Synthetic studies towards new nucleoside analogues: Preparation of (±)-1′,4′-dimethyladenosine

Article abstract of DOI:10.1002/ejoc.201001216

Racemic 1′,4′-dimethyladenosine was prepared according to a stereoselective sequence from 2,5-dimethyl furan and vinylene carbonate. This sequence is short and efficient (nine steps) and relies on the desymmetrization of a meso diol, followed by the glyco

Full text of DOI:10.1002/ejoc.201001216

FULL PAPER
DOI: 10.1002/ejoc.201001216
Synthetic Studies towards New Nucleoside Analogues: Preparation of
(؎)-1Ј,4Ј-Dimethyladenosine
Guilhem Chaubet,[a] Damien Bourgeois,*[a] and Christian Périgaud[a]
Keywords: Nucleosides / Synthesis design / Cycloaddition / Enantioselectivity / Glycosylation
Racemic 1Ј,4Ј-dimethyladenosine was prepared according to
a stereoselective sequence from 2,5-dimethyl furan and vi-
nylene carbonate. This sequence is short and efficient (nine
steps) and relies on the desymmetrization of a meso diol, fol-
lowed by the glycosylation of the anomeric position. The lat-
ter reaction was thoroughly studied, as it is the key step of
the sequence, and very particular reactivity of the 1Ј,4Ј-di-
methylated sugars was observed. Secondary reactions took
place according to original mechanisms and delivered origi-
nal byproducts.
reaction. However, we found it is limited to the synthesis of
pyrimidine analogues, as the acetonide deprotection in the
Introduction
Nucleoside analogues constitute an important class of bio- purine series was unsuccessful, probably because of a
logically active compounds, and the synthesis of new mole- greater sensitivity of these substrates to the required acidic
cules of this class with original substitution patterns re- conditions. It is of key importance to circumvent this severe
mains a challenge for therapeutic chemists.[1] During the drawback, as purine derivatives, and especially adenosine
course of our synthetic studies towards 1Ј,4Ј-disubstituted analogues, are relevant for biological studies.[3] During our
nucleoside analogues, we developed a synthesis of 1Ј,4Ј-di- studies on the Diels–Alder reaction between 1 and 2, we
methyl uridine, starting from 2,5-dimethyl furan (1) and vi- evidenced that prolonged heating leads to clean conversion
nylene carbonate (2).[2] This synthetic route is efficient, as of the endo/exo mixture into exo adduct 3 only, enabling
it overcomes the poor selectivity of the initial Diels–Alder large-scale production of the latter on a preparative scale
Scheme 1. Envisioned pathways towards 1Ј,4Ј-disubstituted adenosine analogues.
after direct crystallization.[4] Therefore, we decided to
[a] Institut des Biomolécules Max Mousseron (IBMM)
UMR 5247 CNRS-UM1-UM2,
exploit this result to design another route to 1Ј,4Ј-dimethyl
Université Montpellier 2, CC1705,
nucleoside analogues, which is both shorter and suitable for
purine synthesis (Scheme 1). It is the purpose of this com-
munication to report our first results regarding this ap-
proach.
Place Eugène Bataillon, 34095 Montpellier Cedex 5, France
Fax: +33-4-67042029
E-mail: damien.bourgeois@univ-montp2.fr
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201001216.
Eur. J. Org. Chem. 2011, 319–326
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
319

G. Chaubet, D. Bourgeois, C. Périgaud
FULL PAPER
Results and Discussion
Our synthetic analysis is very simple (Scheme 1): oxidat-
ive cleavage of the olefinic double bond in 3 should reveal
the 5Ј-hydroxy group, as well as an aldehyde, which would,
after Baeyer–Villiger reaction, lead to a sugar bearing an
activated anomeric center for heterocyclic base introduc-
tion. Thus, exo Diels–Alder adduct 3 possesses the correct
relative configurations at the 2Ј- and 3Ј-oxygen atoms of the
desired nucleoside analogues.
Scheme 3. Preparation of activated sugars 9 and 10.
There are several possibilities to deliver directly the de-
sired aldehyde moiety with an oxidative cleavage. We exam-
ined several ozonolysis methods, including unsymmetrical
versions.[5] However, none gave satisfactory results on our
substrate. Even the classical Me2S- or PPh3-based quenches
did not lead to the corresponding dialdehyde on several
substrates of this family in a clean manner. Therefore, we
performed ozonolysis followed by NaBH4 reduction, which
proved to be very efficient.
On a first attempt, we tried to keep the cyclic carbonate
as a protecting group. However, this group did not with-
stand the reductive conditions, and most of the product was
lost during treatment in the aqueous layer, probably in a
tetraol form. Thus, we hydrolyzed the carbonate of 3 with
potassium carbonate in methanol, then protected the re-
sulting diol as a bis-benzoate with benzoyl chloride in pyr-
idine, with an excellent yield (87%) after crystallization
(Scheme 2). Ozonolysis of 4 in DCM/MeOH at –78 °C fol-
lowed by immediate treatment with NaBH4 delivered diol
5 in 72% yield after crystallization. On a small scale, a bet-
ter yield (87%) can be obtained after purification by silica
gel chromatography.
efficient when silylating agents were used. The second sil-
ylation was more rapid, producing a large amount of bis-
silylated compound. Thus, employing TBSCl (1.0 equiv.),
we isolated desired product 7 in 37% yield, along with start-
ing diol (23%) and bis-silylated product 8 (28%). The cor-
rected yield for 7 is 48%, and the three compounds can be
separated easily by flash column chromatography on silica
gel thanks to the large difference in their Rf values. How-
ever, it is clear that a preparative-scale process cannot be
designed from these sole results, although bis-silylated de-
rivative 8 can be easily recycled into starting diol 5 after
fluoride-mediated deprotection.[7]
With monoprotected alcohols 6 and 7 in hand, we
studied the transformation of the remaining hydroxymethyl
group into a formate group to deliver the protected acti-
vated sugar required for heterocyclic base introduction. The
oxidation of the alcohol moiety into the corresponding al-
dehyde was best effected with IBX in DMSO.[8] The yield
was moderate and variable (60 to 75%) but counterbal-
anced by the quantitative Baeyer–Villiger oxidation on the
same aldehyde (Scheme 3). Furthermore, we demonstrated
that no purification of the intermediate aldehyde is re-
quired. Thus, compound 9 was obtained in 66% yield over
the two steps starting from benzoate 6, and 10 was obtained
in 72% yield from TBS derivative 7. The formate can be
purified by chromatography on silica gel, and it is stable for
several months when stored in the freezer. On the other
hand, analysis by mass spectrometry did not reveal the peak
corresponding to the product. However, under the mass
spectrometry conditions, that is, electrospray ionization
(ESI), the sample must be heated to a minimum of 250 °C,
Scheme 2. Preparation of meso diol 5.
Diol 5 is also a very interesting intermediate: it is a meso a temperature at which the compound is not stable, even in
compound, and several methods are described regarding the neutral mode. The moderate yield for this transforma-
the enantiotopic differentiation of both alcohol groups on tion can be explained by the ease of over-oxidation of the
such substrates. Lipase-catalyzed acylation is envisioned for aldehyde: the latter could not be isolated with a purity
an enantioselective synthesis based on this work.[6] Before greater than 90% and showed decomposition to a baseline
studying this point, we wanted to prove that the synthetic product by 2D TLC analysis. Basic aqueous treatment at
pathway could efficiently deliver the desired nucleoside ana- the Baeyer–Villiger step may then lead to the loss of the
logues. Therefore, we pursued a racemic synthesis, em- corresponding carboxylic acid.
ploying either a benzoate or a tert-butyldimethylsilyl (TBS)
With activated sugars 9 and 10 in hands, we studied the
protecting group (Scheme 3).
introduction of various heterocyclic bases (Scheme 4 and
Monoprotection of diol 5 with a benzoate group was per- Table 1). Among the numerous possibilities, we first focused
formed by using a slight deficiency of the protecting agent, on Vorbrüggen coupling, employing a silylated base with
which allowed almost complete avoidance of bis-benzo- Lewis acid activation.[9] This method is widely applicable,
ylation. Using benzoyl chloride (0.9 equiv.), 22% of the as it usually gives high yield and high selectivity. First, we
starting diol was easily recovered after purification by silica studied the introduction of the uracil moiety. As noticed
gel chromatography, delivering tribenzoate 6 in 83% cor- during our previous studies,[2] we had to work at low tem-
rected yield. On the other hand, the same method was less perature, in contrast to the generally encountered operating
320
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 319–326

Preparation of (Ϯ)-1Ј,4Ј-Dimethyladenosine
Table 1. Glycosylation results on substrates 9 and 10.
Entry
Conditions
Product
Yield [%]
β/α selectivity
1
2
bis-silylated uracil (2.0 equiv.), 9, TMSOTf (1.0 equiv.), –20 °C, 1 h
bis-silylated uracil (2.0 equiv.), 9, TMSOTf (1.0 equiv.), –20 °C to r.t.
11
11
70
35
25
50
27
43
63
48
3:1
3:1
14
3
4
5
6
7
bis-silylated uracil (2.0 equiv.), 9, TMSOTf (1.0 equiv.), –20 to 70 °C
bis-silylated uracil (2.0 equiv.), 10, TMSOTf (1.0 equiv.), –20 °C, 1 h
bis-silylated N6-Bz-Adenine (2.0 equiv.), 9, TMSOTf (1.0 equiv.), –20 °C, 1 h
adenine (2.0 equiv.), SnCl4 (4.0 equiv.), 9, –20 °C, 1 h
14
13
12 (Bz)
12
15
3:1
2:1
1:1
adenine (2.0 equiv.), SnCl4 (4.0 equiv.), 10, –20 °C to r.t.
conditions. Also, to optimize the yield and selectivity of the
reaction, we studied the impact of the different parameters
on the Vorbrüggen reaction outcome.
Scheme 4. Glycosylation of 9 and 10.
Scheme 5. Proposed mechanism of glycosylation of 9 and 10.
The stability of exo-glycal 14 is striking compared to the
The only parameter to have a real influence on the reac-
tion was the temperature (Table 1). We showed that at low
temperature (below –30 °C) that the reaction is sluggish and
that starting material is still present. The best conversion ease of hydrolysis of common exo-glycals on silica gel.[11]
was achieved at –20 °C (70% isolated yield; Table 1, En- 2D TLC analysis showed absolutely no degradation of 14
try 1). When warming to room temperature, a new spot was on standard chromatography silica gel. We propose two
detected after TLC analysis of the reaction mixture. Isola- reasons for this stability: the electron-withdrawing effect of
tion of this compound, followed by spectroscopic analysis three benzoate protecting groups can significantly lower the
revealed the formation of exo methylene 14 (Scheme 4; electrophilicity of the enol ether, and also the steric hin-
Table 1, Entry 2). Further warming of the reaction medium drance on the furanose ring, mainly the presence of the 4Ј-
leads to an increase in the formation of this compound and quaternary center, may disfavor sp3 hybridization of the 1Ј-
to the complete disappearance of Vorbrüggen adducts 11 carbon to avoid 1,3-diaxial interactions. Both reasons are
(Table 1, Entry 3).
also invoked in the discussion on the selectivity of the reac-
Such a compound can result from β-hydrogen elimi- tion.
nation on the intermediate carbocation, probably induced
Regarding the selectivity, these results are somehow dis-
by the Lewis acid conjugated base (Scheme 5). Examining appointing: we observed a modest selectivity in favor of the
the mass spectrum of formate 9, we noticed that the peak β anomer (3:1). The configuration of the anomeric center
corresponding to elimination product 14 is also present, was further confirmed after deprotection of the benzoate
suggesting its easy formation. We eventually showed that protecting groups: the obtained compound was identical to
heating formate 9 at 80 °C in the presence of silica gel leads the derivative prepared by our first route (see below). This
to relatively clean conversion directly into 14. Such a direct behavior during a Vorbrüggen reaction is noteworthy. Usu-
transformation on ketose derivatives was never reported be- ally, the accepted mechanism proposes a thermodynamic
fore to the best of our knowledge, as it usually requires equilibration, favoring the β anomer, through a bridged
activation with a better leaving group and base catalysis.[10] oxonium with the 2Ј-benzoate group.[12] Also, the intramo-
Eur. J. Org. Chem. 2011, 319–326
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
321

G. Chaubet, D. Bourgeois, C. Périgaud
FULL PAPER
lecular participation of the 5Ј-benzoate, through a β-face- tion where the carbonyl group lies above the C1Ј center.
bridged oxonium, which then induces preferred α attack, The last result gives consistence to the above proposed 5Ј-
was already reported on similar substrates.[13] Both interme- benzoate assistance, which leads to a seven-membered ring
diates are depicted in Scheme 5.
(Scheme 5).
The β/α ratio we observe does not vary meaningfully
To complete the synthesis, the three benzoate groups of
with the temperature, suggesting that the reaction stays un- 11 can be cleanly removed simultaneously after treatment
der kinetic control. Any attempt to reach a thermodynamic with NH3 in MeOH (Scheme 6). Desired isomer 16β is ob-
equilibration led instead to exo methylene adduct 14. exo- tained pure after column chromatography on silica gel in
Glycal 14 appears thus to be the thermodynamic well of the 64% yield (β isomer only). NMR spectroscopic analysis
reaction, and we do not think that it is an intermediate en shows identical properties as the compound obtained by
route to protected nucleosides 11–13, as described before our previously described route, bearing all the correct rela-
during the synthesis of complex disaccharides.[14] It is not tive stereochemistries. Thus, the route described in this arti-
possible to completely exclude that compound 14 results cle is three steps shorter (9 steps instead of 12 starting from
from the elimination of the heterocyclic base from com- 1 and 2) and gives slightly higher yields (9% global yield
pound 11, 12, or 13. However, treatment of formate 9 with vs. 7%).
TMSOTf alone in acetonitrile (at –20 °C to r.t.) leads pre-
dominantly to exo-glycal 14, albeit accompanied by several
side products of the same polarity. These conditions may
be slightly too harsh and lead to partial degradation of
compound 14 but also suggest its direct formation from the
intermediate carbocation through β-hydrogen elimination
(Scheme 5). Such a reaction is entropically favored, in con-
trast to base condensation, and we thus propose that any
attempt at α to β equilibration on protected nucleosides 11–
13 will eventually lead to 14. This behavior is specific to our
1Ј,4Ј-dimethylated substrates and has never been observed
before, to the best of our knowledge.
We then studied the same coupling with a 5Ј-silyl group,
a typical nonparticipating group. The reaction was run with
Scheme 6. Completion of the synthesis.
silyl derivative 10, but the selectivity was exactly the same
(3:1) at –20 °C (Table 1, Entry 4). Thus, we think that on
We then tried to apply this route to the synthesis of pur-
the substrates we are studying, the steric bulk on the β face ine nucleoside analogues. With bis-silylated N6-benzoyl-
of the molecules is important and counterbalances to some ated adenine as a base, Vorbrüggen coupling of formate 9
extent the shielding of the α face by the benzoate group. was quite sluggish (Table 1, Entry 5). A poor yield was ob-
We modeled the (simplified) intermediate carbocations with tained (43%) with a low selectivity (2:1). Therefore, we also
DFT calculations by using Gaussian (B3LYP/6-31G**), employed Saneyoshi’s conditions with unsilylated adenine
and it appeared that the 5Ј-group prefers to lie in an axial and SnCl4 catalysis. Unfortunately, there was no β/α selec-
position to avoid 1,2- or 1,3-diaxial interactions between tivity (Table 1, Entry 6). Analysis of the 1H NMR spectrum
the 4Ј-methyl and the 3Ј- and 2Ј-benzoate group, respec- of the crude reaction mixture showed no other signal in the
tively (Figure 1). The shielding of the β face in the case of 4.0 to 6.5 ppm region, which can be attributed to the 2Ј- or
the benzoate group is enhanced by the preferred conforma- 3Ј-protons of other regioisomers such as N7 adducts. TLC
analysis also showed that no starting material remained,
but several low polarity products were detected; amongst
them exo-glycal 14 was present. The yield remains good
regarding this type of coupling and reflects the lower reac-
tivity of the purine heterocycles compared to that of the
pyrimidines.
However, deprotection of the benzoate groups led to a
separable mixture of both β and α anomers. β Anomer 17β
was thus obtained in 31% yield and was fully characterized
(Scheme 6). Nuclear Overhauser effect (nOe) studies con-
firmed the relative configuration of the stereocenters: we
performed nOe difference experiments, which revealed that
the adenine ring, the 5Ј-group, and the 2Ј- and 3Ј-protons
are on the same side of the molecule (Figure 2). As with the
uracil derivative, we did not observe a nOe between the
methyl groups, suggesting a pseudoequatorial conformation
for both groups.
Figure 1. Optimized geometry of the intermediate carbocations en-
compassing a 5Ј-acetate or a 5Ј-trimethylsilyl group (B3LYP/6-
31G**).
322
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 319–326

Preparation of (Ϯ)-1Ј,4Ј-Dimethyladenosine
able to study this new family of 1Ј,4Ј-disubstituted nucleo-
side analogues and their biological applications.
Experimental Section
General Procedures: Unless noted otherwise, all starting materials
and reagents were obtained from commercial suppliers and were
used without further purification. Tetrahydrofuran was distilled
from sodium benzophenone ketyl. Dichloromethane, triethylamine,
acetonitrile, and pyridine were freshly distilled from calcium hy-
dride. All solvents used for routine isolation of products and
chromatography were reagent grade. Reaction flasks were dried at
100 °C. Air- and moisture-sensitive reactions were performed under
an argon atmosphere. Flash column chromatography was per-
formed by using silica gel 60 (230–400 mesh, Merck) with the indi-
cated solvents. Thin-layer chromatography was performed by using
0.25 mm silica gel plates (Merck). Melting points were determined
in open capillary tubes with a Büchi-545. UV spectra were recorded
with a Uvikon 931 (Kontron). Infrared spectra were recorded with
a Perkin–Elmer Paragon 1000 FTIR spectrometer. 1H NMR and
13C NMR spectra were recorded at 300 K with Bruker 300 Avance
and DRX 400 spectrometers in solutions in the indicated solvent.
Chemical shifts are expressed in parts per million downfield from
tetramethylsilane and are referenced to the residual solvent peak
(1H NMR: CHCl3, δ = 7.26 ppm; [D6]DMSO, δ = 2.50 ppm) or
the deuterated solvent peaks (13C NMR: CDCl3, δ = 77.16 ppm;
[D6]DMSO, δ = 39.52 ppm). When possible, hydroxy groups were
identified by D2O swap. FAB mass spectra were recorded in the
positive-ion or negative-ion mode with a JEOL SX 102. The matrix
was a mixture (50:50) of glycerol and thioglycerol (G/T). Electro-
spray ionization (ESI) mass spectra and high-resolution mass spec-
tra were obtained with a Waters Q-TOF.
Figure 2. Observed nOe interactions on compound 17β.
Finally, we also tried Saneyoshi’s conditions[15] with si-
lylated derivative 10 (Table 1, Entry 7). To our surprise, no
coupling adduct was obtained, but a clean cyclization reac-
tion took place to afford bicyclic product 15 (Scheme 4).
The structure of 15 was fully established by IR spec-
troscopy, which showed no free hydroxy band. Other spec-
troscopic data, such as 1H and 13C NMR spectroscopy and
MS, are in full agreement with this structure. Such a cycliza-
tion implies either nucleophilic attack of the TBS ether onto
the intermediate carbocation or prior deprotection of the
silyl ether, followed by addition of the resulting oxide onto
the same carbocation. As TBS groups are described to be
stable under SnCl4 treatment at 0 °C[16] and because they
are poorly nucleophilic, we propose a concerted mecha-
nism, where chloride-mediated desilylation is favored by the
activation of the silicon oxygen bond through oxygen lone
pair interaction with the carbocationic center (Scheme 7).
1,4-Dimethyl-7-oxabicyclo[2.2.1]hept-5-ene-2,3-diyl Dibenzoate (4):
To a solution of carbonate 3 (4.65 g, 25.6 mmol) in methanol
(50 mL) was added K2CO3 (3.90 g, 28.1 mmol, 1.1 equiv.), and the
resulting slurry was stirred for 2 h at room temperature, then fil-
tered, and concentrated in vacuo. The resulting solid was diluted
in DCM (24 mL) and pyridine (16 mL) and cooled to 0 °C. To the
resulting solution was added benzoyl chloride (6.5 mL, 56.3 mmol,
2.2 equiv.) dropwise; a white precipitate appeared. The resulting
mixture was stirred at room temperature for 3 h. After dilution
with AcOEt (50 mL), the mixture was quenched with a saturated
aqueous NH4Cl solution (40 mL), and the layers were separated.
The organic layer was washed successively with water (40 mL), 1 n
HCl (2ϫ40 mL), water (40 mL), and brine (40 mL). It was then
dried with MgSO4, filtered, and concentrated in vacuo. Crystalli-
zation (petroleum ether/ether, 2:1) yielded a white solid (8.15 g,
Scheme 7. Proposed mechanism for the formation of 15.
This cyclization proves that the presence of the methyl
groups tends to induce the adoption of a bicyclic structure;
thus, the 5Ј-protected hydroxy group strongly prefers an ax-
ial position. This point could explain the low selectivity ob-
served during Vorbrüggen coupling: the bridging of the ox-
onium with the 5Ј-benzoate should be favored by the con-
formation of the molecule.
69%). M.p. 136–137 °C. IR (neat): ν = 3061, 2980, 2936, 1728,
˜
1601, 1584, 1491, 1451, 1382, 1354, 1317, 1283, 1177, 1120, 1071,
1040 cm–1. 1H NMR (300 MHz, CDCl3): δ = 1.63 (s, 6 H, Me),
5.32 (s, 2 H, 2-H and 3-H), 6.39 (s, 2 H, 5-H and 6-H), 7.24 (m, 4
H, Ar-H), 7.46 (m, 2 H, Ar-H), 7.89 (m, 4 H, Ar-H) ppm. 13C
NMR (75 MHz, CDCl3): δ = 166.1, 140.8, 133.2, 129.8, 129.5,
128.2, 87.5, 73.2, 14.6 ppm. MS (ESI+): m/z = 365 [M + H]+.
HRMS (ESI+): calcd. for C22H21O5 [M + H]+ 365.1389; found
365.1382.
Conclusions
As a conclusion, we described here another route to
1Ј,4Ј-dimethylated nucleoside analogues. This route is ad-
vantageous, as it enables access to both pyrimidine and pu-
rine derivatives with a short reaction sequence and good
yields. This study also revealed the unique behavior of this
family of substrates during glycosylation. Further studies
will be devoted to the study of the structural parameters
influencing the selectivity of this reaction, as well as to the
2,5-Bis(hydroxymethyl)-2,5-dimethyltetrahydrofuran-3,4-diyl
Di-
benzoate (5): A O3/O2/N2 stream was bubbled at –78 °C into a solu-
tion of dibenzoate 4 (5.4 g, 14.8 mmol) in DCM/MeOH (5:1,
120 mL) with a drop of Sudan red solution (0.6 mg in 2 mL
MeOH) until complete disappearance of the red color (approx.
final design of an enantioselective route. Finally, we will be 30 min). Then, solid NaBH4 (5.6 g, 150 mmol, 10 equiv.) was added
Eur. J. Org. Chem. 2011, 319–326
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
323

G. Chaubet, D. Bourgeois, C. Périgaud
FULL PAPER
portionwise, and the resulting mixture was stirred at room tempera-
ture until H2 bubbling stopped. Then the mixture was diluted with
DCM (150 mL), quenched slowly with a saturated aqueous NH4Cl
solution (60 mL), and the layers were separated. The organic layer
was washed with water (2ϫ30 mL) and brine (30 mL), dried with
Na2SO4, and concentrated in vacuo. The obtained solid was tritu-
rated with petroleum ether/AcOEt (1:1), filtered, and washed with
the same solvent (2ϫ). Diol 5 was obtained as a white solid (3.7 g,
3.45–3.59 (m, 3 H, OH and 2-CH2 and/or 5-CH2), 3.62 (d, J =
9.6 Hz, 1 H, 2-CH2 or 5-CH2), 5.78 (d, J = 5.8 Hz, 1 H, 3-H or 4-
H), 5.88 (d, J = 5.8 Hz, 1 H, 3-H or 4-H), 7.19–7.30 (m, 4 H, Ar-
H), 7.37–7.46 (m, 2 H, Ar-H), 7.79–7.88 (m, 4 H, Ar-H) ppm. 13C
NMR (75 MHz, CDCl3): δ = 165.3, 165.1, 133.3, 133.2, 129.7,
129.6, 129.5, 128.5, 128.4, 84.8, 84.2, 74.7, 73.3, 68.6, 68.5, 26.0,
20.4, 20.0, 18.6, –5.3, –5.5 ppm. MS (ESI+): m/z = 515 [M + H]+.
HRMS (ESI+): calcd. for C28H39O7Si [M + H]+ 515.2465; found
72%). M.p. 124–125 °C. IR (neat): ν = 3422, 3058, 2976, 2924,
515.2463. Data for 8: IR (neat): ν = 2954, 2929, 2857, 1729, 1451,
˜
˜
1726, 1601, 1451, 1315, 1279, 1124, 1067 cm–1. 1H NMR
1314, 1275, 1258, 1092, 1068, 837, 775, 703 cm–1 1H NMR
.
(300 MHz, CDCl3): δ = 1.38 (s, 6 H, Me), 3.71 (s, 4 H, 2-CH2 and (300 MHz, CDCl3): δ = 0.10 (s, 12 H, Si-Me), 0.91 (s, 18 H, Si-
5-CH2), 3.87 (s, 2 H, OH), 5.97 (s, 2 H, 3-H and 4-H), 7.34–7.39 CMe3), 1.43 (s, 6 H, 2-Me and 5-Me), 3.57 (d, J = 10.2 Hz, 2 H,
(m, 4 H, Ar-H), 7.50–7.58 (m, 2 H, Ar-H), 7.90–7.97 (m, 4 H, Ar- 2-CH2 and 5-CH2), 3.67 (d, J = 10.1 Hz, 2 H, 2-CH2 and 5-CH2),
H) ppm. 13C NMR (75 MHz, CDCl3): δ = 165.6, 133.5, 129.8,
129.5, 128.6, 84.6, 74.3, 68.2, 20.2 ppm. MS (ESI+): m/z = 401 [M
+ H]+. HRMS (ESI+): calcd. for C22H25O7 [M + H]+ 401.1600;
found 401.1606.
5.86 (s, 2 H, 3-H and 4-H), 7.32–7.53 (m, 6 H, Ar-H), 7.92–7.95
(m, 4 H, Ar-H) ppm. 13C NMR (75 MHz, CDCl3): δ = 165.1,
133.1, 129.9, 129.7, 128.4, 83.8, 74.7, 69.1, 26.0, 20.4, 18.4, –5.2,
–5.3 ppm. MS (ESI+): m/z = 629 [M + H]+. HRMS (ESI+): calcd.
for C34H53O7Si2 [M + H]+ 629.3330; found 629.3311.
2-[(Benzoyloxy)methyl]-5-(hydroxymethyl)-2,5-dimethyltetrahydro-
furan-3,4-diyl Dibenzoate (6): To a solution of diol 5 (947 mg,
2.4 mmol) in DCM (15 mL) at 0 °C was added NEt3 (1.0 mL,
7.1 mmol, 3.0 equiv.), DMAP (15 mg, 0.12 mmol, 5 mol-%), and
BzCl (250 μL, 2.1 mmol, 0.9 equiv.) dropwise. A white precipitate
formed, and the resulting mixture was stirred at room temperature
for 2 h. The reaction medium was diluted with AcOEt (50 mL) and
quenched with MeOH (2 mL), then with a saturated aqueous
NH4Cl solution (15 mL). The layers were separated, the organic
2-[(Benzoyloxy)methyl]-5-(formyloxy)-2,5-dimethyltetrahydrofuran-
3,4-diyl Dibenzoate (9): To a solution of alcohol 6 (772 mg,
1.5 mmol) in DMSO (15 mL) at 15 °C was added 2-iodoxybenzoic
acid (IBX, 857 mg, 3.0 mmol, 2.0 equiv.) portionwise over 10 min.
The resulting solution was stirred for 2 h at 35 °C, then cooled to
room temperature, quenched with water (30 mL), and diluted with
AcOEt (50 mL). A white precipitate formed, and the mixture was
filtered through a sintered funnel. The layers were separated, and
layer was washed with water (2ϫ15 mL) and brine (15 mL), dried the organic layer was washed with water (3ϫ15 mL) and brine
with Na2SO4, and concentrated in vacuo. Purification by flash (15 mL), dried with MgSO4, and concentrated in vacuo. The crude
chromatography on silica gel (petroleum ether/AcOEt, 70:30 to
50:50) yielded product 6 as a colorless oil (772 mg, 65%) and start-
aldehyde was taken into DCM (25 mL) at 0 °C. Solid NaHCO3
(252 mg, 3.0 mmol, 2.0 equiv.) was added all at once, and then m-
CPBA (70% purity, 550 mg, 2.25 mmol, 1.5 equiv.) was added por-
ing material 5 (212 mg, 22%). IR (neat): ν = 3492, 3063, 2984,
˜
2936, 2976, 1721, 1601, 1451, 1379, 1263, 1176, 1095, 1067, tionwise over 10 min. The resulting mixture was stirred for 15 min
1025 cm–1. 1H NMR (300 MHz, CDCl3): δ = 1.45 (s, 3 H, Me),
1.57 (s, 3 H, Me), 2.80 (dd, J = 8.3, 5.2 Hz, 1 H, OH), 3.54–3.67
(m, 2 H, 5-CH2), 4.41 (d, J = 11.6 Hz, 1 H, 2-CH2), 4.56 (d, J =
11.6 Hz, 1 H, 2-CH2), 5.94 (d, J = 6.1 Hz, 1 H, 3-H or 4-H), 5.98
(d, J = 6.1 Hz, 1 H, 3-H or 4-H), 7.33–7.42 (m, 3 H, Ar-H), 7.43–
7.49 (m, 3 H, Ar-H), 7.51–7.62 (m, 3 H, Ar-H), 7.93–7.99 (m, 4 H,
at room temperature. It was then quenched with an aqueous thio-
sulfate solution (10 mL), and AcOEt (50 mL) was added. The lay-
ers were separated, and the organic layer was washed with an aque-
ous thiosulfate solution (20 mL), an aqueous NaHCO3 solution
(20 mL), and brine (20 mL), dried with MgSO4, and concentrated
in vacuo. Purification by flash chromatography on silica gel (petro-
Ar-H), 8.10–8.15 (m, 2 H, Ar-H) ppm. 13C NMR (75 MHz, leum ether/AcOEt, 80:20) yielded formate 9 as a colorless oil
CDCl3): δ = 166.7, 165.4, 165.2, 133.5, 133.5, 133.4, 129.9, 129.8,
129.7, 129.7, 129.4, 129.3, 128.7, 128.6, 128.6, 84.8, 82.9, 74.3, 73.4,
69.2, 67.4, 20.6, 19.8 ppm. MS (ESI+): m/z = 505 [M + H]+.
HRMS (ESI+): calcd. for C29H29O8 [M + H]+ 505.1862; found
505.1859.
(513 mg, 66%). IR (neat): ν = 3067, 2996, 2943, 1722, 1601, 1451,
˜
1380, 1261, 1175, 1091, 1068, 1024 cm–1 1H NMR (300 MHz,
.
CDCl3): δ = 1.61 (s, 3 H, Me), 1.92 (s, 3 H, Me), 4.42 (d, J =
11.7 Hz, 1 H, 2-CH2), 4.59 (d, J = 11.7 Hz, 1 H, 2-CH2), 6.05 (d,
J = 6.1 Hz, 1 H, 3-H or 4-H), 6.08 (d, J = 6.1 Hz, 1 H, 3-H or 4-
H), 7.30–7.38 (m, 3 H, Ar-H), 7.40–7.48 (m, 3 H, Ar-H), 7.51–7.66
(m, 3 H, Ar-H), 7.85–7.88 (m, 4 H, Ar-H), 8.02–8.06 (m, 2 H, Ar-
H), 8.33 (s, 1 H, Hformate) ppm. 13C NMR (75 MHz, CDCl3): δ =
166.0, 165.0, 164.7, 159.2, 133.8, 133.7, 133.3, 129.8, 129.8, 129.8,
129.7, 129.0, 128.8, 128.6, 128.5, 110.6, 85.5, 76.4, 72.5, 68.6, 20.9,
20.4 ppm. MS (ESI+): m/z = 473 [M + H – HCOOH]+. HRMS
(ESI+): calcd. for C28H25O7 [M + H – HCOOH]+ 473.1600; found
473.1594.
2-{[(tert-Butyldimethylsilyl)oxy]methyl}-5-(hydroxymethyl)-2,5-di-
methyltetrahydrofuran-3,4-diyl Dibenzoate (7): To a solution of diol
5 (452 mg, 1.1 mmol) in dry DMF (4 mL) at 0 °C was added imid-
azole (390 mg, 5.6 mmol, 5.0 equiv.) and TBSCl (170 mg,
1.1 mmol, 1.0 equiv.) portionwise. A white precipitate formed, and
the resulting mixture was stirred at room temperature for 24 h. The
reaction medium was quenched with MeOH (2 mL), then diluted
with AcOEt (50 mL), and a saturated aqueous NH4Cl solution
(10 mL) was added. The layers were separated, and the organic
layer was washed with water (3ϫ10 mL) and brine (10 mL), dried
with MgSO4, and concentrated in vacuo. Purification by flash
chromatography on silica gel (petroleum ether/AcOEt, 80:20 to
40:60) yielded bis-protected compound 8 (229 mg, 28%), product
7 as a colorless oil (192 mg, 37%), and starting material 5 (105 mg,
2-{[(tert-Butyldimethylsilyl)oxy]methyl}-5-(formyloxy)-2,5-dimethyl-
tetrahydrofuran-3,4-diyl Dibenzoate (10): To a solution of alcohol 7
(373 mg, 0.7 mmol) in DMSO (7 mL) at 15 °C was added 2-iodoxy-
benzoic acid (IBX, 408 mg, 1.4 mmol, 2.0 equiv.) portionwise over
5 min. The resulting solution was stirred for 2 h at 35 °C, then co-
oled to room temperature, quenched with water (15 mL), and di-
luted with AcOEt (30 mL). A white precipitate formed, and the
mixture was filtered through a sintered funnel. The layers were sep-
arated, and the organic layer was washed with water (3ϫ10 mL)
23%). Data for 7: IR (neat): ν = 3447, 2931, 2858, 1727, 1451,
˜
1
1314, 1275, 1260, 1122, 1094, 1068, 1025, 837, 705 cm–1. H NMR
(300 MHz, CDCl3): δ = 0.07 (s, 3 H, Si-Me), 0.09 (s, 3 H, Si-Me),
0.87 (s, 9 H, Si-CMe3), 1.26 (s, 3 H, 2-Me or 5-Me), 1.29 (s, 3 H, and brine (10 mL), dried with MgSO4, and concentrated in vacuo.
2-Me or 5-Me), 3.38 (dd, J = 9.6, 3.2 Hz, 1 H, 2-CH2 or 5-CH2), The crude aldehyde was taken into DCM (10 mL) at 0 °C. Solid
324
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 319–326

Preparation of (Ϯ)-1Ј,4Ј-Dimethyladenosine
NaHCO3 (122 mg, 1.4 mmol, 2.0 equiv.) was added all at once, and
then m-CPBA (70 % purity, 270 mg, 1.1 mmol, 1.5 equiv.) was
added portionwise over 5 min. The resulting mixture was stirred for
15 min at room temperature. It was then quenched with an aqueous
thiosulfate solution (5 mL), and AcOEt (30 mL) was added. The
layers were separated, and the organic layer was washed with an
aqueous thiosulfate solution (10 mL), an aqueous NaHCO3 solu-
tion (10 mL), and brine (10 mL), dried with MgSO4, and concen-
trated in vacuo. Purification by flash chromatography on silica gel
(petroleum ether/AcOEt, 85:15) yielded formate 10 as a colorless
3 H, 2-Me), 4.37 (m, 1 H, 5-CH2), 4.42 (d, J = 11.8 Hz, 2 H, 2-
CH2), 4.54 (d, J = 11.8 Hz, 2 H, 2-CH2), 4.62 (m, 1 H, 5-CH2),
5.86 (d, J = 6.0 Hz, 2 H, 3-H or 4-H), 6.32 (d, J = 6.0 Hz, 2 H, 3-
H or 4-H), 7.34–7.41 (m, 2 H, Ar-H), 7.42–7.48 (m, 2 H, Ar-H),
7.51–7.61 (m, 2 H, Ar-H), 7.94–7.98 (m, 2 H, Ar-H), 8.06–8.10 (m,
2 H, Ar-H) ppm. 13C NMR (75 MHz, CDCl3): δ = 166.1, 165.3,
165.2, 157.4, 133.6, 133.5, 133.4, 129.9, 129.6, 129.3, 129.0, 128.6,
128.6, 128.5, 86.5, 85.0, 72.5, 71.3, 68.3, 18.8 ppm. MS (ESI+): m/z
= 473 [M + H]+. HRMS (ESI+): calcd. for C28H25O7 [M + H]+
473.1600; found 473.1590.
oil (277 mg, 72%). IR (neat): ν = 2928, 2857, 1729, 1452, 1256,
˜
Glycosylation According to the Saneyoshi Procedure
1
1093, 1069, 835, 779, 705 cm–1. H NMR (300 MHz, CDCl3): δ =
2-(6-Amino-9H-purin-9-yl)-5-[(benzoyloxy)methyl]-2,5-dimethyl-
tetrahydrofuran-3,4-diyl Dibenzoate (12): To a solution of com-
pound 9 (161 mg, 0.31 mmol) in dry acetonitrile (3 mL) was added
adenine (84 mg, 0.62 mmol, 2.0 equiv.). The resulting solution was
cooled to –20 °C, and SnCl4 (150 μL, 1.25 mmol, 4.0 equiv.) was
added dropwise. The reaction mixture turned yellow and was
stirred at the same temperature for 1 h, then quenched with dry
pyridine (1.5 mL), and diluted with CH2Cl2 (40 mL). A white pre-
cipitate formed, and the mixture was filtered through a Celite pad.
The organic layer was washed with an aqueous NaHCO3 solution
(10 mL), water (10 mL), and brine (3ϫ10 mL), dried with MgSO4,
and concentrated in vacuo. Purification by flash chromatography
on silica gel (DCM/MeOH, 95:5) yielded compound 12 (1:1 mix-
ture of isomers) as a white solid (115 mg, 53%). UV (EtOH): λmax
(ε, m–1 cm–1) = 231 (26500), 261 (9800) nm. 1H NMR (300 MHz,
CDCl3): δ = 1.67 (s, 3 H, 2-Me or 5-Me), 1.68 (s, 3 H, 2-Me or 5-
Me), 2.03 (s, 3 H, 2-Me or 5-Me), 2.17 (s, 3 H, 2-Me or 5-Me),
2.36 (s, 2 H, NH2), 4.42 (d, J = 12.0 Hz, 1 H, 5-CH2), 4.43 (d, J =
11.9 Hz, 1 H, 5-CH2), 4.66 (d, J = 12.0 Hz, 1 H, 5-CH2), 4.71 (d,
J = 11.9 Hz, 1 H, 5-CH2), 5.85 (s, 1 H, 3Ј-H), 5.88 (s, 1 H, 3Ј-H),
6.04 (d, J = 5.5 Hz, 1 H, 3-H or 4-H), 6.15 (d, J = 5.4 Hz, 1 H, 3-
H or 4-H), 6.42 (d, J = 5.4 Hz, 1 H, 3-H or 4-H), 7.10 (d, J =
5.5 Hz, 1 H, 3-H or 4-H), 7.13–7.63 (m, 18 H, Ar-H), 7.69–7.78
(m, 4 H, Ar-H), 7.86–7.91 (m, 3 H, Ar-H), 8.07–8.18 (m, 7 H, Ar-
H), 8.18 (s, 1 H, 8Ј-H), 8.38 (s, 1 H, 8Ј-H) ppm. 13C NMR
(75 MHz, CDCl3): δ = 166.0, 165.9, 165.0, 164.9, 164.8, 164.4,
155.5, 155.4, 152.9, 152.5, 149.9, 149.3, 148.7, 138.7, 138.3, 136.0,
133.7, 133.7, 133.6, 133.6, 133.4, 133.3, 129.9, 129.8, 129.6, 129.4,
129.2, 128.9, 128.9, 128.8, 128.7, 128.6, 128.5, 128.4, 128.3, 120.6,
120.5, 96.5, 95.3, 85.6, 85.2, 76.0, 75.8, 73.3, 72.6, 68.5, 68.2, 27.4,
25.5, 20.3, 19.5 ppm. MS (ESI+): m/z = 608 [M + H]+. HRMS
(ESI+): calcd. for C33H30N5O7 [M + H]+ 608.2145; found 608.2157.
0.00 (s, 3 H, Si-Me), 0.01 (s, 3 H, Si-Me), 0.79 (s, 9 H, Si-CMe3),
1.37 (s, 3 H, 2-Me or 5-Me), 1.78 (s, 3 H, 2-Me or 5-Me), 3.57 (d,
J = 10.6 Hz, 1 H, 2-CH2), 3.64 (d, J = 10.6 Hz, 1 H, 2-CH2), 5.87
(s, 2 H, 3-H and 4-H), 7.17–7.51 (m, 6 H, Ar-H), 7.76–7.79 (m, 2
H, Ar-H), 7.88–7.90 (m, 2 H, Ar-H), 8.20 (s, 1 H, Hformate) ppm.
13C NMR (75 MHz, CDCl3): δ = 166.0, 165.6, 159.8, 133.6, 133.3,
129.8, 129.7, 129.2, 128.6, 128.4, 110.3, 87.5, 77.0, 72.3, 68.3, 25.9,
21.7, 19.9, 18.4, –5.4 ppm. MS (ESI+): m/z = 483 [M + H –
HCOOH]+.
Glycosylation According to the Vorbrüggen Procedure
2-[(Benzoyloxy)methyl]-5-[2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl]-
2,5-dimethyltetrahydrofuran-3,4-diyl Dibenzoate (11): To a solution
of compound 9 (140 mg, 0.27 mmol) in dry acetonitrile (3 mL) was
added uracil (60 mg, 0.54 mmol, 2.0 equiv.) and N,O-bis-trimethyl-
silyl acetamide (280 μL, 1.1 mmol, 4.2 equiv.). The resulting sus-
pension was stirred at 50 °C until clear (1 h). The resulting solution
was cooled to –20 °C, and TMSOTf (50 μL, 0.27 mmol, 1.0 equiv.)
was added dropwise. The temperature was maintained at –20 °C
for 1 h, and the reaction mixture was quenched with an aqueous
saturated NaHCO3 solution (5 mL), and diluted with CH2Cl2
(40 mL). The layers were separated, and the organic layer was
washed with water (10 mL) and brine (10 mL), dried with MgSO4,
and concentrated in vacuo. Purification by flash chromatography
on silica gel (petroleum ether/AcOEt, 50:50) yielded compound 11
(3:1 mixture of isomers) as a colorless oil (110 mg, 70 %). UV
(EtOH): λmax (ε, m–1 cm–1) = 262 (11300) nm. IR (neat): ν = 3269,
˜
3061, 2360, 2341, 1731, 1693, 1601, 1451, 1377, 1267, 1177, 1105,
1069, 1025 cm–1. 1H NMR (300 MHz, CDCl3): δ (major isomer
italicized when identified) = 1.60 (s, 2.25 H, 2-Me or 5-Me), 1.60
(s, 0.75 H, 2-Me or 5-Me), 1.88 (s, 2.25 H, 2-Me or 5-Me), 1.86 (s,
0.75 H, 2-Me or 5-Me), 4.27 (d, J = 12.2 Hz, 0.75 H, 2-CH2), 4.29
(d, J = 12.0 Hz, 0.25 H, 2-CH2), 4.71 (d, J = 12.2 Hz, 0.75 H, 2-
CH2), 4.73 (d, J = 12.2 Hz, 0.25 H, 2-CH2), 5.48 (d, J = 8.4 Hz,
0.75 H, 6Ј-H), 5.83 (d, J = 8.4 Hz, 0.25 H, 6Ј-H), 6.02 (d, J =
5.7 Hz, 0.75 H, 3-H or 4-H), 6.15 (d, J = 5.2 Hz, 0.25 H, 3-H or
4-H), 6.27 (d, J = 5.2 Hz, 0.25 H, 3-H or 4-H), 6.57 (d, J = 5.7 Hz,
0.75 H, 3-H or 4-H), 7.33–7.63 (m, 9 H, Ar-H), 7.77–8.08 (m, 6 H,
Ar-H and 5Ј-H), 8.18 (s, 0.75 H, Ar-H) ppm. 13C NMR (75 MHz,
CDCl3): δ (major isomer italicized when identified) = 165.8, 165.7,
164.8, 164.8, 164.3, 163.7, 163.4, 163.3, 149.5, 149.4, 139.7, 139.2,
133.8, 133.6, 133.6, 133.4, 133.3, 129.7, 129.6, 129.6, 129.4, 129.3,
129.1, 129.0, 128.9, 128.8, 128.8, 128.7, 128.6, 128.5, 128.4, 101.6,
100.9, 98.4, 95.8, 85.9, 84.7, 76.6, 75.6, 73.9, 70.7, 67.8, 67.1, 26.4,
24.4, 19.6, 19.4 ppm. MS (ESI+): m/z = 585 [M + H]+. HRMS
(ESI+): calcd. for C32H29N2O9 [M + H]+ 585.1873; found 585.1871.
1,4-Dimethyl-2,7-dioxabicyclo[2.2.1]heptane-5,6-diyl Dibenzoate
(15): This compound was obtained as a byproduct (white solid) of
this procedure (see Table 1 for details). M.p. 133–135 °C. IR (neat):
ν = 2925, 1731, 1651, 1453, 1288, 1126, 1010 cm–1 1H NMR
.
˜
(300 MHz, CDCl3): δ = 1.56 (s, 3 H, Me), 1.70 (s, 3 H, Me), 3.53
(d, J = 7.2 Hz, 1 H, 3-H), 3.79 (d, J = 7.2 Hz, 1 H, 3-H), 5.38 (d,
J = 6.0 Hz, 1 H, 5-H or 6-H), 5.52 (d, J = 6.0 Hz, 1 H, 5-H or 6-
H), 7.21–7.31 (m, 4 H, Ar-H), 7.44–7.52 (m, 2 H, Ar-H), 7.86–7.91
(m, 4 H, Ar-H) ppm. 13C NMR (75 MHz, CDCl3): δ = 165.5,
165.3, 133.3, 133.3, 129.8, 129.1, 129.0, 128.3, 128.3, 107.6, 85.0,
75.5, 71.6, 14.8, 12.6 ppm. MS (ESI+): m/z = 369 [M + H]+.
HRMS (ESI+): calcd. for C21H21O6 [M + H]+ 369.1338; found
369.1343.
Deprotection with NH3/MeOH
2-[(Benzoyloxy)methyl]-2-methyl-5-methylenetetrahydrofuran-3,4-
diyl Dibenzoate (14): This compound was obtained as a byproduct
(colorless oil) of this procedure (see Table 1 for details). IR (neat):
(؎)-1Ј,4Ј-Dimethyl-1Ј,4Ј-didehydroadenosine (17β): Tribenzoate 12
(83 mg, 0.14 mmol) was introduced into a screw cap vessel, and dry
MeOH saturated with NH3 (4 mL) was added. The vessel was
ν = 3075, 2960, 2925, 2857, 1723, 1601, 1451, 1378, 1259, 1176, closed, and set at room temperature for 16 h. The vessel was slowly
˜
1
1093, 1068, 1025 cm–1. H NMR (300 MHz, CDCl3): δ = 1.63 (s,
opened, and the medium was degassed with argon for 10 min. The
Eur. J. Org. Chem. 2011, 319–326
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
325

G. Chaubet, D. Bourgeois, C. Périgaud
FULL PAPER
[5] a) S. G. Van Ornum, R. M. Champeau, R. Pariza, Chem. Rev.
2006, 106, 2990–3001; b) C. E. Schiaffo, P. H. Dussault, J. Org.
Chem. 2008, 73, 4688–4690; c) S. A. Testero, M. I. Mangione,
A. A. Poeylaut-Palena, M. Gonzalez Sierra, R. A. Spanevello,
Tetrahedron 2007, 63, 11410–11420.
[6] a) M. Candy, G. Audran, H. Bienaymé, C. Bressy, J. M. Pons,
J. Org. Chem. 2010, 75, 1354–1359; b) C. Bonini, R. Racioppi,
L. Viggiani, G. Righi, L. Rossi, Tetrahedron: Asymmetry 1993,
4, 793–799.
resulting product was taken into DCM and concentrated in vacuo.
Purification by flash chromatography on silica gel (DCM/MeOH,
90:10) yielded compound 17β as a white solid (13 mg, 31%) and
isomer 17α still contaminated with a residual amount of 17β. M.p.
191–193 °C (decomp.). UV (EtOH): λmax (ε, m–1 cm–1) = 260
(10600) nm. 1H NMR (300 MHz, [D6]DMSO): δ = 1.19 (s, 3 H, 1Ј-
Me or 4Ј-Me), 1.73 (s, 3 H, 1Ј-Me or 4Ј-Me), 3.25–3.28 (m, 1 H,
5Ј-H), 3.42 (dd, J = 11.8, 5.3 Hz, 1 H, 5Ј-H), 3.96 (t, J = 5.8 Hz, 1
H, 2Ј-H or 3Ј-H), 4.77 (t, J = 5.3 Hz, 1 H, 2Ј-H or 3Ј-H), 4.93 (d, [7] At this stage we did not attempt to ameliorate this yield with
more selective methods, such as: P. G. McDougal, J. G. Rico,
Y. I. Oh, B. D. Condon, J. Org. Chem. 1986, 51, 3388–3390.
[8] a) D. B. Dess, J. C. Martin, J. Org. Chem. 1983, 48, 4156–4158;
b) M. Frigerio, M. Santagostino, S. Sputore, J. Org. Chem.
1999, 64, 4537–4538.
[9] a) H. Vorbrüggen, Acc. Chem. Res. 1995, 28, 509–520; b) H.
Vorbrüggen, C. Ruh-Polenz in Organic Reactions (Ed.: L. Pa-
quette), Wiley, 2004, vol. 55, pp. 1–630.
[10] The base-induced dehydration of the alcohol activated as a
leaving group (mesylate, trifluoroacetate) is well described; see,
for instance: a) A. Caravano, D. Mengin-Lecreulx, J. M. Bron-
dello, S. P. Vincent, P. Sinay, Chem. Eur. J. 2003, 9, 5888–5898;
W. B. Yang, Y. Y. Yang, Y. F. Gu, S. H. Wang, C. C. Chang,
C. H. Linthe, J. Org. Chem. 2002, 67, 3773–3782; acid-cata-
lyzed dehydration of the tertiary alcohol has few precedents;
see, for instance: b) V. N. Kourafalos, P. Marakos, N. Pouli,
L. B. Townsend, Synlett 2002, 1479–1482; dehydration after
heating under vacuum was also described; see, for instance: c)
J. Mulzer, B. List, J. W. Bats, J. Am. Chem. Soc. 1997, 119,
5512–5518.
[11] C. Taillefumier, Y. Chapleur, Chem. Rev. 2004, 104, 263–292.
[12] B. R. Baker in The CIBA Foundation Symposium on the Chem-
istry and Biology of the Purines (Eds.: G. E. W. Wolstenholme,
C. M. O’Connor), Wiley, Chichester, UK, 2008, pp. 120–133.
[13] a) Y. P. Cheng, H. T. Chen, C. C. Lin, Tetrahedron Lett. 2002,
43, 7721–7723; b) A. V. Demchenko, E. Rousson, G. J. Boons,
Tetrahedron Lett. 1999, 40, 6523–6526.
J = 6.1 Hz, 1 H, OH), 5.54 (t, J = 5.8 Hz, 1 H, OH), 5.62 (d, J =
5.3 Hz, 1 H, OH), 7.25 (s, 2 H, NH2), 8.10 (s, 1 H, 3-H), 8.33 (s, 1
H, 8-H) ppm. 13C NMR (100.6 MHz, [D6]DMSO): δ = 156.6,
152.1, 148.3, 139.6, 120.5, 96.9, 87.5, 75.9, 71.1, 67.0, 24.1,
19.9 ppm. MS (ESI+): m/z = 296 [M + H]+. HRMS (ESI+): calcd.
for C12H18N5O4 [M + H]+ 296.1359; found 296.1353.
Supporting Information (see footnote on the first page of this arti-
1
cle): H and 13C NMR spectra of all new compounds.
Acknowledgments
The authors are particularly grateful to Dr. M. Smietana for fruit-
ful discussions regarding the structure elucidation of compound 14.
[1] a) R. De Francesco, A. Carfi, Adv. Drug Delivery Rev. 2007,
59, 1242–1262; b) H. C. Castro, N. I. V. Loreiro, M. Pujol-Luz,
A. M. T. Souza, M. G. Aluquerque, D. O. Santos, L. M. Cab-
ral, I. C. Frugulhetti, C. R. Rodrigues, Curr. Med. Chem. 2006,
12, 313–424.
[2] G. Sautrey, D. Bourgeois, C. Périgaud, Org. Biomol. Chem.
2010, 8, 378–383.
[3] a) L. Cappellacci, G. Barboni, M. Palmieri, M. Pasqualini, M.
Grifantini, B. Costa, C. Martini, P. Franchetti, J. Med. Chem.
2002, 45, 1196–1202; b) P. Franchetti, L. Cappellacci, P. Vita,
R. Petrelli, A. Lavecchia, S. Kachler, K. N. Klotz, I. Marabese,
L. Luongo, S. Maione, M. Grifantini, J. Med. Chem. 2009, 52,
2393–2406; c) L. Cappellacci, P. Franchetti, P. Vita, R. Petrelli,
[14] X. L. Li, H. Ohtake, H. Takahashi, S. Ikegami, Tetrahedron
2001, 57, 4283–4295.
[15] M. Saneyoshi, E. Satoh, Chem. Pharm. Bull. 1979, 27, 2518–
2521.
A. Lavecchia, H. N. Jayaram, P. Saiko, G. Graser, T. Szekeres, [16] For examples of Saneyoshi coupling with a 5Ј-OTBS sugar, see
M. Grifantini, J. Med. Chem. 2008, 51, 4260–4269; d) P.
Franchetti, L. Cappellacci, M. Pasqualini, R. Petrelli, P. Vita,
H. N. Jayaram, Z. Horvath, T. Szekeres, M. Grifantini, J. Med.
Chem. 2005, 48, 4983–4989.
for instance: a) J. H. M. Gould, J. Mann, Chem. Commun.
1997, 243–244; b) G. Rassu, L. Pinna, P. Spanu, F. Ulgheri,
G. Casiraghi, Tetrahedron Lett. 1994, 35, 4019–4022; c) K. H.
Altmann, Tetrahedron Lett. 1993, 34, 7721–7724.
Received: August 27, 2010
[4] C. Taffin, G. Kreutler, D. Bourgeois, E. Clot, C. Périgaud, New
J. Chem. 2010, 34, 517–525.
Published Online: November 22, 2010
326
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 319–326