Design and synthesis of lipophilic phosphoramidate d4T-MP prodrugs expressing high potency against HIV in cell culture: Structural determinants for in vitro activity and QSAR

Article abstract of DOI:10.1021/jm9807104

A series of new substituted-aryl phosphoramidate derivatives of the anti-HIV drug d4T were synthesized as membrane-soluble nucleotide prodrugs, to extend and quantify the SAR observed for an earlier series of related derivatives. All of the compounds were

Full text of DOI:10.1021/jm9807104

4
122
J . Med. Chem. 1999, 42, 4122-4128
Design a n d Syn th esis of Lip op h ilic P h osp h or a m id a te d 4T-MP P r od r u gs
Exp r essin g High P oten cy Aga in st HIV in Cell Cu ltu r e: Str u ctu r a l Deter m in a n ts
for in Vitr o Activity a n d QSAR
†
Adam Q. Siddiqui, Christopher McGuigan,* Carlo Ballatore, Fabio Zuccotto, Ian H. Gilbert, Erik De Clercq, and
J an Balzarini†
Welsh School of Pharmacy, Cardiff University, King Edward VII Avenue, Cardiff CF1 3XF, U.K., and
Rega Institute for Medical Research, Katholieke Universiteit Leuven, B-3000 Leuven, Belgium
Received December 18, 1998
A series of new substituted-aryl phosphoramidate derivatives of the anti-HIV drug d4T were
synthesized as membrane-soluble nucleotide prodrugs, to extend and quantify the SAR observed
for an earlier series of related derivatives. All of the compounds were found to be significantly
more potent against HIV in cell culture than the nucleoside analogue d4T, and most were also
found to be significantly more potent than the parent phosphoramidate. A Hansch type QSAR
analysis was applied to the combined series of 21 compounds. The results of this analysis
revealed anti-HIV activity to be principally dependent on lipophilicity in a quadratic manner,
with terms representing substituent steric bulk and electronic effects having a minimal
significance.
In tr od u ction
Ch a r t 1. Anti-HIV Nucleoside Analogue d4T and
Various Phosphoramidate Prodrugs
Various 2′,3′-dideoxy-2′,3′-didehydro nucleosides are
known to be potent inhibitors of HIV, the causative
agent of AIDS.1 In particular, the 2′,3′-dideoxy-2′,3′-
didehydro analogue of thymidine, d4T (1, Chart 1), has
been shown to exhibit a selective anti-HIV activity in
cell culture and also a reduced toxicity in certain cell
types (e.g. bone marrow progenitor cells) compared to
-6
the antiretroviral agent 3′-azido-2′,3′-dideoxythymidine
AZT).4
,7,8
In common with other dideoxy-didehydro
(
nucleosides, d4T depends on host cell kinase activity to
produce its active triphosphorylated form. Specifically,
it is the monophosphorylation step which is rate-
limiting for d4T, in contrast to AZT where phosphoryl-
ation to the nucleoside 5′-diphosphate is the slowest step
9
,10
in the phosphorylation pathway.
Several of these
types of nucleoside analogues are known to be poor
substrates for nucleoside kinases and 5′-nucleoti-
1
1-15
dases.
These findings have prompted the develop-
ment of prodrug strategies that bypass initial kinase
dependency by the intracellular delivery of the mono-
phosphorylated nucleoside analogue (reviewed by J ones
and Bischofberger16 and recently by Meier ).
17
Previously we have demonstrated the success of an
approach using the aryloxy phosphoramidate deriva-
Chart 1), with a p-chloro substituent on the aryl moiety,
was found to express a 14-fold greater activity compared
to the underivatized parent structure (2) against HIV-
1
8-20
tives of d4A, ddA, and d4T (2, Chart 1).
These
compounds were shown to exhibit greatly enhanced
activity against HIV compared to the parent nucleoside
analogues in vitro, and in contrast to the nucleoside
analogues, full activity was retained in kinase-deficient
cell lines.
1
, in cell culture (EC50[2e] ) 0.005 µM, EC50[2] ) 0.075
µM). Correspondingly, 2e was approximately 10-fold
more lipophilic than 2 based on the 1-octanol/water
partition coefficient (P). It was evident that highly
electron-withdrawing substituents such as p-cyano or
p-nitro had a deleterious effect on antiviral activity. The
observed instability of these compounds in aqueous
buffer (pH 7.6, t1/2 ∼ 150 h) suggested that one possible
explanation was an appreciable decomposition of the
prodrug to less lipophilic derivatives prior to membrane
penetration, with a subsequent reduction in rate of
In recent work, we were able to demonstrate a clear
dependence of antiviral activity on lipophilicity for a
series of aryl-substituted phosphoramidate prodrugs of
2
1
d4T. The most active compound from series I (2e,
*
Author for correspondence: tel/fax, +44 1222 874537; e-mail,
mcguigan@cardiff.ac.uk.
†
21
Katholieke Universiteit Leuven.
diffusion. A carboxy esterase model was introduced to
1
0.1021/jm9807104 CCC: $18.00 © 1999 American Chemical Society
Published on Web 09/21/1999

Phosphoramidate d4T-MP Prodrugs
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 20 4123
Ta ble 1. Substituent Parameters, log P, Anti-HIV-1 and Anti-HIV-2 Activity, and Cytotoxicity of Compounds 2a -2t, 2, and d4T (1)
in Vitro
EC50a (µM)
CC50b (µM)
SIc
compd
MR
σ
log P
CEM/0 (HIV-1)
CEM/0 (HIV-2)
CEM/TK- (HIV-2)
CEM/0
CEM/0 (HIV-1)
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
p
q
r
7.36
6.33
12.87
11.18
6.03
0.92
6.65
7.87
11.18
7.86
12.06
0.92
6.03
8.88
13.94
6.03
8.88
2.097
5.02
5.02
1.03
0.78
0.66
0.45
0.50
0.23
0.06
-0.17
-0.27
0.38
0.35
0.52
0.34
0.23
0.23
0.18
0.37
0.39
0.35
0.54
0.43
0.0
0.90
0.91
1.01
0.90
1.43
1.10
1.19
1.09
0.92
1.86
2.15
1.08
1.37
1.60
1.93
1.42
1.64
1.78
1.77
1.57
1.04
-0.59
0.13
0.19
0.09
0.09
0.025
0.05
73
60
562
400
1240
363
0.15
0.13
0.025
0.11
0.005
0.053
0.04
0.057
0.08
0.014
0.016
0.018
0.018
0.009
0.013
0.025
0.015
0.008
0.01
0.022
0.1
0.007
0.06
0.05
0.053
0.09
0.03
0.03
0.035
0.029
0.016
0.015
0.057
0.016
0.007
0.01
31
40
0.006
0.022
0.025
0.047
0.045
0.03
11
2200
490
26
34
850
35
614
50
625
34
2429
1313
1667
2611
1889
1154
1040
867
0.03
21
0.065
0.065
0.018
0.029
0.04
30
47
17
15
26
0.014
0.016
0.018
0.0054
0.075
33
13
25
3125
1300
1077
>1333
267
s
t
13
0.013
0.075
0.651
0.013
0.075
0.770
14
>100
174
a
EC50 is the 50% effective compound concentration required to protect CEM cells against the cytopathicity of HIV by 50%. Data are
b
the mean of 2-4 independent experiments. CC50 is the 50% cytostatic compound concentration required to inhibit CEM cell proliferation
c
by 50%. SI is the selectivity index or ratio CC50/EC50 for HIV-1 in CEM/0 cells. EC50 and CC50 data for 1 are taken from McGuigan et
1
9
al.
mimic the putative initial bioactivation step for these
phosphoramidates but was rejected as a quantitative
model for biological activity.21 Furthermore, regioiso-
meric substitution with the acetyl group suggested an
independence of antiviral activity on aromatic substitu-
ent position.
nantly equal to or greater than that of the chlorine atom
(π ) 0.71). The order of the measured values of P was
in good agreement with the lipophilic order predicted
by the Hansch values (π) of the aryl substituents. For
the para-halogenated series, the lipophilicity values
varied by approximately 1 order of magnitude (log P[2l]
Here we present the synthesis and biological activity
of a range of novel aryl-substituted analogues of 2,
designed to probe the relationships previously observed
and to provide a range of highly active phosphoramidate
prodrugs. In addition to the data previously published,
the current data provided the basis for QSAR and thus
a powerful tool to predict the antiviral activities of
further prodrugs, designed to satisfy particular phar-
macological requirements.
)
1.08, log P[2o] ) 1.93), and as expected the various
meta-substituted compounds were found to have similar
log P values to those observed for their regioisomers.
Comparative data for d4T indicates significantly in-
creased lipophilicity (ca. 50-500-fold) for all the phos-
phoramidates in series II relative to the nucleoside
analogue.
An tivir a l Activity. All of the phosphoramidates 2j-
2
t were evaluated for their ability to inhibit the replica-
Resu lts a n d Discu ssion
25
tion of HIV, as previously described. The results
obtained using HIV-1- or HIV-2-infected CEM cells are
displayed in Table 1. The tests were also conducted
against HIV-2 in thymidine kinase-deficient (TK ) CEM
cells. Previously obtained data for d4T, 2, and 2a -2i
The synthesis of d4T from thymidine was conducted
2
2
essentially by the method of Horwitz taking into
account the later comments of Mansuri.23 Aryloxyphos-
phorodichloridates and aryloxyaminophosphorochlori-
-
1
9
2
4,25
dates were prepared entirely as previously noted.
are also displayed.
Novel phosphoramidates 2j-2t (series II, Chart 1) were
All of the synthesized derivatives were significantly
more potent than d4T against HIV in cell culture, with
several compounds showing a 10-fold superior selectivity
index (SI) to the nucleoside analogue. The majority of
the derivatives were also significantly more potent
compared to the underivatized parent compound 2, and
all the derivatives expressed a similar selectivity index
to the lead compound (SI[2] > 1333). The elevations in
activity observed for the p-bromo and p-iodo derivatives
compared to the underivatized parent structure (6- and
synthesized according to the method described previ-
1
9
ously. The structure and purity of compounds 2j-2t
were determined by extensive multinuclear NMR and
analytical HPLC using two distinct eluent gradients.
Com p ou n d P a r tition Coefficien ts. The partition
coefficients (P) of compounds 2j-2t were determined in
1
-octanol/aqueous buffer (pH 7.0) (results shown in
Table 1). Each compound was partitioned between
-octanol and aqueous buffer for 5 min. After separation
1
of the two layers the relative proportion of compound
in each layer was determined by comparing the UV
absorption of the octanol phase with that of a control
solution. This method of determining the partition
coefficients allows for a quantitative comparison be-
tween the compounds in this series. The substituents
chosen were those whose Hansch values were predomi-
9
-fold, respectively) were of a similar magnitude to that
observed previously for the p-chloro derivative (14-fold).
It is notable that the increased compound lipophilicity,
conferred by bromo and iodo substituents, did not
enhance antiviral activity beyond that observed for the
p-chloro derivative, indicating a possible second-order

4
124 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 20
Siddiqui et al.
Sch em e 1. Proposed Activation Pathway for Phosphoramidate Prodrugs
polynomial relationship between lipophilicity and an-
tiviral activity. In common with the derivatives from
series I, all the compounds from series II retained full
antiviral activity in thymidine kinase-deficient cells,
which is indicative of an effective kinase bypass, result-
ing from efficient intracellular delivery of the free
nucleotide.
The release of variously substituted phenols upon
prodrug bioactivation may lead to differences in specific
phenol toxicty, for example, via the uncoupling of
oxidative phosphorylation.26,27 The lack of a consistent
trend between the constituent properties of the released
phenol (i.e. lipophilicity and acidity) and the selectivity
index values for the corresponding prodrug suggests
that toxicity is relatively independent of the chemical
nature of the released phenol, within the scope of the
substituents used in the series.
The increased potency of the more lipophilic prodrugs
offers the possibility of reducing the observed clinical
side effects of d4T (peripheral neuropathy and liver
damage/inflammation) by using significantly reduced
concentrations of the prodrug to produce an equivalent
clinical response. This is a desirable goal given the
indefinite time periods over which highly active anti-
retroviral therapies (HAART) for HIV are currently
administered.
substituent position by the putative enzyme(s) respon-
sible for bioactivation. Exceptions to this general ob-
servation are the m-fluoro (2l) and m-chloro (2p )
derivatives which respectively displayed a greater and
lesser antiviral potency than might have been predicted
from the activities of their para-counterparts (2f and
2e, respectively). The o-chloro derivative 2m displayed
a similar antiviral activity to the p-chloro derivative 2e-
within the limits of experimental accuracy-indicating
no apparent o-effect for this particular substituent.
The high in vitro potency of the p-chloro derivative
2e, suggested the synthesis of the 3,4-dichloro derivative
2k according to the SAR methodology outlined by
2
9
Topliss. This was found to be highly active compared
to 2 but certainly no more active than the p-chloro
derivative 2e. This suggests that disubstitution on the
aryl ring is not restrictive in terms of in vitro activity
and opens the way to further manipulate the properties
of the aromatic ring by using combinations of substit-
uents. The superior lipophilicity of the disubstituted
compound compared to 2e (log P[2k ] ) 2.15, log P[2e]
) 1.42) further supports the suggestion of a nonlinear
relationship between antiviral activity and lipophilicity.
QSAR. A Hansch type analysis was conducted on all
21 compounds from series I and II, using physicochem-
ical constants taken from Hansch and Leo (Table 1).3
Correlations were sought between antiviral activity in
cell culture (as measured by EC50[HIV-1]) and various
physicochemical parameters associated with each sub-
0,31
In the putative bioactivation pathway for these phos-
phoramidates (Scheme 1), the initial step is believed to
be the carboxy esterase-mediated hydrolysis of the
2
8
2
amino ester moiety. The extent to which the aryl group
may be derivatized without affecting the ability of the
compound to behave as a substrate is currently un-
known. To further test for constraints in the position of
substitution, the meta-substituted regioisomers of vari-
ous para-substituted derivatives were synthesized. For
the majority of these regioisomeric compounds, no
appreciable differences in antiviral potency were ob-
served, indicating an apparent tolerance to specific aryl
stituent or its effects: log P and (log P) to represent
compound lipophilicity, Hammet σ values to represent
the substituent electronic effects, and MR, the molar
refractivity, as an approximate measure of substituent
steric bulk, taking into account the comments of Hansch
3
1
et al. The two most significant equations generated
are given below. Partial least-squares (PLS) computa-
tional methods were applied using cross-validation to
determine the optimum model composition and ‘boot-

Phosphoramidate d4T-MP Prodrugs
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 20 4125
strapping’ iterative techniques to calculate the standard
errors of the coefficients (given in parentheses). Also
given for each equation are the squared correlation
the pKa values for p-methoxyphenol and p-nitrophenol,
respectively). In previous work, the replacement of the
phenoxy moiety with inferior leaving groups such as
methoxy and ethoxy (methanol pKa ) 15, ethanol pKa
2
coefficients (r ), the standard error of estimate (s), the
2
19
F statistic, and Q , the cross-validated, squared correla-
) 16) produced inactive compounds. The full range of
tion coefficient.
conjugate acid pKa values (particularly the upper limit)
that would characterize a biologically efficient leaving
group remains to be determined, as does the possibility
of replacing the phenoxy group with alternative chemi-
cal structures with acceptable conjugate acid pKa values.
log 1/EC [HIV-1] ) -2.793 + 0.075(0.187)σ +
5
0
0
.009(0.016)MR + 5.588(1.548)log P -
.661(0.561)(log P) (1)
2
1
The high dependence of biological activity on lipophi-
licity alone suggests that the scope of this analysis is
sufficiently broad to allow us to predict the antiviral
activities of a wide variety of substituted-aryl deriva-
tives of 2. In the production of a medicinally viable drug
it is often important to improve the pharmacological
profile of a lead compound without diminishing its in
vitro potency. Stability in vivo, bioavailability, and
water solubility are example areas for potential opti-
mization. The apparent insensitivity to substituents of
modest bulk with variable electron-donating/withdraw-
ing characteristics at the aryl moiety suggests the
potential for optimizing prodrug pharmacological pa-
rameters by simple derivatization at this site. Several
of the compounds from series I were found to have an
enhanced aqueous solubility compared to the lead
compound 2 (data not shown). Of these only the p-ester-
substituted derivative 2c displayed an elevated antiviral
potency compared to 2. Significantly, although 2c is
approximately 12 times more water-soluble than 2
2
n ) 21, s ) 0.208, r ) 0.814, F4
)
,16
2
1
7.519, F4,16R 0.001 ) 7.94, Q ) 0.685
log 1/EC [HIV-1] ) -2.440 + 5.157(1.117)log P -
5
0
2
1
.499(0.395)(log P) (2)
2
n ) 21, s ) 0.200, r ) 0.807, F2
)
,18
2
3
7.735, F2,18R 0.001 ) 10.40, Q ) 0.745
By calculating the t statistic for each of the descriptors
in eq 1, it can be shown that the lipophilic terms log P
and (log P) are both significant at the 1% level, whereas
the terms MR and σ are both rejected at the 50% level.
Additionally, a comparison between the values of Q
reveals eq 2, presenting a simple quadratic relationship
between lipophilicity and antiviral activity, to be the
more robust of the two and thus the most likely to give
accurate predictions. Using this equation, a reasonable
estimate of optimum prodrug lipophilicity can be made
2
2
(
)
saturation concentrations [water]: 2 ) 5.4 mmol/L, 2c
67.0 mmol/L) the lipophilicity of the two compounds
(
log Poptimum ) 1.72).
These results suggest that lipophilicity is the major
is approximately equal.
It should be noted that the inclusion in the QSAR of
data from a range of compounds bearing halogeno
substituents necessarily introduces an element of col-
linearity between the terms used in the analysis, for
example, between MR and log P. The relatively small
contribution made by MR to eq 1 suggests that this is
not a significant source of error. A further series of
compounds bearing aryl substituents with larger values
of MR than those used in this analysis but similar
values of log P and σ would be useful for identifying
trends for substituents of high steric bulk. Similarly,
compounds with comparable values of MR and log P but
varying values of σ would help to examine the influence
of electronic effects.
determinant for in vitro activity. The low significance
of the terms σ and MR in eq 1 demonstrates the
relatively minor contribution to antiviral activity made
by substituent electronic or steric properties, within the
parameters of the analysis. An identical treatment of
the antiviral data for HIV-2 revealed similar quantita-
tive and qualitative relationships to those observed for
HIV-1 (data not shown).
The QSAR analysis is consistent with cellular trans-
port by passive diffusion being a major determinant of
in vitro activity. The absence of a deleterious effect on
in vitro activity when the aryl substituent is relatively
bulky or when the substituent position is varied or when
the aromatic ring is disubstituted suggests that any
interactions between the phenoxy moiety of the prodrug
and the esterase(s) proposed for the first step in intra-
cellular activation are not sufficiently restrictive to
significantly decrease the rate of hydrolysis. Further-
more, the absence of a marked influence on in vitro
activity exerted by the electron-withdrawing/donating
characteristics of the aryl substituent suggests that
although the leaving group ability of the phenoxy group
is important (to allow selectivity between the cleavage
of the P-OAr bond in preference to the P-O nucleoside
bond or the P-N bond) it is not sensitive, unless the
substituent exerts an electron-withdrawing effect of
sufficient magnitude to enable substantial phosphotri-
ester hydrolysis prior to diffusion through cellular
membranes. On the basis of the current study, an
efficient leaving group would be one with a conjugate
acid pKa value within the range 10.2-7.2 (defined by
In conclusion, we have synthesized a range of new,
highly potent anti-HIV phosphoramidate prodrugs, the
majority of which exhibited significantly higher anti-
HIV activity than our lead compound 2 and all of which
were significantly more potent and selective compared
to the parent nucleoside analogue d4T. We have dem-
onstrated antiviral activity to be principally dependent
on lipophilicity and in a quadratic manner, with steric
and electronic effects adding relatively minor contribu-
tions. Additionally, the position of the monosubstituent
had no consistently obvious effect on in vitro antiviral
activity.
By the appropriate use of the developed QSAR, it is
our intention to design and synthesize a range of
pharmacologically optimized prodrugs based on the
phosphoramidate pronucleotide concept, by simple de-
rivatization at the aryl moiety. Work to determine the

4
126 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 20
Siddiqui et al.
extent to which these findings can be applied to the
delivery of other therapeutic nucleoside analogues is
currently in progress.
NH); δ
C
12.80, 12.84 (5-Me), 21.27, 21.41 (d, J 5.66, J 5.64,
Ala-Me), 50.54, 50.62 (Ala-CH), 53.17 (OMe), 67.17, 67.87 (d,
J 0, J 5.0, C-5′), 84.85, 84.95 (C-4′), 90.06, 90.34 (C-1′), 111.68,
1
11.84 (C-5), 120.22, 120.39 (d, J 5.1, Ph), 122.75, 122.89 (d,
J 5.0, Ph), 126.27 (Ph), 128.03, 128.13 (C-2′), 129.50 (Ph),
Exp er im en ta l Section
1
1
1
31.44 (Ph), 133.22, 133.53 (C-3′), 135.89, 136.10 (C-6), 149.45,
49.56 (d, J 6.0, J 6.2, Ph), 151.21 (C-2), 164.13, 164.16 (C-4),
All experiments involving water-sensitive compounds were
conducted under scrupulously dry conditions. Triethylamine
was dried by refluxing over calcium hydride. Anhydrous
tetrahydrofuran
ane were purchased from Aldrich. N-Methylimidazole was
purified by distillation. Nucleosides were dried by storage at
2 5
elevated temperature over P O in vacuo. Proton, carbon, and
phosphorus nuclear magnetic resonance ( H, C, P NMR)
spectra were recorded on a Bruker Avance DPX spectrometer
74.14, 174.32 (d, J 7.2, J 13.4, Ala-CO); MS m/e FAB 534.0631
+
(MH , C20
H
22Cl
2
N
3
O
8
P requires 534.0600); HPLC t
R
35.66,
and
dichlorometh-
3
6.00 min; t
R
38.56, 38.97 min (gradient II).
2
′,3′-Did eh ydr o-2′,3′-d id eoxyth ym id in e-5′-(3-flu or oph e-
n yl m eth oxya la n in ylp h osp h a te) (2l): yield 26%; δ 4.31,
.66; δ 1.31 (3H, m, Ala-Me), 1.84 (3H, d, 5-Me), 3.67 (3H, s,
OMe), 3.89 (2H, m, Ala-NH, Ala-CH), 4.27 (2H, m, H-5′), 4.99
1H, m, H-4′), 5.87 (1H, m, H-3′), 6.28 (1H, m, H2′), 6.91 (4H,
m, 2 Ar-H, H-1′, H-6), 7.22 (3H, m, Ar-H), 8.37 (1H, d, NH);
11.32, 11.37 (5-Me), 19.83, 19.90 (Ala-Me), 51.66 (OMe),
65.61, 66.26 (C-5), 83.55, 83.61 (C-4′), 88.58, 88.84 (C-1′),
P
3
H
1
13
31
(
1
3
operating at 300, 75.5, and 121.5 MHz, respectively. All
C
3
1
and P spectra were recorded proton-decoupled. All NMR
spectra were recorded in CDCl at room temperature (20 ( 3
δ
C
3
13
1
°
C). Chemical shifts for H and C spectra are quoted in parts
1
1
1
07.14, 107.47 (Ph), 110.27, 110.42 (Ph), 111.15, 111.44 (C-5),
15 (Ph), 126.47, 126.61 (C-2′), 129.47, 129.60 (Ph), 131.91,
32.22 (C-3′), 134.52, 134.76 (C-6), 149.80 (C-2), 159.53 (Ph),
per million downfield from tetramethylsilane. Coupling con-
stants are referred to as J values. Signal splitting patterns
are described as singlet (s), doublet (d), triplet (t), quartet (q),
or multiplet (m). Chemical shifts for 31P spectra are quoted in
100.32 (Ph), 162.69, 162.75 (C-4), 172.80, 172.96 (d, J 6.9, J
+
6
.2, Ala-CO); MS m/e FAB 484.1285 (MH , C20
H
24
N
3
O
8
FP
parts per million relative to an external phosphoric acid
standard. Many proton and carbon NMR signals were split
due to the presence of (phosphate) diastereoisomers in the
samples. The mode of ionization for mass spectrometry was
fast atom bombardment (FAB) using MNOBA as matrix.
Column chromatography refers to flash column chromatogra-
phy carried out using Merck silica gel 60 (40-60 µm) as
stationary phase. HPLC (Shimadzu) was conducted on an
SSODS2 reverse-phase column using a water/acetonitrile
requires 484.1285); HPLC t 30.37, 30.69 min; t 32.34, 32.72
R R
min (gradient II).
2′,3′-Dideh ydr o-2′,3′-dideoxyth ym idin e-5′-(2-ch lor oph e-
n yl m eth oxya la n in ylp h osp h a te) (2m ): yield 45%; δ 3.59,
4.34; δ 1.38 (3H, m, Ala-Me), 1.90 (3H, d, 5-Me), 3.76 (3H, d,
P
H
OMe), 4.05 (2H, m, Ala-NH, Ala-CH), 4.42 (2H, m, H5′), 5.09
(1H, m, H4′), 5.96 (1H, m, H3′), 6.38 (1H, m, H2′), 7.07 (1H,
m, H-1′), 7.26 (3H, m, Ph, 1H-6), 7.48 (2H, m, Ph), 8.44 (1H,
(
0
Fisher, HPLC grade) eluent; gradient I (standard gradient):
-80% CH CN (0-60 min), 80-0% CH CN (60-65 min), flow
rate 1 mL/min, UV detection at 265 nm; gradient II: 0-10%
CH CN (0-5 min), 10-70% CH CN (5-55 min), 70-0% CH
C
s, NH); δ 15.04, 15.09 (5-Me), 23.64, 23.82 (d, J 13.1, Ala-
3
3
Me), 52.88, 52.96 (Ala-CH), 55.43 (OMe), 69.80, 70.19 (C-5′),
87.21, 87.32 (C-4′), 92.42, 92.55 (C-1′), 114.21, 114.24 (d, J 2.7,
Ph), 124.35, 124.72 (C-2′), 128.80, 128.90 (Ph), 130.15, 130.23
(C-3′), 130.80 (Ph), 133.34 (Ph), 135.90, 136.05 (C-6), 138.40,
138.55 (d, J 6.8, J 6.0, Ph), 149.33 (Ph), 153.44 (C-2), 166.22,
166.27 (C-4), 176.35, 176.67 (d, J 7.7, J 11.9, Ala-CO); MS m/e
3
3
3
-
CN (55-60 min), flow rate 1 mL/min, UV detection at 265 nm.
Final products showed purities exceeding 99% with undetect-
able levels (<0.02) of parent nucleosides in every case. Parti-
tion coefficients were determined using 1-octanol and phos-
phate buffer, pH 7.0 (Fisons). UV absorptions were determined
using a Unicam Helios R UV spectrometer.
Gen er a l P r oced u r e. Aryl methoxyalaninylphosphorochlo-
ridate (5 mmol) was added to a stirred solution of d4T (1.7
mmol) and N-methylimidazole (5 mmol) in tetrahydrofuran
+
FAB 500.0750 (MH , C H ClN O P requires 500.0990);
2
0
23
3
8
HPLC tR 32.33, 32.88 min; tR 34.78, 35.42 min (gradient II).
′,3′-Dideh ydr o-2′,3′-dideoxyth ym idin e-5′-(4-br om oph e-
n yl m eth oxya la n in ylp h osp h a te) (2n ): yield 46%; δ 4.57,
.93; δ 1.39 (3H, m, Ala-Me), 1.90 (3H, d, 5-Me), 3.71 (3H, s,
OMe), 4.02 (2H, m, Ala-NH, Ala-CH), 4.36 (2H, m, H-5′), 5.08
1H, m, H-4′), 5.93 (1H, m, H-3′), 6.36 (1H, m, H-2′), 7.07 (1H,
m, H-1′), 7.12 (2H, m, meta-Ar), 7.29 (1H, m, H-6), 7.48 (2H,
12.80, 12.84 (5-Me), 21.34
t, J 4.5, Ala-Me), 50.48, 50.60 (Ala-CH), 53.11 (OMe), 67.01,
2
P
3
H
(THF) (20 mL) at ambient temperature. After 16 h, the solvent
(
was removed under reduced pressure. The residual gum was
dissolved in chloroform (50 mL) and washed with 1 M HCl
(
m, ortho-Ar), 9.31 (1H, d, NH); δ
(
C
50 mL), sodium bicarbonate solution (50 mL), and water (50
mL). The organic phase was dried (MgSO ) and the solvent
4
6
9
1
1
1
7.73 (d, J 4.6, J 5.2, C-5′), 84.90, 85.02 (d, J 1.9, J 2.0, C-4′),
0.00, 90.26 (C-1′), 111.68, 111.83 (C-5), 118.51, 118.52, 118.59,
18.61 (Para-Ar), 122.29, 122.35, 122.45, 122.52 (Ar), 127.87,
28.03 (C-2′), 133.15, 133.19 (Ph), 133.34, 133.68 (C-3′), 135.98,
36.26 (C-6), 149.72 149.79, 149.86, 149.95 (ipso-Ar), 151.32
removed under reduced pressure. The residue was purified by
column chromatography on silica with elution by dichlo-
romethane-methanol (97:3). Pooling of appropriate fractions,
followed by removal of solvent under reduced pressure, gave
the product as a brittle white foam.
(
C-2), 164.27, 164.32 (C-4), 174.18, 174.28, 174.38, 174.47 (Ala-
2
′,3′-Did eh yd r o-2′,3′-d id eoxyt h ym id in e-5′-(4-(t r iflu o-
+
CO); MS m/e FAB 544.0476 (MH , C20
00.0484); HPLC t 34.22, 34.65 min; t
gradient II).
′,3′-Did eh yd r o-2′,3′-d id eoxyth ym id in e-5′-(4-iod op h e-
n yl m eth oxya la n in ylp h osp h a te) (2o): yield 50%; δ 4.47,
.83; δ 1.39 (3H, m, Ala-Me), 1.88 (3H, d, 5-Me), 3.77 (3H, s,
H
24
N
3
O
8
PBr requires
r om eth oxy) ph en yl m eth oxyalan in ylph osph ate) (2j): yield
6
3
5
(
R
R
37.54, 39.07 min
5%; δ
P
H
4.18, 4.81; δ 1.34 (3H, m, Ala-Me), 1.84 (3H, d, 5-Me),
.70 (3H, s, OMe), 3.97 (2H, m, Ala-NH, Ala-CH), 4.31 (2H,
2
m, H-5′), 5.10 (1H, m, H-4′), 5.89 (1H, m, H-3′), 6.32 (1H, m,
H-2′), 7.20 (1H, m, H-1′), 7.27 (5H, m, H-6, Ar-H), 9.83 (1H,
P
3
H
d, NH); δ
C
12.72 (5-Me), 21.16 (Ala-Me), 50.56 (Ala-CH), 52.99
), 66.97, 67.72 (C-5′), 84.95 (C-4′), 89.94,
(OMe), 53.85 (CF
3
OMe), 3.99 (2H, m, Ala-NH, Ala-CH), 4.35 (2H, m, H-5′), 5.06
(1H, m, H-4′), 5.95 (1H, m, H-3′), 6.36 (1H, m, H-2′), 7.01 (2H,
m, meta-Ar), 7.07 (1H, m, H-1′), 7.28 (1H, m, H-6), 7.66 (2H,
9
1
1
1
5
3
0.21 (C-1′), 111.63, 111.78 (C-5), 121.81, 121.98 (Ph), 122.64,
22.82 (Ph), 127.90, 127.95 (C-2′), 133.28, 133.61 (C-3′), 135.96,
36.22 (C-6), 146.28 (Ph), 149.02 (Ph), 151.44 (C-2), 164.00,
C
m, ortho-Ar), 9.08 (1H, d, NH); δ 12.81, 12.86 (5-Me), 21.37
64.53 (C-4), 174.20, 174.29 (d, J 6.9, Ala-CO); MS m/e FAB
(t, J 5.0, Ala-Me), 50.48, 50.62 (Ala-CH), 53.13 (OMe), 67.00,
67.76 (C-5′), 85.00, 85.03 (C-4′), 90.01, 90.28 (C-1′), 111.68,
111.84 (C-5), 122.67, 122.74, 122.83, 122.89 (Ar), 127.88,
128.04 (C-2′), 133.35, 133.69 (C-3′), 135.96, 136.24 (C-6),
139.17, 139.20 (Ar), 150.56, 150.63, 150.70, 150.79 (ipso-Ar),
+
50.2110 (MH , C21
H
24
F
3
N
3
O
9
P requires 550.1202); HPLC t
39.36, 39.87 min (gradient II).
′,3′-Did eh yd r o-2′,3′-d id eoxyt h ym id in e-5′-(3,4-d i-
ch lor oph en yl m eth oxyalan in ylph osph ate) (2k): yield 35%;
4.45, 3.87; δ 1.35 (3H, m, Ala-Me), 1.82 (3H, d, 5-Me), 3.70
R
6.42, 36.85 min; t
R
2
δ
P
H
151.19, 151.22 (C-2), 164.09, 164.15 (C-4), 174.14, 174.24,
+
(
(
3H, s, OMe), 3.86 (1H, m, Ala-CH), 4.26 (2H, m, 2H-5′), 5.05
1H, m, H-4′), 5.92 (1H, m, H-3′), 6.32 (1H, m, H-2′), 7.05 (2H,
174.34, 174.43 (Ala-CO); MS m/e FAB 592.0330 (MH , C20
H
24
-
IN
3 8 R R
O P requires 592.0346); HPLC t 34.76, 35.25 min; t 36.89,
m, Ph, H-6), 7.17 (1H, m, H-1′), 7.34 (2H, m, Ph), 8.38 (1H, d,
37.49 min (gradient II).

Phosphoramidate d4T-MP Prodrugs
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 20 4127
2
′,3′-Dideh ydr o-2′,3′-dideoxyth ym idin e-5′-(3-ch lor oph e-
n yl m eth oxya la n in ylp h osp h a te) (2p ): yield 59%; δ 4.38,
.75; δ 1.43 (3H, m, Ala-Me), 1.99 (3H, d, 5-Me), 3.81 (3H, s,
OMe), 4.05 (2H, m, Ala-NH, Ala-CH), 4.41 (2H, m, H-5′), 5.10
1H, m, H-4′), 5.98 (1H, m, H-3′), 6.39 (1H, m, H-2′), 7.10 (1H,
m, H-1′), 7.21 (5H, m, H-6, Ar-H), 8.68 (1H, d, NH); δ 12.74,
2.79 (5-Me), 21.37 (t, J 5.1, Ala-Me), 50.54, 50.63 (Ala-CH),
Deter m in a tion of Com p ou n d P a r tition Coefficien ts.
The following method was used to determine the P values for
compounds 2j-2t. A sample of each compound (2 µmol) was
dissolved in 1-octanol (10 mL) by rapid magnetic stirring for
10 min. The UV absorption of the octanol solution was
determined at 265 nm. Aqueous buffer (pH 7.0) (Fisons) (5 mL)
was added to an aliquot of the octanol solution (5 mL), and
the two phases were mixed by rapid magnetic stirring for 5
min at 25 °C. The two layers were separated by standing for
a further 5 min. An aliquot of the octanol layer was removed
and the UV absorption determined at 265 nm. The P values
were calculated from the ratio of the absorptions of the octanol
aliquot and the original octanol solution. Each experiment was
repeated at least three times.
P
3
H
(
C
1
5
4
3.09 (OMe), 67.19, 67.78 (d, J 0, J 5.3, C-5′), 84.90, 85.01 (C-
′), 90.08, 90.33 (C-1′), 111.69, 111.83 (C-5), 121.12, 122.23 (d,
J 5.3, Ar), 125.93, 125.96, 125.98 (Ar), 127.94, 128.06 (C-2′),
30.92, 130.95 (Ar), 133.33, 133.60 (C-3′), 135.44, 136.91 (C-
), 151.07, 151.08 (C-2), 163.89, 163.94 (C-4), 174.15 (t, J 9.0,
1
6
+
Ala-CO); MS m/e FAB 500.0995 (MH , C20
H
24ClN
3
O
8
P requires
5
00.0990); HPLC t
gradient II).
′,3′-Dideh ydr o-2′,3′-dideoxyth ym idin e-5′-(3-br om oph e-
n yl m eth oxya la n in ylp h osp h a te) (2q): yield 61%; δ 4.35,
.75; δ 1.40 (3H, m, Ala-Me), 1.93 (3H, d, 5-Me), 3.81 (4H,
m, OMe, Ala-NH), 4.08 (1H, m, Ala-CH), 4.41 (2H, m, H-5′),
.10 (1H, m, H-4′), 5.98 (1H, m, H-3′), 6.38 (1H, m, H-2′), 7.08
R
32.72, 32.99 min; t
R
34.88, 35.21 min
(
Deter m in a tion of Com p ou n d Aqu eou s Solu bility. The
following method was used to assess the aqueous solubility of
compounds 2 and 2c. The extinction coefficient (ꢀ) of each
compound was calculated by dissolving a portion of each
compound (ca. 1 mg) in water (15 mL, pH 6.2) and measuring
the UVmax absorption of the resulting solution. The process was
repeated twice more and an average value for ꢀ calculated. A
portion of each compound (10-40 mg) was added to water (500
2
P
3
H
5
(
1H, m, H-1′), 7.30 (5H, m, H-6, Ar-H), 8.60 (1H, d, NH); δ
2.75, 12.79 (5-Me), 21.38 (t, J 5.1, Ala-Me), 50.53, 50.62 (Ala-
CH), 53.12 (OMe), 67.19, 67.79 (d, J 0, J 5.3, C-5′), 84.97, 85.05
C-4′), 90.09, 90.34 (C-1′), 111.69, 111.83 (C-5), 119.29, 119.42,
19.49 (Ar), 124.01 (Ar), 127.97, 128.07 (C-2′), 128.87 (Ar),
31.25 (Ar), 133.32, 133.58 (C-3′), 135.90, 136.10 (C-6), 151.08
C
1
o
µL) and the mixture stirred rapidly for 1 h at 25 C. Following
(
1
1
filtration through glass wool and appropriate dilution, the
UVmax absorption was measured and the extinction coefficient
used to calculate the concentration of the saturated solution.
The results shown are the mean of 3 independent experiments
(
(
C-2), 163.93 (C-4), 174.17 (Ala-CO); MS m/e FAB 546.0461
+
-1
-1
MH , C20
3.51 min; t
′,3′-Did eh yd r o-2′,3′-d id eoxyth ym id in e-5′-(3-iod op h e-
n yl m eth oxya la n in ylp h osp h a te) (2r ): yield 53%; δ 4.33,
.70; δ 1.43 (3H, m, Ala-Me), 1.90 (3H, d, 5-Me), 3.80 (4H,
m, OMe, Ala-NH), 4.03 (1H, m, Ala-CH), 4.40 (2H, m, H-5′),
.09 (1H, m, H-4′), 5.98 (1H, m, H-3′), 6.40 (1H, m, H-2′), 7.16
4H, m, Ar-H, H-6, H-1′), 7.60 (2H, m, Ar-H), 8.58 (1H, d,
NH); δ 12.41, 12.58 (5-Me), 21.02 (t, J 5.8, Ala-Me), 50.14,
0.23 (Ala-CH), 53.11 (OMe), 66.71, 67.34 (d, J 0, J 5.1, C-5′),
H
24BrN
3
O
8
P requires 546.0464); HPLC t
R
33.21,
[ꢀ (l‚mol ‚cm ): 2 ) 7640, 2c ) 6208; saturation concentra-
3
R
35.45, 35.83 min (gradient II).
tion (mmol/L): 2 ) 5.4, 2c ) 67.0].
2
An t ivir a l Act ivit y Det er m in a t ion s. Vir u ses: HIV-
1(strain III ) was kindly provided by Dr. R. C. Gallo (at that
B
P
3
H
time at the National Institutes of Health, Bethesda, MD).
Virus stocks were prepared from the supernatants of HIV-1-
infected MT-4 cells. HIV-2(strain ROD) was a gift from Dr. L.
Montagnier (Pasteur Institute, Paris, France), and virus stocks
were prepared from the supernatants of HIV-2-infected MT-4
cells.
5
(
C
5
8
1
1
1
4.56, 84.67 (C-4′), 89.65, 89.90 (C-1′), 111.32, 111.47 (C-5),
19.63, 119.78 (Ar), 125.88 (Ar), 127.59, 127.68 (C-2′), 129.23,
29.36 (Ar), 132.97, 133.22 (C-3′), 135.56, 135.76 (C-6), 150.58,
An tir etr ovir u s a ssa ys: CEM/0 and CEM/TK- cells were
suspended at 250, 000cells/mL of culture medium and infected
with HIV-1 or HIV-2 at ∼20 and ∼100 CCID50 (50% cell culture
infective dose)/mL, respectively. Then, 100 µL of the infected
cell suspensions were added to 200-µL microtiter plate wells
containing 100 µL of an appropriate dilution of the test
50.71 (C-2), 163.58, 163.53 (C-4), 173.70, 173.92 (Ala-CO); MS
+
m/e FAB 592.0342 (MH , C20
H
24IN
3
O
8
P requires 592.0346);
36.61, 36.99 min (gradient II).
′,3′-Did eh yd r o-2′,3′-d id eoxyt h ym id in e-5′-(4-(t r iflu o-
r om eth yl)p h en yl m eth oxya la n in ylp h osp h a te) (2s): yield
R R
HPLC t 34.31, 34.63 min; t
o
2
compound. After 4 days incubation at 37 C, cell cultures were
examined for giant cell formation.
5
3
4%; δ
P
4.33, 3.72; δ
H
1.40 (3H, m, Ala-Me), 1.90 (3H, d, 5-Me),
The EC50 was defined as the compound concentration (µM)
required to inhibit HIV-induced cytopathicity (giant cell
formation) in CEM cell cultures by 50%. The CC50 was defined
as the compound concentration (µM) required to inhibit CEM
cell proliferation by 50%.
.74 (3H, s, OMe), 3.89 (1H, m, Ala-NH), 4.03 (1H, m, Ala-
CH), 4.37 (2H, m, H-5′), 5.09 (1H, m, H-4′), 5.99 (1H, m, H-3′),
.40 (1H, m, H-2′), 7.09 (1H, m, H-1′), 7.36 (3H, m, H-6, Ar-
H), 7.68 (2H, m, Ar-H), 8.80 (1H, d, NH); δ 12.81, 12.86 (5-
Me), 21.44 (t, J 4.3, Ala-Me), 50.54, 50.63 (Ala-CH), 53.15
OMe), 67.15, 67.88 (d, J 4.3, J 5.8, C-5′), 84.86, 84.97 (C-4′),
6
C
QSAR. Statistical analysis of the relationships between
antiviral activity and substituent properties of interest was
performed by partial least-squares algorithm using the QSAR
(
9
5
3
0.05, 90.32 (C-1′), 111.69, 111.85 (C-5), 120.85, 120.99 (d, J
.0, Ph), 127.62, 127.95 (C-2′), 128.10 (Ar), 133.30, 133.63 (C-
′), 135.93, 136.16 (C-6), 151.14, 151.16 (C-2), 164.00, 164.05
32
module of SYBYL software. The training set used for the
analysis included all 21 compounds presented in Table 1. The
optimal number of components for each PLS model generated
was determined by cross-validation procedures using 6 random
groups. This protocol was repeated 10 times and an average
(C-4), 174.06, 174.16, 174.23 (Ala-CO); MS m/e FAB 534.1236
+
(MH , C21
H
24
F
3
N
3
O
8
P requires 534.1253); HPLC t
R
35.57,
3
5.92 min; t
R
37.89, 38.34 min (gradient II).
2
′,3′-Did eh yd r o-2′,3′-d id eoxyt h ym id in e-5′-(3-(t r iflu o-
2
value of Q calculated.
r om eth yl)p h en yl m eth oxya la n in ylp h osp h a te) (2t): yield
6
3
5
7%; δ
P
H
4.39, 3.83; δ 1.45 (3H, m, Ala-Me), 1.90 (3H, d, 5-Me),
Ack n ow led gm en t. We thank the U.K. Medical
Research Council, the Biomedical Research Programme
of the European Commission, and the Belgian Gecon-
certeerde Onderzoeksacties (Project 95/5) for financial
support. We also thank Mrs. A. Absillis for excellent
technical assistance and Ms. Helen Murphy for secre-
tarial support.
.78 (4H, s, OMe), 4.07 (1H, m, Ala-CH), 4.40 (2H, m, H-5′),
.07 (1H, m, H-4′), 5.98 (1H, m, H-3′), 6.39 (1H, m, H-2′), 7.08
(
1H, m, H-1′), 7.30 (1H, m, H-6), 7.51 (4H, m, Ar), 8.40 (1H, d,
NH); δ 12.74, 12.83 (5-Me), 21.47 (t, J 4.3, Ala-Me), 50.64,
0.61, 50.64 (Ala-CH), 53.12 (OMe), 67.17, 67.85 (d, J 5.1, J
C
5
5
1
.1, C-5′), 84.83, 84.88, 84.95 (C-4′), 90.06, 90.32 (C-1′), 111.69,
11.86 (C-5), 117.74 (d, J 5.0, Ph), 127.98, 128.09 (C-2′), 128.10
(Ar), 130.92 (Ar) 133.30, 133.58 (C-3′), 135.92, 136.11 (C-6),
1
1
50.87, 151.01, 151.05 (C-2), 163.87, 163.91 (C-4), 174.03,
+
Refer en ces
74.144 (Ala-CO); MS m/e FAB 534.1241 (MH , C21
24 3 3 8
H F N O P
requires 534.1253); HPLC t
R
35.77, 35.95 min; t
R
37.96, 38.41
(1) De Clercq, E. In search of a Selective Antiviral Chemotherapy.
min (gradient II).
Clin. Microbiol. Rev. 1997, 10, 674-693.

4
128 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 20
Siddiqui et al.
(
2) Mitsuya, H.; Weinhold, K. J .; Furman, P. A.; St Clair, M. H.;
Nusinoff- Lehrman, S.; Gallo, R. C.; Bolognesi, D.; Barry, D. W.;
Broder, S. 3′- Azido-3′-deoxythymidine (BWA509U): An agent
that inhibits the infectivity and cytopathic effect of human T-cell
lymphotropic virus type III/lymphadenopathy-associated virus
in vitro. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 7096-7100.
3) Mitsuya, H.; Broder, S. Inhibition of the in vitro infectivity and
cytopathic effect of human T-lymphotropic virus, type III/
lymphadenopathy-associated virus (HTLV III/LAV) by 2′,3′-
didoxynucleosides. Proc. Natl. Acad. Sci U.S.A. 1986, 83, 1911-
(16) J ones, R. J .; Bischofberger, W. Minireview - Nucleotide Pro-
drugs. Antiviral Res. 1995, 27, 1-17.
(17) Meier, C. Pro-Nucleotides Recent Advances in the Design of
Efficient Tools for the Delivery of Biologically Active Nucleosides
Monophosphates. Synlett 1998, 233-242.
(
18) McGuigan, C.; Cahard, D.; Sheeka, H. M.; De Clercq, E.;
Balzarini, J . Phosphoramidate derivatives of d4T with improved
anti-HIV efficacy retain full activity in thymidine kinase-
deficient cells. Bioorg. Med. Chem. Lett. 1996, 6, 1183-1186.
19) McGuigan, C.; Cahard, D.; Sheeka, H. M.; De Clercq, E.;
Balzarini, J . Aryl phosphoramidate derivatives of d4T have
improved anti-HIV efficacy in tissue culture and may act by the
generation of a novel intracellular metabolite. J . Med. Chem.
(
(
(
1
915.
4) Balzarini, J .; Kang, G.-J .; Dalal, M.; Herdewijn, P.; De Clercq,
E.; Broder, S.; J ohns, D. G. The anti-HTLV-III (anti-HIV) and
cytotoxic activity of 2′,3′- didehydro-2′,3′-dideoxyribonucleosides.
A comparison with their parental 2′,3′-dideoxyribonucleosides.
Mol. Pharmacol. 1987, 32, 162-167.
5) Herdewijn, P.; Balzarini, J .; De Clercq, E.; Pauwels, R.; Baba,
M.; Broder, S.; Vanderhaeghe, H. 3′-Substituted 2′,3′-dideoxy-
nucleoside analogues as potential anti-HIV (HTLV-III/LAV)
agents. J . Med. Chem. 1987, 30, 1270-1278.
6) Balzarini, J .; Baba, M.; Pauwels, R.; Herdewijn, P.; De Clercq,
E. Anti- retrovirus activity of 3′-fluoro- and 3′-azido-substituted
pyrimidine 2′,3′-dideoxynucleoside analogues. Biochem. Phar-
macol. 1988, 37, 2847-2856.
7) Baba, M.; Pauwels, R.; Herdewifn, P.; De Clercq, E.; Desmyter,
J .; Vandeputte, M. Both 2′,3′-dideoxythmidine and its 2′,3′-
unsaturated derivative (2′,3′-dideoxythymidinene) are potent
and selective inhibitors of human immunodeficiency virus
replication in vitro. Biochem. Biophys. Res. Commun. 1987, 142,
1
996, 39, 1748-1753.
(
20) Balzarini, J .; Kruining, J .; Wedgwood, O.; Pannecouque, C.;
Aquaro, S.; Perno, C. F.; Naesens, L.; Witvrouw, M.; Heijtink,
R.; De Clercq, E.; McGuigan, C. Conversion of 2′,3′-dideoxyad-
enosine (ddA) and 2′,3′- didehydro-2′,3′-dideoxyadenosine (d4A)
to their corresponding aryloxyphosphoramidate derivatives mark-
edly potentiates their activity against human immunodeficiency
virus and hepatitis B virus. FEBS Lett. 1997, 410, 324-328.
(
(
(
(21) Siddiqui, A. Q.; McGuigan, C.; Ballatore, C.; De Clercq, E.;
Balzarini, J . The Presence of Substituents on the Aryl Phos-
phoramidate Derivative of d4T Enhances Anti-HIV Efficacy in
Cell Culture: A Structure-Activity Relationship. J . Med. Chem.
1
999, 42, 393-399.
(
22) Horwitz, J . P.; Chua, J .; DaRooge, M. A.; Noel, M.; Klindt, I. L.
Nucleosides. IX. The formation of 2′,3′-unsaturated pyrimidine
nucleosides via a novel â-elimination rection. J . Org. Chem.
1
28-134.
(
8) Mansuri, M. M.; Hitchcock, M. J . M.; Buroker, R. A.; Bregman,
C. L.; Ghazzouli, I.; Desiderio, J . V.; Starrett J r., J . E.; Sterzycki,
R. Z.; Martin, J . C. Properties and Mouse Toxicities of 3
Thymidine Analogues Active Against Human-Immunodeficiency-
Virus. Antimicrob. Agents Chemother. 1990, 17, 637-641.
9) Furman, P. A.; Fyfe, J . A.; St. Clair, M. H.; Weinhold, K.;
Rideout, J . L.; Freeman, J . A.; Lehrman, S. N.; Bolognesi, D. P.;
Broder, S.; Mitsuya, H.; Barry, D. W. Phosphorylation of
1
996, 31, 205-211.
(
23) Mansuri, M. M.; Starrett J r., J . E.; Ghazzouli, I.; Hitchcock, M.
J . M.; Sterzycki, R. Z.; Brankovan, V.; Lin, T. S.; August, E. M.;
Prussoff, W. H.; Sommadossi, J . P.; Martin, J . C. 1-(2′,3′-dideoxy-
â-D-glycero-pent-2-enfuranosyl)thyimine. A highly Potent and
Selective Anti-HIV Agent. J . Med. Chem. 1989, 32, 462-446.
24) McGuigan, C.; Cahard, D.; Saigado, A.; De Clercq, E.; Balzarini,
J . Phosphoramidates as potent prodrugs of anti-HIV nucle-
otides: studies in the amino region. Antiviral Chem. Chemother.
1996, 7, 31-36.
(
(
3
′-Azido-3′-dideoxythymidine and Selective Interaction of the 5′-
Triphosphate with Human Immunodeficiency Virus Reverse
Transcriptase. Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 8333-
8
337.
(25) McGuigan, C.; Pathirana, R. N.; Balzarini, J .; De Clercq, E.
Intracellular delivery of bioactive AZT nucleotides by aryl
phosphate derivatives of AZT. J . Med. Chem. 1993, 36, 1048-
1052.
(26) Fujita, T.; Miyoshi, H. Quantitative analyses of the uncoupling
activity of substituted phenols with mitochondria from muscles
of house flies. Biochim. Biophys. Acta 1988, 935, 312-321.
(
10) Balzarini, J .; Herdewijn, P.; De Clercq, E. Differential patterns
of intracellular metabolism of 2′,3′-didehydro-2′,3′-dideoxythy-
midine and 3′-azido-2′,3′- dideoxythymidine, two potent anti-
human immunodeficiency virus compounds. J . Biol. Chem. 1989,
2
64, 6127-6133.
(
11) Balzarini, J .; Cooney, D. A.; Dalal, M.; Kang, G.-J .; Cupp, J . E.;
De Clercq, E.; Broder, S.; J ohns, D. G. 2′,3′-Dideoxycytidine:
regulation of its metabolism and anti-retroviral potency by
natural pyrimidine nucleosides and by inhibitors of pyrimidine
nucleotide synthesis. Mol. Pharmacol. 1987, 32, 798-806.
12) Starnes, M. G.; Cheng, Y.-C. Cellular metabolism of 2′,3′-
Dideoxycytidine, a compound active against human immunode-
ficiency virus in vitro. J . Biol. Chem. 1987, 32, 798-806.
13) Hao, Z.; Cooney, D. A.; Farquhar, D.; Perno, C. F.; Zhang, K.;
Masood, R.; Wilson, Y.; Hartman, N. R.; Balzarini, J .; J ohns, D.
G. Potent chain termination activity and selective inhibition of
human immunodeficiency virus reverse transcriptase by 2′,3′-
dideoxyuridine-5′-triphosphate. Mol. Pharmacol. 1990, 37, 157-
(
27) Fujita, T.; Nishioka, T.; Miyoshi, H. Quantitative Relationship
Between Protonophoric and Uncoupling Activities of Substituted
Phenols. Biochim. Biophys. Acta 1987, 891, 194-204.
(28) Valette, G.; Pompon, A.; Girardet, J. L.; Cappellacci, L.; Franchet-
(
(
ti, P.; Grifantini, M.; LaColla, P.; Loi, A. G.; Perigaud, C.;
Gosselin, G.; Imbach, J . L. Decomposition pathways and in vitro
HIV inhibitory effects of isoddA pronucleotides - toward a
rational approach for intracellular delivery of nucleoside 5′-
monophosphates. J . Med. Chem. 1996, 39, 1981-1990.
(29) Topliss, J . G. Utilization of Operational Schemes for Analogue
Synthesis in Drug Design. J . Med. Chem. 1972, 15, 1006-1011.
1
63.
(30) Hansch, C.; Leo, A. Substitution Constants for Correlation
Analysis in Chemistry and Biology; Wiley-Interscience: New
York, 1979.
(
14) J ohnson, M. A.; Ahluwalia, G.; Connelly, M. C.; Cooney, D. A.;
Broder, S.; J ohns, D. G.; Fridland, A. Metabolic pathways for
the activation of the antiretroviral agent 2′,3′-dideoxyadenosine
in human lymphoid cells. J . Biol. Chem. 1988, 263, 15354-
(31) 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.
(32) SYBYL, version 6.5; Tripos Associates: St. Louis, MO, 1998.
1
5357.
(
15) J ohnson, M. A.; Fridland, A. Phosphorylation of 2′,3′-dideoxyi-
nosine by cytosolic 5′-nucleotidase of human lymphoid cells. Mol.
Pharmacol. 1989, 36, 291-295.
J M9807104