Synthesis of Janus type nucleoside analogues and their preliminary bioactivity

Article abstract of DOI:10.1039/c0ob00495b

Novel Janus type nucleoside analogues 1a and 1b were synthesized in seven steps from 2-amino-4,6-dihydroxypyrimidine and 4,6-dihydroxypyrimidine. The base moiety of 1a has one face with a Watson-Crick donor-donor-acceptor (DDA) H-bond array of guanine and

Full text of DOI:10.1039/c0ob00495b

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 1516
www.rsc.org/obc
PAPER
Synthesis of Janus type nucleoside analogues and their preliminary
bioactivity†
Hao-Zhe Yang, Mei-Ying Pan, Da-Wei Jiang and Yang He*
Received 27th July 2010, Accepted 11th November 2010
DOI: 10.1039/c0ob00495b
Novel Janus type nucleoside analogues 1a and 1b were synthesized in seven steps from
2-amino-4,6-dihydroxypyrimidine and 4,6-dihydroxypyrimidine. The base moiety of 1a has one face
with a Watson–Crick donor-donor–acceptor (DDA) H-bond array of guanine and the other face with
an acceptor-acceptor-donor (AAD) array of cytosine, which might lead to its base pairing with either
cytosine or guanine due to the rotating of the glycosyl bond. This property may enable Janus type
nucleoside analogues to act as an antiviral compound in a similar way to ribavirin. Both 1a and 1b were
screened by a vitro HBV DNA replication inhibition test and indeed 1a showed a great potential with
IC50 = 10 mM and SI = 78.9 for antiviral drug development.
to the environments changed by immune response and antiviral
Introduction
drugs.12,13 On the other hand, the higher frequency of mutations in
the viral genome also implies a higher risk of genetic melt-down.14
The maintenance of genetic information of any organism requires
a minimal fidelity level during the process of genome replication.
Excess mutations beyond the upper limit (error threshold) of the
genome would cause the irreversible loss of genetic information
and consequently the loss of specific infectivity of viruses.15
Ribavirin was first synthesized as an ambiguous purine base
analogue.16 Its pseudo-base moiety pairs with cytidine and uridine
by rotating the carboxamide bond (Fig. 1). Crotty and his co-
workers have proved that ribavirin can be templated by either
uridine or cytidine with equal efficiency and it is actually an
RNA virus mutagen.10 Therefore, the dual base pairing property
of ribavirin will cause transition or transversion mutations (A
to G and C to U) and finally lead to its antiviral activity via
lethal mutagenesis by pushing the viral genome beyond the error
threshold.17
However, the efficiency of ribavirin’s incorporating into a viral
genome is relatively low.9 Searching new mutagenic nucleoside
analogues which can be specifically and efficiently incorporated
into a viral genome is a plausible strategy to develop new drugs
for viral disease. Inspired by this consideration, we designed a
Janus guanine-cytosine nucleoside (1a, J-GC) for this purpose.
Janus molecules (from the two-faced Roman god Janus) were first
proposed by Cristol to describe a new symmetrical carbocyclic
system.18 The base moiety of J-GC has one face with a Watson–
Crick donor-donor-acceptor (DDA) H-bond array of guanine and
the other face with an acceptor-acceptor-donor (AAD) array of
cytosine. Rotating the glycosyl bond of J-GC will lead to an anti
or syn conformation (Fig. 2). In principle, J-GC could pair with
cytidine or guanosine via rotating the glycosyl bond in a similar
way to ribavirin’s pairing with U/C via rotating the carboxamide
bond (Fig. 1). J-GC is a mimic of natural nucleosides which
Nucleoside analogues have been used as antiviral agents, antineo-
plastic drugs and immunosuppressive molecules for many years.
This diverse family is considered to compete with physiological
nucleosides through interaction with a large number of intra-
cellular targets. After being transported into a cell by different
membrane carriers like natural nucleosides, these compounds are
usually transformed into phosphorylated derivatives1 and then
behave as antimetabolites via different mechanisms: inhibiting
various enzymes which participate in the synthesis of nucleic acid;
incorporating and altering natural DNA/RNA macromolecules
and terminating the synthesis of the genome.2 Among these nu-
cleoside analogues, ribavirin (1-b-D-ribofuranosyl-1,2,4-triazole-
3-carboximide, also named as virazole) is of particular interest
due to its unique antiviral mechanism.3,4 Nowadays, ribavirin is
used as a preferred drug combined with interferon-a to treat HCV
infection,5,6 single therapy for Lassa fever infections7 and acute
RSV infection.8 However, the exact mechanisms of ribavirin’s
activity has remained controversial since its discovery 30 years
ago. Several distinct indirect or direct mechanisms have been
proposed to explain its antiviral properties.9 Crotty and colleagues
have recently demonstrated that ribavirin’s antiviral activity can
be fully attributed to its lethal mutagenesis.10 Viruses, particularly
RNA viruses, exist as quasispecies, a heterogeneous population
with a master sequence.11 On one hand, the great variability in
genomes of these quasispecies will help viruses to adapt quickly
Laboratory of Ethnopharmacology, Institute for Nanobiomedical Technol-
ogy and Membrane Biology, Regenerative Medicine Research Center, West
China Hospital, West China Medical School, Sichuan University, Chengdu
610041, China. E-mail: heyangqx@yahoo.com.cn; Tel: +86 2885164077
† Electronic supplementary information (ESI) available. CCDC reference
numbers 785478. For ESI and crystallographic data in CIF or other
electronic format see DOI: 10.1039/c0ob00495b
1516 | Org. Biomol. Chem., 2011, 9, 1516–1522
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig. 1 Ribavirin is an ambiguous purine mimic which pairs with cytidine and uridine. The designed J-GC nucleoside analogue is supposed to pair with
either guanosine or cytidine through the specific pattern of hydrogen-bonding patterns (R = ribose).
intermolecular hydrogen bonding interaction may ameliorate this
characteristic. Therefore, another Janus type nucleoside analogue
(1b, J-HC) with a hypoxanthine-cytosine system has also been
prepared (Fig. 2).26–28
Results and discussion
Chemical synthesis
Fig. 2 Ribavirin, J-GC and J-HC nucleoside analogues. The rotation of
There are two major approaches to the synthesis of nucleosides.
glycosyl bond will lead to anti or syn conformation.
The first approach employs the corresponding base derivatives
reacting with an appropriate sugar component (convergent gly-
preserves the pyrimidine system of the canonical base unit and
we hope it could be recognized by viral enzymes more efficiently.
While Janus nucleobase analogues19–24 have been researched,
most studies about these bisfunctionalized molecules were focused
on their applications in nanotechnology: to construct architectures
such as trimers,22 rosettes7,25 and regular noncovalent polymer
arrays19 with different hydrogen bonding codes. Janus guanine-
adenine base analogue was first synthesized by Branda and
colleagues as a non-self-complementary molecule to avoid self-
aggregation.20 This guanine-adenine system was designed as a
base wedge to form cytosine-uracil mismatches, and its interaction
cosylation reaction). The second one involves the construction
of a purine or pyrimidine system from a simple N-glycosylated
precursor (linear method). Initially, we intended to adopt direct
glycosylation reaction through the Vorbrueggen method using cor-
responding silylated Janus type base moieties. However, it was too
difficult to get free bases from some intermediates we synthesized.
For instance, the similar pyrimido[4,5-d]pyrimidine compounds
contain alkyl chains attached to the nitrogen atom,22 which are
hard to remove. In addition, even if we had obtained the free bases,
the four ring nitrogen atoms on the pyrimido[4,5-d]pyrimidine
ring would raise a big problem of regioselectivity during the
glycosylation reaction, because the multiple N-glycosylation sites
will lead to several isomers. For all these reasons, we finally
chose to synthesize compounds 1a and 1b through the route of
Scheme 1. The starting materials 2a and 2b are obtained from
commercially available 2-amino-4,6-dihydroxypyrimidine and 4,6-
dihydroxypyrimidine via Vilsmeier–Haack reaction.29,30 Then,
both 2a and 2b were treated with hydroxylamine hydrochloride
(1.5 equiv.) in acetic acid. After being stirred overnight at room
temperature, the oxime derivative 3a was easily obtained as a
precipitate. Then, acetous impurities were washed out by EtOH
and the pure product 3a was dried in vacuum for further use in
a yield of 83%. However, when 2b was reacted under the same
conditions, there was no precipitate formed in solution. Alterna-
tively, the solution was evaporated and the residue was treated
with ethyl ether. After the insolubles were filtered off, the organic
solution was washed with saturated aqueous sodium bicarbonate.
1
with both U and C was observed by H NMR titration. Then
the same guanine-adenine wedge was proved to also recognize
C-T mismatch.23 After that, Janus guanine-cytosine,24 Janus
adenine-thymine21 base analogues were reported. These two self-
complementary molecules were mostly studied for their self-
association and nano-structure formations.
In order to evaluate the base pairing properties of these two-
hydrogen-pattern-faced systems in RNA/DNA and to see if they
can be used as effective antiviral drugs, their nucleoside derivatives
would be needed. Surprisingly, the synthesis of Janus type
nucleosides with the aforementioned pyrimidopyrimidine as base
moieties has not been reported. Herein, we describe the syntheses
and potential application in antiviral treatment of a Janus type
nucleoside analogue J-GC (1a). Because of its hydrogen self-
complementarity, J-GC easily self-associates in solutions leading
to bad solubility in different kinds of solvents. Reducing one
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1516–1522 | 1517
©

View Article Online
Scheme 1 Sythesis of J-GC and J-HC nucleosides: I. Hydroxylamine hydrochloride, acetic acid, r.t., 16 h. II. SeO2, pyridine (3a), toluene, reflux, 15 min
for 3a and 180 min for 3b. III. iPr2EtN, phthaloyl dichloride, dry THF, r.t., 20 min. IV. 2,3-O-isopropylidene-D-ribofuranosylamine toluene-p-sulfonate,
NaHCO3, iPr2EtN, dry THF, 70 ◦C, 40–60 min. V. iPr2EtN, Ac2O, DMAP, DCM, r.t., 7 min. VI. iPr2EtN, CCI, dry DCM, r.t., 10–30 min; NaOMe,
reflux, 70 min. VII. NaI, TMSi-Cl, acetone, r.t., 20 h, CF3COOH (10% H2O), 0 ◦C, 20 min.
The washing process should be carried out thoroughly otherwise
the traces of remaining acid would catalyze the decomposition of
the product very fast. It was not practicable for further purification
via column chromatography due to the lack of a suitable solvent
system with reasonable solubility of 3b. Fortunately, the unpurified
compound 3b was acceptable for next step.
Strangely, the attempt to dehydrate the aldoxime to give
nitrile failed under the standard acidic or alkaline conditions.
Alternatively, the mild method using selenium dioxide in refluxing
toluene was found to be suitable for the transformation leading to
compounds 4a and 4b. When 3a was reacted with selenium dioxide
(1 equiv.) in boiling toluene, with pyridine (0.5%) added as a base
catalyst, product 4a was formed as pink solid in 73% yield within
15 min. When 3b was treated with selenium dioxide (1.2 equiv.)
in refluxing toluene, no appropriate catalyst was found. So the
reaction lasted longer (~3 h) compared to the formation of 4a and
the reaction remains as a homogeneous solution. The solvent was
evaporated and product 4b (yield 82%) was obtained as a white
powder after column chromatography.
When 4a was treated with 2,3-O-isopropylidene-D-
ribofuranosylamine toluene-p-sulfonate31 directly, no reaction was
detected. We postulated that this is caused by the electron pair of
the 2-amino group which deactivated the electrophilicity at the 6-
position of 4a. To overcome this problem, the amino group should
be protected properly. Finally, compound 4a was treated with
both phthaloyl chloride in dry THF and iPr2EtN as base catalyst
and compound 5a was obtained with a yield of 60% after flash
column chromatography. As expected, protected pyrimidine 5a
can react with 2,3-O-isopropylideneribofuranosylamine toluene-
p-sulfonate successfully, leading the formation of compound 6a
with a yield of 86%. Because compound 4b has no amino group at
the 2-position, it can react with the ribofuranosylamine derivative
directly and compound 6b was obtained in 79% yield via similar
procedures for the preparation of 6a.
The anomeric configurations of compounds 6a and 6b were
firstly determined by the difference in 1H chemical shifts between
the two methyl signals of the isopropylidene group, Dd. According
to Imbach’s rules:32–34 when 0 < Dd < 0.10 ppm, the nucleoside
is in the a-configuration, and when 0.17 < Dd < 0.23 ppm, the
nucleoside is in the b-configuration. The calculated results of both
compounds (0.18 ppm. for 6a and 0.17 ppm. for 6b) suggest that
they are more likely in b-configurations. Then, the single crystals
of these two compounds were obtained from methanol. By single
crystal X-ray analysis, we confirmed both of the nucleosides adopt
b-configurations. The oblique view of crystal structure of 6b is
given here as an example (Fig. 3).
The 5¢-OH groups of compounds 6a and 6b were protected
with acetic anhydride (1.2 equiv.) and the resulting protected
nucleosides 7a and 7b were then treated with N-(chlorocarbonyl)
isocyanate (CCI) to give their urea derivatives. Compared to
compound 7a, 7b reacts with CCI much slower. Due to the
neighboring effect of the oxygen of ribose, the urea group of 7a
and 7b was unstable and decomposed in methanolic ammonia.
We have tried several conditions to carry out the ring cyclization
reaction of 7a and 7b and found NaOMe in methanol was the
only way to get the ring closure product (0.5 M for 6a; 0.15 M
for 6b). During the cyclization reaction both the phthaloyl- and
acetyl-group were removed and the 6-chloride on the base moiety
1518 | Org. Biomol. Chem., 2011, 9, 1516–1522
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Table 1 Inhibition of HBV DNA replication by compounds 1a, 1b and
3TC
Compound
IC50/mM
TC50/mM
SI
1a
10
>1000
0.17
793
3200
>437
79.3
NA
>2500
1b
3TC
SI = TC50/IC50; NA: non-applicable
Antiviral activity
The antiviral activity of the compounds was preliminarily screened
by a HBV DNA replication inhibition assay.36 The cell line
HepG2.2.15 was previously cultured in the standard condition.
Then cells were seeded in a 96-well tissue culture plate for 24 h,
treated with test compounds and positive reference compound
in gradient concentration (0 days). After 4 days treatment, the
culture medium was removed and the test compounds were added
to the cultures again in fresh medium. At day 8, the cultures
were collected and lysed for the intracellular HBV DNA analysis.
The amount of HBV DNA in each condition was measured by
real time PCR. Also, the half maximum cytotoxic concentration
values (TC50) for each compound were measured by calculating
the cell growth rates by a standard procedure.37 In the test,
Lamivudine (3TC) was used as the positive reference compound.
The cytotoxicity was monitored with compound concentrations
up to 3 mM. The half maximum inhibition concentration values
(IC50) and the selected index (SI) were also estimated for all
compounds (Table 1).
Fig. 3 The oblique view of the crystal structure of 6b.
The results show that compound 1a really inhibits the viral DNA
replication with an IC50 of 3.3 mg ml-1. On one hand, compared
to 3TC, the inhibition effect of compound 1a is moderate. The
high efficiency of DNA replication inhibition for 3TC is caused
by the lack of a 3¢-OH. When 3TC is incorporated into viral
DNA, the absence of 3¢-OH will terminate the 3¢–5¢ extension
of the proviral DNA chain. In our case, compound 1a possesses
an entire ribose structure, therefore its effect of inhibiting viral
DNA replication is not due to the chain termination. On the
other hand, compound 1a shows higher efficiency to inhibit viral
DNA replication than ribavirin, which showed no inhibition effect
for HBV DNA replication at a notable high concentration of
50 mg ml-1 under the same assay conditions.38 The assay shows
compound 1b is devoid of viral DNA replication inhibition effect,
which also implies that the canonical cytosine and guanine face in
Janus type nucleoside analogues is recognized by enzymes more
efficiently. According to these results, the Janus GC compound
has strong potential in the application of antiviral drug research.
Its base-pairing properties in oligonucleotides and its mechanism
for DNA replication inhibition are under further investigation.
was replaced by a methoxy group; compounds 8a and 8b were
obtained as products.
The simultaneous removal of both the latent carbonyl function
group and isopropylidene group proceeded as follows: 8a/8b was
dissolved in acetonitrile, NaI (3 equiv.) and TMSi-Cl (3 equiv.)
were added and stirred for 20 h at room temperature. The yellowish
product (yield 95%) was filtered and washed with acetonitrile and
acetone. The product was then added to a solution of CF3COOH
with 10% H2O at 0 ◦C, and reacted for 20 min. After the acid
was evaporated in vacuo, the residue was then co-evaporated with
ethanol several times to remove residual acid. The obtained black
residue was dissolved in hot water, after a few hours a white
precipitate was obtained as our final products 1a and 1b (both
yield 91%).
Also, to estimate the preferred conformation and the energy
differences between syn and anti conformers, we performed
density functional theory (DFT) calculations on 1a and 1b. The
geometries were optimized at the B3LYP/6-31g (d, p) level. The
computational result indicated that the anti conformations are
more stable than the syn conformations with energy differences of
20.49 kJ mol-1 for 1a and 16.90 kJ mol-1 for 1b. These predicted
energy differences are rather close to the energy barriers between
the different syn and anti conformational states of natural purine
and pyrimidine nucleosides (about 25 kJ mol-1).35 These results
indicate that compounds 1a and 1b may rotate their glycosyl
bonds freely to adopt both syn and anti conformations just as
natural nucleosides, which will very likely lead to the formation of
mismatches in viral genomes as supposed.
Conclusion
We have developed a viable route to two novel Janus type nucleo-
side analogues of 4,6-diamino-1-(b-D-ribofuranoyl)pyimido[4,5-
d]pyrimidin-2,5(1H,6H)-dione (J-GC) and 4-amino-1-(b-D-
ribofuranoyl)pyimido[4,5-d]pyrimidin-2,5(1H,6H)-dione (J-HC).
These two compounds are designed as genome mutagens and
their preliminary antiviral activity test shows that only the J-GC
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1516–1522 | 1519
©

View Article Online
nucleoside analogue can be used as an HBV DNA inhibitor. The
disparity in bioactivity between J-GC and J-HC also provided us
with the possibility that in these cases nucleoside analogues with
canonical hydrogen face may be recognized more efficiently.
2-N-Phthaloyl-4-(5¢-O-acetyl-2¢,3¢-O-isopropylidene-b-D-ribo-
furansoyl)amino-5-cyano-6-chloroprymidine (7a). Compound 5a
(1.1 g, 3.45 mmol) was dissolved in dry THF (30 mL), 2,3-O-
isopropylidene-D-ribofuranosylamine toluene-p-sulfonate (2.5 g,
6.9 mmol), NaHCO3 (0.6 g, 6.9 mmol) and iPr2EtN (1.2 mL,
6.9 mmol) were added in turn. The mixture was stirred at
r.t. for 10 min and then heated to 70 ◦C, stirred for another
40 min. Subsequently, the solvent was evaporated, the residue
was dissolved in DCM (50 mL) and washed with H2O (50 mL ¥
2), 0.1 M HCl (50 mL) and finally dried with anhydrous sodium
sulfate. The crude product 6a (2.0 g, 4.2 mmol) was obtained
by evaporation and dissolved in DCM (75 mL). Then, iPr2EtN
(0.9 mL, 5 mmol), Ac2O (0.5 mL, 5 mmol) and DMAP (24.6 mg,
0.2 mmol) were added. The mixture was stirred at r.t. for 7 min,
whereafter, H2O (1 mL) was injected to quench the reaction. The
mixture was diluted to 150 mL with DCM and washed with 0.1 M
HCl (75 mL) and H2O (75 mL). After drying with anhydrous
sodium sulfate, the organic layer was evaporated to afford a crude
foam. FC (DCM/Acet. = 97 : 3 as eluent) was applied to obtain
the pure product 7a (1.65 g, 76%): UV (MeOH): 216 (24226),
236 (21872), 300 (4223). dH (400 MHz; d6-DMSO): 1.36 (3H, s,
CH3), 1.54 (3H, s, CH3), 2.19 (3H, s, CH3CO), 4.05–4.09 (1H,
m, 4¢H), 4.45–4.52 (2H, m, 5¢H), 4.68–4.71 (1H, q, J = 2.3 Hz,
3¢H), 4.87–4.90 (1H, q, J = 2.4 Hz, 2¢H), 6.03–6.06 (1H, t, J =
5.0 Hz, 1¢H), 6.89–6.91 (1H, d, J = 8.0 Hz, NH), 7.81–7.99 (4H,
m, C6H4). dC (600 MHz; d6-DMSO): 20.99, 25.41, 27.15, 64.42,
81.70, 83.22, 87.85, 90.32, 113.13, 124.65, 131.37, 136.02, 152.93,
163.19, 163.51, 164.85, 170.54. HRMS (ESI+) m/z: Calc. for
C23H20ClN5O7: 514.1130 [M+H]+. Found 514.1127 [M+H]+.
Experimental
General
All chemicals were purchased from Aldrich, Sigma, or Fluka. Sol-
vents were of laboratory grade. Thin-layer chromatography (TLC)
was performed on TLC aluminium sheet covered with silica gel 60
F254 (0.2 mm, Merck, Germany). Flash column chromatography
(FC): silica gel 60 (Haiyang chemical company, P. R. China) at 0.4
bar. UV-spectra were recorded on a DU-800 spectrophotometer
(Beckman, US), lmax. in nm, e in dm3 mol-1 cm-1. NMR spectra:
1
AV II spectrometers (Bruker, Germany) at 400 MHz for H and
600 MHz for 13C. The d values in ppm are relative to Me4Si
as internal standard, Elemental analyses were performed by the
analysis and testing center, Sichuan University. High resolution
mass spectra are measured with mass analyzer (Q-TOF-premier,
Waters company, US).
2-Amino-4,6-dichloropyrimidine-5-carbaldehyde oxime (3a).
Compound 2a (10 g, 52 mmol) was firstly dissolved in heated
AcOH (800 mL). After cooling to r.t., hydroxylamine hydrochlo-
ride (5.4 g, 78.1 mmol) in EtOH (800 mL) was added dropwise over
2 h, then the mixture was stirred at r.t. overnight. Subsequently,
the precipitate was filtered and washed with EtOH until neutral.
After vacuum drying, compound 3a (8.8 g, 82.2%) was obtained
as a pale yellow solid: UV (MeOH): 267 (28913). dH (400 MHz;
d6-DMSO): 7.79 (2H, s, NH2), 8.03 (1H, s, CH), 11.62 (1H, s,
OH). dC (600 MHz; d6-DMSO): 111.30, 142.28, 160.34, 161.39.
C5H4Cl2N4O requires C, 29.01; H, 1.95; N, 27.06%; found C, 29.14;
H, 2.13; N, 26.88%.
4,7-Diamino-1-(2¢,3¢-O-isopropylidene-b-D-ribofuranosyl)-5-
methyloxy-pyrimido[4, 5-d]pyrimidine-2(1H)-one (8a). Com-
pound 7a (1.3 g, 2.59 mmol) was dissolved in DCM (46 mL)
at 0 ◦C, iPr2EtN (0.9 mL, 5.18 mmol) was added, 2 min later,
CCI (1.4 g, 13 mmol) was added into the solution. Then the
ice-box was removed and the reaction was stirred at r.t. for
7 min. Subsequently, 0.1 M HCl (6 mL) was injected carefully
to quench the reaction. The resulting mixture was diluted to
150 mL with DCM and washed with H2O (150 mL). Then the
solvent was evaporated and a crude yellowish foam was obtained.
The yellowish foam was dissolved in 0.5 M NaOMe (150 mL)
and stirred under reflux for 75 min. Whereafter, the precipitate
was filtered off when the mixture was still hot and the solvent
was removed by evaporation; a white residue was obtained and
immediately purified by FC (DCM/Methanol = 95 : 5–9 : 1 as
eluent) to afford product 8a (0.3 g, 30.2%): UV (MeOH): 226
(11181), 287 (9281). dH (400 MHz; d6-DMSO): 1.27 (3H, s, CH3),
1.47 (3H, s, CH3), 3.48–3.58 (2H, d, 5¢CH2), 3.92 (1H, s, 4¢H), 3.98
(3H, s, OCH3), 4.67 (1H, s, 3¢H), 4.84 (1H, s, 2¢H), 5.14 (1H, s, OH),
6.85 (1H, s, 1¢H), 7.31–7.58 (3H, m, NH2+H-NH), 7.89 (1H, s,
H-NH). dC (600 MHz; d6-DMSO): 27.73, 54.86, 62.85, 82.53,
84.73, 87.39, 89.00, 112.68, 123.39, 133.08, 134.78, 154.73, 161.31,
162.97, 167.86, 169.68. HRMS (ESI+) m/z: Calc. for C15H20N6O6:
403.1342 [M+Na]+. Found 403.1336 [M+Na]+.
2-Amino-5-cyano-4,6-dichloroprymidine (4a). Compound 3a
(10 g, 48.4 mmol), selenium dioxide (5.4 g, 48.4 mmol) and pyridine
(5 mL) were added in toluene (1200 mL), then the reaction mixture
was brought to 110 ◦C and stirred for 15 min. Whereafter, the
supernatant was evaporated to furnish 4a (6.7 g, 73.2%): UV
(MeOH): 260 (24041). dH (400 MHz; d6-DMSO): 8.63 (2H, s,
NH2). dC (600 MHz; d6-DMSO): 94.96, 114.26, 162.12, 162.89.
C5H2Cl2N4 requires C, 31.77; H, 1.07; N, 29.64%; found C, 31.45;
H, 1.17; N, 29.28%.
2-N-Phthaloyl-5-cyano-4,6-dichloroprymidine
(5a). Com-
pound 4a (6.4 g, 33.8 mmol) was dissolved in dry THF (200 mL),
iPr2EtN (29.2 mL, 170.0 mmol) was added. A few minutes later,
phthaloyl dichloride (9.7 mL, 67.6 mmol) was added dropwise
in 20 min. The mixture was stirred at r.t. for 20 min. After
evaporation, the residue was dissolved in DCM (250 mL) and
washed with H2O (250 mL ¥ 2). The organic layer was finally
dried with anhydrous sodium sulfate and evaporated. Purification
was achieved by FC (eluent: DCM) to give a white solid 5a (6.5 g.
60.3%): UV (MeOH): 210 (42574), 259 (21929), 239 (14642). dH
(400 MHz; d6-DMSO): 7.96–8.08 (4H, m, C6H4). dC (600 MHz;
d6-DMSO): 78.28, 114.39, 125.79, 136.42, 136.64, 163.65, 164.42,
169.00. HRMS (ESI+) m/z: Calc. for C13H4Cl2N4O2: 318.9790
[M+H]+. Found 318.9781 [M+H]+.
4,7-Diamino-1-(b-D-ribofuranoyl)pyrimido[4,5-d]pyrimidine-2,
5(1H,6H)-dione (1a). 8a (3.8 g, 10 mmol) was dissolved in
acetonitrile (30 mL); NaI (4.5 g, 30 mmol) and TMSi-Cl (3.3 g,
30 mmol) were then added. The reaction was stirred at r.t. over
20 h, and the precipitate from the resulting mixture was collected,
1520 | Org. Biomol. Chem., 2011, 9, 1516–1522
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
which was washed with acetonitrile (30 mL ¥ 3) and acetone
(30 mL ¥ 3). The treated precipitate (3.6 g, 95%) was then dissolved
with CF3COOH (10% H2O, 30 mL) and stirred for 20 min at
0 ◦C. After CF3COOH was removed by evaporation, the residue
was co-evaporated with ethanol (30 mL ¥ 3) to remove the residual
acid. The final residue was dissolved in hot water and cooled to r.t.;
8 h later, a white precipitate was found as final product 1a (3.0 g,
91%): UV (MeOH): 228 (2359), 283 (1216). dH (400 MHz; d6-
DMSO): 3.41–3.45 (1H, m, 4¢H), 3.59–3.65 (2H, t, 5¢CH2), 4.19
(1H, s, 3¢H), 4.54–4.56 (1H, d, J = 8.0 Hz, 3¢OH), 4.68 (1H, s,
2¢OH), 4.80–4.81 (1H, bd, J = 2.0 Hz, 2¢H), 5.02 (1H, s, 5¢OH),
6.47 (1H, s, 1¢H), 7.57–8.25 (4H, m, NH2 ¥ 2), 11.72 (1H, br, NH).
HRMS (ESI-) m/z: Calc. for C11H14N6O6: 325.0896 [M - H]-.
Found 325.0891 [M - H]-.
d6-DMSO): 21.06, 25.56, 27.21, 64.48, 81.80, 83.13, 83.84, 87.91,
113.20, 159.97, 162.23, 170.57. HRMS (ESI+) m/z: Calc. for
C15H17ClN4O5: 369.0956 [M+H]+. Found: 369.0973 [M+H]+.
4-Amino-1-(2¢,3¢-O-isopropylidene-b-D-ribofuranoyl)-5-metho-
xylpyrimido[4,5-d]pyrimidine-2(1H)-one (8b). Compound 7b
(3.7 g, 10 mmol) was dissolved in dry DCM (100 mL), and
iPr2EtN (6.5 g, 50 mmol) was added. CCI (5.2 g, 50 mmol) was
then added to the mixture carefully at 0 ◦C. The reaction was
brought to r.t. and stirred for 30 min. Subsequently, 0.1 M HCl
(5 mL) was injected to quench the reaction. After insolubles were
filtered off, the mixture was washed with H2O (50 mL ¥ 2), dried
over anhydrous sodium sulfate and evaporated. The residue was
refluxed with 0.15 M NaOMe (50 mL) for 1 h. and a white powder
8b (1.1 g, 30%) was obtained by purifying the resulting mixture
with FC (DCM/methanol = 97 : 3 as eluent): UV (MeOH): 211
(33559), 273 (6213). dH (400 MHz; d6-DMSO): 1.30 (3H, s, CH3),
1.51 (3H, s, CH3), 3.53–3.65 (2H, q, 5¢CH2), 3.99–4.00 (1H, d,
J = 4.0 Hz, 4¢H), 4.12 (3H, s, OCH3), 4.76 (1H, s, 3¢H), 4.89–4.90
(1H, d, J = 4.8 Hz, 2¢H), 5.22–5.23 (1H, d, J = 6.4 Hz, 5¢OH),
6.87 (1H, s, 1¢H), 7.83 (1H, s, H-NH), 8.55 (1H, s, H-NH), 8.73
(1H, s, CH). dC (600 MHz; d6-DMSO): 25.74, 27.68, 55.84, 62.68,
82.67, 84.53, 88.21, 112.96, 154.09, 159.42, 160.12, 161.10, 166.85.
HRMS (ESI-) m/z: Calc. for C15H19N5O6: 364.1256 [M - H]-.
Found 364.1282 [M - H]-.
4,6-Dichloropyrimidine-5-carbaldehyde oxime (3b). Com-
pound 2b (1.8 g, 10 mmol) was dissolved in acetic acid (20 mL).
Hydroxylamine hydrochloride (1.4 g, 20 mmol) in ethanol (40 mL)
was added dropwise, the mixture was stirred at r.t. for 15 h. After
the solvent was evaporated, the residue was dissolved in ether
(20 mL), insolubles were filtered off and the organic layer was
washed with saturated aqueous NaHCO3 (20 mL ¥ 3), then dried
with anhydrous sodium sulfate. Finally, a white solid 3b (1.6 g,
84%) was furnished by evaporation. UV (MeOH): 256 (7662). dH
(400 MHz; d6-DMSO): 8.21 (1H, s, NCH), 8.91 (1H, s, CH), 12.28
(1H, s, OH). dC (600 MHz; d6-DMSO): 113.63, 142.03, 149.50,
154.25. C5H3Cl2N3O requires C, 31.28; H, 1.57; N, 21.89%; found
C, 31.24; H, 1.46; N, 22.07%.
4-Amino-1-(b-D-ribofuranoyl)pyrimido[4,5-d ]pyrimidine-2,
5(1H,6H)-dione (1b). Compound 8b (3.7 g, 10 mmol) was
dissolved in acetonitrile (30 mL), and NaI (4.5 g, 30 mmol) and
TMSi-Cl (3.3 g, 30 mmol) were then added. The mixture was
stirred for 20 h, whereafter, the resulting precipitate was collected,
and washed with acetonitrile (30 mL ¥ 3) and acetone (30 mL ¥ 3).
Finally, a yellowish powder was obtained (3.3 g, 95%). The powder
was then dissolved in CF3COOH (10% H2O, 30 mL) at 0 ◦C and
stirred at r.t. for 20 min. Subsequently, CF3COOH was removed
by evaporation, and the residue was co-evaporated with ethanol
(30 mL ¥ 3). The final mixture was dissolved in hot water and
cooled to r.t.. After 8 h, the white precipitate was obtained as final
product 1b (2.7 g, 91%): UV (MeOH): 225 (17184), 277 (2670).
4,6-Dichloro-5-cyanopyrimidine (4b). Compound 3b (1.9 g,
10 mmol), and selenium dioxide (1.3 g, 12 mmol) in toluene
(150 mL) were heated under reflux for 3 h. The resulting mixture
was then evaporated and applied to FC (DCM as eluent) affording
a white powder 4b (1.4 g, 82%): UV (MeOH): 235 (3418). dH
(400 MHz; d6-DMSO): 8.53 (1H, s, CH). dC (600 MHz; d6-
DMSO): 101.12, 113.35, 155.25, 163.85. C5HCl2N3 requires C,
34.52; H, 0.58; N, 24.15%; found C, 34.61; H, 0.64; N, 24.25%.
4-(5¢-O-Acetyl-2¢, 3¢-O-isopropylidene-b-D-ribofuransoyl)amino-
5-cyano-6-chloropyrimidine (7b). Compound 4b (1.7 g, 10 mmol)
in dry THF (50 mL) was added to 2,3-O-isopropylidene-
ribofuranosylamine toluene-p-sulfonate (7.2 g, 20 mmol). When
the solution became homogeneous, NaHCO3 (1.7 g, 20 mmol)
and iPr2EtN (3.4 mL, 20 mmol) were added. The mixture was
heated to 60 ◦C and stirred for 1 h. Subsequently, the solvent was
removed under reduced pressure, the residue was dissolved with
DCM (50 mL), washed with 0.1 M HCl (25 mL ¥ 2) and H2O
(25 mL ¥ 2), and finally dried with anhydrous sodium sulfate.
After evaporation, the residue was dissolved in DCM (50 mL);
iPr2EtN (2.6 mL, 15 mmol), acetic anhydride (1.4 mL, 15 mmol)
and DMAP (12.2 mg, 0.1 mmol) were added. The mixture was
stirred at r.t. for 10 min, and washed with 0.1 M HCl (25 mL) and
H2O (25 mL), and finally dried over anhydrous sodium sulfate.
After evaporation, the residue was applied to FC (DCM as eluent)
to furnish compound 7b (2.1 g, 58%): UV (MeOH): 216 (24168),
249 (13910), 298 (3333). dH (400 MHz; d6-DMSO): 1.29 (3H, s,
CH3), 1.46 (3H, s, CH3), 2.02 (3H, s, CH3CO), 4.13–4.18 (3H,
m, 5¢CH2+4¢ H), 4.76–4.78 (1H, q, J = 1.4 Hz, 3¢H), 5.00–5.02
(1H, q, J = 1.2 Hz, 2¢H), 5.87–5.89 (1H, d, J = 8.0 Hz, 1¢H), 8.62
(1H, s, CH), 8.99–9.01 (1H, bd, J = 8.0 Hz, NH). dC (600 MHz;
d
H (400 MHz; d6-DMSO): 3.45 (1H, s, 4¢H), 3.61–3.70 (2H, t, J =
7.6 Hz, 5¢ H), 4.20 (1H, s, 3¢H), 4.56 (1H, s, OH), 4.64 (1H, s, OH),
4.84 (1H, s, 2¢H), 5.04 (1H, s, 5¢OH), 6.51 (1H, s, 1¢H), 8.22–8.42
(3H, m, NH2), 8.40 (1H, s, CH), 13.15 (1H, s, NH). HRMS (ESI-)
m/z: Calc. for C11H13N5O6: 310.0787 [M - H]-. Found 310.0782
[M - H]-.
Acknowledgements
We thank the National Natural Science Foundations of China
(document NO. 20772087) for the financial support. We also thank
Ming-Hai Tang from National Key Laboratory of Biotherapy for
providing us with high quality mass spectra and Peng-Chi Deng
from the analysis and testing center of Sichuan University for
helping us to process NMR spectra. We especially thank Prof.
Mingli Yang for the discussion of computational work.
References
1 C. M. Galmarini, J. R. Mackey and C. Dumontet, Leukemia, 2001, 15,
875.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1516–1522 | 1521
©

View Article Online
2 C. M. Galmarini, J. R. Mackey and C. Dumontet, Lancet Oncol., 2002,
3, 415–424.
3 J. T. Witkowski, R. K. Robins, R. W. Sidwell and L. N. Simon, J. Med.
Chem., 1972, 15, 1150–1154.
22 J. L. Sessler, J. Jayawickramarajah, M. Sathiosatham, C. L. Sherman
and J. S. Brodbelt, Org. Lett., 2003, 5, 2627–2630.
23 D. Chen, S. K. Sharma and L. W. McLaughlin, J. Am. Chem. Soc.,
2004, 126, 70–71.
4 R. W. Sidwell, J. H. Huffman, L. Khare, B. Allen, R. Witkowski and
K. Robins, Science, 1972, 177, 705.
5 J. G. McHutchison, S. C. Gordon, E. R. Schiff, M. L. Shiffman, W. M.
Lee, V. K. Rustgi, Z. D. Goodman, M. H. Ling, S. Cort and J. K.
Albrecht, N. Engl. J. Med., 1998, 339, 1485.
6 G. L. Davis, R. Esteban-Mur, V. Rustgi, J. Hoefs, S. C. Gordon,
C. Trepo, M. L. Shiffman, S. Zeuzem, A. Craxi and M. H. Ling,
N. Engl. J. Med., 1998, 339, 1493.
24 M. Mascal, N. M. Hext, R. Warmuth, J. R. Arnall-Culliford, M. H.
Moore and J. P. Turkenburg, J. Org. Chem., 1999, 64, 8479–8484.
25 H. Fenniri, P. Mathivanan, K. L. Vidale, D. M. Sherman, K. Hallenga,
K. V. Wood and J. G. Stowell, J. Am. Chem. Soc., 2001, 123, 3854–3855.
26 F. H. Martin, M. M. Castro, F. Aboul-ela and I. Tinoco Jr, Nucleic
Acids Res., 1985, 13, 8927.
27 E. Ohtsuka, S. Matsuki, M. Ikehara, Y. Takahashi and K. Matsubara,
J. Biol. Chem., 1985, 260, 2605.
7 J. B. McCormick, I. J. King, P. A. Webb, C. L. Scribner, R. B.
Craven, K. M. Johnson, L. H. Elliott and R. Belmont-Williams,
N. Engl. J. Med., 1986, 314, 20.
28 Y. Takahashi, K. Kato, Y. Hayashizaki, T. Wakabayashi, E. Ohtsuka, S.
Matsuki, M. Ikehara and K. Matsubara, Proc. Natl. Acad. Sci. U. S. A.,
1985, 82, 1931.
29 L. Bell, H. M. McGuire and G. A. Freeman, J. Heterocycl. Chem., 1983,
20, 41–44.
30 A. Gomtsyan, S. Didomenico, C.-H. Lee, M. A. Matulenko, K. Kim,
E. A. Kowaluk, C. T. Wismer, J. Mikusa, H. Yu, K. Kohlhaas, M. F.
Jarvis and S. S. Bhagwat, J. Med. Chem., 2002, 45, 3639–3648.
31 N. J. Cusack, B. J. Hildick, D. H. Robinson, P. W. Rugg and G. Shaw,
J. Chem. Soc., Perkin Trans. 1, 1973, 1720–1731.
8 P. R. Wyde, Antiviral Res., 1998, 39, 63–79.
9 J. D. Graci and C. E. Cameron, Rev. Med. Virol., 2006, 16, 37–48.
10 S. Crotty, C. Cameron and R. Andino, J. Mol. Med., 2002, 80, 86–95.
11 E. Domingo, C. Escarmis, N. Sevilla, A. Moya, S. F. Elena, J. Quer,
I. S. Novella and J. J. Holland, FASEB J., 1996, 10, 859.
12 E. Domingo and J. J. Holland, Annu. Rev. Microbiol., 1997, 51, 151–178.
13 E. Domingo, Virology, 2000, 270, 251–253.
14 C. Ash, Science, 2001, 292, 1969.
32 C. K. Chu, F. M. El-Kabbani and B. B. Thompson, Nucleos. Nucleot.
Nucl., 1984, 3, 1–31.
15 S. Crotty, C. E. Cameron and R. Andino, Proc. Natl. Acad. Sci. U. S. A.,
2001, 98, 6895.
33 B. Rayner, C. Tapiero and J. L. Imbach, Carbohydr. Res., 1976, 47,
16 D. Loakes, Nucleic Acids Res., 2001, 29, 2437.
195–202.
17 S. Crotty, D. Maag, J. J. Arnold, W. Zhong, J. Y. N. Lau, Z. Hong, R.
Andino and C. E. Cameron, Nat. Med., 2000, 6, 1375–1379.
18 S. J. Cristol and D. C. Lewis, J. Am. Chem. Soc., 1967, 89, 1476–1483.
19 A. Marsh, E. G. Nolen, K. M. Gardinier and J. M. Lehn, Tetrahedron
Lett., 1994, 35, 397–400.
20 N. Branda, G. Kurz and J. M. Lehn, Chem. Commun., 1996, 2443–2444.
21 A. Asadi, B. O. Patrick and D. M. Perrin, J. Org. Chem., 2007, 72, 466.
34 J. L. Imbach, Ann. N. Y. Acad. Sci., 1975, 255, 177–184.
35 H. Rosemeyer, G. Toth, B. Golankiewicz, Z. Kazimierczuk, W.
Bourgeois, U. Kretschmer, H. P. Muth and F. Seela, J. Org. Chem.,
1990, 55, 5784–5790.
36 B. E. Korba and G. Milman, Antiviral Res., 1991, 15, 217–228.
37 B. E. Korba and J. L. Gerin, Antiviral Res., 1992, 19, 55–70.
38 C. Ying, E. De Clercq and J. Neyts, Antiviral Res., 2000, 48, 117–124.
1522 | Org. Biomol. Chem., 2011, 9, 1516–1522
This journal is
The Royal Society of Chemistry 2011
©