High-throughput five minute microwave accelerated glycosylation approach to the synthesis of nucleoside libraries

Article abstract of DOI:10.1021/jo061885l

The Vorbrueggen glycosylation reaction was adapted into a one-step 5 min/130 °C microwave assisted reaction. Triethanolamine in acetontrile containing 2% water was determined to be optimal for the neutralization of trimethylsilyl inflate allowing for direct MPLC purification of the reaction mixture. When coupled with a NH₃/methanol deprotection reaction, a high-throughput method of nucleoside library synthesis was enabled. The method was demonstrated by examining the ribosylation of 48 nitrogen containing heteroaromatic bases that included 25 purines, four pyrazolopyrimidines, two 8-azapurines, one 2-azapurine, two imidazopyridines, two benzimidazoles, three imidazoles, three 1,2,4-triazoles, two pyrimidines, two 3-deazapyrimidines, one quinazolinedione, and one alloxazine. Of these, 32 yielded single regioisomer products, and six resulted in separable mixtures. Seven examples provided inseparable regioisomer mixtures of -two to three compounds (16 nucleosides), and three examples failed to yield isolable products. For the 45 single isomers isolated, the average two-step overall yield ± SD was 26 ± 16%, and the average purity ± SD was 95 ± 6%. A total of 58 different nucleosides were prepared of which 15 had not previously been accessed directly from glycosylation/deprotection of a readily available base.

Full text of DOI:10.1021/jo061885l

High-Throughput Five Minute Microwave Accelerated Glycosylation
Approach to the Synthesis of Nucleoside Libraries
Brett C. Bookser* and Nicholas B. Raffaele
Metabasis Therapeutics, Incorporated, 11119 North Torrey Pines Road, La Jolla, California 92037
bookser@mbasis.com
ReceiVed September 12, 2006
The Vorbru¨ggen glycosylation reaction was adapted into a one-step 5 min/130 °C microwave assisted
reaction. Triethanolamine in acetontrile containing 2% water was determined to be optimal for the
neutralization of trimethylsilyl triflate allowing for direct MPLC purification of the reaction mixture.
When coupled with a NH3/methanol deprotection reaction, a high-throughput method of nucleoside library
synthesis was enabled. The method was demonstrated by examining the ribosylation of 48 nitrogen
containing heteroaromatic bases that included 25 purines, four pyrazolopyrimidines, two 8-azapurines,
one 2-azapurine, two imidazopyridines, two benzimidazoles, three imidazoles, three 1,2,4-triazoles, two
pyrimidines, two 3-deazapyrimidines, one quinazolinedione, and one alloxazine. Of these, 32 yielded
single regioisomer products, and six resulted in separable mixtures. Seven examples provided inseparable
regioisomer mixtures of -two to three compounds (16 nucleosides), and three examples failed to yield
isolable products. For the 45 single isomers isolated, the average two-step overall yield ( SD was 26 (
16%, and the average purity ( SD was 95 ( 6%. A total of 58 different nucleosides were prepared of
which 15 had not previously been accessed directly from glycosylation/deprotection of a readily available
base.
Introduction
To more thoroughly pursue therapeutic areas that might be
responsive to treatment with nucleosides, as well as to provide
more building blocks for oligonucleotide synthesis, it would
be beneficial to have a practical method for high-throughput
synthesis of these molecules. Traditional methods of nucleoside
analogue synthesis involve either a glycosylation reaction
Nucleosides and nucleotides have provided a productive area
of chemical and biological research for over 100 years.1,2 The
monomeric units of DNA and RNA are involved in the regu-
lation of a myriad of cellular metabolic pathways and have been
the subject of intense areas of research seeking to identify thera-
peutic agents for a variety of diseases including viral infections,
cancer, cardiovascular diseases, CNS diseases, etc. However,
few drugs have resulted from those efforts,3 which were focused
primarily in areas other than viral4 diseases and leukemias.5
(3) (a) Isono, K. Pharmacol. Ther. 1991, 52, 269-286. (b) Perigaud,
C.; Gosselin, G.; Imbach, J.-L. Nucleosides Nucleotides 1992, 11, 903-
945. (c) Jacobson, K. A.; van Galen, P. J. M.; Williams, M. J. Med. Chem.
1992, 35, 407-422. (d) Appleman, J. R.; Erion, M. D. Expert Opin. InVest.
Drugs 1998, 7, 225-243. (e) Williams, M.; Jarvis, M. F. Biochem.
Pharmacol. 2000, 59, 1173-1185.
(4) (a) Robins, R. K.; Revankar, G. R. In AntiViral Drug DeVelopment:
A Multidisciplinary Approach; De Clercq, E., Walker, R. T., Eds.; Plenum
Press: New York, 1988; p 11. (b) Koszalka, G. W.; Daluge, S. M.; Boyd,
F. L. Annu. Rep. Med. Chem. 1998, 33, 163-171. (c) De Clercq, E. J.
Clin. Virol. 2004, 30, 115-133.
(5) (a) Plunkett, W.; Gandhi, V. Cancer Chemother. Biol. Response
Modif. Annu. 2001, 19, 21-45. (b) Johnson, S. A. Expert Opin. Pharma-
cother. 2001, 2, 929-943.
* To whom correspondence should be addressed. Phone: (858) 622-5512;
fax: (858) 622-5573.
(1) The pioneering work of Emil Fischer in the fields of purines,
carbohydrates, and proteins set the foundation for later advances in the field
of nucleosides. (a) Fischer, E. Chem. Ber. 1897, 30, 1846-1859. (b) Fischer,
E. Chem. Ber. 1898, 31, 104-122. (c) Fischer, E. Chem. Ber. 1899, 32,
267-273.
(2) The term nucleoside was introduced in 1909 by Levene and Jacobs.
Levene, P. A.; Jacobs, W. A. Chem. Ber. 1909, 42, 2474-2478.
10.1021/jo061885l CCC: $37.00 © 2007 American Chemical Society
Published on Web 12/10/2006
J. Org. Chem. 2007, 72, 173-179
173

Bookser and Raffaele
SCHEME 1. Standard Vorbru1ggen Glycosylation Reaction Conditions
between a nitrogenous aromatic base and an activated sugar
intermediate or derivatization of a preformed nucleoside. Only
methods employing the latter strategy have been reported for
nucleoside library synthesis.6-8 These methods have been used
to produce large numbers of nucleosides where the core base
and ribofuranose moieties have been limited to only a few
options.
The glycosylation reaction has remained undeveloped as a
high-throughput process, even though over a period of more
than 90 years, it has been applied to the synthesis of a vast
array of diverse nucleosides.9 Glycosylation protocols include
the silver salt,10 the chloromercury,11 the fusion,12 the sodium
salt,13 the phase transfer,14 and the boron trifluoride etherate15
methods. Vorbru¨ggen combined elements from some of these
methods and introduced the use of Friedel-Crafts catalysts to
promote the glycosylation.16 The broad applicability of the
Vorbru¨ggen reaction made it desirable for development into a
high-throughput reaction.
As illustrated in Scheme 1 with the synthesis of adenosine
(6) as an example, nucleoside synthesis can be considered as a
three-step operation. The typical Vorbru¨ggen reaction requires
presilylation of the base 1 (step 1) and then reaction with
trimethylsilyl triflate activated sugar 3 to form the desired acyl-
protected nucleoside (step 2). One advantage of this method is
that frequently one nitrogen of the base is regioselectively
glycosylated because of thermodynamic equilibration.17 Con-
veniently, this often produces the same regioisomer as preferred
by the natural product. Also, with R-2′-O-acyl-substituted sugar
coupling partners, acyl-oxonium ion18 stabilization of the
anomeric cation, as depicted in formula 4, directs glycosylation
stereospecifically to the â-face of the sugar producing the natural
â-configuration. Furthermore, while the Vorbru¨ggen reaction
is frequently conducted as a two-step operation, it is possible
to combine steps 1 and 2 and perform this reaction by combining
all reagents and heating.19 Finally, deprotection of the acyl-
protecting groups of 5a by ammoniolysis (step 3) can be
performed in a simple operation to provide the nucleoside. The
synthesis of adenosine (6) depicted in Scheme 1 exemplifies
the utility of this method as it has been obtained in 81% overall
yield when base silylation, glycosylation, and deprotection were
each performed as distinct steps.16
(6) Amine substitution of nucleoside heteroaryl-halides: (a) Rodenko,
B; Wanner, M. J.; Koomen, G.-J. J. Chem. Soc., Perkin Trans. 1 2002,
1247-1252. (b) Varaprasad, C. V.; Habib, Q.; Li, D. Y.; Huang, J.; Abt,
J. W.; Rong, F.; Hong, Z.; An, H. Tetrahedron 2003, 59, 2297-2307. (c)
Ding, Y.; Habib, Q.; Shaw, S. Z.; Li, D. Y.; Abt, J. W.; Hong, Z.; An, H.
J. Comb. Chem. 2003, 5, 851-859. (d) Gunic, E.; Amador, R.; Rong, F.;
Abt, J. W.; An, H.; Hong, Z.; Girardet, J.-L. Nucleosides, Nucleotides,
Nucleic Acids 2004, 23, 495-499. (e) Koh, Y.-h.; Landesman, M. B.;
Amador, R.; Rong, F.; An, H.; Hong, Z.; Girardet, J.-L. Nucleosides,
Nucleotides, Nucleic Acids 2004, 23, 501-507.
(7) Palladium-coupling mediated derivatization: Aucagne, V.; Berteina-
Raboin, S.; Guenot, P.; Agrofoglio, L. A. J. Comb. Chem. 2004, 6, 717-
723.
(8) Formation of amides, ureas, and carbamates on the ribose moiety:
(a) Golisade, A.; Bressi, J. C.; Van Calenbergh, S.; Gelb, M. H.; Link, A.
J. Comb. Chem. 2000, 2, 537-544. (b) Herforth, C.; Wiesner, J.; Franke,
S.; Golisade, A.; Jomaa, H.; Link, A. J. Comb. Chem. 2002, 4, 302-314.
(c) Epple, R.; Kudirka, R.; Greenberg, W. A. J. Comb. Chem. 2003, 5,
292-310. (d) Herforth, C.; Wiesner, J.; Heidler, P.; Sanderbrand, S.; Van
Calenbergh, S.; Jomaa, H.; Link, A. Bioorg. Med. Chem. 2004, 12, 755-
762.
Despite the appeal of being able to produce a wide array of
these highly desirable molecules in a stereospecific and regio-
selective manner, several factors have precluded the adoption
of the Vorbru¨ggen reaction as a general high-throughput method.
(9) Review: Vorbru¨ggen, H.; Ruh-Pohlenz, C. Org. React. 2000, 55,
1-630.
(10) (a) Fischer, E.; Helferich, B. Chem. Ber. 1914, 47, 210-235. (b)
Hilbert, G. E.; Johnson, T. B. J. Am. Chem. Soc. 1930, 52, 4489-4494.
(11) Davoll, J.; Lowy, B. A. J. Am. Chem. Soc. 1951, 73, 1650-1655.
(12) Ishido, Y.; Sato, T. Bull. Chem. Soc. Jpn. 1961, 34, 1347-1348.
(13) (a) Kazimierczuk, Z.; Cottam, H. B.; Revankar, G. R.; Robins, R.
K. J. Am. Chem. Soc. 1984, 106, 6379-6382. (g) Kandasamy, R.; Imamura,
N.; Robins, R. K.; Revankar, G. R. Tetrahedron Lett. 1987, 28, 5107-
5110.
(15) Cottam, H. B.; Petrie, C. R.; McKernan, P. A.; Goebel, R. J.; Dalley,
N. K.; Davidson, R. B.; Robins, R. K.; Revankar, G. R. J. Med. Chem.
1984, 27, 1119-1127.
(16) Vorbru¨ggen, H.; Krolikiewicz, K.; Bennua, B. Chem. Ber. 1981,
114, 1234-1255.
(17) Vorbru¨ggen, H.; Hofle, G. Chem. Ber. 1981, 114, 1256-1268.
(18) Baker, B. R.; Joseph, J. P.; Schaub, R. E.; Williams, J. H. J. Org.
Chem. 1954, 19, 1786-1792.
(14) (a) Seela, F.; Winkeler, H.-D. J. Org. Chem. 1982, 47, 226-230.
(b) Ugarkar, B. G.; Castellino, A. J.; DaRe, J. M.; Kopcho, J. J.; Wiesner,
J. B.; Schanzer, J. M.; Erion, M. D. J. Med. Chem. 2000, 43, 2894-2905.
(19) Vorbru¨ggen, H.; Bennua, B. Chem. Ber. 1981, 114, 1279-1286.
174 J. Org. Chem., Vol. 72, No. 1, 2007

Approach to Nucleoside Library Synthesis
These include the following: (1) times reported for conducting
the reaction generally vary from 1 to 24 h depending on the
base; (2) lack of solubility of some bases in the reaction solvent
often results in reactions with poor reproducibility; (3) a
necessary basic aqueous workup to neutralize the acidic mixture
is needed; (4) there is lack of solubility of some products in
the extraction solvent; and (5) there is a necessity for chro-
matographic purification. This paper will explain how several
modern technologies, especially microwave assisted synthesis
and automated silica gel MPLC, have been applied to resolve
these issues and convert this into a high-throughput process.
Repeating this glycosylation with an oil bath heating apparatus
with an internal reaction temperature of 130 °C (150 °C bath
temperature) for 5 min produced 5a in 65% yield as compared
to a 61% yield obtained using microwave heating. Thus, the
microwave advantage is chiefly the safe application of homo-
geneous heat to the reaction process as well as the automation
capabilities inherent within the instrument. Deprotection of the
acyl groups with 6 M ammonia in methanol (50 °C, 16 h)
completed the nucleoside synthesis. Interestingly, the latter
reaction could not be accelerated effectively with the application
of higher microwave induced temperatures. When the ammo-
niolysis was conducted in the microwave reactor at 150 °C, the
pressure limit of the microwave reaction vial (20 bar) was
reached. After 10 min under these conditions, the deprotection
of 5a was incomplete, and 5′-O-benzoyladenosine (5b)25,26 and
adenosine (6) were isolated in 22 and 46% yields, respectively
(Scheme 1).
Chemistry
Elimination of the aqueous workup became an early priority.
The polarity of some bases and products restricted their
solubility and made conducting extractions in parallel cumber-
some and inconsistent across a wide variety of substrates. As
an alternative to extraction, neutralization of the trifluo-
romethanesulfonic acid (TfOH)20 generated from hydrolysis of
trimethylsilyl triflate was performed prior to direct silica gel
chromatographic purification of the reaction mixture. Triethyl-
amine was found to be inadequate since the resulting TfOH
triethylammonium salt contaminated chromatographed products
Results and Discussion
In the original report of the Vorbru¨ggen one-step method,19
adenosine (6), guanosine (15), cytidine (44), and compound 49
were prepared in 63, 44, 59, and 50% yields, respectively. By
comparison, using the 5 min/130 °C microwave Vorbru¨ggen
glycosylation reaction followed by deprotection, nucleosides 6,
15 (as part of a N9/N7 mixture with 16), 44, and 49 were
prepared in 45, 26, 50, and 51% yields, respectively (Table 1).
Other natural nucleosides such as inosine (8 as part of an N9/
N7 isomer mixture with 9) and uridine (46) were also readily
obtained by this method, each in 36% yield. The generality and
high-throughput of the method greatly compensate for the lower
yields observed in certain cases. When coupled with a simple
NH3/methanol deprotection reaction and short automated MPLC
purification processes, a high-throughput synthesis of nucleoside
libraries was realized (see Scheme 2 for method details).
The breadth of the method is evident from the large variety
of bases that was successfully ribosylated (see Table 1 and
Supporting Information). The 48 bases tested as glycosylation
substrates included 25 purines, four pyrazolopyrimidines, two
8-azapurines, one 2-azapurine, two imidazopyridines, two
1
(as observed from H NMR). Next, the ion-exchange resin
Amberlite IRA-400 OH- was found to effectively sequester
TfOH as a resin bound salt, but it was also found to retain certain
hydrophobic products as well. This resulted in decreased yields
with some substrates and a complete loss of the glycosylation
product for one example (entry 4, Table 1). Finally, triethanol-
amine (TEOA) in acetonitrile containing 2% water was deter-
mined to be the most effective acid quenching reagent since
the polar nature of the ammonium salt produced resulted in a
clean chromatographic separation from the glycosylation reac-
tion product in all examples studied.
To accelerate the chromatographic process, a commercially
available automated medium-pressure liquid chromatography
(MPLC) instrument with UV based fraction collection was
utilized for the purification of both the glycosylation and the
deprotection products.21 Moreover, for some very polar nucleo-
sides, MPLC was only necessary for the glycosylation product.
In these examples, the ammoniolysis products precipitated
directly from the reaction solution (entries 5, 7, and 8 in Table
1).
(23) A silica supported fusion glycosylation in a standard kitchen
microwave is the only report of a microwave assisted glycosylation:
Andrzejewska, M.; Kaminski, J.; Kazimierczuk, Z. Nucleosides, Nucleotides,
Nucleic Acids 2002, 21, 73-78.
We investigated conducting the Vorbru¨ggen glycosylation in
the one-step format19 using elevated temperatures generated by
microwave heating22 and observed that reactions reported to
require up to 24 h in refluxing acetonitrile (82 °C) could be
conducted satisfactorily in 5 min at 130 °C (see Table 1 for
examples and the Supporting Information for a more extensive
list).23, 24 At this elevated temperature, even very polar bases
were solubilized as their silylated derivatives. This resulted in
more consistent reactivities across a diverse group of bases.
(24) Use of microwave heating for nucleoside functional group manipu-
lations: (a) Varma, R. S.; Lamture, J. B.; Varma, M. Tetrahedron Lett.
1993, 34, 3029-3032. (b) Kumar, P.; Gupta, K. C. Chem. Lett. 1996, 8,
635-636. (c) Kumar, P.; Gupta, K. C. Nucleic Acids Res. 1997, 25, 5127-
5129. (d) Tschamber, T.; Rudyk, H.; Nouen, D. L. HelV. Chim. Acta 1999,
82, 2015-2019. (e) Gorska, A.; Andrzejewska, M.; Kaminski, J.; Kazim-
ierczuk, Z. Nucleosides, Nucleotides, Nucleic Acids 2003, 22, 13-19. (f)
Paolini, L.; Petricci, E.; Corelli, F.; Botta, M. Synthesis 2003, 1039-1042.
(g) Grunefeld, P.; Richert, C. J. Org. Chem. 2004, 69, 7543-7551.
1
(25) Identified based on comparison of H NMR spectrum (300 MHz,
DMSO-d6) with reported values: (a) Ishido, Y.; Nakazaki, N.; Sakairi, N.
J. Chem. Soc., Perkin Trans. 1 1979, 2088-2098. (b) Maury, G.; Daiboun,
A.; Elalaoui, A.; Genu-Dellac, C.; Perigaud, C.; Bergogne, C.; Gosselin,
G.; Imbach, J.-L. Nucleosides Nucleotides 1991, 10, 1677-1692.
(26) Other reports have documented the selective deprotection of tri-
and tetra-benzoyl-protected nucleosides to yield 5′-O-benzoyl-protected
nucleosides using hydrazine hydrate, ammonia, sodium methoxide, or
lithium 2,2,2-trifluoroethoxide as the deacylating reagent. See ref 25a and
(a) Ferris, J. P.; Devadas, B.; Huang, C.-H.; Ren, W.-Y. J. Org. Chem.
1985, 50, 747-754. (b) Nishino, S.; Rahman, A.; Takamura, H.; Ishido,
Y. Tetrahedron 1985, 41, 5503-5506. (c) Zerrouki, R.; Roy, V.; Hadj-
Bouazza, A.; Krausz, P. J. Carbohydr. Chem. 2004, 23, 299-303. (d)
Nowak, I.; Jones, C. T.; Robins, M. J. J. Org. Chem. 2006, 71, 3077-
3081.
(20) Direct evaporation of the reaction mixture prior to chromatographic
purification decomposed some reaction products, resulting in decreased
yields.
(21) The automated silica gel MPLC process purified most glycosylation
products to homogeneity. However, a few contained impurities related to
excess sugar byproducts or excess base. These were conveniently removed
with automated silica gel MPLC or precipitation of the nucleoside product
after the deacylation reaction so that the final nucleoside product could be
obtained in high purity.
(22) Review: Kappe, C. O. Angew. Chem., Int. Ed. 2004, 43, 6250-
6284.
J. Org. Chem, Vol. 72, No. 1, 2007 175

Bookser and Raffaele
SCHEME 2. High-Throughput Nucleoside Synthesis Method Summary
TABLE 1. Examples of High-Throughput Nucleoside Synthesis
176 J. Org. Chem., Vol. 72, No. 1, 2007

Approach to Nucleoside Library Synthesis
Table 1 (Continued)
J. Org. Chem, Vol. 72, No. 1, 2007 177

Bookser and Raffaele
Table 1 (Continued)
a R ) â-D-ribofuranos-1-yl. b The quench method involves addition of an acid quenching reagent: TEOA ) triethanolamine and resin ) Amberlite
IRA-400 OH-. c Yields were weight based. To determine the yield for a compound represented as part of an isomer mixture, the overall weight based yield
for the mixture was divided into individual component yields based on integral analysis of characteristic protons within the 1H NMR spectrum of the
mixture. d As indicated, these nucleosides and their ribofuranose and glucopyranose analogues have or have not previously been reported as products of a
glycosylation/deprotection sequence conducted with the respective base according to a search of American Chemical Society SciFinder and Beilstein Crossfire
databases. See Supporting Information for complete literature. e Compounds 8 and 9 were obtained as an inseparable mixture in 88% combined overall
purity (two major HPLC peaks summed). f No glycosylation product was isolated from resin workup method. g Compounds 15 and 16 were obtained as an
inseparable mixture in 96% combined overall purity (two major HPLC peaks summed). h Compound isolated by collecting precipitate from NH3/methanol
deprotection reaction mixture. i Compounds 25 and 26 were obtained as an inseparable mixture in 96% combined overall purity (two major HPLC peaks
summed). j Compounds 30-32 were obtained as an inseparable mixture in 100% combined overall purity (three major HPLC peaks summed). k Complex
mixture. Products not isolated. l Compounds 51 and 52 were obtained as an inseparable mixture in 75% combined overall purity (two major HPLC peaks
summed).
benzimidazoles,threeimidazoles,three1,2,4-triazoles,twopyrimi-
dines, two 3-deazapyrimidines, one quinazolinedione, and one
alloxazine. Notably, 15 examples (11, 18, 20, 22, 23, 25, 26,
28, 30-32, 36, 40, 42, and 52) have not been reported as direct
products of glycosylation but were previously synthesized from
multistep base modification of preformed nucleosides (for back-
ground literature, see Supporting Information). Thus, not only
does this method provide faster access to these specific nucleo-
sides, but it provides an impetus to check if other novel nucleo-
sides not previously considered as obtainable as direct products
of glycosylation might be prepared in this simple manner.
Among bases tested as potential novel Vorbru¨ggen glycosyl-
ation substrates, three purine aza-isomers failed to yield isolable
nucleosides (entries 12, 14, and 15 in Table 1). Complex
mixtures were produced that were not amenable to purification
by the general method, even though mass spectral evidence of
product formation (34, 38, and 40, respectively) was obtained
in each case. This is not entirely surprising since one of the
bases that failed, 33, has been documented to produce a mixture
of five glycosylation products.27 Another, 37, has only been
reported to work by a sodium hydride mediated glycosylation.28
Finally the third base, 39, does not appear in the literature as a
glycosylation substrate.
(27) Lichtenthaler, F. W.; Cuny, E. Chem. Ber. 1981, 114, 1610-
1623.
178 J. Org. Chem., Vol. 72, No. 1, 2007

Approach to Nucleoside Library Synthesis
The products were identified by 1H NMR and HPLC
comparison with authentic samples when available and by
comparison with literature H NMR values otherwise. Purity
rapid access to a collection of nucleosides that would have
required considerably more effort were it constructed based only
on previously published methods.
1
was determined by HPLC and 1H NMR (see Supporting
Information for identification method, chromatograms, and
spectra). Of the 48 examples studied, 32 glycosylations yielded
a single regioisomer product, and six (one of which is included
in Table 1, entry 8; see Supporting Information for others)
resulted in separable mixtures of regioisomers. The isolated
single products (totaling 45) were obtained in an average two-
step overall yield ( SD of 26 ( 16% and an average HPLC
purity ( SD of 95 ( 6%.
Mixtures of -two to three regioisomers were obtained from
seven examples (five of which are included in Table 1: entries
2, 5, 9, 11, and 20). In the case of entry 20, base 50 was
previously glycosylated by a different method at lower tem-
peratures where only the N1 isomer 51 was produced.29 The
N3 isomer 52 was prepared previously by a more lengthy
route.30 Other examples have been previously documented to
produce mixtures or are novel glycosylation reactions (see
Supporting Information for complete literature).
The purine series was studied the most thoroughly and
includes examples with base substitution combinations contain-
ing one or more of the following groups at every available
position: H, NH2, NH-alkyl, dO, O-alkyl, SH, S-alkyl, and
methyl (see Supporting Information). A few polar purine bases
such as 2-N-acetylguanine (14), isoguanine (19), and adenine-
8-thiol (21) were essentially insoluble in the reaction medium
at room temperature but became solubilized as trimethylsilyl
derivatives at 130 °C. This indication of a solubility limitation
was used as an advantage. The intermediate glycosylation
products were soluble and readily purified by MPLC. Upon
NH3/methanol deprotection, the respective product nucleosides
(15, 16, 20, and 22) precipitated out of the reaction medium,
thereby simplifying their isolation.
Experimental Procedures
N6-Benzoyladenosine 2′,3′,5′-Tri-O-benzoate (5a). General
Method for Microwave Mediated Vorbru1ggen Glycosylation.
To a 5 mL microwave reaction vial with a stir bar was added 1-O-
acetyl-2,3,5-tri-O-benzoyl-â-D-ribofuranose (3) (150 mg, 0.30
mmol), N6-benzoyladenine (1) (107 mg, 0.45 mmol), 3 mL of
acetonitrile, N,O-bis(trimethylsilyl)acetamide (BSA) (0.328 mL,
1.34 mmol), and trimethylsilyl triflate (TMSOTf) (0.161 mL, 0.90
mmol). The vial was sealed with a mechanically crimped septum
cap and heated in a Biotage microwave reactor at 130 °C for 5
min. The contents were transferred to a 40 mL glass vial and rinsed
with 2 mL of acetonitrile. To this was added 1.33 mL of a 1 M
solution of triethanolamine (1.33 mmol) in acetonitrile with 2%
water, and the vial was shaken at room temperature for 30 min.
The volatiles were removed by evaporation, and the residue was
dissolved in 4 mL of 1:1 MeOH/CH2Cl2 and adsorbed onto 3 g of
silica gel by evaporation. This was transferred to a plastic precolumn
and subjected to automated MPLC through a 12 g column of silica
gel eluting at 30 mL/min with a gradient of 0-15% MeOH/CH2-
Cl2 over 30 min to provide 123 mg (61%) of 5a as an amorphous
solid: 1H NMR (300 MHz, DMSO-d6) δ 4.68 (dd, 1H, J ) 12, 5
Hz), 4.82 (dd, 1H, J ) 12, 4 Hz), 4.89 (q, 1H, J ) 5 Hz), 6.30 (t,
1H, J ) 6 Hz), 6.55 (dd, 1H, J ) 6, 5 Hz), 6.69 (d, 1H, J ) 5 Hz),
7.4-7.7 (m, 12H), 7.9-8.1 (m, 8H), 8.62 (s, 1H), 8.73 (s, 1H),
11.26 (s, 1H); HPLC purity (254 nm) ) 94%.
Alternatively, instead of quenching the acid with triethanolamine,
the crude reaction mixture was added to 1.5 g of Amberlite IRA-
400 (OH-) suspended in 6 mL of CH2Cl2, and the mixture was
shaken for 30 min, filtered, and evaporated. The residue was then
subjected to MPLC as described previously.
Adenosine (6). General Method for NH3/Methanol Depro-
tection. The protected nucleoside 5a (117 mg, 0.171mmol) was
transferred to a 40 mL vial and dissolved in 2 mL of 6 M NH3 in
methanol. This vial was sealed with a screw cap and shaken at
50 °C for 16 h. The solvent was evaporated, and the residue was
dissolved in 1:1 MeOH/CH2Cl2 and adsorbed onto 1 g of silica gel
by evaporation. This was transferred to a plastic precolumn and
subjected to automated MPLC through a 12 g column of silica gel
eluting at 30 mL/min with a gradient of 0-20% MeOH/CH2Cl2
over 30 min to provide 34 mg (74%, 45% overall for two steps) of
adenosine (6) as a solid that had HPLC Rt (purity at 254 nm )
Conclusion
A general, unified two-step nucleoside synthesis method was
developed, which is amenable to the combination of various
bases with various sugars for the rapid preparation of structurally
diverse nucleoside libraries.31 Key features of the method are
the following: (1) a one-step 5 min/130 °C microwave
Vorbru¨ggen glycosylation reaction; (2) single-phase quenching
of TfOH with triethanolamine; and (3) 10-25 min generic,
normal phase silica gel MPLC purification protocols. The
method has been demonstrated with the preparation of a diverse
nucleoside library that contained 15 examples not previously
obtained directly from a glycosylation/deprotection sequence
on a commercially available base. Thus, this method provides
1
100%) and H NMR values identical to an authentic sample.
Compounds 15, 16, 20, and 22 were isolated by filtration from
the crude reaction mixture as precipitated solids.
Acknowledgment. This work was supported in part by NIH
SBIR Grant AI050278.
(28) Eldrup, A. B.; Prhavc, M.; Brooks, J.; Bhat, B.; Prakash, T. P.;
Song, Q.; Bera, S.; Bhat, N.; Dande, P.; Cook, P. D.; Bennett, C. F.; Carroll,
S. S.; Ball, R. G.; Bosserman, M.; Burlein, C.; Colwell, L. F.; Fay, J. F.;
Flores, O. A.; Getty, K.; LaFemina, R. L.; Leone, J.; MacCoss, M.;
McMasters, D. R.; Tomassini, J. E.; Von Langen, D.; Wolanski, B.; Olsen,
D. B. J. Med. Chem. 2004, 47, 5284-5297.
Supporting Information Available: Extended table of ex-
amples including those presented here, with a method of product
identification and related synthesis literature and 1H NMR, HPLC,
and MS data for all deprotected nucleosides. This material is
available free of charge via the Internet at http://pubs.acs.org.
(29) (a) Stout, M. G.; Robins, R. K. J. Org. Chem. 1968, 33, 1219-
1225. (b) Schweizer, M. P.; Banta, E. B.; Witkowski, J. T.; Robins, R. K.
J. Am. Chem. Soc. 1973, 95, 3770-3778. (c) Dunkel, M.; Pfleiderer, W.
Nucleosides Nucleotides 1992, 11, 787-819. (d) Dunkel, M.; Pfleiderer,
W. Nucleosides Nucleotides 1993, 12, 125-137. (e) Bhattacharya, B. K.;
Chari, M. V.; Durland, R. H.; Revankar, G. R. Nucleosides Nucleotides
1995, 14, 45-63.
JO061885L
(31) This process, known as Nucapalooza in our labs, has been applied
successfully to the production of more than 400 nucleosides from the
reaction of six different acylated ribofuranosyl derivatives with a set of
bases which included those listed in Table 1 and the Supporting Information
Table S1.
(30) Girniene, J.; Apremont, G.; Tatibouet, A.; Sackus, A.; Rollin, P.
Tetrahedron 2004, 60, 2609-2619.
J. Org. Chem, Vol. 72, No. 1, 2007 179