A novel synthesis of 4'-thionucleosides and a potential stereospecific route to pyrimidine nucleosides

Article abstract of DOI:10.1016/S0040-4039(00)01796-2

Starting with the L-ascorbate derived epoxide 1, a di-t-butyl dithioacetal cyclisation route to 2'-deoxy-4'-thionucleosides has been developed. Based on an intermediate in this route, a novel and stereospecific route to α- or β-pyrimidine nucleosides has been conceived, but its implementation failed at a key ring-closure step. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)01796-2

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 10099–10105
A novel synthesis of 4%-thionucleosides and a potential
stereospecific route to pyrimidine nucleosides
J. Allen Miller,*,† Ashley W. Pugh, G. Mustafa Ullah and Clare Gutteridge
Department of Medicinal Chemistry, Wellcome Research Laboratories, Langley Court, Beckenham,
Kent BR3 3BS, UK
Received 25 August 2000; accepted 11 October 2000
Abstract
Starting with the
L
-ascorbate derived epoxide 1, a di-t-butyl dithioacetal cyclisation route to 2%-deoxy-
4%-thionucleosides has been developed. Based on an intermediate in this route, a novel and stereospecific
route to a- or b-pyrimidine nucleosides has been conceived, but its implementation failed at a key
ring-closure step. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: 2%-deoxy-4%-thionucleoside synthesis; di-t-butyl dithioacetal cyclisation; epoxide opening.
4%-Thionucleosides have had something of a renaissance over the last decade or so, largely due
to their apparent promise as anti-viral1 and anti-tumour agents,2 although none as yet appear to
have reached the marketplace as therapeutics. In keeping with this interest, a great deal of
ingenuity has recently gone into novel syntheses of 4%-thionucleosides and their sugar precur-
sors,3 and considerable progress has been made since the pioneering, but long and difficult,
routes used by Whistler, Reist and Bobek and their collaborators in the 1960s and 1970s.4
These early routes involved forming the S to C-4 bond of the thiosugar early in the synthesis,
but more recent syntheses, some of which work on a multi-gram scale, have depended upon
formation of the S to C-1 bond first, followed by ring-closure of the sulphur at C-4. This has
now become established as the preferred approach in the 2-deoxypentose series, largely due to
the efforts of Walker et al.1b,5 and this route and its variants have been followed by several other
groups.6 Nevertheless, there are drawbacks associated with 2-deoxypentose starting material
availability, with the conditions used in the cyclisation step (benzyl iodide is a stoichiometric
product!) and with the difficult separation of the roughly equal amounts of the a- and
b-nucleosides formed after the base coupling step. In this paper, we show how the readily
* Corresponding author. Tel: +0044 (0)1625 500555; fax: +0044 (0)1625 500666; e-mail: allen.miller@protherics.com
†
Present address: Protherics PLC, Beechfield House, Lyme Green Business Park, Macclesfield, Cheshire SK11 0JL,
UK.
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)01796-2

10100
available epoxide 1 can be used as starting material for a 2%-deoxy-4%-thionucleoside synthesis
based on a novel di-t-butyl dithioacetal cyclisation (Scheme 1), and we also outline a potentially
general, albeit unfulfilled, strategy for the stereospecific synthesis of a- or b-nucleosides in the
pyrimidine series (Scheme 2).
Scheme 1. Reagents: (i) n-BuLi (R=Et 95%; R=Ph 90%; R=tBu 97%); (ii) NaH, BnBr, DMF (88%); (iii) HCl,
t
MeOH (96%); (iv) BuSi(Ph)2Cl, DMAP, pyridine (96%); (v) MeSO2Cl, DMAP, pyridine (91%); (vi) D, NaI, Et3N,
butan-2-one; (vii) D, Et3N, butan-2-one; (viii) D, DBU (4 equiv.); (ix) 2,4-bis-trimethylsilyloxy-5-ethylpyrimidine,
NBS, MeCN (82%); (x) Bu4NF (92%); (xi) BBr3, CH2Cl2 at −30°C (70%)
Our starting point is the epoxide 1, which can be prepared conveniently from
-ascorbic acid7
L
and one of the attractions of our approach is that the other three stereoisomers of 1 are also
cheap and readily available.7,8 Following earlier work of Trost et al.,9 and that of Kamikawa et
al.,10 epoxide 1 was opened exclusively at the terminal carbon by a range of anions derived by
metalation of the formaldehyde dithioacetals 2, to give the hydroxydithioacetals 3,11 as shown
in Scheme 1. Fairly standard protection–deprotection steps led to the mesylates 4, ready for
ring-closure. However, under a range of conditions the mesylates 4 (R=Ph or Et) did not
cyclise, but yielded a mixture of the E- and Z-vinyl sulphides 5, the result of 1,4-SR shifts
reflecting the prior experiences of Hughes et al., Holzapfel et al. and Blomberg et al.,12 on whose
work Walker’s route5 was based.
It is likely that the success of Walker5 and others6,12 with analogues of 4, (R=Bn) is a
consequence of the relative ease of nucleophilic attack (by iodide ion) on the benzylic carbon in
the presumed sulphonium intermediate, leading to S-debenzylation. We have now found that an
alternative way of achieving cyclisation via a sulphonium salt is to use a di-t-butyl dithioacetal,
e.g. 4, (R=tBu), and a strong base such as 1,8-diazabicyclo[5,4,0]undec-7-ene (DBU). It is
presumed that DBU facilitates de-t-butylation via loss of isobutylene, such that the kinetics of
t
ring-closure to 6 can compete with the 1,4-shift of a BuS group to give 5. This method allows
the protected thiosugar 6 to be obtained as mixed a- and b isomers in 50–60% yields, which
compares reasonably with the 35–40% yields found by others in the 2-deoxy series when R=Bn
and the iodide ion is used as a debenzylating agent.5,6 Finally, the known 4%-thionucleosides 7,

10101
Scheme 2. Reagents: (i) (RS)2CH2, n-BuLi, THF (62–90%); (ii) 2,4-bis-trimethylsilyloxy-5-ethylpyrimidine, NBS,
i
MeCN (79%); (iii) MeSO2Cl, pyridine; (iv) D, Pr2NH, n-propanol (97%, two steps); (v) Bu4NF (92%); (vi) NBS,
MeCN, 0–20°C; (vii) Hg(OAc)2, MeCN
of prime interest because of the powerful anti-herpes activity of the b isomer,13 were obtained
from 6 using standard deprotection procedures, thus confirming the configurational assignments
in 6.
The above synthesis of 7 shares a common feature with those of other 2%-deoxy-4%-thionu-
cleosides—unfavourable b:a ratios. Separation of the anomers can be difficult and laborious,
but is often essential because biological activity is resident only, or largely, in one isomer. In the
last few years, a number of groups, notably that led by Liotta,14,15 have successfully addressed
the anomer problem.16 Several of these approaches rely upon differential facial additions to
double bonds, an iminium bond in the first case14b and a 1,2-glycal in other cases.17 We regarded
chemistry derived from the acetal 3 and its analogues as an opportunity to explore a general a-
or b-specific approach to pyrimidine nucleosides based on stereospecific addition to an iminium
intermediate—see Scheme 2.
The proposed sequence can start from the acetals 3, or the simpler analogues 8,18 and the
detailed sequence for the latter is set out in Scheme 2. The desired outcome was the
anhydronucleoside-like compound 12, which lacks the usual 4%-hydroxymethyl group of a
pentose-derived anhydronucleoside. With acetals 3 as starting material, it was anticipated that a
‘normal’ 2,3%-anhydronucleoside would be produced, and subsequently would bring with it the
usual valued synthetic options19 associated with regiospecific ring-opening of these compounds.

10102
Clearly, the configuration at C-4 in 3, or its isomers (see Scheme 1 for numbering), will
determine the
D
- or -assignment to the 2,3%-anhydronucleoside. Moreover, the relative configu-
L
ration at C-3 and C-4 will determine whether the outcome is an a- or a b-nucleoside, e.g.
epoxide 1 should yield 3 with a 3S,4S-configuration, and this in turn should produce an
a,
L
-2,3%-anhydronucleoside.
This proposed chemistry was initially investigated with the dithioacetal 8, (R=Ph), available
from TBDMS-protected R-glycidol in 90% yield, but it was extended subsequently to the acetals
3. From 8, mono-exchange20 with 5-ethyluracil was achieved in good yield using N-bromosuc-
cinimide (NBS) to give diastereoisomers 9a (ratio varies from 3:2 to 1:1), which were then
converted into the separable methanesulphonates 9b,21 as in Scheme 2. Base-catalysed, quantita-
tive displacement of the methanesulphonates gave 10a, deprotection of which then gave 10b,22
ready, in principle, for the formation of the ‘sugar ring’ by addition of the hydroxy group to the
iminium ion 11. However, using NBS, the key ring-closure to give the anhydronucleoside 12 was
not successful. The lability of the PhS group in 10b was confirmed by mercuric acetate induced
exchange to give acetates 10c, presumably via the iminium species 11. Under these conditions,
the N-3 linked bicyclic aminol 1323 was also formed as a minor product.
The result with mercuric acetate suggests that an intramolecular exchange at C-1% may be
achievable, but the circumstances in our laboratory at that time did not permit a proper
evaluation of this prospect. As far as we are aware, this approach to homochiral nucleosides
remains novel and, we believe, worthy of further scrutiny.
General procedure for preparation of hydroxy-dithioacetals. Compound 3; (R=tBu): Di-(t-
butylthio)methane (422 mg, 2.2 mmol) was dissolved in dry THF (15 ml) under argon. After
cooling to −78°C, n-BuLi (1.6 M in hexane, 1.5 ml, 2.4 mmol) was added. After stirring the
solution for 30 min, epoxide 1 (288 mg, 2.0 mmol) was added dropwise as a solution in dry THF
(5 ml). The solution was slowly allowed to warm to 0°C and stirring continued until TLC
indicated complete consumption of the epoxide 1. Then ammonium chloride solution was added
to the mixture, and this was washed with ethyl acetate (3×15 ml). The combined extracts were
washed with water and then brine, and dried over magnesium sulphate. After filtration, the
solvent was evaporated and the resulting yellow oil was chromatographed (silica gel, 18:1
petrol:ether) to yield the hydroxy-di-t-butylthio acetal 3 (R=tBu) as a pure (TLC), colourless oil
(652 mg, 97%).
Ring-closure of di-t-butyl dithioacetal 4; (R=tBu). Compound 6: Dithioacetal 4 (R=tBu) (703
mg, 1.0 mmol) was added to a solution of DBU (0.61 ml, 4 mmol) in butan-2-one and the
solution refluxed overnight under nitrogen, by which time all of the dithioacetal had been
consumed (TLC in 9:1 petrol:ether). The solvent was removed and the resulting oil was
dissolved in dichloromethane (20 ml) and the solution washed with water (×3), then dilute HCl
(20 ml, 0.05 M). The organic phase was dried, and the solvent evaporated to leave an oily
product which was carefully columned on silica gel using 100:3 petrol:ether. The major product
(330 mg, 60%) was the thiosugar 6. The E- and Z-vinyl sulphides 5, (R=tBu) (15–25% and
1
<10%, respectively) were identified by their characteristic H NMR features (E-: l 6.3 (d, J=15
Hz, 1H, ꢀCHSꢁtBu) and 5.85 (dd, J=15 Hz, 7 Hz, 1H, HCꢀCHS-tBu); Z-: l 6.15 (d, J=11 Hz,
1H, ꢀCHSꢁtBu) and 5.66 (dd, J=11 Hz, 11 Hz, 1H, HCꢀCHSꢁtBu).
Procedure for mono-exchange of dithioacetals. Compound 9a: Bis-2,4-trimethylsilyloxy-5-
ethylpyrimidine was freshly prepared from 5-ethyluracil (308 mg, 2.2 mmol) using hexamethyld-
isilazane and chlorotrimethylsilane in dry acetonitrile (20 ml) under nitrogen at room
temperature.13 Diphenyl dithioacetal 8 (842 mg, 2.0 mmol) in dry acetonitrile (10 ml) was added

10103
to the stirred mixture. N-Bromosuccinimide (390 mg, 2.2 mmol) was then added, and the
mixture stirred until TLC showed consumption of all of the dithioacetal 8. The reaction was
quenched by adding a solution of sodium thiosulphate and conventional extractive work-up led
to a slightly yellow oil which was purified using flash chromatography. The mixture of
diastereoisomers 9a (710 mg, 79%) solidified upon standing overnight.
Procedure for ring-closure of the hydroxy thioaminals 9a. Compound 10a: Methanesulphonyl
chloride (230 mg, 2.0 mmol) in pyridine (5 ml) was added to a solution of the hydroxy
thioaminal 9a (676 mg, 1.5 mmol) in dichloromethane, stirred at 0°C. The temperature of the
solution was gradually allowed to rise to 20°C. After mesylation was complete (TLC), further
dichloromethane was added and conventional acid, then base, extractions yielded the crude
methanesulphonates 9b, which were then heated at reflux in a mixture of di-isopropylamine (1
ml) and n-propanol (10 ml). The methanesulphonates were consumed after 2–3 h (TLC) and
volatile materials were then removed on a rotary evaporator. The residual oil was dissolved in
dichloromethane (10 ml) and underwent acid and then base washes, prior to flash chromatogra-
phy, which yielded compound 10a (630 mg, 97%).
References
1. (a) For an overview, see Wnuk, S. F. Tetrahedron 1993, 49, 9877–9936; (b) Dyson, M. R.; Coe, P. L.; Walker,
R. T. J. Med. Chem. 1991, 34, 2782–2786; Dyson, M. R.; Coe, P. L.; Walker, R. T. J. Chem. Soc., Chem.
Commun. 1991 741–742; (c) For more recent examples, see: Young, R. J.; Shaw-Ponter, S.; Thomson, J. B.;
Miller, J. A.; Cumming, J. G.; Pugh, A. W.; Rider, P. Bioorg. Med. Chem. Lett. 1995, 5, 2599–2604; Yoshimura,
Y.; Kitano, K.; Yamada, K.; Sakata, S.; Miura, S.; Ashida, N.; Machida, H.; Matsuda, A. Nucleic Acids
Symposium Series No. 37, Oxford University Press, 1997, 43–44.
2. Secrist III, J. A.; Parker, W. B.; Tiwari, K. N.; Messini, L.; Shaddix, S. C.; Rose, L. M.; Bennett Jr., L. L.;
Montgomery, J. A. Nucleosides, Nucleotides 1995, 14(3–5), 675–686; Secrist III, J. A.; Tiwari, K. N.; Riordan, J.
M.; Montgomery, J. A. J. Med. Chem. 1991, 34, 2361–2366.
3. Uenishi, J.; Takahashi, K.; Motoyama, M.; Akashi, H.; Sasaki, T. Nucleosides, Nucleotides, 1994, 13(6&7),
1347–1361; Uenishi, J.; Motoyama, M.; Nishiyama, Y.; Wakabayashi, S. J. Chem. Soc., Chem. Commun. 1991,
1421–1422; Uenishi, J.; Kawanami, H.; Kubo, Y. Nucleic Acids Symposium Series No. 29, Oxford University
Press, 1993, 37–38; O’Neil, I. A.; Hamilton, K. M. Synlett 1992, 791–792; O’Neil, I. A.; Hamilton, K. M.; Miller,
J. A. Synlett. 1995, 1053–1054; Yoshimura, Y.; Endo, M.; Miura, S.; Sakata, S. J. Org. Chem. 1999, 64,
7912–7920; Yamada, K.; Sakata, S.; Yoshimura, Y. J. Org. Chem. 1998, 63, 6891–6899.
4. Reist, E. J.; Gueffroy, D. E.; Goodman, L. J. Am. Chem. Soc. 1964, 86, 5658–5663; Whistler, R. L.; Nayak, U.
G.; Perkins Jr., A. W. J. Org. Chem. 1970, 35, 519–521; Nayak, U. G.; Whistler, R. L. Annalen 1970, 741,
131–138; Fu, Y.-L.; Bobek, M. J. Org. Chem. 1976, 41, 3831–3834.
5. Dyson, M. R.; Coe, P. L.; Walker, R. T. Carbohydr. Res. 1991, 216, 237–248.
6. Tiwari, K. N.; Montgomery, J. A.; Secrest III, J. A. Nucleosides, Nucleotides 1993, 12(8), 841–846; Huang, B.;
Hui, Y. Nucleosides, Nucleotides 1993, 12(2), 139–147; Birk, C.; Voss, J.; Wirsching, J. Carbohydr. Res. 1997, 304,
239–247.
7. Abushanab, E.; Bessodes, M.; Antonakis, K. Tetrahedron Lett. 1984, 35(25), 3841–3844; Abushanab, E.;
Vemishetti, P.; Leiby, R. W.; Singh, K.; Mikkilineni, A. B.; Wu, D. C.-J.; Saibaba, R.; Panzica, R. P. J. Org.
Chem. 1988, 53, 2598–2602; Vargeese, C.; Abushanab, E. J. Org. Chem. 1990, 55, 4400–4403; Le Merrer, Y.;
Gravier-Pelletier, C.; Dumas, J.; Depezay, J. C. Tetrahedron Lett. 1990, 31(7), 1003–1006; Teng, K.; Marquez, V.
E.; Milne, G. W. A.; Barchi, J. J.; Kazanietz, M. G.; Lewin, N. E.; Blumberg, P. M.; Abushanab, E. J. Am.
Chem. Soc. 1992, 114, 1059–1070.
8. (a) Via cis-but-2-en-1,4-diol and Sharpless epoxidation: Rama Rao, A. V.; Bose, D. S.; Gurjar, M. K.;
Ravindranathan, T. Tetrahedron 1989, 45(22), 7031–7040; Rama Rao, A. V.; Gurjar, M. K.; Bose, D. S.; Devi,
R. R. J. Org. Chem. 1991, 56, 1320–1321; Neagu, C.; Hase, T. Tetrahedron Lett. 1993, 34(10), 1629–1630;

10104
Azymah, M.; Chavis, C.; Lucas, M.; Imbach, J.-L. J. Chem. Soc., Perkin Trans. 1 1991, 1561–1563; Gurjar, M.
K.; Devi, N. R. Tetrahedron: Asymmetry 1994, 5(4), 755–758; (b) Via the chiral pool: Pottie, M.; Van der Eycken,
J.; Vandewalle, M.; Ro¨per, H. Tetrahedron: Asymmetry 1991, 2(5), 329–330; Gravier-Pelletier, C.; Dumas, J.; Le
Merrer, Y.; Depezay, J. C. Tetrahedron Lett. 1991, 32(9), 1165–1168; Ghosh, A. K.; McKee, S. P.; Lee, H. Y.;
Thompson, W. J. J. Chem. Soc., Chem. Commun. 1992, 273–274.
9. Trost, B. M.; Nubling, C. Carbohydr. Res. 1990, 202, 1–12.
10. Yamagata, K.; Yamagiwa, Y.; Kamikawa, T. J. Chem. Soc., Perkin Trans. 1 1990, 3355–3357.
1
t
11. Hydroxy di-t-butyl dithioacetal 3; (R=tBu): H NMR (CDCl3) l 1.38 (s, 3H, CH3), 1.42 (s, 9H, BuS), 1.44 (s,
9H, tBuS), 1.46 (s, 3H, CH3), 1.68–1.80 (ddd, 1H, 2-CH), 2.05–2.25 (ddd, 1H, 2-CH), 2.5 (solvent and
concentration dependent) (d, J=5 Hz, HOꢁCH,), 3.70–4.05 (m, 4H), 4.23 (dd, J=10 Hz, 4 Hz, 1-CH). MS (m/z)
336 (M+). Anal (C16H32O3S2): C, H.
12. These early studies showed how 1,1-dithioacetals of simple sugars can transfer an RS-group to carbons 4 or 5,
via a cyclic sulphonium ion, which, in turn, can lead to a ring-opened sulphenium ion. In Blomberg et al. and
Hughes et al. this sulphenium ion is trapped by a pendant ꢁOH group, while we observe the formation of vinyl
sulphides 5, because there is no free ꢁOH group in our precursor 4. Hughes, N. A.; Robson, R. J. Chem. Soc.
(C) 1966, 2366–2368; Harness, J.; Hughes, N. A. J. Chem. Soc., Chem. Commun. 1971, 811; Blumberg, K.;
Fuccello, A.; van Es, T. Carbohydr. Res. 1979, 70, 217–232; Bredenkamp, M. W.; Holzapfel, C. W.; Swanepoel,
A. D. Tetrahedron Lett. 1990, 31(19), 2759–2762.
13. Rahim, S. G.; Trivedi, N.; Bogunovic-Batchelor, M. V.; Hardy, G. W.; Mills, G.; Selway, J. W. T.; Snowden, W.;
Littler, E.; Coe, P. L.; Basnak, I.; Whale, R. F.; Walker, R. T. J. Med. Chem. 1996, 39, 789–795.
14. (a) For a comprehensive review of b-selective synthesis, see: Wilson, L. J.; Hager, M. W.; El-Kattan, Y. A.;
Liotta, D. C. Synthesis 1995, 1465–1479; (b) Hager, M. W.; Liotta, D. C. J. Am. Chem. Soc. 1991, 113,
5117–5119.
15. Choi, W.-B.; Wilson, L. J.; Yeola, S.; Liotta, D. C. J. Am. Chem. Soc. 1991, 113, 9377–9379; Hoong, L. K.;
Strange, L. E.; Liotta, D. C.; Koszalka, G. W.; Burns, C. L.; Schinazi, R. F. J. Org. Chem. 1992, 57, 5563–5565;
Kim, H. O.; Jeong, L. S.; Lee, S. N.; Yoo, S. J.; Moon, H. R.; Kim, K. S.; Chun, M. W. J. Chem. Soc., Perkin
Trans. 1 2000, 1327–1329; Mukaiyama, T.; Matsutani, T.; Shimomura, N. Chem. Lett. 1993, 1627–1630;
Mukaiyama, T.; Shimomura, N. Chem. Lett. 1993, 781–784; Mukaiyama, T.; Matsubara, K. Chem. Lett. 1992,
1041–1044; Mukaiyama, T.; Matsubara, K.; Suda, S. Chem. Lett. 1991, 981–984; Mukaiyama, T.; Matsutani, T.;
Shimomura, N. Chem. Lett. 1994, 2089–2092.
16. (a) For intramolecular delivery of the base, see: Xia, X.; Wang, J.; Hager, M. W.; Sisti, N.; Liotta, D. C.
Tetrahedron Lett. 1997, 38(7), 1111–1114; Jung, M. E.; Castro, C. J. Org. Chem. 1993, 58, 807–808; Sugimura,
H.; Taguchi, R.; Sujino, K.; Walker, R. T. Nucleic Acids Symposium Series No. 39, Oxford University Press,
1998, 9–10; (b) For ligand- or ligand:catalyst-directed intermolecular delivery of the base, see: Sugimura, H.;
Osumi, K.; Yamazaki, T.; Yamaya, T. Tetrahedron Lett. 1991, 32(15), 1813–1816; Sujino, K.; Sugimura, H.
Synlett 1992, 553–555; Sugimura, H.; Sujino, K.; Osumi, K. Tetrahedron Lett. 1992, 33(18), 2515–2516.
17. (a) For stereoselective addition to 1,2-glycals, see: Wang, J.; Wurster, J. A.; Wilson, L. J.; Liotta, D. C.
Tetrahedron Lett. 1993, 34(31), 4881–4884; Kim, C. U.; Misco, P. F. Tetrahedron Lett. 1992, 33(39), 5733–5736;
Kim, C. U.; Luh, B. Y.; Martin, J. C. J. Org. Chem. 1991, 56, 2642–2647; Diaz, Y.; El-Laghdach, A.; Castillo´n,
S. Tetrahedron 1997, 53(31), 10921–10938; El-Laghdach, A.; Diaz, Y.; Castillo´n, S. Tetrahedron Lett. 1993,
34(17), 2821–2822; Kassou, M.; Castillo´n, S. Tetrahedron Lett. 1994, 35(30), 5513–5516. (b) For stereoselective
addition to 4-thio-1,2-glycals, see: Miller, J. A.; Pugh, A. W.; Ullah, G. M. Tetrahedron Lett. 2000, 41,
3265–3268; Haraguchi, K.; Nishikawa, A.; Sasakura, E.; Tanaka, H.; Nakamura, K. T.; Miyasaka, T. Tetra-
hedron Lett. 1998, 39, 3713–3716; Haraguchi, K.; Takahashi, H.; Nishikawa, A.; Sasakura, E.; Tanaka, H.;
Nakamura, K. T.; Miyasaka, T. Nucleic Acids Symposium Series No. 39, Oxford University Press, 1998, 17–18.
1
t
18. Hydroxy diphenyl dithioacetal 8, (R=Ph): H NMR (CDCl3) l 0.1 (s, 6H, Me2Si), 1.1 (s, 9H, BuSi), 1.70–1.86
(ddd, 1H, 2-CH), 2.0–2.16 (ddd, 1H, 2-CH), 2.7 (solvent and concentration dependent) (d, J=5 Hz, 1H,
3-CHOH), 3.25 (dd, J=7 Hz, 9 Hz, 1H, 4-CH), 3.42 (dd, J=9 Hz, 4 Hz, 1H, 4-CH), 4.07 (m, 1H, 3-CHOH),
4.75 (dd, J=12 Hz, 3.5 Hz, 1H, 1-CH), and 7.0–7.5 (m, 10H, PhS×2). MS (m/z) 421 (M+).
19. Anhydronucleosides in general are versatile intermediates in pyrimidine nucleoside synthesis, as shown by
Mizuno, Y. The Organic Chemistry of Nucleic Acids; Elsevier/Kodansha Ltd.: Tokyo, 1986; pp. 133–143;
Townsend, L. B. Chemistry of Nucleotides and Nucleosides; Plenum Press: New York, 1988; pp. 49–67.
20. We wish to thank Dr. S. G. Rahim and N. Trivedi for unpublished observations that dithioacetal exchange with
trimethylsilylated pyrimidine bases conveniently stops at the mono-exchange stage.

10105
21. Compound 9 was fully characterised as a TLC-pure, higher Rf (10:1 petrol:ether) diastereoisomer of the
t
methanesulphonates 9b: 1H NMR l (CDCl3) 0.1 (s, 6H, Me2Si), 0.83 (s, 9H, BuSi), 0.98 (t, J=7 Hz, 3H,
5-CH2CH3), 2.1–2.4 (m, 4H, overlapped quartet 5-CH2CH3 and 2%-CH2), 2.94 (s, 3H, S(O)2CH3), 3.73 (dd, J=13
Hz, 5 Hz, 2H, 4%-CH2), 4.7–4.8 (m, 1H, 3%-CH), 6.0 (dd, J=10 Hz, 10 Hz, 1H, 1%-CH), and 7.0–7.4 (bm, 6H, SPh
and 6-CH). MS (m/z) 529 (M+). Anal (C23H36N2O6SiS2): C, H, N.
22. Compound 10 was characterised as the alcohol 10b (following de-silylation of 10a with tetra-n-butylammonium
fluoride): 1H NMR (CD3OD), major diastereomer (ꢀ60%), l 1.19 (t, J=6 Hz, 3H, 5-CH2CH3), 2.46 (q, J=6 Hz,
2H, 5-CH2CH3), 2.2–2.75 (m, 2H, 2%-CH2), 3.7–3.83 (ddd, 2H, 4%-CH2), 4.4–4.52 (m, 1H, 3%-CH), 5.67 (dd, J=9
Hz, 5 Hz, 1%-CH), 7.4–7.55 (m, 5H, PhS), and 7.81 (s, 1H, 6-CH); minor diastereomer (ꢀ40%), l 0.97 (t, J=6
Hz, 3H, 5-CH2CH3), 2.23 (q, J=6 Hz, 2H, 5-CH2CH3), 2.35–2.75 (m, 2H, 2%-CH2), 3.75–4.00 (ddd, 2H, 4%-CH2),
4.87–5.00 (m, 1H, 3%-CH), 5.74 (bs, 1H, 1%-CH) 6.80 (s, 1H, 6-CH), and 7.4–7.6 (m, 5H, PhS). Peak assignments
were enabled by appropriate spin-decoupling and 2D COSY studies. MS (m/z) 318 (M+) on diastereomeric
mixture.
23. The structure of 13 follows from detailed spin-decoupling and NOE difference spectroscopy. The lack of
enhancement between H-6 and H-1% (as compared with 10b and 10c for example), and a significantly different UV
spectrum (umax 280 nm) led to the view that N-3, instead of N-1, was incorporated into the 1,3-oxazine ring of
13, formed as only one diastereomer. A similar cyclisation via N-3 has been observed by Liotta et al. in their
work on the 3%-tether approach to b-nucleosides.16a This study also targeted a 2,3%-anhydronucleoside intermedi-
ate, but the approach is, in detail, quite different to the work described here.
.
.