Synthesis of phosphonate analogues of 2-deoxyribose-1-phosphate

Article abstract of DOI:10.1055/s-1998-1600

Phosphonate derivatives of 2-deoxyribose have been synthesized for the first time, by Lewis acid-mediated reaction of glycosyl donors with diisopropylhydroxymetylphosphonate for compounds 2 or with triisopropyl phosphite for compounds 3. Although, the selectivity of the reaction was moderate, in both cases the α-glycosides were predominantly obtained.

Full text of DOI:10.1055/s-1998-1600

February 1998
SYNLETT
177
Synthesis of Phosphonate Analogues of 2-Deoxyribose-1-Phosphate
a
a
b
a
María-José Rubira, María-Jesús Pérez-Pérez, * Jan Balzarini, and María-José Camarasa
a
Instituto de Química Médica (C.S.I.C.), Juan de la Cierva 3, 28006 Madrid, Spain
b
Rega Institute for Medical Research, Minderbroedersstraat 10, B-3000 Leuven, Belgium
Fax: (int.)+34-1-5644853; E-mail: mjesus@pinar1.csic.es
Received 21 November 1997
Abstract: Phosphonate derivatives of 2-deoxyribose have been
synthesized for the first time, by Lewis acid-mediated reaction of
glycosyl donors with diisopropylhydroxymetylphosphonate for
compounds
2
or with triisopropyl phosphite for compounds 3.
Although, the selectivity of the reaction was moderate, in both cases the
α-glycosides were predominantly obtained.
Glycosyl phosphates are important metabolic intermediates in the
biosynthesis of nucleic acids and glycosides. Among them, 2-deoxy-α−
D-ribose-1-phosphate (1) is
a natural substrate of thymidine
phosphorylase (EC 2.4.2.4), an enzyme involved in the metabolism of
pyrimidine nucleosides and whose increased levels have been
1
associated with the development of solid tumours and metastasis. The
synthesis of stable analogues of glycosyl phosphates (i.e. phosphonate
derivatives) is of great interest since they may act either as inhibitors or
regulators in the metabolic processes where the natural phosphates are
involved.
Recently, we have started a programme directed to the synthesis of
stable analogues of 2-deoxy-α-D-ribose-1-phosphate (1) as potential
inhibitors of thymidine phosphorylase (EC 2.4.2.4). Here we describe
the synthesis of the phosphonate derivatives 2 and 3 as analogues of the
The synthesis of glycosyl phosphonates 3α and 3β was accomplished by
Lewis acid-promoted reaction of glycosyl donors with phosphites
(Scheme 2). Initial attempts to react the methyl derivative 4 with
11
P(OiPr) in the presence of TMSOTf were unsuccessful. However,
natural substrate 1, in which the phosphate group has been replaced by
3
2,3
when 4 was transformed into the more reactive acetate 6 (obtained in
the isosteric phosphonomethoxy moiety.
To the best of our
4
86% yield by reaction of 4 with AcOH/Ac O/H SO ), followed by
knowledge, this is the first time that glycosyl phosphonate derivatives of
2-deoxyribose are reported.
2
2
4
treatment of 6 with P(OiPr) (2.5 eq.) in the presence of TMSOTf (1.1
3
12
eq.) afforded the glycosyl phosphonates 7 in 70% yield, as a mixture
of α and β anomers (5:2 ratio) that could not be separated by
chromatography. When BF •OEt was used as catalyst instead of
3
2
TMSOTf, the yield of 7 was increased up to 86%, but the ratio of
anomers (α:β) remained unchanged. Finally, reaction of 7 with TMSBr
(10 eq.) and 2,6-lutidine (15 eq.) followed by treatment with methanolic
ammonia, yielded the fully deprotected compound 3 (43 %) as a mixture
Figure 1
13
of α and β anomers.
In the synthesis of 2 and 3, special attention was paid to the preparation
of the α-anomers (2α, 3α), that show the same stereochemistry as the
natural substrate 1. Here we also report our attempts to selectively
obtain these 2-deoxy-α-glycosides, together with the unexpected results
encountered.
Compounds 2α and 2β were initially synthesized starting from methyl
4
3,5-bis-O-benzoyl-erythro-pentofuranoside (4) (Scheme 1). To favour
5
the α-anomers,
4
(1 eq.) was reacted with diisopropyl
6
hydroxymethylphosphonate
(3 eq.) in the presence of
trimethylsilyltriflate (1.7 eq., CH CN, 0 °C, 4h) to afford a mixture of
3
7a
7b
5α (51%) and 5β (25%) in a 2:1 ratio, that were separated by flash
column chromatography, using CH Cl : acetone (20:1) as eluent. The
In an attempt to direct the synthesis exclusively to the α-anomers, we
2
2
devised a new synthetic strategy based on the method recently described
by Park and Rizzo, for the stereoselective synthesis of 2'-β-D-
14
stereochemistry of the anomeric center was assigned as α for the major
1
isomer 5α and β for the minor 5β based on H-NMR NOESY
deoxyribonucleosides. They used the assistance of
a
m-
8
experiments. Treatment of 1 eq. of 5α or 5β with bromotrimethylsilane
trifluoromethylbenzoyl group at position 2 of the ribose, both as an
agent directing the stereochemistry in the glycosylation reaction
9
(10 eq.) in dry acetonitrile, in the presence of 2,6-lutidine (15 eq.),
(deprotection of the phosphonate esters), followed by reaction with
saturated methanolic ammonia (deprotection of the benzoyl esters),
(leading to 1,2-trans glycosides) and as
a precursor in the
deoxygenation step. However, when we applied a similar strategy to
10a
10b
15
afforded compounds 2α
disodium salts, by treatment with Dowex 50W-X-4 (Na form), (43 and
and 2β
that were isolated as their
selectively prepare the α-anomer 5α, using 8 as starting material, we
+
obtained the β-anomer 5β (Scheme 3). Thus, reaction of 8 with m-
48% yields, from 5α and 5β, respectively).

178
LETTERS
SYNLETT
trifluoromethylbenzoic acid (2.5 eq.) in dioxane under Mitsunobu
conditions [Ph P (2.5 eq.), DIAD (2.5 eq.), 80°C, 12h] afforded
3
References and Notes
compound 9 in low yield. Addition of 2,6-lutidine to the reaction
(1) a) Yoshimura, A.; Kuwazuru, Y.; Furukawa, T.; Yoshida, H.;
Yamada, K.; Akiyama, S. Biochim. Biophys. Acta 1990, 1034,
107. b) Reynolds, K; Farzaneh, F.; Collins, W.P.; Campell, S.;
Bourne, T.H.; Lawton, F.; Moghaddam, A.; Harris, A.H.;
Bicknell, R. J. Natl. Cancer Inst. 1994, 86, 1234. c) Maeda, K;
Chung, Y.S.; Ogawa, Y.; Takatsuka, S.; Kang, S.M.; Ogawa, M.;
Sawada, T.; Onoda, N.; Kato, Y.; Sowa, M. Brit. J. Cancer, 1996,
73, 884.
16
medium allowed us to increase the yield of
9 up to 70%.
Glycosylation of this sugar derivative 9 with diisopropyl hydroxymethyl
phosphonate (2.5 eq.) in CH Cl using TMSOTf (1.2 eq.) as catalyst
2
2
17
afforded the coupling product 11 in low yield (30%). The yield of 11
was increased up to 66% upon transformation of 9 to the corresponding
4
acetate 10 [AcOH/Ac O/H SO (82% yield)] followed by coupling of
2
2
4
10 with the alcohol HOCH P(O)(OiPr) in the presence of TMSOTf as
2
2
catalyst. The photolysis of 11 carried out through Pyrex using a 450W
medium-pressure mercury lamp in degassed isopropanol:water
(2) Bronson, J.J.; Kim, C.U.; Ghazzouli, I.; Hitchcock, M.J.M.; Kern,
E.; Martin, J.C.; Nucleotide Analogues as Antiviral Agents;
Martin, J.C., Ed. A. C. S.: Washington, DC, 1989, pp 72-87.
a
solution (9:1), and in the presence of Mg(ClO ) and N-methylcarbazole
4 2
14
as photosensitizer at a concentration of 1.4 mM, afforded the 2-
deoxyderivative 5β (47%), together with the 2,3-dideoxysugar 12
18
(3) For a similar approach in pyranosyl nucleosides see : Alexander,
P.; Krishnamurthy, V.V.; Prisbe, E.J. J. Med. Chem. 1996, 39,
1321.
(24%).
The stereochemical outcome of the sequence of reactions described here
strongly suggests that the Mitsunobu reaction in the initial acylation step
of 8, did not proceed, with the expected inversion of the stereochemistry
(4) Gold, A.; Sangaiah, R. Nucleosides & Nucleotides, 1990, 9, 907.
(5) Watanabe, K.A.; Hollenberg, D.H.; Fox, J.J. J. Carbohyd. Chem.
Nucleosides & Nucleotides 1974, 1, 1.
19
at the C-2, but rather with retention, leading to 9. Scarce examples of
esterification of secondary alcohols under Mitsunobu conditions that do
not proceed with inversion are reported.
(6) Phillion, D.P.; Andrews, J.J. Tetrahedron Lett. 1986, 27, 1477.
19
1
(7) a) Physical data of 5α. [α] +63.7 (c 1, methanol). H-NMR data
D
It should be emphasized that attempts to determine the stereochemistry
(CDCl ): δ 1.25-1.40 (m, 12 H, CH ); 2.25-2.63 (m, 2H, H2),
3
3
of 9 or 11 by n.O.e. or NOESY NMR experiments were unsuccessful. A
conformational analysis of these compounds using PSEUROT
3.70-4.10 (m, J = 8.0, J = 11.2 Hz, 2H, OCH P), 4.60 (m, 3H, H4,
2
20
2H5), 4.63-4.85 (m, 2H, OCH(CH ) ), 5.34 (d, J = 5.1 Hz, 1H,
3 2
revealed a preferential E conformation that locates H-1 and H-4, and/or
H1), 5.45 (m, 1H, H3), 7.35-7.81 (m, 10H, aromatic protons).
2
13
H-2 and H-5 at such a distance that no nOe effects could be observed.
The stereochemistry of 9 was chemically determined as follows: when
compound 8 was reacted with m-trifluoromethylbenzoic chloride the
compound obtained was identical to that synthesized under Mitsunobu
conditions 9.
C-NMR data (CDCl ): δ 24.00 (CH ), 24.07 (CH ), 39.22 (C2),
3
3
3
61.70 (d, J = 169 Hz, OCH P), 64.22 (C5), 71.07, 70.96 (d, J
C,P
2
C,P
= 6 Hz, 2 x OCH(CH ) ), 74.61 (C3), 81.57 (C4), 104.92 (d, J
3 2
C,P
= 12 Hz, C1), 128.7-133.24 (C-aromatics), 166.14, 166.40 (2 x
C=O). Anal. Calcd. for C O P: C, 60.00; H, 6.39. Found: C,
H
26 33
9
60.25; H, 6.58. b) Physical data of 5β. [α] -26.7 (c 1, methanol).
D
1
H-NMR data (CDCl ): δ 1.24-1.42 (m, 12H, CH ); 2.35 (m, 1H,
3
3
In summary, we have synthesized, for the first time, phosphonate
derivatives of 2-deoxyribose under mild conditions where no
decomposition of these acid-sensitive 2-deoxysugars has been observed.
Both in the synthesis of 2 and 3, the corresponding α-anomers were
obtained as the major isomers. On the other hand, we have also
observed unexpected results in the esterification of 1,3,5-tri-O-benzoyl-
α-D-ribofuranose under Mitsunobu conditions that, in our hands, indeed
proceed with full retention. This result should be kept in mind when
using derivatives of this easily available and commonly used sugar to
direct the stereochemistry at the anomeric center.
H ), 2.65 (m, 1H, H ), 3.65-4.11 (m, J = 7.2, J = 11.2 Hz, 2H,
2
2
OCH P), 4.52 (m, 3H, H4, 2H5), 4.63-4.85 (m, 2H, OCH(CH ) ),
2
3 2
5.48 (dd, J = 3.3, J = 1.2 Hz, 1H, H1), 5.61 (m, 1H, H3), 7.35-8.20
13
(m, 10H, aromatic protons). C-NMR data (CDCl ): δ 23.89
3
(CH ), 23.98 (CH ), 39.12 (C2), 61.67 (d, J = 168 Hz, OCH P),
3
3
C,P
2
65.25 (C5), 71.00, 71.22 (d, J = 6.5 Hz, 2 x OCH(CH ) ), 75.41
3 2
(C3), 82.10 (C4), 105.37 (d, J = 11.2 Hz, C1), 128.38-133.29
C,P
(C-aromatics), 165.95, 166.07 (2 x C=O). Anal. Calcd. for
C
H O P: C, 60.00; H, 6.39. Found: C, 59.98; H, 6.53.
26 33 9
(8) NOESY experiments carried out with compound 5α showed that
the signal corresponding to H-2a (δ 2.54) correlates with the
signals of both H-3 and H-1, indicating that H-2a, H-3 and H-1
must be in the same side of the molecule, and therefore 5α was
assigned as the α-anomer. On the other hand, NOESY
Acknowledgements. The authors would like to thank the European
Commission for financial support and Mrs. Ria Van Berwaer for
excellent technical assistance. We are very grateful to Dr. C. Rizzo for
helpful discussions.

February 1998
SYNLETT
179
1
experiments in the minor isomer 5β showed a correlation peak
(13) Physical data of 3 (mixture of α and β anomers). H-NMR data
between H-2a (δ 2.60) and H-3, while H-2b (δ 2.35) correlates
with the signal of H-1. This indicates that H-1 and H-3 must be in
opposite sides of the molecule, and therefore 5β was assigned as
the β-anomer.
(D O): δ 1.96 (m, J
= 8.0, J
= 11.0, J
= 6.1 Hz, 1H, H2α), 2.07
2
H1,H2a
H1,H2b
H2a,H3
(m, J
= 6.6, J
= 5.6, J
= 2.0 Hz,
H1,H2a
H2a,H3
H2b,H3
2H, 2H2β), 2.41 (m, J
= 7.9, J
= 7.0 Hz, 1H, H2α),
H1,H2b
H2b,H3
3.45 (dd, J
= 5.6, J
= 12.2 Hz, 1H, H5β), 3.49 (dd,
H5a,H5b
H4,H5a
J
= 5.8, J
,
= 12.2 Hz, 1H, H5α), 3.53 (dd, J
=
(9) The addition of an excess of 2,6-lutidine is to prevent any possible
anomerization or decomposition of the 2-deoxysugars under
treatment with TMSiBr. See: Pérez-Pérez, M.J.; Balzarini, J.;
Rozenski, J.; De Clercq, E.; Herdewijn, P. Bioorg. Med. Chem.
H4,H5a
H5a H5b
H4,H5b
4.1 Hz, 1H, H5β), 3.59 (dd, J
(ddd, J
H4β), 4.05 (dd, 1H, H1α), 4.09 (dd, 1H, H1β), 4.13 (ddd, 1H,
H3α), 4.24 (ddd, 1H, H3β). MS (FAB-thygl) 197.0 (M-H) .
HRMS (gly) Calcd. for C H O P (M-H) 197.0215, Found
197.0222.
= 3,6 Hz, 1H, H5α), 3.83
H4,H5b
= 5.2 Hz, 1H, H4α), 3.85 (ddd, J
= 2.3 Hz, 1H,
H3,H4
H3,H4
-
Lett. 1995, 5, 1115.
-
1
5
10 6
(10) a) Physical data of 2α. [α] +69.4 (c 1, H O). H-NMR data
D
2
(D O): δ 1.85 (dd, J
= 14.9, J = 1.3 Hz, 1H, H2a),
H2a,H3
2
H2a,H2b
(14) Park, M.; Rizzo, C.J. J. Org. Chem., 1996, 61, 6092.
2.18 (ddd, J
= 5.1, J
= 7.7 Hz, 1H, H2b), 3.37 (dd,J =
H2b,H3
H1,H2b
9.9, J = 13.0, 1H, OCH P), 3.46-3.61 (m, J
= 5.1, J =
H4,H5b
(15) Purchased from Pfanstiehl (Europe) Ltd. It can also be prepared
according to the procedure described by Brodfuehrer, P.R.;
Sapino, C. Jr.; Howell, H.G. J. Org. Chem. 1985, 50, 2597.
2
H4,H5a
3.6, J
= 12.3 Hz, 2H, 2H5), 3.70 (dd, 1H, OCH Ph), 3.97
2
H5a,H5b
(m, J
= 5.5 Hz, 1H, H4), 4.09 (m, 1H, H3), 5.11 (d, 1H, H1).
H3,H4
-
-
MS (FAB-thygl) 227.1 (M-H) , 249.1 (M-2H+Na) . HRMS
(16) It has been described that the addition of an exogenous base may
increase the rate of Mitsunobu condensations under alcohol-
limiting conditions [Campbell, D.A.; Bermak, J.C. J Org. Chem.
1994, 59, 658]. In this particular case, the presence of a hindered
base such as 2,6-lutidine, can also prevent the migration of
benzoyl groups in the starting material [Ness, R.K.; Fletcher,
-
(gly:thygl) Calcd. for C H O P (M-H) 227.0320, Found
6
12 7
1
227.0313. b) Data of 2β. [α] -36.5 (c 1, H O). H-NMR data
D
2
(D O): δ 2.11 (ddd, J
= 14.2, J
= 5.2, J
= 5.2
2
H2a,H2b
H1,H2a
H2a,H3
Hz, 1H, H2a), 2.25 (ddd, J
= 2.2, J
= 6.6 Hz, 1H,
H2b,H3
H1,H2b
H2b), 3.46-3.72 (m, J
= 4.0, J
= 6.8 Hz, 4H, 2H5,
H4,H5b
H4,H5a
OCH P), 3.91 (ddd, J
= 4.0 Hz,1H, H4), 4.31 (ddd, 1H, H3),
2
H3,H4
H.G.Jr. J. Am. Chem. Soc. 1956, 78, 4710].
5.23 (dd, 1H, H1). MS (FAB-thygl) 227. HRMS (thygl) Calcd. for
1
(17) Physical data of 11. [α] +11.8 (c 1, CHCl ). H-NMR data
D
3
-
C H O P (M-H) 227.0320, Found 227.0310.
6
12 7
(CDCl ): δ 1.34 (m, 12H, CH ), 3.90 (m, 2H, OCH P), 4.70 (m,
3
3
2
(11) Meuwly, R.; Vasella, A. Helv. Chim. Acta 1986, 69, 25.
5H, H5, H4 OCH(CH ) ), 5.46 (s, 1H, H1), 5.74 (d, J = 4.8 Hz,
,
3 2
1
(12) Physical data of 7 (5:2, mixture of α and β anomers). H-NMR
data (Benzene d ): δ 1.34 (m, 12H, CH ), 2.10-2.58 (m, 2H, H2),
1H, H2), 5.85 (m, 1H, H3), 7.26-8.07 (m, 14H, aromatic protons).
-
- -
EM (FAB-nitrobenzyl alcohol) 708 (M) , 754 (M+NO ) .
6
3
2
1
4.19 (t, H1α), 4.48-4.59 (m, 2H5, H1β, H4β), 4.62-4.90 (m,
(18) Physical data of 12. H-NMR data (CDCl ): δ 1.35 (m, 12H,
3
OCH(CH ) , H4α), 5.28 (m, H3α), 5.44 (d, J = 7.6 Hz, H3β),
3 2
CH ), 2.03 (m, 4H, H2, H3), 3.67 (m, 2H, OCH P), 4.34 (m, 1H,
3 2
H4), 4.43 (m, 2H, H5), 5.73 (m, 2H, OCH(CH ) ), 5.21 (d, J = 2.8
Hz, 1H, H1), 7.40-8.10 (m, 5H, aromatic protons).
13
6.96-8.30 (m, 10H, aromatic protons). C-NMR data (Benzene
3 2
d ): δ 24.06 (CH ), 33.88 (C2α), 34.22 (C2β), 64.09 (C5α), 64.65
6
3
(C5β), 70.46, 71.36 (d, J = 7 Hz, 2 x OCH(CH ) ), 73.82 (d, J =
C,P
3 2
(19) Hughes, D.L. in Organic Reactions: Paquette, L.A., Ed., John
Wiley; New York, 1992, 42, 344.
168 Hz, C1α), 74.20 (d, J = 175, C1β), 75.61 (d, J = 1.5 Hz, C3α),
76.90 (d, J = 10 Hz, C3β), 82.20 (d, J = 6.5 Hz, C4α), 83.96 (d, J =
9.5 Hz, C4β), 129.87-133.12 (C-aromatics), 165.58, 165.91,
(20) de Leeuw, F.A.A.M; Altona, C. J. Comput. Chem. 1983, 4, 428.
166.04, 166.11 (4 x C=O). Anal. Calcd. for C H O P: C, 61.22;
25 31 8
H, 6.37; Found: C, 61.05; H, 6.67.