Self-sacrificing acetylation observed during attempted desilylation of 1-[4-benzenesulfonyl-3-O-(tert-butyldimethylsilyl)-2-deoxy-5-O-methanesulfonyl- α-l-threo-pentofuranosyl]thymine

Article abstract of DOI:10.1016/j.tetlet.2010.06.039

Desilylation of 1-[4-benzenesulfonyl-3-O-(tert-butyldimethylsilyl)-2-deoxy- 5-O-methanesulfonyl-α-l-threo-pentofuranosyl]thymine (4) with Bu ₄NF/THF, when carried out at room temperature, gave four products. Among these, there were 1-[3-O-acetyl-4-benzenesulfonyl-2-deoxy-5-O- methanesulfonyl-α-l-threo-pentofuranosyl]thymine (7) and thymine. A possible reaction mechanism is proposed, which suggests the origin of 3′-O-acetyl group of 7 and thymine as well as structures of the other two products (9a and 9b).

Full text of DOI:10.1016/j.tetlet.2010.06.039

Tetrahedron Letters 51 (2010) 4253–4255
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Self-sacrificing acetylation observed during attempted desilylation
of 1-[4-benzenesulfonyl-3-O-(tert-butyldimethylsilyl)-2-deoxy-5-O-
methanesulfonyl-
a-L-threo-pentofuranosyl]thymine
*
Hisashi Shimada, Yutaka Kubota, Hiromichi Tanaka
School of Pharmaceutical Sciences, Showa University, 1-5-8 Hatanodai, Shinagawa-ku, Tokyo 142-8555, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Desilylation of 1-[4-benzenesulfonyl-3-O-(tert-butyldimethylsilyl)-2-deoxy-5-O-methanesulfonyl-a-L-
threo-pentofuranosyl]thymine (4) with Bu4NF/THF, when carried out at room temperature, gave four
Received 21 April 2010
Revised 3 June 2010
Accepted 7 June 2010
Available online 11 June 2010
products. Among these, there were 1-[3-O-acetyl-4-benzenesulfonyl-2-deoxy-5-O-methanesulfonyl-
a
-
L
-threo-pentofuranosyl]thymine (7) and thymine. A possible reaction mechanism is proposed, which sug-
gests the origin of 30-O-acetyl group of 7 and thymine as well as structures of the other two products (9a
and 9b).
Keywords:
Ó 2010 Elsevier Ltd. All rights reserved.
Thymine nucleoside
Acetyl phenyl sulfone
Acetylation
Thymine-substituted propenal
1. Introduction
ature for 1 h, complete disappearance of 4 was confirmed by TLC
(hexane/EtOAc = 1:3). After reaction overnight, careful TLC analysis
We have recently reported a method for the preparation of 2,1
having a benzenesulfonyl leaving group at the 40-position, from
thymidine-50-aldehyde (1).1,2 Compound 2, after 50-O-silylation,
undergoes stereoselective nucleophilic substitution with organo-
silicon reagents in the presence of SnCl4, as exemplified in Scheme
1 for the preparation of 40-allyl-30,50-bis-O-(tert-butyldimethylsi-
lyl)thymidine (3).1
of the reaction mixture in an alternative solvent system (CH2Cl2/
MeOH = 10:1) revealed that, in addition to three products, thymine
was also formed as a polar product. This was ascertained by 1H
NMR spectroscopy after short column chromatography (CH2Cl2/
MeOH = 10:1) of the reaction mixture, but the actual yield of thy-
mine could not be determined due to contamination by tetrabutyl-
ammonium salts. A mixture of the three other products was
subjected to HPLC separation (CHCl3/MeOH = 25:1).
To apply this chemistry to the synthesis of 40-carbon-substi-
tuted analogues of 20,30-didehydro-30-deoxythymidine (d4T), we
were in need of 6, and thought that desilylation of the 50-O-meth-
anesulfonyl derivative 4 with Bu4NF would lead to the formation of
the oxetane derivative 5 that is anticipated to undergo base-cata-
lyzed elimination to yield 6, as reported in the preparation of
d4T.3 However, upon reacting 4 with Bu4NF in THF at room temper-
ature, an unexpected reaction took place, which is the subject of
the present communication.
The 1H NMR spectrum of the fastest running product (tR
7.4 min) in CDCl3 showed that it has one acetyl group (d 2.18,
3H, singlet) with the 50-O-mesyl group remaining intact. Also, its
H-30 resonance appeared at a lower field (d 6.23) when compared
with that of 4 (d 5.29). These data suggest the structure of this
product to be the 30-O-acetyl derivative 7,4 which was also sup-
ported by FAB-MS (m/z 503, M++H). The isolated yield of 7 was
29%. Where has the 30-O-acetyl group of 7 come from? We would
like to propose a possible reaction mechanism as shown in Scheme
2, which also suggests the structures of the other two products.
2. Results and discussion
O
Me
Compound 2, 1-[4-benzenesulfonyl-3-O-(tert-butyldimethylsi-
NH
Ph
lyl)-2-deoxy-a-L-threo-pentofuranosyl]thymine, was mesylated in
O
O
N
a conventional manner to give the 50-O-mesylate (4) in 82% yield.
When 4 was reacted with Bu4NF (1.2 equiv) in THF at room temper-
S
O
O
MsO
AcO
* Corresponding author.
E-mail address: hirotnk@pharm.showa-u.ac.jp (H. Tanaka).
7
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.06.039

4254
H. Shimada et al. / Tetrahedron Letters 51 (2010) 4253–4255
O
N
O
O
N
Me
Me
Me
NH
NH
NH
Ph
O
N
O
O
O
O
S
SiMe3
SnCl4
TBSO
O
O
O
O
O-silylation
HO
TBSO
TBSO
TBSO
2
3
1
O
N
O
N
O
N
Me
O
Me
O
Me
NH
NH
NH
Ph
O
Ph
O
HO
Ph
O
O
S
O
O
O
S
S
O
O
O
MsO
TBSO
O
4
5
6
Scheme 1.
HPLC. That 7-D contains the 30-O-monodeuterated acetyl group be-
Desilylation of 4 gives rise to the 30-alkoxide A, which under-
goes furanose ring opening leading to B or, through subsequent
elimination of the mesyloxy group, to C. Upon deprotonation of
came clear from its 1H NMR spectrum measured in CDCl3, in which
the characteristic splitting due to the presence of deuterium (JH,D
was present: d 2.17 (2H, triplet, JH,D = 2.0 Hz, COCH2D).
)
the a-hydrogen of the formyl group in C, there could be two possi-
ble elimination pathways. In route-a, the thymine moiety serves as
a leaving group (the formation of thymine was actually observed as
mentioned the above). On the other hand, when elimination takes
place through route-b, acetyl phenyl sulfone (8) as well as 9 should
result. Since 8 is expected to serve as an acetylating reagent,5 its
reaction with A would lead to the formation of 7.
O
O
Me
Me
NH
NH
Ph
O
MsO
Ph
O
MsO
O
N
N
O
O
O
S
S
O
O
To obtain convincing evidence for the proposed mechanism
regarding the formation of 7, the 50-monodeuterated 4 (4-D, a mix-
ture of two diastereomers with a ratio of 2:3, isotopic purity >99%)
was prepared from 1.6 When 4-D was reacted with Bu4NF in THF at
room temperature overnight, 7-D was isolated in 33% yield by
D
TBSO
D
CH2D
O
C
O
4-D
7-D
O
O
N
Me
Me
NH
NH
O
Ph
Ph
O
O
O
N
O
S
S
O
Bu4NF
THF, rt
O
O
4
−
MsO
MsO
−
O
O
A
B
O
O
S
A
C
7
CH3
O
8
O
N
Me
O
NH
Ph
O
O
Thy
Me
O
a
S
route-a
route-b
HN
O
b
O
N
H
9
CHO
H
O
C
Scheme 2.

H. Shimada et al. / Tetrahedron Letters 51 (2010) 4253–4255
4255
gratefully thank Dr. Kikuko Hayamizu (National Institute of
Advanced Industrial Science and Technology, Tsukuba) for helpful
discussion regarding 1H NMR analysis of the deuterated
compounds.
As regard to 9, which was found to be a mixture of two com-
pounds, we initially thought that these were geometrical isomers,
since these compounds after HPLC separation gave identical M++H
(m/z 181) by FAB-MS and their 1H NMR spectra measured in
DMSO-d6 showed for each the presence of the thymine moiety, an
aldehyde proton and two vinylic protons. The 1H NMR data of the
faster running product (9a)7 (tR 8.8 min, obtained in 15% yield) were
found to be in full agreement with those reported for (E)-3-(thymin-
1-yl)propenal.8,9 However, the slower running product (tR 10.3 min,
obtained in 7% yield) gave a slightly larger vinylic coupling constant
(J = 14.8 Hz) than that of 9a (J = 14.6 Hz), which is certainly inconsis-
tent with the (Z)-configuration. Unambiguous structure determina-
tion that the slower running product is (E)-3-(thymin-3-yl)propenal
(9b) came from a HMBC experiment: multiple bond connectivity
was observed between H-3 (d 8.18) of the propenal portion and both
C-2 (d 149.88) and C-4 (d 162.73) carbonyl carbons of the thymine
moiety.10 At the present time, we have no evidence to explain the
mechanism for the formation of 9b.11,12
References and notes
1. Shimada, H.; Kikuchi, S.; Okuda, S.; Haraguchi, K.; Tanaka, H. Tetrahedron 2009,
65, 6008.
2. Giese, B.; Erdmann, P.; Giraud, J.; Göbel, T.; Peyretta, M.; Schäfer, T.; von
Raumer, M. Tetrahedron Lett. 1994, 35, 2683.
3. Horwitz, J. P.; Chua, J.; Da Rooge, M. A.; Noel, M.; Klundt, I. L. J. Org. Chem. 1966,
31, 205.
4. Physical data for 7: 1H NMR (CDCl3) d 2.05 (3H, d, J = 1.2 Hz, 5-Me), 2.18 (3H, s,
Ac), 2.47 (1H, ddd, J = 2.2, 6.3, and 14.4 Hz, H-20), 2.87 (3H, s, SO2Me), 2.84–2.92
(1H, m, H-20), 4.40 and 4.47 (2H, each as d, J = 11.5 Hz, H-50), 6.23 (1H, dd,
J = 2.2 and 6.6 Hz, H-30), 6.77 (1H, dd, J = 6.3 and 9.0 Hz, H-10), 7.61–7.89 (6H,
m, J = 1.2 Hz, H-6 and Ph), 9.17 (1H, br, NH); 13C NMR (CDCl3) d 12.72, 20.68,
37.16, 37.94, 64.59, 72.04, 86.04, 99.21, 113.06, 129.48, 130.31, 134.37, 134.74,
135.26, 150.56, 163.21, 169.01; FAB-MS (m/z) 503 (M++H).
5. Reaction between p-anisoyl p-tolyl sulfone and MeOH has been reported to
give methyl p-anisate: Schank, K.; Werner, F. Liebigs Ann. Chem. 1980, 1477.
6. Reaction sequence for the preparation of 4-D from 1 is shown below. For
detailed experimental procedures, see Ref. 1.
O
O
CHO
Me
Me
NH
N
O
O
Me
Me
NH
NH
N
O
O
N
H
O
N
O
N
O
Et3N
CH2Cl2
NaBD4
MeOH
1)
m-CPBA
PhS
PhS
HO
1
N SPh
4-D
O
O
2) MsCl/pyridine
O
O
D
9b
9a
TBSO
TBSO
CHO
7. 1H NMR (400 MHz) data of 9a and 9b measured in DMSO-d6 are as follows. For
9a: d 1.83 (3H, d, J = 1.2 Hz, 5-Me), 6.47 (1H, dd, J = 7.6 and 14.6 Hz, CH@CH–
CHO), 8.04 (1H, d, J = 1.2 Hz, H-6), 8.16 (1H, d, J = 14.6 Hz, CH@CHCHO), 9.57 (1H,
d, J = 7.6 Hz, CHO), 11.83 (1H, br, NH). For 9b: d 1.81 (3H, d, J = 1.2 Hz, 5-Me), 7.09
(1H, dd, J = 7.8 and 14.8 Hz, CH@CHCHO), 7.44 (1H, d, J = 1.2 Hz, H-6), 8.18 (1H, d,
J = 14.8 Hz, CH@CHCHO), 9.57 (1H, d, J = 7.8 Hz, CHO), 11.34 (1H, br, NH).
8. Johnson, F.; Pillai, K. M. R.; Grollman, A. P.; Tseng, L.; Takeshita, M. J. Med. Chem.
1984, 27, 954.
Finally, it should be mentioned that the formation of 7, 9a, 9b
and thymine from 4 can be avoided by conducting the reaction
at ꢀ78 °C. Thus, after reaction at this temperature for 6 h, 4 gave
the desilylated product in 72% yield. Anticipated formation of the
oxetane 5 (Scheme 1) was not observed.
9. Compound 9a has been isolated as a product of bleomycin-induced strand-
scission of DNA: Giloni, L.; Takeshita, M.; Johnson, F.; Iden, C.; Grollman, A. P. J.
Biol. Chem. 1981, 256, 8608.
10. Additional evidence for the N3-substituted regiochemistry of 9b is that
addition of D2O to the 1H NMR sample sharpened the H-6 signal.
11. One referee pointed out the possibility that 9b may result from addition–
elimination reaction between 9a and N3-anion of thymine. We agree that 9b
derives presumably from 9a. However, as cited in Ref. 12, it is known that
addition of thymine to Michael acceptors in the presence of DBU takes place
exclusively at the N1-position. Therefore, for the formation of 9b, we assume
that the nucleophile reacting with 9a would be not thymine itself but certain
N1-substituted thymine derivative(s).
3. Conclusion
An attempted desilylation of 4 with Bu4NF in THF at room tem-
perature gave unexpected products 7, 9a, 9b and thymine. Their
structures were speculated based on a possible reaction mecha-
nism shown in Scheme 2. Convincing mechanistic evidence for
the formation of 7 was obtained by employing the 50-monodeuter-
ated 4 (4-D).
12. Scheiner, P.; Geer, A.; Bucknor, A.-M.; Imbach, J.-L.; Schinazi, R. F. J. Med. Chem.
1989, 32, 73.
Acknowledgements
Generous financial support (to H.S.) from Showa University Re-
search Grant for Young Researchers is acknowledged. The authors