Hydroxy fatty acids for the delivery of dideoxynucleosides as anti-HIV agents

Article abstract of DOI:10.1016/j.bmcl.2013.12.093

A series of α- and β-carboxylated phospholipid prodrugs of dideoxy nucleosides have been synthesized and evaluated against HIV. An increase in biological effect with a factor of 500 has only been observed for the adenine nucleoside, which suggests that this prodrug approach is base specific.

Full text of DOI:10.1016/j.bmcl.2013.12.093

Bioorganic & Medicinal Chemistry Letters 24 (2014) 817–820
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Hydroxy fatty acids for the delivery of dideoxynucleosides
as anti-HIV agents
Kishore Lingam Gangadhara a, Eveline Lescrinier a, Christophe Pannecouque b, Piet Herdewijn a,
⇑
a Medicinal Chemistry, Department of Pharmaceutical Sciences, Rega Institute for Medical Research, KU Leuven, Minderbroedersstraat-10, 3000 Leuven, Belgium
b Laboratory of Virology and Chemotherapy, Department of Microbiology and Immunology, Rega Institute for Medical Research, KU Leuven, Kapucijnenvoer 33, 3000 Leuven, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of a- and b-carboxylated phospholipid prodrugs of dideoxy nucleosides have been synthesized
Received 19 November 2013
Revised 20 December 2013
Accepted 22 December 2013
Available online 29 December 2013
and evaluated against HIV. An increase in biological effect with a factor of 500 has only been observed
for the adenine nucleoside, which suggests that this prodrug approach is base specific.
Ó 2014 Elsevier Ltd. All rights reserved.
Keywords:
Prodrugs
Nucleotides
Phospholipids
d4A
HIV
20,30-Dideoxynucleosides and 20,30-didehydro-20,30-dideoxynu-
cleosides are nucleoside reverse transcriptase inhibitors (NRTIs).
Some of them have been found effective for the treatment of HIV
infections. These NRTIs1 inhibit the viral replication by terminating
DNA polymerization. One of the efforts to improve the therapeutic
potential of nucleoside analogues for treating viral infections is di-
rected towards prodrug development.
Prodrug approaches have been developed for increasing oral
absorption and cellular penetration of modified nucleos(t)ides. In
addition, prodrugs are also used for tissue specific delivery of
nucleotides, and for kinase bypass strategies. These prodrug
strategies have the aim to mask the anionic phosphate moiety of
nucleotides, and to deliver the modified nucleos(t)ides (in blood
or in cells) involving enzymatic degradation reactions. One of the
approaches that was followed in the past is synthesis of
phospholipid prodrugs of modified nucleosides. Hostetler et al.
has demonstrated that the acyclovir diphosphate dimeristoylglyc-
erol prodrug delivers acyclovir monophosphate inside cells.2 Acyl
nucleotide3 derivatives of d4T and AZT, as well as O-alkyl-5’,50-
dinucleoside phosphates of AZT/cordycepin4 have been synthe-
sized. Generally, the anti-HIV efficacy of the prodrug was lower3,4
or similar3 than that of the parent nucleotide. Many more phos-
phodiester prodrugs of modified nucleosides have been studied5
but generally it is believed that mono alkyl/aryl phosphodiesters
are unsuitable for the intracellular delivery of nucleotides.6 In the
case of d4A and d4C, 50-methyl and 500-phenyl phosphate diesters
shows in vitro anti-HIV activity comparable to those of the parent
nucleoside.7 In the present work we report the synthesis and anti-
viral activity of phospholipid conjugates of d4A, d4T, ddC (Fig. 1).
Two types of lipids have been considered that is ( )-
a-hydroxy
stearic acid and ( )-b-hydroxy stearic acid. It is hypothesized that
the lipid moiety may help cell penetration while the a-carboxylic
acid function may assist in phosphodiester bond cleavage.8 Depen-
dent on which ester bond is cleaved (C50–O–P or CH–O–P), either
the nucleoside or the nucleotide can be delivered in the cell. Addi-
tional factors which may complicate the predictability of the
results are that the stability of 50-phosphodiester conjugates of
O
N
NH2
NH2
N
N
N
NH
N
O
N
O
N
O
HO
HO
HO
O
O
nucleoside with an
a-carboxylate group could be base dependent
2.d4A
Figure 1. Structure of anti-HIV nucleosides.
1.d4T
3.ddC
and that the role of phosphodiesterase enzymes in the cleavage
of the lipid conjugates is unknown. Intramolecular catalysis of
P–O or P–N bond cleavage (to liberate the nucleoside or nucleotide
by chemical degradation of the prodrug) occurs via a 5-membered
⇑
Corresponding author. Tel.: +32 16 337387; fax: +32 16 337340.
E-mail address: piet.herdewijn@rega.kuleuven.be (P. Herdewijn).
(a-COOH) or a 6-membered (b-COOH) intermediate, and both
0960-894X/$ - see front matter Ó 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.12.093

818
K. L. Gangadhara et al. / Bioorg. Med. Chem. Lett. 24 (2014) 817–820
possibilities have been investigated. A prodrug of AZT was not in-
cluded in this study. To evaluate the potential of a new prodrug
concept, results are more pronounced when using a lower active
or almost inactive nucleoside as model compound, rather than a
very potent congener.
O
OH
7
a,
b
Stavudine (d4T) 1 is prepared by a two-step method starting
from thymidine as previously reported.9 The synthesis of 20,30-
didehydro-20,30-dideoxyadenosine 2 has been accomplished as
previously described,10,11 by converting the vicinal diol of adeno-
sine to olefin using Corey–Winter reaction conditions. The nucleo-
OH
O
O
8
side ddC 3 is commercially available. The methyl ( )-a-hydroxy
Scheme 2. Reagents and conditions: (a) CDI, MgCl2, K+ ꢀO2CCH2CO2CH3, THF, rt
overnight; (b) NaBH4, EtOH, 0 °C to rt 15 min.
stearate 5 is prepared in three steps from stearic acid 4 by bromin-
ation12 using Hell–Volhard–Zelinsky conditions followed by
hydrolysis13 and esterification14 in 88% overall yield (Scheme 1).
The methyl ( )-b-hydroxy stearate 8 was synthesized by the
procedure described by Masamune15 which involves homologation
of palmitic acid 7 using in situ generated magnesium monometh-
ylmalonate16 to a preformed acyl imidazole to produce the b-keto
stearic acid methyl ester, which is reduced17 with sodium borohy-
dride in ethanol providing 818 in 77% overall yield (Scheme 2).
The phosphoramidite approach was used for the synthesis of the
lipid-nucleotide conjugates. In the case of thymine nucleoside, the
50-O-phosphoramidite of d4T (9)19 was coupled with the hydroxyl-
ated fatty acid esters 5 and 8 yielding the phosphite intermediates
10 and 11 which were subsequently oxidized using aqueous iodine
in Pyridine/THF resulting in the formation of 12 and 13. (Scheme 3).
In the case of cytosine and adenine nucleotides, the phospho-
N
N
O
N
O
N
R2
NH
NH
a
O
P
O
P
O
O
R1
O
N
O
O
O
O
9
10: R1 = CH3 (CH2)15-, R2= -COOCH3
11: R1= CH3 (CH2)14-, R2= -(CH2)- COOCH3
b
ramidite derivative of the ( )-a-hydroxy stearic methyl ester 6
N
was first synthesized and coupled with the nucleoside analogue.
The phosphoramidite 6 was obtained as a diastereomeric mixture
in 56% yield. The reverse approach (as used for the synthesis of
the thymine analogues) is lower yielding.
O
R2
NH
O
P
O
N
O
R1
O
O
O
The d4A and ddC conjugates were obtained by the reaction of 6
with d4A and ddC, followed by oxidation of the intermediates 14
and 16 to produce 15 and 17 respectively (Scheme 4). Deprotection
of the phosphotriesters 12, 13, 15 and 17 was first carried out by re-
moval of the cyanoethyl protecting group with DBU in THF, followed
by hydrolysis of the methyl ester with 1 N sodium hydroxide in
MeOH. However, the yields were moderate and several side prod-
ucts were formed. Therefore we turned to a one-step deprotection
procedure using an aqueous ammonia/MeOH mixture. (Scheme 5).
The lipid-nucleotide conjugates 18–21 were purified on a Dow-
ex-50 Na+ column followed by lyophilization. Biological activity
was determined on the diastereomeric mixtures.
12:R1 = CH3 (CH2)15-, R2= -COOCH3
13:R1= CH3 (CH2)14-, R2= -(CH2)-COOCH3
Scheme 3. Reagents and conditions: (a) 5 for 10 and 8 for 11, 1H-tetrazole, dry
DCM, rt 4 h; (b) aq iodine in pyridine/THF, rt 30 min.
The activity of the phosphodiester analogues of d4A-20, ddC-21,
d4T-18, 19 against HIV-1 and HIV-2 was determined including the
dideoxy nucleosides d4A, ddC, d4T as reference compounds. The
anti-HIV activity and cytotoxicity of the compounds were evalu-
ated against wild-type HIV-1 and HIV-2 in MT-4 cell cultures
(Table 1). Anti-HIV-2 activity was also determined in CEM cells
as well as in thymidine kinase-deficient CEM (CEM/TK-) cell cul-
O
tures. The
a-hydroxy D4T derivative 18 has an IC50 of 0.4 lM
OH
against HIV-1 and HIV-2 in MT-4 cells, but was not active in CEM
cells. Compound 19 (the b-hydroxy d4T conjugate) is about 30
4
times less active than the a-hydroxy d4T analogue. However both
a,b,c
compound 18 and 19 are less active than the parent nucleoside
(d4T). Similar results are obtained for the ddC lipid conjugate 21
that is the parent nucleoside (ddC) is more active than the poten-
tial prodrug 21, as well in MT-4 cells as in CEM/0 cells.
O
O
OH
5
Between the three a-hydroxy nucleoside conjugates tested, d4A
takes a particular place. The parent d4A itself is an anti-HIV nucleo-
d
side with very low activity as well in MT-4 cells as in CEM/0 cells.
Conjugation of d4A-MP with a-hydroxy stearic acid leads to an in-
O
O
crease in activity of at least a factor 500 in MT-4 cells (Table 1).
The activity is similar against HIV-1 and against HIV-2. Previously,
a methyl phosphate diester was reported as a potential prodrug of
d4A,7 however, with no increase in antiviral potency. The observa-
tion that the activity of 20 in CEM/K- cells is lower than in CEM/0
cells suggests that the prodrug of d4A does not deliver d4A-MP in
the cells.
O
OCE
P
N
6
Scheme 1. Reagents and conditions: (a) PBr3, Br2, 95 °C, 6 h; (b) aq NaOH, 85 °C,
3 h; (c) TMSCl, DMP, MeOH, rt overnight; (d) 2-cyanoethyl N,N-dii-
sopropylchlorophosphoramidite, dry DCM, 1H-tetrazole, 0 °C to rt 15 min.

K. L. Gangadhara et al. / Bioorg. Med. Chem. Lett. 24 (2014) 817–820
819
R2
NH2
N
N
N
O
O
O
R1
O
O
B
O
P
O
b
O
N
P
O
O
O
a
R2
14
15
N
R1
O
P
N
N
14:
R
= CH (CH
3
)
-, R = -COOCH B=adenine-9-yl
1
2 15
2
3,
O
R2
NH2
c
N
O
O
O
R1
B
O
P
O
O
b
N
O
6
O
P
O
N
O
O
O
16
17
N
N
16
:
R
1
= CH (CH ) -, R = -COOCH B=cytosin-1-yl
3
2 15 2 3,
Scheme 4. Reagents and conditions: (a) 2, 1H-tetrazole, dry DCM, rt 4 h; (b) aq iodine in pyridine/THF, rt 30 min; (c) 3, 1H-tetrazole, dry DCM, rt 4 h.
R2
R2
O
O
R1
B
O
P
O
O
a,b
R1
O
B
O
P
O
O
O
Na
Y
X
X
Y
N
12, 13, 15,17
18, 19, 20, 21
18: R1 = CH3 (CH2)15-, R2 = -COONa+, X-Y = CH=CH, B=thymin-1-yl
19
12: R1 = CH3 (CH2)15-, R2 = -COOCH3, X-Y = CH=CH, B=thymin-1-yl
: R1 = CH3 (CH2)14-, R2 = -(CH2)-COONa+, X-Y = CH=CH, B=thym-1-yl
20: R1 = CH3 (CH2)15-, R2 = -COONa+, X-Y = CH=CH, B=adenine-9-yl
21
13: R1 = CH3 (CH2)14-, R2 = -(CH2)-COOCH3, X-Y = CH=CH, B=thymin-1-yl
15
17
: R1 = CH3 (CH2)15-, R2 = -COOCH3, X-Y = CH=CH, B=adenin-9-yl
: R1 = CH3 (CH2)15-, R2 = -COOCH3, X-Y = CH2-CH2, B=cytocin-1-yl
: R1 = CH3 (CH2)15-, R2 = -COONa+, X-Y = CH2-CH2; B=cytosin-1-yl
Scheme 5. Reagents and conditions: (a) aq Ammonia/MeOH (1:0.5), rt 36 h; (b) Ion exchange, Na+ form.
Table 1
Antiviral activity test of compounds against HIV-1 and HIV-2
Analogue
MT-4 cellsa
IC50 M)HIV-2
CEM/0b
CEM/TK-c
IC50 M)HIV-2
IC50
(
lM)HIV-1
(l
CC50
(l
M)
IC50
(lM)HIV-2
CC50
(lM)
(l
18
19
20
21
d4T
d4A
ddC
AZT
0.4 0.1
0.4 0.1
64.7 8.7
>198.2
>39.6
>198.2
92.6 9.5
>198.2
81.3 8.7
91.1 8.3
>267.6
360.3 63.7
178.8 33.0
>467.7
>39.6
>39.6
>39.1
>40.5
P119.7
P81.9
1.9 0.8
>14.9
10.9 1.0
0.19 0.02
9.7 2.5
0.26 0.03
>277.5
16.2 0.2
0.11 0.01
5.7 0.1
0.4 0.1
>277.5
97.8 5.5
94.1 8.8
327.4 32.2
277.5 19.2
324.8 29.4
>93.5
25.3 12.7
7.5 2.5
8.8 6.4
P36.6
0.5 0.3
0.04 0.01
1.0 0.6
0.007 0.002
1.9 0.3
0.007 0.002
a
b
c
HIV-1 (IIIB), HIV-2 (ROD).
CEM/0 cells (ROD).
CEM thymidine kinase deficient cells (ROD).
In conclusion, hydroxy fatty acid conjugates of d4T-MP, d4A-MP
of compounds 18, 19 and 20. All compounds are stable in neutral
and basic (pH12) medium. Degradation of the -COOH compounds
and ddC-MP have been prepared as prodrugs for their potential
ability to deliver the antiviral nucleosides and/or nucleotides into
the cell. Only in the case of d4A, increased antiviral activity against
HIV-1 and HIV-2 has been observed. It therefore seems that this
prodrug approach is base specific. The inactivity of the d4A conju-
gate in CEM/TKꢀ cells suggests that the modified nucleosides
rather than the nucleotide is delivered in the infected cells. We
a
occurs in acidic circumstances, liberating dA or dT and the phos-
phorylated lipid within a few days at pH = 6 and within minutes
at pH = 4. These results suggest that chemical degradation might
not be the reason for delivering d4A inside cells, but this is not con-
clusive. The study in acidic circumstances has been performed in
DMSO (as all compounds from gels in aqueous acetic medium)
which questions the (biological) relevance of this experiment. It
could therefore be possible that 20 is more easily enzymatically
degraded in cells than 18, 19 or 21.
previously observed that the chemical stability of
a-carboxylated
phosphoramidates of nucleosides is base dependent.8 Some initial
stability studies have been performed on the dA and dT analogues

820
K. L. Gangadhara et al. / Bioorg. Med. Chem. Lett. 24 (2014) 817–820
4. Meier, C.; Neumann, J. M.; Andre, F.; Henin, Y.; Huynh Dinh, T. J. Org. Chem.
Acknowledgments
1992, 57, 7300.
5. Wagner, C. R.; Iyer, V. V.; McIntee, E. J. Med. Res. Rev. 2000, 20, 417.
6. Krise, J. P.; Stella, V. J. Adv. Drug Deliv. Rev. 1996, 19, 287.
7. Mullah, K. B.; Rao, T. S.; Balzarini, J.; De Clercq, E.; Bentrude, W. G. J. Med. Chem.
1992, 35, 2728.
8. Maiti, M.; Michielssens, S.; Dyubankova, N.; Maiti, M.; Lescrinier, E.;
Ceulemans, A.; Herdewijn, P. Chem. Eur. J. 2012, 18, 857.
9. Horwitz, J. R.; Chua, J.; Da Rooge, M. A.; Noel, M.; Klundt, I. L. J. Org. Chem. 1966,
31, 205.
Authors are indebted to KU Leuven (GOA 10/013) and FWO for
financial support. We thank J. Rozenski for measuring the MS and
C. Biernaux for editorial help.
Supplementary data
10. Dudycz, L. W. Nucleosides Nucleotides 1989, 8, 35.
11. Chu, C. K.; Bhadti, V. S.; Doboszewski, B.; Gu, Z. P.; Kosugi, Y.; Pullaiah, K. C.;
Vanroey, P. J. Org. Chem. 1989, 54, 2217.
12. Polyakova, S. M.; Belov, V. N.; Yan, S. F.; Eggeling, C.; Ringemann, C.;
Schwarzmann, G.; de Meijere, A.; Hell, S. W. Eur. J. Org. Chem. 2009, 5162.
13. Masuda, Y.; Mori, K. Eur. J. Org. Chem. 2005, 478.
14. Horneff, T.; Herdtweck, E.; Randoll, S.; Bach, T. Bioorg. Med. Chem. 2006, 14,
6223–6234.
15. Brooks, D. W.; Lu, L. D. L.; Masamune, S. Angew. Chem., Int. Ed. Engl. 1979, 18, 72.
16. Ratovelomanana-Vidal, V.; Girard, C.; Touati, R.; Tranchier, J. P.; Hassine, B. B.;
Genêt, J. P. Adv. Synth. Catal. 2003, 345, 261.
17. Yang, D.; Gao, Q.; Zheng, B.-F.; Zhu, N.-Y. J. Org. Chem. 2004, 69, 8821.
18. Bergström, S.; Aulin-Erdtman, G.; Rolander, B.; Stenhagen, E.; Östling, S. Acta
Chem. Scand. 1952, 6, 1157.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2013.12.
093.
References and notes
1. Martin, J. C.; Hitchcock, M. J.; De Clercq, E.; Prusoff, W. H. Antiviral Res. 2010, 85,
34.
2. Hostetler, K. Y.; Parker, S.; Sridhar, C. N.; Martin, M. J.; Li, J. L.; Stuhmiller, L. M.;
van Wijk, G. M.; van den Bosch, H.; Gardner, M. F.; Aldern, K. A. Proc. Natl. Acad.
Sci. U.S.A. 1993, 90, 11835.
3. Bonnaffé, D.; Dupraz, B.; Ughetto-Monfrin, J.; Namane, A.; Henin, Y.; Huynh
Dinh, T. J. Org. Chem. 1996, 61, 895.
19. Abraham, T. W.; Wagner, C. R. Nucleosides Nucleotides 1994, 13, 1891.