New 3'-azido-3'-deoxythymidin-5'-yl O-(omega-hydroxyalkyl) carbonate prodrugs: synthesis and anti-HIV evaluation.

Article abstract of DOI:10.1021/jm001033s

Prodrugs of zidovudine (AZT) have been synthesized in an effort to enhance its uptake by HIV-1 infected cells and its anti-HIV activity. The 5'-OH function of AZT was functionalized with various enzymatically labile alkyl groups using specific procedures. The prodrug moieties included 5'-O-carbonate, 5'-O-carbamate, and 5'-O-ester. Analogues of the 3'-azido-3'-deoxythymidin-5'-yl O-(omega-hydroxyalkyl) carbonate series were particularly interesting since they were rearranged through an intramolecular cyclic process during their enzymatic hydrolysis. Evidence of this prodrug rearrangement was confirmed by comparison of the serum half-lives of 5'-O-carbonate prodrugs with their corresponding 5'-O-ester- and 5'-O-carbamate-AZT prodrugs. Interestingly, the anti-HIV-1 activities (EC(50)) of 3'-azido-3'-deoxythymidin-5'-yl O-(4-hydroxybutyl) carbonate 10 in acutely infected MT-4 cells and in peripheral blood mononuclear cells (PBMCs) were 0.5 nM and 0.78 nM, respectively. Compound 10 was 30 to 50 times more potent than its parent drug AZT. Our results suggest that the specific intramolecular rearrangement associated with the 3'-azido-3'-deoxythymidin-5'-yl O-(omega-hydroxyalkyl) carbonate prodrugs could explain the remarkable anti-HIV-1 activity of this series of AZT prodrugs. Prodrug 10 may therefore have better clinical potential than AZT for the treatment of AIDS.

Full text of DOI:10.1021/jm001033s

J . Med. Chem. 2001, 44, 777-786
777
New 3′-Azid o-3′-d eoxyth ym id in -5′-yl O-(ω-Hyd r oxya lk yl) Ca r bon a te P r od r u gs:
Syn th esis a n d An ti-HIV Eva lu a tion
Patrick Vlieghe,†,‡,§ Fre´de´ric Bihel,†,§ Thierry Clerc,‡ Christophe Pannecouque,| Myriam Witvrouw,|
Erik De Clercq,| J ean-Pierre Salles,‡ J ean-Claude Chermann,§ and J ean-Louis Kraus*,†,§
Laboratoire de Chimie Biomole´culaire, Faculte´ des Sciences de Luminy, 163 avenue de Luminy, case 901,
13288 Marseille Cedex 9, France, Laboratoires LAPHAL, Avenue de Provence, BP 7, 13718 Allauch Cedex, France,
INSERM U322 “Re´trovirus et Maladies Associe´es”, Campus Universitaire de Luminy, BP 33, 13273 Marseille Cedex 9, France,
and Rega Institute for Medical Research, Katholieke Universiteit Leuven, B-3000 Leuven, Belgium
Received J uly 19, 2000
Prodrugs of zidovudine (AZT) have been synthesized in an effort to enhance its uptake by HIV-1
infected cells and its anti-HIV activity. The 5′-OH function of AZT was functionalized with
various enzymatically labile alkyl groups using specific procedures. The prodrug moieties
included 5′-O-carbonate, 5′-O-carbamate, and 5′-O-ester. Analogues of the 3′-azido-3′-deoxy-
thymidin-5′-yl O-(ω-hydroxyalkyl) carbonate series were particularly interesting since they were
rearranged through an intramolecular cyclic process during their enzymatic hydrolysis.
Evidence of this prodrug rearrangement was confirmed by comparison of the serum half-lives
of 5′-O-carbonate prodrugs with their corresponding 5′-O-ester- and 5′-O-carbamate-AZT
prodrugs. Interestingly, the anti-HIV-1 activities (EC50) of 3′-azido-3′-deoxythymidin-5′-yl O-(4-
hydroxybutyl) carbonate 10 in acutely infected MT-4 cells and in peripheral blood mononuclear
cells (PBMCs) were 0.5 nM and 0.78 nM, respectively. Compound 10 was 30 to 50 times more
potent than its parent drug AZT. Our results suggest that the specific intramolecular
rearrangement associated with the 3′-azido-3′-deoxythymidin-5′-yl O-(ω-hydroxyalkyl) carbonate
prodrugs could explain the remarkable anti-HIV-1 activity of this series of AZT prodrugs.
Prodrug 10 may therefore have better clinical potential than AZT for the treatment of AIDS.
In tr od u ction
Intensive efforts are underway worldwide to develop
chemotherapeutic agents effective against the human
immunodeficiency virus (HIV),1,2,3 the etiological agent
of acquired immunodeficiency syndrome (AIDS).4 Among
the current diversity of compounds active against HIV,
the 2′,3′-dideoxynucleosides (ddNs) remain by far the
most potent.5,6 In all cases, the expression of activity
F igu r e 1. Structure of AZT, the first FDA-approved drug for
HIV treatment.
requires conversion into its corresponding 5′-O-tri-
phosphate analogues.7,8,9 These may inhibit the replica-
tion of the virus by competitive inhibition of the viral
nevirapine (BI-RG-587, Viramune),20 delavirdine (PNU-
reverse transcriptase (RT) and/or by incorporation and
90152T, Rescriptor),21 and efavirenz (DMP-266, Sus-
subsequent chain termination of the growing viral DNA
tiva).22,23 Protease inhibitors (PIs) are another family
strand.10 AZT (zidovudine, Figure 1) was the first ddN
of compounds that inhibit the HIV protease. This family
that was approved by the Food and Drug Administra-
includes indinavir (MK-639, Crixivan), ritonavir (ABT-
tion (FDA)11 for the treatment of patients suffering from
538, Norvir), saquinavir (Ro-31-8959, Invirase and
AIDS.12,13 Currently, six drugs belonging to the ddNs
Fortovase), nelfinavir (AG-1343, Viracept), and am-
family are approved by the FDA and are commercially
prenavir (VX-478, Agenerase).6,24,25 Although being the
available: zidovudine (AZT, Retrovir),11 stavudine (d4T,
oldest compound approved by the FDA, AZT is still one
Zerit),14 zalcitabine (ddC, Hivid),15 didanosine (ddI,
of the most potent agents active against HIV. For many
Videx),16 lamivudine (3TC, Epivir),17 and abacavir
years now, AZT in combination with other nucleoside
(1592U89, Ziagen).18 Other families of compounds are
derivatives (such as lamivudine) and/or some NNRTIs
also available and approved by the FDA, including the
(such as Viramune) and/or different PIs (such as indi-
nucleoside analogues such as adefovir dipivoxil (bis-
navir and nelfinavir) has resulted in a decrease in the
(POM)PMEA, Preveon)19 and also the nonnucleoside
mortality rate and frequency of opportunistic infections
reverse transcriptase inhibitors (NNRTIs), which inter-
in AIDS patients.26 However, a significant dose-related
act at a specific site in HIV-1. This family includes
toxicity associated with the administration of AZT,
resulting in anemia and leucopenia, remains a limiting
* Corresponding author. Tel: (33) 491 82 91 41. Fax: (33) 491 82
factor for its use.27,28 Moreover, the necessity of admin-
94 16. E-mail: kraus@luminy.univ-mrs.fr.
† Laboratoire de Chimie Biomole´culaire.
istrating high doses of AZT to achieve adequate con-
‡ Laboratoires LAPHAL.
centrations in the cerebrospinal fluid (CSF) results in
§ INSERM U322 “Re´trovirus et Maladies Associe´es”.
| Katholieke Universiteit Leuven.
bone marrow toxicity.27,28 In attempts to overcome the
10.1021/jm001033s CCC: $20.00 © 2001 American Chemical Society
Published on Web 02/01/2001

778 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5
Vlieghe et al.
F igu r e 3. Structure of the three series of 5′-O-AZT prodrugs.
F igu r e 2. Cyclic intramolecular rearrangement process on
3′-azido-3′-deoxythymidin-5′-yl O-(ω-hydroxyalkyl) carbonate
which recovers the initial prodrug.
the formation of a mixed anhydride using 3,4,5-tri-
methoxybenzoyl chloride in the presence of Et3N in
EtOAc in 86% yield. The resulting anhydride 2 was
converted into its ω-tert-butyl ester 3 by addition of
tBuOH in the presence of DMAP in CH2Cl2 in 77%
yield.45 The preparation of compound 4 was performed
by esterification of AZT with 4-(tert-butoxycarbonyl)-
amino-1-butanoic acid in the presence of DCC and
DMAP in DMF in 90% yield.37 N-Boc deprotection of
compound 4 by TFA in CH2Cl2 led to prodrug 5 in
quantitative yield. As expected, chemical synthesis of
the 4-hydroxy-1-(3′-azido-3′-deoxythymidin-5′-yl) bu-
tanoate was impossible because of its in situ lactonisa-
tion to γ-butyrolactone.
problem of rapid elimination of AZT from the brain and
to increase its therapeutic efficacy, numerous AZT
prodrugs have been reported in the literature.29-33 In
most of the cases, the mechanism of action of these AZT
prodrugs is based on hydrolysis of the enzymatically
labile 5′-O-bonds between the drug (AZT) and its spacer
group. The expected advantages of the AZT prodrugs
can be multiple: synergistic drug interactions,34-36
enhancement of AZT intracellular uptake,37 increase of
AZT brain delivery,38,39 and bypass of the first AZT
phosphorylation step into the cells.40-43
Ca r ba m a te Ser ies (Y ) -NH-C(O)-). 5′-O-Car-
bamate-AZT prodrugs were obtained by condensation
of AZT with N,N-carbonyldimidazole (CDI) in the pres-
ence of various primary amines.46 Carbamates 6, 8, and
9 were obtained by the condensation of AZT with CDI
in DMF or CH3CN in the presence of 4-amino-1-tert-
butyl butanoate, 1,3-diaminopropane, and 3-amino-1-
propanol in 65%, 78%, and 60% yields, respectively.
Hydrolysis of the tert-butyl ester group in compound 6
by TFA in CH2Cl2 led to prodrug 7 in 94% yield.
Ca r bon a te Ser ies (Y ) -O-C(O)-). 5′-O-Carbon-
ate-AZT prodrugs were obtained by condensation of AZT
with CDI in the presence of various ω-alcohols.46 In this
manner, numerous 5′-O-carbonate-AZT prodrugs were
obtained in yields ranging from 49% to 85%. Condensa-
tion of AZT with CDI in DMF in the presence of 1,4-
butanediol gave compound 10 in 85% yield. Its oxidation
using TEMPO‚HCl and m-CPBA in CH2Cl2 led to
compound 11 in 69% yield.47 Esterification of acid 11
was achieved via mixed anhydride 12 which was formed
by reaction of compound 11 with 3,4,5-trimethoxyben-
zoyl chloride in the presence of Et3N in EtOAc in 88%
yield. The resulting anhydride 12 was converted into
the ω-tert-butyl ester 13 by treatment with tBuOH in
the presence of DMAP in CH2Cl2 in 75% yield.45
Condensation of AZT with CDI in CH3CN in the
presence of 3-(tert-butoxycarbonyl)amino-1-propanol led
to the N-protected intermediate 14 in 68% yield.46 Its
N-Boc deprotection by TFA in CH2Cl2 gave prodrug 15
in quantitative yield. The condensation of AZT with CDI
in DMF in the presence of 1,3-propanediol led to
compound 16 in 66% yield.46 Carbonates 17, 18, 19, and
20 were obtained by condensation of AZT with CDI in
In the present paper, we report the specific behavior
of new 3′-azido-3′-deoxythymidin-5′-yl O-(ω-hydroxy-
alkyl) carbonate prodrugs, which are rearranged during
the enzymatic hydrolysis of their 5′-O-carbonate bond,
through an intramolecular cyclic process (Figure 2).
The evidence for this intramolecular cyclic rearrange-
ment, specifically associated with 3′-azido-3′-deoxythy-
midin-5′-yl O-(ω-hydroxyalkyl) carbonate prodrugs, is
confirmed by comparison of their sensitivity to enzy-
matic hydrolysis and biological properties with various
5′-O-ester-, 5′-O-carbamate, and other 5′-O-carbonate-
AZT prodrugs.
With this aim, we have designed and synthesized
several series of AZT prodrugs with the general formula
shown in Figure 3. We then evaluated their anti-HIV-1
activities in cell cultures.
To study the influence of the terminal functional
group X and of the linker moiety Y on the sensitivity of
various prodrugs to enzymatic hydrolysis, we arbitrarily
fixed the parameter n to 3 (Figure 3). We have then
studied the influence of the alkyl chain length (n ) 2,
3, 4, 5, 6, and 8) on the stability against enzymatic
hydrolysis and anti-HIV activity of 3′-azido-3′-deoxythy-
midin-5′-yl O-(ω-hydroxyalkyl) carbonate prodrugs.
Ch em istr y
In this paper, we describe the synthesis of three series
of AZT prodrugs (Figure 3 and Scheme 1).
Ester Ser ies (Y ) -C(O)-). Compound 1 was
obtained directly by condensation of AZT with glutaric
anhydride in the presence of DMAP in DMF in 86%
yield.44 Esterification of acid 1 was achieved through

Synthesis of AZT Prodrugs
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5 779
Sch em e 1a
the alkyl chain length of the spacer bearing a terminal
hydroxyl group was essential for the intramolecular
cyclic rearrangement of the prodrug, we have synthe-
sized various 3′-azido-3′-deoxythymidin-5′-yl O-(ω-hy-
droxyalkyl) carbonate prodrugs (17, 16, 10, 18, 19, and
20). To validate the proposed intramolecular cyclic
rearrangement of these prodrugs, we have also per-
formed the synthesis and studied the biological prop-
erties of various 5′-O-ester- (1, 3, 5), 5′-O-carbamate-
(6-9), and other 5′-O-carbonate-AZT prodrugs (11, 13,
15, 21). In our hypothesis, the rate of the intramolecular
cyclic rearrangement might be under the control of the
methylene chain length of the spacer. As shown in
Figure 2, two different cleavage mechanisms could be
considered during the enzymatic hydrolysis of a prodrug
by human serum hydrolases (carboxyesterases).
(i) In compounds 17 (n ) 2) and 16 (n ) 3): hydrolytic
cleavage according to route a was likely to occur since
the release of AZT would involve the formation of a five-
or six-membered ring, respectively, which was thermo-
dynamically favored. In contrast, hydrolytic cleavage
according to route b, leading to the cyclic rearrangement
of the prodrugs 17 and 16 by an intramolecular nucleo-
philic attack of the terminal hydroxyl function, was not
favored.
(ii) In compound 10 (n ) 4): hydrolytic cleavage
according to route a, leading to the release of AZT, was
thermodynamically unfavorable since this process would
involve formation of a seven-membered ring. In contrast,
hydrolytic cleavage according to route b, leading to the
cyclic rearrangement of the prodrug 10 by an intramo-
lecular nucleophilic attack of the terminal hydroxyl
function, was favored. Under these conditions, the half-
life (t1/2) of enzymatic hydrolysis of this prodrug 10
should be enhanced.
(iii) In compounds 18 (n ) 5), 19 (n ) 6), and 20 (n )
8): hydrolytic cleavage by route a, leading to the release
of AZT via the formation of, respectively, an eight-,
nine-, or eleven-membered ring was thermodynamically
unfavorable. Hydrolytic cleavage by route b, leading to
the cyclic rearrangement of prodrugs 18, 19, or 20,
respectively, by an intramolecular nucleophilic attack
of the terminal hydroxyl function, was also not favored
because of the growth of their alkyl chain length (which
will become too long).
a
Reagents: (a) glutaric anhydride, DMAP, DMF; (b) 3,4,5-
trimethoxybenzoyl chloride, Et3N, AcOEt; (c) tert-butyl alcohol,
DMAP, CH2Cl2; (d) BocNH-(CH2)3-CO2H, DCC, DMAP, DMF; (e)
TFA, CH2Cl2; (f) CDI, DMF then H2N-(CH2)3-CO2tBu, DMF; (g)
CDI, DMF then H2N-(CH2)3-NH2, DMF; (h) CDI, CH3CN then
H2N-(CH2)3-OH, CH3CN; (i) CDI, DMF then HO-(CH2)n-OH, DMF;
(j) m-CPBA, Tempo‚HCl, CH2Cl2; (k) CDI, CH3CN then BocNH-
(CH2)3-OH, CH3CN; (l) CDI, CH3CN then H3C-(CH2)3-OH, CH3CN.
(iv) Compound 21 cannot be cleaved by the intramo-
lecular rearrangement mechanism. This compound did
not bear a terminal hydroxyl group, which could help
the release of AZT or the cyclic rearrangement of the
prodrug by an intramolecular nucleophilic attack. This
prodrug 21 was used as a reference compound for
prodrug 10 in order to validate the concept.
DMF in the presence of 1,2-ethanediol, 1,5-pentanediol,
1,6-hexanediol, and 1,8-octanediol in 56%, 70%, 49%,
and 63% yields, respectively.46 Finally, the butyl car-
bonate 21 was obtained by condensation of AZT with
CDI in CH3CN in the presence of 1-butanol in 64%
yield.46
Ser u m Hyd r ola se Sta bility Stu d ies a n d Lip o-
p h ilicity (Ta bles 1 a n d 2). Taking into account this
hypothesis on chemical reactivity of the carbonate-AZT
prodrugs, we determined the enzymatic hydrolysis half-
lives (t1/2) in human serum of every AZT prodrug
synthesized, by an HPLC method. The data reported
in Tables 1 and 2 suggest the following conclusions:
(i) Generally the 5′-O-carbonate-AZT prodrugs (11, 13,
15, and 16) were most sensitive to enzymatic hydrolysis
(t1/2 values ranging from 8 to 780 min), while the 5′-O-
ester-AZT prodrugs (1, 3, 5) were less sensitive to the
serum hydrolase action, with t1/2 values ranging from
Resu lts a n d Discu ssion
To be clinically useful, a prodrug should fulfill phar-
macokinetic and pharmacodynamic profiles which are
mainly dependent on the kinetics of their conversion to
the parent drug and on its membrane permeation
properties.48 These two properties are related to the
sensitivity of the linkage between the prodrug moiety
and the parent drug to enzymatic hydrolysis and, on
the other hand, to their lipophilicity.
Taking into account that during the course of enzy-
matic hydrolysis of the AZT prodrugs, the influence of

780 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5
Vlieghe et al.
Ta ble 1. Biological Properties of Numerous 5′-O-Ester-,
5′-O-Carbamate-, and 5′-O-Carbonate-AZT Prodrugs
structure
a
c
d
t1/2
EC50 CC50
compd
X
Y
(min) log Pb (nM) (µM)
SIe
1
3
5
/
7
6
8
9
11
13
15
16
AZT
HO2C- -C(O)-
tBuO2C- -C(O)-
1800 -0.52
33 1.17
26 -0.90
-0.95
7
5
20
>50 >7143
>50 >10000
>50 >2500
H2N-
HO-
-C(O)-
-C(O)-
F igu r e 4. Nucleophilic intramolecular attack of the ω-amino
group of compound 15 on its 5′-O-carbonate function leading
to the generation of the 3′-azido-3′-deoxythymidin-5′-yl N-(3-
hydroxypropyl) carbonate 9.
HO2C- -NH-C(O)- >2880 -0.81 100
tBuO2C- -NH-C(O)- >2880 0.88
>50
>500
10
10
>50 >5000
>50 >5000
H2N-
HO-
HO2C- -O-C(O)-
-NH-C(O)- >2880 -1.19
-NH-C(O)- >2880 -1.24 100
>50
>500
780 -0.55
16 1.14
8 -0.93
10 -0.98
-0.88
5
5
20
10
>50 >10000
>50 >10000
>50 >2500
>50 >5000
tBuO2C- -O-C(O)-
their ester counterparts, which were themselves more
stable than their carbonate analogues. This stability of
the carbamate function, in human serum, was in ac-
cordance with literature data.49 The influence of the 5′-
O-terminal groups on the sensitivity to enzymatic
hydrolysis in a prodrug (ester or carbonate) series was
in the following stability order: carboxylic acid . tert-
butyl ester > amine ≈ alcohol (Table 1). The terminal
carboxylic acid group increased the prodrug enzymatic
stability compared to compounds with terminal amino
or hydroxyl groups. This result was not surprising since
it was known that the presence of a carboxylate group
in the structure of a compound inhibits the serum
hydrolase activity.50 Moreover, esterification of the
terminal carboxylic acid by a tert-butyl group decreased
the half-life (t1/2) of the resulting tert-butyl ester, because
of the impossible liberation of carboxylate ions. The
presence of an amino or hydroxyl group at the ω-position
of the 5′-O-spacer enabled the intramolecular nucleo-
philic attack of this terminal group on the 5′-O-ester
(e.g., lactonization of 4-hydroxy-1-(3′-azido-3′-deoxythy-
midin-5′-yl) butanoate to γ-butyrolactone) or 5′-O-
carbonate functions.
(ii) When the 3′-azido-3′-deoxythymidin-5′-yl O-(3-
aminopropyl) carbonate 15 was submitted to serum
hydrolase, we clearly identified AZT (48%) and the 3′-
azido-3′-deoxythymidin-5′-yl N-(3-hydroxypropyl) car-
bamate 9 (52%) on HPLC profile as hydrolysis products.
Compound 9 resulted from the nucleophilic intramo-
lecular attack of the ω-amino group on the 5′-O-
carbonate function within compound 15 (Figure 4). This
result confirmed the two possible hydrolytic cleavages
(routes a or b) in prodrug 15, leading to the release of
AZT or to the generation of the carbamate function
(prodrug 9), respectively. This hypothesis was already
suggested in the case of 3′-azido-3′-deoxythymidin-5′-
yl O-(ω-hydroxyalkyl) carbonate prodrugs (Figure 2).
(iii) We could also observe (Table 2) that compound
10 was the most stable within the 5′-O-carbonate series
(17, 16, 10, 18, 19, 20) in human serum (t1/2 ) 30 min).
In comparison, the reference compound 21 possessed a
shorter half-life (t1/2 ) 9 min). These results support our
hypothesis of an intramolecular cyclic rearrangement
(associated with the intramolecular nucleophilic attack
of the ω-hydroxyl group on 5′-O-carbonate function)
specific to compound 10, which led to the enhancement
H2N-
HO-
-O-C(O)-
-O-C(O)-
H
25 >100 >4000
a
t1/2 (half-life) is the time required for 50% hydrolysis of
prodrugs to AZT at 37 °C upon incubation in human serum
(Normal Human Serum). log P determinations were performed
b
using ACD (Advanced Chemistry Development, Inc.)/log P 1.0 base
calculations. c EC50: concentration in nM required to inhibit the
d
cytopathicity of HIV-1 by 50% in MT-4 cells. CC50: concentration
in µM required to cause 50% death of uninfected MT-4 cells. e SI:
selectivity index ) CC50/EC50
.
Ta ble 2. Biological Properties of Various 5′-O-Carbonate-AZT
Prodrugs, a Study of Their Alkyl Chain Length
structure
a
c
d
t1/2
EC50 CC50
compd
X
n
(min) log Pb (nM) (µM)
SIe
11
13
15
16
17
10
21
18
19
20
AZT
HO2C-
tBuO2C-
H2N-
HO-
HO-
HO-
H-
HO-
HO-
HO-
3
3
3
3
2
4
4
5
6
8
780
16
8
10
<1
30
9
16
7
2
-0.55
5
5
>50
>50
>50
>50
>50
>10000
>10000
>2500
>5000
>5000
1.14
-0.93 20
-0.98 10
-1.20 10
-0.68
1.02
-0.45 10
0.08 10
1.14
-0.88 25
0.5
8
>50 >100000
>50
>50
>6250
>5000
>5000
>20000
>4000
>50
2.5
>50
>100
a
t1/2 (half-life) is the time required for 50% hydrolysis of
prodrugs to AZT at 37 °C upon incubation in human serum
(Normal Human Serum). log P determinations were performed
b
using ACD (Advanced Chemistry Development, Inc.)/log P 1.0 base
calculations. c EC50: concentration in nM required to inhibit the
d
cytopathicity of HIV-1 by 50% in MT-4 cells. CC50: concentration
in µM required to cause 50% death of uninfected MT-4 cells. e SI:
selectivity index ) CC50/EC50
.
26 to 1800 min (Table 1). In contrast, the 5′-O-carbam-
ate-AZT prodrugs (7, 6, 8, 9) appeared to be very stable
since they did not undergo enzymatic hydrolytic cleav-
age for up to 48 h. These results suggested that
sensitivity to enzymatic hydrolysis of the three series
of AZT prodrugs depends on the nature of their 5′-O-
functions. Carbamate prodrugs were more stable than

Synthesis of AZT Prodrugs
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5 781
of its stability (increased t1/2) toward serum hydrolase.
In summary, this cyclic rearrangement process of the
prodrug according to route b was favored in the case of
a four-carbon chain length.
(iv) In contrast, 3′-azido-3′-deoxythymidin-5′-yl O-(ω-
hydroxyalkyl) carbonate prodrugs (17, 16, 18, 19, and
20) whose intramolecular cyclic rearrangement process
according to route b was unfavorable had shorter t1/2
values (<1, 10, 16, 7, and 2 min, respectively) than
compound 10 (30 min).
In conclusion, these serum hydrolase stability studies
were in accordance with our thermodynamic predictions
concerning the specific enzymatic hydrolysis process of
3′-azido-3′-deoxythymidin-5′-yl O-(ω-hydroxyalkyl) car-
bonate prodrugs.
As far as lipophilicity is concerned, it is known that
AZT crosses cell membranes by nonfacilitated diffusion
and that its uptake is insensitive to inhibitors of
nucleoside transport.48 Moreover, it has been reported
that some AZT prodrugs containing a 5′-hydroxyl group
esterified with various amino acid ligands exhibited
anti-HIV activities in peripheral blood lymphocytes
(PBL) greater than that of AZT.37 This indicates that
the lipophilicity of AZT analogues, which is reflected by
their partition coefficient (log P), might have a signifi-
cant role in their diffusion.51 Therefore the log P values
between n-octanol and water were determined for all
compounds, using ACD/log P software from ChemCAD.
The results in Tables 1 and 2 show log P values are in
the range of -1.24 to +1.17 for different series of
prodrugs (the log P value for AZT itself was -0.88).
Moreover, from serum hydrolase stability studies, it is
evident that the lipophilicity of the AZT prodrugs does
not correlate with their sensitivity to enzymatic hy-
drolysis.
An tivir a l Activity Mea su r em en ts. The three series
of AZT prodrugs (compounds 1, 3, 5-11, 13, 15-21)
were evaluated for their inhibitory effects on HIV
replication, monitored by the efficiency of the different
compounds to inhibit the cytopathicity of HIV after
acute infection of MT-4 cell cultures. The 50% effective
concentration (EC50) represented the concentration
required to inhibit by 50% the cytopathicity of HIV for
MT-4 cells. Under these assay conditions, AZT inhibited
HIV replication at EC50 ) 25 nM. Under the same
conditions, the anti-HIV activities of most of the AZT
prodrugs tested were found to be similar or more
pronounced than that of their parent drug (AZT). All of
the AZT prodrugs tested elicited anti-HIV activities with
EC50 values ranging from 0.5 to 100 nM (Tables 1 and
2). The results obtained suggested the following com-
ments:
notable variations in antiviral activity were observed.
It is evident (Tables 1 and 2) that no prodrugs in the
ester or carbamate series were more active than their
corresponding analogues of the carbonate series (except
for compound 8 and its analogue 15, because compound
15 is partly converted into carbamate 9).
(iii) Under assay conditions, among the compounds
bearing a terminal hydroxyl function, the most potent
5′-O-carbonate-AZT prodrug was compound 10 (n ) 4)
with an EC50 ) 0.5 nM. It was more active than its
structural analogues 17, 16, 18, 19, or 20 (with n ) 2,
3, 5, 6, and 8, respectively). These antiviral activities
support the suggested intramolecular cyclic rearrange-
ment specific to this new class of 3′-azido-3′-deoxythy-
midin-5′-yl O-(ω-hydroxyalkyl) carbonate prodrugs. This
result appeared to be in complete agreement with the
above-mentioned thermodynamic predictions (Figure 2)
and with the serum hydrolase studies (Tables 1 and 2).
The increased antiviral activity of compound 10 might
be related to its favorable serum half-life (t1/2 ) 30 min)
associated with the intramolecular rearrangement pro-
cess which enhanced both its hydrolytic stability and
its intracellular uptake under in vitro conditions. More-
over, we cannot rule out that the prodrugs that are more
hydrophobic than its parent drug will have a limited
efflux once inside the target cells. Thus, their increased
half-life in the cell pool will lead to a prolonged release
of AZT from the continuing cleavage of the prodrug.
(iv) In order to confirm the increased antiviral potency
of compound 10 vs AZT, anti-HIV evaluation was also
performed in PBMCs, infected with different virus
strains (HIV-1 IIIB and BaL). PBMCs are an interesting
research tool, being the primary human targets of
HIV. Compound 10 was approximately 10 times more
active than AZT (Table 3). These data confirm the
activity of compound 10 in MT-4 cells infected by HIV-
1, which was found to be 50 times higher compared to
AZT (Table 2).
These results show that a subtle balance between the
rate of intracellular uptake and the sensitivity of
enzymatic hydrolysis of a prodrug under in vitro condi-
tions was required for optimum antiviral activity.
Moreover, the antiviral results for 3′-azido-3′-deoxythy-
midin-5′-yl O-(ω-hydroxyalkyl) carbonate prodrugs were
in complete agreement with the above-mentioned ther-
modynamic predictions for the intramolecular cyclic
rearrangement process and the serum hydrolase stud-
ies. Another very interesting point is the finding that
the antiviral activities of carbamates are similar to AZT.
These compounds seem to be quite stable in the cells,
so that the formation of significant levels of free intra-
cellular nucleoside is unlikely. It is possible that these
carbamates act through another mechanism of action
(perhaps, direct inhibition of HIV-reverse transcriptase).
This finding warrants further investigation.
(i) The prodrugs appeared to be sufficiently lipophilic
to cross the cellular membrane by passive diffusion like
AZT (without excluding a facilitated transport for these
prodrug series) as demonstrated by the antiviral activi-
ties. In fact, from data reported in Tables 1 and 2, it
can be observed that no correlation between partition
coefficient (log P) and anti-HIV activity (EC50) can be
drawn.
(ii) The overall results were quite intriguing since,
depending on both parameters, i.e., (a) the alkyl chain
length of the spacer and (b) the terminal functional
group X (-COOH, -CO2tBu, -NH2, -OH, -CH3),
Cytotoxicity. The cytotoxicity of AZT and its pro-
drugs against MT-4 cells was determined using the MTT
method and in PBMCs by the trypan blue exclusion
method for viability determination. The 50% cytotoxic
concentration (CC50) represented the concentration
required to cause 50% uninfected cell death. The data
reported in Tables 1, 2, and 3 show that most of the
AZT prodrugs tested have selectivity indexes (SI) simi-
lar or better than that of AZT. We also noted that the

782 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5
Vlieghe et al.
Ta ble 3. Anti-HIV Properties of Compound 10 in MT-4 Cells and in PBMCs Acutely Infected with Different Strains of HIV-1 (BRU,
IIIB, and BaL)
MT-4
PBMCs
HIV-1 BRU
HIV-1 IIIB
HIV-1 BaL
SIc
EC50 (nM)
CC50 (µM)
SIc
EC50 (nM)
CC50 (µM)
SIc
a
b
a
b
a
b
compd
EC50 (nM)
CC50 (µM)
10
AZT
0.5
25
>50
>100
>100000
>4000
0.78
7.5
29
8
37200
1067
0.78
11
24
11
30800
1000
a
EC50: concentration in nM required to inhibit the cytopathicity of HIV-1 by 50% on MT-4 cells or to produce inhibition of 50% of
HIV-1 replication on PBMCs. CC50: concentration in µM required to cause 50% death of uninfected MT-4 cells or PBMCs. c SI: selectivity
index ) CC50/EC50
b
.
cytotoxic dose of the compound depends on the cell line
used. For instance, the most active compound 10 ap-
peared to be slightly more toxic than AZT in MT-4 cells,
while in PBMCs it was approximately 2 to 4 times less
cytotoxic than the parent drug. In any cases, the SI of
compound 10 was approximately 30 times higher than
the SI of AZT. These results suggest that the prodrug
10 may have better clinical potential than its parent
drug AZT for the treatment of AIDS.
potential than its parent drug (AZT) for the treatment
of AIDS. However, the mechanism of action of this 3′-
azido-3′-deoxythymidin-5′-yl O-(4-hydroxybutyl) carbon-
ate 10 might be clarified by intramolecular quantifica-
tion of the amount of AZT-MP, AZT-DP, and AZT-TP,
for example.
Consequently, the cyclic intramolecular rearrange-
ment of the carbonate series, specifically associated with
the 5′-O-(4-hydroxybutyl) carbonate moiety, constitutes
a new concept which could be extended to other known
drugs and warrants their further investigation as
potential clinical candidates.
Con clu sion
In the present paper, we report the synthesis, biologi-
cal properties, and the anti-HIV evaluation of a new
series of 3′-azido-3′-deoxythymidin-5′-yl O-(ω-hydroxy-
alkyl) carbonate prodrugs. Among this AZT prodrug
series, compound 10 (3′-azido-3′-deoxythymidin-5′-yl
O-(4-hydroxybutyl) carbonate) was found to be 50 and
30 times more potent than AZT against the replication
of HIV in acutely infected MT-4 cells and PBMCs,
respectively. Serum hydrolase studies have demon-
strated that 3′-azido-3′-deoxythymidin-5′-yl O-(ω-hy-
droxyalkyl) carbonate prodrugs can be recovered through
an intramolecular cyclic rearrangement process (Figure
2). In fact, the initial prodrug is rearranged through an
intramolecular cyclic mechanism resulting from the
nucleophilic attack of the 5′-O-carbonate bond by the
terminal hydroxyl function of the spacer. The optimum
alkyl chain, for this cyclic rearrangement process,
includes 4 methylene groups inserted between the 5′-
O-carbonate bond of the AZT prodrug and the terminal
hydroxyl function (compound 10). This finding is in
complete agreement with our thermodynamic predic-
tions. Prodrug 10 has an enhanced half-life in human
serum (t1/2 ) 30 min) compared with the other 3′-azido-
3′-deoxythymidin-5′-yl O-(ω-hydroxyalkyl) carbonate
prodrugs 16-20 (1 < t1/2 < 16 min, Table 2). This cyclic
rearrangement process could explain the observed im-
provement of anti-HIV activity in prodrug 10. The
increased anti-HIV activity of compound 10 can be also
attributed to its resistance to enzymatic hydrolysis
(increased serum half-life), its increased cellular uptake
(membrane permeation), and its continued cleavage to
AZT inside the target cells (prolonged release). We have
observed that compound 10 is less cytotoxic than AZT
in PBMCs. Our results suggest that this prodrug could
enhance AZT bioavailability and result in better clinical
Exp er im en ta l Section
Nuclear magnetic resonance spectra were recorded with a
Bruker AC-250 spectrometer (1H NMR and 13C NMR); chemi-
cal shifts are expressed as δ units (part per million) downfield
from TMS (tetramethylsilane). Fast atom bombardment (FAB+
or FAB-) mass spectral analyses were obtained by Dr. Astier
(Laboratoire de Mesures Physiques-RMN, USTL, Montpellier,
France) on a J EOL DX-100 using a cesium ion source and
glycerol/thioglycerol (1:1) or m-nitrobenzyl alcohol (NOBA) as
matrix. Mass calibration was performed using cesium iodide.
IR spectra were recorded on a Perkin-Elmer FTIR 1605
spectrophotometer. Microanalyses were carried out by Service
Central d’Analyses du CNRS (Venaison, France) and were
within 0.4% of the theoretical values. Thin-layer chromatog-
raphy (TLC) and preparative layer chromatography (PLC)
were performed using silica gel plates 0.2, 1, or 2 mm thick
(60F254 Merck). Preparative flash column chromatography was
carried out on silica gel (230-240 mesh, G60 Merck). Analyti-
cal HPLC was performed on a Waters 600E instrument with
a M991 detector using the following conditions: 4.6 × 150 mm
column (Waters Spherisorb S5 ODS2, 5 µM); mobile phases:
A ) 0.1% TFA in H2O, B ) CH3OH; flow rate 1 mL/min. All
reagents were of commercial quality (Aldrich Company) from
freshly opened containers.
Cell Lin es, Vir u s, a n d Cell Cu ltu r es. (i) The CEM cell
line and the T-leukemia virus type one (HTLV-I) CD4+ T cell
line, MT-4, were cultured in RPMI/10% FCS, and the medium
replaced twice a week. The laboratory-adapted strain HIVLAV
clade B stock was prepared from the supernatant of infected
CEM cell line, and aliquots were kept frozen at -80 °C until
use.1
(ii) Inhibition of HIV-1 replication was also determined with
PBMCs from seronegative donors. Briefly, 10 × 106 stimulated
PBMCs were infected, respectively, with HIV-1 IIIB and HIV-1
BaL and were incubated for 2 h at 37 °C under 4.5% CO2.
These aliquots were kept frozen at -80 °C until used.
An ti-HIV Activity Assa ys. (i) Anti-HIV activity was
monitored by the efficiency of drug compounds to inhibit the

Synthesis of AZT Prodrugs
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5 783
cytopathicity of HIV in MT-4 cells as already described.52,53
Briefly, 3 × 105 MT-4 cells were first preincubated with 100
µL of various concentrations of drug compounds dissolved in
DMSO or in H2O and then diluted in phosphate buffer saline
solution for 1 h at 37 °C. Then 100 µL of an appropriate virus
dilution was added to the mixture, and cells incubated for 1 h
at 37 °C. After three washes, cells were resuspended in culture
medium in the presence or absence of drug compounds.
Cultures were then continued for 7 days at 37 °C under 5%
CO2 atmosphere, and medium was replaced on day 3 post-
infection with culture medium supplemented or not with drug
compounds. Each culture condition was carried out in dupli-
cate. Viral cytopathicity was followed each day with an
inverted optical microscope. Typically, the virus dilution used
in this assay (multiplicity of infection of 0.1 TCID/cell) led to
cytopathicity on day 5 post-infection. The inhibitory concentra-
tion of drug compounds was expressed as the concentration
that caused 50% inhibition of viral cytopathicity (EC50) without
areas for all the compounds studied, 1, 3, 5-11, 13, and
15-21, in normal human serum are summarized in Tables 1
and 2.
Ch em ica l Syn th eses. Mon o-(3′-a zid o-3′-d eoxyth ym i-
d in -5′-yl) Ester 1,5-P en ta n ed ioic Acid 1. To a solution of
AZT (0.100 g, 0.374 mmol, 1 equiv) in anhydrous DMF (3 mL),
under N2, containing DMAP (0.043 g, 0.374 mmol, 1 equiv)
was added glutaric anhydride (0.047 g, 0.412 mmol, 1.1 equiv).
The reaction mixture was stirred at room temperature for 14
h. Then, the solvent was evaporated under reduced pressure.
The residual oil was dissolved in CH2Cl2 (5 mL) and succes-
sively washed with 1 N HCl (2 × 4 mL) and water (2 × 4 mL).
The combined aqueous solutions were extracted twice with
CH2Cl2. The combined organic layers were dried over Na2SO4,
and the solvent was removed under vacuum. The residue was
purified by flash chromatography on silica gel, using CH3OH/
CH2Cl2 5:95 as eluent, to give the title compound 1 as a white
foam (0.122 g, 86%): Rf ) 0.28 (CH3OH/CH2Cl2 10:90); MS
(FAB+) 382 (M + H)+; HPLC tr ) 15.2 min (method I). Anal.
(C15H19N5O7) C, H, N.
Gen er a l P r oced u r e A for t h e P r ep a r a t ion of Mixed
An h yd r id es. To a solution of ω-carboxylic acid (1 equiv) in
anhydrous EtOAc (5 mL/mmol), under N2, containing Et3N (3
equiv) was added dropwise at 0 °C 3,4,5-trimethoxybenzoyl
chloride (1.1 equiv) in anhydrous EtOAc (5 mL/mmol). The
reaction mixture was then stirred for 4 h at room temperature,
diluted with EtOAc, and washed twice with 0.5 N HCl. After
extraction, the organic layer was washed twice with water and
dried over Na2SO4. EtOAc was removed under reduced pres-
sure, and the product slowly crystallized under vacuum. The
resulting mixed anhydride was used in the next step without
further purification.
Mixed An h yd r id e of 3,4,5-Tr im eth oxyben zoic Acid
a n d Mon o-(3′-a zid o-3′-d eoxyth ym id in -5′-yl) Ester 1,5-
P en ta n ed ioic Acid 2. According to the general procedure A,
the reaction of compound 1 (0.048 g, 0.126 mmol, 1 equiv)
afforded the title compound 2 (0.062 g, 86% yield): MS (FAB+)
576 (M + H)+.
Gen er a l P r oced u r e B for th e P r ep a r a tion of ω-ter t-
Bu tyl Ester s. To a solution of tert-butyl alcohol (5 equiv) in
anhydrous CH2Cl2 (0.2 mL/mmol), under N2, containing DMAP
(1.2 equiv) was added dropwise mixed anhydride (1 equiv)
dissolved in anhydrous CH2Cl2 (5 mL/mmol). The reaction
mixture was stirred for 14 h at room temperature. The solvent
was evaporated under reduced pressure. The residual oil was
dissolved in CH2Cl2 (5 mL), successively washed with 5%
aqueous citric acid (2 × 3 mL), 5% aqueous NaHCO3 (2 × 3
mL), and water (2 × 3 mL). The combined aqueous solutions
were extracted twice with CH2Cl2. The combined organic layers
were dried over Na2SO4, and the solvent was removed under
vacuum. The isolated ω-tert-butyl ester was pure as demon-
strated by its physicochemical characteristics.
direct toxicity for the cells. The cytotoxic concentration (CC50
)
of drug compounds was monitored based on the growth of
noninfected cells by trypan blue exclusion method and cor-
responded to the concentration required to cause 50% cell
death. It should be emphasized that when compounds required
the addition of DMSO to be solubilized in water, the concen-
tration in volume of DMSO used was always less than 10%
with respect to water (the final concentration of DMSO in
MT-4 cells incubation medium being less than 2%). As far as
DMSO could affect the antiviral activity of the tested drugs,54
antiviral assays in which solutions containing equal concen-
trations of DMSO in water were performed and used as
standard assays for each tested drug. EC50 and CC50 reported
values were then calculated from these standard assays.
Moreover, final DMSO concentrations (1/1000) were very far
from the percentages which induced an enhancement of in
vitro HIV-1 infection of T-cells.54
(ii) PBMCs from healthy donors were isolated by density
centrifugation (Lymphoprep; Nycomed Pharma, AS Diagnos-
tics, Oslo, Norway) and stimulated with phytohemagglutin
(PHA) (Sigma Chemical Co., Bornem, Belgium) for 3 days. The
activated cells (PHA-stimulated blasts) were washed with PBS,
and viral infections were done as described by the AIDS clinical
trial group protocol.55 Briefly, PBMCs (2 × 105/200 µL) were
plated in the presence of serial dilutions of the test compound
and were infected with HIV-1 (IIIB) or HIV-1 (BaL) at
1000CCID50 per milliliter. At day 4 post-infection, 125 µL of
the supernatant of the infected cultures was removed and
replaced with 150 µL of fresh medium containing the test
compound at the appropriate concentration. At 7 days after
plating the cells, the p24 antigen was detected in the culture
supernatant by an enzyme-linked immunosorbent assay
(ELISA) (NEN, Paris, France).
Hyd r olysis of th e P r od r u gs 1, 3, 5-11, 13, a n d 15-21
in Hu m a n Ser u m .37 To 990 µL of normal human serum
(NHS) was added 10 µL of a solution of the prodrug (10 mg/
mL in DMSO), and the mixture was incubated at 37 °C in a
water bath. At various time intervals, the samples (100 µL)
were withdrawn and added immediately to ice-cold methanol
(400 µL). The resulting samples were centrifuged (7 min, 3000
rpm). The supernatants were filtered through nylon filters
(0.45 µm) and then analyzed by HPLC using the following
methods. (i) Method I: 20% of solvent B in A to 100% of B in
20 min. (ii) Method II: 40% of solvent B in A to 100% of B in
20 min (A ) 0.1% TFA in H2O, B ) CH3OH). The absorption
maximum for all the studied prodrugs is at 267 nm; therefore,
this wavelength was used for the HPLC detection. Peak
retention times (tr) were 10.0 min for AZT (method I), 15.2
min for 1 (method I), 17.8 min for 3 (method II), 11.5 min for
5 (method I), 13.4 min for 7 (method I), 16.2 min for 6 (method
II), 11.7 min for 8 (method I), 12.9 min for 9 (method I), 15.5
min for 11 (method I), 18.1 min for 13 (method II), 11.9 min
for 15 (method I), 14.2 min for 16 (method I), 12.5 min for 17
(method I), 15.7 min for 10 (method I), 20.4 min for 21 (method
I), 17.1 min for 18 (method I), 17.9 min for 19 (method I), and
20.1 min for 20 (method I). The t1/2 values calculated from peak
1-(3′-Azid o-3′-d eoxyth ym id in -5′-yl),5-(ter t-bu tyl) P en -
ta n ed ioa te 3. According to the general procedure B, the
reaction of compound 2 (0.062 g, 0.108 mmol, 1 equiv) afforded
the title compound 3 as a white foam (0.035 g, 77% yield): Rf
) 0.52 (CH3OH/CH2Cl2 10:90); MS (FAB-) 436 (M - H)-;
HPLC tr ) 17.8 min (method II). Anal. (C19H27N5O7) C, H, N.
4-(ter t-Bu toxyca r bon yl)a m in o-1-(3′-a zid o-3′-d eoxyth y-
m id in -5′-yl) Bu ta n oa te 4. To a solution of 4-(tert-butoxycar-
bonyl)amino-1-butanoic acid (0.285 g, 1.403 mmol, 2.5 equiv)
in anhydrous DMF (3 mL), under N2, cooled in an ice bath to
0 °C was added DCC (0.139 g, 0.674 mmol, 1.2 equiv). The
reaction mixture was stirred at 0 °C for 2 h. Then, a solution
of AZT (0.150 g, 0.561 mmol, 1 equiv) in anhydrous DMF (2
mL), containing DMAP (0.069 g, 0.561 mmol, 1 equiv), was
added dropwise under N2. The reaction mixture was stirred
at room temperature for 24 h. The solvent was evaporated
under reduced pressure. The residual oil was dissolved in
EtOAc (8 mL) and filtered. The filtrate was successively
washed with 5% aqueous citric acid (2 × 7 mL), 5% aqueous
NaHCO3 (2 × 7 mL), and water (2 × 7 mL). The combined
aqueous solutions were extracted twice with EtOAc. The
combined organic layers were dried over Na2SO4, and the

784 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5
Vlieghe et al.
solvent was removed under vacuum. The residue was purified
by flash chromatography on silica gel, using CH3OH/CH2Cl2
5:95 as eluent, to give the title compound 4 as a light yellow
oil (0.228 g, 90% yield): Rf ) 0.29 (CH3OH/CH2Cl2 5:95); MS
(FAB+) 453 (M + H)+.
µL, 0.842 mmol) in DMF afforded the title compound 10 as a
white foam (0.182 g, 85% yield): Rf ) 0.27 (EtOAc/hexane 85:
15); MS (FAB+) 384 (M + H)+; HPLC tr ) 15.7 min (method
I). Anal. (C15H21N5O7) C, H, N.
3′-Azid o-3′-d eoxyth ym id in -5′-yl O-(3-Ca r boxyp r op yl)
Ca r bon a te 11. To a solution of compound 10 (0.100 g, 0.261
mmol, 1 equiv) in anhydrous CH2Cl2 (2 mL), under N2,
containing Tempo‚HCl (40 µL from a 0.2 M solution in CH2-
Cl2, 0.008 mmol, 0.03 equiv) was added dropwise at 0 °C a
solution of m-CPBA (0.081 g, 0.470 mmol, 1.8 equiv) in
anhydrous CH2Cl2 (2 mL). The reaction mixture was stirred
at room temperature for 2 h. The resulting solution was
washed with 5% aqueous citric acid (2 × 5 mL) and water (2
× 5 mL). The combined aqueous solutions were extracted twice
with CH2Cl2. The combined organic layers were dried over Na2-
SO4, and the solvent was removed under vacuum. The residue
was purified by PLC, using CH3OH/CH2Cl2 6:94 as eluent, to
give the title compound 11 as a white foam (0.072 g, 69%): Rf
) 0.22 (CH3OH/CH2Cl2 8:92); MS (FAB+) 398 (M + H)+; HPLC
tr ) 15.5 min (method I). Anal. (C15H19N5O8) C, H, N.
Gen er a l P r oced u r e C for th e Dep r otection of N-Boc
Am in o Gr ou p a n d ω-ter t-Bu tyl Ester s. TFA (25 equiv) was
added by syringe to a solution of N-Boc amino or ω-tert-butyl
ester compound (1 equiv) in anhydrous CH2Cl2 (2 mL), under
N2. The reaction mixture was stirred at room temperature for
4 h. The solvent was evaporated under reduced pressure. The
ω-amino compound was pure as demonstrated by its physico-
chemical characteristics. In the case of ω-carboxylic acid, the
residue was dissolved in CH2Cl2 (5 mL), washed with 1N HCl
(2 × 4 mL) and, water (2 × 4 mL). The combined aqueous
solutions were extracted twice with CH2Cl2, the combined
organic layers were dried over Na2SO4, and the solvent was
removed under vacuum. The ω-carboxylic acid was pure as
demonstrated by its physicochemical characteristics.
4-Am in o-1-(3′-a zid o-3′-d eoxyth ym id in -5′-yl) Bu ta n oa te
5. According to the general procedure C, the reaction of
compound 4 (0.200 g, 0.442 mmol, 1 equiv) afforded the title
compound 5 as a white foam (0.154 g, quantitative yield): Rf
) 0.14 (CH3OH/CH2Cl2 10:90); MS (FAB+) 353 (M + H)+;
HPLC tr ) 11.5 min (method I). Anal. (C14H20N6O5) C, H, N.
Mixed An h yd r id e of 3,4,5-Tr im eth oxyben zoic Acid
a n d 3′-Azid o-3′-d eoxyth ym id in -5′-yl O-(3-Ca r boxyp r op yl)
Ca r bon a te 12. According to the general procedure A, the
reaction of compound 11 (0.112 g, 0.282 mmol, 1 equiv)
afforded the title compound 12 (0.147 g, 88% yield): MS
(FAB+) 592 (M + H)+.
Gen er a l P r oced u r e
D for th e F or m a tion of 5′-O-
Ca r ba m a te- or 5′-O-Ca r bon a te-Bon d s on AZT P r od r u gs.
To a solution of AZT (1 equiv) in anhydrous DMF (compounds
6, 8, 10, 16-20) or in anhydrous CH3CN (compounds 9, 14,
21), N,N-carbonyldiimidazole (CDI, 1.1 equiv) was added under
N2. The reaction mixture was stirred at room temperature for
2 h. Afterward, the ω-amino or ω-hydroxyl compound (1.5
equiv) was added, the reaction mixture was stirred overnight
at room temperature (in the case of carbamate bond formation)
or at 60 °C (in the case of carbonate bond formation). The
solvent was removed under reduced pressure, and the residual
oil was dissolved in CH2Cl2. The organic phase was succes-
sively washed with 5% aqueous citric acid and/or 5% aqueous
NaHCO3 and/or water and then dried over Na2SO4. Evapora-
tion of the solvent under vacuum gave a crude product, which
was then purified by flash chromatography on silica gel or
preparative layer chromatography (PLC, 1 or 2 mm thick),
using hexane/EtOAc 20:80 as eluent.
3′-Azid o-3′-d eoxyth ym id in -5′-yl N-[3-(ter t-Bu toxyca r -
bon yl)p r op yl] Ca r b a m a t e 6. According to the general
procedure D, the reaction of AZT (0.060 g, 0.225 mmol, 1 equiv)
and 4-amino-1-tert-butyl butanoate (0.055 g, 0.338 mmol) in
DMF afforded the title compound 6 as a light yellow oil (0.066
g, 65% yield): Rf ) 0.47 (CH3OH/CH2Cl2 10:90); MS (FAB-)
451 (M - H)-; HPLC tr ) 16.2 min (method II). Anal.
(C19H28N6O7) C, H, N.
3′-Azid o-3′-d eoxyt h ym id in -5′-yl N-(3-Ca r b oxyp r op yl)
Ca r ba m a te 7. According to the general procedure C, the
reaction of compound 6 (0.040 g, 0.088 mmol, 1 equiv) afforded
the title compound 7 as a white foam (0.033 g, 94% yield): Rf
) 0.22 (CH3OH/CH2Cl2 10:90); MS (FAB+) 397 (M + H)+;
HPLC tr ) 13.4 min (method I). Anal. (C15H20N6O7) C, H, N.
3′-Azido-3′-deoxyth ym idin -5′-yl N-(3-Am in opr opyl) Car -
ba m a te 8. According to the general procedure D, the reaction
of AZT (0.150 g, 0.561 mmol, 1 equiv) and 1,3-diaminopropane
(70 µL, 0.842 mmol) in DMF afforded the title compound 8 as
a yellow gum (0.161 g, 78% yield): Rf ) 0.43 (CH3OH/CH2-
Cl2/AcOH 45:45:10); MS (FAB+) 368 (M + H)+; HPLC tr ) 11.7
min (method I). Anal. (C14H21N7O5) C, H, N.
3′-Azid o-3′-d eoxyth ym id in -5′-yl N-(3-Hyd r oxyp r op yl)
Ca r ba m a te 9. According to the general procedure D, the
reaction of AZT (0.060 g, 0.225 mmol, 1 equiv) and 1,3-
diaminopropane (25 µL, 0.338 mmol) in CH3CN afforded the
title compound 9 as a light yellow oil (0.050 g, 60% yield): Rf
) 0.37 (CH3OH/CH2Cl2 10:90); MS (FAB+) 369 (M + H)+;
HPLC tr ) 12.9 min (method I). Anal. (C14H20N6O6) C, H, N.
3′-Azido-3′-deoxyth ym idin -5′-yl O-(4-Hydr oxybu tyl) Car -
bon a te 10. According to the general procedure D, the reaction
of AZT (0.150 g, 0.561 mmol, 1 equiv) and 1,4-butanediol (75
3′-Azid o-3′-d eoxyth ym id in -5′-yl O-[3-(ter t-Bu toxyca r -
bon yl)p r op yl] Ca r bon a te 13. According to the general
procedure B, the reaction of compound 12 (0.115 g, 0.195 mmol,
1 equiv) afforded the title compound 13 as a light yellow oil
(0.067 g, 75% yield): Rf ) 0.33 (CH3OH/CH2Cl2 5:95); MS
(FAB-) 452 (M - H)-; HPLC tr ) 18.1 min (method II). Anal.
(C19H27N5O8) C, H, N.
3′-Azid o-3′-d eoxyth ym id in -5′-yl O-[3-(ter t-Bu toxyca r -
bon yl)a m in op r op yl] ca r bon a te 14. According to the general
procedure D, the reaction of AZT (0.150 g, 0.561 mmol, 1 equiv)
and 3-(tert-butoxycarbonyl)amino-1-propanol (0.147 g, 0.842
mmol) in CH3CN afforded the title compound 14 as a light
yellow oil (0.176 g, 68% yield): Rf ) 0.29 (EtOAc/hexane 80:
20); MS (FAB+) 469 (M + H)+.
3′-Azido-3′-deoxyth ym idin -5′-yl O-(3-Am in opr opyl) Car -
bon a te 15. According to the general procedure C, the reaction
of compound 14 (0.165 g, 0.352 mmol, 1 equiv) afforded the
title compound 15 as a white foam (0.129 g, quantitative
yield): Rf ) 0.18 (CH3OH/CH2Cl2 10:90); MS (FAB+) 369 (M
+ H)+; HPLC tr ) 11.9 min (method I). Anal. (C14H20N6O6) C,
H, N.
3′-Azid o-3′-d eoxyth ym id in -5′-yl O-(3-Hyd r oxyp r op yl)
Ca r bon a te 16. According to the general procedure D, the
reaction of AZT (0.120 g, 0.449 mmol, 1 equiv) and 1,3-
propanediol (50 µL, 0.674 mmol) in DMF afforded the title
compound 16 as a light yellow oil (0.109 g, 66% yield): Rf )
0.20 (EtOAc/hexane 85:15); MS (FAB+) 370 (M + H)+; HPLC
tr ) 14.2 min (method I). Anal. (C14H19N5O7) C, H, N.
3′-Azido-3′-deoxyth ym idin -5′-yl O-(2-Hydr oxyeth yl) Car -
bon a te 17. According to the general procedure D, the reaction
of AZT (0.120 g, 0.449 mmol, 1 equiv) and 1,2-ethanediol (40
µL, 0.674 mmol) in DMF afforded the title compound 17 as a
light yellow oil (0.089 g, 56% yield): Rf ) 0.18 (EtOAc/hexane
85:15); MS (FAB+) 356 (M + H)+; HPLC tr ) 12.5 min (method
I). Anal. (C13H17N5O7) C, H, N.
3′-Azid o-3′-d eoxyth ym id in -5′-yl O-(5-Hyd r oxyp en tyl)
Ca r bon a te 18. According to the general procedure D, the
reaction of AZT (0.120 g, 0.449 mmol, 1 equiv) and 1,5-
pentanediol (70 µL, 0.674 mmol) in DMF afforded the title
compound 18 as a white foam (0.124 g, 70% yield): Rf ) 0.30
(EtOAc/hexane 85:15); MS (FAB+) 398 (M + H)+; HPLC tr )
17.1 min (method I). Anal. (C16H23N5O7) C, H, N.
3′-Azid o-3′-d eoxyt h ym id in -5′-yl O-(6-H yd r oxyh exyl)
Ca r bon a te 19. According to the general procedure D, the
reaction of AZT (0.120 g, 0.449 mmol, 1 equiv) and 1,6-
hexanediol (0.080 g, 0.674 mmol) in DMF afforded the title
compound 19 as a white foam (0.089 g, 49% yield): Rf ) 0.35

Synthesis of AZT Prodrugs
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5 785
(EtOAc/hexane 85:15); MS (FAB+) 412 (M + H)+; HPLC tr )
17.9 min (method I). Anal. (C17H25N5O7) C, H, N.
(12) Yarchoan, R.; Broder, S. Anti-retroviral therapy of AIDS and
related disorders: general principles and specific development
of dideoxynucleotides. Pharmacol. Ther. 1989, 40, 329-348.
(13) Dournon, E.; Matheron, S.; Rozenbaum, W.; Gharakhalnian, S.;
Michon, C.; Girard, P. M.; Perronne, C.; Salmon, D.; De Truchis,
P.; Leport, C.; Bouvet, E.; Dazza, M. C.; Levacher, M.; Regnier,
B. Effects of zidovudine in 365 consecutive patients with AIDS
of AIDS-related complex. Lancet 1988, ii, 1297-1302.
(14) 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′-dideoxyribonucleo-
sides: a comparison with their parental 2′,3′-dideoxyribonucleo-
sides. Mol. Pharmacol. 1987, 32, 162-167.
(15) Cooney, D. A.; Dalal, M.; Mitsuya, H.; Mc Mahon, J . B.;
Nadkarni, M.; Balzarini, J .; Broder, S.; J ohns, D. G. Initial
studies on the cellular pharmacology of 2′,3′-dideoxycytidine, an
inhibitor of HTLV-III infectivity. Biochem. Pharmacol. 1986,
35, 2065-2068.
(16) Ahluwalia, G.; Cooney, D. A.; Mitsuya, H.; Fridland, A.; Flora,
K. P.; Hao, Z.; Dalal, M.; Broder, S.; J ohns, D. G. Initial studies
on the cellular pharmacology of 2′,3′-dideoxyinosine, an inhibitor
of HIV infectivity. Biochem. Pharmacol. 1987, 36, 3797-3800.
(17) Coates, J . A. V.; Cammack, N. S.; J enkinson, H. J .; J owett, A.
J .; J owett, M. I.; Pearson, B. A.; Penn, C. R.; Rouse, P. L.; Viner,
K. C.; Cameron, J . M. (-)-2′-deoxy-3′-thiacytidine is a potent,
highly selective inhibitor of human immunodeficiency virus type
1 and type 2 replication in vitro. Antimicrob. Agents Chemother.
1992, 36, 733-739.
3′-Azido-3′-deoxyth ym idin -5′-yl O-(8-Hydr oxyoctyl) Car -
bon a te 20. According to the general procedure D, the reaction
of AZT (0.120 g, 0.449 mmol, 1 equiv) and 1,8-octanediol (0.066
g, 0.674 mmol) in DMF afforded the title compound 20 as a
white foam (0.125 g, 63% yield): Rf ) 0.48 (EtOAc/hexane 85:
15); MS (FAB+) 440 (M + H)+; HPLC tr ) 20.1 min (method
I). Anal. (C19H29N5O7) C, H, N.
3′-Azid o-3′-d eoxyth ym id in -5′-yl O-Bu tyl Ca r bon a te 21.
According to the general procedure D, the reaction of AZT
(0.100 g, 0.374 mmol, 1 equiv) and 1-butanol (52 µL, 0.561
mmol) in CH3CN afforded the title compound 21 as a light
yellow oil (0.084 g, 64% yield): Rf ) 0.65 (EtOAc/hexane 80:
20); MS (FAB+) 368 (M + H)+; HPLC tr ) 20.4 min (method
I). Anal. (C15H21N5O6) C, H, N.
Ack n ow led gm en t. This research was supported by
grants from LAPHAL Laboratories and PACA Regional
Council (P.V.). INSERM is acknowledged for financial
support. We thank Dr. Toni Williamson for correcting
the English, Cindy Heens for excellent technical as-
sistance with the PBMCs assays, and Karine Barral for
synthesis of compounds 18 and 19.
(18) Crimmins, M. T.; King, B. W. An efficient asymmetric approach
to carbocyclic nucleosides: asymmetric synthesis of 1592U89, a
potent inhibitor of HIV reverse transcriptase. J . Org. Chem.
1996, 61, 4192-4193.
Su p p or tin g In for m a tion Ava ila ble: NMR and IR spec-
troscopic data of the desired compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
(19) Starrett, J . E., J r.; Tortolani, D. R.; Russell, J .; Hitchcock, M. J .
M.; Whiterock, V.; Martin, J . C.; Mansuri, M. M. Synthesis, oral
bioavailability determination, and in vitro evaluation of prodrugs
of the antiviral agent 9-[2-(phosphonomethoxy)ethyl]adenine
(PMEA). J . Med. Chem. 1994, 37, 1857-1864.
Refer en ces
(1) Barre´-Sinoussi, F.; Chermann, J . C.; Rey, F.; Nugeyre, M. T.;
Chamaret, S.; Gruest, J .; Dauguet, C.; Axler-Blin, C.; Brun-
Ve´zinet, F.; Rouzioux, C.; Rozenbaum W.; Montagnier, L. Isola-
tion of a T-lymphotropic retrovirus from a patient at risk for
Acquired Immune-Deficiency Syndrome (AIDS). Science 1983,
220, 868-871.
(2) Montagnier, L.; Dauguet, C.; Axler, C.; Chamaret, S.; Gruest,
J .; Nugeyre, M. T.; Rey, F.; Barre´-Sinoussi, F.; Chermann, J . C.
A new type of retrovirus isolated from patients presenting with
lymphadenopathy and AIDS: structural and antigenic related-
ness with equine infectious anemia virus. Ann. Inst. Pasteur,
Virol. 1984, 135E (nE1), 119-134.
(3) Gallo, R. C.; Salahuddin, S. Z.; Popovic, M.; Shearer, G. M.;
Kaplan, M.; Haynes, B. F.; Palker, T. J .; Redfield, R.; Oleske,
J .; Safai, B.; White, G.; Foster, P.; Markham, P. D. Frequent
detection and isolation of cytopathic retroviruses (HTLV-III)
from patients with AIDS and at risk of AIDS. Science 1984, 224,
500-503.
(4) Volsky, D. J .; Sakai, K.; Stevenson, M.; Dewhurst, S. Retroviral
etiology of the acquired immune deficiency syndrome (AIDS).
AIDS Res. 1986, 2 (suppl. 1), S35-48.
(5) 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′-
dideoxynucleosides. Proc. Natl. Acad. Sci. U.S.A. 1986, 83,
1911-1915.
(6) De Clercq, E. In search of a selective antiviral chemotherapy.
Clin. Microbiol. Rev. 1997, 10, 674-693.
(7) Furman, P. A.; Fyfe, J . A.; St. Clair, M. H.; Weinhold, K.;
Rideout, J . L.; Freeman, G. A.; Nusinoff-Lehrman, S.; Bolognesi,
D. P.; Broder, S.; Mitsuya H.; Barry, D. W. Phosphorylation of
3′-azido-3′-deoxythymidine and selective interaction of the 5′-
triphosphate with HIV reverse transcriptase. Proc. Natl. Acad.
Sci. U.S.A. 1986, 83, 8333-8337.
(8) Hao, Z.; Cooney, D. A.; Hartman, N. R.; Perno, C. F.; Fridland,
A.; De Vico, A. L.; Sarngadharan, M. G.; Broder, S.; J ohns, D.
G. Factors determining the activity of 2′,3′-dideoxynucleosides
in suppressing human immunodeficiency virus in vitro. Mol.
Pharmacol. 1988, 34, 431-435.
(20) Skoog, M. T.; Hargrave, K. D.; Miglietta, J .J ; Kopp, E. B.;
Merluzzi, V. J . Inhibition of HIV-1 reverse transcriptase and
virus replication by
a non nucleoside dipyridodiazepinone
BI-RG-587 (Nevirapine). Med. Res. Rev. 1992, 12, 27-40.
(21) Romero, D. L.; Morge, R. A.; Genin, M. J .; Biles, C.; Busso, M.;
Resnick, L.; Althaus, I. W.; Reusser, F.; Thomas, R. C.; Tarpley,
W. G. Bis(heteroaryl)piperazine(BHAP) reverse transcriptase
inhibitors: structure-activity relationships of novel substituted
indole analogues and the identification of 1-[(5-methanesulfon-
amido-1H-indol-2-yl)carbonyl]-4-[3-[(1-methylethyl)amino]pyridi-
nyl]piperazine monomethane sulfonate (U-90152S), a second-
generation clinical candidate. J . Med. Chem. 1993, 36, 1505-
1508.
(22) Young, S. D.; Britcher, S. F.; Tran, L. O.; Payne, L. S.; Lumma,
W. C.; Lyle, T. A.; Huff, J . R.; Anderson, P. S.; Olsen, D. B.;
Carroll, S. S. and al. L-743, 726 (DMP-266): a novel, highly
potent nonnucleoside inhibitor of the human immunodeficiency
virus type 1 reverse transcriptase. Antimicrob. Agents Chemo-
ther. 1995, 39, 2602-2065.
(23) De Clercq, E. The role of nonnucleoside reverse transcriptase
inhibitors (NNRTIs) in the therapy of HIV-1 infection. Antivir.
Res. 1998, 38, 153-179.
(24) Alterman, M.; Bjo¨rsne, M.; Mu¨hlman, A.; Classon, B.; Kvarn-
stro¨m, I.; Danielson, H.; Markgren, P. O.; Nillroth, U.; Unge,
T.; Hallberg, A.; Samuelsson, B. Design and synthesis of new
potent C2-symmetric HIV-1 protease inhibitors. Use of L-
mannaric acid as a peptidomimetic scaffold. J . Med. Chem. 1998,
41, 3782-3792.
(25) Miller, S. L. Amprenavir approved for HIV treatment. Am. J .
Health-Syst. Pharm. 1999, 56, 1057-1058.
(26) De Clercq, E. Toward improved anti-HIV chemotherapy: thera-
peutic strategies for intervention with HIV infections. J . Med.
Chem. 1995, 38, 2491-2517.
(27) Fischl, M. A.; Richman, D. D.; Grieco, M. H.; Gottlieb, M. S.;
Volberding, P. A.; Laskin, O. L.; Leedom, J . M.; Groopman, J .
E.; Mildvan, D.; Schooley, R. T.; J ackson, G. G.; Durack, D. T.;
King, D. The efficacy of azidothymidine (AZT) in the treatment
of patients with AIDS and AIDS-related complex. A double blind,
placebo-controlled trial. N. Engl. J . Med. 1987, 317, 185-191.
(28) Richman, D. D.; Fischl, M. A.; Grieco, M. H.; Gottlieb, M. S.;
Volberding, P. A.; Laskin, O. L.; Leedom, J . M.; Groopman, J .
E.; Mildvan, D.; Hirsch, M. S.; J ackson, G. G.; Durack, D. T.;
Nusinoff-Lehrman, S. The toxicity of azidothymidine (AZT) in
the treatment of patients with AIDS and AIDS-related complex.
A double blind, placebo-controlled trial. N. Engl. J . Med. 1987,
317, 192-197.
(9) Balzarini, J .; De Clercq, E. Nucleoside and nonnucleoside reverse
transcriptase inhibitors active against HIV. In Textbook of AIDS
Medicine; Merigan, T. C., Bartlett, J . G., Bolognesi, D., Eds.;
Williams and Wilkings: Baltimore, 1999; pp 815-847.
(10) Cheng, Y. C.; Dutschman, G. E.; Bastow, K. F.; Sarngadharan,
M. G.; Ting, R. Y. C. HIV reverse transcriptase. General
properties and its interactions with nucleoside triphosphate
analogues. J . Biol. Chem. 1987, 262, 2187-2189.
(11) Mitsuya, H.; Weinhold, K. J .; Furman, P. A.; St. Clair, M. H.;
Nusinoff-Lehrman, S.; Gallo, R. C.; Bolognesi, D. P.; 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.
(29) Rosowsky, A.; Fu, H.; Pai, N.; Mellors, J .; Richman, D. D.;
Hostetler, K. Y. Synthesis and in vitro activity of long chain
5′-O-[(a lkoxyca r bon yl)ph osph in yl]-3′-a zido-3′-deoxyt h ym i-
dines against wild-type and AZT- and forscarnet-resistant
strains of HIV-1. J . Med. Chem. 1997, 40, 2482-2490.

786 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 5
Vlieghe et al.
(30) Kuipers, M. E.; Swart, P. J ., Hendriks, M. M. W. B.; Meijer, D.
F. K. Optimization of the reaction conditions for the synthesis
of neoglycoprotein-AZT-monophosphate conjugates. J . Med.
Chem. 1995, 38, 883-889.
(31) Desseaux, C.; Huynh-Dinh, T. Synthesis of phosphodiester and
triester derivatives of AZT with tethered N-methyl piperazine
and N,N,N′-trimethylethylenediamine. Bioorg. Med. Chem. Lett.
1993, 3, 1547-1550.
(32) Henin, Y.; Gouyette, C.; Schwartz, O.; Debouzy, J . C.; Neumann,
J . M.; Huynh-Dinh, T. Lipophilic glycosyl phosphotriester de-
rivatives AZT: synthesis, NMR, transmembrane transport
study, and antiviral activity. J . Med. Chem. 1991, 34, 1830-
1837.
(33) Piantadosi, C.; Marasco, C. J ., J r.; Morris-Natschke, S. L.; Meyer,
K. L.; Gumus, F.; Surles, J . R.; Ishaq, K. S.; Kucera, L. S.; Iyer,
N.; Wallen, A. C.; Piantadosi, S.; Modest, E. J . Synthesis and
evaluation of novel ether lipid nucleoside conjugates for anti-
VIH-1 activity. J . Med. Chem. 1991, 34, 1408-1414.
(34) Kimura, T.; Matsumoto, H.; Matsuda, T.; Hamawaki, T., Akaji,
K.; Kiso, Y. A new class of anti-HIV agents: synthesis and
activity of conjugates of HIV protease inhibitors with a reverse
transcriptase inhibitor. Bioorg. Med. Chem. Lett. 1999, 9, 803-
806.
(35) Dessolin, J .; Galea, P.; Vlieghe, P.; Chermann, J . C.; Kraus, J .
L. New bicyclam-AZT conjugates: design, synthesis, anti-HIV
evaluation, and their interaction with CXCR-4 coreceptor. J .
Med. Chem. 1999, 42, 229-241.
(36) Mourier, N.; Trabaud, C.; Graciet, J . C.; Simon, V.; Niddam, V.;
Faury, P.; Charvet, A. S.; Camplo, M.; Chermann, J . C.; Kraus,
J . L. Peptide-nucleoside conjugates: synthesis and anti-HIV
activities. Nucleosides Nucleotides 1995, 14, 1393-1402.
(37) Aggarwal, S. K.; Gogu, S. R.; Rangan, S. R. S.; Agrawal, K. C.
Synthesis and biological evaluation of prodrugs of zidovudine.
J . Med. Chem. 1990, 33, 1505-1510.
(38) Chu, C. K.; Bhadti, V. S.; Doshi, K. J .; Etse, J . T.; Gallo, J . M.;
Boudinot, F. D.; Schinazi, R. F. Brain targeting of anti-HIV
nucleosides: synthesis and in vitro and in vivo studies of
dihydropyridine derivatives of 3′-azido-2′,3′-dideoxyuridine and
3′-azido-3′-deoxythymidine. J . Med. Chem. 1990, 33, 2188-2192.
(39) Tadayoni, B. M.; Friden, P. M.; Walus, L. R.; Musso, G. F.
Synthesis, in vitro kinetics, and in vivo studies on protein
conjugates of AZT: evaluation as a transport system to increase
brain delivery. Bioconjugate Chem. 1995, 4, 139-145.
(40) Lefebvre, I.; Perigaud C.; Pompon, A; Aubertin, A. M.; Girardet,
J . L.; Kirn, A.; Gosselin, G. and Imbach, J . L. Mononucleoside
phosphodiester derivatives with S-acyl-2-thioethyl bioreversible
phosphate-protecting group: intercellular delivery of 3′-azido-
2′,3′-dideoxythymidine 5′-monophosphate. J . Med. Chem. 1995,
38, 3941-3950.
(41) Meier, C.; De Clercq, E.; Balzarini, J . Nucleotide delivery from
cyclosaligenyl-3′-azido-3′-deoxythymidine monophosphates (cy-
closal-AZTMP). Eur. J . Org. Chem. 1998, 5, 837-846.
(42) Siddiqui, A. Q.; Balatore, C.; Mc Guigan, C.; De Clercq, E.;
Balzarini, J . The presence on the aryl moiety of the aryl
phosphoramidate derivative of d4T enhances anti-HIV efficacy
in cell culture: a structure-activity relationship. J . Med. Chem.
1999, 42, 393-399.
(43) Saboulard, D.; Naesens, L.; Cahard, D.; Salgado, A.; Pathirana,
R.; Velazquez, S.; Mc Guigan, C.; De Clercq, E.; Balzarini, J .
Characterization of the activation pathway of phosphoramidate
triester prodrugs of stavudine and zidovudine. Mol. Pharmacol.
1999, 56, 693-704.
(44) Tamamura, H.; Ishihara, T.; Oyake, H.; Imai, M.; Otaka, A.;
Ibuka, T.; Arakaki, R.; Nakashima, H.; Murakami, T.; Waki, M.;
Matsumoto, A.; Yamamoto, N.; Fujii, N. Convenient one-pot
synthesis of cysteine-containing peptides using the trimethylsilyl
chloride-dimethyl sulfoxide/trifluoroacetic acid system and its
application to the synthesis of bifunctional anti-HIV compounds.
J . Chem. Soc., Perkin Trans. 1 1998, 495-500.
(45) Greene, T. W.; Wuts, P. G. M. Protection for carboxyl group. In
Protective Groups in Organic Synthesis, 3rd ed.; J ohn Wiley &
Sons Inc.: New York, 1999; pp 369-453.
(46) Pochet, S.; Kansal, V.; Destouesse, F.; Sarfati, S. R. Alkylgly-
coside carbonates of 3′-azido-3′-deoxythymidine. Tetrahedron
Lett. 1990, 31, 6021-6024.
(47) Cella, J . A.; Kelley, J . A.; Kenehan, E. F. Nitroxide-catalyzed
oxidation of alcohols using meta-chloroperbenzoic acid, a new
method. J . Org. Chem. 1975, 40, 1860-1862.
(48) Zimmerman, T. P.; Mahony, W. B.; Prus, K. L. 3′-azido-3′-
deoxythymidine. An unusual nucleoside analogue that perme-
ates the membrane of human-erythrocytes and lymphocytes by
non facilitated diffusion. J . Biol. Chem. 1987, 262, 5748-5754.
(49) Silverman, R. B. Prodrugs and drug delivery systems. In The
organic chemistry of drug design and drug action; Academic
Press Inc.: San Diego, 1992; pp 352-401.
(50) Meyer, W.; Follmann, H. A study of the substrate and inhibitor
specificities of AMP aminohydrolase, 5′-nucleotidase, and aden-
ylate kinase with adenosine carboxylates of variable chain
length. Z. Naturforsch., C. 1980, 35, 273-278.
(51) Hansch, C.; Steward, A. R.; Anderson, S. M.; Bentley, D. The
parabolic dependence of drug action upon lipophilic character
as revealed by a study of hypnotics. J . Med. Chem. 1968, 11,
1-11.
(52) Rey, F.; Barre´-Sinoussi, F.; Schmidtmayerova, H.; Chermann,
J . C. Detection and titration of neutralizing antibodies to HIV
using an inhibition of the cytopathic effect of the virus on MT4
cells. J . Virol. Methods 1987, 16, 239-249.
(53) Rey, F.; Donker, G.; Hirsch, I.; Chermann, J . C. Productive
infection of CD4+ cells by selected HIV strains is not inhibited
by anti-CD4 monoclonal antibodies. Virology 1991, 181, 165-
171.
(54) Seki, J .; Ikeda, R.; Hoshino, H. Dimethyl sulfoxide and related
polar compounds enhance infection of human T cells with HIV-1
in vitro. Biochem. Biophys. Res. Commun. 1996, 227, 724-729.
(55) J apour, A. J .; Mayers, D. L.; J ohnson, V. A.; Kuritzkes, D. R.;
Beckett, L. A.; Arduino, J . M.; Lane, J .; Black, R. L.; Reichelder-
fer, P. S.; D’Aquila, R. T.; Crumpacker, C. S. Standardized
peripheral blood mononuclear cell culture assay for determina-
tion of drug susceptibilities of clinical human immunodeficiency
virus type 1 isolates. Antimicrob. Agents Chemother. 1993, 37,
1095-1101.
J M001033S