Anti versus syn opening of epoxides derived from 9-(3-deoxy-β-D- glycero-pent-3-enofuranosyl)adenine with Me3Al: Factors controlling the stereoselectivity

Article abstract of DOI:10.1021/jo052243l

Upon epoxidation with dimethyldioxirane, the 2′,5′-bis-O-silyl derivatives of 9-(3-deoxy-β-D-glyceropent-3-enofuranosyl)adenine gave the respective "3′,4′-up" epoxides exclusively. Reaction between these epoxides and Me₃Al was investigated in d

Full text of DOI:10.1021/jo052243l

Anti versus Syn Opening of Epoxides Derived from
9-(3-Deoxy-â-D-glycero-pent-3-enofuranosyl)adenine with Me3Al:
Factors Controlling the Stereoselectivity
Yutaka Kubota, Kazuhiro Haraguchi, Mayumi Kunikata, Mari Hayashi,
Masaomi Ohkawa, and Hiromichi Tanaka*
School of Pharmaceutical Sciences, Showa UniVersity, 1-5-8 Hatanodai, Shinagawa-ku,
Tokyo 142-8555, Japan
hirotnk@pharm.showa-u.ac.jp
ReceiVed October 27, 2005
Upon epoxidation with dimethyldioxirane, the 2′,5′-bis-O-silyl derivatives of 9-(3-deoxy-â-D-glycero-
pent-3-enofuranosyl)adenine gave the respective “3′,4′-up” epoxides exclusively. Reaction between these
epoxides and Me3Al was investigated in detail. It was found that the stereoselectivity of epoxide ring
opening (anti versus syn) varied significantly upon changing the amount of Me3Al, the solvent, the O-silyl
protecting group, and the reaction temperature. A possible reaction mechanism is proposed.
Introduction
can also be used for the same purpose. Furthermore, such
substrates upon reacting with organoaluminum reagents are
expected to provide informative experimental results to further
the understanding of the stereochemistry of epoxide ring
opening. In the present study, reaction of Me3Al with the 3′,4′-
epoxides derived from 9-(3-deoxy-â-D-glycero-pent-3-enofura-
nosyl)adenine (1) was investigated in detail.
Ring opening of epoxides with organoaluminum reagents is
a useful operation for forming C-C bonds in organic synthesis.
It is generally recognized that simple epoxides undergo anti
opening with organoaluminum reagents;1-3 however, there seem
to be no clear explanations available for the stereochemical
outcome, except for one report in the reaction of epoxy
alcohols.4
Results and Discussion
Our recent synthetic endeavor has been focused on the
preparation of 4′-carbon-substituted nucleosides based on ring
opening of their epoxy intermediates.5-7 A previous report has
demonstrated that reaction of organoaluminum reagents with a
nucleoside 4′,5′-epoxide constitutes an efficient entry to the
analogues bearing sp3-, sp2-, and sp-hybridized carbon substit-
uents.6 It was readily anticipated that nucleoside 3′,4′-epoxides
Epoxidation of 9-(3-Deoxy-â-D-glycero-pent-3-enofurano-
syl)adenine Derivatives. The 3′,4′-unsaturated nucleoside 1
employed as a starting material in the present study was prepared
from adenosine according to the procedure reported by Moffatt
and co-workers: reaction of adenosine with 2-acetoxyisobutyryl
bromide followed by DBN treatment of the resulting mixture
of 2′,3′-vicinal halo acetates.8 Compound 1 was silylated at the
2′- and 5′-hydroxyl groups to give 2-4, which were further
converted to the N6-pivaloyl derivatives 5-7 to avoid N-oxide
formation9 during epoxidation.
(1) Mole, T.; Jeffery, E. A. Organoaluminium Compounds; Elsevier
Publishing Company: Amsterdam; The Netherlands, 1972.
(2) Nagata, W.; Yoshioka, M.; Okamura, T. Tetrahedron Lett. 1966, 847.
(3) Fried, J.; C.-H. Lin; Ford, S. H. Tetrahedron Lett. 1969, 634 and
1379.
(4) Suzuki, T.; Saimoto, H.; Tomioka, H.; Oshima, K.; Nozaki, H.
Tetrahedron Lett. 1982, 23, 3597.
(5) Haraguchi, K.; Takeda, S.; Tanaka, H. Org. Lett. 2003, 5, 1399.
(6) Haraguchi, K.; Takeda, S.; Tanaka, H.; Nitanda, T.; Baba, M.;
Dutschman, G. E.; Cheng, Y.-C. Bioorg. Med. Chem. Lett. 2003, 13, 3775.
(7) Haraguchi, K.; Kubota, Y.; Tanaka, H. J. Org. Chem. 2004, 69, 1831.
10.1021/jo052243l CCC: $33.50 © 2006 American Chemical Society
Published on Web 01/07/2006
J. Org. Chem. 2006, 71, 1099-1103
1099

Kubota et al.
SCHEME 1
SCHEME 2
For the preparation of the desired 3′,4′-epoxide from 5-7,
one would anticipate that use of nucleophilic reagents and/or
solvents should be avoided, since the presence of an enol ether
structure renders their epoxide fairly susceptible to nucleophilic
ring opening. This appeared to be the case as shown in Scheme
1. When 5 was reacted with m-CPBA (1.5 equiv) in CH2Cl2
for 1.5 h at room temperature, there were formed several
products as evidenced by TLC (hexane/EtOAc ) 1/1), among
which one product was assumed to be the corresponding
epoxide. However, silica gel column chromatography of the
reaction mixture allowed only the isolation of 9-[4-(3-chlo-
robenzoyloxy)xylofuranosyl]adenine (8: 16%).10
In contrast to the reaction of m-CPBA, epoxidation of 5 with
an acetone solution of dimethyldioxirane (DMDO)11,12 carried
out in CH2Cl2 at -30 °C for 0.5 h gave the corresponding
epoxide 9 in quantitative yield, simply by evaporation of the
solvents. Although 9 was not stable enough for further purifica-
tion, the 1H NMR spectrum warranted its sufficient purity and
the NOE data13 clearly supported the depicted “3′,4′-up”
structure. In a similar manner, the epoxides 10 and 11 were
prepared from 6 and 7, respectively.
TABLE 1. Reaction of 9 with Me3Al by Varing the Amount of
Reagenta
equiv of
Me3Al
combined yield (%)
ratio of
entry
of 12 plus 13
12/13b
1
2
3
4
1.0
3.0
4.0
6.0
10
69
76
80
90
93
59
2/1
2/1
3/1
5/1
6/1
only 13
5
6c
6.0
a All reactions were carried out in CH2Cl2 at -30 °C for 4.5 h. b The
ratios were determined by 1H NMR spectroscopy. c The Me3Al used in this
entry was prepared by reacting a THF solution of MeMgCl with AlCl3.
with AlCl3. The observed exclusive formation in entry 6 of the
syn-opened product 9-[2,5-bis-O-(tert-butyldimethylsilyl)- 4-C-
methyl- R-L-arabinofuranosyl]-N6-pivaloyladenine (13)15 was,
therefore, assumed to be due to a complex formation between
THF and Me3Al during preparation of the reagent. In support
of this assumption is the fact that the reaction of 9 with the
commercially available Me3Al (6 equiv), when carried out in
THF, also gave 13 as the sole product, albeit in a lower yield
(33%). The use of Et2O as a reaction solvent led to essentially
the same result (the yield of 13, 25%). The highest yield of 13
(90%) was seen when the reaction was carried out in 1,4-dioxane
at room temperature.
There have been several precedents for the reaction of Me3-
Al with epoxides derived from cyclic enol ethers, 3,4-dihydro-
2H-pyran,16 and glycal17,18 derivatives. In these instances, syn
opening of the epoxide ring has been reported to be a dominant
or exclusive reaction pathway. To see if the presence of the
N6-pivaloyladenine moiety in 9 has any influence on the reaction
pathway, methyl 2,5-bis-O-(tert-butyldimethylsilyl)-3-deoxy-â-
D-glycero-pent-3-enofuranoside (14) was prepared from the
known methyl 3,5-di-O-benzoyl-â-D-xylofuranoside19 in 4
steps.20 Upon DMDO oxidation of 14 and successive reaction
with Me3Al under the conditions in entry 4 in Table 1, there
was formed solely the syn-opened product 1521 (90%) (Scheme
3).On the basis of the above experimental results, we assumed
that the reaction mechanism between 9 and Me3Al could be
illustrated as shown in Scheme 4 (N6-pivaloyladenine moiety
is omitted for simplicity). Highly oxygenophilic Me3Al would
prefer coordination to the 3′,4′-epoxy structure of 9 to give A,
Reaction of the 3′,4′-Epoxides with Me3Al. The above-
prepared “3′,4′-up” epoxide 9 was reacted in CH2Cl2 with a
commercially available hexane solution of Me3Al that is free
from ethereal solvents (Scheme 2). In Table 1 are summarized
the results obtained by varying the amount of Me3Al. There
can be seen an apparent trend that formation of the anti-opened
product 9-[2,5-bis-O-(tert-butyldimethylsilyl)-4-C-methyl-â-D-
xylofuranosyl]-N6-pivaloyladenine (12)14 becomes a favorable
event upon increasing the amount of Me3Al (entries 1-5).
However, a dramatic change occurred in entry 6, where Me3Al
employed was prepared by reacting a THF solution of MeMgCl
(8) Jain, T. C.; Jenkins, I. D.; Russell, A. F.; Verheyden J. P. H.; Moffatt,
J. G. J. Org. Chem. 1974, 39, 30.
(9) Oxidation of the base moiety of adenosine derivatives with dimeth-
yldioxirane has been reported: Saladino, R.; Crestini, C.; Bernini, R.;
Mincione, E.; Ciafrino, R. Tetrahedron Lett. 1995, 36, 2665.
(10) NOE data for 8: H-1′/H-3′ (0.7%); H-2′/H-5′b (1.6%); H-2′/3′-OH
(8.1%). Of the two protons at the 5′-position, the one that appears at a
higher field is designated as H-5′a, and the other as H-5′b, throughout the
text and the Experimental Section.
(11) For the preparation of dimethyldioxirane: Adam, W.; Bialas, J.;
Hadjiarapoglou, L. Chem. Ber. 1991, 124, 2377.
(12) For epoxidation of glycals with DMDO: Halcomb, R. L.; Dan-
ishefsky, S. J. J. Am. Chem. Soc. 1989, 111, 6661.
(15) The stereochemistry of 13 was confirmed based on the following
NOE data: H-5′a/H-1′ (2.0%); H-5′a/H-3′ (2.0%); H-5′b/H-1′ (4.9%); H-5′b/
H-3′ (6.0%); H-3′/H-1′ (1.0%).
(16) Bailey, J. M.; Craig, D.; Gallagher, P. T. Synlett 1999, 132.
(17) Rainier, J. D.; Allwein, S. P.; Cox, J. M. Org. Lett. 2000, 2, 231.
(18) Rainier, J. D.; Cox, J. M. Org. Lett. 2000, 2, 2707.
(19) Collins, P. M.; Hurford, J. R.; Overend, W. G. J. Chem. Soc., Perkin
Trans. 1 1975, 2163.
(13) NOE data for 9: H-3′/H-5′a (0.8%); H-3′/H-5′b (1.1%); H-3′/H-1′
(0.6%).
(14) The stereochemistry of 12 was confirmed based on the following
NOE data: H-5′b/H-8 (0.9%); H-3′/H-1′ (1.9%); H-2′/3′-OH (4.1%); H-1′/
H-3′ (1.7%); H-8/H-3′ (0.3%).
(20) For the preparation of 14 from methyl 3,5-di-O-benzoyl-â-D-
xylofuranoside, see the Supporting Information.
(21) NOE data for 15: 4-Me/1-OMe (0.8%); 4-Me/H-2 (0.5%); 4-Me/
3-OH (1.0%); H-3/H-1 (0.3%); H-5b/H-3 (3.9%); H-5b/H-1 (0.5%); 1-OMe/
H-2 (0.1%).
1100 J. Org. Chem., Vol. 71, No. 3, 2006

Factors Controlling the StereoselectiVity
TABLE 2. Reactions of 9-11 with Me3Ala
SCHEME 3 a
combined yield (%) ratio of â-D-isomer/
entry epoxide products
of the two isomers
R-L-isomerb
1
2
3
9
10
11
12 and 13
16 and 17
18 and 19
90
89
94
12/13 ) 5/1c
16/17 ) 10/1
18/19 ) 1/7
a All reactions were carried out in CH2Cl2 at -30 °C for 4.5 h by using
Me3Al (6.0 equiv). b The ratios were determined by 1H NMR spectroscopy.
c Data taken from Table 1
a R ) TBDMS
SCHEME 4 a
TABLE 3. Reaction of 9 with Me3Al by Varying the Temperaturea
temp,
°C
combined yield (%)
ratio of
entry
of 12 plus 13
12/13b
1
2
3
4
rt
0
-30
-80
92
98
90
96
0.7/1
1.4/1
5/1c
30/1
a All reactions were carried out in CH2Cl2 for 4.5 h by using Me3Al
(6.0 equiv). b The ratios were determined by 1H NMR spectroscopy. c Data
taken from Table 1.
of the methyl group from tetramethylaluminate had been
suppressed. The influence of the 2′-O-silyl group was assessed
by employing the 3′,4′-epoxides 9 (R ) TBDMS), 10 (R )
SiEt3), and 11 (R ) TBDPS) as substrates. The results are
summarized in Table 2. Compared with the ratio previously
observed in 9 (entry 1, 12/13 ) 5/1), a significant increase of
less hindered attack (anti opening) was seen by the use of 10
(entry 2, 16/17 ) 10/1) having the less bulky triethylsilyl group.
The fact that 11 containing the more bulky TBDPS group gave
reversed stereoselectivity (entry 3, 18/19 ) 1/7) exceeded our
expectations. These experimental results may constitute further
support to the proposed mechanism.
a R ) TBDMS
which subsequently undergoes epoxide ring opening to form
an oxonium ion that carries an alkoxyaluminate at the 3′-
position. Two extreme conformers can be depicted for the
oxionium ion through rotation of the 3′-O-Al bond: in one
conformer B, Al is located above the furanose ring, and in the
other conformer C, it is outside the ring avoiding either steric
or electronic repulsion with the adenine base. In the case where
the amount of the remaining Me3Al is limited or it is complexed
with an ethereal solvent, intramolecular attack of the methyl
ligand from B (syn opening) would inevitably take place to give
E, which is finally converted to 13. On the other hand, when
noncoordinated Me3Al is sufficiently available in CH2Cl2, there
is a good opportunity for C to transfer its methyl ligand to Me3-
Al, yielding tetramethylaluminate and D.22 Under such circum-
stances, the presence of the adenine base as well as the 3′-
alkoxyaluminum substituent in D would render the stereochem-
ical bias in favor of less hindered attack (anti opening) to lead
to the dominant formation of F, which gives 12 after workup.
If this mechanism is correct, the following two additional
assumptions can be made: (1) formation of tetramethylaluminate
(C f D) leading to anti-ring opening would be affected by the
concentration of Me3Al in the reaction medium, and (2) anti
opening (D f F) would be encouraged by employing a less
bulky 2′-O-protecting group. When the reaction in entry 4 in
Table 1 (12/13 ) 5/1) was carried out in a 50-fold diluted
medium (Me3Al, 3.56 mM in the reaction medium), the ratio
of 12/13 became 1.5/1, suggesting that intermolecular delivery
Finally, the effect of the reaction temperature was examined.
As shown in Table 3 from the reaction of 9, it appeared that
the lower the reaction temperature, the higher the selectivity
for anti opening. This suggests that, at a lower temperature, the
equilibrium between B and C (Scheme 4) had shifted toward
C. Since stereoselectivity of the reaction of Me3Al with the
epoxide derived from 14, lacking the nucleobase, was not
affected by reaction temperature, forming the syn-opened
product 15 exclusively at -80 °C (73%) as well as at room
temperature (78%), the presence of the N6-pivaloyladenine
moiety may be crucial for this equilibrium shift. By combining
two factors directing the stereoselectivity in combination, the
reaction temperature and steric bulk of the 2′-O-silyl group, the
highest selectivity of 50/1 for â-D-isomer/R-L-isomer was
attained in the reaction of 10 (R ) SiEt3) at -80 °C (combined
yield of 16 plus 17: 90%).
Conclusion
(22) Transfer of an ethyl group from the weaker acid M[Et3AlOR] (M
) K or Na) to the stronger acid Et3Al is known to proceed quantitatively
to give M[Et4Al]: Lehmkuhl, H. Angew. Chem., Int. Ed. 1964, 3, 107.
The “3′,4′-up” epoxides (9-11) of 9-(3-deoxy-â-D-glycero-
pent-3-enofuranosyl)adenine (1), having a cyclic enol ether
J. Org. Chem, Vol. 71, No. 3, 2006 1101

Kubota et al.
J ) 1.8 and 1.3 Hz), 6.47 (1H, d, J ) 1.8 Hz), 8.02 (1H, s), 8.48
(1H, br), 8.77 (1H, s); 13C NMR (CDCl3) δ -5.42, -5.39, -4.6,
18.0, 18.3, 25.7, 25.8, 27.4, 40.4, 58.5, 80.2, 92.1, 99.9, 122.8,
139.9, 149.5, 151.4, 152.4, 161.8, 175.7; FAB-MS (m/z) 562 (M+
+ H). Anal. Calcd for C27H47N5O4Si2: C, 57.72; H, 8.43; N, 12.46.
Found: C, 57.70; H, 8.61; N, 12.47.
structure, were efficiently prepared by oxidation with DMDO.
Stereoselectivity of epoxide ring opening with Me3Al, anti
versus syn, was investigated by changing the amount of Me3-
Al, the solvent, the O-silyl protecting group, and the reaction
temperature. Reaction of the 3′,5′-bis-O-TBDMS epoxide 9
carried out in CH2Cl2 in a range of 0 to -80 °C uniformly
resulted in the preferential formation of the anti-ring-opened
product 12, which is in sharp contrast to the reported stereo-
selectivity for epoxides derived from 3,4-dihydro-2H-pyran and
glycal derivatives.
The fact that 14 lacking the adenine base, upon reacting in
CH2Cl2 at -30 °C, gave solely the syn-opened product 15
suggests an important role of the nucleobase for the observed
anti-selectivity. By employing 9, it was also possible to change
stereochemical bias toward exclusive formation of the syn-
opened product 13, simply by carrying out the reaction in an
ethereal solvent, such as THF, Et2O, and 1,4-dioxane. These
results led us to propose a reaction mechanism depicted in
Scheme 4. Based on this mechanism, the highest anti-selectivity
(anti/syn ) 50/1) was accomplished by reacting the 2′-O-SiEt3
epoxide 10 in CH2Cl2 at -80 °C.
9-[2,5-Bis-O-(tert-butyldimethylsilyl)-4-(3-chlorobenzoyloxy)-
â-D-xylofuranosyl]-N6-pivaloyladenine (8). To a CH2Cl2 (5 mL)
solution of 5 (200 mg, 0.36 mmol) was added a CH2Cl2 (20 mL)
solution of m-CPBA (purity minimum 65%, 280 mg, 1.07 mmol).
After being stirred for 1.5 h at room temperature, the reaction
mixture was treated with Et3N (0.15 mL) and partitioned between
CH2Cl2 and H2O. Evaporation of the organic layer followed by
column chromatography (hexane/EtOAc ) 2/1) gave 8 (foam, 42.9
mg, 16%): UV (MeOH) λmax 272 nm (ꢀ 21 200), λmin 248 nm (ꢀ
1
12 800); H NMR (CDCl3) -0.14, -0.02, 0.05, and 0.07 (12H,
each as s), 0.72 and 0.86 (18H, each as s), 1.42 (9H, s), 4.34 and
4.38 (2H, each as d, J ) 11.1 Hz), 4.69 (1H, dd, J ) 3.1 and 7.3
Hz), 4.73 (1H, dd, J ) 3.1 and 3.8 Hz), 5.59 (1H, d, J ) 7.3 Hz),
6.25 (1H, d, J ) 3.8 Hz), 7.38-7.42, 7.55-7.58, 7.94-7.96, and
8.03-8.04 (4H, m), 8.23 (1H, s), 8.59 (1H, br), 8.77 (1H, s); 13C
NMR (CDCl3) δ -5.4, -5.1, -4.8, 17.8, 18.4, 25.5, 25.9, 27.5,
40.7, 62.0, 80.4, 81.9, 91.2, 112.6, 123.5, 128.1, 129.8, 129.9, 131.9,
133.5, 134.6, 142.2, 150.2, 150.7, 152.6, 163.9, 175.7; FAB-MS
(m/z) 734 (M+ + H). Anal. Calcd for C34H52ClN5O7Si2: C, 55.60;
H, 7.14; N, 9.54. Found: C, 55.55; H, 7.25; N, 9.46.
Since 9-(â-D-xylofuranosyl)adenine has long been known as
a biologically active nucleoside analogue23 and since nucleosides
having a carbon substituent at the 4′-position constitute a
promising class of antiviral agents,24,25 we believe that the
present study may be useful for the development of new
biologically active nucleoside analogues.
The Epoxide 9 Formed by DMDO Oxidation of 5. This
compound was not fully characterized due to its instability. Spectral
data of this compound are as follows: 1H NMR (CDCl3) δ 0.09,
0.11, and 0.13 (12H, each as s), 0.92 (18H, s), 1.41 (9H, s), 3.72
(1H, s), 4.17 and 4.27 (2H, each as d, J ) 12.3 Hz), 4.63 (1H, s),
6.54 (1H, s), 8.38 (1H, s), 8.56 (1H, br), 8.73 (1H, s); FAB-MS
(m/z) 578 (M+ + H).
Experimental Section
9-[2,5-Bis-O-(tert-butyldimethylsilyl)-3-deoxy-â-D-glycero-
pent-3-enofuranosyl]adenine (2). To a pyridine (20 mL) solution
of 1 (2.0 g, 8.06 mmol) was added TBDMSCl (3.6 g, 24.2 mmol)
at 0 °C. The mixture was stirred at 0 °C for 5 min and then at
room temperature for 24 h. The reaction mixture was partitioned
between CH2Cl2 and saturated aq NaHCO3. Evaporation of the
organic layer followed by column chromatography (CHCl3/MeOH
) 60/1) gave 2 (foam, 3.69 g, 96%): UV (MeOH) λmax 260 nm (ꢀ
Reaction of Me3Al with 9 Listed in Entry 4 of Table 1 as a
Typical Procedure: Formation of 9-[2,5-Bis-O-(tert-butyldim-
ethylsilyl)-4-C-methyl-â-D-xylofuranosyl]-N6-pivaloyladenine (12)
and 9-[2,5-Bis-O-(tert-butyldimethylsilyl)-4-C-methyl-R-L-ara-
binofuranosyl]-N6-pivaloyladenine (13). To a CH2Cl2 (12 mL)
solution of 5 (200 mg, 0.36 mmol) was added an acetone solution
of DMDO (ca. 0.04M, 13.2 mL, 0.53 mmol) under positive pressure
of dry Ar at -30 °C. After the mixture was stirred for 0.5 h at
-30 °C, the solution containing 9 was evaporated and dried under
diminished pressure. Compound 9 thus prepared was dissolved in
CH2Cl2 (12 mL) and cooled to -30 °C. To this was added Me3Al
(1.03 M hexane solution, 2.12 mL, 2.14 mmol). The reaction
mixture was stirred for 4.5 h at -30 °C, quenched by adding
saturated aq NH4Cl, and filtered through a Celite pad. The filtrate
was extracted with CH2Cl2. Column chromatography (hexane/
EtOAc ) 1/1) of the organic layer gave a mixture of 12 and 13
(foam, 190.1 mg, 90%, 12/13 ) 5/1). HPLC (hexane/EtOAc )
1/1) separation of the mixture gave analytical pure 12 (foam, tR
10.9 min) and 13 (foam, tR 13.0 min).
Physical date for 12: UV (MeOH) λmax 272 nm (ꢀ 18 100), λmin
235 nm (ꢀ 4 300); 1H NMR (CDCl3) δ -0.08, 0.04, 0.09, and 0.10
(12H, each as s), 0.85 and 0.92 (18H, each as s), 1.41 (9H, s), 1.43
(3H, s), 3.77 and 3.88 (2H, each as d, J ) 10.6 Hz), 4.04 (1H, d,
J ) 1.8 Hz), 4.60 (1H, dd, J ) 3.5 and 1.8 Hz), 4.93 (1H, br), 6.04
(1H, d, J ) 3.5 Hz), 8.29 (1H, s), 8.59 (1H, br), 8.76 (1H, s); 13C
NMR (CDCl3) δ -5.5, -5.3, -5.0, -4.8, 17.9, 18.3, 21.5, 25.6,
26.0, 27.5, 40.6, 67.3, 83.0, 84.3, 87.1, 90.2, 123.3, 142.2, 149.8,
151.1, 152.5, 175.7; FAB-MS (m/z) 594 (M+ + H). Anal. Calcd
for C28H51N5O5Si2: C, 56.63; H, 8.66; N, 11.79. Found: C, 56.53;
H, 8.84; N, 11.76.
1
15 200), λmin 231 nm (ꢀ 4 100); H NMR (CDCl3) δ 0.06, 0.09,
0.10, and 0.11 (12H, each as s), 0.89 and 0.92 (18H, each as s),
4.26 and 4.30 (2H, each as d, J ) 14.6 Hz), 5.18 (1H, d, J ) 1.3
Hz), 5.25 (1H, dd, J ) 1.8 and 1.3 Hz), 5.58 (2H, br), 6.41 (1H, d,
J ) 1.8 Hz), 7.85 (1H, s), 8.39 (1H, s); 13C NMR (CDCl3) δ -5.40,
-5.38, -4.54, -4.51, 18.1, 18.3, 25.7, 25.8, 58.6, 80.1, 91.9, 99.8,
119.5, 137.7, 149.7, 153.4, 155.4, 161.8; FAB-MS (m/z) 478 (M+
+ H). Anal. Calcd for C22H39N5O3Si2: C, 55.31; H, 8.23; N, 14.66.
Found: C, 55.27; H, 8.36; N, 14.62.
9-[2,5-Bis-O-(tert-butyldimethylsilyl)-3-deoxy-â-D-glycero-
pent-3-enofuransyl]-N6-pivaloyladenine (5). To a CH2Cl2 (20 mL)
solution of 2 (5.8 g, 12.1 mmol) was added i-Pr2NEt (3.2 mL, 18.2
mmol) and pivaloyl chloride (1.8 mL, 14.6 mmol) under positive
pressure of dry Ar at 0 °C. After being stirred for 2 h at 0 °C, the
reaction mixture was partitioned between CH2Cl2 and saturated aq
NaHCO3. Evaporation of the organic layer followed by column
chromatography (hexane/EtOAc ) 5/1) gave 5 (foam, 5.85 g,
86%): UV (MeOH) λmax 272 nm (ꢀ 17 800), λmin 236 nm (ꢀ 6 000);
1H NMR (CDCl3) δ 0.05, 0.08, 0.10, and 0.11 (12H, each as s),
0.89 and 0.92 (18H, each as s), 1.41 (9H, s), 4.26 and 4.31 (2H,
each as d, J ) 14.6 Hz), 5.17 (1H, d, J ) 1.3 Hz), 5.26 (1H, dd,
(23) For an example, see: Ellis, D. B.; LePage, G. A. Mol. Pharmacol.
1965, 1, 231.
(24) For an excellent review, see: Hayakawa, H.; Kohgo, S.; Kitano,
K.; Ashida, N.; Kodama, E.; Mitsuya, H.; Ohrui, H. AntiViral Chem.
Chemother. 2004, 15, 169.
(25) For a recent example, see: Tanaka, H.; Haraguchi, K.; Kumamoto,
H.; Baba, M.; Cheng, Y.-C. AntiViral Chem. Chemother. 2005, 16, 217.
Physical data for 13: UV (MeOH) λmax 272 nm (ꢀ 18 300),
λ
min 237 nm (ꢀ 6 300). 1H NMR (CDCl3) δ -0.04, 0.06, and 0.08
(12H, each as s), 0.87 and 0.92 (18H, each as s), 1.39 (3H, s), 1.41
(9H, s), 3.62 and 3.81 (2H, each as d, J ) 9.7 Hz), 4.15 (1H, d, J
) 1.4 Hz), 4.72 (1H, dd, J ) 2.4 and 1.4 Hz), 5.72 (1H, br), 5.80
(1H, d, J ) 2.4 Hz), 8.07 (1H, s), 8.55 (1H, br), 8.77 (1H, s); 13C
1102 J. Org. Chem., Vol. 71, No. 3, 2006

Factors Controlling the StereoselectiVity
NMR (CDCl3) δ -5.5, -5.4, -5.0, -4.8, 17.8, 18.2, 25.6, 25.8,
27.4, 40.5, 66.4, 79.2, 85.0, 89.7, 92.7, 124.0, 142.6, 150.1, 152.1,
175.6; FAB-MS (m/z) 594 (M+ + H). Anal. Calcd for C28H51N5O5-
Si2: C, 56.63; H, 8.66; N, 11.79. Found: C, 56.43; H, 8.83; N,
11.68.
Y. Odanaka (Center for Instrumental Analysis, Showa Univer-
sity) for technical assistance with NMR, MS, and elemental
analyses.
Supporting Information Available: General Experimental
Section, scheme, and procedures for the preparation of 14,
characterization data for compounds 3, 4, 6, 7, 10, 11, 14-16, 17-
19, and 13C NMR spectra of compounds 14, 15, and those involved
in the preparation of 14. This material is available free of charge
via the Internet at http://pubs.acs.org.
Acknowledgment. This study was supported by grants from
the Japan Health Science Foundation (SA14804 to H.T.) and
Grant-in-Aid (KAKENHI) from JSPS (Japan Society for the
Promotion of Science), No. 17590093 to H.K. and No. 17590094
to H.T. The authors are also grateful to Ms. K. Shiobara and
JO052243L
J. Org. Chem, Vol. 71, No. 3, 2006 1103