The first example of a new type of acyclic, achiral nucleoside analogue: 1-[3-hydroxy-2-(hydroxymethyl)prop-1-enyl]thymine

Article abstract of DOI:10.1039/b101759o

Various preparative routes to 1-(dihydroxyalk-1-enyl)thymines, which are acyclic, achiral nucleoside analogues, have been examined, and a successful synthesis of 1-[3-hydroxy-2-(hydroxymethyl)prop-1-enyl]thymine (1, B = T) has been devised.

Full text of DOI:10.1039/b101759o

View Article Online / Journal Homepage / Table of Contents for this issue
The first example of a new type of acyclic, achiral nucleoside
analogue: 1-[3-hydroxy-2-(hydroxymethyl)prop-1-enyl]thymine
Daniel Sejer Pedersen,† Thomas Boesen, Anne B. Eldrup,‡ Benedikte Kiær, Christian Madsen,
Ulla Henriksen and Otto Dahl*
1
Department of Chemistry, University of Copenhagen, The H. C. Ørsted Institute,
Universitetsparken 5, DK-2100, Copenhagen, Denmark
Received (in Cambridge, UK) 23rd February 2001, Accepted 29th May 2001
First published as an Advance Article on the web 18th June 2001
Various preparative routes to 1-(dihydroxyalk-1-enyl)thymines, which are acyclic, achiral nucleoside analogues, have
been examined, and a successful synthesis of 1-[3-hydroxy-2-(hydroxymethyl)prop-1-enyl]thymine (1, B = T) has been
devised.
Introduction
Nucleoside analogues are of considerable interest both as anti-
viral and antitumour agents,1 and as monomer precursors for
the generation of antisense oligonucleotides, which are short
single-stranded nucleic acids that bind to a specific messenger
RNA and thereby interfere with its translation to protein.2 Well
Fig. 1 Target molecules and known analogues.
known examples of antiviral nucleoside analogues in clinical
use are the acyclic herpes virus drug acyclovir3 and the cyclic
HIV drug 3Ј-azido-2Ј,3Ј-dideoxythymidine (AZT).4 In the anti-
sense field many nucleoside analogues have been examined,
although so far only one deoxyoligonucleoside phosphoro-
thioate, containing only unmodified nucleosides, has been
approved as a drug.5 Promising nucleoside analogues for anti-
sense use which confer improved binding to RNA and high
nuclease resistance on the oligonucleotide are the acyclic PNA,6
the cyclic anhydrohexitol nucleosides,7 and the recently intro-
duced bicyclic LNA.8 We became interested in nucleoside
analogues 1 (Fig. 1), where the ribose sugar is substituted by a
simple acyclic, achiral unit. According to Dreiding models this
unit positions the nucleobase and the two hydroxy groups close
to the positions of the nucleobases and the 3Ј- and 5Ј-OH
groups in ribonucleosides in the North conformation prevailing
in A-type double helices.9 The double bond of 1 restricts the
possible conformations of the acyclic unit, and oligonucleotides
prepared from 1 might therefore be preorganised to form A-
type double helices with RNA, which usually result in increased
duplex stability.10 In addition, compounds 1 might have inter-
esting antiviral properties. To the best of our knowledge, no
compounds with the structure 1 or substituted compounds 2
are known, apart from two acetal-protected compounds 311
(Fig. 1). However, many structurally related non-sugar
analogues with one or two hydroxy groups are known, e.g. the
alk-1-enyl nucleobases 4–8, the allenes 9 and 10, the alk-2-enyl
compounds 11 and 12, the cyclopropane derivatives 13–16,
and the cyclobutane compounds 17 and 18, analogues of the
natural compound oxetanocin 19 (Fig. 2). Most of these com-
pounds have been screened for antiviral properties, but we are
only aware of antisense studies for compounds 17–19.26–28 We
Fig. 2 Known nucleoside analogues structurally related to 1.
present here our synthetic efforts to prepare compounds with
the structure 1 or 2, including the successful preparation of
1-[3-hydroxy-2-(hydroxymethyl)prop-1-enyl]thymine (1, B = T).
Results and discussion
Many routes can be envisaged for the preparation of a trisub-
stituted alkene such as 1. By analogy with methods published
† Current address: Cureon A/S, Fruebjergvej 3, DK-2100 Copenhagen,
Denmark.
‡ Current address: ISIS Pharmaceuticals, Medicinal Chemistry, 2292
Faraday Avenue, Carlsbad, CA 92008, USA.
for the preparation of 4–7,12–16 early attempts were based on the
alkylation of thymine with properly substituted alkyl halides.
However, attempted reactions of the vicinal dibromide 2029
1656
J. Chem. Soc., Perkin Trans. 1, 2001, 1656–1661
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b101759o

View Article Online
(Scheme 1) with thymine in the presence of K2CO3 in DMF,
or with bis(trimethylsilyl)thymine in CH3CN, failed to give any
of either the alkylated product 21 or its eliminated counterpart
22 under various conditions. This is perhaps unsurprising
given that 20 is a very hindered primary bromide (cf. neopentyl
bromide); however, a similar reaction was possible on a cyclo-
propyl analogue.15 Alkylation of either thymine or bis(trimethyl-
silyl)thymine using 2-chloromethyl-1,3-dichloropropene 2330
was slightly more successful and yielded small amounts of 24.
The tendency of thymine to undergo dialkylation and the ease
with which both allylic chloro atoms of 23 underwent sub-
stitution undoubtedly contributed to the low yield of this reac-
tion. In any event, this route was abandoned when all attempts
to isomerise 24 to the target compound 25 failed. Another route
to a potential precursor of 1, i.e. addition of thymine or bis-
(trimethylsilyl)thymine to methoxymethylene Meldrum’s acid
26,31 gave no substitution product 27, even under forcing con-
ditions. The low nucleophilicity of bis(trimethylsilyl)thymine
towards 26 was evident from the isolation of the acetamido-
substituted product 28 instead of 27 when bis(trimethyl-
silyl)thymine was prepared from bis(trimethylsilyl)acetamide
(BSA). Attempts to build up the thymine base from the known
urea derivative 2932 were aborted since 29 did not form the
corresponding uracil derivative 30 with 3-methoxypropenoyl
chloride. A recent method described by one of us33 to prepare
N-(alk-1-enyl)thymines from aldehydes and thymine in the
presence of trimethylsilyl trifluoromethanesulfonate failed with
the aldehyde 31,31 which self-condensed instead. Another recent
method from our group to obtain N-(alk-1-enyl)thymines or
adenines is by Horner or Horner–Wadsworth–Emmons reac-
tions.34 However, several hydroxy-protected dihydroxyacetone
derivatives, e.g. 3235 and 1-(diphenylphosphorylmethyl)thymine
3334 (or the 3-benzoyl derivative), in the presence of NaH
or LDA gave none of the desired alkene product 34 due to
self-condensation of 32 under the reaction conditions. The
high tendency of 32 to give self-aldol products has been
observed earlier.36 We thought that the Horner reaction failed
because the anion of 33 is too strong a base, and we decided to
study condensations of hydroxy-protected dihydroxyacetone
derivatives with less basic anionic thymine reagents. Since we
were unable to prepare 1-(triphenylphosphoniomethyl)thymine
we turned to the known 1-(methoxycarbonylmethyl)thymine.37
Attempted routes to 1 and 2 from 1-(methoxycarbonylmethyl)-
thymine
The anion prepared from 3-benzyloxymethyl-1-(methoxy-
carbonylmethyl)thymine 35 and LDA at Ϫ78 ЊC and protected
dihydroxyacetone 32 gave the addition product 36 in high yield
(Scheme 2). Removal of benzaldehyde with aqueous acid was
accompanied by lactonisation to give the lactone 37, which,
after removal of the benzyloxymethyl (BOM) group and
protection of the primary hydroxy group with a tert-butyl-
dimethylsilyl group, gave 38. Elimination of water to give 39
occurred partly during silica column purification of 38, and was
complete after stirring with silica in chloroform for 48 h at rt.
The compound 39 and the unprotected compound 40 were
prepared as potential precursors of 2, X = CONH2. However,
treatment of 39 with ammonia gave none of the desired 41, but
probably the furan 42. Attempts to open the lactone ring of
the dimethoxytrityl (DMT) protected analogue of 39, or the
unprotected lactone 40, with ammonia or methylamine also
Scheme 1 Attempted routes to 1.
J. Chem. Soc., Perkin Trans. 1, 2001, 1656–1661
1657

View Article Online
subsequent elimination to unwanted regioisomeric alkenes
analogous to 49 (benzaldehyde acetal). The tertiary alcohol
group of 45 was slowly mesylated by methanesulfonyl chloride
in pyridine to give 46 in 72% yield after 5 days at 5 ЊC. The
corresponding diethyl thiophosphate 47 could be prepared in
85% yield much faster, via phosphitylation of 45 with diethyl
chlorophosphite followed by oxidation with sulfur. The mesyl
derivative 46 underwent elimination with potassium tert-
butoxide in THF at rt to give a mixture of 48 and the regio-
isomers 49, which isomerised to nearly pure 48 after standing
for 2 days at rt. The thiophosphate 47 likewise gave 48 and 49,
but several unknown compounds were formed as well, which
resulted in a lower yield of 48. The last step, removal of the
benzyl protecting groups from 48, was expected to be trouble-
some because the ethers are allylic as well as benzylic. However,
deprotection with boron trichloride in dichloromethane at
Ϫ78 ЊC, following the procedure of Zemlicka,17 gave 1 (B = T)
in 42% yield after purification. That the thymine is alkylated
at N-1 was verified on 48 by substantial NOE enhance-
ments between the irradiated H-6 of thymine and both the
alkene proton at 6.71 ppm and the CH C᎐CH protons at
᎐
2
4.02 ppm.
Conclusion
Several preparative routes to the new nucleoside analogues 1
and 2 have been examined with thymine as the nucleobase.
Aldol-type couplings between a protected dihydroxyacetone
(e.g. 32) and a protected thyminylacetic ester (e.g. 35) gave
promising precursors (36–40) to 1 and 2, but this route has not
yet led to the target compounds. A route via ring-opening of a
properly substituted epoxide (e.g. 44) with the anion of thymine
was more successful and led to the first synthesis of 1 (B = T).
Work is in progress to develop this route to prepare 1 (B = A, C,
and G), and to evaluate 1 and 2 in an antisense and antiviral
context.
Scheme
2
Condensations with 3-benzyloxymethyl-1-(methoxy-
carbonylmethyl)thymine 35.
failed. Treatment of 38 with ammonia likewise failed to give a
ring-opened product, probably because elimination to 39 was
faster than attack at the lactone carbon. The new compounds
36–40 might be valuable precursors of 2 with other X groups,
e.g. after reduction of the ester group, and 36 after saponifi-
cation should be able to eliminate CO2 and OHϪ to give pro-
tected 1. Work is in progress to evaluate such routes to 1
and 2.
The epoxide route to 1
Experimental
Nucleobases or their anions have been shown to react with
epoxides to give N-(2-hydroxyalkyl)nucleobase derivatives.17,38,39
Subsequent mesylation of the 2-hydroxy group and treatment
with potassium tert-butoxide was recently shown to give some
N-(alk-1-enyl)adenine derivatives 8 in reasonable yields.17 We
decided to examine this route and found that the known
epoxide 44,40 which we prepared from the easily obtainable
alkene 43,23 and excess thymine in the presence of NaH gave a
70% yield of 45 (Scheme 3). Protected 2,2-bis(hydroxymethyl)-
epoxides other than 44 were also tried, but with less success,
due to partial removal of the protecting groups (acetyl) or
1-(Methoxycarbonylmethyl)thymine,37
benzyloxymethyl
chloride,41 2-phenyl-1,3-dioxan-5-one monohydrate (32),35
and 1,1-bis(benzyloxymethyl)ethylene (43),23 were prepared
according to literature procedures. NaH was ca. 80% in mineral
oil from Aldrich. Lithium diisopropylamide (LDA) was pre-
pared fresh from diisopropylamine in dry THF and BuLi
(1.4 M in hexane, Aldrich) at Ϫ78 ЊC. Other chemicals were
97–99% pure from Aldrich, Fluka, or Merck. Solvents were
HPLC grade from LABSCAN, of which DMF and pyridine
were dried over molecular sieves (Grace 4 Å) and THF was
freshly distilled from Na–benzophenone to a water content
below 30 ppm, measured on a Metrohm 652 KF-Coulometer.
TLC was run on Merck 5554 silica 60 aluminium sheets,
column chromatography on Merck 9385 silica 60 (0.040–0.063
mm). NMR spectra (reference tetramethylsilane for δH and
δC, external 85% H3PO4 for δP, J values are given in Hz) were
run on a Varian Mercury 300 MHz spectrometer, and FAB MS
data obtained on a JEOL HX 110/110 Mass Spectrometer with
m-nitrobenzyl alcohol (MNBA) as the matrix unless otherwise
noted.
3-Benzyloxymethyl-1-(methoxycarbonylmethyl)thymine (35)
1-(Methoxycarbonylmethyl)thymine (2.00 g, 10.1 mmol) was
dried by evaporation from dry pyridine (2 × 50 ml) and dry
DMF (100 ml) was added under N2. NaH (0.35 g, 70% in
mineral oil, 10.1 mmol) was added and the reaction mixture
stirred at rt for 20 min. Benzyloxymethyl chloride (1.60 ml,
11.5 mmol) was added and stirring continued at rt for 2.5 h. The
reaction was quenched with H2O (100 ml) and the mixture
Scheme 3 The epoxide route to 1.
1658
J. Chem. Soc., Perkin Trans. 1, 2001, 1656–1661

View Article Online
was extracted with CHCl3 (2 × 100 ml). The organic phase was
dried (MgSO4) and evaporated in vacuo to give a yellow oil that
was purified by column chromatography, eluted with EtOAc–
heptane (70 : 30 v/v). The fractions containing the product were
concentrated in vacuo to give a clear oil that was crystallized
from EtOAc–heptane to give pure 35 (2.44 g, 76%) as colourless
crystals, mp 60–62 ЊC. TLC Rf 0.39 (EtOAc–heptane 80 : 20
v/v). NMR (CDCl3): δH 7.38–7.23 (5H, m, Ph), 6.90 (1H, q,
J 1.2, T-H6), 5.51 (2H, s, NCH2O), 4.69 (2H, s, OCH2Ph), 4.44
(2H, s, CH2CO2CH3), 3.80 (3H, s, CO2CH3), 1.94 (3H, d, J 1.2,
T-CH3). δC 167.84, 163.40, 151.38, 138.78, 137.65, 128.04,
127.48, 127.42, 110.31, 71.92, 70.44, 52.57, 49.21, 12.76. FAB+
MS: 319.0 (M + H+ calc. 319.1) (Found: C, 60.4; H, 5.9; N, 8.8.
Calc. for C16H18N2O5: C, 60.4; H, 5.7; N, 8.8%).
1-[4-(tert-Butyldimethylsilyloxymethyl)-4-hydroxy-2-oxotetra-
hydrofuran-3-yl]thymine (38)
A solution of 37 (2.86 g, 7.6 mmol) in MeOH (50 ml) and
20% Pd(OH)2/C (0.5 g) was stirred under H2 at 1 atm and rt for
24 h. The solution was filtered through Celite, the Celite washed
with MeOH (100 ml), and the solvent removed in vacuo. The
residue was purified on a column, eluted with CH2Cl2–MeOH
9 : 1 v/v, to give 1-[4-hydroxy-4-(hydroxymethyl)-2-oxotetra-
hydrofuran-3-yl]thymine (0.76 g, 39%). To this intermediate
(0.76 g, 3.0 mmol) in dry DMF (5 ml) was added imidazole
(0.41 g, 6 mmol) and tert-butyldimethylsilyl chloride (0.55 g,
3.6 mmol) and the mixture was stirred for 20 h. Evaporation
in vacuo gave a yellow syrup that was purified by column
chromatography, eluted with EtOAc–heptane (6 : 4 v/v) to give
38 (Rf = 0.22), contaminated with ca. 10% of the elimination
product 39 (Rf = 0.44) (1.08 g, 75%, 29% from 37). NMR
(DMSO-d6) for 38: δH 11.46 (1H, s, NH), 7.19 (1H, s, T-H6),
5.89 (1H, s, OH), 5.7 (1H, br s, NCHC(O)O), 4.34 (2H, AB
system, ∆ = 73.4 Hz, JAB 9.2, CH2OC(O)), 3.60 (2H, AB system,
∆ = 22.1 Hz, JAB 10.3, SiOCH2), 1.76 (3H, s, T-CH3), 0.86 (9H,
s, t-Bu), 0.05 (6H, s, Si-CH3). δC 172.4, 163.9, 151.4, 140.5,
108.2, 77.9, 77.0, 74.4, 64.1, 57.5, 25.6, 18.0, 12.2. FAB+ MS:
371.1 (M + H+ calc. 371.2) (Found: C, 52.5; H, 6.8; N, 7.5.
Calc. for C16H26N2O6Si: C, 51.9; H, 7.1; N, 7.6%).
3-Benzyloxymethyl-1-[(5-hydroxy-2-phenyl-1,3-dioxan-5-yl)-
(methoxycarbonyl)methyl]thymine (36)
Compound 35 (2.00 g, 6.3 mmol) was dried by evaporation
from dry pyridine (2 × 25 ml), dry THF (50 ml) was added
under N2 and the mixture cooled to Ϫ78 ЊC. Freshly prepared
LDA (6.3 mmol in THF–hexane) was added to give a yellow
solution which was stirred at Ϫ78 ЊC for 30 min. 2-Phenyl-1,3-
dioxan-5-one monohydrate (32, 1.23 g, 6.3 mmol), dried by
evaporation from dry pyridine (3 × 15 ml) and dissolved in dry
THF (10 ml), was added dropwise and the reaction mixture
stirred at Ϫ78 ЊC for 2 h. The reaction was quenched with sat.
aq. NH4Cl (10 ml) and transferred to a separation funnel with
CHCl3 (100 ml) and H2O (100 ml), extracted with CHCl3
(2 × 50 ml), and the combined organic phases washed with H2O
(100 ml) and dried (Na2SO4). Removal of the solvent in vacuo
gave a yellow solid (2.84 g, 91%, 36, slightly contaminated
with 35). Recrystallization from EtOAc–heptane gave pure
36 (2.34 g, 75%), colourless crystals, mp 187.5–188 ЊC. TLC
Rf 0.26 (heptane–EtOAc 40 : 60 v/v). 1H NMR (CDCl3): δH 7.75
(1H, q, J 1.1, T-H6), 7.52–7.14 (10H, m, Ph), 6.19 (1H, s, OH),
6.17 (1H, s, CHPh), 5.56 (1H, s, CH CO2CH3), 5.39 (2H, s,
NCH2O), 4.55 (1H, dd, J 11.3 and 2.9, OCH2C(OH)), 4.52 (2H,
s, OCH2Ph), 3.76 (1H, d, J 11.3, OCH2C(OH)), 3.76 (3H, s,
CO2CH3), 3.75 (2H, AB of ABX system, ∆ 58.8 Hz, JAB 11.5,
JAX 2.9, OCH2C(OH)), 1.87 (3H, d, J 1.1, T-CH3). 13C NMR
(DMSO-d6): δC 168.8, 162.6, 151.5, 139.3, 137.9, 137.6, 129.0,
128.1, 128.0, 127.4, 127.2, 126.7, 108.1, 101.7, 72.2, 70.9, 70.5,
68.6, 57.8, 52.8, 13.0. FAB+ MS: 496.9 (M + H+ calc. 497.2)
(Found: C, 63.0; H, 6.0; N, 5.7. Calc. for C26H28N2O8: C, 62.9;
H, 5.7; N, 5.6%).
1-[4-(tert-Butyldimethylsilyloxymethyl)-2-oxo-2,5-dihydrofuran-
3-yl]thymine (39)
A suspension of 38 (0.37 g, 1 mmol, contaminated with ca. 10%
of 39) and silica (Merck 9385 silica 60, 30 g) in CHCl3 (150 ml)
was stirred at rt for 48 h. The suspension was filtered, the silica
washed with CH2Cl2–MeOH (9 : 1 v/v) (2 × 50 ml), and the
combined solutions evaporated in vacuo to a colourless solid
residue, which was nearly pure 39 (0.22 g, 62%). NMR (DMSO-
d6): δH 11.54 (1H, s, NH), 7.44 (1H, q, J 1.2, T-H6), 5.07 (2H, s,
CH2OC(O)), 4.66 (2H, s, SiOCH2), 1.77 (3H, d, J 1.2, T-CH3),
0.83 (9H, s, t-Bu), 0.04 (6H, s, Si-CH3). FAB+ MS: 353.1
(M + H+ calc. 353.2) (Found: C, 55.0; H, 7.1; N, 8.0. Calc. for
C16H24N2O5Si: C, 54.5; H, 6.9; N, 7.95%).
1-[4-(Hydroxymethyl)-2-oxo-2,5-dihydrofuran-3-yl]thymine (40)
To a solution of 39 (0.18 g, 0.5 mmol) in dry THF (15 ml) was
added Et3Nؒ3HF (0.40 ml, 2.5 mmol) and the mixture stirred at
rt for 3 h. The residue after evaporation in vacuo was purified on
a column, eluted with CH2Cl2–MeOH (9 : 1 v/v), to give 40
(0.11 g, 92%) as a colourless solid. NMR (DMSO-d6): δH 11.54
(1H, s, NH), 7.43 (1H, s, T-H6), 5.45 (1H, t, J 5.3, OH), 5.09
(2H, s, CH2OC(O)), 4.43 (2H, d, J 5.3, HOCH2), 1.78 (3H, s,
T-CH3). δC 169.0, 164.1, 163.5, 149.1, 139.6, 120.9, 109.5,
69.7, 56.6, 11.9. FAB+ MS (matrix DMSO): 239.2 (M + H+
calc. 239.1) (Found: C, 48.6; H, 4.1; N, 11.0. Calc. for
C10H10N2O5+0.5 H2O: C, 48.6; H, 4.5; N, 11.3%).
3-Benzyloxymethyl-1-[4-hydroxy-4-(hydroxymethyl)-2-oxotetra-
hydrofuran-3-yl]thymine (37)
A suspension of 36 (0.30 g, 0.60 mmol) in a mixture of MeOH
(20 ml) and 0.5 M H2SO4 (8 ml, 8 mmol) was heated to reflux
for 1 h to give a clear solution. The cooled reaction mixture was
transferred to a separation funnel with H2O (25 ml) and
extracted with CHCl3 (3 × 50 ml). The combined organic
phases were dried (Na2SO4) and evaporated to give a colourless
residue that was purified by column chromatography, eluted
with EtOAc–heptane (9 : 1 v/v). The fractions containing the
product were evaporated in vacuo to give pure 37 (0.18 g, 78%)
as a colourless powder, mp 63–68 ЊC. TLC Rf 0.25 (EtOAc–
heptane 9 : 1 v/v). NMR (DMSO-d6): δH 7.37–7.26 (6H, m,
Ph + T-H6), 5.92 (1H, br s, COH), 5.80 (1H, br s, CHC(O)),
5.36 (2H, s, NCH2O), 5.13 (1H, br s, CH2OH ), 4.59 (2H, s,
OCH2Ph), 4.53 (1H, d, J 9.3, C(O)OCH2), 4.23 (1H, d, J 9.3,
C(O)OCH2), 3.48 (2H, AB of ABX system, ∆ = 20.5 Hz, JAB
11.2, JAX = JBX 5.2, CH2OH), 1.83 (3H, s, T-CH3). 13C NMR
(CDCl3): δC 170.7, 163.2, 152.7, 138.9, 137.0, 128.3, 127.9,
127.7, 109.8, 78.3, 73.9, 72.3, 70.9, 62.7, 59.1, 12.9. FAB+ MS:
377.1 (M + H+ calc. 377.1) (Found: C, 57.4; H, 5.6; N, 7.1.
Calc. for C18H20N2O7: C, 57.4; H, 5.4; N, 7.4%).
2,2-Bis(benzyloxymethyl)oxirane (44)
A mixture of 1,1-bis(benzyloxymethyl)ethylene (43, 7.00 g, 26.1
mmol) and m-chloroperbenzoic acid (70%, 7.71 g, 31.3 mmol)
in CHCl3 (200 ml) was refluxed for 90 min. After cooling to rt
10% aq. Na2S2O5 (60 ml) and more CHCl3 (100 ml) were added,
and the mixture was extracted with sat. aq. NaHCO3 (4 × 100
ml) and brine (150 ml). The organic phase was dried (Na2SO4),
and the solvent removed in vacuo to give a yellow oil that was
purified by column chromatography, eluted with heptane–
EtOAc (7 : 3 v/v), to give pure 44 (4.45 g, 60%) as a colourless
oil. TLC Rf 0.35 (heptane–EtOAc 7 : 3 v/v). NMR (CDCl3):
δH 7.40–7.33 (10H, m, Ph), 4.60 (4H, AB system, ∆ = 8.4 Hz,
JAB 12.0, OCH2-oxirane), 3.73 (4H, s, PhCH2), 2.83 (2H, s,
oxirane). 13C NMR: δC 137.7, 128.2, 127.5, 73.2, 69.9, 57.5,
48.5. FAB+ MS: 285.3 (M + H+ calc. 285.1) (Found: C, 75.95;
H, 7.0. Calc. for C18H20O3: C, 76.0; H, 7.1%).
J. Chem. Soc., Perkin Trans. 1, 2001, 1656–1661
1659

View Article Online
1-[3-Benzyloxy-2-(benzyloxymethyl)-2-hydroxypropyl]thymine
(45)
10 mmol). After stirring for 20 min at 0 ЊC the solution was kept
at rt for 3 days, followed by neutralisation with 4 M aq. HCl at
0 ЊC and removal of the solvents and t-BuOH in vacuo. The
residue in CHCl3 (100 ml) was extracted with sat. aq. NaHCO3
(2 × 50 ml), brine (50 ml), and the organic phase dried
(MgSO4), followed by removal of the solvent in vacuo. Crystal-
lisation of the residue from EtOAc–hexane gave pure 48 (0.80 g,
51%) as colourless crystals, mp 93–93.5 ЊC.
To a mixture of thymine (8.86 g, 70 mmol) and NaH (70% in
oil, 0.68 g, 20 mmol) in dry DMF (250 ml) under N2 was added
a solution of 44 (4.00 g, 14.1 mmol) in dry DMF (25 ml), and
the mixture was stirred at 110 ЊC for 48 h. After cooling to rt
sat. aq. NH4Cl (50 ml) was added, followed by H2O (500 ml),
and CHCl3 (250 ml). The aqueous phase was separated and
extracted with CHCl3 (3 × 300 ml), and the combined organic
phases were washed with brine, dried (Na2SO4) and evaporated
in vacuo to give a light brown oil. Purification by column chro-
matography, eluted with CH2Cl2–MeOH (95 : 5 v/v), followed
by recrystallisation from EtOAc–heptane gave 45 (4.05 g, 70%)
as colourless crystals, mp 110–111 ЊC. TLC Rf 0.40 (CH2Cl2–
MeOH 9 : 1 v/v). NMR (CDCl3): δH 8.6 (1H, s, NH), 7.40–7.25
(10H, m, Ph), 7.11 (1H, q, J 1.2, T-H6), 4.52 (4H, s, PhCH2),
3.94 (2H, s, NCH2), 3.48 (4H, s, OCH2C(OH)), 1.84 (3H, d,
J 1.2, T-CH3). δC 163.9, 152.0, 142.1, 137.4, 128.3, 127.8, 127.6,
109.8, 74.3, 73.5, 71.7, 51.3, 12.1. FAB+ MS: 411.2 (M + H+
calc. 411.2) (Found: C, 67.3; H, 6.3; N, 6.8. Calc. for
C23H26N2O5: C, 67.3; H, 6.4; N, 6.8%).
Method B. Starting from 47 (4.50 g, 8.0 mmol) and t-BuOK
(2.24 g, 20.0 mmol) in dry THF (150 ml), the same procedure as
above gave a less pure crude product. Purification by column
chromatography, eluted with CH2Cl2–EtOAc (6 : 4 v/v),
followed by crystallisation from EtOAc–heptane, gave 48 (1.26
g, 40%) as colourless crystals, mp 92–94 ЊC. TLC Rf = 0.35
(CH2Cl2–EtOAc 6 : 4 v/v). NMR (CDCl3): δH 8.6 (1H, s, NH),
7.36–7.26 (10H, m, Ph), 7.23 (1H, q, J 1.1, T-H6), 6.71 (1H, br
s, N-CH᎐C), 4.55 and 4.50 (2 × 2H, 2 × s, PhCH ), 4.16 (2H, d,
᎐
2
J 1.2, CH C᎐CH), 4.02 (2H, s, CH C᎐CH), 1.86 (3H, d, J 1.1,
᎐
᎐
2
2
T-CH3). δC 164.0, 150.1, 140.8, 138.0, 137.5, 132.2, 128.8, 128.7,
128.3, 128.13, 128.10, 128.05, 126.2, 110.6, 73.7, 72.7, 70.1,
64.8, 12.5. FAB+ MS: 392.9 (M + H+ calc. 393.2) (Found:
C, 69.65; H, 6.3; N, 7.3. Calc. for C23H24N2O4: C, 70.4; H, 6.2;
N, 7.1%).
1-[3-Benzyloxy-2-(benzyloxymethyl)-2-(methylsulfonyloxy)-
propyl]thymine (46)
To a stirred solution of 45 (3.70 g, 9.0 mmol) in dry pyridine
(45 ml) under N2 at 0 ЊC was added dropwise methane-
sulfonyl chloride (3.5 ml, 45 mmol), and the solution kept in a
refrigerator at 5 ЊC for 5 days. Excess methanesulfonyl chloride
was hydrolysed by dropwise addition of MeOH–H2O (2 : 1 v/v,
10 ml) at 0 ЊC, followed by stirring for 1 h at rt before removal
of the solvents in vacuo. The product was extracted from the
residue with boiling EtOAc (3 × 100 ml), followed by purifi-
cation on a column, eluted with CH2Cl2–MeOH (9 : 1 v/v), to
give 46 (3.15 g, 72%). An analytical sample was obtained by
recrystallisation from EtOAc–hexane, mp 114–115 ЊC. TLC
Rf 0.44 (CH2Cl2–MeOH 9 : 1 v/v). NMR (CDCl3): δH 10 (1H,
br s, NH), 7.36–7.26 (10H, m, Ph), 7.18 (1H, q, J 1.2, T-H6),
4.53 (4H, s, PhCH2), 4.22 (2H, s, NCH2), 3.90 (4H, AB system,
∆ = 27.3 Hz, JAB 10.6, OCH2C(OMs)), 3.03 (3H, s, Ms), 1.84
(3H, d, J 1.2, T-CH3). δC 164.0, 151.4, 141.1, 136.8, 128.3,
127.9, 127.7, 110.2, 91.3, 73.6, 69.2, 50.0, 40.2, 12.0. FAB+ MS:
489.2 (M + H+ calc. 489.2) (Found: C, 58.8; H, 5.8; N, 5.9;
S, 6.5. Calc. for C24H28N2O7S: C, 59.0; H, 5.8; N, 5.7; S, 6.6%).
1-[3-Hydroxy-2-(hydroxymethyl)prop-1-enyl]thymine (1, B ؍ T)
To a stirred solution of 48 (0.39 g, 1.0 mmol) in dry CH2Cl2
(35 ml) at Ϫ78 ЊC under N2 was added dropwise BCl3 (1 M
in CH2Cl2, 6.0 ml, 6 mmol) during 5 min. Stirring was con-
tinued at Ϫ78 ЊC for 4 h, followed by dropwise addition of
MeOH–CH2Cl2 (1 : 1 v/v, 8 ml) at Ϫ78 ЊC. The cooling bath
was removed and the solvents removed in vacuo to give a
residue that was stirred for 2 h with a mixture of MeOH (15 ml)
and solid NaHCO3 (0.40 g, 6 mmol). The solids were removed
by filtration and washed with MeOH–CH2Cl2 (1 : 1 v/v, 2 ×
10 ml). The combined filtrates were concentrated in vacuo and
the residue purified by column chromatography, eluted with
CH2Cl2–MeOH (9 : 1 v/v), to give pure 1, B = T, (89 mg, 42%)
as colourless crystals, mp 166–169 ЊC. TLC Rf 0.10 (CH2Cl2–
MeOH 9 : 1 v/v). NMR (DMSO-d6): δH 11.3 (1H, s, NH), 7.41
(1H, s, T-H6), 6.41 (1H, s, N-CH᎐C), 5.01 and 4.94 (2 × 1H,
᎐
2 × t, J 5 and 5, OH), 4.08 and 3.92 (2 × 2H, 2 × d, J 5 and 5,
CH2), 1.76 (3H, s, T-CH3). δC 164.1, 150.1, 141.1, 137.8, 121.6,
108.4, 60.5, 55.8, 11.9. FAB+ MS: 213.0, FABϪ MS: 211.1
(M calc. 212.1) (Found: C, 50.8; H, 5.7; N, 12.9. Calc. for
C9H12N2O4: C, 50.9; H, 5.7; N, 13.2%).
1-[3-Benzyloxy-2-(benzyloxymethyl)-2-(diethoxythiophosphoryl-
oxy)propyl]thymine (47)
To a stirred solution of 45 (4.50 g, 11.0 mmol) in dry pyridine
(50 ml) under N2 at rt was added dropwise diethyl chlorophos-
phite (1.96 ml, 11.5 mmol). After 45 min S8 (0.38 g, 12 mmol S)
was added, and the mixture was stirred for 1.5 h, followed by
evaporation in vacuo. The residue was dissolved in CHCl3
(400 ml) and the solution extracted with sat. aq. NaHCO3 (2 ×
150 ml), brine (80 ml), dried (Na2SO4), and evaporated in vacuo.
Crystallisation of the residue from EtOAc–heptane gave pure
47 (5.25 g, 85%) as colourless crystals, mp 88.5–89.5 ЊC. NMR
(DMSO-d6): δH 11.26 (1H, s, NH), 7.32–7.26 (11H, m, Ph +
T-H6), 4.49 (4H, s, PhCH2), 4.13 (2H, s, NCH2), 4.01–3.91
(4H, m, Et), 3.81 (4H, AB system, ∆ = 19.1 Hz, JAB 10.3,
OCH2C(OP)), 1.68 (3H, s, T-CH3), 1.10 (6H, t, J 7.0, Et).
δC 164.0, 151.4, 141.9, 137.7, 128.1, 127.5, 108.0, 86.2 (d, J 10),
72.6, 69.3, 63.9 (d, J 6), 49.2, 15.5 (d, J 8), 12.0. δP 58.6. FAB+
MS: 563.4 (M + H+ calc. 563.2) (Found: C, 57.6; H, 6.35;
N, 5.1; S, 5.3. Calc. for C27H35N2O7PS: C, 57.6; H, 6.3; N, 5.0;
S, 5.7%).
Acknowledgements
The Danish Natural Science Research Council and The Danish
Technical Research Council are thanked for financial support.
References
1 Nucleosides and Nucleotides as Antitumour and Antiviral Agents,
ed. C. K. Chu and D. C. Baker, Plenum Press, New York, 1993.
2 P. H. Seeberger and M. H. Caruthers, Modified Oligonucleotides as
Antisense Therapeutics, Ch. 3 in Applied Antisense Oligonucleotide
Technology, ed. C. A. Stein and A. M. Krieg, Wiley-Liss, New York,
1998.
3 H. J. Schaeffer, L. Beauchamp, P. de Miranda, G. B. Elion,
D. J. Bauer and P. Collins, Nature, 1978, 272, 583.
4 H. Mitsuya, K. J. Weinhold, P. A. Furman, M. H. St. Clair,
S. N. Lehrman, R. C. Gallo, D. Bolognesi, D. W. Barry and
S. Broder, Proc. Natl. Acad. Sci. USA, 1985, 82, 7096.
5 A. Persidis, Nat. Biotechnol., 1999, 17, 403.
1-[3-Benzyloxy-2-(benzyloxymethyl)prop-1-enyl]thymine (48)
6 P. E. Nielsen, M. Egholm, R. Berg and O. Buchardt, Science, 1991,
254, 1497.
7 A. Van Aerschot, I. Verheggen, C. Hendrix and P. Herdewijn,
Angew. Chem., 1995, 107, 1483.
Method A. To a stirred solution of 46 (1.95 g, 4.0 mmol) in
dry THF (75 ml) under N2 at 0 ЊC was added t-BuOK (1.12 g,
1660
J. Chem. Soc., Perkin Trans. 1, 2001, 1656–1661

View Article Online
8 S. K. Singh, P. Nielsen, A. A. Koshkin and J. Wengel, Chem.
Commun., 1998, 455.
9 C. Altona and M. Sundaralingam, J. Am. Chem. Soc., 1972, 94,
8205.
10 P. Herdewijn, Liebigs Ann., 1996, 1337; E. T. Kool, Chem. Rev.,
1997, 97, 1473.
11 R. Beugelmans, A. Lechevallier, T. Frinault, T. Gharbaoui and
R. Benhida, Synlett, 1994, 513.
12 P. Ciapetti and M. Taddei, Tetrahedron, 1998, 54, 11305;
D. A. Adams, A. S. F. Boyd, R. Ferguson, D. S. Grierson and
C. Monneret, Nucleosides Nucleotides, 1998, 17, 1053; H.-J. Gi,
Y. Xiang, R. F. Schinazi and K. Zhao, J. Org. Chem., 1997, 62, 88;
J. Zhou and P. B. Shevlin, Synth. Commun., 1997, 27, 3591;
M. D’Auria and A. Vantaggi, J. Heterocycl. Chem., 1994, 31, 375.
13 M. Jokic and V. Skaric, Tetrahedron Lett., 1994, 35, 2937.
14 S. Phadtare and J. Zemlicka, Tetrahedron Lett., 1990, 31, 43.
15 Y.-L. Qiu and J. Zemlicka, Synthesis, 1998, 1447; Y.-L. Qiu,
M. B. Ksebati, R. G. Ptak, B. Y. Fan, J. M. Breitenbach, J.-S. Lin,
Y.-C. Cheng, E. R. Kern, J. C. Drach and J. Zemlicka, J. Med.
Chem., 1998, 41, 10.
25 T. Sekiyama, S. Hatsuya, Y. Tanaka, M. Uchiyama, N. Ono,
S. Iwayama, M. Oikawa, K. Suzuki, M. Okunishi and T. Tsuji,
J. Med. Chem., 1998, 41, 1284.
26 J.-M. Henlin, K. Jaekel, P. Moser, H. Rink, E. Spieser and
G. Baschang, Angew. Chem., Int. Ed. Engl., 1992, 31, 482.
27 N. Katagiri, Y. Morishita, I. Oosawa and M. Yamaguchi,
Tetrahedron Lett., 1999, 40, 6835; S. Honzava, S. Ohwada,
Y. Morishita, K. Sato, N. Katagiri and M. Yamaguchi, Tetrahedron,
2000, 56, 2615.
28 S. Niitsuma, Y. Ichikawa and T. Takita, Stud. Nat. Prod. Chem.,
1992, 10, 585; A. Kakefuda, A. Masuda, Y. Ueno, A. Ono and
A. Matsuda, Tetrahedron, 1996, 52, 2863.
29 G. Riehs, E. Urban and H. Voellenkle, Tetrahedron, 1996, 52, 8725.
30 N. Stuhr-Hansen, L. Henriksen and J. T. Kodra, Synth. Commun.,
1996, 26, 3345.
31 H. A. Bates, J. Farina and M. Tong, J. Org. Chem., 1986, 51, 2637.
32 R. Cassis, R. Tapia and J. A. Valderrama, Synth. Commun., 1984,
14, 961.
33 U. Henriksen, Nucleosides Nucleotides, 2000, 19, 1093.
34 T. Boesen, C. Madsen, U. Henriksen and O. Dahl, J. Chem. Soc.,
Perkin Trans. 1, 2000, 2015.
35 D. C. Forbes, D. G. Ene and M. P. Doyle, Synthesis, 1998, 879;
P. H. J. Carlsen, K. Sørbye, T. Ulven and K. Aasbø, Acta Chem.
Scand., 1996, 50, 185.
16 C. Cheng, T. Shimo, K. Somekawa and M. Kawaminami,
Tetrahedron Lett., 1997, 38, 9005.
17 H.-P. Guan, M. B. Ksebati, E. R. Kern and J. Zemlicka, J. Org.
Chem., 2000, 65, 5177.
18 J. Zemlicka, in Nucleosides and Nucleotides as Antitumour and
Antiviral Agents, ed. C. K. Chu and D. C. Baker, Plenum Press,
New York, 1993, pp. 73–100.
19 J. Zemlicka, Nucleosides Nucleotides, 1997, 16, 1003.
20 W. T. Ashton, L. C. Meurer, C. L. Cantone, A. K. Field, J. Hannah,
J. D. Karkas, R. Liou, G. F. Patel, H. C. Perry, A. F. Wagner,
E. Walton and R. L. Tolman, J. Med. Chem., 1988, 31, 2304.
21 D. R. Haines, C. K. H. Tseng and V. E. Marquez, J. Med. Chem.,
1987, 30, 943.
22 M. G. Lee, J. F. Du, M. W. Chun and C. K. Chu, J. Org. Chem.,
1997, 62, 1991.
23 T. Izawa, S. Nishiyama, S. Yamamura, K. Kato and T. Takita,
J. Chem. Soc., Perkin Trans. 1, 1992, 2519.
36 M. Majewski, D. M. Gleave and P. Novak, Can. J. Chem., 1995, 73,
1616.
37 K. L. Dueholm, M. Egholm, C. Behrens, L. Christensen, H. F.
Hansen, T. Vulpius, K. H. Petersen, R. H. Berg, P. E. Nielsen and
O. Buchardt, J. Org. Chem., 1994, 59, 5767.
38 J. C. Martin, D. F. Smee and J. P. H. Verheyden, J. Org. Chem., 1985,
50, 755.
39 B. Allart, R. Busson, J. Rozenski, A. Van Aerschot and
P. Herdewijn, Tetrahedron, 1999, 55, 6527; B. Allart, K. Khan,
H. Rosemeyer, G. Schepers, C. Hendrix, K. Rothenbacher, F. Seela,
A. Van Aerschot and P. Herdewijn, Chem. Eur. J., 1999, 5, 2424.
40 K. K. Ogilvie, N. Nguyen-Ba and R. G. Hamilton, Can. J. Chem.,
1984, 62, 1622.
24 N. Gauvry and F. Huet, Tetrahedron, 1999, 55, 1321; N. Katagiri,
41 D. S. Connor, G. W. Klein, G. N. Taylor, R. K. Boeckman, Jr. and
J. B. Medwid, Org. Synth., 1973, Coll. Vol. VI, 101.
H. Sato and C. Kaneko, Chem. Pharm. Bull., 1990, 38, 3184.
J. Chem. Soc., Perkin Trans. 1, 2001, 1656–1661
1661