Are 5′-O-Carbamate-2′,3′-dideoxythiacytidine new anti-HIV and anti-HBV nucleoside drugs or prodrugs?

Article abstract of DOI:10.1016/S0960-894X(03)00496-7

In contrast to 5′-O-carbonate 3TC derivatives (23, 24), which are clearly 3TC prodrugs, the corresponding 3TC carbamates (15-21 and 25), found to be very stable compounds with respect to enzymatic hydrolysis (cellular lysates and culture cell media) and s

Full text of DOI:10.1016/S0960-894X(03)00496-7

Bioorganic & Medicinal Chemistry Letters 13 (2003) 2459–2463
Are 50-O-Carbamate-20,30-dideoxythiacytidine New Anti-HIV
and Anti-HBV Nucleoside Drugs or Prodrugs?
Carole Anastasi,a,b Patrick Vlieghe,b Olivier Hantz,c Olivier Schorr,c
Christophe Pannecouque,d Myriam Witvrouw,d Erik De Clercq,d
Pascal Clayette,e,f Nathalie Dereuddre-Bosquet,e,f Dominique Dormont,e
Francoise Gondois-Rey,g Ivan Hirschg and Jean-Louis Krausa,*
a
´
´
´
´
Laboratoire de Chimie Biomoleculaire, INSERM U-382, IBDM, Universite Mediterranee,
Parc Scientifique de Luminy, 163 avenue de Luminy, case 901, 13288 Marseille Cedex 9, France
bLaboratoires LAPHAL, Avenue de Provence, B.P. 7, 13718 Allauch Cedex, France
c
´
´
´
INSERM U-271, Unite de Recherche sur les Hepatites, le Sida et les Retrovirus Humains,
151 cours A. Thomas, 69424 Lyon cedex 3, France
dRega Institute for Medicinal Research, Katholieke Universiteit Leuven, B-3000 Leuven, Belgium
eCEA, Service de Neurovirologie, DSV/DRM, CRSSA, 60-68 avenue de la division Leclerc, B.P. 6,
92265 Fontenay aux Roses Cedex, France
fSPI-BIO, 2 rue du Buisson aux Fraises, Z.I. de la Bonde, 91741 Massy Cedex, France
g
´
´
`
INSERM U-372, Unite de Pathogenie des Infections a Lentivirus, Parc Scientifique de Luminy,
163 avenue de Luminy, B.P 178, 13276 Marseille Cedex 9, France
Received 3 April 2003; revised 14 May 2003; accepted 15 May 2003
Abstract—In contrast to 50-O-carbonate 3TC derivatives (23, 24), which are clearly 3TC prodrugs, the corresponding 3TC carba-
mates (15–21 and 25), found to be very stable compounds with respect to enzymatic hydrolysis (cellular lysates and culture cell
media) and still active on both HIV-1 and HBV infected cells, may not be 3TC prodrugs. The antiviral properties as well as the
mechanism of action of 3TC analogues have been studied and evaluated.
# 2003 Elsevier Ltd. All rights reserved.
Most of the nucleoside prodrugs described so far were
designed for various purposes: to decrease toxicity
associated with nucleoside drugs, to release active spe-
cies by the degradative metabolism of the prodrugs, and
to allow higher quantities of drug to enter into the cell.
For this last purpose lipophilic chemical moieties were
linked to the 50-O position of the nucleoside through
enzymatically hydrolysable functions, such as ester1 or
carbonate bonds.2,3 Although some 50-O-carbamoyl-
20,30-dideoxynucleoside derivatives were reported inac-
tive as antiviral compounds because of the enzymatic
stability of the carbamate bond,4 recently various amino-
acids coupled to AZT through 50-O-carbamate bonds
have been found to be active in HIV-1 infected PBMCs
(peripheral blood mononuclear cells).5 Furthermore,
their antiviral activities were suggested to be unrelated
to the release of AZT in the cell, suggesting that these
50-O-carbamate analogues may act through an original
pathway.2,5
To date only (ꢀ)-20-deoxy-30-thiacytidine or 3TC (Fig.
1) has been licensed as both an anti-HBV and an anti-
HIV drug (Epivir1 against HIV infection, and Zeffix1
as anti-hepatitis drug), so we have synthesized new 3TC
carbamate analogues (Fig. 2). In this paper we report
the intriguing in vitro antiviral properties (HIV and
HBV) associated with this new class of nucleosides,
their enzymatic stabilities, and some preliminary studies
on their mechanism of action in comparison with their
corresponding carbonate analogues.
The syntheses of 3TC carbamate and carbonate analo-
gues are given in Scheme 1. The 3TC protected
*Corresponding author. Fax: +33-491-829416; e-mail: kraus@
luminy.univ-mrs.fr
0960-894X/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00496-7

2460
C. Anastasi et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2459–2463
nucleosides were treated with phosgene6ꢀ8 in dichloro-
methane (DCM) at room temperature in the presence of
diisopropylethylamine (DIEA) for several hours and
then selected o-functionalised primary amines were con-
densed. The yield of the corresponding 50-O-3TC-carba-
mates ranged from 40 to 70%. The deprotection of
N-Boc (tert-butyloxycarbonyl) group by trifluoroacetic
acid (TFA) in DCM led to the final compounds in
quantitative yields.
In parallel, we also accessed the 50-O-carbamates
through the use of o-aminoalkyl-50-O-3TC-carbonates.
As shown in Figure 3, we demonstrated for the first time
that 3TC-carbonate derivatives (23 and 24) could rear-
range into the corresponding 50-O-3TC carbamates
through the intramolecular nucleophilic attack of the o-
amino group on the 50-O-carbonate function. We found
that o-aminoalkyl-50-O-3TC-carbonates rearranged
partially into their corresponding carbamates in biolo-
gical media (extra and intracellular media) as indicated
in studies on stability and lipophilicity, but are chemi-
cally stable enough to be isolated in the experimental
conditions of synthesis.
Figure 1. Structure of 3TC, the only nucleoside drug licensed for HIV
as well as HBV treatment.
50-O-3TC-carbamate drugs (compounds 15 to 22 and
25) showed a less pronounced activity to inhibit the
cytopathicity of HIV after MT-49 cells and PBMCs10
acute infection, compared to that of the parent drug
Figure 2. Structure of carbamate and carbonate analogues of 3TC.
Scheme 1. Synthesis of the carbamate and carbonate analogues of 3TC. Reagents: (a) Boc2O, anhydrous CH2Cl2, rt; (b) iPr2NEt, Phosgene, anhy-
drous CH2Cl2, 0 ꢁC and then rt; (c) iPr2NEt, mono-tert-butoxycarbonyl diaminoalkyl, anhydrous CH2Cl2, rt; (d) iPr2NEt, aminoalcohol, anhydrous
CH2Cl2, rt; (e) iPr2NEt, alkylamine, anhydrous CH2Cl2, rt; (f) diethylphophonoaminomethyl, anhydrous CH2Cl2, rt; (g) iPr2NEt, N-tert-butoxy-
carbonyl aminoalcohol, anhydrous CH2Cl2, rt; (h) TFA, CH2Cl2; (i) Bromotrimethylsilane, anhydrous CH2Cl2, rt/MeOꢀNa+, MeOH, rt.

C. Anastasi et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2459–2463
2461
Figure 3. Nucleophilic intramolecular attack of o-amino group of compounds 23 and 24 on its 50-O-carbonate function leading to the generation of
the 50-O-carbamate (route b) analogues 17 and 18.
3TC (Table 1). The tested 3TC-carbamate drugs showed
anti-HIV activities within EC50 values ranging from 15
to 254 mM in MT-4 cell cultures and from 1 to 10 mM in
PBMCs (Table 1). In contrast anti-HIV activities of
50-O-3TC-carbonate drugs (23 and 24) were very similar
to that of the parent drug 3TC (EC50 values of 0.30 and
0.84 mM in MT-4 cells, 0.01 and 0.1 mM in PBMCs)
(Table 1). This last result, strongly suggests that 50-O-
3TC-carbonate nucleosides act as 3TC prodrugs,
knowing that the carbonate bond is highly sensitive to
enzymatic hydrolysis.
HepG2.2.15 infected cells using a known standard pro-
cedure.11 Most of the tested compounds belonging to
both series exhibited significant in vitro anti-HBV
activities with EC50 values ranging from 1 to 5mM
without significant toxicity up to 200 mM (Table 1).
Interestingly similar anti-HBV activities were observed
for any 50-O-functions (carbamate or carbonate) linking
the nucleoside core to the 50-O side-chain and for any
chemical structure of the side-chain. In contrast to their
anti-HIV activities, 3TC-nucleoside-carbamates, which
were found to be two orders of magnitude less potent
than the corresponding carbonate analogues, demon-
strated potent anti-HBV activities similar to that of
their carbonate analogues or their parent drug 3TC.
Anti-HBV activities and cytotoxicities of both 3TC car-
bamate and carbonate series were tested in the
Table 1. Antiviral potencies of 50-O-carbamate and 50-O-carbonate 3TC analogues
No.
Structure
Y
HIV-1 IIIB, MT-4
HIV-1 LAI, PBMCs
HBV, HEPG2.2.15
a
b
a
X
n
EC50 (mM)
CC50 (mM)
SIc
EC50 (mM)
CC50b (mM)
SIc
EC50d,e (mM)
15
16
17
18
19
20
21
22
23
24
25
3TC
NH
NH
NH
NH
NH
NH
NH
NH
O
NH2
NH2
OH
OH
OH
CH3
CH3
PO(OEt)2
NH2
NH2
3
4
3
4
6
3
2
1
3
4
1
27ꢂ1.4
15ꢂ2.3
>254
>282
>218
>281
>217
>208
>283
>285
>233
>17
>10
>14
n.c.g
n.c.g
>3
>11
>9
>2
>57
>260
n.c.g
>107
1ꢂ0.02
10ꢂ0.01
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>1
>1
n.c.g
>1
>2
>1
>1
>1000
>100
n.c.g
>1000
1ꢂ0.06
1ꢂ0.06
1ꢂ0.06
5ꢂ0.1
>200
>10
75ꢂ30
26ꢂ5.8
32ꢂ6.8
126ꢂ28
0.3ꢂ0.1
0.8ꢂ0.2
n.a.f
10ꢂ0.02
5ꢂ0.03
10ꢂ0.01
10ꢂ0.01
0.01ꢂ0.01
0.1ꢂ0.02
>10
5ꢂ0.1
1ꢂ0.06
5ꢂ0.1
>5 (19%)
0.5ꢂ0.05
1ꢂ0.06
5ꢂ0.1
O
NH
>218
>322
>44
PO(OH)2
0.4ꢂ0.3
0.01ꢂ0.004
0.4ꢂ0.02
aConcentration in mM required to inhibit by 50% the cytopathicity of HIV-1 on MT-4 cells or to inhibit by 50% HIV-1 replication on PBMCs,
followed by standard deviation.
bConcentration in mM required to cause 50% death of uninfected MT-4 cells or PBMCs.
cSelectivity index=CC50/EC50
.
dConcentration in mM required to inhibit 50% of HBV DNA secretion by HEPG2.2.15 stably transfected by HBV, followed by standard deviation.
eNo apparent cytotoxicity was observed below 200 mM for each compound; no SI was given since the CC50 values could not be accurately deter-
mined in the conditions used for the antiviral assay.
fNo activity.
gNot calculable.

2462
C. Anastasi et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2459–2463
Next we investigated the stabilities of these carbamate
nucleosides in HIV and HBV cell culture media, and
also in HepG2.2.15 and MT-4 cell lysates. Enzymatic
hydrolysis half-lives (t1/2) of the drugs conversion into
the parent drug 3TC were determined by a HPLC
method2,3 (not shown). Our results indicated that the
50-O-carbamate-3TC nucleosides, compounds 15 to 22
and 25, were very stable under the different biological
conditions tested. After 2 days of treatment no detect-
able conversion into the parent drug 3TC was detected.
Furthermore, in order to determine the upper stability
limit, selected compound 17 was incubated for 5 days in
the above incubation conditions and release of 3TC was
not observed. The enhanced stabilities of 50-O-carba-
mate-3TC drugs are consistent with already reported
literature data for 50-O-AZT-carbamate drugs.2,5 In
contrast, 50-O-carbonate-3TC drugs (compounds 23 and
24) incubated in the same conditions were found to be
highly sensitive to enzymatic hydrolysis; their half-lives
were found to be between 13 min and 10 h for com-
pound 23, and 5 h and 24 h for compound 24 (not
shown).
warrant further investigations in order to understand
the mechanism of action of 50-O-carbamate nucleosides.
In conclusion, 3TC-carbamate nucleosides are quite
stable compounds in cell lysates or culture cell media, so
that the release or formation of significant levels of free
intracellular 3TC drug may be unlikely. Thus, hydro-
lysis of the carbamoyl moiety would not be a pre-
requisite for anti-HIV or anti-HBV activities. Since the
carbamate analogues did not inhibit directly the HIV-1
RT activity this suggests that RT is not their biological
target. Consequently, these compounds cannot be con-
sidered as classical 3TC prodrugs. Therefore, the anti-
viral activity of these compounds remains unclear.
Nevertheless, 50-O-3TC-carbamates have shown potent
antiviral activities on HIV and particularly on HBV
replication.
Acknowledgements
We gratefully acknowledge Pr. K. Dudley (INSERM
U-382, Universite Mediterranee, Luminy-Marseille,
France) for his English corrections of the manuscript.
This research was supported by grants from LAPHAL
Laboratories and PACA regional Council (C.A.).
INSERM is acknowledged for financial support. The
following reagent was obtained through the AIDS
Research and Reference Reagent Program, NIAID,
NIH: Reverse Transcriptase from the University of
Alabama at Birmingham, Center for AIDS Research,
Gene Expression Core Facility.
When (ꢀ)-20-deoxy-30-thiacytidin-50-yl-O-(3-aminopro-
pyl) carbonate 23 was incubated in biological media,
formation of both the parent drug (3TC) and (ꢀ)-20-
deoxy-30-thiacytidin-50-yl-O-(hydroxypropyl) carbamate
17 was clearly identified by HPLC in the enzymatic
extract, by comparison with the HPLC profiles of com-
pounds 3TC and 17 (not shown). Formation of this
later compound resulted from nucleophilic intramole-
cular attack of the o-amino group on the 50-O-carbon-
ate function within the side-chain of compound 23 (Fig.
3). This result is consistent with the two possible
hydrolytic cleavages (routes a or b) of carbonate pro-
drugs leading to the release of 3TC or to the generation
of the carbamate function. Such rearrangements of o-
aminoalkyl carbonate into o-hydroxyalkyl carbamate
have only been reported in the case of AZT2,3 and more
recently in the case of HIV protease inhibitors.12
References and Notes
1. Aggarwal, S. R.; Gogu, S. R.; Rangan, S. R.; Agrawal,
K. C. J. Med. Chem. 1990, 33, 1505.
2. Vlieghe, P.; Bihel, F.; Clerc, T.; Pannecouque, C.; Witv-
rouw, M.; De Clercq, E.; Salles, J.-P.; Chermann, J.-C.;
Kraus, J.-L. J. Med. Chem. 2001, 44, 777.
3. Vlieghe, P.; Clerc, T.; Pannecouque, C.; Witvrouw, M.; De
Clercq, E.; Salles, J.-P.; Chermann, J.-C.; Kraus, J.-L. J. Med.
Chem. 2001, 44, 3014.
4. Perez-Perez, M. J.; Balzarini, J.; De Clercq, E.; Camarasa,
M. J. Bioorg. Med. Chem. 1993, 1, 279.
5. Chang, S.-L.; Griesgraber, G.; Abraham, T. W.; Garg, T.;
Song, H.; Zimmerman, C. L.; Wagner, C. R. Nucleosides,
Nucleotides and Nucleic Acids 2000, 19, 87.
6. Fisher, M. J.; Giese, U.; Harms, C. S.; Kinnick, M. D.;
Lindstrom, T. D.; McCowan, J. R.; Mest, H.-J.; Morin, J. M.,
Jr.; Mullaney, J. T.; Paal, M.; Rapp, A.; Ruhter, G.; Ruter-
bories, K. J.; Sall, D. J.; Scarborough, R. M.; Schotten, T.;
Stenzel, W.; Towner, R. D.; Um, S. L.; Utterback, B. G.;
Wyss, V. L.; Jakubowski, J. A. Bioorg. Med. Chem. Lett. 2000,
10, 385.
7. Martin, S. F.; Follows, B. C.; Hergenrother, P. J.; Franklin,
C. L. J. Org. Chem. 2000, 65, 4509.
8. Greenberg, M. M.; Matray, T. J.; Kahl, J. D.; Yoo, D. J.;
Mc Minn, D. L. J. Org. Chem. 1998, 63, 4062.
9. Pauwels, R.; Balzarini, J.; Baba, M.; Snoeck, R.; Schols,
D.; Herdewjin, P.; Desmyter, J.; De Clercq, E. J. Virol. Meth-
ods 1988, 20, 309.
10. 100 000 PBMCs were pre-treated for 30min with mole-
cules and then infected with 75TCID50 of HIV-1-LAI strain. 7
days after infection supernatants were collected and HIV
Results from our antiviral activities and stability studies
on o-functionalised alkyl-50-O-3TC-carbamate drugs
suggest that their antiviral activity mechanism could be
different from that of 3TC (RT nucleoside inhibitor).
However it must be underlined that the stability of these
compounds in cell culture media or cell extracts can be
different from that in intact cells, most nucleoside pro-
drugs being cleaved in these biological test conditions.
Another hypothesis could be that these compounds are
transported across the cellular membrane in their intact
form, and that their antiviral potencies may not be
associated with the release of free 3TC. Such hypothesis
has been already suggested by Wagner et al.13 in the
case of phosphonoester amidates of nucleoside pro-
drugs. The accumulation of the unmodified nucleoside
drug accumulated inside the cell might be at the origin
of the observed antiviral effect. To this end we investi-
gated the capacity of 3TC-carbamate drugs to inhibit
exogenous RT, using standard HIV RT assays.5,14 We
showed that 3TC-carbamate drugs do not directly inhi-
bit HIV-1 RT at concentrations as high as 5 mM (not
shown). Taken together, the reported observations

C. Anastasi et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2459–2463
2463
replication was measured by dosage of RT activity in cell cul-
ture supernatant.
American Chemical Society, Orlando, FL, April 7–11,
2002.
11. Chu, C.; Ma, T.; Shanmuganathan, K.; Wang, C.; Xiang,
Y.; Pai, S.; Yao, G.; Sommadossi, J. P.; Cheng, Y. Antimicrob.
Agents Chemother. 1995, 39, 979.
12. Kazmierski, W., Bevans, P., Furfine, E., Porter, D.,
Spaltenstein, A., Yang, H., 223rd National Meeting of the
13. Mc Intee, E.; Remmel, R.; Schinazi, R.; Abraham, T.;
Wagner, C. J. Med. Chem. 1997, 40, 3323.
14. Von Janta-Lipinsky, M.; Costisella, B.; Ochs, H.; Hub-
scher, U.; Hafkemeyer, P.; Matthes, E. J. Med. Chem. 1998,
41, 2040.