Synthesis and antiviral activity of novel 5-(1-cyanamido-2-haloethyl) and 5-(1-hydroxy(or methoxy)-2-azidoethyl) analogues of uracil nucleosides

Article abstract of DOI:10.1021/jm010226s

A new class of 5-(1-cyanamido-2-haloethyl)-2′-deoxyuridines (4-6) and arabinouridines (7, 8) were synthesized by the regiospecific addition of halogenocyanamides (X-NHCN) to the 5-vinyl substituent of the respective 5-vinyl-2′-deoxyuridine (2) and 2′-arab

Full text of DOI:10.1021/jm010226s

J . Med. Chem. 2001, 44, 3531-3538
3531
Syn th esis a n d An tivir a l Activity of Novel 5-(1-Cya n a m id o-2-h a loeth yl) a n d
5
-(1-Hyd r oxy(or m eth oxy)-2-a zid oeth yl) An a logu es of Ur a cil Nu cleosid es
Rakesh Kumar,* Dinesh Rai, Sanjay K. Sharma, Holly A. Saffran, Ryan Blush, and D. Lorne J . Tyrrell
Department of Medical Microbiology and Immunology, Faculty of Medicine, University of Alberta,
Edmonton, AB, Canada T6G 2H7
Received May 22, 2001
A new class of 5-(1-cyanamido-2-haloethyl)-2′-deoxyuridines (4-6) and arabinouridines (7, 8)
were synthesized by the regiospecific addition of halogenocyanamides (X-NHCN) to the 5-vinyl
substituent of the respective 5-vinyl-2′-deoxyuridine (2) and 2′-arabinouridine (3). Reaction of
2
with sodium azide, ceric ammonium nitrate, and acetonitrile-methanol or water afforded
the 5-(1-hydroxy-2-azidoethyl)-(10) and 5-(1-methoxy-2-azidoethyl)-2′-deoxyuridines (11). In vitro
antiviral activities against HSV-1-TK (KOS and E-377), HSV-1-TK , HSV-2, VZV, HCMV,
and DHBV were determined. Of the newly synthesized compounds, 5-(1-cyanamido-2-iodoethyl)-
+
-
2
′-deoxyuridine (6) exhibited the most potent anti-HSV-1 activity, which was equipotent to
acyclovir and superior to 5-ethyl-2′-deoxyuridine (EDU). In addition, it was significantly
inhibitory for thymidine kinase deficient strain of HSV-1 (EC50 ) 2.3-15.3 µM). The 5-(1-
cyanamido-2-haloethyl)-2′-deoxyuridines (4-6) all were approximately equipotent against
HSV-2 and were ∼1.5- and 15-fold less inhibitory for HSV-2 than EDU and acyclovir,
respectively. Compounds 4-6 were all inactive against HCMV but exhibited appreciable
antiviral activity against VZV. Their anti-VZV activity was similar or higher to that of EDU
and approximately 5-12-fold lower than that of acyclovir. The 5-(1-cyanamido-2-haloethyl)-
(7,8) analogues of arabinouridine were moderately inhibitory for VZV and HSV-1 (strain KOS),
whereas compounds 10 and 11 were inactive against herpes viruses. Compounds 5 and 6 also
demonstrated modest anti-hepatitis B virus activity against DHBV (EC50 ) 19.9-23.6 µM).
Interestingly, the related 5-(1-azido-2-bromoethyl)-2′-deoxyuridine (1n ) analogue proved to be
markedly inhibitory to DHBV replication (EC50 ) 2.6-6.6 µM). All compounds investigated
exhibited low host cell toxicity to several stationary and proliferating host cell lines as well as
mitogen-stimulated proliferating human T lymphocytes.
In tr od u ction
deoxyuridines (1i-l)5,6 was reported, which exhibited
significant in vitro activity against HSV-1 (EC50 ) 0.24-
New development in the synthesis of 2′-deoxyuridine
and arabinouridine that possess novel 2-carbon substit-
uents at the C-5 position and exhibit potent antiviral
activity represents an important area of drug design.
Among many 5-substituted pyrimidine nucleosides that
3
1.2 µM range). In further studies, the 5-(1-azido-2-
7
haloethyl) derivatives of 2′-deoxyuridine (1m -o) were
found to exhibit broad-spectrum antiviral activity against
herpes viruses [HSV-1 (EC50 ) 0.06-6.4 µM range),
HSV-2 (EC50 ) 2.9-13.2 µM range), VZV (EC50 ) 4.5-
12.3 µM range), and EBV (EC50 ) 4.0-11.1 µM range)].
These compounds were more active than the 5-(1-
hydroxy (or methoxy)-2-haloethyl) analogues (1i-l)
against HSV-1 and HSV-2. These studies indicate that
an azido substituent at C-1 in the 5-(1-azido-2-haloethyl)
side chain is an important determinant for potent
antiviral activity against several herpes viruses. To
synthesize more effective broad-spectrum antiviral
agents, a cyanamido (NHCN) moiety located at the C-1
position of the 5-substituent attracted our attention. The
NHCN moiety is structurally and electronically related
to N3 group and appears to be an efficient bioisostere
have been investigated, (E)-5-(2-halovinyl)- (1a , IVDU;
1
1
b, BVDU; 1c, CVDU) , and 5-(2-chloroethyl)- (1d ,
2
CEDU) 2′-deoxyuridines are the most potent and
selective in their action against herpes simplex virus
type-1 (HSV-1). CEDU is effective against systemic
HSV-1 infection and HSV-1 encephalitis in mice at
-15-fold lower dose than BVDU.3 Herpes simplex
5
type-2 virus (HSV-2) replication is affected only at
considerably higher concentrations of BVDU or CEDU.
The less potent 5-ethyl-2′-deoxyuridine (1e, EDU) is
1
approximately equiactive against HSV-1 and HSV-2.
In contrast, 5-(hydroxyethyl)-2′-deoxyuridine (1f) is an
2
inactive antiviral agent.
8
,9
4
of both N3 and OH groups. The cyanamido group, like
the azido group, retains nitrogen atoms at the N-R and
N-γ positions. The resonance contribution for the cy-
anamido group may resemble the π-electron delocaliza-
tion for azido resonance species. Bioisosteres modulate
biological activity by virtue of subtle differences in
physicochemical properties and define some of the
essential requirements of the pharmacophores. It was
therefore of interest to investigate the hitherto unknown
Shiau et al. reported that the 5-(hydroxymethyl)- (1g)
and 5-(azidomethyl)- (1h ) 2′-deoxyuridines have good
affinity for HSV-1 encoded thymidine kinase (TK) and
are strong inhibitors of HSV-1 replication. However,
they were cytotoxic to host cells. In earlier studies, the
synthesis of 5-[1-hydroxy (or methoxy)-2-haloethyl]-2′-
*
To whom correspondence should be addressed. Tel: (780) 492-
7
545. Fax: (780) 492-7521. E-mail: rakesh.kumar@ualberta.ca.
1
0.1021/jm010226s CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/13/2001

3
532 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 21
-(1-cyanamido-2-haloethyl)-2′-deoxyuridines (4-6) to
Kumar et al.
Sch em e 1a
5
determine whether the substitution of the N3 at the C-1
by a NHCN group could result in a new class of
compounds more potent and selective than 5-(1-azido-
2
-haloethyl)-2′-deoxyuridines (1m -o).
To continue our efforts to study the structure-activity
relationships (SARs) of modified nucleosides, besides
novel 5-(C-1) substituted derivatives (4-6), we were
interested in investigating the effect of bioisosteric azido
moiety at the C-2 carbon of 5-substituent on biological
activity. Therefore, we have now synthesized novel 5-(1-
hydroxy-2-azidoethyl)- (10) and 5-(1-methoxy-2-azido-
ethyl)- (11) 2′-deoxyuridines. The rationale for the
synthesis of 10 and 11 came from the observation that
the halogen atom at the C-2 position of the 5-substituent
1
-3
contributes significantly to the anti-herpes activity.
We also observed previously that 5-(1-azido-2-haloethyl)
analogues (1m -o) exhibited potent antiviral activity in
contrast to the 5-(1-azidoethyl) analogue of 2′-deoxyuri-
7
a
dine (1q), which was devoid of antiviral activity. The
2 3
Reagents: (i) N-chlorosuccinimide, NH CN, CH CN, 0-25 °C
electronic (inductive effect, F value), steric (molecular
refractive value), and lipophilic effect (π-value) of an
azido group are close to that of halogens, viz: N3, 0.30,
(4); N-bromosuccinimide (5, 7), N-iodosuccinimide (6, 8), NH2CN,
DME, 0 °C; (ii) ceric ammonium nitrate, NaN3, CH3CN-H2O,
-5 °C [10 (method A)]; ceric ammonium nitrate, NaN3, dry
CH3CN, -15 °C [10 (method B), 11].
1
0
0.2, 0.46; Cl, 0.41, 6.03, 0.71; Br, 0.44, 8.88, 0.86; I,
.40, 13.94, 1.12, respectively.10 Thus the azido sub-
amido-2-haloethyl)- (5, 6) derivatives were found to
exhibit appreciable anti-HBV activity (EC50 ) 19.9-23.6
µM). These results suggested that 5-(1-azido-2-haloeth-
yl) analogues (1m -o) could be potential anti-HBV
agents and encouraged us to resynthesize 1m -o to
evaluate them for their anti-HBV activity. Interestingly,
among these compounds, 5-(1-azido-2-bromoethyl)-2′-
deoxyuridine (1n ) was found to be markedly inhibitory
for DHBV replication (EC50 ) 2.6-6.6 µM). These
results further prompted us to resynthesize the related
stituent possesses desirable physicochemical properties
to be considered as a bioisosteric replacement of halogen
substituents (Cl, Br, I) at the C-2 position of the
5
-substituent. Furthermore, the azido group may be
capable of electrostatic binding to an enzyme, such as
TK, that is not possible with a halogen such as I, Br, or
Cl. An interesting property of azido substituents con-
taining pyrimidine nucleosides is that they can provide
highly reactive functional groups upon exposure to UV
light or radiation, which can interact with the target
enzyme susceptible to photoaffinity labeling, resulting
in enzyme inactivation.11
5
-(1-methoxy-2-haloethyl)-2′-deoxyuridines (1j-l) in or-
der to determine their anti-HBV activity and to identify
structure-activity correlationship. However, they did
not prove to be inhibitory to DHBV replication at
2
5 µM.
Ch em istr y
The target 5-(1-cyanamido-2-haloethyl) derivatives of
2
8
′-deoxyuridine (4-6, 15-42%), and arabinouridine (7,
, 18-19%) were synthesized by the reaction of respec-
tive 5-vinyl analogues (2, 3) with N-halosuccinimide and
1
3
cyanamide as illustrated in Scheme 1. The C NMR (J
modulation) spectrum of 4 provided conclusive evidence
for the regiospecific addition of XNHCN (X ) Cl, Br, I)
across the 5-vinyl substituent of 2 and 3. For example,
the chlorine atom in 4 is attached to a methylene carbon
that exhibited resonance at δ 45.4, whereas the cyan-
amido substituent is attached to a chiral methine carbon
that exhibited dual resonances at δ 55.9 and 56.2. The
regiospecific addition is consistent with reports that
unsymmetrical olefins, capable of halonium ion forma-
tion, were found to favor an unsymmetrical bridged
intermediate of the type illustrated in Scheme 1, even
in solvents having a high dipole moment.13 Similar
In this report we describe the synthesis and antiviral
activity for the new series of 5-(1-cyanamido-2-haloeth-
yl)-2′-deoxyuridines (4-6), arabinouridines (7, 8) and
5
-[1-hydroxy (or methoxy)-2-azidoethyl]- (10, 11) ana-
logues of 2′-deoxyuridine. It was postulated that com-
regioselectivity has been also observed for the addition
1
2
14,15
pounds (4-8, 10, 11), in contrast to EDU (1e), would
be resistant to metabolic hydroxylation at the C-1
position of the 5-substituent due to obstructive substitu-
tion by the cyanamido, hydroxy, or methoxy substitu-
ents. Of the newly synthesized compounds, 5-(1-cyan-
of hypobromous acid (BrOH),
methyl hypobromite
1
6
17
(BrOCH3), and bromoazide (BrN3) to unsymmetri-
cally substituted alkenes. Compounds 4-8 are mixtures
of two diastereomers (1:1 ratio), which differ in config-
uration (R and S) at the 1-carbon atom of the 5-(1-

Synthesis of Deoxyuridines and Arabinouridines
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 21 3533
cyanamido-2-haloethyl) moiety. Attempts to separate
the diastereomers of 4-8 by flash column chromatog-
raphy or multiple-development TLC techniques were
unsuccessful.
respective products that were isolated. The intermediacy
of related carbonium ions has been previously pro-
posed.
2
0
Resu lts a n d Discu ssion
In the reaction of 2 with N-chlorosuccinimide and
cyanamide in acetonitrile, a mixture of (E)-5-(2-chloro-
vinyl)-2′-deoxyuridine (1c) and 5-(1-methoxy-2-chloro-
ethyl)-2′-deoxyuridine (1j) was also produced in addition
to the target 5-(1-cyanamido-2-chloroethyl) product (4).
Further, reactions of 2 and 3 employing N-halosuccin-
imide and cyanamide provided complex reaction mix-
tures of starting materials and unidentified compounds,
which were difficult to separate, leading to low yields
of the desired compounds.
The antiviral activities for the new class of 5-(1-
cyanamido-2-haloethyl)- (4-8) and 5-(1-hydroxy(or meth-
oxy)-2-azidoethyl)- (10, 11) uracil nucleosides were
determined in culture against herpes simplex virus type
1
(HSV-1, strains KOS and E-377), herpes simplex virus
type 2 (HSV-2, strain MS), varicella zoster virus (VZV,
strain Ellen), and human cytomegalovirus (HCMV,
strain AD-169) infected human foreskin fibroblast (HFF)
or Vero cells. In addition, the compounds were also
Reaction of 5-vinyl-2′-deoxyuridine (2) with 2 equiv
of ceric ammonium nitrate and 1 equiv of sodium azide
in CH3CN-H2O yielded 5-(1-hydroxy-2-azidoethyl)-2′-
deoxyuridine (10, 31.5% (Scheme 1)). In a related
reaction of 2 in dry acetonitrile, followed by quenching
with methanol, 5-(1-methoxy-2-azidoethyl)-2′-deoxyuri-
dine (11, 25%) was obtained. Although a similar reaction
of 2 in dry acetonitrile proceeded instantaneously,
showing one major product by analytical TLC, a second
less polar compound was present after the reaction
mixture was quenched with aqueous acetonitrile. Sepa-
ration of this reaction mixture by silica gel column
chromatography afforded 5-(1-hydroxy-2-azidoethyl)-2′-
deoxyuridine (10, 32%). The less polar compound which
arose after the aqueous acetonitrile quench was identi-
-
evaluated against a thymidine kinase-deficient (TK )
mutant HSV-1 strain (KOSSB) in Vero cells. The results
are summarized in Table 1. The 5-(1-cyanamido-2-
haloethyl)-2′-deoxyuridines (4-6) exhibited significant
inhibitory activity against HSV-1 (strains KOS and
E-377), HSV-2, and VZV replication. The most potent
compound of this series 5-(1-cyanamido-2-iodoethyl)-2′-
deoxyuridine (6) inhibited HSV-1 strains (EC50 ) 1.6-
9
.4 µM range) at concentrations which compared favor-
ably with that of reference drug acyclovir (EC50 ) 2.2-
4
5
.4 µM range) and exhibited superior activity to that of
-ethyl-2′-deoxyuridine (1e) (EC50 ) 19.5-42.9 µM
range). The 5-(1-cyanamido-2-bromoethyl)-2′-deoxyuri-
dine (5) also proved to be a potent inhibitor of HSV-1
replication, it is 3-10-fold (depending on the strain and
cell line) less active than acyclovir, and approximately
equiactive to EDU against HSV-1, whereas the corre-
sponding chloro (4) analogue was much less inhibitory,
i.e., 17-48-fold less inhibitory than acyclovir and 2-5
times less active than EDU. The 5-(1-cyanamido-2-
bromoethyl)- (5) and 5-(1-cyanamido-2-iodoethyl)- (6)
derivatives were found to exhibit 16- and 6-fold superior
anti-HSV-1 activity, respectively, relative to their C-1
1
cal ( H NMR) to the compound 10 as described before.
Compounds 10 and 11 were mixtures of two diastere-
1
omers. The H NMR spectrum of 10 and 11 exhibited
overlapping multiplets for the CH2N3 resonances at δ
3
.30 (10) and 3.56 (11), a multiplet assigned to the
CHOH resonance at δ 4.60 (10), a multiplet assigned
to the CHOMe resonance at δ 4.48 (11), a singlet
assigned to the CHOH resonance at δ 5.72 (exchanges
with deuterium oxide) (10), two closely spaced singlets
for the OCH3 protons at δ 3.34 and 3.36 (11), and two
closely spaced singlets for H-6 at δ 7.80 and 7.82 (10)
and 7.90 and 7.92 (11). The 13C NMR spectrum of 10
and 11 provided further evidence for the assigned
structures based on the resonances at δ 55.01 (CH2N3
of 10), 53.84 (CH2N3 of 11), 66.14 (CHOH of 10), 76.54
7
azido counterparts (1n ,o), when compared with acy-
clovir as a reference. In contrast, the anti-HSV-1 activity
of 5-(1-cyanamido-2-chloroethyl)- (4) was significantly
reduced relative to the 5-(1-azido-2-chloroethyl)-2′-de-
7
oxyuridine (1m ). The compounds 5 and 6 also exhibited
superior anti-HSV-1 activity relative to their corre-
sponding C-1 methoxy (1k ,l) and C-1 hydroxy deriva-
tives (1i, X ) Br, I), compounds 5,6 being 1-10-fold less
active than acyclovir whereas 1k ,l were 10-100 fold less
active and 1i (X ) Br, I) was 1000-times less active than
(CHOCH3 of 11), and 57.46 and 57.52 (OCH3 of 11). The
signals assigned for protons at the C-1 substituent and
the carbon (C-1) bearing the hydroxy or methoxy group
are consistent with the shifts reported for 5-[1-hydroxy
5
,6
acyclovir. In the 5-(1-cyanamido-2-haloethyl)- (4-6)
series of compounds, the relative anti-HSV-1 activity
order parallels that of the 5-(1-methoxy-2-haloethyl)-
(
or methoxy)-2-haloethyl]-2′-deoxyuridines.5,6 The IR
spectrum of 10 and 11 showed very strong bands at
-
1
5,6
2
118 and 2104 cm , respectively, for the azide group,
(1j-l) series, (I > Br > Cl), and differs from that of
whereas the bands for the nitrate group were not
present. In both of these instances (10 and 11), there
was no evidence for the presence of the 5-(1-nitrato-2-
azidoethyl)-2′-deoxyuridine (9). This is in contrast to a
previous report where reaction of olefins with ceric
ammonium nitrate and sodium azide in dry or aqueous
acetonitrile produced R-azido-â-nitratoalkanes as addi-
tion products.1
5-(1-azido-2-haloethyl)-2′-deoxyuridines (1m -o) (Cl >
7
I > Br). These results suggest that the C-1 substituent,
in conjunction with the halogen atom at the C-2 of the
5-side chain, collectively determine the anti-HSV-1
activity.
Acyclovir is a widely used drug for the treatment of
opportunistic herpes virus infections in immunocom-
promised patients. However, resistance to acyclovir is
increasing, due to altered or deficient thymidine kinase
enzymes in the resistant viruses. Herpes virus thy-
midine kinase (HSV-TK) leads to initial selective phos-
phorylation of acyclovir resulting in monophosphate
derivative of the drug. However, the herpes virus
8,19
A possible mechanism for the formation of 10 and 11
involves the decomposition of 5-(1-nitrato-2-azidoethyl)-
2
1
2
′-deoxyuridine (9), which is formed at lower tempera-
ture, to a carbonium ion intermediate at 25 °C which
underwent reaction with methanol or water to yield the

3
534 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 21
Kumar et al.
Ta ble 1. In Vitro Anti-Herpes Activity of 5-Substituted 2′-Deoxyuridines and Arabinouridines
EC50 (µM)a
HSV-1
HSV-1b
KOSSB
(TK )
HSV-1b
KOS
(TK )
E-377
HSV-2c
MS
VZVc
Ellen
HCMVc
AD-169
-
+
+
no.
R
R1
(TK )
4
5
6
7
8
1
1
1
CH(NHCN)CH2Cl
CH(NHCN)CH2Br
CH(NHCN)CH2I
CH(NHCN)CH2Br
CH(NHCN)CH2I
CH(OH)CH2N3
H
H
H
OH
OH
H
H
H
>150
>133
2.3-15.3
>125
>114
>161
>150
>780
>150
177.6
ND
37.5
6.6
1.6
62.5
22.8
>80.0
>75.0
19.5
1.5
210d
42.5
9.4
>256
>228
>322
>300
42.9
2.7
105d
87.7
31.5d
20.7
14.6
32.0
11.4
>322
>300
32.7
0.003
2.6
274
d
>266
>236
>256
>228
>322
>300
>390
>150
ND
d
77.8c
>256
>228
>322
>300
58.5
ND
0
1
e
CH(OMe)CH2N3
e
CH2CH3(EDU)
f
BVDU
ACV
GCV
CHdCHBr
H
g
2.2
ND
4.4
ND
6.2
ND
h
i
ND
0.23
a
Mean of two to three assays. b The concentration (µM) required to inhibit plaque formation in monolayers of Vero cells by 50%. c The
drug concentration (µM) required to reduce the viral cytopathic effect (CPE) in infected human foreskin fibroblast (HFF) cell monolayers
to 50% of untreated infected controls. Plaque reduction assay. e 5-Ethyl-2′-deoxyuridine. f (E)-5-(2-Bromovinyl)-2′-deoxyuridine. Acyclovir.
d
g
h
i
Ganciclovir. ND ) not determined.
infected cells still contain host encoded TK, and this is
the reason that nucleoside analogues, which work as
anti-herpes agents independent of HSV-TK, can be
identified and potentially used for acyclovir resistant
HSV infections. Therefore, we examined the activity of
newly designed compounds 4-8, 10, and 11 in HSV-
TK negative strain of HSV-1 (KOSSB). Interestingly,
the compound 5-(1-cyanamido-2-iodoethyl)-2′-deoxyuri-
Thus the above observations indicate that a C-1
cyanamido functional group in the ethyl side chain at
the 5-position of the pyrimidine ring not only influences
the antiviral potency but also modulates the antiviral
spectrum.
The 5-(1-cyanamido-2-haloethyl)- (7, 8) analogues of
arabinouridine were inactive for HSV-1 (strain E-377),
HSV-2, and HCMV. However, they were moderately
inhibitory for VZV (EC50 ) 11.4-32.0 µM range) and
HSV-1 (strain KOS) (EC50 ) 22.8-62.5 µM range). Thus
the anti-HSV-1 activity of these compounds is depend-
ent on the virus strain used in the assays.
-
dine (6) exhibited marked activity against HSV-TK
HSV-1 infected cells (EC50 ) 2.3-15.3 µM), as compared
to acyclovir (EC50 ) 177.6 µM). The exact mechanism
of action of the compound 6 is not yet known. However,
it can be postulated that compound 6 may be phospho-
rylated by cellular kinases followed by selective inhibi-
tion of viral DNA polymerase.
Intriguingly, 5-(1-hydroxy-2-azidoethyl)- (10) and 5-(1-
methoxy-2-azidoethyl)- (11) 2′-deoxyuridines, in contrast
to 5-(1-hydroxy (or methoxy)-2-haloethyl)-2′-deoxyuridines
5
,6
The anti-HSV-2 activity of compounds 4-6 is about
(1i-l), did not show increased antiviral activity but
were found to be inactive. This indicates that an azido
group is not preferred to a halogen atom at the C-2
position of the 5-substituent. The observed lack of
1
.5- and ∼15-fold less than that of EDU and acyclovir,
respectively, but they exhibited greater anti-HSV-2
activity than the corresponding 5-(1-azido-2-haloethyl)
series of analogues (1m -o), which were 32-146-fold
+
-
activity of 10, 11 against TK herpes viruses and TK
7
less potent than acyclovir. The halogen atom was not
herpes virus may be attributed to their insufficient
recognition by cellular/herpes thymidine kinases and/
or poor affinity of their 5′-triphosphate derivatives for
viral DNA polymerases.
a determinant of anti-HSV-2 activity in compounds 4-6
since they were all equiactive against HSV-2, whereas
7
in 5-(1-azido-2-haloethyl)- (1m -o) series of compounds,
the relative anti-HSV-2 activity order was Br g Cl > I.
The 5-(1-cyanamido-2-haloethyl)-2′-deoxyuridines (4-
The anti-HBV activity for new compounds 4-8, as
5
-7
well as 1j-o,
along with the reference antiviral
6
), like 5-(1-azido-2-haloethyl)-2′-deoxyuridines (1m -
compound (-)-â-L-2′,3′-dideoxy-3′-thiacytidine (3-TC),
was also assessed in confluent cultures of primary duck
hepatocytes obtained from chronically infected Pekin
ducks. These cells chronically produce DHBV DNA,
therefore antiviral activity was determined by analysis
of viral DNA by dot blot hybridization (Table 2). In the
new class of 5-(1-cyanamido-2-haloethyl)-2′-deoxyuridines,
compounds 5 and 6 exhibited considerable inhibition of
DHBV replication (EC50 ) 19.9-23.6 µM) in vitro, in
contrast to their 2′-arabinouridine analogues (7 and 8),
which were not active at 25 µM. These results are
consistent with the previous observations that changing
o), were inactive against HCMV and exhibited appre-
ciable activity against VZV (EC50 ) 14.6-31.5 µM
range). The anti-VZV activity of compounds 4-6 was
similar or higher than that of EDU (EC50 ) 32.7 µM),
approximately 5-12-fold lower than that of acyclovir
(
EC50 ) 2.6 µM), and approaching that of 5-(1-azido-2-
haloethyl) derivatives (1m -o) (EC50 ) 4.5-12.3 µM
range). The nature of C-1 or C-2 substituents in the
7
compounds 4-6 and 1m -o did not alter their relative
7
anti-VZV activity, since compounds 4-6 and 1m -o
were almost equiactive.

Synthesis of Deoxyuridines and Arabinouridines
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 21 3535
Ta ble 2. In Vitro Anti-Hepatitis B Virus Activity of
the highest concentration tested (100 µM) (data not
shown), indicating that novel compounds 4-8 do not
have cytotoxicity against various human cell lines nor
any anti-cancer activity.
5
-Substituted 2′-Deoxyuridines and Arabinouridines
The compounds 4-8, 10, and 11 were tested in vitro
for their toxicity against several other cell lines. None
of these compounds displayed significant in vitro toxicity
against stationary phase cells [Vero cells (CC50 >228
to >600 µM), HFF cells (CC50 >228 to >322 µM)] and
proliferating [HFF (IC50 ) >118 to >322 µM)] cell lines
up to the highest concentration tested. In rapidly
proliferating fresh human T lymphocyte cell culture,
cellular DNA synthesis, as monitored by the incorpora-
EC50 (µM)
no.
R
R1
DHBVa
jb
k
l
m
n
o
1
1
1
1
1
1
4
5
6
7
8
1
1
1
3
CH(OMe)CH2Cl
CH(OMe)CH2Br
CH(OMe)CH2I
CH(N3)CH2Cl
CH(N3)CH2Br
CH(N3)CH2I
CH(NHCN)CH2Cl
CH(NHCN)CH2Br
CH(NHCN)CH2I
CH(NHCN)CH2Br
CH(NHCN)CH2I
CH(OH)CH2N3
CH(OMe)CH2N3
H
H
H
H
H
H
H
H
>30.0
>30.0
>25.0
15.0
2.6-6.6
11.8
>30.0
19.9
23.6
3
tion of [methyl- H]-thymidine into DNA, was also not
affected by compounds 4-8, 10, and 11 up to concentra-
tions of >114 to >150 µM.
The observation that compounds 4-6 exhibited sig-
nificant anti-HSV-1, HSV-2, and VZV activity and no
toxicity to several host cells suggests that they may be
selectively phosphorylated in virus infected cells by
HSV-TK or VZV-TK, respectively, and inhibit viral DNA
polymerases. Alternatively, compounds 4-6 may be
phosphorylated by cellular kinases in both virus infected
and uninfected cells, but selectively inhibit viral DNA
polymerase, since they show activity against TK-
deficient strain of HSV-1 (6) as well as DHBV (5, 6).
H
OH
OH
H
H
H
>25.0
>25.0
NDd
0
1
e
-TC
ND
29.2
0.04
c
CH2CH3(EDU)
e
a
The drug concentration (µM) required to reduce the viral DNA
in HBV infected primary duck hepatocytes to 50% of untreated
b
infected controls. Compounds 1j-o are reported in refs 5, 6, and
c
d
e
8
. 5-Ethyl-2′-deoxyuridine. ND ) not determined. (-)-â-L-2′,3′-
Su m m a r y
Dideoxy-3′-thiacytidine.
The hitherto unknown 5-(1-cyanamido-2-haloethyl)
analogues of 2′-deoxyuridine (4-6) represent novel
inhibitors of HSV-1, HSV-2, VZV, and DHBV in cell
culture. 5-(1-Cyanamido-2-iodoethyl)-2′-deoxyuridine (6)
was found to be higher or equally active for HSV-1, 1.3-
and 12.5-fold less active for HSV-2, 2-fold more to 5-fold
less active for VZV, compared to the reference drugs
EDU and acyclovir, respectively. It is noteworthy that
compound 6 was also active against thymidine kinase
negative strain of HSV-1 and therefore could serve as
a useful lead compound for the development of improved
anti-herpes drugs. It was observed that replacement of
the C-1 azido moiety in 1m -o with a cyanamido group
the 2′-deoxyribose in 5′-tri-phosphate derivative of
5
-propenyl-2′-deoxyuridine to an arabinose moiety re-
sulted in reduced inhibitory activity against DHBV
DNA polymerase.22 The corresponding 5-(1-methoxy-2-
haloethyl)-2′-deoxyuridine derivatives (1j-l) exhibited
less than 50% inhibitory activity against DHBV replica-
tion at 25-30 µM. Surprisingly, the 5-(1-azido-2-halo-
ethyl)- (1m -o) series of compounds demonstrated a
significant anti-HBV activity with 50% effective con-
centration (EC50) of 2.6-15.0 µM, the most potent being
5
-(1-azido-2-bromoethyl) derivative (1n ) with an EC50
of 2.6-6.6 µM. The marked anti-HBV activity of 5-(1-
azido-2-bromoethyl)-2′-deoxyuridine (1n ) is in contrast
to other 5-substituted-2′-deoxyuridines. For example,
BVDU (1b) is a potent anti-herpes agent but did not
show anti-HBV activity (EC50 ) >150 µM) as measured
in the DHBV infected primary duck hepatocyte cul-
tures.23 Thus, it was noted that the 5-(1-azido-2-halo-
ethyl) moiety at the 5-position of the pyrimidine nu-
cleosides contributes significantly to the potent anti-
HBV activity. Although the exact mechanism of action
of these compounds is not known, it is possible that they
may be phosphorylated by host cell kinases to their tri-
phosphate derivatives, which then inhibit DHBV DNA
polymerase like other nucleoside analogues. The com-
pounds 4-8 and 1j-o did not have any visual effect on
the morphology and viability of uninfected duck hepa-
tocytes up to a concentration of 150 µM.
(4-6) at the 5-position of the pyrimidine ring led to
compounds with an altered antiviral spectrum, specif-
ically increased anti-HSV-1 and HSV-2 activities. In
contrast, it is encouraging that 5-(1-azido-2-haloethyl)-
2
′-deoxyuridine analogues (1m -o) emerged as a new
series of potent anti-HBV agents that exhibited superior
anti-HBV activity compared to those of 5-(1-cyanamido-
2
-haloethyl) analogues (4-6). Our observations reported
here support the design and investigation of novel
bioisosteric molecules. Further structure-activity re-
lationship studies are ongoing on these new series of
compounds.
Exp er im en ta l Section
Melting points were determined with a Buchi capillary
1
13
apparatus and are uncorrected. H NMR and C NMR spectra
The cytotoxic activities of compounds 4-8 were also
were determined for solutions in Me SO-d , MeOH-d , or D O
2
6
4
2
determined by the National Cancer Institute (NCI)
on a Bruker AM 300 spectrometer using Me
4
Si as an internal
standard ( H NMR). The assignment of all exchangeable
protons (OH, NH) was confirmed by the addition of the D O.
C NMR spectra were acquired using the J modulated spin-
1
using an in vitro assay24 in which compounds were
2
evaluated against approximately 60 human tumor cell
lines (e.g., melanoma, leukemia, nonsmall cell lung,
small cell lung, colon, central nervous system, ovarian,
prostate, and renal cancers). None of the compounds
showed any activity or selectivity in these assays up to
1
3
echo technique where methyl and methine carbon resonances
appear as positive peaks, and methylene and quaternary
carbons appear as negative peaks. Infrared spectra were
recorded using a Nicolet Magna 750 FT spectrometer. Micro-

3
536 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 21
Kumar et al.
analyses were within (0.4% of theoretical values for all
elements listed, unless otherwise indicated. Preparative thin-
layer chromatography (PTLC) was performed using Whatman
PLK 5F plates, 1.0 mm in thickness, and silica gel column
chromatography was carried out using Merck 7734 silica gel
dimethoxyethane (50 mL) during a period of 5 min. The
reaction was allowed to proceed for 2 h at 0 °C with stirring,
and the solvent was removed in vacuo. The product was
purified by PTLC using chloroform-methanol (90:10, v/v) as
the development solvent to yield 5-(1-cyanamido-2-bromo-
1
(
5
100-200 µM particle size). 5-Vinyl-2′-deoxyuridine (2) and
ethyl)-arabinouridine (7) as a syrup (36 mg, 18%): H NMR
6
,25
-vinylarabinouridine (3)
were prepared by using the
(MeOH-d
4
) (mixture of two diastereomers in a ratio of 1:1) δ
Br), 3.80 (m, 2H, H-5′), 3.92 (m, 1H, H-4′),
.08 (m, 1H, H-3′), 4.15 (m, 1H, H-2′), 4.70 (m, 1H, CHNHCN),
.14 (d, J 1′,2′ ) 3.0 Hz, 1H, H-1′), 7.90 and 7.92 (2s, 1H total,
Br) C, H, N.
published procedures.
-(1-Cyan am ido-2-ch lor oeth yl)-2′-deoxyu r idin e (4), (E)-
-(2-Ch lor ovin yl)-2′-d eoxyu r id in e (1c), a n d 5-(1-Meth -
3.55 (m, 2H, CH
4
6
2
5
5
H-6). Anal. (C12
15 4 6
H N O
oxy-2-ch lor oeth yl)-2′-d eoxyu r id in e (1j). N-Chlorosuccin-
imide (79.8 mg, 0.6 mmol) was added slowly to a pre-cooled
5-(1-Cya n a m id o-2-iod oet h yl)-a r a bin ou r id in e (8). N-
Iodosuccinimide (119 mg, 0.53 mmol) was added slowly with
stirring to a pre-cooled (0 °C) solution of 5-vinyl-arabinouridine
(143 mg, 0.53 mmol) and cyanamide (88.2 mg, 2.1 mmol) in
dimethoxyethane (100 mL), and the reaction was allowed to
proceed for 2 h at 0 °C. Removal of the solvent in vacuo gave
a residue, which was purified by PTLC using chloroform-
methanol (90:10, v/v) as the development solvent. Extraction
of the ultraviolet visible spot with chloroform:methanol (88:
(0-5 °C) solution of 5-vinyl-2′-deoxyuridine (101 mg, 0.4 mmol)
and cyanamide (67.2 mg, 1.6 mmol) in acetonitrile (25 mL).
The resulting reaction mixture was stirred for 4 h at 0-5 °C
and then stirred for 20 h at room temperature. Removal of
the solvent in vacuo gave a residue that was purified by silica
gel column chromatography using chloroform-methanol (92:
8
, v/v) as eluent. The first fraction eluted (20 mg) was found
to be a mixture of (E)-5-(2-chlorovinyl)-2′-deoxyuridine (1c) and
5
1
-(1-methoxy-2-chloroethyl)-2′-deoxyuridine (1j) in a ratio of
12, v/v) yielded 5-(1-cyanamido-2-iodoethyl)-arabinouridine (8)
1
1
:2, based on the H NMR spectral data. The compounds 1c
as a syrup (45 mg, 19.2%): H NMR (D
2
O) (mixture of two
I), 3.80 (m,
1
5,26
and 1j were identical ( H NMR) to authentic samples.
diastereomers in a ratio of 1:1) δ 3.50 (m, 2H, CH
2
Further elution using chloroform-methanol (90:10, v/v)
2H, H-5′), 3.95 (m, 1H, H-4′), 4.06 (m, 1H, H-3′), 4.20 (m, 1H,
H-2′), 4.42 (m, 1H, CHNHCN), 6.20 and 6.22 (two overlapping
d, J 1′,2′ ) 3.0 Hz, 1H total, H-1′), 7.88 and 7.90 (two s, 1H total,
solvent yielded 5-(1-cyanamido-2-chloroethyl)-2′-deoxyuridine
(
(
1
4) as a viscous syrup (25 mg, 19%); H NMR (Me
2
SO-d
6
)
mixture of two diastereomers in a ratio of 1:1) δ 2.14 (m, 2H,
15 4 6
H-6). Anal. (C12H N O I) C, H, N.
H-2′), 3.60 (m, 2H, H-5′), 3.80 (m, 1H, H-4′), 3.90 (m, 2H, CH
Cl), 4.28 (m, 2H, H-3′, CHNHCN), 5.0-5.40 (br, 2H, 3′-OH,
′-OH), 6.14 and 6.18 (two overlapping t, J ) 6.0 Hz, 1H total
H-1′), 7.50 (m, 1H, NHCN; exchanges with deuterium oxide),
.98 and 8.0 (2s, 1H total, H-6), 11.60 (s, 1H, NH, exchanges
2
-
5-(1-Hyd r oxy-2-a zid oeth yl)-2′-d eoxyu r id in e (10). Meth -
od A. Ceric ammonium nitrate (550 mg, 1.0 mmol) was added
slowly to a pre-cooled suspension (-5 °C) prepared by mixing
a solution of 5-vinyl-2′-deoxyuridine (130 mg, 0.51 mmol) and
sodium azide (39 mg, 0.6 mmol) in acetonitrile (80 mL) and
water (0.5 mL). The reaction mixture was stirred for 2 h at
-5 °C, and the solvent was removed in vacuo. Purification of
the product by elution from a silica gel column using chloro-
form-methanol (90:10, v/v) as the eluent afforded 5-(1-
5
7
1
3
with deuterium oxide); C NMR (MeOH-d
CH
2
4
) 41.73 (C-2′), 45.48
Cl), 55.99 and 56.29 (CHNHCN), 62.71 (C-5′), 72.08 and
2.30 (C-3′), 86.90 (C-1′), 89.16 (C-4′), 111.20 (C-5), 116.73
(
7
(
4
CN), 140.85 and 140.92 (C-6), 151.69 (C-2, CdO), 164.19 (C-
, CdO). Anal. (C12 Cl) C, H, N.
-(1-Cya n a m id o-2-br om oeth yl)-2′-d eoxyu r id in e (5). N-
H
15
N
4
O
5
hydroxy-2-azidoethyl)-2′-deoxyuridine (10) as a syrup (50 mg,
1
5
31.5%): H NMR (Me
2
SO-d
6
) (mixture of two diastereomers
),
Bromosuccinimide (35.6 mg, 0.2 mmol) was added slowly to a
2 3
in a ratio of 1:1) δ 2.08 (m, 2H, H-2′), 3.30 (m, 2H, CH N
pre-cooled (0 °C) solution of 5-vinyl-2′-deoxyuridine (50.8 mg,
3.54 (m, 2H, H-5′), 3.78 (m, 1H, H-4′), 4.22 (m, 1H, H-3′), 4.60
(m, 1H, CHOH), 5.0 and 5.28 (two s, 1H each, 3′-OH, 5′-OH,
exchanges with deuterium oxide), 5.72 (s, 1H, CHOH, ex-
changes with deuterium oxide), 6.18 and 6.20 (two overlapping
t, J ) 6.0 Hz, 1H total H-1′), 7.80 and 7.82 (two s, 1H total,
0
.2 mmol) and cyanamide (33.6 mg, 0.8 mmol) in dimethoxy-
ethane (20 mL). The initial yellow color produced upon addition
of each NBS aliquot quickly disappeared when all the NBS
had reacted. The reaction mixture was stirred (2 h) at 0 °C,
and the solvent was removed in vacuo. The resulting residue
was purified by silica gel column chromatography using
chloroform-methanol (90:10, v/v) as eluent to afford 5-(1-
1
3
H-6), 11.45 (br, 1H, NH, exchanges with deuterium oxide);
C
NMR (Me SO-d ), 61.62
2
6
2 3
) δ 39.78 and 40.06 (C-2′), 55.01 (CH N
and 61.90 (C-5′), 66.14 (CHOH), 70.74 and 70.80 (C-3′), 84.59
cyanamido-2-bromoethyl)-2′-deoxyuridine (5) as a syrup (32
and 84.64 (C-1′), 87.51 and 87.56 (C-4′), 114.32 (C-5), 137.67
1
-1
mg, 42.5%): H NMR (Me
2
SO-d
6
) (mixture of two diastereo-
(C-6), 150.40 (C-2, CdO), 164.5 (C-4, CdO); IR 2118 (N
Anal. (C11 ) C, H, N.
3
) cm .
mers in a ratio of 1:1) δ 2.14 (m, 2H, H-2′), 3.58 (m, 2H, H-5′),
15 5 6
H N O
3
4
.70 (m, 2H, CH
.30 (m, 1H, CHNHCN), 5.06 and 5.28 (2 br s, 1H each, 3′-
2
Br), 3.82 (m, 1H, H-4′), 4.25 (m, 1H, H-3′),
5-(1-Hyd r oxy-2-a zid oeth yl)-2′-d eoxyu r id in e (10). Meth -
od B. Ceric ammonium nitrate (876 mg, 1.6 mmol) was added
slowly to a cooled (ca. -15 °C) flask containing a suspension
of 5-vinyl-2′-deoxyuridine (203 mg, 0.8 mmol) and sodium azide
(57 mg, 0.88 mmol) in dry acetonitrile (200 mL) with stirring.
The reaction was allowed to proceed stirred for 1 h at -15 °C.
The reaction mixture was evaporated in vacuo, and the residue
obtained was redissolved in aqueous acetonitrile (50 mL).
Purification of the product by elution from a silica gel column
using chloroform-methanol (90:10, v/v) as eluent yielded 5-(1-
hydroxy-2-azidoethyl)-2′-deoxyuridine (10) as a syrup (80 mg,
OH, 5′-OH), 6.16 and 6.18 (two overlapping t, J ) 6.0 Hz, 1H
total H-1′), 7.50 (m, 1H, NHCN; exchanges with deuterium
oxide), 7.98 and 8.0 (2s, 1H total, H-6), 11.60 (s, 1H, NH,
exchanges with deuterium oxide). Anal. (C12
H, N.
15 4 5
H N O Br) C,
5
-(1-Cya n a m id o-2-iod oeth yl)-2′-d eoxyu r id in e (6). N-
Iodosuccinimide (116 mg, 0.51 mmol) was added slowly to a
pre-cooled (0 °C) solution of 5-vinyl-2′-deoxyuridine (120 mg,
0
.47 mmol) and cyanamide (80 mg, 1.9 mmol) in dimethoxy-
ethane (20 mL). The reaction mixture was stirred for 1 h at 0
C. The reaction was completed, and the product was purified
as described for 2 to yield 5-(1-cyanamido-2-iodoethyl)-2′-
1
13
32%). The H NMR, C NMR, and FT-IR spectra for 10 were
°
identical to the spectral data described in the method A.
5
-(1-Meth oxy-2-a zid oeth yl)-2′-d eoxyu r id in e (11). Ceric
1
2
deoxyuridine (6) as a syrup (45 mg, 22.5%): H NMR (D O)
ammonium nitrate (219 mg, 0.4 mmol) was added slowly to a
(
mixture of two diastereomers in a ratio of 1:1) δ 2.38 (m, 2H,
cooled (ca. -15 °C) flask containing a suspension of 5-vinyl-
2
H-2′), 3.60-3.82 (complex m, 4H, H-5′, CH I), 4.08 (m, 1H,
2
0
′-deoxyuridine (50.4 mg, 0.2 mmol) and sodium azide (13 mg,
.2 mmol) in dry acetonitrile (50 mL) with stirring. The
H-4′), 4.50 (m, 2H, H-3′, CHNHCN), 6.30 and 6.32 (two
overlapping t, J ) 6.0 Hz, 1H total H-1′), 8.06 and 8.08 (2s,
reaction mixture was stirred for 30 min at -15 °C. The reaction
mixture was evaporated in vacuo, and the residue obtained
was dissolved in methanol (1 mL) and purified on a PTLC
using chloroform-methanol (90:10, v/v) as the development
solvent. Extraction of the ultraviolet visible spot with chloro-
form:methanol (85:15, v/v) yielded 5-(1-methoxy-2-azidoethyl)-
1
H total, H-6). Anal. (C12
H
15
N
4
O
5
I) C, H, N.
5
-(1-Cya n a m id o-2-br om oeth yl)-a r a bin ou r id in e (7). N-
Bromosuccinimide (89 mg, 0.5 mmol) was added slowly with
stirring to a pre-cooled (0 °C) solution of 5-vinyl-arabinouridine
(135 mg, 0.5 mmol) and cyanamide (84 mg, 2.0 mmol) in

Synthesis of Deoxyuridines and Arabinouridines
′-deoxyuridine (11) as a viscous syrup (16 mg, 25%): 1H NMR
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 21 3537
2
NaOH/NaCl at room temperature for 30 min and neutralized
in Tris-HCl/NaCl. The filters were exposed to ultraviolet
irradiation for 3 min. DNA hybridization was initiated by
(
(
2
2
D
2
O) (mixture of two diastereomers in a ratio of 1:1) δ 2.42
m, 2H, H-2′), 3.34 and 3.36 (two s, 3H total, OCH ), 3.56 (m,
), 3.80 (m, 2H, H-5′), 4.08 (m, 1H, H-4′), 4.48 (m,
H, H-3′, CHOCH ), 6.30 and 6.32 (two overlapping t, J ) 6.0
3
3
2
6
H, CH
2
N
3
adding a recently prepared DHBV ( P) DNA probe at 10
3
CPM/mL and incubating overnight. Filters were washed twice
in 1xSSC (20xSSC in 3 M NaCl plus 0.3 M sodium citrate, pH
7.0)-0.1% SDS at 65 °C for 2 h and 1xSSC at room temper-
ature for 30 min. After an autoradiographic image was
obtained, the filters were exposed in a phosphoimaging screen
for 1-2 h and samples were quantitated by a Fujix BAS1000
and the percentage density of phosphoimaging units were
calculated.28 3-TC was used as the reference compound. Tests
were repeated 2-3 times, and the data for each test compound
were compared with a positive and negative control performed
at the same time under identical conditions. For the com-
pounds where the EC50 obtained from three experiments was
within 10% standard deviation, average values are shown,
otherwise a range of EC50 values are shown. Percent inhibition
was calculated by using the following formula: (untreated
positive control - treated test sample) × 100/untreated
positive control.
1
3
Hz, 1H total H-1′), 7.90 and 7.92 (two s, 1H total, H-6);
NMR (D O) δ 39.89 and 40.01 (C-2′), 53.84 (CH ), 57.46 and
7.52 (OCH ), 61.46 and 61.79 (C-5′), 71.13 and 71.20 (C-3′),
6.54 (CHOCH ), 86.48 and 86.55 (C-1′), 87.62 and 87.67 (C-
′), 111.63 (C-5), 140.52 (C-6), 152.12 (C-2, CdO), 165.23 (C-
C
2
2 3
N
5
7
4
4
3
3
-
1
, CdO); IR 2104 (N
3 17 5 6
) cm . Anal. (C12H N O ) C, H, N.
In Vitr o An tivir al Assays [HSV-1 (E-377), HSV-2, HCMV,
VZV]. Plaque reduction and cytopathic effect (CPE) inhibition
assays for HSV-1, HSV-2, HCMV, and VZV were performed
under the NIH Antiviral Research Branch Testing Program
7
using standard procedures described previously.
+
In Vitr o An tivir a l Assa ys [HSV-1 (KOS, TK ), (KOSSB,
-
TK )]. African green monkey kidney (Vero) cells were grown
in DMEM supplemented with 5% FBS. Cells were seeded into
2
4-well plates 1 day prior to the assay. Wild-type HSV-1, KOS,
2
7
and a thymidine kinase deficient mutant, KOSSB, were used
to infect at ∼100 PFU’s per well for 1 h at 37 °C. After
infecting, the inoculum was replaced with serial dilutions of
media containing diluted drug. All drug dilutions were done
Cell Cytotoxicity: Neu tr a l Red Up ta k e Assa y. Cytotox-
icity of test compounds on human foreskin fibroblast (HFF)
cells was determined using neutral red uptake assay as
reported previously.7
in duplicate, initially using the following concentrations: 25,
+
MTT Assa y. Cell viability was measured using the cell
proliferation kit 1 (MTT; Boehringer Mannheim) as per
manufacturer’s instructions. Briefly, a 96-well plate was
1
0, 5, and 1 µg/mL (for TK virus) and 50, 25, 10, and 1 µg/
-
mL (for TK virus). Once an activity range was determined,
compounds were again tested at 1:2 serial dilutions to obtain
a more precise EC50. Controls included infected wells that were
not treated with drugs, as well as infected wells treated with
4
seeded with Vero cells at a density of 2.5 × 10 cells per well.
Cells were allowed to attach for 6-8 h, the media was replaced
with media containing drugs at concentrations of 200, 100, 50,
25, 12.5, 6.3, and 1.5 µg/mL. DMSO was also included as
control. Plates were incubated for 3 days at 37 °C. The color
reaction involved adding 10 µL MTT reagent per well, incubat-
ing 4 h at 37 °C, and then adding 100 µL of solubilization
reagent. Plates were read on an ELISA plate reader (Abs 560-
+
acyclovir at 5 and 1 µg/mL (for TK virus) and 50, 25, 10, and
-
1
°
µg/mL (for TK virus). Plates were incubated for 48 h at 37
C. To visualize plaques, wells were fixed by incubation with
methanol for 10 min at room temperature. The methanol was
aspirated and replaced with 1× Giemsa stain (Sigma) for 1 h
at room temperature. Plaques were counted and compared to
the number of plaques in the no-drug controls in order to
6
50 nm) following an overnight incubation at 37 °C.
Cell P r olifer a tion Assa y: HF F Cells. The effect of test
calculate EC50
.
compounds on proliferation of HFF cells was determined
In Vit r o An t ivir a l Assa y (Du ck Hep a t it is B Vir u s,
DHBV). Pekin duck eggs were obtained from a duck colony
maintained at the University of Alberta farm and were stored
in a 37 °C egg incubator until hatching occurred. Newly
hatched ducklings were infected with a 50 µL intravenous
injection of duck serum containing DHBV. Persistently in-
fected ducks were identified by detection of DHBV DNA in
according to previously reported procedure.7
Hu m a n T Cells. Enriched T cell populations were purified
from buffy coats obtained from normal Canadian Blood Service
donors, by using nylon wool columns according to published
procedures.30 Test compounds dissolved in DMSO at 10 mg/
mL were prepared as stock solutions. In 96-well flat bottom
plates, test compounds were plated in triplicate wells starting
from 50 µg/mL final concentrations. Serial dilutions of com-
pounds were prepared at 1:2, 1:5, or 1:10 dilution for up to
eight dilutions. All of the dilutions were made with serum free
AIM V media for lymphocytes (Life Technologies, ON). Control
wells with the same DMSO concentration as those with test
2
8
sera by Dot hybridization.
Primary cultures of duck hepatocytes were prepared from
to 14 day old DHBV-infected ducklings using a modified
method. Cells were cultured in 60 mm cell culture dishes in
mL of L-15 medium containing 5% fetal bovine serum,
9
2
9
4
penicillin G Sodium (10 IU/mL), streptomycin sulfate (10 µg/
mL), and nystatin (25 U/mL). The test compounds were always
added in triplicate to the hepatocyte cultures on day 2 and
were maintained in culture with media changed every second
day until day 12. Cells were harvested at day 14. Initially,
the test compounds were screened at a 10 µg/mL final
concentration. Inhibition in DHBV replication at 10 µg/mL was
calculated as the average of triplicate wells. Standard devia-
tions were within 10% of the average values. After initial
testing, test compounds were serially diluted to determine
more precise anti-DHBV EC50 values. The hepatocytes were
lysed with 1.0 mL of lysis buffer containing 10 mM Tris-HCl
and 1% SDS. The lysate was digested with 0.5 mg/mL
proteinase K and extracted with an equal volume of phenol
saturated with Tris-HCl EDTA and 0.1% 8-hydroxyquinoline,
followed by extraction with chloroform. Concentrated NaCl
compounds were also prepared in triplicate wells. A total of
5
2
× 10 T cells in AIM V media were added to each of these
prepared wells. This was followed by the addition of phytohe-
magglutinin (PHA, Sigma Chemical Company) to make up a
final concentration of 1 µg/mL of PHA. Positive control wells
with PHA and T cells, and negative control wells with T cells
in media alone, were also prepared on each plate. The cultures
2
were incubated for 3-4 days in 5% CO and 95% humidity at
3
3
7 °C. After 3-4 days, [ H] thymidine (1 µCi/well) was added
to each well. Eighteen hours after pulsing with radioactive
thymidine, the cells were harvested onto nylon filters, and [ H]
3
thymidine incorporation into the DNA of proliferating cells was
measured by liquid scintillation counting. CPM of wells
containing test compounds were compared with CPM of the
wells containing stimulated T cells and corresponding DMSO
concentration.
(5M) was added to the aqueous phase to yield a final concen-
tration of 0.5 M NaCl, and the DNA was precipitated with two
volumes of 95% ethanol. The DNA pellet was washed with 70%
ethanol and dried. The dried DNA was dissolved in 100 µL of
a solution containing Tris-HCl EDTA.
DNA samples were applied to a nylon filter (Hybond-N,
Amersham) using a Bio-Dot (Bio-Rad Laboratories) microfil-
tration apparatus. DNA on the filter was denatured with
Ack n ow led gm en t. We are grateful to the Alberta
Heritage Foundation for Medical Research (AHFMR,
Canada) for an establishment grant (R.K.) and the
Medical Research Council of Canada/Canadian Insti-
tutes of Health Research for an operating grant (MOP-
36393) (R.K. and D.L.J .T.) for the financial support of

3
538 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 21
Kumar et al.
this research. R.K. is a recipient of an AHFMR Medical
Scholar Award. We also thank the United States
National Institutes of Health Antiviral Research Branch,
which provided part of the antiviral test results. We
thank Ms. J yy-Shiang Huang for providing excellent
technical assistance.
(15) Patai, S. The Chemistry of Double-bonded Functional Groups;
Wiley and Sons: New York, 1977; pp 725-912.
16) Dubois, J . E.; Chretien, J . R. Reactivity of Methyl-substituted
Ethylenebromonium Ions. Correlation Between Charge Distri-
bution and Regio- or Chemoselectivity. J . Am. Chem. Soc. 1978,
(
1
00, 3506-3513.
(
17) Van Ende, D.; Krief, A. A New Reagent for Stereospecific
Synthesis of Aziridines from Olefins. Angew. Chem., Int. Ed.
Engl. 1974, 13, 279-280.
Refer en ces
(18) Trahanovsky, W. S.; Robbins, M. D. Oxidation of Organic
Compounds with Cerium (IV). XIV. Formation of R-Azido-â-
nitratoalkanes from Olefins, Sodium azide and Ceric ammonium
nitrate. J . Am. Chem. Soc. 1971, 93, 5256-5258.
(
1) De Clercq, E. Antiviral Activity of 5-Substituted Pyrimidine
Nucleoside Analogs. Pure Appl. Chem. 1983, 55, 623-626.
2) Griengl, H.; Bodenteich, M.; Hayden, W.; Wanek, E.; Streicher,
W.; Stutz, P.; Bachmayer, H.; Ghazzouli, I.; Rosenwirth, B.
(
(
(
(
(
19) Clive, D. L. J .; Etkin, N. Synthesis of R-Amino Acids by Addition
of Putative Azido Radicals to a-Methoxy Acrylonitriles Derived
from Aldehydes and Ketones. Tetrahedron lett. 1994, 35, 2459-
5
-(Haloalkyl)-2′-deoxyuridines: A Novel Type of Potent Antiviral
Nucleoside Analogue. J . Med. Chem. 1985, 28, 1679-1684.
3) De Clercq, E.; Rosenwirth, B. Selective in vitro and in vivo
Activities of 5-(2-Haloalkyl) Pyrimidine Nucleoside Analogs,
Particularly 5-(2-Chloroethyl)-2′-deoxyuridine, Against Herpes
Simplex Virus. Antimicrob. Agents Chemother. 1985, 28, 246-
2
462.
(
20) J ones, A. S.; McClean, M. J .; Slater, M. J .; Walker, R. T.
Synthesis of 5-(1-Substituted ethyl)uracil Derivatives and Some
of Their Chemical and Biological Properties. J . Chem. Soc.,
Perkin Trans. 1 1987, 457-464.
21) Andrei, G.; Snoeck, R.; De Clercq, E. Susceptibilities of Several
Drug-resistant Herpes Simplex Virus Type 1 Strains to Alterna-
tive Antiviral Compounds. Antimicrob. Agents Chemother. 1995,
2
51.
(
4) Shiau, G. T.; Schinazi, R. F.; Chen, M. S.; Prusoff, W. H.
Synthesis and Biological Activities of 5-(Hydroxymethyl, azi-
domethyl, or aminomethyl)-2′-deoxyuridine and Related 5′-
Substituted Analogues. J . Med. Chem. 1980, 23, 127-33.
5) Kumar, R.; Wiebe, L. I.; Hall, T. W.; Knaus, E. E.; Tovell, D. R.;
Tyrrell, D. L.; Allen, T. M.; Fathi-Afshar, R. Synthesis of 5-[1-
Hydroxy(or methoxy)-2-bromo(or chloro)ethyl]-2′-deoxyuridines
and Related Halohydrin Analogues with Antiviral and Cytotoxic
Activity. J . Med. Chem. 1989, 32, 941-944.
3
9, 1632-1635.
22) Tao, P.-Z.; Lofgren, B.; Lake-Bakaar, D.; J ohansson, N. G.;
Datema, R.; Oberg, B. Inhibition of Human Hepatitis B Virus
DNA Polymerase and Duck Hepatitis B Virus DNA Polymerase
by Tri-Phosphates of Thymidine Analogs and Pharmacokinetic
Properties of the Corresponding Nucleosides. J . Med. Virol. 1988,
26, 353-362.
(
(
6) Kumar, R.; Xu, L. H.; Knaus, E. E.; Wiebe, L. I.; Tovell, D. R.;
Tyrrell, D. L.; Allen, T. M. Synthesis and Antiviral and Cytotoxic
Activity of Iodohydrin and Iodomethoxy Derivatives of 5-Vinyl-
(23) Yokota, T.; Konno, K.; Chonan, E.; Mochizuki, S.; Kozima, K.;
Shigeta, S.; De Clercq, E. Comparative Activities of Several
Nucleoside Analogs Against Duck Hepatitis B Virus in vitro.
Antimicrob. Agents.# Chemother. 1990, 34, 1326-1330.
(24) Boyd, M. R.; De Vita, V. T.; Hellman, S.; Rosenberg. S. A., Eds.;
Cancer: In Principles and practice of oncology update; Lippin-
cott: Philadelphia, PA, 1989; Vol. 3, pp 1-12.
(25) Rahim, S. G.; Duggan, M. J . H.; Walker, R. T.; J ones, A. S.; Dyer,
R. L.; Balzarini, J .; DeClercq, E. Synthesis and Biological
Properties of 2′-Deoxy-5-vinyluridine and 2′-Deoxy-5-Vinylcyti-
dine. Nucleic Acids Res. 1982, 10, 5285-5295.
2
′-deoxyuridines, 2′-Fluoro-2′-deoxyuridine, and Uridine. J . Med.
Chem. 1990, 33, 717-723.
(
7) Kumar, R.; Wiebe, L. I.; Knaus, E. E. Synthesis and Antiviral
Activity of Novel 5-(1-Azido-2-haloethyl) and 5-(1-Azido-, amino-,
or methoxyethyl) analogues of 2′-deoxyuridine. J . Med. Chem.
1
993, 36, 2470-2474.
(
8) Morin, K. W.; Kumar, R.; Knaus, E. E.; Wiebe, L. I. Design and
Synthesis of the 3′-Aminocarbonylamino, Aminothiocarbonyl-
amino and N-(Hydroxy)guanidinyl Derivatives of Thymidine
as Potential Anti-HIV Agents. J . Heterocycl. Chem. 1991, 28,
(
(
(
26) J ones, A. S.; Verhelst, G.; Walker, R. T. The Synthesis of the
Potent Anti-herpes Virus Agents, E-5-(2-Bromovinyl)-2′-deoxy-
uridine and Related Compounds. Tetrahedron Lett. 1979, 45,
8
07-809.
(
9) Patany, G. A.; LaVoie, E. J . Bioisosterism: A Rationale Approach
in Drug Design. Chem. Rev. 1996, 96, 3147-3176.
4
415-4418.
(
10) Hansch, C.; Leo, A.; Unger, S. H.; Kim, K. H.; Nikaitani, D.;
Lien, E. J . “Aromatic” Substituent Constants for Structure-
activity Correlations. J . Med. Chem. 1973, 16, 1207-1216.
11) Balzarini, J .; Andrei, G.; Kumar, R.; Knaus, E. E.; Wiebe, L. I.;
De Clercq, E. The Cytostatic Activity of 5-(1-Azidovinyl)-2′-
deoxyuridine (AZVDU) Against Herpes Simplex Virus Thymi-
dine Kinase Gene-transfected FM3A Cells is Due to Inhibition
of Thymidylate Synthase and Enhanced by UV light (lambda )
27) Varmusza, S. L.; Smiley, J . R. Signals for Site-specific Cleavage
of HSV-DNA: Maturation Involves Two Separate Cleavage
Events at Sites Distal to the Recognition Sequences. Cell 1985,
(
4
1, 793-802.
28) Lee, B.; Luo, W.; Suzuki, S.; Robins, M. J .; Tyrrell, D. L. J . In
vitro and in vivo Comparison of the Abilities of Purine and
Pyrimidine 2′,3′-Dideoxynucleosides to Inhibit Duck Hepadnavi-
rus. Antimicrob. Agents Chemother. 1989, 33, 336-339.
2
54 nm) Exposure. FEBS Lett. 1995, 373, 41-44.
(
29) Tuttleman, J . S.; Pugh, J . C.; Summers, J . W. In vitro Experi-
mental Infection of Primary Duck Hepatocyte Cultures with
Duck Hepatitis B Virus. J . Virol. 1986, 58, 17-25.
30) Agrawal, B.; Reddish, M. A.; Longenecker, B. M. CD30 Expres-
sion on Human CD8+ T cells Isolated from Peripheral Blood
Lymphocytes of Normal Donors, J . Immunol. 1996, 157,
(
12) Szinai, E.; Bihari, M.; Ujszaszy, K.; Veres, Z.; Szaboles, A.; Holly,
S.; Gacs-Baitz, E.; Otves, L. Bioorganic Heterocycles; Vander
Plas, H. C., Otves, L., Simonyi, M., Eds.; Elsevier Press:
Amsterdam, 1985; p 335.
(
(
13) Dalton, D. R.; Dutta, V. P.; J ohns, D. C. Bromohydrin Formation
in Dimethyl sulfoxide. J . Am. Chem. Soc. 1968, 90, 5498-5501.
14) Freeman, F. Possible Criteria for Distinguishing Between Cyclic
and Acyclic Activated Complexes and among Cyclic Activated
Complexes in Addition Reactions. Chem. Rev. 1975, 75, 439-
3
229-3234.
(
4
90.
J M010226S