CycloSal-PMEA and cycloAmb-PMEA: Potentially new phosphonate prodrugs based on the cyclosal-pronucleotide approach

Article abstract of DOI:10.1021/jm050641a

Two new classes of lipophilic prodrugs of the antiviral active phosphonate 9-[2-phosphonomethoxyethyl]adenine (PMEA 1) have been prepared and were studied with regard to their hydrolysis properties and biological activity. A first series of compounds was

Full text of DOI:10.1021/jm050641a

J. Med. Chem. 2005, 48, 8079-8086
8079
cycloSal-PMEA and cycloAmb-PMEA: Potentially New Phosphonate Prodrugs
Based on the cycloSal-Pronucleotide Approach
Chris Meier,*,# Ulf Go¨rbig,# Christian Mu¨ller,# and Jan Balzarini†
Institute of Organic Chemistry, University of Hamburg, Martin-Luther-King-Platz 6, D-20146 Hamburg, Germany, and
Rega Institute for Medical Research, K.U. Leuven, Minderbroedersstraat 10, B-3000 Leuven, Belgium
Received July 5, 2005
Two new classes of lipophilic prodrugs of the antiviral active phosphonate 9-[2-phospho-
nomethoxyethyl]adenine (PMEA 1) have been prepared and were studied with regard to their
hydrolysis properties and biological activity. A first series of compounds was prepared on the
basis of the cycloSal nucleotide approach. Because of the surprisingly low hydrolysis stability
of these cycloSal-PMEA derivatives, more stable derivatives have to be developed. Instead of
using salicyl alcohol, in cycloAmb-PMEA derivatives 2-aminobenzyl alcohols were attached to
PMEA 1. The latter compounds showed a considerably higher stability compared to the cycloSal
counterparts. Stability studies revealed that all lipophilic prodrugs delivered PMEA selectively
by chemical means. All compounds proved to be noninhibiting to acetyl- and butyrylcholinest-
erase, and some of the phosphonate diesters were found to be more active against HIV compared
to the parent PMEA.
Introduction
The acyclic nucleoside phosphonates 9-[2-phospho-
nomethoxyethyl]adenine (PMEA; Adefovir 1; Figure 1),
(R)-9-[2-phosphonomethoxypropyl]adenine (PMPA; Teno-
fovir 2), and (S)-9-[3-hydroxy-2-phosphonomethoxypro-
pyl]cytosine (HPMPC; Cidofovir 3) show broad antiviral
activity against DNA and retroviruses not only in vitro
Figure 1. Structural formulas of the acyclic nucleoside
but also in-vivo.1 For example, PMEA 1 has demon-
phosphonate analogues PMEA, PMPA, and HPMPC.
strated antiviral activity against human immunode-
ficiency virus (HIV), Rauscher murine sarcoma virus
for example, d4TMP intracellularly, and thus improve
(R-MuLV), herpes simplex virus (HSV), cytomegalovirus
the biological activity.6 The most remarkable advantage
(CMV), simian immunodeficiency virus (SIV), feline
of the cycloSal nucleotides is the use of only one masking
immunodeficiency virus (FIV), Epstein-Barr virus (EBV),
unit (a salicyl alcohol) per nucleotide and the purely pH-
and hepatitis B virus (HBV).
driven delivery of the nucleotides. Recently, we attached
However, the therapeutic use of these nucleoside
esterase-cleavable side chains to the aromatic ring of
phosphonates is limited due to their poor oral bioavail-
the cycloSal moiety to trap the cycloSal derivatives
ability caused by the negatively charged phosphonate
intracellularly.8
moiety that is present at physiological pH. PMEA was
Because of the interesting properties of nucleoside
reported to have <1% oral bioavailability in monkeys
phosphonates, the cycloSal approach was applied to
and <11% in rats.2 Therefore, neutral and membrane-
PMEA. Here, we report on the synthesis of two new
permeable prodrugs were synthesized. Bis(POM)-PMEA,3
classes of PMEA prodrugs (cycloSal-PMEA 5 and
bis(DTE)-PMEA, and bis(SATE)-PMEA,4 for example,
cycloAmb-PMEA 6, Figure 2), their hydrolysis proper-
can penetrate cellular membranes and deliver the
ties, and their biological evaluation.
parent PMEA by enzymatic activation. In the last
several years, we have successfully developed the
cycloSal approach for the delivery of nucleoside mono-
Results and Discussion
Chemistry. In contrast to prototype cycloSal-phos-
phate triesters of type 4, the synthesis of the phospho-
nate cannot be achieved by our previously developed
P(III)-route using cycloSal-chlorophosphites prepared
from salicyl alcohols 7 and PCl3. This led to the
formation of cycloSal-phosphites that were subsequently
oxidized by t-BuOOH.5-8 For the phosphonates, diethyl
ester of PMEA 8 was used as starting material. First,
the exocyclic amino group was blocked with the mono-
methoxytrityl group in 89% yield.9 Treatment of com-
pound 9 with trimethylsilyl bromide and pyridine in
CH3CN gave the intermediate bis(trimethylsilyl) ester
that was immediately converted into the corresponding
phosphates (pronucleotides). The structure of the pro-
totype cycloSal compounds 4 is shown in Figure 2. The
cycloSal approach has already been applied successfully
to several nucleoside analogues,5 for example, the anti-
HIV 2′,3′-dideoxy-2′,3′-didehydrothymidine (d4T)6 and
the anti-herpes-simplex virus 5-[(E)-2-bromovinyl]-2’-
deoxyuridine (BVDU).7 We have proven that cycloSal
triesters are able to penetrate cell membranes, release,
* To whom correspondence should be addressed. Tel: +49-40-
42838-4324. Fax:
+49-40-42838-2495. E-mail:
chris.meier@
chemie.uni-hamburg.de.
# University of Hamburg.
† Rega Institute for Medical Research.
10.1021/jm050641a CCC: $30.25 © 2005 American Chemical Society
Published on Web 11/10/2005

8080 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 25
Meier et al.
Scheme 1. Synthetic Pathways toward cycloSal-PMEAs
5 (a) and cycloAmb-PMEAs 6 (b)a
Figure 2. Structural formulas of the cycloSal-phosphate
triesters 4 as well as the cycloSal- and cycloAmb-PMEA
derivatives 5 and 6.
dichloride10 that was reacted with four different salicyl
alcohols 7 in the presence of triethylamine in CH2Cl2
to give cycloSal-PMEA diesters 10a-d in 31-48% yield
after purification on silica gel chromatography. Finally,
the MMTr group was cleaved by treatment of 10 with
TFA in CH3CN, which led to the target cycloSal-PMEA
derivatives 5a-d in 53-82% yield (Scheme 1a). The
different substitution pattern in the cycloSal moiety has
been selected due to their influence on the hydrolysis
properties of the cycloSal system as reported before.5
Attempts to prepare the cycloSal-PMEAs without pro-
tection of the exocylic amino group led to a complex
mixture of reaction products. Moreover, different modi-
fications for the above reaction sequence were tried. We
tried to attach the t-Boc group instead of the MMTr
group of the exocyclic amino function. However, the
reaction of diethyl PMEA 7 with Boc anhydride or the
active ester “Boc-on” in the presence of different bases
and solvents failed. In contrast, the introduction of the
fluorenylmethoxycarbonyl (FMOC) group could be
achieved partly. Reaction of diethyl PMEA 8 with
FMOC-Cl in pyridine gave a mixture of the N,N-di-
FMOC and the N-FMOC derivative in 78 and 18%,
respectively. With both compounds the above-mentioned
reaction sequence (TMS-Br; oxalyl chloride or PCl5;
salicyl alcohol) led to the formation of the N-FMOC
protected intermediate analogous to diester 10. How-
ever, the yields were found to be only 5-12%. The
cleavage of the FMOC groups was only achieved by
addition of NEt3 in pyridine in poor 20% yield. Other
cleavage conditions such as DBU/CH3CN, NEt3/CH3CN,
piperidine/THF, morpholine/DMF, TBAF/THF, or NEt3‚
HF/THF did not led to the deprotected cycloSal-PMEAs
5 most presumably due to their low hydrolytic stability
(see hydrolysis section). Besides the use of the MMTr
group, the more stable trityl group also was introduced.
The analogous reaction sequence to the cycloSal-PMEAs
gave the diesters in 9% yield. In this reaction, oxalyl
chloride was used for the chlorination reaction instead
of PCl5. However, oxalyl chloride caused more acidic
reaction conditions so that in addition to the tritylated
cycloSal-PMEA diester 10 the cleavage of the trityl
a Reaction conditions: (a) MMTrCl, NEt3, DMAP, CH2Cl2, rt, 4
h; (b) 1. TMSBr, pyridine, CH3CN, rt, 16 h; 2. PCl5, CH2Cl2, rt,
2.5 h; 3. Salicyl alcohol 7a-d, NEt3, CH2Cl2, 0 °C, 1 h; (c) TFA,
CH3CN, rt, 1 h; (d) 1. (COCl)2, N,N-diethylformamide, CH2Cl2; 2.
pyridine, CH2Cl2; 3. 2-aminobenzyl alcohol 11a-c, DIPEA, CH2Cl2,
CH3CN, rt, 1 h.
group was detected by TLC. As expected, this effect was
more pronounced when the MMTr group was used
instead so that PMEA diesters could not be isolated.
In addition to the use of the dichlorides of protected
PMEA for the esterification, it has been published that
esterification of PMEA 9 salts with strong activation
agents is also a suitable route. Thus, N-MMTr-diethyl
PMEA 9 was treated with TMSBr in CH3CN and then
quenched by addition of aqueous triethylammonium
bicarbonate buffer. The formed bis(triethylammonium)
salt of N-MMTr-PMEA was subsequently reacted with
the salicyl alcohols 7 in the presence of mesitylenesulfo-
nyl-3-nitrotriazole (MSNT) in pyridine. After purifica-
tion, the target N-MMTr-cycloSal-PMEA diesters 10
was isolated in 23% yield. Finally, the yields obtained
by this route are markedly lower as compared to those
obtained starting from PMEA dichloride. Moreover, the
purification was tedious.
In addition, a second group of lipophilic PMEA
prodrugs has been prepared. By using 2-aminobenzyl
alcohols 11 instead of salicyl alcohols, the phenol oxygen
atom in the cycloSal moiety is replaced by a nitrogen
atom. Thus, so-called cycloAmb(cycloaminobenzyl)-
PMEAs 6a-c were prepared. In this case, PMEA 1 was
used as starting material. First, PMEA was converted
into the dichloride by the reaction with oxalyl chloride,
N,N-diethylformamide (DEF) in CH2Cl2. Under these
reaction conditions, the exocyclic amino group is tem-
porarily blocked as a formamidine group.11 Then, 2-ami-

cycloSal-PMEA and cycloAmb-PMEA
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 25 8081
Table 1. Hydrolysis, BChE- and Anti-HIV Data of cycloSal- and cycloAmb-PMEA Derivatives
biological activity in CEM/O cells
EC50c [µM]
chem stabilitya
comp
X
pH 7.3 [h]
pH 2.0 [h]
BChEb IC50 [µM]
HIV-1
HIV-2
CC50d [µM]
5a
H
0.09
0.56
4.05
3.05
1.3
n.d.e
n.d.e
23.7
9.6
>50
>50
>50
>50
>50
>50
>50
n.d.e
n.d.e
n.d.e
5b
3-Me
3-t-Bu
3.5-t-Bu
6-F
n.d.e
n.d.e
n.d.e
5c
3.0 ( 1.4
5.5 ( 0.7
28 ( 0.0
20 (7.1
29 ( 2.9
0.18 ( 0.11
10 ( 6.4
4.5 ( 0.7
5.3 ( 0.5
26 ( 2.1
21 ( 9.2
56 ( 43.4
0.075 ( 0.007
10
29 ( 6.5
17 ( 0.6
244 ( 8.5
183 ( 24.0
>250
0.74 ( 0.11
50 ( 13
5d
6a
n.d.e
9.6
6b
H
3-Me
4.0
6c
21.3
14.7
29.0
15f
n.d.e
PMEA 1
a Hydrolysis in 25 mM phosphate buffer at 37 °C or 25 mM citrate buffer, pH 2.0 at 37 °C; half-lives (t1/2) were determined from the
decreasing peak of the starting phosphate triester and are the mean of duplicate experiments. b Inhibition of human butyrylcholinesterase.
c Anti-HIV-1 activity (EC50): 50% effective concentration in wild-type CEM cell cultures. d CC50: cytotoxic concentration. e Not determined
due to low chemical stability. f Bis(POM)PMEA.
Scheme 2. Hydrolysis Mechanism of the cycloSal- and
cycloAmb-PMEA Derivatives
nobenzyl alcohols 11a-c were used for the preparation
of the target cycloAmb-PMEA derivatives 6a-c that
were isolated in 25-42% yield. Additionally, a reaction
sequence starting from diethyl PMEA 8 as described for
the cycloSal-PMEAs 5 was tested. However, although
the target compounds 6 can be isolated, the yields were
very poor (3-7%). Exchange of the base (DIPEA instead
of NEt3) and the solvent (CH2Cl2 to CH3CN) did not
improve the yields. The reaction sequence is sum-
marized in Scheme 1b.
The enantiomeric mixtures of the PMEA esters 5 and
6 were characterized by means of 1H, 13C, and 31P NMR
spectroscopy, by analytical reversed-phase high-perfor-
mance liquid chromatography (RP-HPLC), and by FAB
or ESI mass spectrometry and UV spectroscopy. After
lyophilization, phosphonate diesters 5 and 6 were
obtained as white fluffy solids.
The lipophilicity (logP) of the new phosphonate de-
rivatives can be determined experimentally by parti-
tioning the compound between n-octanol/water or can
be evaluated qualitatively by their retention time on the
HPLC. Balzarini et al. showed some years ago that the
retention time on a HPLC column using the same
gradient can be correlated with the logP value of a
compound.12 Here, some of the compounds are too labile
to be used in this experimental procedure (see below).
Therefore, a combination of both methods was used. The
retention times and the determined logP values of
cycloSal-phosphonates 5 were 9.4 min (5a), 11.5 min
(5b), 12.1 min and logP 0.47 (5c), and 16.4 min and logP
1.72 (5d), respectively. The logP values for compounds
5a,b were not available due to their low chemical
stability (Table 1). In comparison, cycloAmb-phospho-
nate esters 6 showed retention times of 11.9 min and a
logP of 0.49 (X ) F; 6a), 11.0 min and logP 0.23 (X ) H;
6b), and 12.1 min and logP 0.45 (X ) 3-Me; 6c). These
data suggest that the cycloAmb-phosphonate diesters
are slightly more lipophilic as compared to cycloSal-
phosphonates 5. However, compared to the cycloSal-
phosphate esters 4, both phosphonates are more hydro-
philic: 12.3 min (X ) H; logP ) 0.27); 13.7 min (X )
6-F; logP ) 0.69); and 13.6 min (X ) 3-Me; logP ) 0.70).
Thus, 3-tBu-cycloSal-PMEA 5c has more or less the
same lipophilicity as 3-Me-cycloAmb-PMEA 6c and
cycloSal-d4TMP 4.
chemically induced tandem or cascade reaction.5 In
contrast to other prodrug concepts that are based on
enzymatically triggered activation, our approach in-
volves the successive coupled cleavage of the phenyl and
benzyl esters of the cycloSal-phosphate triester. This
chemical degradation pathway has been proven by
different methodologies (e.g., 31P NMR or hydrolysis in
labeled water). The intracellular delivery of the nucleo-
tides has additionally been proven by the observed
biological activity in different cell lines and in thymidine
kinase-deficient CEM cells. Additionally, cell incubation
studies with radiolabeled cycloSal-phosphate triesters
demonstrated increased intracellular amounts of phos-
phorylated metabolites.13 Here we investigated (i)
whether the new cycloSal- or cycloAmb-PMEA phos-
phonate diesters are able to release PMEA 1 and (ii)
whether the mechanism of this release is identical to
that found in the case of cycloSal-phosphate triesters
(Scheme 2).
First, the chemical stability was determined in phos-
phate buffer, pH 7.3 (Table 1). The series of cycloSal-
PMEA derivatives 5 showed an unexpected 30-fold
decrease of the stability with respect to the correspond-
ing prototype cycloSal-phosphate esters of d4T 4. The
half-lives of the unsubstituted- (5a) and 3-methyl-
cycloSal-PMEA 5b were 0.1 and 0.6 h only. Thus, they
are too labile to be useful as prodrugs. However, due to
the introduction of stabilizing substituents such as the
3-tert-butyl- in 5c or the 3,5-di-tert-butyl groups in 5d
Hydrolysis Studies. The original cycloSal-phosphate
triester approach has been designed to release the
nucleotides and the masking group selectively by a

8082 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 25
Meier et al.
the half-lives were increased to 3-4 h. The considerable
lower stability points to a much higher electrophilicity
of the phosphorus atom in the phosphonates as com-
pared to the phosphates. This may be due to lack of pπ-
dπ-back-bonding of the additional oxygen atom in the
phosphate esters.14 This interpretation is supported by
the observation that dimethyl methylphosphonate hy-
drolyzed 20-fold faster than trimethyl phosphate in
to the amino group a second electron-donating methyl
group is present in the aromatic ring. This favors the
cleavage of the benzyl-C-O-bond to release PMEA.
One major advantage of the phosphonate prodrugs of
PMEA and related derivatives is their oral bioavailabil-
ity. To address this property in part the stability of some
of the new cycloSal-PMEA and cycloAmb-PMEA deriva-
tives in acidic media was determined in citrate-buffer
at pH 2.0 (Table 1). All studied compounds showed
markedly higher stability at this pH value compared
to the stability values found at pH 7.3. The cycloAmb-
compounds 6 were found to be 1.3-3-fold more stable.
Interestingly, the cycloSal-PMEA compounds 5 proved
to be 3-6-fold more stable. It cannot be excluded that
this difference is caused by an N-protonation in the case
of the cycloAmb-derivatives.
In addition to hydrolysis studies under pure chemical
conditions, the compounds were also incubated with
CEM/O cell extracts and in 20% human serum. Because
of the very low stability of cycloSal-PMEAs 5a,b only
the more stable cycloSal-PMEA derivatives 5c and 5d
were studied. Under both conditions the two compounds
showed half-lives of 3 to 4 h and the rapid release of
PMEA. Thus, the stability was nearly identical to that
observed in the aqueous system. This result points to a
pure chemically driven hydrolysis mechanism. Such a
behavior has been observed already for the cycloSal-
phosphate triesters.5 Additionally, cycloAmb-PMEAs 6
were incubated in CEM cell extracts. As shown by
means of HPLC, compounds 6 proved to be much more
stable than cycloSal-diesters 5 and they release the
benzyl phosphonate ester of type 12 (Z ) NH2; Scheme
2) first. From this intermediate PMEA is slowly re-
leased. After 8 h of incubation of cycloAmb-PMEA 6a
in the extracts, 74% of the starting material, 16% of the
intermediate 13, and 10% of PMEA was detected. It
cannot be excluded that intermediate 12 is or can also
be cleaved enzymatically by a phosphoramidase present
in the CEM cell extracts. This enzyme has been sug-
gested for the cleavage of nucleoside phosphoramidite
monoesters formed from phosphoramidites pronucle-
otides.18
alkaline solution.15 The activation barrier of the (EAct
)
phosphate hydrolysis was found to be 2.7 kcal/mol
higher compared to the EAct of methylphosphonate.16
Nevertheless, 31P NMR studies in imidazole-HCl
buffer, pH 7.3 revealed that the hydrolysis product of
the cycloSal-PMEA diester 5 is solely PMEA 1. Thus,
the hydrolysis mechanism via benzylphosphonate ester
12 as shown in Scheme 2 is entirely identical to that of
the cycloSal-phosphates. Interestingly, the selectivity of
the cleavage is much higher. In the case of 3,5-di-tert-
butyl-cycloSal-d4TMP about 33% of the phenyl phos-
phate diester (analogous to 13, Scheme 2) instead of the
benzyl phosphate diester was formed as a side product.
The phenyl monoester 13 was not formed in one of the
cycloSal-PMEA hydrolyses! 31P-NMR spectra in Sup-
porting Information show the hydrolysis of 3-methyl-
cycloSal-PMEA 5b (18.5 ppm) and 3,5-di-tert-butyl-
cycloSal-PMEA 5d (19.8 ppm) in imidazole/HCl buffer
at pH 7.3. Only in the spectra of compound 5b can the
benzyl PMEA monoester 12b be detected in small
amounts because the first hydrolysis is much faster
compared to the hydrolysis of diester 5d. At the end of
the hydrolyses in both cases PMEA (15.4 ppm) was
observed as the product. In the phosphate triester
series, 3-methyl-cycloSal derivatives led to the formation
of 6-8% of the phenyl phosphate diester.17
The unexpected low stability of the cycloSal-PMEA
derivatives 5 led to the development of the cycloAmb-
PMEA derivatives 6 because the less electronegative
nitrogen reduces the electrophilicity of the phosphorus
atom and consequently should therefore increase the
stability. In 31P NMR experiments, the selective deg-
radation to PMEA from 3-methyl-cycloAmb-PMEA 6b
was observed again (Supporting Information).
Anti-Cholinesterase Activity. As the cycloSal-
PMEAs, all three cycloAmb-PMEAs were found to be
noninhibitory to human acetyl- (data not shown)19 and
butyryl cholinesterase (Table 1). Obviously, indepen-
dently from the chemical stability of the lipophilic
PMEA compounds, these structures are not substrates
for this class of enzymes. In more detailed previous
studies, it was shown that nucleosides bearing an
adenine heterocycle had even better substrate properties
compared to guanine or pyrimidines.20 Thus, the acyclic
“glycon” structure of PMEA seems to determine the lack
of cholinesterase-inhibiting properties.
Antiviral Evaluation. The phosphonate prodrugs
have been tested for their antiviral potency against
HIV-1 and HIV-2 in CEM/O cells. The results of these
tests are summarized in Table 1. PMEA 1 was found to
be active against both virus strains at 10 µM. Interest-
ingly, although cycloSal-PMEA derivatives 5c,d have
only half-lives of 3-4 h in the hydrolyses studies, both
of them proved to be slightly more antivirally potent
compared to the parent compound (3-5 µM). CycloSal-
PMEAs 5a,b were not tested because of their extreme
More importantly, the stability of the compounds was
dramatically higher compared to the cycloSal-PMEAs,
for example, a 35-fold increase was observed for 3-meth-
yl-cycloAmb-PMEA 6c compared to 3-methyl-cycloSal-
PMEA 5b (Table 1). The half-lives of the cycloAmb-
phosphonate esters were very similar to those of the
corresponding cycloSal-d4TMP phosphate triesters. Thus,
by simply exchanging the phenol oxygen by a nitrogen
atom, the higher electrophilicity of the phosphonate
P-atom was compensated. However, a further conse-
quence was observed in the 31P NMR studies. In the
case of the cycloSal-d4TMP triesters 4 the rate-limiting
step of the delivery of the d4TMP was the cleavage of
the phosphate triester to the intermediate phosphate
diester. However, in the case of the cycloAmb-phospho-
nate diesters 6 the rate-limiting step is the cleavage of
the intermediate benzyl phosphonate monoester 11 to
PMEA. This was observed particularly for 6-fluoro-
cycloAmb-PMEA 6a and cycloAmb-PMEA 6b. In the
former case, the formation of PMEA was very slow. In
contrast, 3-methyl-cycloAmb-PMEA 6c led to a com-
paratively fast formation of PMEA because in addition

cycloSal-PMEA and cycloAmb-PMEA
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 25 8083
distilled from Na prior to use. The solvents for the HPLC were
obtained from Merck (acetonitrile, HPLC grade). Ion pairing
buffer solution was prepared by mixing 6.6 mL of tetrabutyl-
ammonium hydroxide with 1000 mL of water. The pH value
was adjusted to 3.8 by adding concentrated phosphoric acid
(buffer I). To 60 mL of buffer I solution 1000 mL of water was
added (buffer II). Evaporation of solvents was carried out on
a rotary evaporator under reduced pressure or using a high-
vacuum pump. Chromatography: Chromatotron (Harrison
Research 7924), silica gel 60Pf (Merck, “gipshaltig”); UV
detection at 254 nm. TLC: analytical thin-layer chromatog-
raphy was performed on Merck precoated aluminum plates
60 F254 with a 0.2-mm layer of silica gel containing a fluores-
cence indicator; sugar-containing compounds were visualized
with the sugar spray reagent (0.5 mL of 4-methoxybenzalde-
hyde, 9 mL of ethanol, 0.5 mL of concentrated sulfuric acid,
and 0.1 mL of glacial acectic acid) by heating with a fan or a
hot plate. HPLC: (Merck-Hitachi) analytical HPLC, LiChro-
CART 250-3 with LiChrospher 100 RP-18 endcapped (5 µm),
gradient I 12-80% CH3CN (0-20 min), 12% CH3CN (20-35
min), flow 0.6 mL, UV detection at 250 nm; gradient II 8-100%
CH3CN (0-22 min), 100% CH3CN (22-27 min), 8% CH3CN
(27-33 min), flow 0.6 mL, UV detection at 250 nm; gradient
III same as II, instead of water the ion pairing buffer solution
was used. NMR spectra were recorded using (1H NMR) Bruker
AC 250 at 250 MHz, Bruker WM 400 at 400 MHz, Bruker
AMX 400 at 400 MHz or Bruker DMX 500 at 500 MHz (CDCl3
or DMSO as internal standard); (13C NMR) Bruker WM 400
at 101 MHz, Bruker AMX 400 at 101 MHz or Bruker DMX
500 at 123 MHz (CDCl3 or DMSO as internal standard); (31P
NMR) Bruker AMX 400 at 162 MHz or Bruker DMX 500 at
lability. For comparison, bis(POM)PMEA 15 was also
tested. The antiviral activity was found to be about
0.075-0.18 µM in all three assays. Thus, compound 15
is 16-fold more active compared to 3-tert-butyl-cycloSal-
PMEA 5c. On the other hand, the SI value of the same
compounds are 9.7 for 3-tert-butyl-cycloSal-PMEA 5c
while that of bis(POM)PMEA was 4.1! Thus, the higher
antiviral activity of the bis(POM) derivative was coun-
teracted by a markedly increased cytotoxicity. This
parallel increase in toxicity with increased antiviral
potency has also been observed in the case of the bis-
(SATE)-PMEAs4 and the phosphoramidates of PMEA.18
The higher antiviral activity of the cycloSal-PMEAs 5
may be attributed to the low chemical stability (t1/2 3-4
h, releasing quickly the parent drug) and/or the lower
lipophilicity as compared to, for example, bis(POM)-
PMEA.
Although the chemical stability has been increase
considerably, cycloAmb-PMEAs 6 surprisingly exhibited
antiviral activity at an EC50 of 20-29 µM only. This is
2-3-fold lower compared to the parent and 10-fold lower
than the cycloSal-PMEAs. However, the three cycloAmb-
derivatives had CC50 values as high as 186 µM to >250
µM. Thus, these compounds are even less toxic than
PMEA 1 (CC50: 50 µM) and the SI value is as before
about 10. The lower antiviral activity may be attributed
to the slow release of PMEA from the cycloAmb-PMEAs.
As described in the hydrolysis studies section, the rate-
limiting step in the PMEA delivery of the cycloAmb-
derivatives is the cleavage of the intermediate benzyl
phosphonate ester.
1
202 MHz (H3PO4 as external standard). All H and 13C NMR
chemical shifts (δ) are quoted in parts per million (ppm)
downfield from tetramethylsilane, (CD3)(CD2H)SO being set
at δH 2.49 as a reference. 31P NMR chemical shifts are quoted
in ppm using H3PO4 as external reference. The spectra were
The interesting different behavior of the two groups
of prodrugs will be studied in antiviral assays of
different DNA viruses, for example, HSV-1 and HBV.
recorded at room temperature, and all 13C and 31P NMR were
recorded in proton-decoupled mode. UV spectra were taken
with a Varian Cary 1E UV/Vis spectrometer. Infrared spectra
were recorded with a Perkin-Elmer 1600 Series FT-IR or a
ATI Mattson Genesis Series FT-IR spectrometer in KBr
pellets. Mass spectra were obtained with a Finnigan electro-
spray MAT 95 Trap XL (ESI) or a VG Analytical VG/70-250 F
spectrometer (FAB, matrix was m-nitrobenzyl alcohol). The
test compounds were isolated as mixtures of enantiomers
arising from the mixed stereochemistry at the phosphate
center. The resulting lyophilized triesters did not gave useful
microanalytical data most probably due to incomplete combus-
tion of the compound, but were found to be pure by HPLC
analysis, high-field multinuclear NMR spectroscopy, and mass
spectroscopy.
Conclusion
In summary, from the hydrolytic and antiviral data
disclosed here, cycloSal-PMEAs 5 and cycloAmb-PMEAs
6 provide a method to deliver the nucleoside phospho-
nate PMEA intracellularly. It was proven that the
delivery mechanism is the same as for cycloSal-d4TMP
4. Interestingly, the selectivity of the degradation
pathway was even better as compared to the cycloSal-
phosphate triesters because in the hydrolysis studies
none of the new derivatives led to the formation of the
benzyl phosphonate ester 12. The stability was modu-
lated considerably by replacing the phenol oxygen by a
nitrogen atom. The antiviral evaluation showed inter-
esting properties among the two different groups of
lipophilic prodrugs. While cycloSal-PMEAs improved
the antiviral activity as compared to PMEA, the
cycloAmb-derivatives reduced the cytotoxicity notice-
able. Further work to tune the properties of these two
interesting classes of prodrug compounds are currently
underway in our laboratories.
9-[2-Diethylphosphonylmethoxyethyl]adenine 7. Di-
ethyl-PMEA 8 was prepared according to a protocol published
by Holy et al.21 and the modifications introduced by Bronson
et al.22
N6-4-Monomethoxytrityl-9-[2-diethylphosphonyl-
methoxyethyl]adenine 9. The compound was prepared
according to Puech et al.9 Yield: 2.13 g (3.54 mmol, 89%,
colorless foam); the analytical data were found to be identical
to those published.
General Procedure for the Preparation of the N6-
MMTr-cycloSal-PMEAs 10a-d. One equivalent of diethyl
PMEA 9 was coevaporated twice with dry CH3CN and subse-
quently solubilized in CH3CN in an inert atmosphere. After
addition of 50 µL of pyridine, 2.5 equiv of bromotrimethylsilane
(TMSBr) were added slowly and the resulting mixture was
stirred for 16 h at ambient temperature. The volatiles were
distilled off in a high vacuum and the residual was coevapo-
rated twice with toluene. The remaining foam was solubilized
in dry CH2Cl2 and 2.1 equiv of PCl5 were added. After 2.5 h of
stirring at room temperature, the solvent was removed in high
vacuum. The remaining yellowish foam was again coevapo-
rated with toluene and then CH2Cl2 was added. The solution
was cooled to 0 °C and within 15 min a solution of 2 equiv of
Experimental Section
All experiments involving water-sensitive compounds were
conducted under scrupulously dry conditions (argon atmo-
sphere). Solvents: anhydrous methylene chloride (CH2Cl2),
anhydrous tetrahydrofurane (THF), and anhydrous acetoni-
trile (CH3CN) were obtained in a Sure/Seal bottle from Fluka
and stored over 4 Å molecular sieves; ethyl acetate, methylene
chloride, and methanol employed in chromatography were
distilled before used. Ethyldiisopropylamine (DIPEA) was

8084 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 25
Meier et al.
the salicyl alcohol 7 and 3.8 equiv of NEt3 in dry CH2Cl2 were
added dropwise. After complete addition, the solution was
warmed to room temperature and the reaction was followed
by TLC (CH2Cl2/MeOH 9:1 v/v; 0.1% HOAc). After complete
evaporation of the solvent, the residues were purified by
chromatography on a Chromatotron on silica gel plates using
a gradient of MeOH (0-4%) in CH2Cl2.
(m, 4H), 7.07-7.02 (m, 2H), 6.84 (d, 2H), 5.27 (dd, 1H), 5.22
(dd, 1H), 4.19 (dd, 2H), 4.17 (dd, 1H), 4.10 (dd, 1H), 3.83 (dd,
1H), 3.80 (dd, 1H), 3.71 (s, 3H), 1.26 (s, 9H), 1.26 (s, 9H); 13C
NMR (101 MHz, DMSO-d6) δ 157.9, 153.5, 151.2, 148.7, 148.2,
145.9, 145.3, 141.6, 137.5, 137.3, 129.9, 128.6, 127.8, 126.7,
123.9, 123.5, 121.5, 120.3, 113.1, 70.5, 70.1, 67.3, 63.9, 55.2,
42.6, 34.5, 34.4, 31.3, 29.7; 31P NMR (202 MHz, DMSO-d6) δ
16.98; MS (FAB) m/z calc. 745.34, found 746.2 (M + H+).
General Procedure of the Deprotection of N6-MMTr-
cycloSal-PMEAs 10. The MMTr-protected cycloSal-PMEA
derivative was solubilized in dry CH2Cl2 or CH3CN in an inert
atmosphere. Then 20 vol % of trifluoroacetic acid (TFA) was
added dropwise. After 1 h of stirring at room temperature, the
reaction mixture was diluted with water, and the resulting
mixture was freeze-dried. The residual was purified by chro-
matography on a Chromatotron on silica gel plates using a
gradient of MeOH (0-12%) in CH2Cl2 containing 0.1% of
HOAc. The cycloSal-PMEAs 5 were isolated as their TFA salts.
9-[2-cycloSal-phosphonylmethoxyethyl]adenine
(cycloSal-PMEA) 5a. Reagents: 71.0 mg (0.11 mmol) of N-4-
monomethoxytrityl-9-[2-cycloSal-phosphonylmethoxyethyl]ad-
enine 10a, 250 µL of TFA, 1 mL of dry CH2Cl2. Yield: 28.4
mg (0.06 mmol, 53%; colorless solid); 1H NMR (500 MHz,
DMSO-d6) δ 8.47 (br. s, 1H), 8.43 (br. s, 1H), 8.39 (s, 1H), 8.17
(s, 1H), 7.45 (dd, 1H), 7.34 (d, 1H), 7.27 (dd, 1H), 7.16 (d, 1H),
5.47 (dd, 1H), 5.43 (dd, 1H), 4.42-4.39 (m, 1H), 4.32 (dd, 1H),
4.29 (dd, 1H), 4.01 (dd, 1H), 3.97 (dd, 1H); 13C NMR (101 MHz,
DMSO-d6) δ 155.2, 152.8, 149.6, 148.0, 142.6, 129.9, 126.2,
124.2, 122.5, 120.5, 117.9, 70.4, 67.1, 63.5; 31P NMR (202 MHz,
DMSO-d6) δ 16.62; 19F NMR (471 MHz, DMSO-d6) δ -74.40;
MS (ESI) m/z calc. 361.09, found 362.26 (M + H+), 384.28 (M
+ Na+).
N6-4-Monomethoxytrityl-9-[2-cycloSal-phosphonyl-
methoxyethyl]adenine (N6-MMTr-cycloSal-PMEA) 10a.
Reagents: 500 mg (0.83 mmol) of N6-4-monomethoxytrityl-9-
[2-diethylphosphonylmethoxyethyl]adenine 9, 269 µL (318 mg,
2.08 mmol) of bromotrimethylsilane (TMSBr), 5 mL of dry
CH3CN; 5 mL of dry CH2Cl2, 364 mg (1.75 mmol) of PCl5; 2.5
mL of dry CH2Cl2, 207 mg (1.66 mmol) of salicyl alcohol 7a,
440 µL (319 mg, 3.16 mmol) of dry NEt3, 2.5 mL of dry
1
CH2Cl2. Yield: 244 mg (0.39 mmol, 46%; colorless foam); H
NMR (500 MHz, DMSO-d6) δ 7.89 (s, 1H), 7.3 (s, 1H), 7.32-
7.26 (m, 9H), 7.22-7.19 (m, 5H), 7.14 (d, 1H), 7.01 (dd, 1H),
7.01 (d, 1H), 6.84 (d, 2H), 5.27 (dd, 1H), 5.23 (dd, 1H), 4.22-
4.15 (m, 3H), 4.13 (dd, 1H), 3.83 (dd, 1H), 3.79 (dd, 1H), 3.70
(s, 3H); 13C NMR (101 MHz, DMSO-d6) δ 157.9, 153.6, 151.2,
149.6, 148.8, 145.3, 141.6, 137.3, 130.0, 129.9, 128.6, 127.8,
126.7, 126.3, 124.2, 122.6, 120.4, 118.0, 113.2, 70.5, 70.1, 67.1,
64.0, 55.2, 42.5; 31P NMR (202 MHz, DMSO-d6) δ 16.56; MS
(FAB) m/z calc. 633.21, found 634.5 (M + H+).
N6-4-Monomethoxytrityl-9-[2-(3-methyl-cycloSal)-phos-
phonylmethoxyethyl]adenine (N6-MMTr-3-Me-cycloSal-
PMEA) 10b. Reagents: 300 mg (0.50 mmol) of N6-4-monometh-
oxytrityl-9-[2-diethylphosphonylmethoxyethyl]adenine 9, 161
µL (191 mg, 1.25 mmol) of bromotrimethylsilane (TMSBr), 3
mL of dry CH3CN; 4 mL of dry CH2Cl2, 302 mg (1.05 mmol) of
PCl5; 2.0 mL of dry CH2Cl2, 138 mg (0.99 mmol) of 3-methyl-
salicyl alcohol 7b, 264 µL (192 mg, 1.89 mmol) of dry NEt3,
2.0 mL of dry CH2Cl2. Yield: 102 mg (0.16 mmol, 32%; colorless
9-[2-(3-Methyl-cycloSal)-phosphonylmethoxyethyl]ad-
enine (3-Me-cycloSal-PMEA) 5b. Reagents: 88.0 mg (0.14
mmol) of N-4-monomethoxytrityl-9-[2-(3-methyl-cycloSal-phos-
phonylmethoxyethyl]adenine 10b, 500 µL of TFA, 2 mL of dry
1
foam); H NMR (500 MHz, DMSO-d6) δ 7.85 (s, 1H), 7.84 (s,
1H), 7.31-7.19 (m, 13H), 7.15-7.12 (m, 1H), 7.00-6.94 (m,
2H), 6.86-6.83 (m, 2H), 5.25 (dd, 1H), 5.26 (dd, 1H), 4.20-
4.13 (m, 3H), 4.09 (dd, 1H), 3.81 (dd, 1H), 3.77 (dd, 1H), 3.71
(s, 3H), 2.09 (s, 3H); 13C NMR (101 MHz, DMSO-d6) δ 157.9,
153.5, 151.2, 148.7, 148.1, 145.4, 141.5, 137.3, 131.1, 130.0,
128.6, 127.8, 126.8, 126.7, 123.8, 123.7, 122.8, 120.4, 113.2,
70.5, 70.0, 67.0, 64.0, 55.1, 42.5, 15.0; 31P NMR (202 MHz,
DMSO-d6) δ 17.46; MS (FAB) m/z calc. 647.23, found 648.4
(M + H+).
1
CH2Cl2. Yield: 44.6 mg (91.1 µmol, 67%; colorless solid); H
NMR (500 MHz, DMSO-d6) δ 8.31 (br. s, 1H), 8.28 (br. s, 1H),
8.24 (s, 1H), 7.98 (s, 1H), 7.17 (d, 1H), 7.03-6.98 (m, 2H), 5.28
(dd, 1H), 5.23 (dd, 1H), 4.27 (dd, 1H), 4.22 (dd, 1H), 4.16 (dd,
1H), 4.12 (dd, 1H), 3.83 (t, 2H), 2.11 (s, 3H); 13C NMR (125
MHz, DMSO-d6) δ 156.1, 152.5, 149.6, 148.2, 140.9, 131.1,
126.8, 123.8, 123.8, 122.8, 118.7, 70.6, 67.0, 64.1, 42.5, 15.0;
31P NMR (202 MHz, DMSO-d6) δ 17.49; 19F NMR (471 MHz,
DMSO-d6) δ -74.46; MS (FAB) m/z calc. 375.11, found 376.2
(M + H+).
N6-4-Monomethoxytrityl-9-[2-(3-tert-butyl-cycloSal)-
phosphonylmethoxyethyl]adenine
(N6-MMTr-3-t-Bu-
cycloSal-PMEA) 10c. Reagents: 500 mg (0.83 mmol) of N6-
4-monomethoxytrityl-9-[2-diethylphosphonylmethoxyethyl]-
adenine 9, 269 µL (318 mg, 2.08 mmol) of bromotrimethylsi-
lane (TMSBr), 5 mL of dry CH3CN; 5 mL of dry CH2Cl2, 364
mg (1.75 mmol) of PCl5; 2.5 mL of dry CH2Cl2, 300 mg (1.66
mmol) of 3-tert-butylsalicyl alcohol 7c, 440 µL (319 mg, 3.16
mmol) of dry NEt3, 2.5 mL of dry CH2Cl2. Yield: 274 mg (0.40
9-[2-(3-tert-Butyl-cycloSal)-phosphonylmethoxyethyl]-
adenine (3-t-Bu-cycloSal-PMEA) 5c. Reagents: 191 mg
(0.28 mmol) of N-4-monomethoxytrityl-9-[2-(3-tert-butyl-
cycloSal-phosphonylmethoxyethyl]adenine 10c, 750 µL of TFA,
2 mL of dry CH3CN. Yield: 120 mg (227 µmol, 82%; colorless
1
solid); H NMR (500 MHz, DMSO-d6) δ 8.09 (s, 1H), 7.81 (s,
1H), 7.46 (s, 2H), 7.28 (d, 1H), 7.11-7.05 (m, 2H), 5.27 (dd,
1H), 5.23 (dd, 1H), 4.21 (dd, 2H), 4.20-4.17 (m, 1H), 4.14 (dd,
1H), 3.86 (dd, 1H), 3.82 (dd, 1H), 1.27 (s, 9H); 13C NMR (100
MHz, DMSO-d6) δ 155.1, 151.3, 149.4, 148.6, 141.4, 138.3,
127.3, 124.6, 124.0, 123.9, 118.6, 70.6, 67.0, 64.0, 42.7, 34.4,
29.7; 31P NMR (202 MHz, DMSO-d6) δ 17.06; 19F NMR (471
MHz, DMSO-d6) δ -74.32; MS (HR-ESI) m/z calc. 440.1464,
found 440.1469 (M + Na+).
1
mmol, 48%; colorless foam); H NMR (500 MHz, DMSO-d6) δ
7.84 (s, 1H), 7.81 (s, 1H), 7.31-7.26 (m, 9H), 7.22-7.19 (m,
5H), 7.07-7.02 (m, 2H), 6.84 (d, 2H), 5.24 (dd, 1H), 5.20 (dd,
1H), 4.19 (dd, 2H), 4.19 (dd, 1H), 4.13 (dd, 1H), 3.83 (dd, 1H),
3.80 (dd, 1H), 3.70 (s, 3H), 1.26 (s, 9H); 13C NMR (101 MHz,
DMSO-d6) δ 157.9, 153.5, 151.2, 148.7, 145.3, 141.6, 138.3,
137.3, 129.9, 128.6, 127.8, 127.3, 126.6, 124.5, 123.8, 120.4,
120.3, 113.2, 70.5, 70.4, 67.0, 64.0, 55.1, 42.7, 34.4, 29.7; 31P
NMR (202 MHz, DMSO-d6) δ 16.93; MS (FAB) m/z calc. 689.28,
found 690.5 (M + H+).
N6-4-Monomethoxytrityl-9-[2-(3,5-Di-tert-butyl-cycloSal)-
phosphonylmethoxyethyl]adenine (N6-MMTr-3,5-di-t-
Bu-cycloSal-PMEA) 10d. Reagents: 500 mg (0.83 mmol) of
N6-4-monomethoxytrityl-9-[2-diethylphosphonylmethoxyethyl]-
adenine 9, 269 µL (318 mg, 2.08 mmol) of bromotrimethylsi-
lane (TMSBr), 5 mL of dry CH3CN; 5 mL of dry CH2Cl2, 364
mg (1.75 mmol) of PCl5; 2.5 mL of dry CH2Cl2, 392 mg (1.66
mmol) of 3,5-di-tert-butylsalicyl alcohol 7d, 440 µL (319 mg,
3.16 mmol) of dry NEt3, 2.5 mL of dry CH2Cl2. Yield: 193 mg
(0.26 mmol, 31%; colorless foam); 1H NMR (500 MHz, DMSO-
d6) δ 7.85 (s, 1H), 7.76 (s, 1H), 7.30-7.26 (m, 9H), 7.22-7.19
9-[2-(3,5-Di-tert-butyl-cycloSal)-phosphonylmeth-
oxyethyl]adenine (3,5-Di-t-Bu-cycloSal-PMEA) 5d. Re-
agents: 145 mg (0.19 mmol) of N-4-monomethoxytrityl-9-[2-
(3,5-di-tert-butyl-cycloSal-phosphonylmethoxyethyl]adenine
10d, 500 µL of TFA, 2 mL of dry CH3CN. Yield: 75.0 mg (0.13
1
mmol, 66%; colorless solid); H NMR (500 MHz, DMSO-d6) δ
8.07 (s, 1H), 7.78 (s, 1H), 7.29 (d, 1H), 7.26 (s, 2H), 7.16 (d,
1H), 5.29 (dd, 1H), 5.24 (dd, 1H), 4.19 (dd, 2H), 4.18 (dd, 1H),
4.11 (dd, 1H), 3.86 (dd, 1H), 3.81 (dd, 1H), 1.29 (s, 9H), 1.26
(s, 9H); 13C NMR (100 MHz, DMSO-d6) δ 155.7, 152.0, 149.5,
146.0, 146.0, 141.2, 137.6, 123.9, 123.6, 121.5, 118.7, 70.6, 67.4,
64.0, 42.6, 34.5, 34.4, 31.3, 29.7; 31P NMR (202 MHz, DMSO-
d6) δ 17.49; 19F NMR (471 MHz, DMSO-d6) δ -74.46; MS
(HR-ESI) m/z calc. 496.2090, found 496.2082 (M + Na+).

cycloSal-PMEA and cycloAmb-PMEA
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 25 8085
2-Amino-6-fluoro-benzyl alcohol 11a. 1.00 g (6.40 mmol)
of 2-amino-6-fluoro-benzoic acid dissolved in 30 mL of dry THF
was added dropwise to a suspension of 0.73 g (19.0 mmol) of
lithiumaluminum hydride in 60 mL of dry THF. The resulting
mixture was stirred for 1 h at room temperature and then
refluxed for 2 h. After cooling, the mixture was neutralized
with 2 N HCl and extracted with EtOAc four times. The
combined organic phases were filtered through Celite and then
concentrated under reduced pressure. The crude product was
recrystallized from petrolether/EtOAc 99:1. Yield 64%; mp
89-90 °C,1H NMR (400 MHz, DMSO-d6) δ 7.04 (ddd, 1H),
6.49-6.41 (m, 2H), 4.79 (d, 2H); 13C NMR (101 MHz, DMSO-
d6) δ 161.2, 148.4, 129.9, 112.9, 111.6, 104.8, 55.4; 19F NMR
(471 MHz, CDCl3) δ -120.8; Anal. Calcd for C7H8FNO: C
59.57 H 5.71 N 9.92, found C 59.67 H 5.81 N 9.89.
General Procedure for the Preparation of the
cycloAmb-nucleoside Phosphonates 6a-c. 3.5 equiv of
oxalyl chloride were added slowly to a slurry of PMEA 1 (1.0
equiv) and N,N-diethylformamide (1.1 equiv) in dry CH2Cl2
and the resulting mixture was refluxed for 3 h. The reaction
was cooled to room temperature and concentrated in vacuo.
The resulting foam was dissolved in dry CH2Cl2, cooled to 0
°C, and treated slowly with 2 equiv of pyridine. The resulting
solution was then added slowly to a solution of 2-aminoben-
zyl alcohol (1.0 equiv) and DIPEA (6.3 equivalents) in dry
CH2Cl2 at -78 °C. The reaction mixture was slowly warmed
to room temperature, stirred overnight and was then concen-
trated in vacuo. The crude product was dissolved in EtOH,
treated with HOAc (7.0 equiv), and warmed to reflux. After 2
h, the reaction was cooled to room temperature and concen-
trated. The residue was purified by silica gel chromatography
three times [(1) 10% MeOH in EtOAc on a column; (2) on a
Chromatotron on silica gel plates using a gradient of MeOH
(3-15%) in CH2Cl2; (3) on a Chromatotron on silica gel plates
using a gradient of MeOH (3-12%) in EtOAc]. Finally, the
products were freeze-dried.
9-[2-(6-Fluoro)-cycloAmb-phosphonylmethoxyethyl]-
adenine (6-F-cycloAmb-PMEA) 6a. Reagents: 200 mg (0.73
mmol) of PMEA 1, 88 µL (80 mg, 0.79 mmol) of N,N-
diethylformamide, 0.23 mL (0.35 g, 2.5 mmol) of oxalyl
chloride, 5.5 mL of CH2Cl2; 0.13 mL (0.13 g, 1.5 mmol) of
pyridine, 5 mL of CH2Cl2; 103 mg (0.73 mmol) of 2-amino-6-
fluorobenzyl alcohol 11a, 0.79 mL (0.60 g, 4.6 mmol) of DIPEA,
4 mL of CH2Cl2. Yield: 68 mg (0.18 mmol, 25%; colorless foam);
1H NMR (400 MHz, DMSO-d6) δ 8.77 (d, 1H), 8.08 (s, 1H),
7.79 (s, 1H), 7.24-7.14 (m, 3H), 6.71 (dd, 1H), 6.66 (d, 1H),
5.26 (dd, 1H), 5.12 (dd, 1H), 4.18 (dt, 2H), 3.95 (dd, 1H), 3.89
(d, 1H), 3.79 (t, 2H); 13C NMR (101 MHz, DMSO-d6) δ 158.1,
155.9, 152.3, 149.4, 141.8, 140.8, 130.0, 118.5, 112.7, 110.0,
107.1, 70.3, 65.8, 60.9, 42.3; 19F NMR (471 MHz, DMSO-d6) δ
-120.6; 31P NMR (202 MHz, DMSO-d6) δ 16.68; MS (HR-ESI)
m/z calc. 378.1006, found 401.0897 (M + Na+).
9-[2-cycloAmb-phosphonylmethoxyethyl]adenine
(cycloAmb-PMEA) 6b. Reagents: 200 mg (0.73 mmol) of
PMEA 1, 88 µL (80 mg, 0.79 mmol) of N,N-diethylformamide,
0.23 mL (0.35 g, 2.5 mmol) of oxalyl chloride, 5.5 mL of
CH2Cl2; 0.13 mL (0.13 g, 1.5 mmol) of pyridine, 5 mL of
CH2Cl2; 90 mg (0.73 mmol) of 2-aminobenzyl alcohol 11b, 0.79
mL (0.60 g, 4.6 mmol) of DIPEA, 4 mL of CH2Cl2. Yield: 86
mg (0.24 mmol, 33%; colorless foam); 1H NMR (400 MHz,
DMSO-d6) δ 8.52 (d, 1H), 8.09 (s, 1H), 7.73 (s, 1H), 7.22-7.09
(m, 3H), 7.12 (d, 1H), 6.88 (dd, 1H), 6.85 (d, 1H), 5.18-5.03
(m, 2H), 4.14 (t, 2H), 3.87 (d, 1H), 3.85 (d, 1H), 3.76 (t, 2H);
13C NMR (101 MHz, DMSO-d6) δ 155.9, 152.3, 149.4, 140.9,
139.9, 128.9, 125.6, 122.5, 120.6, 118.5, 116.7, 70.2, 66.5, 65.9,
42.3; 31P NMR (202 MHz, DMSO-d6) δ 17.49; MS (HR-ESI)
m/z calc. 360.1100, found 361.1177 (M + H+).
DIPEA, 4 mL of CH2Cl2. Yield: 117 mg (0.31 mmol, 42%;
colorless foam); 1H NMR (400 MHz, DMSO-d6) δ 8.08 (s, 1H),
7.82 (d, 1H), 7.66 (s, 1H), 7.16 (s, 2H), 7.08 (d, 1H), 7.00 (d,
1H), 6.82 (dd, 1H), 5.13-5.97 (m, 2H), 4.16-4.07 (m, 2H), 3.85
(d, 2H), 3.73 (t, 2H), 2.16 (s, 3H); 13C NMR (101 MHz, DMSO-
d6) δ 155.9, 152.3, 149.3, 140.8, 138.1, 130.3, 129.2, 125.5,
123.4, 120.7, 118.5, 70.1, 66.1, 65.9, 42.4, 17.1; 31P NMR (202
MHz, DMSO-d6) δ 19.59; MS (HR-ESI) m/z calc. 374.1256,
found 397.1151 (M + Na+).
Kinetic Data. (a) Aqueous Buffers. 11.4 µL of DMSO
stock solutions (50 mM) of the triesters were diluted in 300
µL of water or water/DMSO (c ) 1.9 mM). 0.3 mL of this
solution were added to 0.3 mL of aqueous buffer (50 mM
phosphate buffer, pH 7.3 or 50 mM citrate buffer, pH 2.0)
containing 5 µL of an aqueous AZT solution (AZT as internal
standard) at 37 °C. The final concentrations were 0.96 mM
for the triesters and 24.8 mM for the aqueous buffer. Aliquots
of 60 µL of the hydrolysis mixture were taken, and the
hydrolysis was stopped by addition of 5 µL of glacial acetic
acid and frozen in liquid air. After thawing, samples were
analyzed by analytical HPLC [Merck LiChroCART column,
LiChrospher 100 reversed-phase silica gel RP-18 endcapped
(5 µm); UV detection at 250 nm]. The hydrolysis of the
compounds 5 and 6 was followed by integration of the
decreasing peak areas in the HPLC chromatograms. The rate
constants k were determined from slope of the logarithmic
degradation curve. The half-lives (t1/2) were calculated by using
the rate constants k.
(b) CEM Cell Extract. 1.5 mM stock solution of the
triesters in DMSO were prepared. 20 µL of this stock solution
was mixed with 100 µL of cell extract and 20 µL of a 70 mM
magnesium chloride solution. The hydrolysis process was
stopped after 8 h by addition of 300 µL of acidic methanol and
storage for 5 min at 0 °C. The mixtures were centrifuged by
13000 rpm for 10 min, filtered (Schleicher-Schuell Spartan
13/30, 0.2 µm), and the supernatant was analyzed as men-
tioned above.
(c) Human Sera. The studies were performed as described
in (b) but instead of cell extracts 20% of human serum in
phosphate buffer, pH 6.8 was used, and the data were collected
in the same way.
31P NMR Hydrolysis Studies of the cycloSal-Phosphate
Triesters. Approximately 7 µmol of the cycloSal- or cycloAmb-
phosphonates were dissolved in 300 µL of deuterated DMSO
and 700 µL of a 50 mM imidazole/HCl buffer (pH 7.3). The
resulting solutions were transferred immediately into a NMR
tube and investigated by 31P NMR spectroscopy (proton
decoupled, 202 MHz, 1024 scans each sample). All samples
were stored at room temperature. Furthermore, some proton
coupled 31P NMR spectra (202 MHz, 1024 scans each sample)
were recorded for the identification of the hydrolysis products.
Cholinesterase Assay. The assay has been carried out as
published previously19 using isolated human acetyl- and
butyrylcholinesterase as well as human serum.
Determination of the Partition Coefficients (logP
Values). logP values were determined as described in ref 6a.
Antiretroviral Evaluation. Human immunodeficiency
virus type 1 (HIV-1) was originally obtained from a persis-
tently HIV-infected H9 cell line, as described previously, and
was kindly provided by Dr. R. C. Gallo (then at the National
Institutes of Health, Bethesda, MD). Virus stocks were pre-
pared from the supernatants of HIV-1-infected MT-4 cells.
HIV-2 (strain ROD) was kindly provided by Dr. L. Montagnier
then at the Pasteur Institute, Paris, France, and virus stocks
were prepared from the supernatants of HIV-2-infected MT-4
cells. CEM cells were obtained from the American Tissue
Culture Collection (Rockville, MD). CEM cells were infected
with HIV as previously described.23 Briefly, 4 × 105 cells/mL
were infected with HIV-1 or HIV-2 at ∼100 CCID50 (50% cell
culture infective dose) per mL of cell suspension. Then, 100
µL of the infected cell suspension was transferred into 96-well
microtiter plate wells and mixed with 100 µL of the appropri-
ate dilutions of the test compounds. After 4-5 days, giant cell
9-[2-(3-Methyl)-cycloAmb-phosphonylmethoxyethyl]-
adenine (3-Me-cycloAmb-PMEA) 6c. Reagents: 200 mg
(0.73 mmol) of PMEA 1, 88 µL (80 mg, 0.79 mmol) of N,N-
diethylformamide, 0.23 mL (0.35 g, 2.5 mmol) of oxalyl
chloride, 5.5 mL of CH2Cl2; 0.13 mL (0.13 g, 1.5 mmol) of
pyridine, 5 mL of CH2Cl2; 101 mg (0.73 mmol) of 2-amino-3-
methylbenzyl alcohol 11c, 0.79 mL (0.60 g, 4.6 mmol) of

8086 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 25
Meier et al.
(7) Meier, C.; Lomp, A.; Meerbach, A.; P. Wutzler, P. CycloSal-
BVDUMP pronucleotides - How to convert an anti-EBV-inactive
nucleoside analogue into a bioactive compound. J. Med. Chem.
2002, 45, 5157-5172.
(8) Meier, C.; Ruppel, M.; Vukadinovic, D.; Balzarini, J. Second
generation of cycloSal-pronucleotides with esterase-cleavable
sites - the “lock-in”-concept. Nucl., Nucl., Nucleic Acids 2004,
89-115.
(9) Puech, F.; Gosselin, G.; Lefebvre, I.; Pompom, A.; Aubertin, A.-
M.; Kirn, A.; Imbach, J.-L. Intracellular delivery of nucleoside
monophosphates through a reductase-mediated activation pro-
cess. Antiviral Res. 1993, 22, 155-174.
(10) (a) Morita, T.; Okamoto, Y.; Sakurai, H. A mild and facile
synthesis of alkyl- and arylphosphonyl dichlorides under neutral
conditions. Reaction of bis(trimethylsilyl) phosphonates with
PCl5. Chem. Lett. 1980, 435-438; (b) Serafinowska, H. T.;
Ashton, A. J.; Bailey, S.; Harnden, M. R.; Jackson, S. M.; Sutton,
D. Synthesis and in vivo evaluation of prodrugs of 9-[2-(phospho-
nomethoxy)ethoxy]adenine. J. Med. Chem. 1995, 38, 1372-1379.
(11) Erion, M. D.; Reddy, K. R.; Boyer, S. H.; Matelich, M. C.; Gomez-
Galeno, J.; Lemus, R. H.; Ugarkar, B. G.; Colby, T. J.; Schanzer,
J.; Poelje, P. D. Design, synthesis and characterization of a series
of cytochrome P450 3A-activated prodrugs (HepDirect prodrugs)
useful for targeting phosph(on)ate-based drugs to the liver. J.
Am. Chem. Soc. 2004, 126, 5154-5163.
(12) Balzarini, J.; Cools, J. M.; De Clercq, E. Estimation of the
lipophilicity of anti-HIV nucleoside analogues by determination
of the partition coefficient and retention time on a LiChrosphere
60 RP-8 HPLC column. Biochem. Biophys. Res. Commun. 1989,
158, 413-422.
(13) (a) Balzarini, J.; Aquaro, S.; Knispel, T.; Rampazzo, C.; Bianchi,
V.; Perno, C.-F.; De Clercq, E.; Meier, C. Cyclosaligenyl-2′,3′-
didehydro-2′,3′-dideoxythymidine monophosphate: Efficient in-
tracellular delivery of d4TMP. Mol. Pharmacol. 2000, 58, 928-
935. (b) Balzarini, J.; Naesens, L.; Aquaro, S.; Knispel, T.; Perno,
C.-F.; De Clercq, E.; Meier, C. Intracellular metabolism of
cycloSaligenyl 3′-azido-2′,3′-dideoxythymidine monophosphate,
a prodrug of 3′-azido-2′,3′-dideoxythymidine (Zidovudine). Mol.
Pharmacol. 1999, 56, 1354-1361.
(14) Cox, Jr., J. R.; Ramsay, B. Mechanisms of nucleophilic substitu-
tion in phosphate esters. Chem. Rev. 1964, 64, 317-352.
(15) Hudson, R. F.; Keay, L. The hydrolysis of phosphonate esters.
J. Chem. Soc. 1956, 2463-2469.
(16) Westheimer, F. H.; Huang, S.; Covitz, F. Rates and mechanisms
of hydrolysis of esters of phosphorous acid. J. Am. Chem. Soc.
1988, 110, 181-185.
(17) Meier, C.; Renze, J.; Ducho, C.; Balzarini, J. CycloSal-d4TMP
pronucleotides - Structural variations, mechanistic insights and
antiviral activity. Curr. Top. Med. Chem. 2002, 2, 1111-1121.
(18) Balatore, C.; McGuigan, C.; De Clercq, E.; Balzarini, J. Synthesis
and evaluation of novel amidate prodrugs of PMEA and PMPA.
Bioorg. Med. Chem. Lett. 2001, 11, 1053-1056.
(19) Ducho, C.; Balzarini, J.; Meier, C. Noninhibition of acetylcho-
linesterase by cycloSal-nucleotides. Nucl., Nucl., Nucleic Acids
2003, 22, 841-843.
(20) Meier, C.; Ducho, C.; Go¨rbig, U.; Esnouf, R.; Balzarini, J.
Interaction of cycloSal-pronucleotides with cholinesterases from
different origins - A structure-activity relationship. J. Med.
Chem. 2004, 47, 2839-2852.
(21) Holy, A.; Rosenberg, I. Synthesis of 9-[2-(phosphonylmethoxy)-
ethyl]adenine and related compounds. Collect. Czech. Chem.
Commun. 1987, 52, 2801-2809.
(22) Bronson, J. J.; Ghazzouli, I., Hitchcock, M. J. M.; Webb, R. R.,
II.; Martin, J. C. Synthesis and antiviral activity of the nucle-
otide analogue (S)-1-[3-hydroxy-2-(phosphonylmethoxy)propyl]-
cytosine. J. Med Chem. 1989, 32, 1457-1463.
(23) Popovic, M.; Sarngadharan, M. G.; Read, E.; Gallo, R. C.
Detection, Isolation, and continuous production of cytopathic
retroviruses (HTLV-III) from patients with AIDS and pre-AIDS.
Science 1984, 224, 497-500.
formation was recorded microscopically in the HIV-infected
cell cultures.
Acknowledgment. We gratefully acknowledge the
support of the Deutsche Forschungsgemeinschaft, the
Vth Framework Program of the European Commission,
the Fonds voor Wetenschappelijk Onderzoek (FWO)-
Vlaanderen and the Belgian Geconcerteerde Onderzo-
ecksacties. We thank Mrs. Ann Absillis and Mrs. Lizette
van Berckelaer for excellent technical help.
Supporting Information Available: Detailed character-
ization including high-resolution mass spectrometry of com-
pounds 10a-d, 5a-d, 11a and 6a-c. Additionally, 31P NMR
spectra are given showing the selective release of PMEA from
cycloSal- and cycloAmb-phosphonates. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
References
(1) (a) Naesens, L.; Balzarini, J.; De Clercq, E. Therapeutical
potential of PMEA as an antiviral drug. Med. Virol. 1994, 4,
147-159. (b) De Clercq, E. Broad-spectrum anti-DNA virus and
anti-retrovirus activity of phosphonylmethoxyalkylpurines and
-pyrimidines. Biochem. Pharmacol. 1991, 42, 963-972. (c) De
Clercq, E.; Holy, A.; Rosenberg, I.; Sakuma, T.; Balzarini, J.;
Maudgal, P. C. A novel selective broad-spectrum anti-DNA virus
agent. Nature 1986, 323, 464-467. (d) De Clercq, E. Cidofovir
in the treatment of poxvirus infections. Antiviral Res. 2002, 55,
1-13. (e) Arimilli, M. N.; Dougherty, J. P.; Cundy K. C.;
Bischofberger, N. Orally bioavailable acyclic nucleoside phos-
phonate prodrugs: Adefovir dipivoxil and bis(POC)PMPA. In
Advances in Antiviral Drug Design; De Clercq E., Ed.; JAI
Press: Stamford, CT, 1999; pp 69-91. (f) Robbins, B. L.;
Srinivas, R. V.; Kim, C.; Bischofberger, N.; Fridland, A. Anti-
Human Immunodeficiency Virus Activity and Cellular Metabo-
lism of a Potential Prodrug of the Acyclic Nucleoside Phospho-
nate 9-R-(2-Phosphonomethoxypropyl)adenine (PMPA), Bis(isopro-
pyloxymethylcarbonyl)PMPA. Antimicrob. Agents Chemother.
1998, 42, 612-617.
(2) Srinivas, R. V.; Robbins, B. L.; Connelly, M. C.; Gong, Y. F.;
Bischofberger, N.; Fridland, A. Metabolism an in vitro Antiret-
roviral Activities of Bis(Pivaloyloxymethyl)-Prodrugs of Acyclic
Nucleoside Phosphonates. Antimicrob. Agents Chemother. 1993,
37, 2247-2250.
(3) Starrett Jr., J. E.; Tortolani, D. R.; Hitchcock, M. J. M.; Martin,
J. C.; Mansuri, M. M. Synthesis and in vitro evaluation of a
phosphonate prodrug: bis(pivaloylosymethyl) 9-(2-phosphonyl-
methoxyethyl)adenine. Antiviral Res. 1992, 19, 267-273.
(4) Benzaria, S.; Pe´licano, H.; Johnson, R.; Maury, G.; Imbach, J.
L.; Aubertin, A. M.; Obert, G.; Gosselin, G. Synthesis, in vitro
antiviral evaluation, and stability studies of bis(S-acyl-2-thio-
ethyl) ester derivatives of 9-(2-phosphonometoxy)ethyladenine
(PMEA) as potential PMEA prodrugs with improved oral bio-
availability. J. Med. Chem. 1996, 39, 4958-4965.
(5) (a) Meier, C. CycloSal-pronucleotides - design of chemical trojan
horses. Mini Rev. Med. Chem. 2002, 2, 219-234. (b) Meier, C.;
Ruppel, M.; Vukadinovic, D.; Balzarini, J. “Lock-in”-cycloSal-
pronucleotides - a new generation of chemical trojan horses.
Mini Rev. Med. Chem. 2004, 4, 383-394. (c) Meier, C. CycloSal-
pronucleotides - design of the concept, chemistry and antiviral
acticity. In Advances in Antiviral Drug Design; De Clercq, E.,
Ed.; JAI Press: Stamford, CT, 2004; Vol. 4, pp 147-213.
(6) (a) Meier, C.; Lorey, M.; De Clercq, E.; Balzarini, J. CycloSal-
2′,3′-dideoxy-2′,3′-didehydrothymidine monophosphate (cycloSal-
d4TMP): Synthesis and antiviral evaluation of a new d4TMP
delivery system. J. Med. Chem. 1998, 41, 1417-1427. (b) Ducho,
C.; Wendicke, S.; Go¨rbig, U.; Balzarini, J.; Meier, C. 3,5-tert-
Butyl-6-fluoro-cycloSal-d4TMP - a pronucleotide with a highly
optimized masking group. Eur. J. Org. Chem. 2003, 4786-4791.
JM050641A