6-(2-Alkylimidazol-1-yl)purines undergo regiospecific glycosylation at N9

Article abstract of DOI:10.1021/ol051573p

(Chemical Equation Presented) Regioselective control of glycosylation of purines at N9 (versus N7) has been a continuing challenge. We now report Lewis acid catalyzed regiospecific glycosylations of 6-(2-alkylimidazol-1-yl)purines at N9. The 6-(2-alkyl)im

Full text of DOI:10.1021/ol051573p

ORGANIC
LETTERS
2005
Vol. 7, No. 21
4601-4603
6-(2-Alkylimidazol-1-yl)purines Undergo
Regiospecific Glycosylation at N9†
Minghong Zhong, Ireneusz Nowak, and Morris J. Robins*
Department of Chemistry and Biochemistry, Brigham Young UniVersity,
ProVo, Utah 84602-5700
morris_robins@byu.edu
Received July 5, 2005
ABSTRACT
Regioselective control of glycosylation of purines at N9 (versus N7) has been a continuing challenge. We now report Lewis acid catalyzed
regiospecific glycosylations of 6-(2-alkylimidazol-1-yl)purines at N9. The 6-(2-alkyl)imidazole moiety also functions as a versatile leaving group
that can be replaced by nucleophiles (SNAr) and aryl groups (Suzuki cross-coupling).
The most challenging aspects of nucleoside synthesis by
glycosylation involve the achievement of concomitant regio-
and stereocontrol. Alkylation or glycosylation of purines is
rarely regiospecific, and both usually produce mixtures of
N9 and N7 (and even N3 and/or N11) products. Factors
known to influence N9/N7 ratios include the composition
of the alkylating reagent or glycosyl donor, the purine, and
the reaction conditions.2,3 Numerous studies have pursued
ways to maximize N9 glycosylation, but few have resulted
in systematic improvements. Regioselective coupling of
adenine, 6-chloropurine, and 2,6-dichloropurine with ribo-
furanose donors under catalysis by SnCl4 was reported by
Saneyoshi and Satoh,4 who did not detect N1, N3, or N7
isomers by TLC. However, reaction times were lengthy, and
emulsion formation during workup was problematic. Bulky
protecting groups on C6 substituents have been used to
improve selectivity for N9 glycosylation. Geen et al. reported
that N9/N7 ratios were increased significantly by expansion
from an O6-methyl to an O6-isopropyl group.5 Benner and
co-workers have employed 2-N-isobutyryl-6-O-[p-nitro-
phenyl)ethyl]guanine for regioselective alkylation and glyco-
sylation at N9.6 Robins et al.7 and Timar et al.8 have achieved
highly regioselective N9 glycosylations and alkylations with
the 6-O-diphenylcarbamoyl (6-O-DPC) group. However, the
sensitive 6-O-DPC group was less effective for smaller
electrophiles and for acidic glycosyl donors under Lewis acid
conditions.9 Therefore, a serious need for methods to exclude
N7 alkylation and glycosylation remained.
Five-membered heteroaryl rings have been introduced at
C6 of purines for temporary protection of the N6 amino
group, for leaving groups, and for biological studies.10
Imidazole and 1,2,4-triazole rings have been used as replace-
able substituents at C6 of purines and C4 of pyrimidines.
(5) Geen, G. R.; Grinter, T. J.; Kincey, P. M.; Jarvest, R. L. Tetrahedron
1990, 46, 6903-6914.
(6) (a) Jenny, T. F.; Previsani, N.; Benner, S. A. Tetrahedron Lett. 1991,
32, 7029-7032. (b) Jenny, T. F.; Benner, S. A. Tetrahedron Lett. 1992,
33, 6619-6620.
(7) (a) Zou, R.; Robins, M. J. Can. J. Chem. 1987, 65, 1436-1437.
(b) Robins, M. J.; Zou, R.; Guo, Z.; Wnuk, S. F. J. Org. Chem. 1996, 61,
9207-9212.
† Nucleic Acid Related Compounds. 128. For part 127, see: Nowak, I.;
Conda-Sheridan, M.; Robins, M. J. J. Org. Chem. 2005, 70, 7455-7458.
(1) For a recent example, see: Guillarme, S.; Legoupy, S.; Bourgougnon,
N.; Aubertin, A.-M.; Huet, F. Tetrahedron 2003, 59, 9635-9639.
(2) (a) Rasmussen, M.; Chan, J. H.-S. Aust. J. Chem. 1975, 28, 1031-
1047. (b) Liu, M.-C.; Kuzmich, S.; Lin, T.-S. Tetrahedron Lett. 1984, 25,
613-616.
(8) Timar, Z.; Kovacs, L.; Kovacs, G.; Schmel, Z. J. Chem. Soc., Perkin
Trans. 1 2000, 19-26.
(9) Zhong, M.; Robins, M. J. Tetrahedron Lett. 2003, 44, 9327-9330
and references therein.
(10) Estep, K. G.; Josef, K. A.; Bacon, E. R.; Carabateas, P. M.; Rumney,
S., IV.; Pilling, G. M.; Krafte, D. S.; Volberg, W. A.; Dillon, K.; Dugrenier,
N.; Briggs, G. M.; Canniff, P. C.; Gorczyca, W. P.; Stankus, G. P.; Ezrin,
A. M. J. Med. Chem. 1995, 38, 2582-2595 and references therein.
(3) Vorbru¨ggen, H.; Ruh-Pohlenz, C. Org. React. 2000, 55, 1-630.
(4) Saneyoshi, M.; Satoh, E. Chem. Pharm. Bull. 1979, 27, 2518-2521.
10.1021/ol051573p CCC: $30.25
© 2005 American Chemical Society
Published on Web 09/15/2005

Recently, we have elaborated the 6-amino group of adenine-
type nucleosides into 6-(1,2,4-triazol-4-yl)11 and 6-(2,
5-dimethylpyrrol-1-yl) moieties12 and employed a modified
Appel reaction for introduction of an (imidazol-1-yl) ring at
C6 of purine nucleosides.13 Both theoretical calculations and
X-ray crystal structures indicate that the two heteroaromatic
rings of 6-(imidazol-1-yl)purines are essentially coplanar.
This causes apparent shielding of N7 by H5′ of the imidazole
ring, which should result in highly regioselective N9 glyco-
sylation.
Table 1. Glycosylations of 6-(2-Propylimidazol-1-yl)purinea
The purine sodium salt glycosylation procedure14 is an
important nucleoside-forming methodology, but N7 isomer
formation is problematic. Our attempts to couple sodium salts
of certain substituted 6-(imidazol-1-yl)purines with either
2,3,5-tri-O-benzoyl-D-ribofuranosyl chloride or 2,3,4,6-tetra-
O-benzoyl-R-D-glucopyranosyl bromide were not promising.
We then examined Vorbru¨ggen-type glycosylations of 6-(im-
idazol-1-yl)purine (1a) and 1-O-acetyl-2,3,5-tri-O-benzoyl-
â-D-ribofuranose (2a) with SnCl4 in acetonitrile4 or pretri-
methylsilylated 1a (BSA/DCE) with TMSOTf in toluene
(Scheme 1). The SnCl4-catalyzed glycosylation went to
a Reactions were performed by the general procedure (Supporting
Information). b Hours. c Percent isolated. d No reaction.
Scheme 1. Glycosylation of 6-(Imidazol-1-yl)purines 1 with
2a
same conditions (entry 7), which is likely attributable to the
less favorable formation of an oxocarbenium ion with
pyranosides and different activation energy barriers.15 Con-
version of the anomeric acetate of 2d to the pyranosyl
bromide16 of 2e, and use of SnCl4 as catalyst, resulted in
successful coupling to give the desired product 3d (83%) in
good yield upon extension of the reaction time to 9 h (entry
9).
We also subjected the reactants to coupling with TMSOTf
as catalyst. This one-step, one-pot procedure employed
TMSOTf (in DCE) both for trimethylsilylation of 6-(2-
propylimidazol-1-yl)purine (1b) and for catalytic activation
of the glycosyl donors. Couplings were complete within 1.5
h at ambient temperature with furanosyl donors 2a-c (entries
2, 4, and 6). The isolated yields in these cases (∼80%) were
comparable to those under catalysis with SnCl4 and were
completed in less than half the time. However, TMSOTf-
catalyzed coupling was not observed with the pyranosyl
donors 3d (entry 8) and 3e (entry 10).
It is noteworthy that glycosylations of silylated purines
under Vorbru¨ggen conditions typically give lower (60-70%)
yields and mixtures of regioisomers,17 whereas the Vor-
bru¨ggen methodology usually gives high yields of the desired
N1 isomer with silylated pyrimidines.3 Our present proce-
dures combine ambient temperature conditions, good yields,
and versatility with regiospecific coupling of purines. Ad-
ditionally, we have developed a modified workup of reactions
that employ SnCl4 to avoid formation of the troublesome
completion, and the coupling product 3 (R ) H) was formed
in high yield as single regioisomer. However, addition of
water to the imidazole ring occurred to give 4 (diastereomeric
mixture). Chromatography of a TMSOTf-catalyzed reaction
mixture gave a salt of 3 (R ) H) that produced the hydrate
4 (81%) upon neutralization with NaHCO3/H2O.
We reasoned that replacement of imidazole per se by a
2-alkylimidazole at C6 of the purine ring might prevent
hydration of the 6-(2-alkylimidazol-1-yl)purine nucleoside
intermediates 3. Thus, 6-(2-propylimidazol-1-yl)purine (1b)
and 2a in acetonitrile with SnCl4 as catalyst gave 3a (R )
Pr) in 79% isolated yield (Table 1, entry 1). The xylosyl,
2b, and arabinosyl, 2c, donors coupled equally well under
these conditions to give nucleosides 3b (78%) (entry 3) and
3c (81%) (entry 5). In contrast, 1,2,3,4,6-penta-O-acetyl-R-
D-glucopyranose (2d) did not undergo coupling under the
(11) (a) Samano, V.; Miles, R. W.; Robins, M. J. J. Am. Chem. Soc.
1994, 116, 9331-9332. (b) Miles, R. W.; Samano, V.; Robins, M. J.
J. Am. Chem. Soc. 1995, 117, 5951-5957.
(15) Bochkov, A. F.; Zaikov, G. E. Chemistry of the O-Glycosidic
(12) Nowak, I.; Robins, M. J. Org. Lett. 2003, 5, 3345-3348.
(13) (a) Lin, X.; Robins, M. J. Org. Lett. 2000, 2, 3497-3499. (b) Janeba,
Z.; Lin, X.; Robins, M. J. Nucleosides Nucleotides Nucleic Acids 2004, 23,
137-147.
Bond: Formation and CleaVage; Pergamon: Elmsford, NY, 1979; pp 196-
197.
(16) Koenigs, W.; Knorr, E. Ber. Dtsch. Chem. Ges. 1901, 34, 957-
981.
(14) Kazimierczuk, Z.; Cottam, H. B.; Revankar, G. R.; Robins, R. K.
J. Am. Chem. Soc. 1984, 106, 6379-6382.
(17) Lichtentaler, F. W.; Voss, P.; Heerd, A. Tetrahedron Lett. 1974,
2141-2144.
4602
Org. Lett., Vol. 7, No. 21, 2005

emulsions.18 A 10-fold excess of solid NaHCO3 and a tiny
amount of H2O was stirred with the concentrated reaction
mixture containing SnCl4. The resulting suspension was
filtered using Celite, and the filter cake was washed
thoroughly. The combined filtrate was flash evaporated, and
the residue was purified by chromatography.
Table 2. Methanolysis or Deprotection of 3a
The imidazole moiety in 6-(2-alkylimidazol-1-yl)purines
can function as a leaving group for SNAr reactions with
nitrogen, oxygen, and sulfur nucleophiles13a and as a cross-
coupling partner with arylboronic acids for Suzuki reac-
tions.19 Our present methodology expands the generality of
these substitution/cross-coupling applications from examples
derived from naturally occurring purine nucleosides and
2′-deoxynucleosides13,19 to the unlimited variety of purine
and sugar combinations available through sugar-base cou-
pling.
Selective O-deacylation without displacement of the imi-
dazole moiety (Scheme 2) was achieved readily in methanolic
Scheme 2. Methanolysis or Deprotection of 3aa
a Reactions were performed at ambient temperature as described in the
Supporting Information. b Isolated yields.
both the furanosyl donors (4 h) and the pyranosyl bromide
(9 h). A modified workup procedure avoids formation of
emulsions usually encountered with SnCl4, which markedly
enhances the convenience of reactions with this catalyst. Both
procedures are mild (ambient temperature) and provide
∼80% isolated yields of only the N9 isomer, whereas
the sodium salt and Vorbru¨ggen glycosylation methods
with halopurines give both N9 and N7 regioisomers. The
6-(2-alkylimidazol-1-yl)purine nucleoside derivatives un-
dergo selective deacylation readily, and the imidazole moiety
can be displaced by nucleophiles (SNAr) or aryl groups
(Suzuki cross-coupling). Biological evaluations of these new
compounds are in progress.
a Reactions were performed at ambient temperature as described
in the Supporting Information.
ammonia at ambient temperature. Heating 3a in the same
solution at 80 °C produced mixtures of 5a and adenosine
(adenosine can be obtained exclusively at higher tempera-
tures) (Table 2). Clean formation of the 6-methoxypurine
nucleoside 5a was effected by stirring 3a with Dowex 1 ×
2 (OH-) resin in MeOH at ambient temperature.11 Analogous
substitution with other alkoxy groups can be effected with
Dowex 1 × 2 (OH-) in a solution of a 6-(2-alkylimidazol-
1-yl)purine intermediate in a different primary alcohol.
In conclusion, we have employed a (2-alkylimidazol-1yl)
substituent at C6 of the purine ring to direct glycosylation
to N9. Vorbru¨ggen procedures provide the N9 isomers
regiospecifically with SnCl4/acetonitrile or TMSOTf/DCE.
Either Lewis acid catalyst works well with the furanosyl
donors, and reactions with TMSOTf/DCE were completed
in less than half the time (1.5 h). Coupling with the pyranosyl
donors was not observed with TMSOTf/DCE under these
conditions, but the SnCl4/acetonitrile system works well for
Acknowledgment. We gratefully acknowledge NIH
Grant GM029332, pharmaceutical company gift funds
(M.J.R.), and Brigham Young University for support of this
research. We thank Professor J. F. Cannon for X-ray crystal
structure determinations.
Supporting Information Available: Experimental pro-
cedures and spectral data and 13C NMR spectra of com-
pounds 3a-d, 5a-d, and 6a-d. This material is available
free of charge via the Internet at http://pubs.acs.org.
(18) For other workup applications that avoid emulsions, see ref 13.
(19) Liu, J.; Robins, M. J. Org. Lett. 2004, 6, 3421-3423.
OL051573P
Org. Lett., Vol. 7, No. 21, 2005
4603