An efficient approach to the synthesis of nucleosides: Gold(I)-catalyzed N-glycosylation of pyrimidines and purines with glycosyl ortho-alkynyl benzoates

Article abstract of DOI:10.1002/anie.201100514

Persuaded with gold: The title reaction in the presence of [Ph ₃PAuNTf₂] (Tf=trifluoromethanesulfonyl) led conveniently to the corresponding nucleosides with excellent regioselectivity (see scheme). Even purine derivatives underwent this transformation owing to the mild conditions, which enabled the use of protecting groups that would not usually be compatible with N-glycosylation conditions. Copyright

Full text of DOI:10.1002/anie.201100514

DOI: 10.1002/anie.201100514
Nucleoside Synthesis
An Efficient Approach to the Synthesis of Nucleosides: Gold(I)-
Catalyzed N-Glycosylation of Pyrimidines and Purines with Glycosyl
ortho-Alkynyl Benzoates**
Qingju Zhang, Jiansong Sun,* Yugen Zhu, Fuyi Zhang, and Biao Yu*
The synthesis of nucleosides has continuously been a topical
subject in efforts to develop new therapeutic agents (e.g.,
antitumor and antiviral drugs),[1] to manipulate genetic
processes (e.g., antisense oligonucleotides and RNA interfer-
ence),[2] and to expand the genetic code and understand the
scope and limits of Watson–Crick base pairing.[3] However,
the key technique for such syntheses, namely, the N-glycosidic
coupling of sugars and nitrogen heterocycles, is rather
conventional. A Vorbrꢀggen-type reaction[4] involving the
coupling of sugar acetates with trimethylsilylated nucleobases
under the action of strong Lewis acids (mostly stoichiometric
amounts of trimethylsilyl triflate and SnCl4) is still the
predominant method.[5] The coupling yields are not always
high. In particular, when purines are used, low coupling yields
(or even failure of the reaction to occur) and moderate N9/N7
regioselectivity have been encountered.[6,7] Protecting groups
and temporary substituents have thus been introduced onto
purines to enhance their reactivity and hinder the reaction of
nonglycosylated nitrogen atoms.[8] However, the choice of
protecting groups is limited by the harsh conditions required
for Vorbrꢀggen-type reactions. Variation of the anomeric
leaving group of the glycosyl donors could enable N-glyco-
sylation under milder conditions. However, only limited
success has been reported for nucleoside synthesis with
privileged O-glycosylation donors, including glycosyl chlor-
ides/bromides,[5,9] trichloroacetimidates,[10] phosphites,[11] sulf-
oxides,[12] thioglycosides,[13] n-pentenyl glycosides,[14] and
sugar 1,2-anhydrides.[15] One rationale is that nucleobases
are poorly nucleophilic (and often poorly soluble) and thus
compete unfavorably for glycosidation with other nucleo-
philic species that occur in a glycosylation system, such as
those derived from the leaving groups and promoters.
These considerations led us to try glycosyl N-phenyl-
trifluoroacetimidate (PTFAI) donors[16] for nucleoside syn-
thesis, since these donors have been shown to be advanta-
geous for the N-glycosylation of amides owing to the poor
competitiveness of the N-phenyltrifluoroacetamide leaving
entity.[17] PTFAI donors could be coupled smoothly with
silylated pyrimidines;[18] however, they underwent decompo-
sition when much more poorly nucleophilic purine derivatives
were used as acceptors. In such circumstances, the stability of
the donors is not only beneficial to handling but is also
demanding for an effective glycosidic coupling to proceed
(prior to decomposition). It is probably for this reason that
sugar acetates are still the prevailing donors for nucleoside
synthesis (i.e., the Vorbrꢀggen-type reaction). Our recently
introduced glycosyl ortho-alkynyl benzoates are as stable as
sugar acetates, yet can be activated for glycosidation under
mild conditions by catalysis with a gold(I) complex (e.g.,
commercially available and shelf-stable [Ph3PAuNTf2]; Tf =
trifluoromethanesulfonyl).[19] Furthermore, the leaving entity
(an isocoumarin) and the promoter introduce no competitive
nucleophilic species.[20] Thus, we aimed to tackle the problem
of purine glycosylation with this new glycosylation protocol.
We first examined the N-glycosylation of pyrimidine
nucleobases with glycosyl ortho-hexynylbenzoates
1
(Schemes 1 and 2). Thus, uridine (2a) was silylated with
N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) in aceto-
nitrile and then subjected to coupling with the perbenzoyl
ribofuranosyl ortho-hexynylbenzoate 1a in the presence of
[Ph3PAuNTf2] (0.1 equiv) at room temperature. The reaction
was slow and required 3 days to reach completion. Never-
theless, it provided the desired nucleoside 3 cleanly (85%).
The coordination of acetonitrile to the gold(I) catalyst
accounts for the slow reaction rate.[19b] With nitromethane
as the solvent, the equivalent reaction was complete within
24 hours. Other solvents often used for O-glycosylation, such
as dichloromethane, 1,2-dichloroethane, and toluene, are
poor solvents for the bases and therefore led to failure of
the coupling reaction.
[*] Q. Zhang, Dr. J. Sun, Y. Zhu, Prof. B. Yu
State Key Laboratory of Bioorganic and Natural Products Chemistry
Shanghai Institute of Organic Chemistry
Chinese Academy of Sciences
345 Lingling Road, Shanghai 200032 (China)
Fax: (+86)21-6416-6128
E-mail: byu@mail.sioc.ac.cn
Q. Zhang, Dr. F. Zhang
Department of Chemistry, Zhengzhou University (China)
[**] This research was financially supported by the National Natural
Science Foundation of China (20932009 and 20921091), the
National Basic Research Program of China (2010CB529706), and
the E-Institute of Shanghai Municipal Education Commission
(E09013).
Without optimizing the present N-glycosylation condi-
tions, we examined the coupling of the three pyrimidine
nucleobases, that is, uridine (2a), thymine (2b), and N4-
benzoylcytosine (2c), with peracyl furanosyl and pyranosyl
ortho-hexynylbenzoates 1a, 1b, and 1d–g. The coupling
reactions with furanosyl donors 1a and 1b provided the
desired nucleosides 3–8 in excellent yields (85–96%). With
the peracetyl galactopyranosyl and rhamnopyranosyl donors
Supporting information for this article, including experimental
details, characterization data, and NMR spectra for all new
compounds, is available on the WWW under http://dx.doi.org/10.
1002/anie.201100514.
Angew. Chem. Int. Ed. 2011, 50, 4933 –4936
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4933

Communications
yield. Nevertheless, upon replacement of the peracetyl donor
1d with its perbenzoyl counterpart 1e, the coupling yields
increased significantly, with nucleosides 12 and 13 formed in
about 90% yield. Moreover, we later found that the addition
of HOTf (0.1 equiv) to the reaction mixture could greatly
increase the reaction rate. Thus, the coupling of 1a with
pyrimidine nucleobases 2a–c was complete within 3–6 hours
to provide nucleosides 3–5 in 92–98% yield (HOTf itself
could not promote these glycosylation reactions). The cou-
pling of uridine (2a) with the ribofuranosyl ortho-hexynyl-
benzoate 1a was readily scaled up to a gram scale, which
enabled us to quickly assemble the nucleoside lipid derivative
JBIR-68. This compound was recently identified from Actino-
bacteria as a potent anti-influenza agent.[22,23]
We then explored the real challenge of purine N-glyco-
sylation. However, under similar conditions to those used for
the above pyrimidine N-glycosylation, the coupling of ribo-
furanosyl ortho-hexynylbenzoate 1a or 1b with routinely
protected purine nucleobases, that is, N6-benzoyladenine
(2d) and N2-acetyl-6-O-(diphenylcarbamoyl)guanine (2e;
Scheme 1),[8e,f,24] led to either low yields of the coupling
products (about 28% yield for the coupling with 2d) or
marginal N9/N7 regioselectivity (1.3:1 for the coupling with
2e in a total yield of about 70%). Fortunately, the mild
glycosylation conditions with ortho-alkynyl benzoates as
donors would be compatible with a large variety of protecting
groups on purines, including protecting groups that would not
survive under conventional N-glycosylation conditions. Our
immediate choice was to use the tert-butoxycarbonyl (Boc)
group, one of the most frequently used N-protecting groups,
to protect the exocyclic amino group of purines. In fact, N6-
bis(tert-butoxycarbonyl)adenine (2 f) and N2-tert-butoxycar-
bonyl-2-amino-6-chloropurine (2g) have already been con-
veniently prepared and effectively used in Mitsunobu-type
N-alkylation reactions (with high N9/N7 regioselectivity).[25,26]
The good solubility of purine derivatives 2 f and 2g enabled
their glycosylation to be carried out in dichloromethane and
without the need for prior silylation. The coupling of 2 f with
1a in the presence of [Ph3PAuNTf2] (0.1 equiv) in dichloro-
methane proceeded smoothly at room temperature to give
the N9 nucleoside 20 in good yield (77%) within 12 hours
(Scheme 3). Only a trace amount of the mono-Boc-protected
product was isolated as a side product,[23] and the correspond-
ing N7 nucleoside was not detected at all.
Scheme 1. Coupling partners for N-glycosylation. Bz=benzoyl.
Glycosylation of the N2-Boc-2-amino-6-halopurines 2g–i
with 1a led to the corresponding N9 nucleosides 21–23 in
better yields (ca. 85%). The perbenzoyl arabinofuranosyl
donor 1c[21] performed better than the ribofuranosyl counter-
part 1a in coupling with purines 2 f–h: the N9 nucleoside
products 24–26 were obtained in 80–90% yield. The perben-
zoyl glucopyranosyl and peracetyl rhamnopyranosyl ortho-
hexynylbenzoates 1e and 1g also underwent coupling with
Boc-protected purines 2 f and 2g, although the desired
coupling products 27–30 were formed in lower yields (48–
80%) than those observed with furanose donors. Neverthe-
less, the reactions were still clean; remaining starting materi-
als mainly accounted for the lower yields, and no N7-
nucleoside products were detected. The N-Boc groups on
the resulting nucleosides 20–30 were removed selectively in
Scheme 2. N-glycosylation of pyrimidines 2a–c with glycosyl ortho-
hexynylbenzoates.
1 f and 1g, the coupling yields were still satisfactory; the
reactions led to the nucleoside products 14–19 in 74–95%
yield. Only when the peracetyl glucopyranosyl ortho-hexy-
nylbenzoate 1d was used as the donor did the coupling yields
become lower, with nucleosides 9–11[21] formed in 67–76%
4934 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 4933 –4936

Angew. Chem. 1996, 108, 2020 – 2036; Angew. Chem. Int. Ed.
Engl. 1996, 35, 1894 – 1909.
[3] a) I. Luyten, P. Herdewijn, Eur. J. Med. Chem. 1998, 33, 515 –
576; b) A. J. A. Cobb, Org. Biomol. Chem. 2007, 5, 3260 – 3275.
[4] a) H. Vorbrꢀggen, Acc. Chem. Res. 1995, 28, 509 – 520; b) H.
Vorbrꢀggen, C. Ruh-Pohlenz, Org. React. 2000, 55, 1 – 630.
[5] a) S. Knapp, Chem. Rev. 1995, 95, 1859 – 1876; b) L. J. Wilson,
M. W. Hager, Y. A. El-Kattan, D. C. Liotta, Synthesis 1995,
1465 – 1479; c) C. Mathꢁ, C. Pꢁrigaud, Eur. J. Org. Chem. 2008,
1489 – 1505; d) G. Romeo, U. Chiacchio, A. Corsaro, P. Merino,
Chem. Rev. 2010, 110, 3337 – 3370.
[6] For recent examples, see: a) F. Marcelo, J. Jimꢁnez-Barbero, J.
Marrot, A. P. Rauter, P. Sinaꢂ, Y. Blꢁriot, Chem. Eur. J. 2008, 14,
10066 – 10073; b) C. S. Stauffer, A. Datta, J. Org. Chem. 2008, 73,
4166 – 4174.
[7] The diastereoselectivity issue in N-glycosylation reactions with
2-deoxysugars is not considered herein, which resorts mostly to
the sodium-salt method, see: Z. Kazimierczuk, H. B. Cottam,
G. R. Revankar, R. K. Robins, J. Am. Chem. Soc. 1984, 106,
6379 – 6382; for a recent report, see: J. W. Arico, A. K. Calhoun,
K. J. Salandria, L. W. McLaughlin, Org. Lett. 2010, 12, 120 – 122.
[8] a) M. Saneyoshi, E. Satoh, Chem. Pharm. Bull. 1979, 27, 2518 –
2521; b) R. Zou, M. J. Robins, Can. J. Chem. 1987, 65, 1436 –
1437; c) G. R. Green, T. J. Grinter, P. M. Kincey, R. L. Jarvest,
Tetrahedron 1990, 46, 6903 – 6914; d) T. F. Jenny, N. Previsani,
S. A. Benner, Tetrahedron Lett. 1991, 32, 7029 – 7032; e) T. F.
Jenny, S. A. Benner, Tetrahedron Lett. 1992, 33, 6619 – 6620;
f) M. J. Robins, R. Zou, Z. Guo, S. F. Wnuk, J. Org. Chem. 1996,
61, 9207 – 9212; g) M. Zhong, M. J. Robins, Tetrahedron Lett.
2003, 44, 9327 – 9330; h) M. Zhong, I. Nowak, M. J. Robins, Org.
Lett. 2005, 7, 4601 – 4603.
[9] a) E. Fischer, B. Helferich, Ber. Dtsch. Chem. Ges. 1914, 47, 210 –
235; b) G. E. Hilbert, T. B. Johnson, J. Am. Chem. Soc. 1930, 52,
4489 – 4494.
Scheme 3. N-glycosylation of Boc-protected purines 2 f–i with glycosyl
ortho-hexynylbenzoates. MS=molecular sieves.
[10] a) L. Chanteloup, N. T. Thuong, Tetrahedron Lett. 1994, 35, 877 –
880; b) K. i. Shohda, T. Wada, M. Sekine, Nucleosides Nucleo-
tides 1998, 17, 2199 – 2210.
[11] S.-I. Hashimoto, J. Inagaki, H. Sakamoto, A. Sano, M. Nakajima,
Heterocycles 1997, 46, 215 – 220.
[12] a) L. Chanteloup, J.-M. Beau, Tetrahedron Lett. 1992, 33, 5347 –
5350; b) D. DꢃAlonzo, A. Guaragna, A. Van Aerschot, P.
Herdewijn, G. Palumbo, J. Org. Chem. 2010, 75, 6402 – 6410.
[13] a) S. Knapp, W.-C. Shieh, Tetrahedron Lett. 1991, 32, 3627 – 3630;
b) H. Sugimura, I. Muramoto, T. Nakamura, K. Osumi, Chem.
Lett. 1993, 169 – 172; c) S. Hanessian, G. Huang, C. Chenel, R.
Machaalani, O. Loiseleur, J. Org. Chem. 2005, 70, 6721 – 6734;
d) S. Knapp, V. V. Thakur, M. R. Madduru, K. Malolanarasim-
han, G. J. Morriello, G. A. Doss, Org. Lett. 2006, 8, 1335 – 1337.
[14] M.-C. Chapeau, L. J. Marnett, J. Org. Chem. 1993, 58, 7258 –
7262.
[15] K. Chow, S. Danishefsky, J. Org. Chem. 1990, 55, 4211 – 4214.
[16] a) B. Yu, H. Tao, Tetrahedron Lett. 2001, 42, 2405 – 2407; b) B.
Yu, J. Sun, Chem. Commun. 2010, 46, 4668 – 4679.
[17] a) H. Tanaka, Y. Iwata, D. Takahashi, M. Adachi, T. Takahashi, J.
Am. Chem. Soc. 2005, 127, 1630 – 1631; b) K. Tanaka, T.
Miyagawa, K. Fukase, Synlett 2009, 1571 – 1574.
[18] a) J. Liao, J. Sun, B. Yu, Tetrahedron Lett. 2008, 49, 5036 – 5038;
b) J. Liao, J. Sun, B. Yu, Carbohydr. Res. 2009, 344, 1034 – 1038.
[19] a) Y. Li, Y. Yang, B. Yu, Tetrahedron Lett. 2008, 49, 3604 – 3608;
b) Y. Li, X. Yang, Y. Liu, C. Zhu, Y. Yang, B. Yu, Chem. Eur. J.
2010, 16, 1871 – 1882; c) Y. Yang, Y. Li, B. Yu, J. Am. Chem. Soc.
2009, 131, 12076 – 12077.
greater than 90% yield under the effect of BuOH/H2O[27] or
CF3COOH/ClCH2CH2Cl;[28] the acyl groups on the sugar
units were not affected.[23] The 6-halopurine nucleosides
synthesized are useful precursors to the corresponding
guanine nucleosides or their 6-substituted derivatives.[26,29]
In conclusion, readily accessible and shelf-stable glycosyl
ortho-hexynylbenzoates have been shown to be superior
donors for the N-glycosylation of nucleobases under the
catalysis of [Ph3PAuNTf2]. The success of this highly efficient
and regioselective N9-glycosylation of purines can be attrib-
uted to the mild glycosylation conditions that enable Boc-
protected purine derivatives to be used as coupling partners
for the first time. Further improvement of the coupling
efficiency through careful tuning of the reaction parameters
for individual coupling partners and the application of this
efficient method to nucleoside synthesis are anticipated.
Received: January 21, 2011
Published online: April 15, 2011
Keywords: glycosylation · gold · homogeneous catalysis ·
.
nucleosides · purines
[20] Y. Li, B. Yu, Chem. Commun. 2010, 46, 6060 – 6062.
[21] CCDC 807785 (1cb) and 807784 (11) contain the supplementary
crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
[1] a) J. Stagg, M. J. Smyth, Oncogene 2010, 29, 5346 – 5358; b) E.
De Clercq, Med. Res. Rev. 2010, 30, 667 – 707.
[2] a) E. Uhlmann, A. Peyman, Chem. Rev. 1990, 90, 543 – 584; b) U.
Englisch, D. H. Gauss, Angew. Chem. 1991, 103, 629 – 646;
Angew. Chem. Int. Ed. Engl. 1991, 30, 613 – 629; c) M. Egli,
Angew. Chem. Int. Ed. 2011, 50, 4933 –4936
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4935

Communications
[22] M. Takagi, K. Motohashi, A. Nagai, M. Izumikawa, M. Tanaka,
[26] a) X. Q. Yin, W. K. Li, S. W. Schneller, Tetrahedron Lett. 2006,
47, 9187 – 9189; b) B. Y. Michel, P. Strazewski, Tetrahedron 2007,
63, 9836 – 9841; c) A. Porcheddu, G. Giacomelli, I. Piredda, M.
Carta, G. Nieddu, Eur. J. Org. Chem. 2008, 5786 – 5797.
[27] J. Wang, Y. L. Liang, J. Qu, Chem. Commun. 2009, 5144 – 5146.
[28] N. Srinivasan, A. Yurek-George, A. Ganesan, Mol. Diversity
2005, 9, 291 – 293.
S. Fuse, T. Doi, K. Iwase, A. Kawaguchi, K. Nagata, T.
Takahashi, K. Shin-ya, Org. Lett. 2010, 12, 4664 – 4666.
[23] See the Supporting Information for details.
[24] a) F. W. Lichtenthaler, P. Voss, A. Heerd, Tetrahedron Lett. 1974,
15, 2141 – 2144; b) J. Sengupta, A. Bhattacharjya, J. Org. Chem.
2008, 73, 6860 – 6863.
[25] a) S. Dey, P. Garner, J. Org. Chem. 2000, 65, 7697 – 7699; b) S.
Fletcher, V. M. Shahani, A. J. Lough, P. T. Gunning, Tetrahedron
2010, 66, 4621 – 4632.
[29] P. Lagisetty, L. M. Russon, M. K. Lakshman, Angew. Chem.
2006, 118, 3742 – 3745; Angew. Chem. Int. Ed. 2006, 45, 3660 –
3663.
4936 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 4933 –4936