Efficient large scale synthesis of 2′-O-alkyl pyrimidine ribonucleosides

Article abstract of DOI:10.1021/op990100t

An efficient process to synthesize 2′-O-alkyl pyrimidine ribonucleosides in high yield has been described. The inexpensive method was used on a multikilogram-scale synthesis and optimized reaction conditions have been investigated.

Full text of DOI:10.1021/op990100t

Organic Process Research & Development 2000, 4, 170−171
Efficient Large Scale Synthesis of 2′-O-Alkyl Pyrimidine Ribonucleosides
Saroj K. Roy*,† and Jin-yan Tang
Hybridon Inc., 155 Fortune BouleVard, Milford, Massachusetts 01757
Scheme 1
Abstract:
An efficient process to synthesize 2′-O-alkyl pyrimidine ribo-
nucleosides in high yield has been described. The inexpensive
method was used on a multikilogram-scale synthesis and
optimized reaction conditions have been investigated.
Results and Discussion
The presence of 2′-O-alkyl ribonucleosides in oligonucle-
otides has been shown both in vitro and in vivo to provide
a better stability and also strong binding capability to the
complementary gene targets. Particularly 2′-O-methyl modi-
fied nucleosides have been widely used in oligonucleotides
synthesis for the purpose of research and future therapeutic
application.1-3 Several 2′-O-methyl nucleoside modified
oligonucleotides are at present in clinical trials including the
antiangiogenic ribozyme molecule, RPI 4610.4
We require a large quantity of 2′-O-methyl uridine (10
kg) for phase II trials of GEM 231.5 2′-O-methyl uridine
has been prepared by using following two methods (Schemes
1 and 2). Scheme 1 would involve N3-protection by benzoyl
or toluyl, and 3′,5′-O-protected with TIPDSi-Cl ribonucleo-
sides have been subjected to alkylation of the ribose hydroxyl
group using methyl iodide and silver oxide.6-9 The overall
yield after deportation of N3 and 3′,5′-O-TIPDSi ribonucleo-
sides was 13%. The three purifications at intermediate level
made the process labor-intensive.
The second method also needed the 3′,5′-O-protection
with TIPDSi-Cl followed by Pummerer rearrangement. The
recovery of 6 after borohydride reduction and deportation
of 9 was 18% (Scheme 2). These methods thereby increase
Scheme 2
* Author for correspondence. E-mail: saroj_roy@pbio.com.
† Present address: PE Biosystems, 500 Old Connecticut Path, Framingham,
MA 01701.
(1) (a) Cummins, L.; Owens, S.; Risen, L.; Lesnik, E.; Freier, S.; McGee, D.;
Guinosso, C.; Dan Cook, P. Nucleic Acids Res. 1995, 23, 2019-2024. (b)
Lesnick, E.; Guinosso, C.; Kawasaki, A.; Sasmor, H.; Zounes, M.;
Cummins, L.; Ecker, D.; Dan Cook, P.; Freier, S. Biochemistry 1993, 32,
7832-7838. (c) Kawai, G.; Yamamoto, Y.; Kamimura, T.; Masegi, T.;
Sekine, M.; Hata, T.; Iimori, T.; Watanabe, T.; Miyazawa, T.; Yokoyama,
S. Biochemistry 1992, 31, 1040-1046. (d) Sproat B.; Lamond A.; Beijer
B.; Neuner P.; Ryder U. Nucleic Acids Res. 1989, 17, 3373-3386.
(2) Crooke, S. T., Lebleu, B., Eds. Antisense Research and Applications; CRC
Press: Boca Raton, Florida, 1993.
(3) (a) Goodchild, J. Nucleic Acids Res. 1992, 20, 4607-4612. (b) Agrawal,
S., Ed. Antisense Therapeutics; Humana Press: Totowa, New Jersey, 1996.
(4) Sandberg, J.; Bouhana, K.; Gallegos, A.; Agrawal, A.; Grimm, S.; Wincott,
F.; Reynolds, M.; Pavco, P.; Parry, T. Antisense Nucleic Acid Drug DeV.
1999, 9, 271-277.
the time and expense of synthesis while decreasing its
efficiency and overall yield. The poor recovery, expensive
reagents, and labor-intensive purification caused us to
investigate high-yielding, inexpensive, and practical method
amenable to a large-scale operation.
In this paper we report a simple, practical, and optimized
method for the synthesis of 6 in large scale. The synthesis
has been done in two steps via anhydronucleoside opening
by a nucleophile (Scheme 3). This method provides efficient
(5) Wang, H.; Cai, Q.; Zeng, X.; Yu, D.; Agrawal, S.; Zhang, R. Proc. Natl.
Acad. Sci. U.S.A. 1999, 96, 13989-13904.
(6) (a) Haines, A. Tetrahedron 1973, 29, 2807-2810. (b) Robins, M.; Naik,
S.; Lee, S. J. Org. Chem. 1974, 39, 1891-1899.
(7) Yamauchi, K.; Nakagima, T.; Kinoshita, M. J. Chem. Soc., Perkin Trans.
1 1980, 2787-2792.
Scheme 3
(8) Wagner, E.; Oberhauser, B.; Holzner, A.; Brunar, H.; Issakides, G.;
Schaffner, G.; Cotten, M.; Knollmuller, M.; Noe, C. Nucleic Acids Res.
1991, 19, 5965-5971.
(9) (a) Srivastva, S.; Roy, S. U.S. Patent 5,214,135, 1993. (b) Beijer, B.; Grotli,
M.; Douglas, M.; Sproat, B. Nucleosides Nucleotides 1994, 13, 1905-
1927. (c) Hodge, R.; Sinha, N. Tetrahedron Lett. 1995, 36, 2933-2936.
170
•
Vol. 4, No. 3, 2000 / Organic Process Research & Development
10.1021/op990100t CCC: $19.00 © 2000 American Chemical Society and The Royal Society of Chemistry
Published on Web 02/16/2000

synthesis of 2′-O-CH3 substituted pyrimidine ribonucleosides.
Synthesis of 10 has been carried out using diphenyl carbonate
in DMF and with sodium bicarbonate in catalytic amount,
which selectively gave the desired product in 90%. The
product crystallized out from the reaction mixture during 12
h of stirring at room temperature. The crystallized material
was 98% pure by HPLC. The second step was the ring
opening of 2,2′-anhydrouridine using 4 equiv of magnesium
methoxide, which afforded the desired product.
the starting material and the formation of a polar product
was observed by in-process analysis with HPLC. The product
slowly crystallized out from the reaction mixture. The
mixture was stirred overnight at rt. The solid was filtered
and washed with methanol. The product was 98% pure by
RP-HPLC. The material was dried in a high-vacuum oven
to its constant weight 1.69 kg (90% yield). The product was
identified by comparison with an authentic sample purchased
from Aldrich.
Magnesium methoxide is a weak base, which eliminates
the possibility of forming the epoxide and consequently the
formation of the arabinofuranosyl structure. Hence, the only
possible reaction is the attack at C-2′ and the formation of
a ribofuranosyl structure. 2,2′-Anhydrocytidine HCl was also
reacted with magnesium methoxide to give the 2′-O-methyl
cytidine in >76% recovery.
Preparation of 2′-O-Methyl Uridine (6). 2,2′-Anhy-
drouridine (10, 1.5 kg, 6.63 mol) was added to a freshly
prepared solution of 12% magnesium methoxide (18.64 L,
25.88 mol) in methanol. The mixture was refluxed for 5 h.
The commercially available magnesium methoxide did not
work efficiently. The reaction was monitored by HPLC. The
starting material was consumed in 5 h and the formation of
a comparatively nonpolar peak observed by in-process
analysis. The mixture was cooled to room temperature and
then to 5 °C. The pH was adjusted to 7 by the addition of
glacial acetic acid. The solvent was striped off to foam. The
resultant foam was refluxed with ethyl alcohol (15 L) for 2
h to separate the desired product from inorganic impurities.
The solid (magnesium salts) was filtered off and discarded.
The filtrate was striped off to a solid mass to give compound
6. The isolated yield of the compound 6 was 1.46 kg (92%).
The product was >95% pure by HPLC. The yield after
Conclusions
The method comprises the preparation of 2,2′-anhydro-
uridine and reaction with freshly prepared Mg(OCH3)2 in
the corresponding alcohols to produce 2′-O-CH3-substituted
uridine and cytidine. Commercially available Mg(OCH3)2
was not successful (see Experimental Section). A further
advantage of the present method is that the alkylation is
essentially complete within 5 h.10-12
This improved method of alkylation through ring opening
of anhydro ribonucleosides makes it possible to synthesize
a broad range of 2′-O-substituted pyrimidine nucleosides via
a divalent alkoxide acting as a nucleophile of the corre-
sponding alcohol.
1
quantitation against uridine was 89%. H NMR (300 MHz,
D2O) δ 7.85 (d, J ) 8.09), 5.90 (d, J ) 3.36), 5.82 (d, J )
7.93), 4.27 (m, 1H), 4.02 (m, 1H), 3.97 (m, 1H), 3.45 (s,
3H), 3.76-3.84 (m, 2H).
Preparation of 2′-O-Methyl Cytidine. 2′-O-Methyl cy-
tidine was synthesized similarly by using 2,2′-anhydrocyti-
dine HCl (3.1 g, 3.82 mmol) and freshly prepared solution
of 10% magnesium methoxide (16.25 mL, 22.17 mmol) in
methanol.
The mixture was refluxed to 5 h. The workup was similar
to that described above, and the isolated yield after quanti-
tation against cytidine was 752 mg (76%). The product was
>93% pure by HPLC. 1H NMR (300 MHz, D2O) δ 7.82 (d,
J ) 7.48), 5.91 (d, J ) 3.33), 6.01 (d, J ) 7.33), 4.24 (m,
1H), 4.04 (m, 1H), 3.95 (m, 1H), 3.47 (s, 3H), 3.76-3.85
(m, 2H).
Experimental Section
All of the starting materials are commercially available
from Aldrich (Madison, WI), except that 12% magnesium
methoxide that was prepared freshly. HPLC analysis was
performed on a Delta-Pak C18 Column (300 Å, 3.9 × 150
mm) by using a linear gradient of A (0.1 M ammonium
acetate in H2O) and B (80% CH3CN in 20% buffer A) at a
flow rate 1.0 mL/min with UV detection at 240 nm.
Preparation of 2,2′-Anhydrouridine (10). Diphenyl
carbonate (1.928 kg, 9.0 mol) was added to the uridine (1.2
kg, 8.19 mol) in 4.9 L of DMF under inert atmosphere
followed by sodium bicarbonate (20.7 g, 0.245 mol). The
mixture was heated for 3 h at 80 °C. The disappearance of
Acknowledgment
The authors thank Bonnie Cauley and Jianbo Zhang for
their provision of COSY experiments.
(10) During the course of this work, a paper published by McGee, D.; Zhai, Y.
Nucleosides Nucleotides 1996, 15, 1797-1803 reported selective alkylation
of 5′-O-dimethoxytrityl anhydrouridine with magnesium or calcium alkoxide
in DMF at 100 °C. Their attempts to react the unprotected anhydrouridine
with Mg(OCH3)2 under the same conditions gave no product.
Received for review November 19, 1999.
OP990100T
(11) Verheyden, J.; Wagner, D.; Moffatt, J. J. Org. Chem. 1971, 36, 250-254.
(12) Roy, S.; Tang, J. U.S. Patent 5,739,314, 1998.
Vol. 4, No. 3, 2000 / Organic Process Research & Development
•
171