limited the generality of the method. Therefore, the develop-
ment of a general and high-yielding method for the synthesis
of 1,2-trans-glycosides would considerably increase the gen-
erality of both the iterative as well as conventional glyco-
sylations to access to certain complex oligosaccharides.
Here we report the first use of dialkylphosphates as
stereodirecting groups for the synthesis of 1,2-trans-glyco-
sides. As the phosphates can be removed after glycosylation,
they can be used as stereodirecting protecting groups
(Scheme 1). In addition, we found that these protecting
groups can be used in iterative glycosylation reactions.
Therefore, a variety of oligosaccharides that possess the 1,2-
trans-glycosidic linkage can be synthesized under a set of
glycosylation conditions. Although dialkylphosphinamides
have been used for the protection of amines,[9] there is no
Scheme 2. Effects of dialkylphosphates. Reagents and conditions: a) 1
(1.0 equiv), BSP (1.1 equiv), Tf2O (1.4 equiv), DTBMP (2.0 equiv),
CH2Cl2, MS (4 Š) ,À608C for 30 min, then cyclohexanol (1.5 equiv),
À608C for 30 min, 2a; 76% (>99% b), 2b; 84% (>99% b), 2c;
77% (>99% b), 2d; 84% (89% b) . b)1a (1.0 equiv), NIS (1.1 equiv),
TfOH (0.1 equiv), cyclohexanol (1.5 equiv), CH2Cl2, MS (4 Š) ,À458C,
60 min, 79% (99% b) . c)2 (1.0 equiv), NaOH (10 equiv), EtOH/H2O,
608C, 1 h. Bn=benzyl, Cy=cyclohexyl, NIS=N-iodosuccinimide,
Tf =trifluoromethylsulfonyl.
[*] K. Takemura, Prof. S. Yamago[+]
Division of Molecular Materials Science
Graduate School of Science, Osaka City University
Osaka 558-8585 (Japan)
Dr. T. Yamada,[+] Prof. J.-i. Yoshida
Department of Synthetic Chemistry and Biological Chemistry
Graduate School of Engineering, Kyoto University
Kyoto 615-8510 (Japan)
orthophosphorate product was not detected. The same
glycosylation of 1a by reaction with N-iodosuccinimide and
triflic acid activators in the presence of cyclohexanol also
afforded b-2a exclusively in 79% yield.[12] The DMTM
phosphate group in 2a was removed by treatment with
sodium hydroxide (10 equiv) in an ethanol/water mixture at
608C for 1.0 h to give the free alcohol 3 in 95% yield.
[+] Current address:
Institute for Chemical Research, Kyoto University
Kyoto 611-0011 (Japan)
Fax: (+81)774-38-3067
We next examined the structural effects of the phosphates
on the selectivity and deprotection efficiency. Trimethylene
phosphate 1b reacted with similar efficiencies to 1a in both
the glycosylation (84%, > 99% b) and the deprotection
(84%). Diphenyl phosphate 1c also showed high glycosyla-
tion efficiencies (77%, > 99% b), but the deprotection was
E-mail: yamago@scl.kyoto-u.ac.jp
[**] This work was supported in part by a Shiseido Grant for Science
Research. The authors thank Prof. Kawabata of Kyoto University for
the measurement of optical rotation.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2006, 45, 7575 –7578
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7575

Communications
sluggish (< 30% yield). The selectivity eroded in
the glycosylation of diethyl phosphate 1d (89%
b). We selected DMTM phosphate for the follow-
ing studies since in the protection of the C2
hydroxy group it was more efficient than the
corresponding trimethylene analogue (see the
Supporting Information).
The synthetic scope of the current phosphate
protecting group has been examined by changing
the structure of donors as well as acceptors
(Table 1, Scheme 3). We employed a binary
activation system using BSP and Tf2O[11] because
these conditions could be applied for the iterative
glycosylation of thioglycosides derived from 2-
deoxy-2-aminoglycosides.[7a,b] The glycosylation
of 1a with 4e, which has a hydroxy group at C6,
proceeded smoothly and afforded the desired O-
glycoside with complete b selectivity in quantita-
tive yield (entry 1, Table 1). The glycosylation of
1a with 4 f and 4g, which have the sterically
hindered hydroxyl groups at C3 and C4, respec-
tively, also afforded 6 and 7 in good yields
(entries 2 and 3, Table 1). While small quantities
of the 1,2-cis isomers were formed in both cases,
sufficiently high levels of selectivity were
observed (> 96%). Thioglycosides with
a
DMTM protecting group at C2 could also be
used as glycosyl acceptors. Thus, the coupling of
1a with 8 afforded b-9h in quantitative yield
(entry 4, Table 1). Iterative glycosylation using 9h
as a donor and 8 as an acceptor afforded
trisaccharide 9i (entry 5, Table 1), which was
further glycosylated with 8 to give tetrasaccharide
9j (entry 6, Table 1). The desired 1,2-trans-glyco-
sides formed in good yields with excellent selec-
tivities in all cases. As the glycosyl donor was
activated prior to the coupling reaction, the
armed glycoside 10 could be used as a glycosyl
acceptor (entry 7, Table 1).[13,14] As all the prod-
Scheme 3. Examples of donors and acceptors used in the glycosylation and the
resulting products; see Table 1. Bz=benzoyl, NAP=2-naphthylmethyl,
(RO)2 =OCH2C(CH3)2CH2O, TCPN=N-tetrachlorophthalimide, TBS=tert-butyl-
dimethylsilyl.
ucts are thioglycosides, they could be directly used for the
next iterative or armed–disarmed glycosylation reaction as
glycosyl donors.
This method could be extended to the use of glycosides
other than glucose derivatives. Galactose derivative 12
afforded exclusively b isomer 13 (entry 8, Table 1). Mannose
donor 14k gave exclusively a isomers 15 and 16 upon coupling
with acceptors 4 f and 14l, respectively, as a result of the 1,2-
trans-directing effect of the axial DMTM group at C2
(entries 9 and 10, Table 1).
Potential intermediates in the current glycosylation reac-
tion were analyzed to understand the remarkable 1,2-trans
selectivity. Thus, 1a was activated by BSP and Tf2O in CD2Cl2
at À808C, and the resulting reaction mixture was analyzed by
1H, 13C, and 31P NMR spectroscopy at the same temperature.
We were surprised to find that the a-glycosyl triflate
17formed as the sole product.[15] The
Table 1: Stereoselective glycosylation of thioglycosides.[a] For structures
see Scheme 3.
Entry
Donor
Acc.
Prod.
Yield [%]
Sel.
(a:b)
1
2
3
4
5
6
7
8
9
10
1a
1a
1a
1a
9h
9i
1a
12
14k
14k
4e
4 f
4g
8
8
8
10
4e
4 f
14l
5
6
7
9h
9i
9j
11
13
15
16
98
75
71
100
80
80
85
100
81
76
1:>99
2:98
4:96
1:>99
1:99
1:>99
2:98
2:98
99:1
96:4
anomeric carbon in 17 resonates at d =
105 ppm in the 13C NMR spectra, which
strongly indicates the formation of a neu-
tral species rather than a cationic spe-
[a] Donor (1.0 equiv), BSP (1.1 equiv), Tf2O (1.4 equiv), DTBMP
(2.0 equiv), CH2Cl2, MS (4 Š) ,À608C, 10–30 min, then acceptor
(1.5 equiv)at À608C, 15–60 min.
cies.[15,7a] The a stereochemistry was sug-
gested by the 1H NMR spectra, in which the
7576
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7575 –7578

Angewandte
Chemie
DTBMP (41.1 mg, 0.20 mmol), and molecular sieves 4 Š (ca. 100 mg)
C1 proton gives rise to a doublet at d = 6.40 ppm with a
coupling constant of 2.4 Hz. The phosphorus signal in 17
occurs at d = À8.9 ppm, a chemical shift virtually identical to
that of 1a (d = À8.8 ppm). The results clearly indicate that the
phosphorus atom in 17 does not bear a positive charge.[16]
These results are in sharp contrast to the formation of the
orthoester cation intermediate in C2 acyl-protected glycosyl
donors.[17] Therefore, the lack of the formation of orthophos-
phate product must be attributed to the lack of positive
charge on the phosphorus atom in 17.
in CH2Cl2 (1.0 mL) was treated with Tf2O (38.4 mg, 0.14 mmol) at
À608C. After 10 min, glycosyl acceptor 4e (112 mg, 0.15 mmol) was
added at this temperature. The reaction mixture was at this temper-
ature for 15 min, then quenched by addition of Et3N (0.1 mL), and
warmed to room temperature. To this solution was added saturated
aqueous NaHCO3 solution, and organic phase was separated. The
aqueous phase was extracted three times with ethyl acetate, and the
combined organic extracts were washed with saturated aqueous NaCl
solution, dried with MgSO4, filtered, and concentrated under reduced
pressure by rotary evaporator to give the crude product. Purification
by flash column chromatography (silica gel 10 g; elution with 30%
ethyl acetate in hexane) afforded 5 in 98% (92.1 mg, a:b = < 1:99) as
a white amorphous substance. [a]2D0 = + 59.8 (c 0.4, CHCl3); IR (KBr):
n˜ = 1732(s), 1389, 1354, 1301, 1271, 1068, 1010, 740, 711; 1H NMR
(400 MHz, CDCl3): d = 0.81 (s, 3H, CH3), 1.17 (s, 3H, CH3), 3.44 (dt,
J = 9.7, 3.1 Hz, 1H, H-5’), 3.58–3.68 (m, 3H, H-4’, H-6’) 3.83 (t, J =
9.0 Hz, 1H, H-3’), 3.89–4.18 (m, 6H, H-6, P(O){OCH2C-
(CH3)2CH2O}), 4.24–4.36 (m, 2H, H-2’, H-5), 4.41 (d, J = 12.4 Hz,
1H, CH2Ph), 4.50 (d, J = 12.4 Hz, 1H, CH2Ph), 4.81 (d, J = 10.8 Hz,
1H, CH2Ph), 4.56 (t, J = 10.4 Hz, 1H, H-2), 4.71 (d, J = 8.0 Hz, 1H, H-
1’), 4.77 (d, J = 10.8 Hz, 1H, CH2Ph), 4.78 (d, J = 10.4 Hz, 1H,
CH2Ph), 4.96 (d, J = 10.4 Hz, 1H, CH2Ph), 5.46 (t, J = 9.8 Hz, 1H, H-
4), 5.88 (d, J = 10.4 Hz, 1H, H-1), 6.17 (dd, J = 9.8, 9.4 Hz, 1H, H-3),
7.09–7.53 (m, 26H, Ar), 7.69–7.74 (m, 2H, Ar), 7.85–7.91 ppm (m,
2H, Ar); 13C NMR (100 MHz, CDCl3): d = 20.41 (CH3), 21.49 (CH3),
32.05 (C, JCP = 5.3 Hz), 54.49 (CH, C2), 67.53 (CH2, C-6), 68.37 (CH2,
C6’), 69.54 (CH, C4), 72.11 (CH, C3), 73.42 (CH2, CH2Ph), 74.85
(CH2, CH2Ph), 75.02 (CH, C5’), 75.28 (CH2, CH2Ph), 77.60 (CH, C4),
77.59–77.70 (CH2, 2C, P(O){OCH2C(CH3)2CH2O}), 78.60 (CH, C5),
A possible explanation for the observed 1,2-trans selec-
tivity is the “SN2-like” attack of the acceptor alcohol on the a-
triflate intermediate (Scheme 4, path A) in analogy to the b-
mannoside synthesis developed by Crich.[15] However, Crich
79.27 (CH, JCP = 6.9 Hz, C2’), 82.04 (CH, C1), 83.04 (CH, JCP
=
3.0 Hz, C3’), 100.71 (CH, JCP = 3.1 Hz, C1’), 126.75 (C), 127.05 (C),
127.54 (CH), 127.61 (CH), 127.72 (CH, 2C), 127.81 (CH, 2C), 128.16
(CH, 2C), 128.16 (C), 128.27 (CH, 2C), 128.30 (CH, 4C), 128.32 (CH,
4C), 128.38 (CH, 2C), 128.59 (C), 129.25 (CH, 2C), 129.72 (CH, 2C),
129.80 (CH, 2C), 129.91 (C), 130.02 (C), 130.90 (C), 132.30 (CH, 2C),
133.46 (CH), 133.49 (CH), 137.90 (C), 137.98 (C), 138.11 (C), 140.37
=
=
=
(C), 140.61 (C), 162.05 (C O), 163.20 (C O), 165.31 (C O), 165.90
=
(C O); HRMS (FAB) m/z: calcd for C66H61Cl4NO16PS: 1326.2203
[M+H]+; found: 1326.2195.
Scheme 4. Plausible mechanism.
Received: July 7, 2006
Revised: August 15, 2006
Published online: October 20, 2006
also reported that a-triflate intermediates, which possess a
nonparticipative protecting group at C2, showlowor
moderate 1,2-trans selectivity.[18] We also reported that a b-
triflimide intermediate of N-phthaloyl-protected glucosamine
exclusively afforded the b-glycosides,[7a] suggesting that the a-
triflates do not always react in an “SN2-like” manner. An
alternative and more plausible mechanism involves inter-
mediate 19, which would arise through neighboring group
participation of the phosphorus ester. This intermediate
would be short-lived and exist in equilibrium with 18. The
alcohol reacts at less hindered side in 19 to give 1,2-trans-
glycoside (path B). While there is no direct evidence for the
involvement of 19, we believe that the latter is more plausible
for the formation of 1,2-trans-glycosides.[19,20] Further syn-
thetic and mechanistic elaborations of this newstereodirect-
ing protecting group are currently underway.
Keywords: carbohydrates · glycosylation · oligosaccharides ·
.
protecting groups · stereoselective synthesis
[1] a) Carbohydrates in Chemistry and Biology (Eds.: B. Ernst,
G. W. Hart, P. Sinaþ), Wiley-VCH, Weinheim, 2000; b) A. Varki,
Glycobiology 1993, 3, 97.
[2] Other stereodirecting groups except acyl group have been
recently reported. a) J.-H. Kim, H. Yang, J. Park, G.-J. Boons, J.
Am. Chem. Soc. 2005, 127, 12090; b) J. T. Smoot, P. Pornsur-
iyasak, A. V. Demchenko, Angew. Chem. 2005, 117, 7285;
Angew. Chem. Int. Ed. 2005, 44, 7123.
[3] Selected recent examples: a) P. H. Seeberger, M. Eckhardt, C. E.
Gutteridge, S. J. Danishefsky, J. Am. Chem. Soc. 1997, 119,
10064; b) Y. Konda, T. Toida, E. Kaji, K. Takeda, Y. Harigaya,
Carbohydr. Res. 1997, 301, 123; c) X.-S. Ye, C.-H. Wong, J. Org.
Chem. 2000, 65, 2410; d) F. W. Lichtenthaler, M. Oberthür, S.
Peters, Eur. J. Org. Chem. 2001, 3849; e) Y. Du, M. Zhang, F.
Kong, J. Chem. Soc. Perkin Trans. 1 2001, 2289; f) M. Mach, U.
Schluester, F. Mathew, B. Fraser-Raid, K. C. Hazen, Tetrahedron
2002, 58, 7345; g) H. Amer, A. Hofinger, P. Kosma, Carbohydr.
Res. 2003, 338, 35; h) A. Fürstner, F. Jeanjean, P. Razon, C.
Experimental Section
Typical glycosylation procedure (synthesis of 5): A solution of a
glycosyl donor 1a (69.1 mg, 0.10 mmol), BSP (23.0 mg, 0.11 mmol),
Angew. Chem. Int. Ed. 2006, 45, 7575 –7578
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7577

Communications
Wirtz, R. Mynott, Chem. Eur. J. 2003, 9, 320; i) J. Ning, W.
Zhang, Y. Yi, G. Yang, Z. Wu, J. Yi, F. Kong, Bioorg. Med. Chem.
2003, 11, 2193; j) N. Pleuss, H. Kunz, Angew. Chem. 2003, 115,
3282; Angew. Chem. Int. Ed. 2003, 42, 3174; k) H. A. Orgueira,
A. Bartolozzi, P. Schell, R. E. J. N. Litjens, E. R. Palmacci, P. H.
Seeberger, Chem. Eur. J. 2003, 9, 140; l) S. Dziadek, C. Brocke,
H. Kunz, Chem. Eur. J. 2004, 10, 4150; m) S. M. Rele, S. S. Iyer, S.
Baskaran, E. L. Chaikof, J. Org. Chem. 2004, 69, 9159; n) X.
Zhu, B. Yu, Y. Hui, R. R. Schmidt, Eur. J. Org. Chem. 2004, 965.
[4] a) A. Harreus, H. Kunz, Liebigs Ann. Chem. 1986, 717; b) K. W.
Neumann, J. Tamura, T. Ogawa, Bioorg. Med. Chem. 1995, 3,
1637; c) O. J. Plante, E. R. Palmacci, R. B. Andrade, P. H.
Seeberger, J. Am. Chem. Soc. 2001, 123, 9545; d) J. D. C.
Codꢀe, J. van den Bos, R. E. J. N. Litjens, H. S. Overkleeft,
J. H. van Boom, G. A. van der Marel, Org. Lett. 2003, 5, 1947.
[5] a) T. Ogawa, K. Katano, M. Matsui, Carbohydr. Res. 1978, 64,
C3; b) K. Zegelaar-Jaarsveld, H. I. Duynstee, G. A. van der
Marel, J. H. van Boom, Tetrahedron 1996, 52, 3575.
[6] Y. Zeng, J. Ning, F. Kong, Tetrahedron Lett. 2002, 43, 3729.
[7] a) S. Yamago, T. Yamada, T. Maruyama, J. Yoshida, Angew.
Chem. 2004, 116, 2197; Angew. Chem. Int. Ed. 2004, 43, 2145;
b) T. Yamada, S. Kinjo, J. Yoshida, S. Yamago, Chem. Lett. 2005,
34, 1556; c) S. Yamago, T. Yamada, O. Hara, H. Ito, Y. Mino, J.
Yoshida, Org. Lett. 2001, 3, 3867; d) S. Yamago, T. Yamada, H.
Ito, O. Hara, Y. Mino, J. Yoshida, Chem. Eur. J. 2005, 11, 6159.
[8] a) J. D. C. Codꢀe, R. E. J. N. Litjens, L. J. van der Bos, H. S.
Overkleeft, G. A. van der Marel, Chem. Soc. Rev. 2005, 34, 769;
b) See also, X. Huang, L. Huang, H. Wang, X.-S. Ye, Angew.
Chem. 2004, 116, 5333; Angew. Chem. Int. Ed. 2004, 43, 5221.
[9] Protective Groups in Organic Synthesis, 3rd ed. (Eds.: T. W.
Greene, P. G. M. Wuts), Wiley, NewYork, 1999, pp. 598 – 600.
[10] a) L. Zervas, S. Konstas, Chem. Ber. 1960, 93, 435; b) J. Banoub,
P. Boullanger, D. Lafont, Chem. Rev. 1992, 92, 1167.
[11] D. Crich, M. Smith, J. Am. Chem. Soc. 2001, 123, 9015. Several
attempts to utilize 2-acyl protected thioglycoside donors for the
iterative glycosylation under Crichꢁs conditions failed.
[12] a) P. Konradsson, U. E. Udodong, B. Fraser-Reid, Tetrahedron
Lett. 1990, 31, 4313; b) G. H. Veeneman, S. H. van Leeuwen,
J. H. van Boom, Tetrahedron Lett. 1990, 31, 1331.
[13] Competition experiments indicated that 1a could be classified
into disarmed glycoside (See Supporting Information).
[14] D. R. Mootoo, P. Koradsson, U. Udodong, B. Fraser-Reid, J. Am.
Chem. Soc. 1988, 110, 5583.
[15] D. Crich, S. Sun, J. Am. Chem. Soc. 1997, 119, 11217.
[16] The 31P NMR resonances shift downfield when trialkylphos-
phates are converted to tetraalkyl phosphonium ions. J. H.
Finley, D. Z. Denney, D. B. Denney, J. Am. Chem. Soc. 1969, 91,
5826.
[17] D. Crich, Z. M. Dai, S. Gastaldi, J. Org. Chem. 1999, 64, 5224.
[18] a) D. Crich, W. Cai, J. Org. Chem. 1999, 64, 4926; b) D. Crich,
T. K. Hutton, A. Banerjee, P. Jayalath, J. Picione, Tetrahedron:
Asymmetry 2005, 16, 105.
[19] It is unlikely that the excellent selectivity is attributed to the
steric effect because the use of steric hindered alcohols showed
lower selectivity than one of less hinder alcohols (Table 1,
entries 1–3).
[20] Thiomannoside 14k gave the corresponding a-triflate inter-
mediate upon activation by BSP and Tf2O as judged by the low
temperature 1H NMR study, and the intermediate selectively
afforded a-glycoside upon treatment of alcohol. The result
absolutely denies the path A and is consistent with the path B.
7578
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7575 –7578