A. Patel et al. / Tetrahedron 71 (2015) 7339e7343
7341
2. Results and discussion
synthesized by introducing the fluorine atom at an earlier step. The
-fluoro- -ketoester 13, prepared by nucleophilic fluorination of
the corresponding -diazo-
-ketoester,20 underwent cyclization
a
b
2.1. 6-Arylisocytosines
a
b
with guanidine carbonate to give 5-fluoro-6-phenylisocytosine 14
6-Substituted isocytosines are readily available by condensation
albeit in only 11% yield. A methyl substituent at the C-5 position
of
b
-ketoesters with guanidine. Thus 6-phenylisocytosine 6 was
was introduced in a similar manner starting from a-methyl-b-
synthesized by reaction of guanidine carbonate with ethyl ben-
zoylacetate in 70% yield.15 Likewise, the known 4-nitrophenyl de-
rivative 9 and its 3-nitrophenyl analogue 10 were obtained from the
ketoester 15 and gave the 5-methyl-6-phenylisocytosine 16, again
in poor yield.
corresponding
b
-ketoesters (Scheme 1).16 In addition to the aryl
2.3. Linked 6-arylisocytosines
ring at the 6-position of isocytosine, the incorporation of ferrocene
was investigated. Ferrocene derivatives have been of interest in
medicinal chemistry in recent years, including ferrocene-linked
thymine and uracil derivatives, which undergo supramolecular
packing through hydrogen bonding.17 The synthesis of 6-
ferrocenylisocytosine 11 was described by Cooke and co-workers
in 2007, and their methodology was followed to give the iso-
cytosine 11 in 44% (Scheme 1).18
Two families of linked isocytosines were investigated, ligated
either through the 6-aryl substituent or through a urea group.
Two examples of each were prepared starting from the appro-
priately substituted
b-ketoesters. 1,4-Diacetylbenzene was used
as a starting point for the synthesis of bis-isocytosine 18 in two
steps (Scheme 3). Treatment of the diketone with sodium hy-
dride and diethyl carbonate gave a 1:0.4 mixture of the bis-b-
ketoester 17 and its mono-enolized tautomer in 50% yield. Sub-
sequent reaction with guanidine carbonate produced bis-
isocytosine 18 in 66%. The same chemistry starting with 1,3-
diacetylbenzene was unsatisfactory, so isophthalic acid was
converted into bis-b-ketoester 19 by reaction with CDI, followed
by displacement of the acylimidazole with potassium monoethyl
malonate in 59% yield. 1H NMR spectroscopy showed 19 to exist
in a keto-enol mixture that underwent cyclization with guani-
dine carbonate to give benzene-1,3-bis-isocytosine 20 in 56%
yield (Scheme 3).
A similar strategy involving late stage introduction of the
heterocyclic ring from a suitable
b-ketoester was used for the
urea linked isocytosines 23 and 8 (Scheme 4). Commercially
available ethyl 4-nitrobenzoylacetate and the 3-nitro analogue
were subjected to catalytic hydrogenation as described in the
literature for the 4-nitro compound.21 However, this resulted in
reduction of both the nitro group and the ketone to the corre-
sponding alcohol, as confirmed by 1H and 13C NMR spectroscopy
and mass spectrometry. Reduction of ethyl 4-nitrobenzoylacetate
to the desired ethyl 4-aminobenzoylacetate 21 with iron and
ammonium chloride proceeded in 73% yield, but the same con-
ditions with ethyl 3-nitrobenzoylacetate gave a disappointing
11% of ethyl 3-aminobenzoylacetate 24. Instead, sodium
dithionite was employed to give the desired aniline 24 in
Scheme 1. Synthesis of 6-arylisocytosines.
2.2. 5-Substituted-6-arylisocytosines
a modest 38% yield. The anilino b-ketoesters 21 and 24 were then
coupled with CDI to give ureas 22 and 25. The final step of iso-
cytosine formation proceeded in 40% and 63% for the 1,4- and
1,3-substituted aryl isomers, respectively, to give the desired urea
linked isocytosines 23 and 8 (Scheme 4). The 1,3-substituted
isomer 8 is of particular interest due to its potential to form
hydrogen bonds analogous to bropirimine (Fig. 3C), which would
lead to a cyclic dimer reminiscent of ureidopyrimidinones sys-
tems previously reported.22 These cyclic structures would be of
interest if they prove to be effective as DNA G-quadruplex li-
gands, as they provide a means to an extended, near-planar
network from a relatively small organic compound.
Bromination of 6-phenylisocytosine 6 using bromine in acetic
acid gave bropirimine 7 in 60%,16 while chlorination with N-chlor-
osuccinimide was achieved in 58% to give 5-chloro analogue 12
(Scheme 2).19 The corresponding 5-fluoro analogue was
2.4. X-ray crystallography
One aim of this project was to see whether the hydrogen
bonding behaviour of bropirimine 7 could be replicated.14 Un-
fortunately, single crystals of our new isocytosines proved very
difficult to grow due to the poor solubility of these compounds
in a variety of organic solvents, and only two examples were
obtained. 6-Phenylisocytosine 6 was crystallized from formam-
ide, and shown to exist in the same lattice that was described by
Fowler et al., with hydrogen bonding between 1H-tautomers
only (Fig. 3A).13 In contrast, 6-(4-nitrophenyl)isocytosine 9 was
Scheme 2. Synthesis of 5-substituted-6-phenylisocytosines.