Efficient synthesis of 5-(thioalkyl)uridines via ring opening of α-ureidomethylene thiolactones

Full text of DOI:10.1021/jo960966m

5708
J . Org. Chem. 1996, 61, 5708-5709
Sch em e 1
Efficien t Syn th esis of 5-(Th ioa lk yl)u r id in es
via Rin g Op en in g of r-Ur eid om eth ylen e
Th iola cton es
Sengen Sun, Xiao-Qing Tang, Aziz Merchant,
P. S. R. Anjaneyulu, and J oseph A. Piccirilli*
Howard Hughes Medical Institute and Department of
Biochemistry and Molecular Biology and Department of
Chemistry, The University of Chicago, 5841 South
Maryland Avenue, Room N101, Chicago, Illinois 60637
Received J une 4, 1996
Disulfide cross-linking of thiols is a potentially valuable
tool for probing RNA structure and function. The ap-
proach has proven useful in the study of DNA1 and
DNA-protein interactions,2 and chemical strategies for
its application to RNA are beginning to emerge.3 Disul-
fide bonds have been site-specifically engineered into
RNA molecules to constrain a stem-loop conformation,3a
to determine spatial proximity of hammerhead ribozyme-
substrate complexes,3b and to engineer a minimal sub-
strate for an RNA glycosylase.3c While these examples
use the disulfide crosslink to impose conformational
constraints, it may also be possible to obtain information
about folding pathways and dynamics of RNA molecules,
analogous to studies with proteins.4 Extension of these
approaches to RNA requires the incorporation of thiol
groups at specific regions in an RNA tertiary structure
with minimal perturbation of native conformation. This
requirement necessitates the synthesis of appropriate
nucleoside monomers.
Sch em e 2
5-(Thioalkyl)uridines may be particularly useful for
this purpose. Substituents at the C-5 position of pyri-
midine nucleosides are likely to have little effect on
glycosidic torsion angle and overall nucleoside conforma-
tion5 and do not interfere directly with the hydrogen-
bonding functional groups, thereby allowing Watson-
Crick base pairing and other tertiary contacts. Synthesis
of 5-(thioalkyl)-2′-deoxyuridines from 5-(hydroxyalkyl)-
2′-deoxyuridines has recently been reported by Goodwin
and Glick.6 They introduced hydroxyalkyl substituents
at C-5 of 2′-deoxyuridines using palladium-catalyzed
addition of alkenes or alkynes to either 5-halogenated
or 5-mercurio-2′-deoxyuridines, followed by appropriate
redox chemistry. The hydroxyalkyl groups were then
transformed to thioalkyl groups in several steps. Syn-
thesis of the corresponding ribonucleosides has not yet
been described.
peracylated ribose with pyrimidines is highly efficient
and proceeds with high regio- and stereoselectivity in the
preparation of pyrimidine nucleosides,7 and (2) R-ure-
idomethylene lactones are readily isomerized to 5-(hy-
droxyalkyl)uracils.8 The possibility that the correspond-
ing thiolactones might analogously rearrange to 5-
(thioalkyl)uracils suggests a convenient route to (thio-
alkyl)uridines that does not require multistep thiol-
incorporation reaction sequences.6,9
Ureidomethylene thiolactones were prepared from
thiolactones and subsequently isomerized to uracil de-
rivatives according to Scheme 2. Butyrothiolactone (2b)
is commercially available, and valeryl lactone (2c)10 and
thiepanone (2d )11 can be easily prepared using literature
methods. While the Claisen condensations of the corre-
sponding normal lactones with methyl formate worked
adequately with sodium methoxide as a base to give the
We report here an efficient preparation of 5-(thioalkyl)-
uridines from simple, readily available starting materi-
als. Our approach (Scheme 1) is based on two observa-
tions: (1) the Hilbert-J ohnson glycosylation reaction of
* To whom correspondence should be addressed. Tel.: (312) 702-
9312. Fax: (312) 702-0271. Email: jpicciri@midway.uchicago.edu.
(1) (a) Ferentz, A. E.; Verdine, G. L. Nucleic Acids Mol. Biol. 1994,
8, 14. (b) Chaudhuri, N. C.; Kool, E. T. J . Am. Chem. Soc. 1995, 117,
10434. (c) Wang, H.; Osborne, S. E.; Zuiderweg, E. R. P.; Glick, G. D.
J . Am. Chem. Soc. 1994, 116, 5021.
(2) Stanojevic, D.; Verdine, G. L. Nature Struct. Biol. 1995, 2, 450.
(3) (a) Goodwin, J . T.; Glick, G. D. Tetrahedron Lett. 1994, 35, 1647.
(b) Sigurdsson, S. T.; Tuschl, T.; Eckstein, F. RNA 1995, 1, 575. (c)
Allerson, C. R.; Verdine, G. L. Chem. Biol. 1995, 2, 667. (d) Gao, H.;
Yang, M.; Patel R.; Cook, A. F. Nucleic Acids Res. 1995, 23, 2025.
(4) (a) Careaga, C. L.; Sutherland, J .; Sabeti, J .; Falke, J . J .
Biochemistry 1995, 34, 3048. (b) Creighton, T. E. Prog. Biophys. Mol.
Biol. 1978, 33, 231. (c) Creighton, T. E. Biochem. J . 1990, 270, 1. (d)
Weissman, J . S.; Kim, P. S. Science 1991, 253, 1368.
(7) (a) Niedballa, U.; Vorbruggen, H. J . Org. Chem. 1974, 39, 3654,
3664, 3668, 3672; (b) 1976, 41, 2084. (c) Vorbruggen, H.; Hofle, G.
Chem. Ber. 1981, 114, 1256.
(8) (a) Denny, G. H.; Ryder, M. A. J . Med. Chem. 1974, 17, 1230.
(b) Fissekis, J . D.; Sweet, F. J . Org. Chem. 1973, 28, 264.
(9) Cosstick, R.; Vyle, J . S. Nucl. Acids Res. 1990, 18, 829.
(10) Korte, F.; Lohmer, K. H. Chem. Ber. 1958, 91, 1397.
(11) Overberger, C. G.; Weise, J . K. J . Am. Chem. Soc. 1968, 90,
3533.
(5) Hillen, W.; Egert, E.; Lindner, H. J .; Gassen, H. G. Biochemistry
1978, 17, 5314.
(6) Goodwin, J . T.; Glick, G. D. Tetrahedron Lett. 1993, 34, 5549.
S0022-3263(96)00966-8 CCC: $12.00 © 1996 American Chemical Society

Communications
J . Org. Chem., Vol. 61, No. 17, 1996 5709
sodium salt of R-hydroxymethylene lactone,8 no reaction
of 2b with methyl formate was observed under identical
conditions. A stronger base, magnesium diisopropyla-
mide, was reported to afford low yields for the condensa-
tion reactions of 2b and 2c.12 We observed that lithium
diisopropylamide (LDA) improved the condensations
considerably. Thus, addition of LDA to a mixture of 2b
and methyl formate in dry THF produced the lithium salt
3b, which precipitated from the reaction mixture in 90%
yield (crude). Without further purification, compound 3b
was added to a stirred solution of urea in 3 N HCl at 0
°C. After the solution was stirred for 1 h at room
temperature and allowed to stand overnight, a solid
formed that was collected by filtration to give 4b in 67%
yield.
Treatment of 4b with potassium hydroxide in methanol
at reflux for 4 h under nitrogen yielded a solid precipitate.
To this suspension were added water (to dissolve the
solid) and 1-(tert-butylthio)hydrazine 1,2-dicarboxymor-
pholide.13 The mixture was stirred for 1 h at room
temperature and allowed to stand overnight to give again
a solid precipitate, which was collected by filtration to
yield 5b in 79% yield.
Sch em e 3
sulfonyl chloride in pyridine gave the pyridinium salt 8
in 69% yield after recrystallization from methanol.
Heating an aqueous mixture of 8 and sodium thioben-
zoate at 80 °C for 15 min resulted in precipitation of the
benzoylthio derivative 9 in greater than 90% yield.
Debenzoylation of 9 with lithium hydroxide in methanol
in the presence of 1-(tert-butylthio)hydrazine 1,2-dicar-
boxymorpholide resulted in efficient conversion to the
tert-butyl disulfide 5a (93%).
The modified bases 5a -d were converted to the cor-
responding nucleosides using the Hilbert-J ohnson reac-
tion (Scheme 2).7 These bases 5a -d were heated in
hexamethyldisilazane at reflux, subsequently treated in
situ with 1-O-acetyl-2,3,5-tri-O-benzoyl-â-D-ribofuranose
and tin chloride in acetonitrile. Although these reactions
were extremely air- and moisture-sensitive, products
6a -d were obtained in yields ranging from 75-93%.
Debenzoylation by treatment with 25% sodium methox-
ide in methanol at room temperature, followed by acidi-
fication with Dowex-resin (50WX8-200), afforded the
unprotected nucleosides 1a -d in yields exceeding 90%.
The condensation of 2c with methyl formate under the
same conditions as for 2b failed to give 3c as a precipi-
tate, possibly because of increased solubility of the
product in THF due to the extra methylene group. Thus,
this condensation reaction mixture was concentrated and
directly treated with urea in 3 N HCl. The product 4c
precipitated from the mixture, was collected by filtration,
and was found to be analytically pure by elemental
1
analysis and H NMR. The overall yield for the two-step
conversion of 2c to 4c was 48%. Compound 4c was
converted to 5c in 75% yield using the same procedure
as that for 4b to 5b.
Analogous transformations of thiepanone 2d using
procedures similar to those described for conversion of
2b and 2c to 5b and 5c, respectively, generated product
5d in very low overall yields. This is not surprising as
base-induced polymerization of 2d has been well docu-
mented11 and can compete with the desired condensation
between 2d and methyl formate. We tested a different
base, lithium bis(trimethylsilyl)amide at -78 °C, which
proved to be very effective for the desired condensation,
presumably by minimizing polymerization and other side
reactions. The condensation reaction mixture was con-
centrated to a solid and treated with a solution of urea
in absolute ethanol. After the mixture was refluxed
overnight, water and 1-(tert-butylthio)hydrazine 1,2-
dicarboxymorpholide were sequentially added at room
temperature. Two hours of stirring gave the product 5d
as a solid. It is likely that condensation intermediate
3d reacts with urea, leading directly to formation of the
pyrimidine ring. While possible intermediates 3d and
4d were not isolated, the overall procedure was simpli-
fied, and the overall yield from 2d to 5d was 56%, higher
than those from 2b and 2c to 5b and 5c, respectively.
We are currently investigating whether it is possible to
further improve the conversions of 5b and 5c according
to this modified procedure and whether it can be applied
to higher homologs of the thiolactones.
In summary, we have described a short, practical route
to 5-(thioalkyl)uridines from readily available starting
materials. These modified uridines 1a -d have been
converted to the phosphoramidite products 10a -d using
standard procedures14,15 and have been incorporated site-
specifically into RNA molecules with coupling yields
identical to those of normal nucleoside phosphoramidites.
Furthermore, it has been possible to convert nucleosides
6a -d directly to the corresponding cytidine derivative
via the 4-triazole intermediate.16 These results will be
reported in due course.
Ack n ow led gm en t. We thank Professors William D.
Wulff and Viresh Rawal for helpful comments on the
manuscript. A.M. thanks the Howard Hughes Medical
Institute for an Undergraduate Summer Research Fel-
lowship.
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails and characterizations of all new compounds (9 pages).
Attempts to convert the 4-membered thiolactone 2a to
5a using these procedures proved unsuccessful. How-
ever, 5a was easily prepared from commercially available
5-(hydroxymethyl)uracil according to Scheme 3. Thus,
treatment of 5-(hydroxymethyl)uracil with methane-
J O960966M
(14) Hakimelahi, G. H.; Proba, Z. A.; Ogilvie, K. K. Can. J . Chem.
1982, 60, 1106.
(15) Usman, N.; Ogilvie, K. K.; J iang, M.-Y.; Cedergren, R. J . J . Am.
Chem. Soc. 1987, 109, 7845.
(16) Divakar, K. J .; Mottoh, A.; Reese, C. B.; Sanghvi, Y. S. J . Chem.
Soc., Perkin Trans. 1 1990, 969.
(12) Korte, F.; Buchel, K. H. Chem. Ber. 1960, 93, 1021.
(13) Wunsch, E.; Moroder, L.; Romani, S. Physiol. Chem. 1982, 363,
1463.