A Ferrier-type allylic rearrangement of 3′-deoxy-3′,4′- didehydronucleosides mediated by DMF dimethyl acetal: Direct access to 4′-alkoxy-2′,3′-didehydro-2′,3′-dideoxynucleosides

Article abstract of DOI:10.1021/ol201519a

A straightforward synthesis of epimeric 4′-alkoxy-substituted 2′,3′-didehydro-2′,3′-dideoxynucleosides via a DMF dimethyl acetal mediated allylic rearrangement of 3′-deoxy-3′, 4′-didehydronucleosides is described.

Full text of DOI:10.1021/ol201519a

ORGANIC
LETTERS
2011
Vol. 13, No. 16
4200–4203
A Ferrier-Type Allylic Rearrangement of
30-Deoxy-30,40-didehydronucleosides
Mediated by DMF Dimethyl Acetal: Direct
Access to 40-Alkoxy-20,30-didehydro-20,30-
dideoxynucleosides
ꢀ
Magdalena Petrova, Milos Budesınsky, Eva Zbornıkova, Pavel Fiedler, and
ꢁ
ꢁꢁ
ꢀ
ꢀ
´
´
Ivan Rosenberg*
Institute of Organic Chemistry and Biochemistry, Academy of Sciences of the Czech
ꢀ
Republic, v.v.i., Flemingovo nam. 2, 166 10 Praha 6, Czech Republic
ivan@uochb.cas.cz
Received June 7, 2011
ABSTRACT
A straightforward synthesis of epimeric 40-alkoxy-substituted 20,30-didehydro-20,30-dideoxynucleosides via a DMF dimethyl acetal mediated allylic
rearrangement of 30-deoxy-30,40-didehydronucleosides is described.
Nucleoside analogues possessing an unsaturated sugar
moiety are an important class of biologically active com-
pounds; d4T (Stavudine) has been approved for AIDS
treatment, and β-L-d4FC (Elvucitabine) and β-D-20-Fd4C
are in clinical trials as anti-HIV and/or anti-HBV agents.1
The recently reported 40-ethynyl d4T derivative was more
potent against HIV than stavudine and less toxic to CEM
cell growth, whereas the 40-cyano d4T analogue was 5
times less active than stavudine.2 Various approaches at
HIV-1.4 The design, synthesis, and properties of 40-sub-
stituted ribonucleosides as inhibitors of HCV replication
have been previously reported; in particular, 40-azidocyti-
dine has been shown to act as a potent and highly selective
inhibitor.5
Epimeric 40-alkoxy substituted derivatives have been
previously prepared from O2,40-anhydronucleosides,
40,50-epoxy-, or 30,40-epoxynucleosides, in the presence of
alcohol.6 40-C-Branched 20,30-didehydro-20,30-dideoxyri-
bonucleosides have been prepared by the treatment of
40,50-epoxynucleosides with various organosilicon or
synthesizing
20,30-didehydro-20,30-dideoxynucleosides
were reviewed,3 as was the potential of 40-C-substituted
nucleosides (in ribo, 20-deoxyribo, 20,30-dideoxy, and 20,30-
didehydro-20,30-dideoxy series) for use in the treatment of
(5) Smith, D. B.; Martin, J. A.; Klumpp, K.; Baker, S. J.; Blomgren,
P. A.; Devos, R.; Granycome, C.; Hang, J.; Hobbs, Ch.J.; Jiang, W.-R.;
Lakton, C.; Le Pogam, S.; Leveque, V.; Ma, H.; Maile, G.; Merrett,
J. H.; Pichota, A.; Sarma, K.; Smith, M.; Swallow, S.; Symons, J.; Vesey,
D.; Najera, I.; Cammack, N. Bioorg. Med. Chem. Lett. 2007, 17, 2570–
2576.
(1) Komiotis, D.; Manta., S.; Tsoukala, E.; Tzioumaki, N. Anti-
Infect. Agents Med. Chem. 2008, 7, 219–244.
(2) (a) Haraguchi, K.; Takeda, S.; Tahala, H.; Nitanda, T.; Baba, M.;
Dutschman, G. E.; Cheby, Y.-C. Bioorg. Med. Chem. Lett. 2003, 13,
3775–3777. (b) Haraguchi, K.; Itoh, Y.; Takeda, S.; Honma, Y.;
Tanaka, H.; Nitanda, T.; Baba, M.; Dutschman, G. E.; Cheng, Y.-Ch.
Nucleosides, Nucleotides Nucleic Acids 2004, 23, 647–654. (c) Kubota,
Y.; Ishizaki, N.; Haraguchi, K.; Hamasaki, T.; Baba, M.; Tanaka, H.
Bioorg. Med. Chem. 2010, 18, 7186–7192.
(3) Len, Ch.; Postel, D. Curr. Org. Synth. 2006, 3, 261–281.
(4) Hayakawa, H.; Kohgo, S.; Kitano, K.; Ashida, N.; Kodaky, E.;
Mitsuya, H.; Ohriu, H. Antiviral Chem. Chemother. 2004, 15, 169–187.
(6) (a) Cook, S. L.; Secrist, J. A., III. J. Am. Chem. Soc. 1979, 101,
1554–1564. (b) Tarasu, H.; Sajiki, H.; Sirota, K. Heterocycles 2005, 65,
2991–2999. (c) Kubota, Y.; Ishizaki, N.; Kaneda, Y.; Haraguchi, K.;
Odznaky, Y.; Tanaka, H.; Kato, N.; Baba, M.; Balzarini, J. Antiviral
Chem. Chemother. 2009, 19, 201–212. (d) Verheyden, J. P. H.; Moffat,
J. G. J. Am. Chem. Soc. 1975, 97, 4386–4396. (e) Richards, C. N.;
Verheyden, J. P. H.; Moffat, J. G. Carbohydr. Res. 1982, 100, 315–329.
(7) (a) Kubota, Y.; Haraguchi, K.; Kunikata, M.; Hayashi, M.;
Ohkawa, M.; Tanaka, H. J. Org. Chem. 2006, 71, 1099–1103. (b)
Haraguchi, K.; Takeda, S.; Tahala, H. Org. Lett. 2003, 5, 1399–1402.
r
10.1021/ol201519a
Published on Web 07/18/2011
2011 American Chemical Society

organoaluminum reagents,7 or by the allylic substitution
of 30,40-unsaturated nucleosides having a leaving group at
the 20-position.8 Other synthetic approaches employ cou-
pling reactions of sugar synthons with a nucleobase.9
Given their broad interest and important applications
discussed above, simplified synthetic pathways leading to
various 40-alkoxy substituted 20,30-unsaturated nucleoside
derivatives would have a great impact, providing more
facile access to potentially biologically active 40-alkoxy
derivatives. Herein, we describe a new, straightforward
synthesis of epimeric 40-alkoxy-substituted 20,30-didehy-
dro-20,30-dideoxynucleosides from easily available 30-
deoxy-30,40-didehydronucleosides.10
Dialkylformamide acetals, in general, are useful re-
agents in nucleoside chemistry, e.g., for the selective in-
troduction of the amidine-type protecting group for
nucleobases,11 for the transition protection of cis-diols in
ribonucleosides,12 and for the alkylation of the amide
linkage of nucleobases.13 The role of DMF acetals as
leaving groups is well documented, for example, in the
intramolecular displacement of the 50-O-amidoacetal
group by the 3-N atom of the xanthine nucleobase provid-
ing 3-N,50-cyclo-xanthosine,14 or in the decarboxylative
elimination of the 30-O-amidoacetal grouping of nucleo-
side-50-carboxylate leading to dihydrofuryl nucleosides.15
Upon heating, allylic DMF acetals undergo intramolecu-
lar rearrangement to β,γ-unsaturated dimethylamides.16
Having previously reported a straightforward, high-
yielding synthesis of 30-deoxy-30,40-didehydronucleosides
3 and the respective 50-aldehydes 2 from nucleoside orthoe-
sters 1 (Scheme 1),10 we decided to further study their
chemical transformations.17 Surprisingly, our attempt to
selectively protect the 6-N-amino group of 30-deoxy-30,40-
didehydroadenosine 3a with a dimethylaminomethylene
moiety using N,N-dimethylformamide dimethyl acetal
(DMF-DMA)18 in methanol resulted, apart from the
formation of adenine, in the formation of 40-epimeric β-
D- and R-L-20,30-didehydro-20,30-dideoxy-40-methoxyade-
nosine (5a) (see Table 1, entry 1). To the best of our
knowledge, no DMF-DMA-mediated allylic rearrangement
of nucleoside derivatives, as described above, has been
reported so far. Typically, an acid catalyzed allylic rear-
rangement of optionally acylated 3-hydroxy glycals invol-
ving C-3 dehydroxylation with nucleophilic substitution is
known as a Ferrier-type rearrangement and is applied to
the transformation of glycals into 2,3-unsaturated glycosyl
derivatives.19 Alcohols, with the exception of phenols,
generally undergo reaction with allyl acetates upon trans-
esterification, and allylic alkylation is achieved only with
complex metal catalysis.20
A series of 30-deoxy-30,40-didehydronucleosides 3bꢀe
were treated with DMF-DMA in methanol (Scheme 1),21
and 40-methoxy derivatives 5bꢀe were obtained in good to
moderate yields (Table 1, entries 2ꢀ5). Whereas the trans-
formation of pyrimidine derivatives proceeded smoothly
(entries 2ꢀ4), providing better isolated yields, the use of
purines (entries 1 and 5) was accompanied by substantial
cleavage of the nucleoside bond (up to 50% according to
the LCMS analyses of the crude reaction mixtures).
Scheme 1. Scope of Syntheses of 40-Alkoxy-20,30-didehydro-
20,30-dideoxynucleosides 5ꢀ13
In addition, it was found that the rearrangement also
proceeded with 30,40-didehydronucleoside-50-aldehyde 2 as
the starting compound; after in situ reduction using
NaBH4 in methanol and a subsequent treatment with
DMF-DMA, 20,30-didehydro-20,30-dideoxy-40-methoxy-
5-methyluridine (5b) was prepared in good yield (entry 6).
The recently published one-pot synthesis of 30-deoxy-30,40-
didehydronucleosides 3 starting from 20,30-O-methoxy-
methylideneribonucleosides 110 encouraged us to adopt a
one-flask approach (Scheme 1), including (a) the oxidation
of 20,30-O-orthoester 1 with DMF-DMSO-EDC-pyridine-
TFA tothe50-aldehyde, (b) the Et3N-mediated elimination
upon formation of the 30,40-unsaturated 50-aldehyde 2, (c)
the reduction of the 50-aldehyde group using NaBH4, (d)
DMF-DMA-mediated allylic rearrangement of the 30,
(8) Haraguchi, K.; Tanaka, H.; Itoh, Y.; Yamaguchi, K.; Miyasaka,
T. J. Org. Chem. 1996, 61, 851–858.
(9) (a) Albert, M.; De Souza, D.; Feiertag, P.; Honig, H. Org. Lett.
2002, 19, 3251–3254. (b) Hegedus, L. S.; Hervert, K. L.; Matsui, S.
J. Org. Chem. 2002, 67, 4076–4080. (c) Hegedus, L. S.; Geisler, L.;
Riches, A. G.; Walkman, S. S.; Umbricht, G. J. Org. Chem. 2002, 67,
7649–7655.
ꢀ
ꢁꢁ
(10) Petrova, M.; Budesı
ꢀ
nsky, M.; Rosenberg, I. Tetrahedron Lett.
´
2010, 51, 6874–6876.
ꢁ
(11) Tocı
´
k, Z.; Arnold, L.; Smrt, J. Nucleic Acids Symp. Ser. 1987, 18,
193–196.
(12) Oyelere, A. K.; Strobel, S. A. J. Am. Chem. Soc. 2000, 122,
10259–10267.
ꢁ
ꢁ
(13) Zemlicka, J. Collect. Czech. Chem. Commun. 1970, 35, 3572–
3583.
(14) Zemlicka, J. Collect. Czech. Chem. Commun. 1968, 33, 3796–
3802.
ꢁ
ꢁ
(19) (a) Ferrier, R. J. J. Chem. Soc. 1964, 5443–5449. (b) Ferrier,
R. J.; Prasad, N. J. Chem. Soc. C 1969, 570–575. (c) Ferrier, R. J.;
Zubkov, O. A. Org. React. 2004, 569–736.
(20) Nakagawa, H.; Hirabayashi, T.; Sakaguchi, S.; Ishii, Y. J. Org.
Chem. 2004, 69, 3474–3477.
(21) 5 Molar excess of DMF-DMA over the starting nucleoside was
used, with subsequent removal of the exocyclic dimethylaminomethy-
lene group from the cytosine and guanine nucleobase using aq ammonia
treatment.
ꢁ
ꢁ
(15) Zemlicka, J.; Gasser, R.; Freisler, J. V.; Horwitz, J. P. J. Am.
Chem. Soc. 1972, 94, 3213–3218.
€
€
(16) Buchi, G.; Cushman, M.; Wuest, H. J. Am. Chem. Soc. 1974, 96,
5563–5565.
(17) Petrova, M.; Budesı
ꢀ
ꢁꢁ
ꢀ
ꢀꢁ
ꢀ
´
nsky, M.; Klepetarova, B.; Rosenberg, I.
Tetrahedron 2011, 67, 4227–4235.
(18) Abdulla, R. F.; Brinkmeyer, R. S. Tetrahedron 1979, 35, 1675–
1735.
Org. Lett., Vol. 13, No. 16, 2011
4201

To examine the role of the primary 50-hydroxy group of
the starting nucleoside 3 in the course of the rearrange-
ment, we protected the 50-hydroxy group with TBDPS.23
As expected, the 20,50-bis-O-silylated compound remained
intact when mixing with DMF-DMA and methanol,
whereas the 50-O-silyl derivative 4a reacted smoothly,
giving the 40-methoxyadenosine derivative 6a in very good
yield (Table 1, entry 14). Next, the reactivity of allyl
alcohol, trifluoroethanol, and 2-azidoethanol toward 4a
was examined, and after deprotection,24 the desired epi-
merically pure 40-alkoxyadenosine derivatives 10aꢀ12a
were obtained, although in low yields (Table 1, entries
15ꢀ17). In these cases, according to LCMS analyses, a
significant amount of furan derivative 14 was identified, as
were minor amounts of the 40-methoxyadenosine deriva-
tives 5a, as discussed above.
As a representative of pyrimidine derivatives more prone
to this rearrangement, 30-deoxy-30,40-didehydro-5-methylur-
idine (3b) was subjected to reaction with 2-methylthioetha-
nol, propargyl alcohol, allyl alcohol, trifluoroethanol, and
2-methoxyethanol. According to LCMS analyses of the
reaction mixtures, both thymine and the 40-methoxy bypro-
duct 5b were identified, with the desired 40-alkoxy derivatives
8bꢀ11b being the major products (entries 18ꢀ21). In con-
trast to all of the adenine 40-alkoxy derivatives, which were
conveniently separated by RP HPLC, the separation of the
40-epimeric thymine derivatives proved to be more laborious;
for example, the epimeric 40-(2-methoxyethoxy) derivative 7b
was resolved into pure epimers only after a 50-O-TBDPS
group had been introduced (entry 23).25
Table 1. Scope of Syntheses of 40-Alkoxy-20,30-didehydro-20,30-
dideoxynucleosides 5ꢀ13
starting
compound
product
isolated
yield
entry
no.
B
no.
B
R
β-D/R-L
1
3a
3b
3c
3d
3e
2b
1a
1c
1d
1e
3a
3a
3a
4a
4a
4a
4a
3b
3b
3b
3b
3b
7b
A
T
U
C
G
T
5a
A
T
U
C
G
T
A
U
C
G
A
A
A
A
A
A
A
T
T
T
T
T
T
CH3
29/10
39/32
28/21
42/13
16/10
38/28
37/19
20/7
2
5b
CH3
3
5c
CH3
4
5d
5e
CH3
5
CH3
6
5b
CH3
7
ABz
5a
CH3
8
U
5c
CH3
9
CBz
5d
5e
CH3
19/15
19/17
8/5
10
11
12
13
14
15
16
17
18
19
20
21
22
23
GDMAM
CH3
A
A
A
A
A
A
A
T
T
T
T
T
T
7a
CH3OCH2CH2
CH3SCH2CH2
HCꢁCCH2
CH3
8a
20/17
26/17
38/22
29/19
6/5
9a
6a
10a
11a
12a
8b
H2C=CHCH2
F3CCH2
N3CH2CH2
CH3SCH2CH2
HCꢁCCH2
H2C=CHCH2
F3CCH2
CH3OCH2CH2
CH3OCH2CH2
13/8
16/13
16/8
9b
10b
11b
7ba
13b
16/12
25/11
52
10/8b
a According to LCMS analysis, a nonseparable mixture of β-D/R-L.
b Overall isolated yield starting from 3b
Scheme 2. Proposed Reaction Mechanisms
40-didehydronucleosides 3, and eventually (e) the final depro-
tection of the nucleobase exocyclic amino group using aq
ammonia. Thus, pure epimers 5 of both purine and pyrimi-
dine nucleosides were obtained in good isolated yields (over
five reaction steps) (Table 1, entries 7ꢀ10). The only draw-
back of this one-flask approach was a rather laborious
isolation of the final compounds from the reaction mixture,
wherein both silica gel and reversed-phase chromatography
had to be employed. To examine the possibility of introducing
other 40-alkoxy groups, the reaction of 30-deoxy-30,40-didehy-
droadenosine (3a) with a series of alcohols was checked, and
the desired epimeric 40-alkoxy derivatives were obtained,
although in lower yields, depending on the reactivity of the
corresponding alcohol (Table 1, entries 11ꢀ13).22 In all cases,
LCMS analysis of the reaction mixture revealed, apart from
the presence of the nucleobase and desired 40-epimeric pro-
ducts, minor formation of epimeric 40-methoxy derivatives 5a,
which likely arose due to the presence of methanol coming
from the transacetalization reaction. An attempt to use DMF
dineopentyl acetal as a less reactive reagent was unsuccessful;
no desired product was formed, and substantial cleavage of
the nucleosidic bond was instead observed.
As has been suggested by the diversity of the reaction
byproducts obtained, the rearrangement reaction mechan-
ism is rather complicated, involving several pathways
leading to the various products. We propose that transa-
cetalization of DMF-DMA with the unprotected 20- and 50-
hydroxy groups of a 30-deoxy-30,40-didehydronucleoside
gives the 20,50-bis-O-dimethylamino(alkoxy)methyl inter-
mediate Im1, which undergoes two main reactions
(23) TBDPSCl in pyridine gave the 20,50-bis-O-silylated derivative as
the main product; TBDPSCl in DMF with Et3N provided the 50-O-silyl
derivative 4a in 60% isolated yield (with only 10% of the bis-silylated
compound).
(24) Using aq ammonia followed by 0.5 M TBAF in THF.
(25) Separation conditions are discussed in detail in the Supporting
Information
(22) The respective alcohol was used as a solvent, or in a mixture
with DMF to improve the solubility of the starting 30,40-
didehydronucleoside.
4202
Org. Lett., Vol. 13, No. 16, 2011

(Scheme 2): the desired allylic rearrangement (blue arrows),
and nucleoside bond cleavage (red arrows). A “pushꢀpull”
mechanism, the simultaneous attack of the alcohol hydroxy
group on a partially positively charged C40 atom, and the
withdrawal of the 20-dimethylamino(alkoxy)methoxy group
from Im1, seems to be the driving force of this allylic
rearrangement. The 50-O-dimethylamino(alkoxy)methyl
group in Im1 appears to be responsible for the nucleoside
bond cleavage (red arrows). This is strongly supported by the
finding that the 50-O-TBDPS group-protected nucleoside
intermediate Im2 provided only traces of nucleobase even in
the absence of alcohol. The extent of the nucleoside bond
cleavage in Im1 was found to depend strongly on the
nucleophilicity of the alcohol used. The less reactive the
alcohol, the more readily the nucleobase is formed. Treat-
ment of the nucleoside 3a with DMF-DMA in the absence of
alcohol led to almost complete nucleoside bond cleavage. To
support the proposed mechanism of the nucleoside bond
cleavage (Scheme 2, the red route), we treated the 20-O-
TBDPS derivative of 3a with DMF-DMA at rt for 16 h
expecting the nucleoside bond cleavage. No nucleobase
formation, however, was observed. This finding seems ap-
parently to be in contradiction to the proposed mechanism
(Scheme 2, the red route), but it is obvious that the nucleo-
base elimination proceeds only in connection with the allylic
rearrangement. In contrast to pyrimidine nucleosides, the
purine nucleosides were found to be more susceptible to
nucleoside bond cleavage. The amount of the purine nucleo-
base formed depended on alcohol reactivity and varied from
several to 50% (LCMS analyses).
The introduction of the 50-O-TBDPS group reduced the
nucleobase formation significantly; however, a new, furan-
based nucleoside byproduct 14 was identified and presum-
ably formed by a cyclic elimination mechanism of the Im2
(Scheme 2, green arrows). We observed that furan 14
formation decreased with the increasing reactivity of the
alcohol used, as was similarly observed for nucleoside
bond cleavage.
As obvious from Table 1, the DMF-DMA mediated
rearrangement proceeds with a moderate stereoselectivity
for the β-D epimer. We can speculate that the 30,40-unsa-
turated sugar ring with the C40-hydroxymethyl moiety is
not strictly planar but can acquire two conformations with
the O40 atom oriented up and below the plane formed of
the sugar ring carbon atoms. In this case, the dihedral
angles C20;C30dC40;O40 of both conformers are differ-
ent from zero, and thus, the C40-hydroxymethyl substitu-
ent could be oriented either up or below the plane. Very
likely, the “upper” orientation of the hydroxymethyl moi-
ety is more advantageous, and thus, a preferential attack of
the alcohol hydroxy group on the C40 is from the less
hindered “bottom” side, giving rise to a slight excess of the
β-D epimer. The highest β-D/R-L ratio (>2:1) was achieved
for 5a, 5d, 5c, and 11b.
carbon atom C-40 of nucleosides 5ꢀ13 was derived from
NOE contacts detected in 2D-H,H-ROESY spectra. Ob-
servedNOE contacts of H-6 (pyrimidines) orH-8 (purines)
with the H-atoms of the 40-substituent proved its cis-
orientation relative to the base (above the dihydrofuran
ring), whereas the observed contacts of H-10 with the
H-atoms of the 40-substituent indicated cis-orientation of
that substituent relative to H-10 (directed below the dihy-
drofuran ring). Very small vicinal coupling (J(10,20) = 1.5
Hz) and large allylic coupling (J(10,30) = 2 Hz), were
observed in the whole series of C40-epimers 5ꢀ13, in
agreement with the preferred OE form. Between the con-
figuration at carbon atom C-40 and chemical shifts of
hydrogen H-10, and carbon atoms C-10 and C-40, the
0
following relation was found: δ H-1 [40S] > δ H-1 [40R]; δ
0
[40S] > δ C-1 [40R]; and δ C-4 [40S] > δ C-4 [40R].
0
0
0
C-10
Significant differences in the IR spectra of all the studied
nucleoside epimer pairs were observed in the region
1000ꢀ1200 cmꢀ1, with several bands of the β-D epimers
occurring at higher frequencies than those of the R-L
epimers.27
In conclusion, we have found that the exposure of
30-deoxy-30,40-didehydronucleosides to DMF-DMA and
alcohol at rt results in a Ferrier-type glycal allylic rearran-
gement upon formation of the C40-epimeric 40-alkoxy-
20,30-didehydro-20,30-dideoxynucleosides. Although the
formation of these compounds is accompanied by nucleo-
side bond cleavage and/or formation of a furan nucleoside,
pure β-D and R-L epimers were obtained by means of
preparative reversed-phase and/or silica gel chromatogra-
phy. Moreover, a direct, one-pot synthesis of 40-methoxy-
20,30-didehydro-20,30-dideoxynucleosides was successfully
elucidated, starting from the readily available 20,30-O-
methoxymethylideneribonucleosides. We expect that this
simple approach toward a wide range of 40-alkoxy-20,30-
didehydro-20,30-dideoxynucleosides will become a useful
tool in the search for biologically active compounds.
Despite therelatively low yields of the 40-alkoxy derivatives
in some cases, the reaction as introduced is worth further
study because of the ready availability of the starting 30,40-
unsaturated nucleosides. Studies on antiviral and antic-
ancer properties of the prepared nucleoside analogues are
in progress, and results will be reported in due course.
Acknowledgment. Financial support was provided by
Grants 203/09/0820 and 202/09/0193 (Czech Science
Foundation) and by Research Centers KAN200520801
(Acad. Sci. CR) and LC06077 (Ministry of Education,
CR), under the Institute Research Project Z40550506. The
authors are indebted to the staff of the Mass Spectrometry
ꢁ
Group of this Institute (Dr. Josef Cvacka, Head) for
measurements of HR-MS spectra.
Supporting Information Available. Full experimental
and compound characterization data. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
The structures of all of the nucleosides were confirmed
1
by H and 13C NMR spectra.26 The configuration at
(26) See Tables 1S and 2S in the Supporting Information
(27) As illustrated in detail in the Supporting Information
Org. Lett., Vol. 13, No. 16, 2011
4203