Efficient synthesis of 2-deoxy L-ribose from L-arabinose: mechanistic information on the 1,2-acyloxy shift in alkyl radicals.

Article abstract of DOI:10.1021/ol990838v

[formula: see text] Conversion of the inexpensive L-arabinose 1 into the ethylthio ortho ester 7 followed by generation of the dialkoxyalkyl radical III produces the desired 2-deoxy-L-ribose triester 4 in excellent overall yield. It has been shown that th

Full text of DOI:10.1021/ol990838v

ORGANIC
LETTERS
1999
Vol. 1, No. 10
1517-1519
Efficient Synthesis of 2-Deoxy L-Ribose
from L-Arabinose: Mechanistic
Information on the 1,2-Acyloxy Shift in
Alkyl Radicals
Michael E. Jung* and Yue Xu
Department of Chemistry and Biochemistry, UniVersity of California,
Los Angeles, California 90095-1569
jung@chem.ucla.edu
Received July 20, 1999
ABSTRACT
Conversion of the inexpensive L-arabinose 1 into the ethylthio ortho ester 7 followed by generation of the dialkoxyalkyl radical III produces
the desired 2-deoxy-L-ribose triester 4 in excellent overall yield. It has been shown that the similar dialkoxyalkyl radical IV is not an intermediate
in the 1,2-acyloxy shift of anomeric radical I.
L-Nucleosides and their analogues have become useful agents
for the treatment of viral diseases due in part to their good
antiviral activity and generally low toxicity.1 Either normal
(L-RNA) or 2′-deoxy (L-DNA) L-nucleosides may also be
of value in antisense oligonucleotide therapy as materials to
bind pieces of D-m-RNA.2 For these reasons, new methods
for the preparation of L-nucleoside analogues and the
carbohydrates from which they are derived are an important
synthetic goal. We recently reported an efficient synthesis
of L-ribose and 2-deoxy L-ribose from D-ribose and a fast
preparation of the latter from L-arabinose3 as well as a
conceptually different approach to prepare 2-deoxy L-ribose
from penta-1,4-dien-3-ol.4 We have reexamined the first
approach and report herein a very efficient synthesis of
2-deoxy L-ribose which uses a radical rearrangement of an
unusual cyclic monothio ortho ester.5,6 We also offer
experimental evidence that the 1,2-acyloxy shift in alkyl
radicals, first reported by Surzur7 and Tanner,8 and used
extensively in carbohydrates by Giese,9 does not proceed via
a cyclic dialkoxyalkyl radical.
(4) Jung, M. E.; Nichols, C. J. Tetrahedron Lett. 1998, 39, 4615.
(5) For syntheses of L-ribose, see: (a) Matteson, D. S.; L., P. M. J. Org.
Chem. 1987, 52, 5116. (b) Wulff, G.; Hansen, A. Carbohydr. Res. 1987,
164, 123. (c) Austin, W. C.; Humoller, F. L. J. Am. Chem. Soc. 1932, 54,
4749. (d) Acton, E. M.; Ryan, K., J.; Goodman, L. J. Am. Chem. Soc. 1964,
86, 5352. (e) Abe, Y.; Takizawa, T.; Kunieda, T. Chem. Pharm. Bull. 1980,
28, 1324. (f)Visser, G. M.; van Westrenen, J.; van Boeckel, C. A. A.; van
Boom, J. H. Recl. TraV. Chim. Pays-Bas 1986, 105, 528. (g) Batch, A.;
Czernecki, S. J. Carbohydr. Chem. 1994, 13, 935. (h) Chelain, E.; Floch,
O.; Czernecki, S. J. Carbohydr. Chem. 1995, 14, 1251. (i) Fischer, E.; Piloty,
O. Chem. Ber. 1891, 24, 521. (j) Mori, K.; Kikuchi, H. Liebigs Ann. Chem.
1989, 1267.
(6) Walker, T. E.; Hogenkamp, H. P. C. Carbohydr. Res. 1974, 32, 413.
(7) Surzur, J. M.; Teissier, P. C. R. Hebd. Seances Acad. Sci., Ser. C
1967, 264, 1981. Surzur, J.-M.; Teissier, P. C. R. Bull. Soc. Chim. Fr. 1970,
3060.
(8) Tanner, D. D.; Law, F. C. P. J. Am. Chem. Soc. 1969, 91, 7535.
(9) Giese, B.; Gilges, S.; Gro¨ninger, K. S.; Lamberth, C.; Witzel, T.
Liebigs Ann. Chem. 1988, 615.
(1) For leading references, see: (a) Okabe, M.; Sun, R.-C.; Tam, S. Y.-
K.; Todaro, L. J.; Coffen, D. L. J. Org. Chem. 1988, 53, 4780. (b) Schinazi,
R. F.; Gosselin, G.; Faraj, A.; Korba, B. E.; Liotta, D. C.; Chu, C. K.;
Mathe´, C.; Imbach, J.-L.; Sommadossi, J.-P. Antimicrob. Agents Chemother.
1994, 38, 2172. (c) Chu, C. K.; Ma, T.; Shanmuganathan, K.; Wang, C.;
Xiang, Y.; Pai, S. B.; Yao, G.-Q.; Sommadossi, J.-P.; Cheng, Y.-C.
Antimicrob. Agents Chemother. 1995, 39, 979. (d) Lin, T.-S.; Luo, M.-Z.;
Liu, M.-C.; Pai, B.; Dutschman, G. E.; Cheng, Y.-C. J. Med. Chem. 1994,
37, 798 and the references listed in refs 3 and 4.
(2) For leading references, see the following. (a) L-DNA: Damha, M.
J.; Giannaris, P. A.; Marfey, P. Biochemistry 1994, 33, 7877. Hashimoto,
Y.; Iwanami, N.; Fujimori, S.; Shudo, K. J. Am. Chem. Soc. 1993, 115,
9883. Fujimori, S.; Shudo, K.; Hashimoto, Y. J. Am. Chem. Soc. 1990,
112, 7436. (b) L-RNA: Ashley, G. W. J. Am. Chem. Soc. 1992, 114, 9731.
(3) Jung, M. E.; Xu, Y. Tetrahedron Lett. 1997, 38, 4199.
10.1021/ol990838v CCC: $18.00 © 1999 American Chemical Society
Published on Web 10/06/1999

To improve our earlier synthesis,4 we first prepared methyl
L-arabinofuranoside. After esterification with toluoyl chloride
to facilitate the later preparation of the desired R-chloride,
we converted it to the anomeric bromide 2a and thioether
2b in 67% and 95% yields, respectively. However, radical-
irrespective of the mechanism of the rearrangement, genera-
tion of III should produce the desired product 4.
promoted rearrangement-reduction of either 2a or 2b with
tributyltin hydride and AIBN gave a mixture of the undesired
“direct reduction” 1-deoxy product 3 and the desired 2-deoxy
sugar 4. This ratio varied widely depending on conditions,
but was never better than 2.8:1 in favor of the rearranged
product 4, which was isolated in about 50% yield at best.10
Presumably the 1,2-acyloxy shift of the initially formed
radical I to give II is slower in the arabino than in the ribo
series (perhaps due to steric acceleration of the ribo case).
Thus the radical I is reduced competitively with its rear-
rangement to II (and reduction to 4). Hence we used the
Stork catalytic hydride conditions,11 e.g., catalytic Bu3SnH
with stoichiometric sodium cyanoborohydride, with, how-
ever, an unusual result, namely the complete formation of
the 1,2-acetal 5 (the same product was formed without the
tin reagent present indicating an ionic mechanism). This
Solvolysis of the bromide 2a in nitromethane in the
presence of thiophenol and a hindered base gave the
phenylthio ortho ester 6, which proved unstable for use in
our system as it readily rearranged to the R-thiophenyl
arabinoside 2b.12 However, the ethylthio analogue 7, prepared
in 86% yield from 2a and ethanethiol, was more stable. It
was treated with tributyltin hydride in hot toluene to give
the desired 2-deoxy L-ribose derivative 4 in 85% yield, along
with small amounts (2-5%) each of the 1-deoxy compound
3, the acetal 5, and the product of solvolysis with internal
trapping 8. Thus generation of the dialkoxyalkyl radical III
from the ethylthio compound 7 produced the desired radical
II, in a much greater amount as compared to I as predicted,
and from it the 2-deoxy L-ribose 4. Hydrolysis of 4 furnished
selective formation of a 1,2-acetal in carbohydrates, while a
very useful synthetic transformation in itself, provided a
possible solution to the problem of radical reduction due to
a sluggish rearrangement. We hypothesized that if one
prepared the dialkoxyalkyl radical III, it would open to the
more stable ester radicals. Molecular mechanics calculations
(PM3) indicated that II was ∼3 kcal/mol more stable than I
(the extensive results of Giese9 also supported this energy
difference). If one generated III from a readily available
precursor, it should open selectively to give II which would
be reduced to 4. It is also possible that the mechanism for
the rearrangement of I to II proceeds via III. However,
(10) Recently there have been reports of the use of Lewis acids to
accelerate these rearrangements, for example: Lacoˆte, E.; Renaud, P. Angew.
Chem., Int. Ed. Engl. 1998, 37, 2259. Also there has been a theoretical
study on the effectiveness of protic acids for this acceleration: Zipse, H. J.
Am. Chem. Soc. 1997, 119, 1087.
(11) Stork, G.; Sher, P. J. Am. Chem. Soc. 1986, 108, 303.
1518
Org. Lett., Vol. 1, No. 10, 1999

2-deoxy L-ribose 9 in nearly quantitative yield, thereby
ending a short six-step synthesis of 9 from 1 in 49% overall
yield. Treatment of 4 with HCl in acetic acid gave the desired
crystalline R-chloride 10 in 70% yield,13 from which the
various L-nucleosides 11 were prepared in two simple steps
and then their 5′-DMTr-protected phosphoramidites for use
in solid-phase synthesis of L-DNA.
Finally we decided to test one possible mechanism for the
rearrangement of I to give II, the intermediacy of the
dialkoxyalkyl radical III. Ingold and others have proposed
seven possible transition states for this rearrangement but
exclude the dialkoxyalkyl radical, based mainly on ESR
evidence.14 Yet this system allowed us to provide strong
experimental evidence for whether the dialkoxyalkyl radical
III might be an intermediate (or transition state) in this
rearrangement. Hence we prepared the two substrates, the
thioether 12a and the bromide 12b bearing cyclopropyl-
carboxylates from L-arabinose 1, in good yield. Treatment
of 12a under normal radical rearrangement conditions gave
the expected 2:1 mixture of the 2-deoxy and 1-deoxy
products, 13 and 14, also in good yield. Solvolysis of the
bromide 12b with trapping with ethanethiol gave the second
substrate 15. Treatment of 15 under conditions identical to
those for 12a gave a very fast formation of a 1:1 mixture of
the two stereoisomeric alkenes 16ab resulting from rapid
opening of the cyclopropylcarbinyl radical IV. We can now
say unequivocally that the rearrangement of 12a to give the
deoxy products 13 and 14 does not proceed via the
dialkoxyalkyl radical IV since when it is generated separately
by a different route, the same products are not formed. Thus
the 1,2-acyloxy shift does not proceed via the intermediacy
of the dialkoxyalkyl radical IV (III in the earlier discussion
of 3 and 4).15
Application of this chemistry to the synthesis of modified
L-nucleosides and oligomers of L-DNA is currently underway
in our laboratories.
(12) (a) Balan, N. F.; Bakinovskii, L. V.; Kochetkov, N. K. Bioorg. Khim.
1980, 6, 1657. (b) Backinowsky, L. V.; Tsvetkov, Y. E.; Balan, N. F.;
Byramova, N. E.; Kochetkov, N. K. Carbohydr. Res. 1980, 85, 209.
(13) Selective crystallization of the R-anomer of the bis(toluoyl) chloride
was already known for the D-enantiomer. Hoffer, M. Chem. Ber. 1960, 93,
2777.
Acknowledgment. We thank the National Institutes of
Health (GM47228) and the Universitywide AIDS Research
Program (R97-LA-139) for generous financial support.
(14) (a) Barclay, L. R. C.; Griller, D.; Ingold, K. U. J. Am. Chem. Soc.
1982, 104, 4399. (b) Barclay, L. R. C.; Lusztyk, J.; Ingold, K. U. J. Am.
Chem. Soc. 1984, 106, 1793. (c) Kocovsky, P.; Stary, I.; Turecek, F.
Tetrahedron Lett. 1986, 27, 1513. (d) Korth, H.-G.; Sustmann, R.;
Gro¨ninger, K. S.; Leisung, M.; Giese, B. J. Org. Chem. 1988, 53, 4364. (e)
Beckwith, A. L. J.; Duggan, P. J. J. Chem. Soc., Perkin Trans. 2 1992,
1777. (f) Beckwith, A. L. J.; Duggan, P. J. J. Chem. Soc., Perkin Trans. 2
1993, 1673. (g) Crich, D.; Filzen, F. J. Org. Chem. 1995, 60, 4834. (h)
Beckwith, A. L. J.; Duggan, P. J. J. Am. Chem. Soc. 1996, 118, 12838. (i)
Zipse, H. J. Chem. Soc., Perkin Trans. 2 1997, 1797. (j) For an excellent
review, see: Beckwith, A. L. J.; Crich, D.; Duggan, P. J.; Yao, Q. Chem.
ReV. 1997, 97, 3273.
Supporting Information Available: Proton and carbon
NMR spectra for all new compounds and procedures for the
preparation of the thio ortho ester 7 and its conversion into
4. This material is available free of charge via the Internet
at http://pubs.acs.org.
OL990838V
(15) For the preparation and reaction of a different dialkoxyalkyl radical,
see: Jung, M. E.; Kiankarimi, M. J. Org. Chem. 1995, 60, 7013.
Org. Lett., Vol. 1, No. 10, 1999
1519