Aryl H-phosphonates. 14. Synthesis of new nucleotide analogues with phosphonate-phosphate internucleosidic linkage

Article abstract of DOI:10.1021/ol035166u

(Equation presented) Aryl nucleoside H-phosphonates 3 and aryl nucleoside P-acylphosphonates 4, generated in situ from the appropriate H-phosphonate 1 and acylphosphonate monoesters 2, respectively, reacted rapidly in the presence of tertiary amines to pr

Full text of DOI:10.1021/ol035166u

ORGANIC
LETTERS
2003
Vol. 5, No. 20
3571-3573
Aryl H-Phosphonates. 14. Synthesis of
New Nucleotide Analogues with
Phosphonate−Phosphate
Internucleosidic Linkage
Marzena Szymczak,† Agnieszka Szyman´ska,† Jacek Stawin´ski,†,‡
Jerzy Boryski,† and Adam Kraszewski*,†
Institute of Bioorganic Chemistry, Polish Academy of Sciences, 61-704 Poznan´,
Poland, and Department of Organic Chemistry, Arrhenius Laboratory,
Stockholm UniVersity, S-106 91 Stockholm, Sweden
akad@ibch.poznan.pl
Received June 24, 2003
ABSTRACT
Aryl nucleoside H-phosphonates 3 and aryl nucleoside P-acylphosphonates 4, generated in situ from the appropriate H-phosphonate 1 and
acylphosphonate monoesters 2, respectively, reacted rapidly in the presence of tertiary amines to produce in high yields the extended,
pyrophosphate analogues, diaryl dinucleoside phosphonate
−phosphate derivatives 6. These, depending on a substituent on the
r
-carbon of
the phosphonate moiety, either underwent transformation into the dinucleoside phosphonate−phosphate 7 or afforded nucleoside H-phosphonates
8 and aryl nucleoside phosphate 9.
Searching for new nucleoside kinase bypass lipophilic
pronucleotides,1 we considered diaryl dinucleoside phospho-
nate-phosphate (6, Scheme 1) as unique vehicles for
delivering biologically active nucleotides into the cell. A
distinctive feature of this type of analogues is that their
lipophilicity and susceptibility to hydrolysis are expected to
be tuneable through a proper choice of substituent R on the
R-carbon of the phosphonate center and the aryl ester
moieties. In addition, the electronic structure of R can be
instrumental in controlling possible degradation pathways
of phosphonate-phosphate 6.
analogues, have not been investigated yet. This is most likely
due to synthetic problems connected with the preparation of
gem-diphosphonates of type 5 and harsh reaction conditions
that are required for their rearrangement into the respective
phosphonate-phosphates 6.4
To overcome these problems we designed a new synthesis
of aryl esters of dinucleoside phosphonate-phosphates 6 by
(1) (a) Siddiqui, A.; Ballatore, C.; McGuigan, C.; DeClerq, E.; Balzarini,
J. J. Med. Chem. 1999, 42, 393-399. (b) Peyrottes, S.; Schlienger, N.;
Beltran, T.; Lefebvre, I.; Pompon, A.; Gosselin, G.; Aubertin, A.-M.;
Imbach, J.-L.; Perigaud, C. Nucleosides Nucleotides Nucleic Acids 2001,
20, 315-321. (c) Meier, C.; Lorey, M.; DeClerq, E.; Balzarini, J. J. Med.
Chem. 1998, 41, 1417-1427. (d) Parang, K.; Wiebe, L. I.; Knaus, E. E.
Curr. Med. Chem. 2000, 7, 995-1039.
Although simple tetraalkyl phosphonate-phosphates are
known compounds2 and have already found several thera-
peutic applications,3 the corresponding nucleotide derivatives
of type 6, which can be viewed as extended pyrophosphate
(2) (a) Nicholson, D. A.; Vaughn, H. J. Org. Chem. 1971, 36, 3843-
3845. (b) Gancarz, R.; Gancarz, I. Tetrahedron Lett. 1993, 34, 145-148.
(3) (a) Nguyen, L. M.; Niesor, E.; Bentzen, C. L. J. Med. Chem. 1987,
30, 1426-1433. (b) Turhanen, P. A.; Ahlgren, M. J.; Jarvinen, T.;
Vepsalainen, J. Synthesis 2000, 633-637.
† Polish Academy of Sciences.
‡ Stockholm University.
(4) Fitch, S. J.; Moedritzer, K. J. Am. Chem. Soc. 1962, 84, 1876-1879.
10.1021/ol035166u CCC: $25.00
© 2003 American Chemical Society
Published on Web 09/04/2003

elaborated an efficient method involving condensation of
nucleoside P-acylphosphonate of type 27 with 4-chlorophenol
(1.2 molar equiv) in the presence of diphenyl chlorophos-
phate (DPCP, 1.5 molar equiv). This reaction proceeded
smoothly in methylene chloride-pyridine (9:1, v/v), and after
2 h the desired 4-chlorophenyl (or phenyl) nucleoside
P-acylphosphonate 48 (two diastereoisomers, for 4a, δP )
-6.50 and -6.95 ppm, 2t, 3JHP ) 6.5 Hz) was usually formed
as the sole nucleotidic product (31P NMR spectroscopy).
To produce phosphonate-phosphate 6a (Scheme 1),
equimolar amounts of aryl H-phosphonate 3a and aryl
P-acylphosphonate 4a were allowed to react in methylene
chloride-pyridine (9:1, v/v) in the presence of N,N-diiso-
propylethylamine (DIEA, 10 molar equiv). The reaction was
rapid (<5 min) and afforded a new compound, which gave
rise to two multiplets centered at δP ) 16.39 and -6.22 ppm
in the 31P NMR spectrum. Integrals of these two groups of
signals (ca 1:1) together with their chemical shifts indicated
the presence of two kinds of phosphorus centers in the
molecule (a phosphate triester center, δP ) -6.22 ppm ,and
an aromatic C-phosphonate center, δP ) 16.39 ppm), and
suggested a phosphonate-phosphate structure of type 6a.9
Compound 6a was stable during silica gel column chroma-
tography [gradient of propanol-2 (0-10%) in methylene
chloride] and its structure was unambiguously confirmed by
1H, 31P, and HRMS spectroscopy (vide infra).
Scheme 1. Synthesis and Degradation of Dinucleoside
Phosphonate-Phosphatesa
By monitoring the progress of this reaction by 31P NMR
spectroscopy we could not detect a putative intermediate
involved in this transformation, namely gem-diphosphonate
5. Since tetraalkyl gem-diphosphonates are known to be
rather stable species that require harsh reaction conditions
for the rearrangement,2a this may indicate that the presence
of aromatic moieties in 5 significantly facilitates the conver-
sion into phosphonate-phosphate product 6.
Since a separate handling of aryl H-phosphonates 3 and
aryl P-acylphosphonates 4 may cause, due to the high
reactivity of these compounds, some experimental inconve-
nience, we attempted to develop a four-component one-pot
a Reagents and conditions: (i) conc aq NH3, 50 °C; (ii) CH3CN/
Et3N/H2O 2:1:1 (v/v) or phosphate buffer, pH 7.4, 37 °C.
reaction for the synthesis of phosphonate-phosphate 6. 31
P
NMR spectroscopy revealed that by reacting equimolar
amounts of H-phosphonate monoester 1a and P-acylphos-
phonate 2a with 4-chlorophenol in the presence of a
condensing agent (DPCP), the expected aryl esters 3a and
4a were formed cleanly and in equal amounts. Addition of
DIEA to this reaction mixture triggered rapid (<5 min)
formation of the desired phosphonate-phosphate 6a, which
was isolated as described above. This modification signifi-
cantly simplified the synthetic protocol and, due to sup-
making use of nucleoside aryl H-phosphonates 3 and the
corresponding P-acylated derivatives 4, as shown in Scheme
1. The inspiration for this came from our previous observa-
tion that phosphonate-phosphate derivatives of type 6 were
formed as side-products during aminolysis of reaction
mixtures containing nucleoside aryl H-phosphonates and
pivaloyl chloride.5
Starting materials for this new synthetic protocol are read-
ily accessible. Aryl H-phosphonates 3 can be prepared from
the corresponding nucleoside H-phosphonates 1 and the
appropriate phenol as described previously.6 For the prepara-
tion of the second component, acylphosphonate 4, we
(7) Compounds 2 were obtained by modification of the method of van
Boom et al. (de Vroom, E.; Spierenburg, M. L.; Dreef, C. E.; Van der Marel,
G. A.; Van Boom, J. H. Recl. TraV. Chim. Pays-Bas 1987, 106, 65-66)
that involved acylation of a nucleoside bis(trimethylsilyl)phosphite with
respective acyl chloride.
(8) Compounds 4a-d turned out to be rather unstable and underwent
rapid hydrolysis to acylphosphonate monoester of type 2 upon the addition
of water. Their structure was assigned on the basis of the spectral data
(chemical shift and splitting pattern of signals in the 31P NMR spectra) and
the observed chemical reactivity.
(5) (a) Sobkowska, A.; Sobkowski, M.; Cies´lak, J.; Kraszewski, A.; Kers,
I.; Stawin´ski, J. J. Org. Chem. 1997, 62, 4791-4794. (b) Cies´lak, J.;
Jankowska, J.; Sobkowski, M.; Kers, A.; Kers, I.; Stawin´ski, J.; Kraszewski,
A. Collect. Symp. Ser. 1999, 2, 63-68.
(6) Cies´lak, J.; Szymczak, M.; Wenska, M.; Stawin´ski, J.; Kraszewski,
A. J. Chem. Soc. Perkin Trans. 1 1999, 3327-3331.
(9) Due to the presence of three chiral centers in the phosphonate-phos-
phate backbone, compounds 6 are formed as a mixture of eight diastereomers
and usually gave a complex pattern of signals in the 31P NMR spectra.
3572
Org. Lett., Vol. 5, No. 20, 2003

pressing possible interference from adventitious water, made
the method highly reproducible.
phonate 2 with phenols in the presence of diphenyl chloro-
phosphate, followed by treatment with N,N-diisopropyleth-
ylamine. The method is experimentally simple, making use
of readily accessible starting materials, and permits an easy
introduction of structural variations into the phosphonate-
phosphates of type 6. By a proper choice of ester groups
and a substituent on the R-carbon of the phosphonate moiety,
one can also control stability and the decomposition pathways
of these compounds. Since di-AZT phosphonate-phosphate
6d under mild conditions can generate AZT 5′-H-phospho-
nate 8 and AZT 5′-phenylphosphate 9, this type of com-
pounds can be considered as potential lipophilic prodrugs
for the delivery of one or two kinds of nucleoside mono-
phosphates into the cells. Evaluation of antiviral activity of
di-AZT phosphonate-phosphates and related compounds
will be reported in due course.13
The efficacy of this approach was further assessed in the
synthesis of phosphonate-phosphate 6b (Scheme 1) bearing
a nucleoside analogue with established antiviral properties,
3′-azido-3′-deoxythymidine (AZT). The replacement of thy-
midine by AZT did not affect the efficiency of the method
and di-AZT phosphonate-phosphates 6b was obtained in a
yield comparable with that of 6a.
To expand the array of structural variations in phospho-
nate-phosphate 6, we prepared phenyl and 4-methoxyphenyl
derivatives of 6 (6c, R ) phenyl, and 6d, R ) 4-methoxy-
phenyl) by condensing the respective aryl H-phosphonate 3
(3b or 3c) with different aryl P-acylphosphonates 4 (4c and
4d, respectively). The syntheses were uneventful and pro-
duced the expected phosphonate-phosphates 6c and 6d as
the sole nucleotidic products (31P NMR). In the instance of
the 4-methoxyphenyl derivative of 6 bearing the 4-chlo-
rophenyl phosphoester group, we observed a partial decom-
position during silica gel chromatography. Fortunately, the
replacement of the 4-chlorophenyl by unsubstituted phenyl
alleviated this problem and permitted the preparation of
diAZT phosphonate-phosphate 6d in high yield and of high
purity. Compounds 6 after precipitation from methylene
chloride with an excess of n-hexane, filtration, and drying
were obtained as amorphous white powders with purity
higher than 98%.
Acknowledgment. Financial support from the State Com-
mittee for Scientific Research, Republic of Poland (project
no PBZ-KBN-059/T09/19) is gratefully acknowledged.
Supporting Information Available: 1H, 31P NMR, and
HRMS analytical data for compounds 6a-d and 7a-c. This
material is available free of charge via the Internet at
http://pubs.acs.org.
OL035166U
(10) The initial products of the degradation of 6d are probably 9 and
AZT phenyl R-hydroxy(4-methoxyphenyl)methylenephosphonate. The latter
one is apparently in equilibrium with 4-anisaldehyde and AZT phenyl
H-phosphonate diester, but due to irreversible hydrolysis of the phenyl
H-phosphonate diester into AZT H-phosphonate 9, the equilibrium is driven
to the right and depletes the amount of R-hydroxyphosphonate in the reaction
mixture. Since no intermediates were observed by 31P NMR spectroscopy,
probably the rate determining step in this process is the formation of the
hydroxyphosphonate derivative. More detail studies on this are currently
being carried out in our laboratory.
To assess the stability of a phosphonate-phosphate
backbone, compounds 6 were subjected to treatment with
33% aq ammonia at 50 °C for 48 h. Phosphonate-
phosphates 6 with R ) tert-butyl or 4-chlorophenyl (6a, 6b,
and 6c) under these conditions underwent only ester group
hydrolysis and produced exclusively (31P NMR analysis)
unprotected phosphonate-phosphates 7a, 7b, and 7c. On the
other hand, AZT phosphonate-phosphate 6d with R )
4-methoxyphenyl afforded a rather complex mixture of
products. However, under milder conditions, e.g. in aceto-
nitrile-water-triethylamine (2:1:1, v/v, rt), 6d underwent
rapid and clean degradation (>5 min, 31P NMR) to afford
an equimolar amount of AZT H-phosphonate 810 and AZT
phenyl phosphate diester 9 exclusively.11
(11) The identity of compounds 8 and 9 was confirmed by comparison
with authentic samples obtained in other ways. (a) Cies´lak, J.; Jankowska,
J.; Sobkowski, M.; Wenska, M.; Stawin´ski, J.; Kraszewski, A. J. Chem.
Soc., Perkin Trans. 1 2002, 31-37. (b) Jankowska, J.; Sobkowski, M.;
Stawin´ski, J.; Kraszewski, A. Tetrahedron Lett. 1994, 35, 3355-3358.
(12) Meier, C. Angew. Chem., Int. Ed. Engl. 1993, 32 (12), 1704-1706.
(13) Typical Procedure for the Synthesis of Dinucleoside Phospho-
nate-Phosphates of Type 6. Nucleoside H-phosphonate 1 (1 molar equiv),
P-acylphosphonate 2 (1 molar equiv), and the appropriate phenol (2.5 molar
equiv) were rendered anhydrous by repeated evaporation of added pyridine
(3 × 20 mL/1 mmol) and then dissolved in methylene chloride/pyridine
9:1 (v/v) (10 mL/1 mmol). To this solutionwas added diphenyl chloro-
phosphate (2.5 molar equiv), and when the formation of aryl phosphonate
esters 3 and 4 was complete (ca. 2 h), N,N-diisopropylethylamine (10 molar
equiv) was added. After 5 min the reaction mixture was diluted with
methylene chloride (10 times the initial volume) and washed with 5%
aq NaHCO3. The organic layer was dried (Na2SO4) and evaporated, and
the oily residue was applied on a silica gel column preequilibrated with
CH2Cl2. Products 6 were isolated by using a stepwise gradient of propanol-2
(0-10%) in methylene chloride. Fractions containing pure products were
evaporated and the residue precipitated with an excess of petroleum ether.
After filtration and drying under vacuum, compounds 6 were obtained as
amorphous white powders with a purity higher than 98% (1H NMR).
Synthesis of Dinucleoside Phosphonate-Phosphates of Type 7. To
products 6a-c dissolved in pyridine (0.1 g/2 mL) was added aqueous
concentrated ammonia (33%; 4 mL) and the reaction mixture was kept for
24 h at 50 °C. This time was usually sufficient to complete the deprotection
of phosphonate-phosphates 6a-c (31P NMR). Ammonia and solvents were
removed by evaporation, and the residue was dissolved in propanol-2/conc
NH3 95:5 (v/v) and applied to a silica gel column equilibrated with the
same solvent. Compounds 7a-c were isolated by using a linear gradient
of water [0-10% (v/v)] in propanol-2/conc NH3 95:5 (v/v). Fractions
containing pure compounds were collected and evaporated with an added
excess of propanol-2 to afford products 7a-c (ammonium salts) as
amorphous white powder.
The same degradation pathway of 6d but with different
kinetics (t1/2 ca. 60 min) was also observed in 0.1 M
phosphate buffer (pH 7.4, 37 °C). Since phosphonate-
phosphates 6b and 6c were completely stable under the same
conditions (overnight), it seems that the presence of a strong
electron-donating R group in 6 favors the C-O bond scission
of a phosphonate-phosphate backbone, most likely due to
stabilization of a putative intermediate carbocation.12 This
feature can make dinucleoside phosphonate-phosphates of
type 6d potentially useful vehicles for delivery of pronucle-
otides into the cell, where they will undergo further chemical
and enzymatic transformations into the corresponding,
nucleoside triphosphates of antiviral activity.
In conclusion, we developed a new and efficient protocol
for the synthesis of novel dinucleoside phosphonate-
phosphate analogues. It consisted of a one-pot reaction of
the nucleoside H-phosphonate 1 and nucleoside acylphos-
Org. Lett., Vol. 5, No. 20, 2003
3573