Synthesis of the first homo-3′,5′-cyclic nucleotide

Article abstract of DOI:10.1039/a607699h

Syntheses of 2′,3′-dideoxy-3′-hydroxymethylcytidine-6′- diphenylphosphate and the corresponding 5′,6′-cyclicphenylphosphate from 5(S)-tert-butyldimethylsilyl-oxymethylfuran-2(5H)-one are achieved.

Full text of DOI:10.1039/a607699h

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis of the first homo-3A,5A-cyclic nucleotide
Jayne H. M. Gould and John Mann*
Department of Chemistry, Reading University, Whiteknights, Reading, UK RG6 6AD
Syntheses of 2A,3A-dideoxy-3A-hydroxymethylcytidine-6A-
diphenylphosphate and the corresponding 5A,6A-
cyclicphenylphosphate from 5(S)-tert-butyldimethylsilyl-
oxymethylfuran-2(5H)-one are achieved.
reacted with bis-trimethysilylcytosine in the presence of tin(iv)
chloride to yield the protected nucleoside analogue 7a as a ca.
1:1 anomeric mixture, which could not be separated by flash
chromatography. Selective removal of the benzoate was
In the search for drugs with activity against HIV reverse
transcriptase, novel analogues of the natural nucleosides have
always been the most popular targets for synthesis. Recent
discoveries include biologically potent l -series nucleosides1
and variously protected phosphates like 1, which are prodrugs
of AZT.2 Our synthesis of 2A,3A-dideoxy-3A-hydroxymethyl
nucleosides of general structure 2 demonstrated that the cytidine
derivative (B = cytosine) and 5-fluorocytidine derivative
(B = 5-fluorocytosine) had particularly potent activity against
a range of viruses,3 and this encouraged us to try to prepare the
monophosphate and cyclic phosphate derivatives for biological
evaluation. Here we describe the successful attainment of these
goals.
We have previously described the photoinduced addition of
methanol to 5(S)-tert-butyldimethylsilyloxymethylfuran-
2(5H)-one 3 to produce adduct 4,4 but for the present work this
reaction was carried out on the 30 g scale with yields in the
range 45–50%. Protection of the free hydroxy as its benzoate
ester 5a allowed selective reduction of the lactone (disiamyl
borane in THF) and derivatisation of the resultant lactols to
yield the anomeric acetates 6a (overall yield 52%). These were
O
Me
HN
B
O
O
O
N
HO
HO
O
P
O
RO
OR
2
4
N3
1 R = subst. aryl
or RCOSCH2CH2
O
O
ButMe2SiO
HO
O
O
O
ButMe2SiO
3
O
O
OAc
ButMe2SiO
RO
ButMe2SiO
RO
5a R = Bz
6a R = Bz
b R = ButMe2Si
b R = ButMe2Si
NH2
N
NH2
N
NH2
N
N
N
O
O
O
O
O
N
O
ButMe2SiO
ButMe2SiO
O
PhO
P
O
O
RO
7a,b
(PhO)2P(O)O
8
9
O
O
HN
O
NO2
HN
O
NO2
N
N
O
O
O
O
N
N
ButMe2SiO
ButMe2SiO
HO
HO
10
11
O
HN
O
NO2
NH2
N
N
N
O
O
O
N
O
O
O
PhO
PhO
P
P
O
O
O
O
12
13
Chem. Commun., 1997
243

View Article Online
J5A,P 24.9, Jgem 11.7, J5A,3A 3.7 Hz, 5A-H), 4.47 (t, J 6.6 Hz, 2 3 CH2-npeoc),
4.50–4.55 (m, 5A-H), 4.59 (ddd, J5A,P 19.2, Jgem 10.3, J5A,4A 4.2 Hz, 5A-H), 6.02
(t, J1A,2A 5.9 Hz, 1A-H), 6.09 (t, J1A,2A 6.6 Hz, 1A-H), 7.20–7.27 (m, 2 3 5-H,
4 3 Ar), 7.38 (t, J 7.7 Hz, 6 3 Ar), 7.46 (d, J 8.8 Hz, 4 3 Ar npeoc), 7.95
(d, J6,5 7.7 Hz, 6-H), 7.99 (d, J6,5 7.3 Hz, 6-H) and 8.19 (d, J 8.8 Hz, 4 3
Ar npeoc); dC (100.4 MHz, CDCl3 and CD3OD) 34.8 (CH2-npeoc), 35.7 (C-
2A), 36.6 (C-2A), 42.7 (C-3A), 42.8 (C-3A), 62.4 (C-6A), 65.0 (C-6A), 65.4 (CH2-
npeoc), 67.2 (C-5A), 68.4 (C-5A), 81.5 (C-4A), 84.4 (C-4A), 88.3 (C-1A), 88.7
(C-1A), 119.7 (2 3 C-5), 123.6 (Ar), 123.8 (Ar), 125.7 (Ar), 129.8 (Ar),
142.8 (Ar), 142.9 (Ar), 143.3 (C-6), 143.4 (C-6), 145.2 (C–NO2), 146.9 (C-
quart), 150.0 (C-quart), 152.8 (2 3 C-4), 155.6 (2 3 CNO, npeoc) and 163.2
(2 3 C-2).
For the b-anomer of 12: dH (400 MHz, CDCl3), integration values for the
discrete diastereoisomers have not been given but were in the approximate
ratio of 5:1 with respect to the signals for the discrete diastereoisomers,
2.29–2.43 (m, 4 3 2-HA), 2.56–2.64 (m, 2 3 3-HA), 3.11 (t, J 6.6 Hz, 2 3
CH2-npeoc), 4.09 (dd, Jgem 12.8, J6A,3A 10.3 Hz, 6A-H), 4.14–4.24 (m, 3 3 6A-
H), 4.32 (dd, Jgem 11.7, J5A,4A 3.7 Hz, 5A-H), 4.34–4.42 (m, 2 3 4A-H, 5A-H),
4.45 (t, J 6.6 Hz, 2 3 CH2-npeoc), 4.49–4.58 (m, 5A-H), 4.67 (ddd, J5A,P 18.3,
Jgem 10.6, J5A,4A 4.4 Hz, 5A-H), 6.05 (dd, J1A,2A 6.8, J1A,2A 1.6 Hz, 1A-H), 6.12 (d,
J1A,2A 7.0 Hz, 1A-H), 7.22 (d, J 7.7 Hz, 4 3 Ar), 7.32–7.38 (m, 2 3 5-H, 6 3
Ar), 7.41 (d, J 8.4 Hz, 4 3 Ar npeoc), 7.76 (d, J6,5 7.3 Hz, 6-H), 7.89 (d, J6,5
6.6 Hz, 6-H) and 8.17 (d, J 8.4 Hz, 4 3 Ar npeoc); dP (81 MHz, CDCl3)
25.2 and 25.7; m/z 572.1294 (M+) (C25H25N4O10P requires 572.1308).
‡ We thank A. Jahans and Dr M. G. B. Drew for the X-ray structure
determination, which will be described in detail elsewhere. Suffice to say,
the compound crystallised in an extended conformation.
§ Selected spectroscopic data for 13: dH (400 MHz, CD3OD), integration
values for the discrete diastereoisomers have not been given but were in the
approximate ratio of 2:1 with respect to the signals for the discrete
diastereoisomers, 2.16–2.21 (m, 4 3 2-HA), 2.71–2.78 (m, 2 3 3-HA), 4.0
(dd, J5A,P 20.9, Jgem 11.0 Hz, 5A-H), 4.05–4.17 (m, 2 3 4A-H, 2 3 5A-H), 4.23
(ddd, J6A,P 22.3, Jgem 10.3, J6A,3A 8.2 Hz, 6A-H), 4.32–4.44 (m, 3 3 6A-H), 4.49
(ddd, J5A,P 20.0, Jgem 10.3, J5A,4A 3.9 Hz, 5A-H), 5.81 (d, J5,6 7.7 Hz, 5-H), 5.83
(d, J5,6 7.7 Hz, 5-H), 5.99 (dd, J1A,2A 6.6, J1A,2A 3.7 Hz, 1A-H), 6.03 (dd, J1A,2A
6.9, J1A,2A 2.9 Hz, 1A-H), 7.15 (t, J 7.9 Hz, 6 3 Ar), 7.30 (t, J 7.9 Hz, 4 3 Ar),
7.57 (d, J 7.7 Hz, 6-H) and 7.59 (d, J 7.7 Hz, 6-H); m/z 380.1011 [(M + H)+]
(C16H19N3O6P requires 380.1011).
possible using base (1% NaOH in MeOH) and the free hydroxy
was selectively phosphorylated using a modification of the
method of Uchiyama et al.5 This involved reaction of the
nucleoside with 1 equiv. of tert-butylmagnesium chloride in the
presence of diphenylphosphorochloridate to produce the de-
sired 6A-diphenylphosphate 8 (63% yield). We had hoped to
convert this compound into the cyclic phosphate 9, but despite
considerable efforts, this goal eluded us.
We thus turned our attention to the bis-silylated adduct 5b
which was reduced (DIBAL-H) and derivatised to provide
lactol esters 6b and thence the protected nucleoside anomers 7b
(52% overall). Protection of the free amino group of the
cytosine as its 2-(p-nitrophenyl)ethyl carbamate6 yielded the
fully-protected nucleoside anomers 10, and to our delight the
anomers could now be separated (with relative ease) by flash
chromatography. (Indeed we have used this protecting group to
facilitate separation of anomeric mixtures of several substituted
cytidines.) The silyl ether groups of these individual anomers
were now removed (toluene-p-sulfonic acid in MeOH) and the
resultant diols 11 were reacted in pyridine with bis(benzotri-
azolyl)phenyl phosphate7 (from 2 equiv. of 1-hydroxy-
benzotriazole and phenylphosphodichloridate in the presence of
triethylamine), to produce the desired cyclic phosphates 12
(35% a and 31% b after chromatography.† The a-anomer was
more easily recrystallised than the b-anomer, and an X-ray
crystallographic study confirmed the structure of this com-
pound.‡ Finally, the carbamate group was removed from the b-
anomer using triethylamine to yield the novel homo-cyclic
nucleotide 13,§ which was submitted for biological evaluation.
We had hoped that the compound would act as a prodrug from
compound 2 (B = cytosine), a compound with good activity
against several viruses.3 It did in fact show modest activity
against HIV-1 (EC50 350 mmolar) and is presently being
evaluated in other systems.
All of the reactions as far as compound 11 have been carried
out routinely on the half-gram scale (though the cyclic
phosphate formation has only been accomplished thus far on the
100 mg scale). One obvious extension of this work is to replace
the phenyl group with other aryl groups or by S-acyl-2-thioethyl
groups as in prodrugs like 1, in an attempt to improve the
bioavailability of the compounds. However, given the pivotal
importance of cyclic-3A,5A-AMP and -GMP in biological
processes, a more urgent task is to extend this methodology to
the purine series in order to make homologues of these very
important compounds.
J. G. thanks the MRC for a studentship under the AIDS
Initiative Programme and we thank Dr Naheed Mahmood
(MRC) for the (ongoing) biological evaluation. We are also
very grateful for the advice given by Professor Colin Reese and
for the numerous FAB mass spectra provided by the EPSRC
Mass Spectrometry Centre at the University of Swansea.
References
1 J. W. Beach, L. S. Jeong, A. J. Alves, D. Pohl, H. O. Kim, C.-N. Chang,
S.-L. Doong, R. F. Schirazi, Y.-C. Cheng and C. K. Chu, J. Org. Chem.,
1992, 57, 2217; T.-S. Lin, M.-Z. Luo, M.-C. Liu, Y.-L. Zhu, E. Gullen,
G. E. Dutschman and Y.-C. Cheng, J. Med. Chem., 1996, 39, 1757;
T.-S. Lin, M.-Z. Luo and M.-C. Liu, Tetrahedron, 1995, 51, 1055; T.-
S. Lin, M.-Z. Luo, M.-C. Liu, S. B. Pai, G. E. Dutschman and Y.-
C. Cheng, J. Med. Chem., 1994, 37, 798.
2 X. Pannecouke, G. Parmentier, G. Schmitt, F. Dolle and B. Luu,
Tetrahedron, 1994, 50 1173; I. Lefebre, C. Perigaud, A. Pompon,
A.-M. Aubertin, J.-L. Girardet, A. Kirn, G. Gosselin and J.-L. Imbach,
J. Med. Chem., 1995, 38, 3941; C. McGuigan, R. N. Pathirana,
J. Balzarini and E. De Clerq, J. Med. Chem., 1993, 36, 1048;
C. McGuigan, R. N. Pathirana, M. P. H. Davies, J. Balzarini and E. De
Clerq, Bioorg. Med. Chem. Lett., 1994, 4, 427.
3 J. Mann and A. C. Weymouth-Wilson, J. Chem. Soc., Perkin Trans. 1,
1994, 3141.
4 J. Mann and A. C. Weymouth-Wilson, Synlett, 1992, 67.
5 M. Uchiyama, Y. Aso, R. Noyori and Y. Hayakawa, J. Org. Chem., 1993,
58, 373.
6 F. Himmelsbach, B. S. Schulz, T. Trichtinger, R. Charubala and
W. Pfleiderer, Tetrahedron, 1984, 40, 59.
Footnotes
† Selected spectroscopic data for the a-anomer of 12: dH (400 MHz, CDCl3
and CD3OD), integration values for the discrete diastereoisomers have not
been given but were in the approximate ratio of 2:1 with respect to the
signals for the discrete diastereoisomers, 1.75–1.84 (m, 2 3 2A-H),
2.45–2.51 (m, 2A-H), 2.80–2.87 (m, 2 3 3A-H), 2.93–2.99 (m, 2A-H), 3.14 (t,
J 6.6 Hz, CH2-npeoc), 3.51 (dd, Jgem 11.0, J6A,3A 6.6 Hz, 6A-H), 3.59 (dd, Jgem
11.0, J6A,3A 5.1 Hz, 6A-H), 3.76 (dd, Jgem 11.0, J6A,3A 4.4 Hz, 6A-H), 3.82 (dd,
Jgem 11.0, J6A,3A] 4.0 Hz, 6A-H), 4.12–4.25 (m, 5A-H, 2 3 4A-H), 4.40 (ddd,
7 G. van der Marel, C. A. A. van Boeckel, G. Wille and J. H. van Boom,
Tetrahedron Lett., 1981, 22, 3887.
Received, 12th November 1996; Com. 6/07699H
244
Chem. Commun., 1997