Nucleoside H-phosphonates. Part 19: Efficient entry to novel nucleotide analogues with 2-pyridyl- and 4-pyridylphosphonothioate internucleotide linkages

Article abstract of DOI:10.1016/j.tet.2003.11.010

Synthetic and ³¹P NMR spectroscopy studies resulted in the development of efficient protocols for the stereospecific synthesis of a novel type of nucleotide analogues, 2-pyridyl- and 4-pyridylphosphonothioates. The underlying chemistry involves formation of the P-C bond via a base-promoted reaction of suitably protected dithymidine H-phosphonothioates with N-methoxypyridinium tosylate in acetonitrile, or with trityl chloride in pyridine, to produce high yields of nucleotide analogues with a 2-pyridyl- or 4-pyridyl moiety directly bound to the phosphorus centre.

Full text of DOI:10.1016/j.tet.2003.11.010

Tetrahedron 60 (2004) 389–395
Nucleoside H-phosphonates. Part 19: Efficient entry to novel
nucleotide analogues with 2-pyridyl- and
4-pyridylphosphonothioate internucleotide linkages
*
Tommy Johansson and Jacek Stawinski
Arrhenius Laboratory, Department of Organic Chemistry, Stockholm University, S-106 91 Stockholm, Sweden
Received 14 August 2003; revised 16 October 2003; accepted 6 November 2003
Abstract—Synthetic and 31P NMR spectroscopy studies resulted in the development of efficient protocols for the stereospecific synthesis of
a novel type of nucleotide analogues, 2-pyridyl- and 4-pyridylphosphonothioates. The underlying chemistry involves formation of the P–C
bond via a base-promoted reaction of suitably protected dithymidine H-phosphonothioates with N-methoxypyridinium tosylate in
acetonitrile, or with trityl chloride in pyridine, to produce high yields of nucleotide analogues with a 2-pyridyl- or 4-pyridyl moiety directly
bound to the phosphorus centre.
q 2003 Elsevier Ltd. All rights reserved.
1. Introduction
and 2-pyridylphosphonate16 analogues. As an extension of
thesestudies, weattemptedtodevelopasyntheticmethodfora
new nucleotide analogue in which the phosphoryl oxygen
atom of the pyridylphosphonate moiety was replaced by
sulfur. Since a 2-pyridylphosphonate moiety may act as a
bidentate chelating ligand17 for transition metals, the replace-
ment of oxygen by sulfur may significantly change its affinity
for particular types of metal cations and thus be of importance
in designing new artificial nucleases,18 and reporter groups for
investigation of electron transfer (ET) phenomena in nucleic
acids.19
Pyridylphosphonate derivatives represent an important class
of organophosphorus compounds with a wide range of
technical applications, for example, corrosion inhibitors,1
sensitisers for photovoltaic cells,2 dispersing and emulsify-
ing agents,3 lubricant additives,4 etc. Apart from these, they
also show diverse biological activity and are known as
potent insecticides,5 fungicides,6 and herbicides.7 Recently,
2-pyridylphosphonates emerged as broad spectrum drugs
with anti-proliferating and anti-PAF activity,8 as inhibitors
of fructose-1,6-bisphosphates9 (and thus of potential use in
diabetics therapy), and as lucitropic agents for the treatment
of cardiovascular diseases.10 These can probably be traced
back to a structural similarity of 2-pyridylphosphonates to
a-aminophosphonates11 (P-analogues of a-amino acids),
that are potent inhibitors of proteases12 and exhibit
pronounced antineoplastic activity.13
In contrast to pyridylphosphonates, their thiophosphonate
counterparts are rare compounds. Also their biological
properties are largely unexplored, apart from a few reports
in the patent literature where these compounds have been
advocated as activators and enhancers in certain compo-
sitions of herbicides and insecticides.20 These most limited
applications are probably due to the lack of convenient
methods for the preparation of pyridylphosphonothioates.
The only method of synthetic value affords the target
thiophosphonates in mediocre yields,21 and requires lengthy
thiation of their oxygen congeners with P2S5 in toluene
under reflux. Recently, an approach based on the metalation-
induced rearrangement of 3-pyridyl phosphorothioates was
reported,22 but this produced mixtures of 2- and 4-pyridyl-
phosphonothioates (3:1) in rather low yields (ca. 30%).
This diverse array of biological activity of simple dialkyl
pyridylphosphonates recently prompted us to explore the
possibility of incorporating this functionality into nucleic
acid fragments, kindled with the hope that it may confer
novel properties that could be useful when designing new
antisense and antigene therapeutics. For this purpose, we
have developed efficient methods for the conversion of
dinucleoside H-phosphonates into 4-pyridyl-,14 3-pyridyl-,15
In this paper, 31P NMR spectroscopy investigations
and synthetic studies on the formation of 2-pyridyl- and
4-pyridylphosphonothioates from the corresponding
H-phosphonothioate diesters, are described.
Keywords: H-Phosphonates; H-Phosphonothioates; Pyridylphosphono-
thioates.
*
Corresponding author. Tel.: þ46-8-16-24-85; fax: þ46-8-15-49-08;
e-mail address: js@organ.su.se
0040–4020/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2003.11.010

390
T. Johansson, J. Stawinski / Tetrahedron 60 (2004) 389–395
3
2. Results and discussion
1JPH¼675, 674 Hz, JPH¼9.8 Hz, dq) was treated in
pyridine with Tr-Cl (1.2 equiv.) in the presence of TEA
(2.4 equiv.), the reaction was still rapid (,10 min) and
produced 1,4-dihydropyridine derivatives 2 (dP¼93.9,
94.5 ppm) practically quantitatively (Scheme 1). No
dinucleoside phosphorothioate diesters or any other side
products could be detected by 31P NMR spectroscopy. Here,
however, an unexpected problem arose: 1,4-dihydro-
pyridylphosphonothioate intermediate 2 turned out to be
rather stable and, in contrast to the 1,4-dihydropyridyl-
phosphonate, that collapsed spontaneously to the corre-
sponding pyridylphosphonate derivative,14 it underwent
only slow conversion into the product, 4-pyridylphos-
phonothioate diester 3 (dP¼83.0, 84.3 ppm; ca. 20%
conversion after a few hours).
As viable means to synthesize 4-pyridyl- and 2-pyridylphos-
phonothioates, we considered 1,8-diazabicylo[5.4.0]undec-
7-ene (DBU)-promoted reactions of dinucleoside H-phos-
phonothioates with the pyridine-trityl chloride reagent
system and with N-methoxypyridinium salts, respectively,
analogously to the recently developed methods for the
preparation of pyridylphosphonate derivatives.14 Despite
apparent similarities between H-phosphonate and H-phos-
phonothioate diesters, the latter ones are more prone to side
reactions and, under basic conditions, undergo a ligand
exchange process23 that may scramble substituents around
the phosphorus center.
We assumed that synthesis of 4-pyridylphosphonothioates 3
could be effected via in situ generation of N-tritylpyridinium
cation (from e.g., pyridine and trityl chloride), followed by
an attack of a phosphorus nucleophile (e.g., H-phos-
phonothioate) at the less sterically hindered C-4 carbon of
the pyridine ring (Scheme 1). Indeed, the reaction of
dinucleoside H-phosphonothioate 1 in pyridine with trityl
chloride (Tr-Cl, 1.2 equiv.) in the presence of DBU
(2.4 equiv.) was rapid (ca. 10 min, 31P NMR spectroscopy)
but, apart from the expected 1,4-dihydropyridine inter-
mediate 2 (70%, 2 signals at ca. 94 ppm) and the des0ired
product 3 (10%, 2 signals at ca. 84 ppm), also afforded 5 -O-
dimethoxytritylthymidin-30-yl 30-O-dimethoxytritylthymi-
din-50-yl phosphorothioate diesters (20%, signals at 56.5
and 56.5 ppm), apparently due to oxidation of the starting
material 1. Use of more DBU (8 equiv.) speeded up the
reaction and reduced the amount of the oxidation product to
ca. 7%, but simultaneously another side products appeared
(ca. 6%; signals at ca. 87 and 88 ppm). We hypothesized
that formation of the phosphorothioate diesters under the
reaction conditions was probably due to air oxidation of a
highly reactive tervalent phosphite form generated from
H-phosphonothioate 1 in the presence of a strong base
(DBU). To remedy this problem, we replaced DBU by
triethylamine (TEA), assuming that in the presence of a
weaker base a concentration of a phosphite form generated
from H-phosphonothioate 1 should be lower, and thus
formation of the oxidation product should be suppressed.
To facilitate rearomatization of the 1,4-dihydropyridine
intermediate and to convert it into 4-pyridylphosphono-
thioate 3, we elaborated in situ oxidation of 2 with iodine.
Although phosphorothioate diesters undergo rapid desulfur-
ization in the presence of iodine,24 uncharged phosphorus
compounds bearing the thiophosphoryl function are resist-
ant towards desulfurization (e.g., phosphorothioate tri-
esters,24 H-phosphonothioate diesters 125) and thus these
reaction conditions should not affect the integrity of the
produced pyridylphosphonothioates 3. To check the efficacy
of this approach, the reaction mixture containing 1,4-
dihydropyridine intermediate 2 (obtained as described
above) was treated with iodine (2 equiv.) in pyridine. 31P
NMR spectroscopy revealed rapid formation of the desired
product 3 (ca. 50% after 10 min), however, with time,
signals from dinucleoside phosphorothioate diesters (18%;
2 signals at ca. 56 ppm)26 also appeared in the 31P NMR
spectrum. In this instance, the phosphorothioate diesters
could be formed due to a possible reversibility of the
dihydropyridine intermediate 2 formation14 that can gen-
erate small equilibrium amounts of the starting material 1.
This, in the presence of iodine and adventitious water could
produce, via the intermediacy of phosphorothioiodidates,
the corresponding dinucleoside phosphorothioates.
We thought that this problem could be overcome by using
more iodine for the reaction. The higher iodine concen-
tration should increase the rate of rearomatization of 2 and
thus lessen the problem of reversibility of the 1,4-
dihydropyridine intermediate formation. It was rewarding
Indeed, when dinucleoside H-phosphonothioate 1 (a
diastereomeric mixture, 1:1; dP¼69.9 and 71.6 ppm,
Scheme 1.

T. Johansson, J. Stawinski / Tetrahedron 60 (2004) 389–395
391
to see that addition of 4 equiv. of iodine to the reaction
mixture containing 1,4-dihydrointermediate 2, afforded,
after 15 min the desired 4-pyridylphosphonothioate diester
3 exclusively. In a preparative run, using the developed
reaction conditions, 4-pyridylphsophonothioates 3 were
obtained in ca. 80% yield, after silica gel column
chromatography (see the Section 4).
phonothioate 1, severely decreases the formation of the
desired 2-pyridylphosphonothioate derivatives, apparently
due to the known instability of N-alkoxypyridinium salts
under basic conditions.29
To remedy these problems, we first wanted to identify
structures of the side products formed. On the basis of their
chemical shift values (that were close to those of dinucleo-
side pyridylphosphonothioates 5), multiplicity of the signals
in the 31P NMR spectra, and by considering viable reaction
pathways, we assumed that the observed side products were,
most likely, isomeric nucleoside methyl 2-pyridyl-
phosphonothioates.30 This was consistent with a putative
decomposition pathway of N-methoxy-1,2-dihydropyridine
intermediate 4 (Scheme 2) that along with 2-pyridyl-
phophonothioates 5 should also generate a methoxide
anion. Since, H-phosphonothioate diesters are susceptible
to nucleophilic substitution at the phosphorus center,23 this
could cause a partial replacement of the 50- or 30-nucleosidic
unit in H-phosphonothioate 1 by the methoxide anion, and
ultimately lead to the formation of the observed side
products.
The stereochemical course of the reaction sequence as in
Scheme 1 was investigated by performing this transform-
ation on the separate diastereomers of dithymidine H-phos-
phonothioate 1.27 It was found that H-phosphonothioate
diester 1a (RP diastereomer27,28 resonating at higher field in
the 31P NMR spectrum) afforded pyridylphosphonothioate
3a (the diastereomer resonating at higher field) with the
intermediacy of 1,4-dihydropyridylphosphonothioate 2a,
while the diastereomer 1b (SP diastereomer,27,28 resonating
at lower field in the 31P NMR spectrum) gave pyridyl-
phosphonothioate 3b (resonating at lower field), with the
intermediacy of 2b. Thus, the transformation was found to
be stereospecific and, assuming the reaction pathway shown
in Scheme 1, it most likely proceeded with an overall
retention of configuration. On this basis we tentatively
assigned the configuration at the phosphorus center in
pyridylphosphonothioate 3a as RP, and that in 3b dia-
stereomer, as SP.
Various approaches were tried to suppress the transesteri-
fication phenomena of 1 by the generated methoxide anion.
The most successful one consisted of adding DBU before
N-methoxypyridinium p-toluenesulfonate and decreasing
the amount of the latter one to 1.2 equiv.31 Under these
conditions, the formation of methyl nucleoside pyridyl-
phosphonothioates was completely eliminated, and 2-pyr-
idylphosphonothioates 5 were formed as major products
(95%, 31P NMR spectroscopy). No further attempts were
made to eliminate the formation of 4-pyridylphos-
phonothioates 3 (ca. 5%) as these were easily removed
during silica gel chromatography.
As to the synthesis of dinucleoside 2-pyridylphosphono-
thioates 5, a protocol consisting of a reaction of H-phos-
phonothioate diesters
1
in acetonitrile with
N-methoxypyridinium p-toluenesulfonate (2 equiv.) in the
presence of DBU (4 equiv.), was designed (Scheme 2). The
reaction was rapid (,5 min), but along with the desired
2-pyridylphosphonothioates 5 (dP¼78.2 and 79.3 ppm; ca.
50%), 31P NMR spectroscopy revealed the presence in the
reaction mixture of the isomeric 4-pyridylphosphonates 3
(dP¼83.0 and 84.3 ppm; ca. 15%) and also other side
products resonating close to 2-pyridylphosphonatates 5 (two
signals at ca. 78 ppm and two signals a ca. 80 ppm; total,
25%). When the order of addition of N-methoxypyridinium
p-toluenesulfonate and DBU was changed that is DBU was
added before the pyridinium salt, the reaction showed
higher chemoselectivity (65% of 2-pyridylphosphono-
thioates 5 and 5% of 4-pyridyl isomers 3), however, the
amount of the unidentified side products increased (ca.
30%). Premixing of N-methoxypyridinium p-toluene-
sulfonate with DBU, followed by the addition of H-phos-
A mechanism for the investigated reaction (Scheme 2) is
probably similar to that of 4-pyridylphosphonothioates 3
formation, and involves intermediacy of the corresponding
1,2-dihydropyridine derivatives 4. However, in contrast to
1,4-dihydropyridylphosphonate 2, 1,2-dihydropyridine
intermediate 4 could not be detected by 31P NMR
spectroscopy, probably due to its high lability under basic
conditions. Since intermediate 4 spontaneously collapsed to
2-pyridylphosphonothioate 5 with expulsion of a methoxide
group, this reaction did not require a separate rearomatiza-
tion step. When carried out on a preparative scale under the
Scheme 2.

392
T. Johansson, J. Stawinski / Tetrahedron 60 (2004) 389–395
optimised conditions, the reaction of H-phosphonothioate 1
with N-methoxypyridinium salt in the presence of DBU
afforded dinucleoside 2-pyridylphosphonothioates 5 in ca.
80% yield, after silica gel column chromatography.
The methods make use of easily available starting materials,
H-phosphonothioate diesters, and afford the target pyridyl-
phosphonothioates in high yields under mild reaction
conditions. The underlying chemical reactions are stereo-
specific and can be extended to the preparation of other
types of biologically important phosphorus compounds
bearing these types of modifications.
In the context of regioselectivity of the above reaction, that
is formation of 2-pyridyl- versus 4-pyridylphosphonothioate
derivatives (vide supra), an interesting observation was
made. In all instances, when the reactions of H-phos-
phonothioates 1 with N-methoxypyridinium p-toluene-
sulfonate were promoted by DBU, the major products
formed were 2-pyridyl derivatives 5. The other positional
isomer, 4-pyridylphosphonothioates 3, was detected only in
small amounts (5–15%, vide supra) in these reaction
mixtures. By contrast, use of triethylamine instead of
DBU, afforded 4-pyridylphosphonothioates 3 as the major
products (80%; 31P NMR spectroscopy).32 Although the
precise factors that produce the strikingly different results in
the two reactions are not known, it seems that kinetic versus
thermodynamic control may be responsible for this
phenomenon. Since, the base participates, most likely, in
the collapse of N-methoxy-1,2-dihydropyridine derivative
4, this process should be fast in the presence DBU, and thus
formation of the kinetic products,33 2-pyridylphos-
phonothioate 5, should be favored. In the presence of
triethylamine, however, the collapse of 1,2-dihydropyridine
derivative 4, is expected to be slower, and thus, this initially
formed intermediate may isomerise to the thermodynami-
cally more stable one, the 1,4-dihydropyridyl derivative,33
from which elimination of methoxide anion can occur.
This ultimately will lead to the formation of isomeric
4-pyridylphosphonothioates 3.
4. Experimental
4.1. Material and methods
1H and 31P NMR spectra were recorded on a Varian Unity
400 BB VT spectrometer. The 31P NMR spectroscopy
experiments were carried out at 25 8C in 5 mm tubes using
0.1 M concentrations of phosphorus-containing compounds
in appropriate solvents (0.6 mL), and the spectra were
referenced to 2% H3PO4 in D2O (external standard). TLC
analyses were carried out on Merck silica gel 60 F254
precoated plates using the following solvent systems: (A)
CH3Cl/CH3OH 8:2 (v/v); (B) toluene/ethyl acetate 8:2
(v/v). Pyridine (LabScan Ltd.) and anhydrous acetonitrile
˚
(LabScan Ltd.) were stored over molecular sieves 4 A. 1,8-
Diazabicylo[5.4.0]undec-7-ene (DBU) and triethylamine
(from Aldrich) were freshly distilled. The starting materials
for the synthesis, dinucleoside H-phosphonothioates 1 were
obtained according to published procedure.34
The assignments of signals in the 31P NMR spectra to
particular products or intermediates were carried out on the
basis of their chemical shifts, multiplicity of the signals in
1H-coupled and 1H-decoupled spectra, by spiking the
reaction mixtures with appropriate species and, if possible,
by isolation of a compound in question from reaction
mixtures. The assignment of proton and carbon resonances
of 3 and 5 was carried out on the basis of known or expected
The stereochemical course of the formation of 2-pyridyl-
5 (Scheme 2) was elucidated
phosphonothioates
analogously to that of 4-pyridylphosphonothioates using
31P NMR spectroscopy. The exclusive formation of
2-pyridylphosphonothioate 5a (‘fast’ diastereomer, resonat-
ing at higher field in the 31P NMR spectrum) from
H-phosphonothioate 1a (RP diastereomer), and 5b (‘slow’
diastereomer, resonating at lower field in the 31P NMR
spectrum) from H-phosphonothioate 1b (SP diastereomer),
established this reaction as stereospecific. Assuming a
mechanism as proposed in Scheme 2, the reaction in this
instance also most likely occurs with overall retention of
configuration at the phosphorus center and the produced
pyridylphosphonothioates 5a and 5b, should have RP and SP
configurations, respectively.
1
1
chemical shifts in conjunction with H–1H, H–13C, and
DEPT correlated NMR spectroscopy. Subscripts ‘a’ and ‘b’
1
in the H NMR spectroscopy0 data refer to the protons in
nucleosid-30-yl and nuclosid-5 -yl units, respectively.
4.2. General procedure for the preparation of 4-
pyridylphosphonothioates 3a and 3b
The separate diastereomers of 50-O-dime0thoxytrityl-
thymidin-30-yl 30-O-dimethoxytritylthymidin-5 -yl H-phos-
phonothioate 1 (0.26 mmol) were rendered anhydrous by
repeated evaporation of added acetonitrile (3£10 mL), and
the residue was treated in pyridine (5 mL) with trityl
chloride (1.2 equiv.) and triethylamine (2.4 equiv.). When
the starting material 1 disappeared (,10 min, TLC
analysis), iodine was added (4 equiv.), and after 15 min
the reaction mixture was concentrated, partitioned between
a solution of aq. Na2S2O3/NaHCO3 (3:1, v/v; 20 mL) and
CH2Cl2 (20 mL), and the aqueous phase was washed with
CH2Cl2 (2£20 mL). The organic layer was dried over
anhydrous Na2SO4, concentrated and the residue was
purified by silica gel column chromatography using a
stepwise gradient of methanol (1–2%) in toluene/ethyl
acetate (1:1, v/v) containing 0.01% TEA. Purity of the
isolated compounds .98% (1H NMR spectroscopy).
To facilitate spectral characterization of the produced 4- and
2-pyridylphosphonothioate diesters, compounds 3a, 3b, 5a
and 5b, were subjected to detritylation with 80% aqueous
acetic acid. The unprotected compounds, 3c, 3d, 5c and 5d,
respectively, showed the expected pattern of signals in the
1H NMR spectra, characteristic for the respective isomeric
4- and 2-pyridylphosphonothioate derivatives.
3. Conclusions
We have developed simple and efficient methods for the
preparation of new types of nucleotide analogues bearing 2-
and 4-pyridylphosphonothioate internucleotide linkages.

T. Johansson, J. Stawinski / Tetrahedron 60 (2004) 389–395
393
4.2.1. 50-O-Dimethoxytritylthymidin-30-yl 30-O-
dimethoxytritylthymidin-50-yl pyridyl-4-phosphonothio-
ate 3a (from faster moving H-phosphonate diastereomer
1a). White solid (0.263 mg), yield 82%. HRMS [MþH]þ
found: 1227.4069; C67H66N5O14PS requires: 1227.4065.
J¼8.4 Hz, Cb40), 78.90 (d, J¼4.6 Hz, Ca30), 74.03 (Cb30),
66.85 (Cb50), 63.21 (Ca50), 55.39 and 55.36 (4£ CH3O),
39.15 (2£C20), 12.60 (Cb5–CH3), 11.87 (Ca5–CH3).
4.2.3. Thymidin-30-yl thymidin-50-yl pyridyl-4-phos-
phonothioates 3c and 3d. The diastereomers of fully
protected 4-pyridylphosphonothioates 3a and 3b
(0.130 mmol) were dissolved separately in 80% acetic
acid (aq) (15 mL) and were left while stirring for 4 h. Water
(15 mL) was then added and the mixture was washed with
diethyl ether (3£30 mL). The aqueous layer was separated,
evaporated to dryness and the product was freeze-dried from
benzene-methanol (4:1, v/v).
1H NMR. dH (in ppm, CDCl3) 8.97 and 8.96 (2£s, 2H,
2£NH), 8.71 (m, 2H, pyr-H2, pyr-H6), 7.52–7.57 (m, 3H,
Ha6, pyr-H3, pyr-H5), 7.44–7.14 (m, 18H, ArH), 7.09 (s,
1H, Hb6), 6.85–6.82 (m, 8H, ArH ortho to OMe), 6.36 (m,
1H, Ha10), 6.23 (m, 1H, Hb10), 5.46 (m, 1H, Ha30), 4.24 (d,
J¼6.2 Hz, 1H, Hb30), 3.95 (m, 2H, 2£H40), 3.93–3.70 (m,
2H, Hb50), 3.80, 3.79, 3.77 (3£s, 12H, 4£CH3O), 3.35–3.23
(m, 2H, Ha50), 2.49–2.30 (m, 2H, Ha20), 2.03–1.98, 1.77–
1.72 (2£m, 2H, Hb20), 1.77 (s, 3H, Cb5–CH3), 1.47 (s, 3H,
Ca5–CH3).
Compound 3c (from faster moving diastereomer 3a). White
solid (0.069 g), 85%. Anal. calcd for C25H30N5O10PS: C,
48.15; H, 4.85; N, 11.23. Found: C, 47.95; H 4.95; N 11.00.
31P NMR. dP (CDCl3) 83.04 ppm.
1H NMR. dH (in ppm, CD3OD), 8.78 (m, 2H, pyr-H2, pyr-
H6), 7.89 (m, 2H, pyr-H3, pyr-H5), 7.81 (d, J¼1.3 Hz, 1H,
H6), 7.52 (d, J¼1.1 Hz, 1H, H6), 6.34 (m, 1H, Ha10), 6.27 (t,
J¼6.7 Hz, 1H, Hb10), 5.42 (m, 1H, Ha30), 4.46–4.42 (m, 1H,
Hb30), 4.46–4.33 (m, 1H, Hb50), 4.15–4.10 (m, 2H, 2£H40),
3.78–3.69 (m, 2H, Ha50), 2.63–2.58, 2.43–2.36 (2£m, 2H,
Ha20), 2.33–2.28 (m, 2H, Hb20), 1.90 (d, J¼1.1 Hz, 3H, C5–
CH3), 1.83 (d, J¼1.1 Hz, 3H, C5–CH3).
13C NMR. dC (in ppm, CDCl3) 163.87 and 163.84 (2£C4),
158.95, 158.93 and 158.91 (4C of DMT), 150.57, 150.44,
150.39, 150.33 (pyr-C2, pyr-C6, 2£C2), 144.91 and 144.19
(2C of DMT), 141.15 (d, J¼151 Hz, pyr-C4), 136.02 (2C of
DMT), 135.78 (Cb6), 135.21 (Ca6), 135.09 and 135.05 (2C
of DMT), 130.33, 130.27, 130.13, 128.22, 128.19, 128.12
(16C of DMT), 127.35 (2C of DMT), 123.88 (d, J¼9.9 Hz,
pyr-C3, pyr-C5), 113.58 and 113.48 (8C of DMT), 111.88
and 111.27 (2£ C5), 87.59 and 87.34 (2£C DMT), 86.51
(Cb10), 84.57 and 84.51 (Ca40, Ca10), 84.19 (d, J¼8.4 Hz,
Cb40), 78.40 (d, J¼4.6 Hz, Ca30), 74.30 (Cb30), 66.78
(d, J¼6.1 Hz, Cb50), 63.20 (Ca50), 55.38 (4£CH3O),
39.31 (Ca20), 39.08 (Cb20), 12.48 (Cb5–CH3), 11.91
(Ca5–CH3).
31P NMR. dP (CD3OD) 82.40 ppm.
Compound 3d (from slower moving diastereomer 3b).
White solid (0.065 g), 80%. Anal. calcd for
C25H30N5O10PS: C, 48.15; H, 4.85; N, 11.23. Found: C,
47.99; H 4.98; N 11.06.
4.2.2. 50-O-Dimethoxytritylthymidin-30-yl 30-O-
dimethoxytritylthymidin-50-yl pyridyl-4-phosphonothio-
ate 3b (from slower moving H-phosphonate diastereo-
mer 1b). White solid (0.245 g), yield 77%. HRMS [MþH]þ
found: 1227.4071; C67H66N5O14PS requires: 1227.4065.
1H NMR. dH (in ppm, CD3OD), 8.76 (m, 2H, pyr-H2, pyr-
H6), 7.89 (m, 2H, pyr-H3, pyr-H5), 7.80 (d, J¼0.9 Hz, 1H,
H6), 7.42 (d, J¼1.3 Hz, 1H, H6), 6.30 (m, 1H, Ha10), 6.22 (t,
J¼6.9 Hz, 1H, Hb10), 5.40 (m, 1H, Ha30), 4.42–4.36 (m, 3H,
Hb30, Hb50), 4.32 (m, 1H, Ha40), 4.10 (m, 1H, Hb40), 3.84 (m,
2H, Ha50), 2.42–2.29 (2£m, 2H, Ha20), 2.31–2.20 (m, 2H,
Hb20), 1.89 (d, J¼1.1 Hz, 3H, C5–CH3), 1.83 (d, J¼1.1 Hz,
3H, C5–CH3).
1H NMR. dH (in ppm, CDCl3) 8.77–8.74 (m, 3H, pyr-H2,
pyr-H6, NH), 8.58 (s, 1H, NH), 7.58–7.52 (m, 3H, Ha6, pyr-
H3, pyr-H5), 7.42–7.14 (m, 18H, ArH), 6.93 (s, 1H, Hb6),
6.86–6.79 (m, 8H, ArH ortho to OMe), 6.40 (m, 1H, Ha10),
6.19 (m, 1H, Hb10), 5.40 (m, 1H, Ha30), 4.22 (s, 1H, Ha40),
4.11 (d, J¼6.6 Hz, 1H, Hb30), 3.95 (s, 1H, Hb40), 79, 3.76,
3.75 (3£s, 12H, 4£CH3O), 3.78–3.70 (m, 2H, Hb50), 3.48–
3.39 (m, 2H, Ha50), 2.36–2.32 (m, 2H, Ha20), 1.84–1.76,
1.46–1.38 (2£m, 2H, Hb20), 1.76 (s, 3H, Cb5–CH3), 1.45 (s,
3H, Ca5–CH3).
31P NMR. dP (CD3OD) 83.22 ppm.
4.3. General procedure for the preparation of 2-pyridyl-
phosphonothioates 5a and 5b
To a solution of separate diastereomers of 50-O-dimetho0xy-
tritylthymidin-30-yl 30-O-dimethoxytritylthymidin-5 -yl
H-phosphonothioate 1 (0.26 mmol) in acetonitrile (10 mL)
DBU (4 equiv.) and N-methoxypyridinium tosylate
(1.2 equiv.) were added. Caution: the reagents have to be
added swiftly and in the above mentioned order to ensure
the efficient formation of 2-pyridylphosphonothioate 5.
After 5 min (31P NMR) the reaction mixture was concen-
trated, partitioned between 10% aq. NaHCO3 (20 mL) and
CH2Cl2 (20 mL), and the aqueous layer was extracted with
CH2Cl2 (2£20 mL). The organic layer was dried over anhyd
Na2SO4, concentrated and the residue was purified by silica
gel column chromatography using toluene/ethyl acetate/
methanol (49:49:2, v/v/v). Purity of the isolated compounds
.98% (1H NMR spectroscopy).
31P NMR. dP (CDCl3) 84.32 ppm.
13C NMR. dC (in ppm, CDCl3) 163.81 and 163.64 (2£C4),
158.93 (4C of DMT), 150.64, 150.49, 150.37, 150.29 (pyr-
C2, pyr-C6, 2£C2), 144.83 and 144.20 (2C of DMT),
141.06 (d, J¼150 Hz, pyr-C4), 135.95 and 135.94 (2C of
DMT), 135.26, 135.12, 135.05 (2C of DMT, 2£C6), 130.28,
130.24, 130.15, 128.27, 128.23, 128.20, 128.15 (16C of
DMT), 127.39 and 127.34 (2C of DMT), 123.99 (d,
J¼9.2 Hz, pyr-C3, pyr-C5), 113.53 (8C of DMT), 111.93
and 111.32 (2£C5), 87.55 and 87.48 (2£C DMT), 85.39
(Cb10), 85.32 (d, J¼3.1 Hz, Ca40), 84.50 (Ca10), 83.96 (d,

394
T. Johansson, J. Stawinski / Tetrahedron 60 (2004) 389–395
4.3.1. 50-O-Dimethoxytritylthymidin-30-yl 30-O-
dimethoxytritylthymidin-50-yl pyridyl-2-phosphonothio-
ate 5a (from faster moving H-phosphonate diesters 1a).
White solid (0.253 g), yield 79%. HRMS [MþH]þ found:
1227.4070; C67H66N5O14PS requires: 1227.4065.
4.3.3. Thymidin-30-yl thymidin-50-yl pyridyl-2-phos-
phonothioate 5c and 5d. Separate diastereomers of
2-pyridylphosphonothioates 5a and 5b were deprotected
analogously as it was described above to the 4-pyridyl
derivatives 3a and 3b.
1H NMR. dH (in ppm CDCl3) 8.55 and 8.48 (2£s, 2H,
2£NH), 8.43 (d, 1H, J¼4.8 Hz, pyr-H6), 7.81 (m, 1H, pyr-
H3), 7.71 (m, 1H, pyr-H4), 7.54 (s, 1H, Ha6), 7.45–7.15 (m,
20H, ArH, pyr-H5, Hb6), 6.83 (m, 8H, ArH ortho to OCH3),
6.41–6.34 (m, 2H, 2£ H10), 5.47 (m, 1H, Ha30), 4.34 (d,
J¼6.2 Hz, 1H, Hb30), 4.14 (s, 1H, Ha40), 4.05–3.67 (m, 2H,
Hb50), 3.85 (s, 1H, Hb40), 3.79, 3.77 (2£s, 12H, 4£ CH3O),
3.33 (m, 2H, Ha50), 2.45–2.31 (m, 2H, Ha20), 2.07–1.87 (m,
2H, Hb20), 1.73 (s, 3H, Cb5–CH3), 1.43 (s, 3H, Ca5–CH3).
Compound 5c (from faster moving diastereomer 5a). White
solid (0.068 g), yield 84%. Anal. calcd for C25H30N5O10PS:
C, 48.15; H, 4.85; N, 11.23. Found: C, 48.03; H 4.99; N
11.11.
1H NMR. dH (in ppm CD3OD) 8.76 (d, 1H, J¼4.4 Hz, pyr-
H6), 8.12 (m, 1H, pyr-H3), 7.99 (m, 1H, pyr-H4), 7.82, 7.67
(2£s, 02H, 2£6), 7.60 (m,01H, pyr-H5), 6.36–6.27 (m, 2H,
2£H1 ), 5.43 (m, 1H, Ha3 ), 4.52–4.39 (m, 3H, Hb30, Hb50),
4.22 (s, 1H, Ha40), 4.15 (s, 1H, Hb40,), 3.80 (m, 2H, Ha50),
2.61–2.35 (m, 2H, Ha20), 2.33–2.22 (m, 2H, Hb20), 1.89,
1.80 (2£ s, 6H, 2£ C5–CH3).
31P NMR. dP (CDCl3) 78.15 ppm.
13C NMR. dC (in ppm, CDCl3) 163.86 and 163.78 (2£C4),
158.91 (4C of DMT), 154.55 (d, J¼192 Hz, pyr-C2),
150.49, 150.43, 150.20 (2£C2, pyr-C6), 145.04 and 144.30
(2C of DMT), 136.51 (d, J¼13.0 Hz, pyr-C4), 136.25 and
136.44 (2C of DMT), 136.00 (Cb6), 135.34 (2C of DMT),
135.21 (Ca6), 130.31, 130.18, 128.40, 128.22, 128.18 (16C
of DMT), 127.32 (2C of DMT), 127.04 (d, J¼29.8 Hz, pyr-
C3), 126.28 (pyr-5H), 113.58 and 113.46 (8C of DMT),
111.75 and 111.15 (2£C5), 87.57 and 87.29 (2£C DMT),
85.80 (Cb10), 84.80 (Ca40), 84.52 (Ca10, Cb40) 78.50 (Ca30),
74.91 (Cb30), 67.02 (Cb50), 63.44 (Ca50), 55.40 (4£CH3O),
39.42 (Ca20, Cb20), 12.34 (Cb5–CH3), 11.82 (Ca5–CH3).
31P NMR. dP (CDCl3) 78.81 ppm.
Compound 5d (from slower moving diastereomer 5b).
White solid (0.065 g), yield 80%. Anal. calcd for
C25H30N5O10PS: C, 48.15; H, 4.85; N, 11.23. Found: C,
47.99; H 4.93; N 11.05.
1H NMR. dH (in ppm, CD3OD) 8.77 (d, 1H, J¼4.4 Hz, pyr-
H6), 8.15 (m, 1H, pyr-H3), 7.99 (m, 1H, pyr-H4), 7.82, 7.65
(2£s, 02H, 2£6), 7.60 (m, 01H, pyr-H5), 6.36–6.28 (m, 2H,
2£H1 ), 5.42 (m, 1H, Ha3 ), 4.55–4.34 (m, 2H, Hb50), 4.47
(m, 2H, Hb30), 4.31 (m, 1H, Ha40), 4.15 (s, 1H, Hb40), 3.85
(m, 2H, Ha50), 2.50–2.27 (m, 2H, Ha20), 2.24–2.19 (m, 2H,
Hb20), 1.89, 1.83 (2£s, 6H, 2£C5–CH3).
4.3.2. 50-O-Dimethoxytritylthymidin-30-yl 30-O-
dimethoxytritylthymidin-50-yl pyridyl-2-phosphonothio-
ate 5b (from slower moving H-phosphonate diesters 1b).
White solid (0.255 mg), yield 80%. HRMS [MþH]þ found:
1227.4067; C67H66N5O14PS requires: 1227.4065.
31P NMR. dP (CDCl3) 79.26 ppm.
1H NMR. dH (in ppm, CDCl3) 8.58 (d, 1H, J¼4.4 Hz, pyr-
H6), 8.35, 8.24 (2£s, 2H, 2£NH), 7.94 (m, 1H, pyr-H3),
7.79 (m, 1H, pyr-H4), 7.56 (s, 1H, Ha6), 7.41–7.17 (m, 20H,
ArH, pyr-H5, Hb6), 6.82 (m, 8H, ArH ortho to OCH3),
6.43–6.33 (m, 2H, 2£H10), 5.46 (m, 1H, Ha30), 4.23 (d,
J¼5.6 Hz, 1H, Hb30), 4.15 (s, 1H, Ha40), 3.79–3.69 (m, 3H,
Hb40, Hb50), 3.79, 3.76 (2£s, 12H, 4£CH3O), 3.41 (m, 2H,
Ha50), 2.44–2.25 (m, 2H, Ha20), 1.93–1.63 (m, 2H, Hb20),
1.76 (s, 3H, Cb5–CH3), 1.56 (s, 3H, Ca5–CH3).
Acknowledgements
We are indebted to Professor P. J. Garegg for his interest in
this work. Financial support from the Swedish Natural
Science Research Council and the Swedish Foundation for
Strategic Research is gratefully acknowledged.
References and notes
31P NMR. dP (CDCl3) 79.29 ppm.
1. Redmore, D. (Petrolite Corp.). U.S. 3,673,196, 1972.
¨
2. Ruile, S.; Kohle, O.; Pechy, P.; Gratzel, M. Inorg. Chim. Acta
13C NMR. dC (in ppm, CDCl3) 163.71 (2£C4), 158.93 and
158.87 (4C of DMT), 154.62 (d, J¼192 Hz, pyr-C2),
150.51, 150.38, 150.27 (2£C2, pyr-C6), 145.02 and 144.25
(2C of DMT), 136.58 (d, J¼13.0 Hz, pyr-C4), 136.22 and
136.13 (2C of DMT), 135.88 (Cb6), 135.34 (2C of DMT),
135.17 (Ca6), 130.34, 130.28, 130.21, 128.35, 128.22
(16C of DMT), 127.37, 127.27 (2C of DMT), 127.33
(d, J¼31.3 Hz, pyr-C3), 126.45 (pyr-5H), 113.52 (8C of
DMT), 111.65 and 111.37 (2£C5), 87.52 and 87.41 (2£C
DMT), 85.17 (d, J¼4.6 Hz, Ca40), 84.98 (Cb10), 84.60
(Ca10), 84.42 (d, J¼8.4 Hz, Cb40), 78.67 (Ca30), 74.62
(Cb30), 66.65 (Cb50), 63.20 (Ca50), 55.41 and 55.38
(4£CH3O), 39.25 (Ca20, Cb20), 12.35 (Cb5–CH3), 11.81
(Ca5–CH3).
¨
1997, 261, 129–140. Gratzel, M.; Kohle, O.; Nazeeruddin,
M. K.; Pechy, P.; Rotzinger, F. P.; Ruile, S.; Zakeeruddin,
S. M. PCT Int. Appl. 9,529,924, Ecole Polytechnique Federale
de Lausanne (epfl), Switz., 1995.
3. Redmore, D. (Petrolite Corp.). U.S. 3,775,057, 1973.
4. Redmore, D. (Petrolite Corp.) U.S. 3,810,907, 1974.
5. Protopopova, G. V.; Reidalova, L. I.; Pavlenko, A. F.;
Sologub, L. S.; Kukhar, V. P.; Petrenko, V. S. Fiziol. Akt.
Veshchestva 1980, 12, 16–18.
6. Kukhar, V. P.; Cherepenko, T. I.; Pavlenko, A. F. Dokl. Akad.
Nauk Ukr. RSR, Ser. B: Geol., Khim. Biol. Nauki 1982, 60–63.
Akiba, K.; Tsuzuki, K. (Toyo Soda Mfg. Co., Ltd, Japan). Jpn
Kokai Tokkyo Koho 61,221,102, 1986..

T. Johansson, J. Stawinski / Tetrahedron 60 (2004) 389–395
395
7. Malhotra, S. K.; Evoy, I. L. (Dow Chemical Co., USA). Dow
Chemical Co. 4,606,757, 1986.
Geterotsikl. Soedin. 1974, 1354–1356. Haase, M.; Gunther,
W.; Gorls, H.; Anders, E. Synthesis 1999, 2071–2081.
22. Onys’ko, P. P.; Suvalova, E. A.; Chudakova, T. I.; Sinitsa,
A. D. Zh. Obshch. Khim. 1994, 64, 610–612.
8. Murray, K. J.; Porter, R. A.; Prain, H. D.; Warrington, B. H.
(Smith Kline and French Laboratories Ltd, UK). PCT Int.
Appl. 9,117,987, 1991.
23. Zain, R.; Stawinski, J. J. Org. Chem. 1996, 61, 6617–6622.
24. Connolly, B. A.; Potter, B. V. L.; Eckstein, F.; Pingoud, A.;
Grotjahn, L. Biochemistry 1984, 23, 3443–3453.
9. Erion, M. D.; Vanpoelje, P. (Metabasis Therapeutics, Inc.,
USA). PCT Int. Appl. 0,038,666, 2000. Dang, Q.; Kasibhatla,
S. R.; Reddy, K. R.; Erion, M. D.; Reddy, M. R.; Agarwal, A.
(Metabasis Therapeutics, Inc., USA). PCT Int. Appl. 0,014,
095, 2000.
25. Bollmark, M.; Stawinski, J. Chem. Commun. 1997, 991–992.
26. After 1 h, the reaction mixture consisted of intermediate 2
(21%), 4-pyridylphosphonothioate 3 (61%) and a dinucleoside
phosphorothioate diester (18%).
10. Murray, K. J.; Porter, R. A.; Warrington, B. H.; Lahouratate, P.
(SmithKline Beckman Corp., UK Smithkline Beecham
Laboratoires Pharmaceutiques). PCT Int. Appl. 9,319,754,
1993.
¨
27. Stawinski, J.; Stromberg, R.; Zain, R. Tetrahedron Lett. 1992,
33, 3185–3188.
28. Stawinski, J.; Thelin, M. Tetrahedron Lett. 1992, 33,
3189–3192.
11. Kafarski, P.; Lejczak, B. Phosphor. Sulfur Silicon 1991, 63,
193–215.
29. Katritzky, A. J. Chem. Soc. 1956, 2404–2408. Feely, W.;
Lehn, W. L.; Boekelheide, V. J. Org. Chem. 1957, 22, 1135.
Coats, N. A.; Katritzky, A. J. Org. Chem. 1959, 24,
1836–1837.
12. Bergin, C.; Hamilton, R.; Walker, B.; Walker, B. J. Chem.
Commun. 1996, 1155–1156.
13. Chen, R. Y.; Mao, L. J. Phosphor. Sulfur Silicon 1994, 89,
97–104.
30. This assignment was supported by an independent synthesis of
methyl 50-O-dimethoxytritylthymidin-30-yl and methyl 30-O-
dimethoxytritylthymidin-50-yl 2-pyridylphosphonothioates,
and their comparison with the side products formed in the
investigated reaction.
14. Kers, A.; Stawinski, J. Tetrahedron Lett. 1999, 40,
4263–4266.
15. Johansson, T.; Stawinski, J. Chem. Commun. 2001,
2564–2565.
16. Johansson, T.; Kers, A.; Stawinski, J. Tetrahedron Lett. 2001,
42, 2217–2220.
31. The amount of DBU used in this reaction (4 equiv.) was also
critical for efficient formation of pyridylphosphonothioates 5.
When the reaction of H-phosphonothioate 1 was carried out
with 1.2 equiv. of N-methoxypyridinium p-toluenesulfonate in
the presence of only 2 equiv. of DBU, complex reaction
mixture was formed (pyridylphosphonate 5, ca. 70% methyl
nucleoside 2-pyridylphosphonothioates, ca. 4% dinucleoside
hydroxymethylphosphonates, ca. 18% and 4-pyridylphos-
phonothioates 3, ca. 8%).
17. Huang, Y. S.; Chaudret, B.; Bellan, J.; Mazieres, M. R.
Polyhedron 1991, 10, 2229–2233. Speca, A. N.; Pytlewski,
L. L.; Karayannis, N. M. J. Inorg. Nucl. Chem. 1972, 34,
3671–3679.
¨
¨
¨
18. Haner, R.; Hall, J.; Pfutzer, A.; Husken, D. Pure Appl. Chem.
1998, 70, 111–116. Konevetz, D. A.; Beck, I. E.; Beloglazova,
N. G.; Sulimenkov, I. V.; Silnikov, V. N.; Zenkova, M. A.;
Shishkin, G. V.; Vlassov, V. V. Tetrahedron 1999, 55,
503–512.
32. The reaction was carried out in acetonitrile using H-phos-
phonothioate 1, 2 equiv. of N-methoxypyridinium p-toluene-
sulfonate and 4 equiv. of triethylamine (rt, 5 min). Further
studies on regioselectivity in this type of reactions are in
progress.
19. Barton, J. K. Pure Appl. Chem. 1998, 70, 873–879. Kelley,
S. O.; Jackson, N. M.; Hill, M. G.; Barton, J. K. Angew. Chem.,
Int. Ed. 1999, 38, 941–945. Schuster, G. B. Acc. Chem. Res.
2000, 33, 253–260.
33. Katritzky, A. R.; Lunt, E. Tetrahedron 1969, 25, 4291–4305.
34. Stawinski, J.; Thelin, M. J. Org. Chem. 1994, 59, 130–136.
Stawinski, J.; Thelin, M.; Zain, R. Tetrahedron Lett. 1989, 30,
2157–2160.
20. Pitt, L. S.; Large, G. B.; MacDonald, A. A. (Stauffer Chemical
Co., USA). Ger. Offen. 2,738,194, 1978.. Gray, R. A.; Hyzak,
D. L. (Stauffer Chemical Co., USA). U.S. 4,652,302, 1987.
21. Eliseenkov, V. N.; Bureva, N. V.; Ivanov, B. E. Khim.