A new concept for the preparation of beta-L- and beta-D-2',3'-dideoxynucleoside analogues.

Article abstract of DOI:10.1021/ol026460+

[reaction: see text] A new method for the synthesis of 2',3'-dideoxynucleoside analogues has been developed. An electrochemical activation of 2-substituted furans is followed by the coupling with a pyrimidine or purine base. This gives planar furyl nucleosides as key intermediates, which are hydrogenated cis-selectively to give the corresponding beta-2',3'-dideoxynucleosides as racemic mixtures. An enzymatic kinetic resolution gives rise to beta-D- and beta-L-configured derivatives in high optical purity. This is exemplified by the synthesis of beta-D- and beta-L-3'-deoxythymidine.

Full text of DOI:10.1021/ol026460+

ORGANIC
LETTERS
2002
Vol. 4, No. 19
3251-3254
A New Concept for the Preparation of
â-L- and â-D-2′,3′-Dideoxynucleoside
Analogues
Martin Albert,* Dominic De Souza, Petra Feiertag, and Helmut Ho1nig*
Institut fu¨r Organische Chemie, Technische UniVersita¨t Graz, A-8010 Graz, Austria
albert@orgc.tu-graz.ac.at
Received July 3, 2002
ABSTRACT
A new method for the synthesis of 2′,3′-dideoxynucleoside analogues has been developed. An electrochemical activation of 2-substituted
furans is followed by the coupling with a pyrimidine or purine base. This gives planar furyl nucleosides as key intermediates, which are
hydrogenated cis-selectively to give the corresponding â-2′,3′-dideoxynucleosides as racemic mixtures. An enzymatic kinetic resolution gives
rise to â-D- and â-L-configured derivatives in high optical purity. This is exemplified by the synthesis of â-D- and â-L-3’-deoxythymidine.
Since the first synthesis of a pyrimidine 2′,3′-dideoxynucleo-
side, 3′-deoxythymidine, described by Michelson and Todd
in 1955,1 a huge number of 2′,3′-dideoxynucleoside ana-
logues (ddNs) has been prepared. The reason for interest in
this family of compounds is based on the finding that ddNs
are potentially effective therapeutic agents for the treatment
of acquired immune deficiency syndrome (AIDS) and other
virus-caused diseases.2 Four of the six nucleoside analogues
that are approved anti-HIV drugs, namely, Zidovudine
(AZT),3 Didanosine (ddI),4 Zalcitabine (ddC),5 and Stavudine
(d4T),6a re derivatives of the naturally occurring nucleosides.
These compounds all have one thing in common, namely, a
â-D-configuration in their sugar moieties. All of these drugs
and other nucleoside analogues are believed to have a similar
mechanism of HIV inhibition, in which the analogues are
progressively phosphorylated by cytoplasmic enzymes to
nucleoside 5′-triphosphates. These then compete with the
natural nucleoside triphosphate substrate for binding to
cellular DNA polymerase and the viral reverse transcriptase.
More recently, a number of nucleoside analogues, pos-
sessing the unnatural â-L-configuration, have emerged as
potential antiviral agents. Research was encouraged by the
fact that L-nucleosides are generally endowed with lower host
toxicity while maintaining good antiviral/antibacterial activ-
ity.7 Lamivudine (3TC) is the first compound out of that
family that has been approved for use in combination therapy
against HIV and HBV.8 Other promising candidates with
notedly good antiviral profiles are â-L-ddC,9 â-L-Fd4C,10 or
â-L-FddC11 (Figure 1).
(1) Michelson, A. M.; Todd, A. R. J. Chem. Soc. 1955, 816-823.
(2) Mansour, T. S.; Storer, R. Curr. Pharm. Des. 1997, 3, 227-264.
(3) (a) De Clercq, E. AdV. Drug. Res. 1988, 17, 1-59. (b) Levantis, P.;
Stein, C.; Oxford, J. AntiViral Chem. Chemother. 1993, 4, 131-139 and
references therein.
(4) (a) Yarchoan, R.; Mitsuya, H.; Thomas, R. V.; Pluda, J. M.; Hartman,
N. R.; Perno, C.-F.; Marczyk, K. S.; Allain, J.-P.; Johns, D. G.; Broder, S.
Science (Washington D.C.) 1989, 245, 412-415. (b) Perry, C. M.; Balfour,
J. A. Drugs 1996, 52, 928-962.
(5) (a) Mitsuya, H.; Broder, S. Proc. Natl. Acad. Sci. U.S.A. 1986, 83,
1911-1915. (b) Skowron, G. AdV. Exp. Med. Biol. 1996, 394, 257-269.
(6) (a) Mansuri, M. M.; Starett, J. E.; Ghazzouli, I.; Hitchcock, M. J.
M.; Sterzycki, R. Z.; Brankovan, V.; Lin, T.-S.; August, E. M.; Prusoff,
W. H.; Sommadossi, J.-P.; Martin, J. C. J. Med. Chem. 1989, 32, 461-
466. (b) Riddler, S. A.; Anderson, R. E.; Mellors, J. W. AntiViral Res. 1995,
27, 189-203. (c) Lin, T. S.; Schinazi, R. F.; Prusoff, W. H. Biochem.
Pharmacol. 1987, 36, 2713-2178.
(7) (a) Nair, V.; Jahnke, T. S. Antimicrob. Agents Chemother. 1995, 39,
1017-1029. (b) Wang, P.; Hong, J. H.; Cooperwood, J. S.; Chu, C. K.
AntiViral Res. 1998, 40, 19-44.
(8) (a) Chang, C.-N.; Doong, S.-L.; Zhou, J. H.; Beach, J. W.; Jeong, L.
S.; Chu, C. K.; Tasi, C.-H.; Cheng, Y.-C. J. Biol. Chem. 1992, 267, 13939-
13942. (b) Doong, S.-L.; Tasi, C.-H.; Schinazi, R. F.; Liotta, D. C.; Cheng,
Y.-C. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 8495-8499.
(9) (a) Mansuri, M. M.; Farina, V.; Starrett, J. E., Jr.; Benigni, D. A.;
Brankovan, V.; Martin, J. C. Bioorg. Chem. Med. Lett. 1991, 1, 65-68.
(b) Gosselin, G.; Mathe´, C.; Bergogne, M.-C.; Aubertin, A.-M.; Somma-
dossi, J.-P.; Schinazi, R.; Imbach, J.-L. Nucleosides Nucleotides 1995, 14,
611-617.
10.1021/ol026460+ CCC: $22.00 © 2002 American Chemical Society
Published on Web 08/29/2002

Bearing that in mind, we anticipated that furoic acid,
furfural, or furyl methanol derivatives, which are available
in large scale, represent ideal starting materials for such a
synthesis, since they already consist of the five-carbon
skeleton of the final product. The coupling of a suitable furan
derivative with a heterocycle should yield prochiral planar
furyl nucleosides as key intermediates. The high reactivity
of the furan moiety should allow a chemoselective reduction
in the presence of an unsaturated N-heterocycle. By choosing
a proper catalyst the hydrogenation should proceed in a
highly cis-selective manner, thus fixing the desired â-con-
figuration and yielding 2′,3′-dideoxynucleosides as a racemic
mixture (Scheme 1).
Scheme 1. Retrosynthetic Analysis
Figure 1. Structures of relevant nucleoside analogues.
A common structural feature of all active nucleoside
analogues is determined through the cis-arrangement of the
hydroxymethyl side chain to the nucleobase (â-configura-
tion). Syntheses for D-configured nucleoside analogues,
starting from naturally occurring nucleosides, are well
elaborated. The hydroxy groups at C-2′ and C-3′ can be
removed by a number of reactions without affecting the
glycosyl bond. Rather expensive starting materials and a
limited range of accessible products narrow this approach.12
Since there was no general method for the synthesis of
furyl nucleosides starting from non-nucleoside-derived pre-
cursors,15 we were looking for a convenient route to obtain
this class of compounds. Initial attempts to couple 2-bromo-
5-substituted furans or 5-substituted furyl-2-boronic acids or
esters, respectively, with pyrimidine or purine derivatives,
employing palladium16 or copper17 catalysis, either failed or
gave unsatisfying results. Because of the obvious restriction
of the catalyzed (and uncatalyzed) nucleophilic aromatic
substitution at position 2 of furans, we decided to use well
elaborated acetal chemistry for the coupling reaction. In 1952
Clauson-Kaas described the electrochemical oxidative
dimethoxylation of some 2-substituted furans.18 The reaction
tolerates both furoic acid (1a) and furyl methanol derivatives
(1b) and can easily be performed in a 100 g scale to give
2-substituted-2,5-dimethoxy-2,5-dihydrofurans (2a and b) in
very good yields (Scheme 2).
Alternatively, the fusion of a suitable sugar part or a carba-
or heterocyclic equivalent with an activated base can give
rise to the corresponding D- and L-configurated nucleoside
analogues and other unnatural nucleoside derivatives. Thereby,
problems concerning the R/â selectivity of the coupling
reaction are usually encountered. This can be overcome by
the introduction of an anchimerically assisting group at
position 2 or 3 of the sugar moiety that has to be removed
at the end of the synthesis, thus rendering the sequence
laborious.13
A de novo approach to D- and L-nucleosides starting from
furan was described by Trost and Shi.14 Palladium-catalyzed
desymmetrization of a cis-2,5-diacyloxy-2,5-dihydrofuran
(prepared by reaction of lead tetraacyloxylates with furan)
enables the subsequent introduction of the nucleobase and
the side chain, thereby allowing access to both enantiomeric
series.
Scheme 2. Electrochemical Activation of Furans According to
Clauson-Kaas
However, all the existing methods for the synthesis of 2′,3′-
dideoxynucleosides suffer from at least one of these draw-
backs: expensive starting materials, toxic reagents, lack of
control during the glycosylation, and lengthy synthesis.
To our delight, it turned out that compound 2a can easily
be coupled with TMS-thymine in the presence of TMSOTf
to give 3a as a mixture of diastereoisomers in 90% yield.19
(10) Lin, T.-S.; Luo, M.-Z.; Liu, M.-C.; Zhu, Y.-L.; Gullen, E.;
Dutschman, G. E.; Cheng, Y.-C. J. Med. Chem. 1996, 39, 1757-1759.
(11) (a) Lin, T.-S.; Luo, M.-Z.; Liu, M.-C.; Pai, S. B.; Dutschman, G.
E.; Cheng, Y.-C. J. Med. Chem. 1994, 37, 798-803. (b) Van Draanen, N.
A.; Tisdale, N.; Parry, N. R.; Jansen, R.; Dornsife, R. E.; Tuttle, J. V.;
Averett, D. R.; Koszalka, G. W. Antimicrob. Agents Chemother. 1994, 38,
868-871.
(12) (a) Huryn, D. M.; Okabe, M. Chem. ReV. 1992, 92, 1745-1768.
(b) Robins, M.; Wilson, J. S.; Madej, D.; Low, N. H.; Hansske, F.; Wnuk,
S. F. J. Org. Chem. 1995, 60, 7902-7908 and references therein.
(13) (a) Chu, C. K.; Babu, J. R.; Beach, J. W.; Ahn, S. K.; Huang, H.-
Q.; Jeong, L. S.; Lee, S. J. J. Org. Chem. 1990, 55, 1418-1420. (b) Bolon,
P. J.; Wang, P.; Chu, C. K. Bioorg. Med. Chem. 1996, 6, 1657-1662. (c)
Wilson, L. J.; Liotta, D. Tetrahedron Lett. 1990, 31, 1815-1818. (d) Chen,
S.-H.; Li, X.; Li, J.; Niu, C.; Carmichael, E.; Doyle, T. W. J. Org. Chem.
1997, 62, 3449-3452.
(14) Trost, B. M.; Shi, Z. J. Am. Chem. Soc. 1996, 118, 3037-3038.
3252
Org. Lett., Vol. 4, No. 19, 2002

By treatment of 3a with TMSOTf (1 equiv) in acetonitrile,
elimination together with rearomatization occurred conse-
quently. Thereby, one single product was obtained from four
stereoisomers (94% yield). Since these furyl nucleosides can
exist as a pair of atropisomers (see Figure 2), we started
Scheme 4. Kinetic Enzymatic Resolution of Racemic 5aa
a Reagents and conditions: (a) pig liver esterase, pH 7.4, 0.1 M
NaH2PO4 buffer; (b) NaBH4, H2O.
Figure 2. Hyperchem calculations indicate a rotational barrier of
8 kJ/mol.
The extension of this principle to 2,5-dimethoxy-2,5-
dihydrofuran 2b as starting material and other pyrimidine
derivatives (in addition to thymine) was subsequently
investigated. As outlined in Scheme 5, the tandem reaction
some dynamic 1H NMR experiments with 4b (for synthesis,
see Scheme 5) with and without chiral additives.20
In contrast to results obtained by Akhrem et al.,15c who
claimed to have isolated optically active furyl nucleosides
at room temperature (from cyclic nucleoside derived precur-
sors), no direct observation of atropisomers, even at -100
°C, could be observed in our case. Additional molecular
modeling calculations suggest a rotational barrier of 8 kJ/
mol, which corroborates our results.
Scheme 5a
Compound 4a could finally be hydrogenated cis-selectively
to the corresponding â-D- and â-L-dideoxynucleoside deriva-
tives (5a) with Rh/Al2O3 in methanol in 94% yield (Scheme
3). The double bond of the pyrimidine ring remained
unaffected.
Scheme 3. Coupling, Elimination, and Hydrogenationa
a Reagents and conditions: (a) 0.2 equiv of TMSOTf, CH3CN,
-30 °C to rt; (b) TMSBr, CH3CN, rt; (c) see text.
consisting of coupling and elimination (with TMSBr as
catalyst) leads, via 3b-d, to the desired furyl nucleosides
(15) This class of compounds was exclusively obtained from nucleosides
(mainly as side product): (a) Haraguchi, K.; Tanaka, H.; Itoh, Y.;
Yamaguchi, K.; Miyasaka, T. J. Org. Chem. 1996, 61, 851-858. (b) Robins,
M. J.; Jones, R. A.; Mengel, R. J. Am. Chem. Soc. 1976, 98, 8213-8217.
(c) Timoshchuk, V. A.; Akhrem, A. A. Dokl. Akad. Nauk. SSSR 1998, 302,
105-107. (d) Joshi, B. V.; Reese, C. B. J. Chem. Soc., Perkin Trans. 1
1992, 441-443.
(16) Reviews: (a) Hartwig, J. F. Synlett 1997, 329-340. (b) Hartwig, J.
F. Angew. Chem. 1998, 110, 2154-2177. (c) Hartwig, J. F. Acc. Chem.
Res. 1998, 31, 852-860. (d) Wolfe, J. P.; Wagaw, S.; Marcoux, J.-F.;
Buchwald, S. L. Acc. Chem. Res. 1998, 31, 805-818.
(17) (a) Lam, P. Y. S.; Clark, C. G.; Saubern, S.; Adams, J.; Winters,
M. P.; Chan, D. M. T.; Combs, A. Tetrahedron Lett. 1998, 39, 2941-
2944. (b) Combs, A. P.; Saubern, S.; Rafalski, M.; Lam, P. Y. S.
Tetrahedron Lett. 1999, 40, 1623-1626. (c) Kiyomori, A.; Marcoux, J.-F.;
Buchwald, S. L. Tetrahedron Lett. 1999, 40, 2657-2660.
(18) Clauson-Kaas, N.; Limborg, F.; Dietrich, P. Acta Chem. Scand. 1952,
6, 545-550.
a Reagents and conditions: (a) TMSOTf, CH3CN, -30 °C to rt;
(b) TMSOTf, CH3CN, rt; (c) H2 (1 bar), Rh/Al2O3 (5%), MeOH.
An enzymatic screening with a series of commercially
available enzymes using racemic 5a as substrates revealed
that pig liver esterase selectively hydrolyzed â-L-5a, thus
allowing the separation of the two enantiomers. The reduction
with NaBH4 completed the synthesis of â-D-ddT and â-L-
ddT in good (unoptimized) yields and high optical purity.
Org. Lett., Vol. 4, No. 19, 2002
3253

catalyzed hydrolysis of acyl-protected or an enzyme-
catalyzed acyl transfer to unprotected analogues of 5b-d
should enable access to the corresponding enantiomerically
pure â-L- and â-D-derivatives.
Scheme 6a
The coupling of 2b with 6-chloro-purine gives rise to the
coupling product 3e in good yield. The elimination to the
furyl nucleoside 4e proved to be difficult. BCl3 in acetonitrile
gave the desired elimination product in 40% yield, which
was the best we could attain so far.
An optimization of this reaction and attempts for the
hydrogenation of 4e are currently under investigation.
This new synthetic principle, combining “furan chemistry”
with “nucleoside chemistry” (via hydrogenation), is highly
promising because it opens new routes to known as well as
new nucleoside analogues and avoids some difficulties
encountered in the classical approaches. Antiviral activity
and cytotoxic evaluations of selected compounds are in
progress.
a Reagents and conditions: (a) 0.2 equiv of TMSOTf, CH3CN,
-30 °C to rt; (b) 1 equiv of BCl3, CH3CN, rt.
4b-d in good to excellent yields. The hydrogenation of 4b
and 4d was successful when Pd/C in MeOH was used as
catalyst. However, this system led to the reduction of both
the furan and the pyrimidine in the case of the uracil-
derivative 4c. This problem has been overcome by using less
reactive Pd/C in EtOAc, yielding racemic 5c in good yield
(80%). Some hydrogenolytic cleavage of the C-N bond was
observed as a side reaction in each case. An enzyme-
Acknowledgment. We thank Prof. J. Weber for his help
in the conformational studies of compound 4b by dynamic
1H NMR. Prof. A. Fu¨rstner, Max-Planck-Institut fu¨r Kohlen-
forschung, Germany, is thanked for helpful discussions.
Supporting Information Available: Experimental pro-
cedures and spectroscopic and analytical data for all new
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
(19) Compare: Feringa, B. L.; Dannenberg, W. Recl. TraV. Chim. Pays-
Bas 1984, 103, 192-196.
(20) Gasparrini, F; Grilli, S.; Leardini, R.; Lunazzi, L.; Mazzanti, A.;
Nanni, D.; Pierini, M.; Pinamonti, M. J. Org. Chem. 2002, 67, 3089-3095.
OL026460+
3254
Org. Lett., Vol. 4, No. 19, 2002