Sulfur Extrusion with Tin Radical: Synthesis of 4′,5′ -Didehydro-5′-deoxy-5′-(tributylstannyl)adenosine, an Intermediate for Potential Inhibitors against S-Adenosyl Homocysteine Hydrolase

Article abstract of DOI:10.1021/jo030256y

A new approach has been developed for the synthesis of potential inhibitors of S-adenosyl-L-homocysteine (AdoHcy) hydrolase. The key intermediate 9-[2,3-bis-O-(tert-butyldimethylsilyl)-5-(Z)-(tributylstannyl) -5-deoxy-β-D-erythro-pent-4-enofuranosyl] adenine (12) was prepared by sulfur extrusion reaction of 4′,5′- didehydro-5′ -deoxy-5′-(phenylthio)adenosine (11) with tributyltin radical. It was found that this reaction proceeds stereoselectively, forming 12 irrespective of the geometry of 11. Compound 12 readily underwent iodination, bromination, and chlorination with retention of configuration, whereas fluorination gave both (Z)- and (E)-isomers of vinyl fluoride. Because of the susceptibility of 12 to protodestannylation, the (Z)-vinyl iodide (13), prepared in quantitative yield from 12, was used as a substrate for C-C bond formation. Various types of carbon substituents (phenyl, vinyl, trifluorovinyl, ethynyl, and cyano) were introduced to the 5′-position of the 5-deoxy-β -D-erythro-pent-4-enofuranosyl structure to open up a new route to potential inhibitors of AdoHcy hydrolase.

Full text of DOI:10.1021/jo030256y

Su lfu r Extr u sion w ith Tin Ra d ica l: Syn th esis of
4′,5′-Did eh yd r o-5′-d eoxy-5′-(tr ibu tylsta n n yl)a d en osin e, a n
In ter m ed ia te for P oten tia l In h ibitor s a ga in st S-Ad en osyl
Hom ocystein e Hyd r ola se
Hiroki Kumamoto, Sayoko Onuma, and Hiromichi Tanaka*
School of Pharmaceutical Sciences, Showa University, 1-5-8 Hatanodai,
Shinagawa-ku, Tokyo 142-8555, J apan
hirotnk@pharm.showa-u.ac.jp
Received August 11, 2003
A new approach has been developed for the synthesis of potential inhibitors of S-adenosyl-L-
homocysteine (AdoHcy) hydrolase. The key intermediate 9-[2,3-bis-O-(tert-butyldimethylsilyl)-5-
(Z)-(tributylstannyl)-5-deoxy-â-D-erythro-pent-4-enofuranosyl]adenine (12) was prepared by sulfur
extrusion reaction of 4′,5′-didehydro-5′-deoxy-5′-(phenylthio)adenosine (11) with tributyltin radical.
It was found that this reaction proceeds stereoselectively, forming 12 irrespective of the geometry
of 11. Compound 12 readily underwent iodination, bromination, and chlorination with retention of
configuration, whereas fluorination gave both (Z)- and (E)-isomers of vinyl fluoride. Because of the
susceptibility of 12 to protodestannylation, the (Z)-vinyl iodide (13), prepared in quantitative yield
from 12, was used as a substrate for C-C bond formation. Various types of carbon substituents
(phenyl, vinyl, trifluorovinyl, ethynyl, and cyano) were introduced to the 5′-position of the 5-deoxy-
â-D-erythro-pent-4-enofuranosyl structure to open up a new route to potential inhibitors of AdoHcy
hydrolase.
In tr od u ction
S-Adenosyl-L-homocysteine (AdoHcy) hydrolase is one
of the target enzymes for antiviral chemotherapy,1 since
deactivation of this enzyme leads to accumulation of
AdoHcy (A) that inhibits methylation of viral mRNA as
a potent feedback inhibitor. Scheme 1 shows that hy-
drolysis with AdoHcy hydrolase begins with oxidation of
the 3′-hydroxyl group of A to the 3′-keto derivative B with
the enzyme-bound NAD+. Spontaneous elimination of
Hcy (D) from B yields the 4′,5′-didehydro intermediate
C, which in turn is hydrated to form the 3′-keto adenosine
E which finally is reduced to adenosine (F ) with the
enzyme-bound NADH. The finding2 that this enzyme
converts 4′,5′-didehydro-5′-deoxyadenosine (1) to A and
F , presumably through C, provides convincing support
for this reaction sequence.
Synthetic endeavors have been devoted to the search
for AdoHcy hydrolase inhibitors based on the above
mechanism.3-5 The designed inhibitors have the general
structure 2 with X being the leaving group orientated in
the Z-configuration so that, after oxidation of the 3′-
hydroxyl group, a nucleophilic site of the enzyme can
react through an addition-elimination pathway.
We recently reported that the 3′-benzenesulfonyl de-
rivative of 2′,3′-didehydro-2′,3′-dideoxyadenosine (d4A)
undergoes ipso-substitution with tributyltin radical to
give the stannylated product 3 and exemplified the
usefulness of this vinylstannane for the preparation of a
various types of 3′-substituted d4A analogues.6 Such tin
radical-mediated extrusion of sulfur atoms (sulfides and
sulfones) bound to sp2-carbon has been exploited on many
occasions in synthetic chemistry7,8 although the actual
mechanism of these reactions is not clear.
In the present study, we planned to synthesize a
vinylstannane 5 from 4 (Scheme 2) with the expectation
that 5 would serve as a common precursor for a various
types of potential AdoHcy hydrolase inhibitors.
Syn th esis of (Z)-4′,5′-Did eh yd r o-5′-d eoxy-5′-(tr ibu -
tylsta n n yl)a d en osin e. Since 5 is not a simple vinyl-
* Corresponding author.
(1) De Clercq, E. Biochem. Pharmacol. 1987, 36, 2567.
(2) Palmer, J . L.; Abeles, R. H. J . Biol. Chem. 1979, 254, 1217.
(3) J arvi, E. T.; McCarthy, J . R.; Mehdi, S.; Matthews, D. P.;
Edwards, M. L.; Prakash, N. J .; Bowlin, T. L.; Sunkara, P. S.; Bey, P.
J . Med. Chem. 1991, 34, 647.
(4) Wnuk, S. F.; Dalley, N. K.; Robins, M. J . J . Org. Chem. 1993,
58, 111.
(6) Onuma, S.; Kumamoto, H.; Kawato, M.; Tanaka, H. Tetrahedron
2002, 58, 2497.
(7) For the reaction of vinyl sulfides: (a) Schmidt, R. R.; Betz, R.
Angew. Chem., Int. Ed. Engl. 1984, 23, 430. (b) Tanaka, H.; Hayakawa,
H.; Obi, K.; Miyasaka, T. Tetrahedron Lett. 1985, 26, 6229. (c) Tanaka,
H.; Hayakawa, H.; Obi, K.; Miyasaka, T. Tetrahedron 1986, 42, 4187.
(d) Pallenberg, A. J .; White, J . D. Tetrahedron Lett. 1986, 27, 5591.
(e) Hollingworth, G. J .; Perkins, G.; Sweeney, J . J . Chem. Soc., Perkin
Trans. 1 1996, 1913.
(5) Robins, M. J .; Neschadimenko, V.; Ro, B.-O.; Yuan, C.-S.;
Borchardt, R. T.; Wnuk, S. F. J . Org. Chem. 1998, 63, 1205.
10.1021/jo030256y CCC: $27.50 © 2004 American Chemical Society
Published on Web 12/06/2003
72
J . Org. Chem. 2004, 69, 72-78

Sulfur Extrusion with Tin Radical
SCHEME 1
SCHEME 2
SCHEME 4
SCHEME 3. Rea ction s of 6 a n d 8 w ith Bu 3Sn H/
AlBNa
of the solvent, 1H NMR spectrum of the resulting mixture
showed complete conversion to the desired vinylstannane
7 as evidenced by J Sn,H: (E)-isomer, δ 4.63 (HR, J (119Sn-
1H) ) J (117Sn-1H) ) 36.0 Hz), 6.22 (Hâ, J (119Sn-1H) )
J (117Sn-1H) ) 33.6 Hz); (Z)-isomer, δ 4.50 (HR, J (119Sn-
1H) ) J (117Sn-1H) ) 47.6 Hz), 6.78 (Hâ, J (119Sn-1H) )
98.0 Hz, J (117Sn-1H) ) 95.6 Hz).10 The observed shift in
the E/Z ratio from 1/2.5 of 6 to 1.3/1 of 7 suggests that
partial inversion of the configuration occurred during this
sulfur extrusion reaction. The instability to TLC of the
stannylated product was also apparent for the reaction
of 2,3-dihydro-4-phenylthiofuran (8), but again formation
a
Key: i-Pr2NEt/benzene/80 °C/4 h.
stannane but it has an enol ether structure, we needed
to see if the sulfur extrusion with tin radical also works
with vinyl sulfides having a â-alkoxy substituent, and if
the resulting products are stable enough for further
manipulations. The initial study was, therefore, carried
out by using simple â-alkoxy-substituted vinyl sulfides
69 and 8 as model compounds (Scheme 3).
1
of 9 was confirmed by H NMR spectrum of the reaction
mixture: δ 6.08 (Hâ, J (119Sn-1H) ) J (117Sn-1H) ) 14.0
Hz).
A 0.1 M solution of 6 in benzene containing Bu3SnH
(2.0 equiv), AIBN (0.2 equiv), and i-Pr2NEt (3.0 equiv)
was heated at 80 °C for 4 h. Despite the disappearance
of 6, no distinct product by TLC analysis (hexane) of the
reaction mixture was observed. However, after removal
The preparation of a nucleosidic vinyl sulfide to be used
as a substrate for the tin radical reaction was next carried
out (Scheme 4). The starting material 1 was prepared
from adenosine according to the published method.11
Silylation of its 2′- and 3′-hydroxyl groups with TBDMSCl
turned out to be unsatisfactory, resulting in a low yield
of product with poor reproducibility. Use of TBDMSOTf
gave 10 in 85% yield. Electrophilic addition of PhSCl was
carried out by treating 11 with a mixture of PhSH (2.5
(8) For the reaction of vinyl sulfones: (a) Watanabe, Y.; Ueno, Y.;
Araki, T.; Endo, T.; Okawara, M. Tetrahedron Lett. 1986, 27, 215. (b)
Dubois, E.; Beau, J .-M. Tetrahedron Lett. 1990, 31, 5165. (c) McCarthy,
J . R.; Matthews, D. P.; Stemerick, D. M.; Huber, E. W.; Bey, P.; Lippert,
B. J .; Snyder, R. D.; Sunkara, P. S. J . Am. Chem. Soc. 1991, 113, 7439.
(d) Wnuk, S. F.; Robins, M. J . Can. J . Chem. 1993, 71, 192. (e)
McCarthy, J . R.; Huber, E. W.; Le, T.-B.; Laskovics, F. M.; Matthews,
D. P. Tetrahedron 1996, 52, 45. (f) Kumamoto, H.; Onuma, S.;
Tsuchiya, K.; Egusa, Y.; Tanaka, H.; Satoh, T. Nucleosides Nucleotides
Nucleic Acids 2002, 21, 275.
(10) For NMR data of vinylstannanes, see: Leusink, A. J .; Budding,
H. A.; Drenth, W. J . Organomet. Chem. 1967, 9, 295.
(11) McCarthy, J . R., J r.; Robins, R. K.; Robins, M. J . J . Am. Chem.
Soc. 1968, 90, 4993.
(9) Vlattas, I.; Della Vecchia, L.; Lee A. O. J . Am. Chem. Soc. 1976,
98, 2008.
J . Org. Chem, Vol. 69, No. 1, 2004 73

Kumamoto et al.
F IGURE 1. Possible stereochemical pathway for the selective formation of 12 from 11.
SCHEME 5
The observed sole formation of the stannylated (Z)-
isomer (12) from 11 would be explicable based on the
conformational preference of the intermediate C4′-radical
as depicted in Figure 1. In the case of â-face attack of
tributyltin radical, two conformers G and H are possible
upon departure of thiyl radical. Since G is considered to
be less favored because of the sterically constrained
accommodation of the 3′-O-TBDMS group, surrounded
by two bulky substituents SnBu3 and SPh, elimination
of the thiyl radical would take place from H. The same
applies to conformers I and J which result from R-face
attack of tributyltin radical. Overall, irrespective of the
face-selectivity of tin radical attack, only the (Z)-isomer
(12) is formed in this sulfur extrusion reaction. In support
of the above proposed mechanism is the fact that a
mixture of 11 and its (E)-isomer (E/ Z ) 1.0/1.4),14 when
subjected to the tin radical reaction, gave 12 as the sole
product in 86% yield.
TABLE 1. Rea ction of 11 w ith Tr ibu tyltin Ra d ica l
i-Pr2NEt
product
(isolated yield, %)
entry
(equiv)
purification
1
2
3
silica gel
Florisil
Florisil
12 (47), 10 (45)
12 (84), 10 (9)
12 (94)
3.0
equiv) and NCS (2.5 equiv) in CH2Cl2 at 0 °C. Subsequent
elimination proceeded simply by adding DBN (2.0 equiv.)
and then refluxing the reaction mixture. The product (Z)-
4′,5′-didehydro-5′-deoxy-5′-(phenylthio)adenosine (11)12
was isolated upon short column chromatography followed
by crystallization in 68% yield (contaminated with suc-
Syn th esis of P oten tia l In h ibitor s a ga in st S-Ad -
en osyl Hom ocyst ein e Hyd r ola se. Stannyl groups
bound to sp2-carbon atom can be manipulated in various
ways.15 Simple vinylstannanes are known to undergo
halogenation with retention of configuration.16 Halogena-
tion was first examined by using (Z)-4′,5′-didehydro-5′-
deoxy-5′-(tributylstannyl)adenosine (12). Iodination car-
ried out with iodine in THF gave the (Z)-vinyl iodide 1317
in quantitative yield within 1 h at room temperature.
Under similar conditions, NBS also works effectively to
give 14 in 97% yield. The depicted Z-configuration of
these compounds was confirmed based on NOE experi-
ment: 13, H-5′/H-3′ (7.5%); 14, H-5′/H-3′ (8.7%). Chlori-
nation of 12 with NCS, on the other hand, proceeded
rather sluggishly, and required heating at 60 °C for 12
h. The product 15 (76% yield) was desilylated with TBAF
(tetrabutylammonium fluoride) in THF to give the free
nucleoside 16, which gave an identical 1H NMR spectrum
1
cinimide, the yield was calculated by H NMR spectro-
scopy).
Reaction of 11 with tributyltin radical was investigated
(Scheme 5). The procedure employed is very simple,
simply refluxing a mixture of 11 (0.1 M in toluene), Bu3-
SnH (3.0 equiv), and AIBN (0.5 equiv) for 4 h either in
the presence or absence of i-Pr2NEt (3.0 equiv). The
stannylated product 12 was sufficiently stable13 to permit
its isolation by flash column chromatography (Table 1).
As shown in entry 1, even by flash chromatography, silica
gel caused a significant extent of protodestannylation
leading to 10. Use of Florisil as an absorbent greatly
improved the isolated yield of 12 (entry 2), but a small
amount of 10 was still isolated. The best result was
obtained by conducting the reaction in the presence of
i-Pr2NEt (entry 3). The vinylstannane structure of 12 was
readily elucidated from a characteristic splitting of H-5′
(δ 4.81) due to tin isotopes (J (119Sn-1H) ) J (117Sn-1H)
) 37.6 Hz). Its (Z)-stereochemistry was determined based
on NOE experiment: H-5′/H-3′ (6.1%).
(14) This mixture was obtained by HPLC purification (hexane/
EtOAc ) 2/3, tR 17.4-18.2 min) of the filtrate resulting from crystal-
lization of 11.
(15) Pereyre, M.; Quintard, J .-P.; Rahm, A. Tin in Organic Synthesis;
Butterworth: London, 1987.
(16) (a) Collins, P. W.; J ung, C. J .; Gasiecki, A.; Pappo, R. Tetrahe-
dron Lett. 1978, 3187. (b) Chen, S.-M. L.; Schaub, R. E.; Grudzinskas,
C. V. J . Org. Chem. 1978, 43, 3450.
(17) Electrophilic addtion-elimination of AgF/iodine to N6,N6-diben-
zoyl-4′,5′-didehydro-5′-deoxy-2′,3′-O-isopropylideneadenosine has been
reported to yield the corresponding (E)-5′-iodo derivative: J enkins, I.
D.; Verheyden, J . P. H.; Moffatt, J . G. J . Am. Chem. Soc. 1976, 98,
3346.
(12) The Z-configuration of 11 was determined on the basis of NOE
experiments: H-5′/3′-O-SiMe (4.1%), H-5′/H-3′ (5.8%).
(13) For complete conversion of the isolated 12 to 10, it took 57 h in
3 M AcOH in THF at room temperature.
74 J . Org. Chem., Vol. 69, No. 1, 2004

Sulfur Extrusion with Tin Radical
(measured in DMSO-d6) to the reported (Z)-isomer.4
TABLE 2. Ca r bon -Ca r bon Bon d -F or m in g Rea ction s
Usin g 13a
reaction
time
product
(isolated yield, %)
entry
reagent (equiv)
T (°C)
1
2
3
4
Bu3SnPh (3.0)
60
60
60
70
15 h
24 h
40 min
45 min
18 (48)
19 (76)
20 (71)
21 (67)
Bu3SnCHdCH2 (4.0)
Bu3SnCFdCF2 (5.0)
Me3SiCtCH (3.0)b
a
All reactions were carried out in DMF in the presence of CuI
b
(0.2 equiv) and (PPh3)2PdCl2 (0.2 equiv). Triethylamine (3.0
equiv) was added to the reaction mixture.
TABLE 3. Tr a n sfor m a tion of 13 to th e Vin yln itr ile 22a
Fluorination with XeF2 according to the reported
method18 gave an intractable mixture of products. Fluo-
rination was more successful using 1-chloromethyl-4-
fluoro-1,4-diazoniabicyclo[2.2.2]octane bis(tetrafluorobo-
rate) (Selectfluor)19 as an electrophilic fluorinating agent
(in CH3CN at room temperature for 15 min). However,
there was obtained an inseparable mixture of the vinyl
fluoride 17 [two geometrical isomers, 19F NMR (CDCl3)
δ (relative to CFCl3): -66.7 (d, J ) 78.8 Hz) and -61.8
(d, J ) 75.2 Hz), combined yield of 43% and diastereo-
meric ratio of ca. 3:1 calculated by 1H NMR] and the
destannylated byproduct 10 (53% by 1H NMR). From the
reported chemical shift of H-5′,4,20 we assume that the
(E)-isomer is dominant over the (Z)-isomer in this
mixture.
entry
(PPh3)2PdCl2 (equiv)
CuI (equiv)
yield (%) of 22
1
2
3
4
5
0.2
0.2
0.5
0.2
-
2
9
69
80
64
0.2
0.5
1.0
1.0
a
All reactions were carried out in DMF by using Bu3SnCN (4.0
equiv) at 120 °C for 24 h.
Introduction of carbon substituents was next investi-
gated. Although ample precedents of the Stille reaction21
are available, vinylstannanes having an alkoxy group at
the â-position have scarcely been used as substrates.22
When the Stille coupling between 12 and PhI/(Pd(PPh3)4/
DMF) was carried out at room temperature, no reaction
occurred irrespective of the presence or absence of CuI.
At an elevated temperature of 80 °C, in the presence of
CuI, protodestannylation to yield 10 was the sole observ-
able event. These results led us to use the vinyl iodide
13 for C-C bond-forming reactions.
Although 13 remained unchanged upon reacting with
SnPh4 in the presence of Pd(PPh3)4 and CuI (in DMF, 80
°C, for 24 h), replacement of the catalyst with (PPh3)2-
PdCl2 gave the desired product 18 in 12% yield.
Due to the poor solubility of SnPh4, the reaction was
reexamined by employing Bu3SnPh. This gave a higher
yield of 18 as shown in entry 1 of Table 2. In Table 2 are
also included the results of other C-C bond-forming
reactions. The reaction with Bu3SnCHdCH2 under simi-
lar conditions gave the conjugated diene 19 in good
yield (entry 2). Use of Bu3SnCFdCF2 considerably
shortened the reaction time to give 20 (entry 3), as can
be expected from the highly electron-withdrawing nature
of the trifluorovinyl group.24 In entry 4 is exemplified a
coupling reaction with a terminal alkyne by the formation
of 21.25
23
Finally, introduction of a cyano group was carried out.
Nair et al. reported the first example of a Pd-catalyzed
coupling between Bu3SnCN and an aryl iodide, 2-iodoad-
enosine.26 Compound 13 remained unchanged (recovery
90%) when reacted with Bu3SnCN (3.0 equiv.) under
Nair’s conditions, Pd(PPh3)4/DMF/120 °C/20 h. By chang-
ing the catalyst to (PPh3)2PdCl2, however, a small amount
of the vinylnitrile 22 was formed as shown in entry 1 in
Table 3. It was observed that addition of CuI to the
reaction medium increased the yield of 22 (entries 2-4),
although not exactly proportional. Quite unexpectedly,
(18) Tius, M. A.; Kawakami, J . K. Synthetic Commun. 1992, 22,
1461.
(19) Matthews, D. P.; Miller, S. C.; J arvi, E. T.; Sabol, J . S.;
McCarthy, J . R. Tetrahedron Lett. 1993, 34, 3057.
(23) For the preparation of this reagent, see: Burdon, J .; Coe, P.
L.; Haslock, I. B.; Powell, R. L. Chem. Commun. 1996, 49.
(24) Electron-withdrawing groups on the phenyl ring of Bu3SnCH2-
Ph accelerate the coupling reaction with PhCOCl; see: Stille, J . K.
Angew. Chem., Int. Engl. 1986, 25, 508.
(25) (a) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett.
1975, 4467. (b) Heck, R. F. Acc. Chem. Res. 1979, 12, 146.
(26) Nair, V.; Buenger, G. S. J . Am. Chem. Soc. 1989, 111, 8502.
(20) The 1H NMR spectrum of this mixture measured in CDCl3
showed two H-5′ resonances corresponding to 17: 6.85 (d, J ) 78.8
Hz) for major isomer; 6.39 (d, J ) 75.2 Hz) for minor isomer.
(21) For a review, see: Mitchell, T. N. Synthesis 1992, 803.
(22) To the best of our knowledge, there has been only one report
available for the Stille reaction of â-alkoxyvinylstannanes, see: Piers,
E.; Lu, Y.-F. J . Org. Chem. 1988, 53, 926.
J . Org. Chem, Vol. 69, No. 1, 2004 75

Kumamoto et al.
CH). For 7Z: 1H NMR (CDCl3) δ 0.89-0.95 (9H, m, SnBu),
1.15-1.24 (9H, m, SnBu and CH2CH3), 1.28-1.35 (6H, m,
SnBu), 1.49-1.66 (6H, m, SnBu), 3.76 (2H, q, J ) 6.8 Hz, CH2-
CH3), 4.50 (1H, J ) 7.2 Hz, J Sn,H ) 47.6 Hz, SnCH)CH), 6.78
(1H, d, J ) 7.2 Hz, J Sn,H ) 98.0, 95.6 Hz, SnCHdCH).
a significant yield was seen even in the absence of the
catalyst (entry 5).27
One might speculate that CuCN formed in this reaction
could be the actual reactive species for this transforma-
tion, since there is a precedent that CuCN effects such
reaction of a vinyl iodide.28 This, however, still remains
questionable, because no reaction was observed when 13
was treated with CuCN (3.0 equiv) in DMF at 120 °C for
24 h.
2,3-Dih yd r o-4-p h en ylth iofu r a n (8). To a THF (15 mL)
solution of 2,3-dihydrofuran (500 µL, 6.6 mmol) was added
dropwise freshly prepared PhSCl29 (950 µL) at -70 °C under
positive pressure of dry Ar. After being stirred for 15 min, the
mixture was further treated with KOBu-t (1.11 g, 9.9 mmol),
and allowed to warm to room temperature over 1 h. The
resulting mixture was partitioned between hexane and satur-
ated aqueous NaHCO3. The organic layer was dried (Na2SO4),
evaporated, and purified by column chromatography (hexane/
EtOAc ) 50/1). This gave 8 (830 mg, 71%) as an oil: 1H NMR
(CDCl3) δ 2.70 (2H, dt, J ) 9.6, 2.0 Hz, H-3), 4.53 (2H, t, J )
9.6 Hz, H-2), 6.70 (1H, t, J ) 2.0 Hz, H-5), 7.14-7.18 (1H, m,
Ph), 7.27-7.29 (4H, m, Ph); FAB-MS m/z 178 (M+ + H).
2,3-Dih yd r o-4-(t r ib u t ylst a n n yl)fu r a n (9). This com-
pound was prepared from 8 by the procedure used for the
preparation of 7 from 6: 1H NMR (CDCl3) δ 0.87-0.97 (9H,
m, SnBu), 1.05-1.16 (6H, m, SnBu), 1.26-1.36 (6H, m, SnBu),
1.47-1.62 (6H, m, SnBu), 2.59-2.68 (2H, m, H-3), 4.20 (2H,
t, J ) 9.2 Hz, H-2), 6.08 (1H, t, J ) 2.0 Hz, J Sn,H ) 14.0 Hz,
H-5).
9-[2,3-Bis-O-(ter t-bu tyldim eth ylsilyl)-5-deoxy-â-D-er yth -
r o-p en t-4-en ofu r a n osyl]a d en in e (10). To a mixture of 1
(4.84 g, 19.2 mmol) and imidazole (6.61 g, 97.1 mmol) in DMF
(150 mL) was added TBDMSOTf (16.0 mL, 68 mmol) at 0 °C.
The reaction mixture was stirred at rt for 48 h and then was
partitioned between EtOAc and saturated aqueous NaHCO3.
Column chromatography (hexane/EtOAc ) 1/1) of the organic
layer gave 10 (7.9 g, 85%) as a foam. For physical data of 10,
see the supporting information of ref 30.
Con clu sion
To develop a new approach to potential inhibitors to
S-adenosyl-L-homocysteine (AdoHcy) hydrolase, sulfur
extrusion reaction of 4′,5′-didehydro-5′-deoxy-5′-(phen-
ylthio)adenosine was examined with tributyltin radical.
As a result, exclusive formation of 9-[2,3-bis-O-(tert-
butyldimethylsilyl)-5-(Z)-(tributylstannyl)-5-deoxy-â-D-
erythro-pent-4-enofuranosyl]adenine (12) was observed,
irrespective of the geometry of the substrate. Although
12 was stable enough to be isolated by flash column
chromatography, its further manipulation was possible
only for halogenation. Carbon-carbon bond-forming re-
actions at the 5′-position of the 5-deoxy-â-D-erythro-pent-
4-enofuranosyl structure were, therefore, carried out by
using the vinyl iodide (13) which was readily prepared
from 12 in quantitative yield. Various types of carbon-
substituents (phenyl, vinyl, trifluorovinyl, and ethynyl)
were introduced based on palladium-catalyzed cross-
coupling reaction. It was found that, in the case of
introducing a cyano group, the presence of the Pd-catalyst
was not necessarily required, although the actual reactive
species involved in this reaction still remains to be
identified.
9-[2,3-Bis-O-(tert-butyldimethylsilyl)-5-deoxy-5-(Z)-(phen-
ylth io)-â-D-er yth r o-p en t-4-en ofu r a n osyl]a d en in e (11). To
a CH2Cl2 (15 mL) solution of NCS (349 mg, 2.61 mmol) was
added PhSH (270 µL, 2.31 mmol) at 0 °C. After the mixture
was stirred for 0.5 h, 10 (500 mg, 1.05 mmol) was added and
stirring was continued further 0.5 h at 0 °C. After addition of
DBN (260 µL, 2.1 mmol), the mixture was refluxed for 15 h.
The reaction mixture was partitioned between CH2Cl2 and
saturated aqueous NaHCO3. Column chromatography (hexane/
EtOAc ) 1/1) of the organic layer gave a crude mixture of
products, which was crystallized from ether-hexane to give
11 (441 mg). This material was contaminated with succinimide
Exp er im en ta l Section
Melting points are uncorrected. 1H and 13C NMR spectra
were recorded at 500 MHz. Chemical shifts are reported
relative to Me4Si. 19F NMR spectra were measured at 400 MHz
with CFCl3 as an internal standard. Mass spectra (MS) were
taken in FAB mode with m-nitrobenzyl alcohol as a matrix.
Column chromatography was carried out on silica gel (silica
gel 60, Merck) unless otherwise noted. Thin-layer chromatog-
raphy (TLC) was performed on silica gel (precoated silica gel
plate F254, Merck). HPLC was carried out on a Shimadzu LC-
6AD with a Shim-pack PREP-SIL (H)‚ KIT column (2 × 25
cm).
Rea ction of 6 w ith Tr ibu tyltin Ra d ica l: F or m a tion of
Tr ibu tyl(2-eth oxyvin yl)tin (7). A mixture of 6 (297 mg, 1.65
mmol), Bu3SnH (890 µL, 3.3 mmol), AIBN (54 mg, 0.33 mmol),
and i-Pr2NEt (862 µL, 4.95 mmol) in benzene (16.5 mL) was
heated at 80 °C for 4 h under positive pressure of dry Ar.
Evaporation of the reaction mixture was followed by drying
under reduced pressure. The resulting oily residue was
analyzed by 1H NMR spectroscopy to confirm quantitative
conversion to 7. For 7E: 1H NMR (CDCl3) δ 0.89-0.95 (9H,
m, SnBu), 1.15-1.24 (9H, m, SnBu and CH2CH3), 1.28-1.35
(6H, m, SnBu), 1.49-1.66 (6H, m, SnBu), 3.79 (2H, q, J ) 6.8
Hz, CH2CH3), 4.63 (1H, d, J ) 15.6 Hz, J Sn,H ) 36.0 Hz,
SnCHdCH), 6.22 (1H, d, J ) 15.6 Hz, J Sn,H ) 33.6 Hz, SnCHd
1
(ca. 5%) and the actual yield of 11 was 68% based on H NMR
spectroscopy. An analytical sample of 11 was obtained by
HPLC (hexane/EtOAc ) 1/2, tR ) 17.6 min) purification: mp
210-213 °C; UV (MeOH) λmax 258 nm (ꢀ 26 700), λmin 232 nm
(ꢀ 12 600); 1H NMR (CDCl3) δ -0.29, -0.04, 0.16 and 0.17
(12H, each as s, SiMe), 0.78 and 0.96 (18H, each as s, Bu-t),
4.79 (1H, d, J ) 4.0 Hz, H-3’), 5.20 (1H, dd, J ) 6.0, 4.0 Hz,
H-2’), 5.44 (1H, s, H-5’), 5.63 (2H, br, NH2), 6.19 (1H, d, J )
6.0 Hz, H-1’), 7.14-7.18 (1H, m, Ph), 7.24-7.28 (2H, m, Ph),
7.34-7.36 (2H, m, Ph), 7.87 and 8.35 (2H, each as s, H-2 and
H-8); FAB-MS m/z 586 (M+ + H). Anal. Calcd for C28H43N5O3-
Si2: C, 57.40; H, 7.40; N, 11.95. Found: C, 57.38; H, 7.48; N,
12.11.
9-[2,3-Bis-O-(ter t-b u t yld im et h ylsilyl)-5-(Z)-(t r ib u t yl-
st a n n yl)-5-d eoxy-â-D-er yth r o-p en t -4-en ofu r a n osyl]a d e-
n in e (12). A mixture of 11 (1.0 g, 1.7 mmol), Bu3SnH (1.38
mL, 5.12 mmol), AIBN (140 mg, 0.85 mmol), and i-Pr2NEt (890
µL, 5.12 mmol) in toluene (17 mL) was refluxed for 4 h under
positive pressure of dry Ar. The reaction mixture was evapor-
ated, and the residue was purified by flash Florisil column
chromatography (hexane/EtOAc ) 4/1). This gave 12 (1.23 g,
94%) as an oil: UV (MeOH) λmax 259 nm (ꢀ 17 200) λmin 240
(27) The NOE correlations (measured in CDCl3) between H-5′ and
H-3′ of 18-20 and 22 are as follows: 18 8.6%; 19 6.6%; 20 7.2%; 22
5.9%. The Z-stereochemistry of 21 was confirmed after converting to
the corresponding free nucleoside (H-5′/H-3′, 1.8% in DMSO-d6).
(28) Kitano, Y.; Matsumoto, T.; Wakasa, T.; Okamoto, S.; Shimazaki,
T.; Kobayashi, Y.; Sato, F.; Miyaji, K.; Arai, K. Tetrahedron Lett. 1987,
28, 6351.
(29) Paquette L. A.; Brand, S.; Behrens, C. J . Org. Chem. 1999, 64,
2010.
(30) Haraguchi, K.; Takeda, S.; Tanaka, H. Org. Lett. 2003, 5, 1399.
76 J . Org. Chem., Vol. 69, No. 1, 2004

Sulfur Extrusion with Tin Radical
nm (ꢀ 11 000); 1H NMR (CDCl3) δ -0.31, -0.07, 0.13 and 0.15
(12H, each as s, SiMe), 0.75 and 0.93 (18H, each as s, Bu-t),
0.78-0.82 (15H, m, SnBu), 1.14-1.26 (6H, m, SnBu), 1.34-
1.44 (6H, m, SnBu), 4.49 (1H, d, J ) 4.4 Hz, H-3’), 4.81 (1H,
J (119Sn-1H) ) J (117Sn-1H) ) 37.6 Hz, H-5’), 5.03 (1H, dd, J
) 6.0, 4.4 Hz, H-2’), 5.64 (2H, br, NH2), 6.09 (1H, d, J ) 6.0
Hz, H-1’), 7.86 and 8.35 (2H, each as s, H-2 and H-8); FAB-
MS m/z 769 (M+ + H). Anal. Calcd for C34H66N5O3Si2Sn: C,
53.19; H, 8.66; N, 9.12. Found: C, 53.27; H, 8.36; N, 9.09.
9-[2,3-Bis-O-(ter t-bu tyldim eth ylsilyl)-5-deoxy-5-(Z)-iodo-
â-D-er yth r o-p en t-4-en ofu r a n osyl]a d en in e (13). A mixture
of 12 (722 mg, 0.94 mmol) and iodine (359 mg, 1.41 mmol as
I2) in THF (10 mL) was stirred for 1 h at rt. The reaction
mixture was partitioned between CHCl3 and saturated aque-
ous Na2S2O3. Column chromatography (hexane/EtOAc ) 1/1)
of the organic layer gave 13 (565 mg, 100%) as a solid: mp
251-254 °C; UV (MeOH) λmax 259 nm (ꢀ 17 500), λmin 238 nm
(ꢀ 9500); 1H NMR (CDCl3) δ -0.31, -0.07, 0.14 and 0.15 (12H,
each as s, SiMe), 0.75 and 0.94 (18H, each as s, Bu-t), 4.74
(1H, d, J ) 4.0 Hz, H-3’), 5.15 (1H, dd, J ) 6.4, 4.0 Hz, H-2’),
5.23 (1H, s, H-5’), 5.61 (2H, br, NH2), 6.21 (1H, d, J ) 6.4 Hz,
H-1’), 7.90 and 8.36 (2H, each as s, H-2 and H-8); FAB-MS
m/z 604 (M+ + H). Anal. Calcd for C22H38IN5O3Si2: C, 43.77;
H, 6.35; N, 11.60. Found: C, 44.05; H, 6.34; N, 11.59.
5.23 (1H, m, H-2′), 5.15 (1H, d, J ) 7.2 Hz, H-1′), 6.39 (1H, d,
J ) 75.2 Hz, H-5′), 7.89 and 8.36 (2H, each as s, H-2 and H-8);
FAB-MS for 17 m/z 497 (M+ + H).
9-[2,3-Bis-O-(ter t-b u t yld im e t h ylsilyl)-5-d e oxy-5-(Z)-
p h en yl-â-D-er yth r o-p en t-4-en ofu r a n osyl]a d en in e (18). A
mixture of 13 (200 mg, 0.33 mmol), PhSnBu3 (323 µL, 0.99
mmol), (PPh3)2PdCl2 (49 mg, 0.07 mmol), and CuI (13 mg, 0.07
mmol) in DMF (1.5 mL) was heated at 60 °C for 15 h under
positive pressure of dry Ar. The reaction mixture was parti-
tioned between EtOAc and saturated aqueous NaHCO3.
Column chromatography (hexane /EtOAc ) 1/1) of the organic
layer gave 18 (88 mg, 48%) as a solid: mp 222-224 °C; UV
(MeOH) λmax 263 nm (ꢀ 38 700), λmin 229 nm (ꢀ 9900); 1H NMR
(CDCl3) δ -0.30, -0.06, 0.17 and 0.20 (12H, each as s, SiMe),
0.76 and 0.96 (18H, each as s, Bu-t), 4.68 (1H, d, J ) 4.4 Hz,
H-3’), 5.10 (1H, dd, J ) 6.8, 4.4 Hz, H-2’), 5.55 (1H, s, H-5’),
5.62 (2H, br, NH2), 6.34 (1H, d, J ) 6.8 Hz, H-1’), 7.11-7.15
(1H, m, Ph), 7.22-7.26 (2H, m, Ph), 7.45-7.47 (2H, m, Ph),
7.95 and 8.36 (2H, each as s, H-2 and H-8); FAB-MS m/z 554
(M+ + H). Anal. Calcd for C28H43N5O3Si2: C, 60.72; H, 7.83;
N, 12.65. Found: C, 60.52; H, 8.06; N, 12.54.
9-[2,3-Bis-O-(ter t-bu tyldim eth ylsilyl)-5-deoxy-5-(Z)-vin yl-
â-D-er yth r o-p en t-4-en ofu r a n osyl]a d en in e (19). A mixture
of 13 (400 mg, 0.66 mmol), tributylvinyltin (775 µL, 2.65
mmol), (PPh3)2PdCl2 (98 mg, 0.13 mmol), and CuI (26 mg, 0.13
mmol) in DMF (3 mL) was heated at 60 °C for 24 h under
positive pressure of dry Ar. The reaction mixture was parti-
tioned between EtOAc and saturated aqueous NaHCO3.
Column chromatography (hexane/EtOAc ) 1/1) of the organic
layer gave 19 (252 mg, 76%) as a solid: mp 211-215 °C; UV
(MeOH) λmax 248 nm (ꢀ 35 500), λmin 222 nm (ꢀ 11 500), 260
nm (shoulder); 1H NMR (CDCl3) δ -0.31, -0.08, 0.13 and 0.15
(12H, each as s, SiMe), 0.75 and 0.92 (18H, each as s, Bu-t),
4.56 (1H, d, J ) 4.0 Hz, H-3’), 4.96 (1H, dd, J ) 10.8, 2.0 Hz,
CHdCH2), 5.05 (1H, dd, J ) 6.4, 4.0 Hz, H-2’), 5.13 (1H, dd, J
) 17.2, 2.0 Hz, CHdCH2), 5.32 (1H, d, J ) 10.8 Hz, H-5’), 5.74
(2H, br, NH2), 6.19 (1H, d, J ) 6.4 Hz, H-1’), 6.50 (1H, dt, J )
17.2, 10.8 Hz, CHdCH2), 7.92 and 8.37 (2H, each as s, H-2
and H-8); FAB-MS m/z 504 (M+ + H). Anal. Calcd for
9-[5-(Z)-Br om o-2,3-b is-O-(ter t-b u t yld im et h ylsilyl)-5-
d eoxy-â-D-er yth r o-p en t-4-en ofu r a n osyl]a d en in e (14). A
mixture of 12 (785 mg, 1.02 mmol) and NBS (273 mg, 1.54
mmol) in THF (10 mL) was stirred for 1 h at rt. The reaction
mixture was partitioned between CHCl3 and saturated aque-
ous NaHCO3. Column chromatography (hexane/EtOAc ) 1/1)
of the organic layer gave 14 (533 mg, 94%) as a solid: mp 249-
251 °C; UV (MeOH) λmax 259 nm (ꢀ 17 600), λmin 231 nm (ꢀ
1
8000); H NMR (CDCl3) δ -0.31, -0.07, 0.45 and 0.16 (12H,
each as s, SiMe), 0.75 and 0.95 (18H, each as s, Bu-t), 4.69
(1H, d, J ) 4.4 Hz, H-3’), 5.16 (1H, dd, J ) 6.8, 4.4 Hz, H-2’),
5.39 (1H, s, H-5’), 5.87 (2H, br, NH2), 6.21 (1H, d, J ) 6.8 Hz,
H-1’), 7.92 and 8.34 (2H, each as s, H-2 and H-8); FAB-MS
m/z 556 and 558 (M+ + H). Anal. Calcd for C22H38BrN5O3Si2:
C, 47.47; H, 6.88; N, 12.58. Found: C, 47.76; H, 6.95; N, 12.44.
C
24H41N5O3Si2: C, 57.22; H, 8.20; N, 13.90. Found: C, 57.30;
9-[2,3-Bis-O-(ter t-b u t yld im et h ylsilyl)-5-(Z)-ch lor o-5-
d eoxy-â-D-er yth r o-p en t-4-en ofu r a n osyl]a d en in e (15). A
mixture of 12 (148 mg, 0.19 mmol) and NCS (40 mg, 0.29
mmol) in THF (5 mL) was heated at 60 °C for 12 h. The
reaction mixture was partitioned between CHCl3 and satur-
ated aqueous NaHCO3. Column chromatography (hexane/
EtOAc ) 1/1) of the organic layer gave 15 (74 mg, 76%) as a
solid: mp 243-245 °C; UV (MeOH) λmax 259 nm (ꢀ 16 200),
H, 8.44; N, 13.78.
9-[2,3-Bis-O-(ter t-bu tyldim eth ylsilyl)-5-d eoxy-5-(Z)-(tr i-
flu or ovin yl)-â-D-er yth r o-pen t-4-en ofu r an osyl]aden in e (20).
A mixture of 13 (400 mg, 0.66 mmol), tributy(trifluorolvinyl)-
tin (1.23 g, 3.31 mmol), (PPh3)2PdCl2 (98 mg, 0.13 mmol), and
CuI (26 mg, 0.13 mmol) in DMF (3 mL) was heated at 60 °C
for 40 min under positive pressure of dry Ar. The reaction
mixture was partitioned between EtOAc and saturated aque-
ous NaHCO3. Florisil column chromatography (hexane/EtOAc
) 2/1) of the organic layer gave 20 (263 mg, 71%) as a solid:
mp 152-154 °C; UV (MeOH) λmax 249 nm (ꢀ 23 400), λmin 225
nm (ꢀ 17 000); 1H NMR (CDCl3) δ -0.21, -0.04, 0.15 and 0.16
(12H, each as s, SiMe), 0.79 and 0.96 (18H, each as s, Bu-t),
4.80 (1H, d, J ) 4.4 Hz, H-3’), 5.03 (1H, dd, J H,F ) 22.8, 3.6
Hz, H-5’), 5.10 (dd, J ) 5.2, 4.4 Hz, H-2’), 5.60 (2H, br, NH2),
6.17 (1H, d, J ) 5.2 Hz, H-1’), 7.87 and 8.35 (2H, each as s,
H-2 and H-8); 13C NMR (CDCl3) δ -5.2, -4.7, -4.5 and -4.4
(SiMe), 17.9 and 18.0 (CMe3), 25.6 and 25.8 (CMe3), 73.0 (C-
3’), 73.6 (C-2’), 87.4 (dd, J C, F ) 20.7, 5.2 Hz, C-5’), 90.6 (C-1’),
120.5 (C-5), 125.6 (ddd, J C,F ) 231.7, 50.7, 20.7 Hz, C-6’), 140.0
(C-8), 149.8 (C-4), 152.3 (ddd, J C,F ) 290.7, 281.4, and 47.6
Hz, C-7’), 153.2 (C-2), 155.6 (C-6), 156.5 (dd, J C,F ) 9.3, 4.1
Hz, C-4’); 19F NMR (CDCl3) δ -175.5 (ddd, J F,F) 110.9, 30.2
1
λmin 232 nm (ꢀ 5400); H NMR (CDCl3) δ -0.32, -0.07, 0.14
and 0.15 (12H, each as s, SiMe), 0.75 and 0.95 (18H, each as
s, Bu-t), 4.68 (1H, d, J ) 4.0 Hz, H-3’), 5.20 (1H, dd, J ) 6.8,
4.0 Hz, H-2’), 5.40 (1H, s, H-5’), 5.68 (2H, br, NH2), 6.20 (1H,
d, J ) 6.8 Hz, H-1’), 7.90 and 8.36 (2H, each as s, H-2 and
H-8); FAB-MS m/z 513 (M+ + H). Anal. Calcd for C22H38
-
ClN5O3Si2: C, 51.59; H, 7.48; N, 13.67. Found: C, 51.83; H,
7.71; N, 13.65.
9-[2,3-Bis-O-(ter t-bu tyld im eth ylsilyl)-5-d eoxy-5-flu or o-
â-D-er yth r o-p en t-4-en ofu r a n osyl]a d en in e (17). A mixture
of 12 (139 mg, 0.18 mmol) and Selectfluor (96 mg, 0.27 mmol)
in CH3CN (5 mL) was stirred at room temperature for 15 min
under positive pressure of dry Ar. The reaction mixture was
partitioned between CHCl3 and saturated aqueous NaHCO3.
Column chromatography (hexane/EtOAc ) 1/1) of the organic
layer gave a mixture of 17Z, 17E, and 10 (84 mg, 17Z/17E/10
) 1/3/5). Yields of the products were calculated by 1H NMR
spectroscopy by integrating H-5′: 17, 43%; 10, 53%: 19F NMR
(CDCl3) δ -66.7 (d, J ) 78.8 Hz), -61.8 (d, J ) 75.2 Hz).
Partial 1H NMR data for 17E: 1H NMR (CDCl3) δ -0.37,
-0.15, and 0.04 (12H, each as s, SiMe), 0.59 and 0.80 (18H,
each as s, Bu-t), 4.93 (1H, q, J ) 4.4, 2.0 Hz, H-3′), 5.23 (1H,
dd, J ) 8.0, 4.4 Hz, H-2′), 6.10 (1H, d, J ) 8.0 Hz, H-1′), 6.85
(1H, d, J ) 78.8 Hz, H-5′), 7.87 and 8.37 (2H, each as s, H-2
Hz, J H,F) 22.8 Hz), -117.3 (ddd, J F,F ) 110.9, 69.2 Hz, J H,F
)
3.6 Hz), -103.7 (dd, J F,F ) 69.2 and 30.2 Hz). FAB-MS m/z
558 (M+ + H). Anal. Calcd for C24H38F3N5O3Si2: C, 51.68; H,
6.87; N, 12.56. Found: C, 51.88; H, 6.95; N,12.63.
9-[2,3-Bis-O-(ter t-bu tyldim eth ylsilyl)-5-deoxy-5-(Z)-(tr i-
m eth ysilyleth yn yl)-â-D-er yth r o-p en t-4-en ofu r a n osyl]a d -
en in e (21). A mixture of 13 (700 mg, 1.16 mmol), trimethyl-
silylacetylene (491 µL, 3.48 mmol), (PPh3)2PdCl2 (161 mg, 0.23
mmol), CuI (44 mg, 0.23 mmol), and i-Pr2NEt (606 µL, 3.48
1
and H-8). Partial H NMR data for 17Z: 1H NMR (CDCl3) δ
J . Org. Chem, Vol. 69, No. 1, 2004 77

Kumamoto et al.
mmol) in DMF (10 mL) was heated at 70 °C for 45 min under
positive pressure of dry Ar. The reaction mixture was parti-
tioned between EtOAc and saturated aqueous NaHCO3.
Column chromatography (hexane/EtOAc ) 1/1) of the organic
layer gave 21 (446 mg, 67%) as a solid: mp 150-154 °C; UV
(MeOH) λmax 254 nm (ꢀ 35 900), λmin 224 nm (ꢀ 12 500); 1H
NMR (CDCl3) δ -0.21, -0.03, 0.14 and 0.15 (12H, each as s,
SiMe), 0.14 (9H, s, TMS), 0.79 and 0.65 (18H, each as s, Bu-t),
4.82 (1H, s, H-5’), 4.83 (1H, d, J ) 4.4 Hz, H-3’), 4.96 (1H, dd,
J ) 5.2, 4.4 Hz, H-2’), 5.63 (2H, br, NH2), 6.23 (1H, d, J ) 5.2
Hz, H-1’), 7.91 and 8.36 (2H, each as s, H-2 and H-8); FAB-
MS m/z 574 (M+ + H). Anal. Calcd for C27H47N5O3Si3: C, 56.50;
H, 8.25; N, 12.20. Found: C, 56.73; H, 8.50; N, 12.14.
9-[2,3-Bis-O-(ter t-b u t yld im e t h ylsilyl)-5-(Z)-cya n o-5-
d eoxy-â-D-er yth r o-p en t-4-en ofu r a n osyl]a d en in e (22). A
mixture of 13 (200 mg, 0.33 mmol), tributyltin cyanide (420
mg, 1.33 mmol), (PPh3)2PdCl2 (55 mg, 0.07 mmol), and CuI
(63 mg, 0.33 mmol) in DMF (1.5 mL) was heated at 120 °C for
24 h under positive pressure of dry Ar. The reaction mixture
was partitioned between EtOAc and saturated aqueous NaH-
CO3. Column chromatography (hexane/EtOAc ) 1/3) of the
organic layer gave 22 (132 mg, 80%) as a solid, which was
crystallized from Et2O-hexane: mp 249-251 °C; UV (MeOH)
λmax 257 nm (ꢀ 15 700) and 233 nm (ꢀ 18 500), λmin 248 nm (ꢀ
15 600) and 223 nm (ꢀ 17 000); 1H NMR (CDCl3) δ -0.09, 0.03,
0.17, and 0.18 (12H, each as s, SiMe), 0.84 and 0.96 (18H, each
as s, Bu-t), 4.59 (1H, d, J ) 1.2 Hz, H-5’), 5.15 (1H, dd, J )
4.8, 3.2 Hz, H-2’), 5.39 (1H, dd, J ) 4.8, 1.2 Hz, H-3’), 5.95
(2H, br, NH2), 6.16 (1H, d, J ) 3.2 Hz, H-1’), 7.88 and 8.31
(2H, each as s, H-2 and H-8); 13C NMR (CDCl3) δ -5.0, -4.7,
-4.6 and -4.5 (SiMe), 17.9 and 18.1 (CMe3), 25.6 and 25.7
(CMe3), 70.5 (C-5’), 72.5 (C-3’), 73.0 (C-2’), 92.3 (C-1’), 114.9
(CN), 120.5 (C-5), 140.3 (C-8), 149.3 (C-4), 153.3 (C-2), 155.9
(C-6), 173.7 (C-4’); FAB-MS m/z 503 (M+ + H). Anal. Calcd
for C23H38N6O3Si2: C, 54.95; H, 7.62; N, 16.72. Found: C,
54.93; H, 7.76; N, 16.66.
Ack n ow led gm en t. We are grateful to J SPS (J apan
Society for the Promotion of Science) for support of this
work through KAKENHI Grant Nos. 15790075 (to H.K.)
and 15590020 (to H.T.).
J O030256Y
78 J . Org. Chem., Vol. 69, No. 1, 2004