3,5-Di-(tert-Butyl)-6-fluoro-cycloSal-d4TMP - A Pronucleotide with a Considerably Improved Masking Group

Article abstract of DOI:10.1002/ejoc.200300537

A new, considerably improved cycloSal masking group has been developed. This new group combines four desirable properties and has been attached to the anti-HIV drug 2′,3′-dideoxy-2′,3′-didehydrothymidine (d4T, 1) to give 3,5-(di-tert-butyl)-6-fluoro-cycloSal-d4TMP (2i). This phosphate triester has a reasonable chemical half-life, highly selectively released d4TMP, has poor - if any - inhibitory effect on butyrylcholinesterase (BChE), and achieved the TK-bypass. ( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003).

Full text of DOI:10.1002/ejoc.200300537

FULL PAPER
3
,5-Di-(tert-Butyl)-6-fluoro-cycloSal-d4TMP ؊ A Pronucleotide with a
Considerably Improved Masking Group
[a]
[a]
[a]
[b]
[a]
Christian Ducho, Silke Wendicke, Ulf Görbig, Jan Balzarini, and Chris Meier*
Keywords: Antiviral agents / DNA / Nucleosides
A new, considerably improved cycloSal masking group has
triester has a reasonable chemical half-life, highly selectively
been developed. This new group combines four desirable released d4TMP, has poor − if any − inhibitory effect on
properties and has been attached to the anti-HIV drug 2Ј,3Ј-
dideoxy-2Ј,3Ј-didehydrothymidine (d4T, 1) to give 3,5-(di-
tert-butyl)-6-fluoro-cycloSal-d4TMP (2i). This phosphate
butyrylcholinesterase (BChE), and achieved the TK-bypass.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
Introduction
The properties of cycloSal nucleotides have been ‘‘fine-
tuned’’ through selection of different substitution patterns
[
4a]
in the salicyl alcohol residue 4 (cycloSal moiety). Despite
the successful TK-bypass, the structural variations were in
some cases accompanied by unwanted properties in the
phosphate triesters, including very high resistance to hy-
drolysis or reduced selectivity in the d4TMP delivery. Here
we report on a considerably improved cycloSal masking unit
attached to d4TMP.
Nucleoside analogues are widely used as antiviral and/or
antitumor agents. Unlike many other drugs, however, nu-
cleoside analogues are not bioactive as such, a prerequisite
for their biological action being their metabolism to the cor-
responding nucleoside triphosphates by cellular or viral kin-
[
1]
ases. Often, one of the three kinase-catalyzed steps is the
metabolism-limiting step, particularly the first conversion
into thymine-nucleoside monophosphate (nucleotide;
[2]
NMP), catalyzed by thymidine kinase (TK). The use of
nucleotides as therapeutics, however, is not possible, be-
cause they are not able to penetrate the cell membrane due
to their negative charges. Moreover, nucleotides are ef-
ficiently catabolized in the blood stream by unspecific
nucleotidases. In such a case, intracellular delivery of the
nucleotide from a lipophilic precursor (TK-bypass) would
Chemistry and Discussion
Four properties are desirable for cycloSal-d4TMP pronu-
cleotides:
i) They should have reasonable stability,
ii) They should deliver d4TMP highly selectively,
iii) They should achieve the TK-bypass, and
iv) They should not be inhibitors of cholinesterases such
as acetylcholinesterase (AChE) or butyrylcholinesterase
bypass these limitations.[3] It has recently been shown that
cycloSal-pronucleotides 2 are a successful class of delivery
systems for antiviral nucleotide analogues.[4] Mechan-
(BChE).
istically, the delivery of the nucleotide from the cycloSal
triesters is the result of a chemically induced releasing pro-
cess. The cycloSal approach has been applied successfully to
various nucleoside analogues, such as 2Ј,3Ј-dideoxy-2Ј,3Ј-
So far, the best cycloSal-d4TMP triesters have fulfilled at
least two of these criteria, but never all four. From our ear-
lier work it had become apparent that the introduction of
donor substituents (e.g., methyl or tert-butyl groups) into
triester derivatives provided reasonable stability and full re-
tention of the anti-HIV activity found in wild-type CEM
cells in the TK-deficient CEM cell line (Table 1).
example, 3-methyl-cycloSal-d4TMP (2b) showed a half-life
t1/2 ϭ 16 h at pH 7.3 in 25 m phosphate buffer and was
[5]
didehydrothymidine (d4T, 1), (E)-5-bromovinyl-2Ј-deox-
[
6]
[7]
[8]
yuridine, acyclovir, 2Ј,3Ј-dideoxyadenosine, and 2Ј-
[9]
ribo-fluoro-2Ј,3Ј-dideoxyadenosine. We have shown that
the corresponding cycloSal-d4TMP triesters 2 achieved the
TK-bypass through the intracellular delivery of d4TMP 3.
The cleavage mechanism is summarized in Scheme 1.
[
4,5]
As an
Ϫ
found to retain full activity in HIV-infected CEM/TK
[
[
a]
b]
Institute of Organic Chemistry, University of Hamburg,
Martin-Luther-King-Platz 6, 20146 Hamburg, Germany
Rega-Institute for Medical Research, K. U. Leuven,
Minderbroedersstraat 10, 3000 Leuven, Belgium
cells. Although its half-life is only 8 h, the same is true for
the 5-methyl derivative 2c, and retention of antiviral activity
was also observed for 3,5-dimethyl-cycloSal-d4TMP (2d,
4786
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejoc.200300537
Eur. J. Org. Chem. 2003, 4786Ϫ4791

A Pronucleotide with a Highly Optimized Masking Group
FULL PAPER
Scheme 1
Table 1. Properties of different cycloSal-d4TMP phosphate triesters 2 in relation to d4T (1)
[a]
Product ratio[b]
3:5
IC50[c]
[µM]
BChE
EC50 [µM][d]
CEM/O
HIV-1
CC50[e]
[µM]
Substituent
X
t
1/2
[h]
Ϫ
CEM/TK
HIV-2
HIV-2
2
2
2
2
2
2
2
2
2
a
b
c
d
e
f
g
h
i
H
4.4
16
8.0
29
96
8.6
73
1.1
6.2
99:1
94:6
96:4
85:15
92:8
94:6
66:34
100:0
100:0[
Ϫ
0.8
1.2
4.6
46
4.2
1.4
Ͼ 50
0.4
48
0.28Ϯ0.1
0.62Ϯ0.38
0.12Ϯ0.03
0.34Ϯ0.22
0.17Ϯ0.1
0.65Ϯ0.07
0.90Ϯ0.14
1.2Ϯ0.0
0.38Ϯ0.32
0.33Ϯ0.11
1.35Ϯ1.63
0.50Ϯ0.17
0.093Ϯ0.023
0.18Ϯ0.06
0.08Ϯ0.01
0.33Ϯ0.11
1.5Ϯ0.7
46.9Ϯ22.2
20.9Ϯ5.9
37.1Ϯ16.9
17.4Ϯ4.7
34.4Ϯ7.5
33.9Ϯ3.0
27.0Ϯ7.4
72.1Ϯ6.0
41.1Ϯ0.78
78.6Ϯ31.6
3-Me
5-Me
3,5-Me
3-tBu
5-tBu
3,5-tBu
6-F
0.087Ϯ0.012
0.18Ϯ0.03
0.093Ϯ0.023
0.18Ϯ0.11
0.14Ϯ0.02
1.1Ϯ0.14
0.17Ϯ0.12
0.16Ϯ0.06
1.68Ϯ1.12
2.0Ϯ0.0
0.85Ϯ0.25
0.60Ϯ0.23
33.3Ϯ20.8
f]
3,5-tBu; 6-F
Ϫ
d4T (1)
Ϫ
Ϫ
[
a]
[b]
Hydrolysis stability in 25 m phosphate buffer, pH 7.3 at 37 °C.
Product ratio d4TMP 3: phenyl phosphate diester 5 observed in
3
1
[c]
P NMR studies in imidazole/HCl buffer, pH 7.3.
Inhibitory potency: 50 % inhibitory concentration against butyrylcholinesterase
[
d]
[e]
[f]
from human serum. Antiviral activity: 50 % effective concentration. Cytostatic activity: 50 % cytostatic concentration. After six
weeks in the NMR tube.
t
ϭ 29 h). Detailed studies of the hydrolysis products by human serum. The physiological importance of this enzyme
/2
P NMR spectroscopy, however, revealed that the use of is still unknown. In contrast, though, human acetylcholine-
1
3
1
triester 2b gave rise to the formation of phenyl phosphate sterase (AChE) is a highly important enzyme that belongs
diester 5 as a side product in 6 % yield in addition to to the same class of hydrolases. Surprisingly, all the triesters
d4TMP (3, Scheme 1), while the 5-methyl derivative gave 2 described here were devoid of any inhibitory potency
[
11]
only 4 % of the corresponding side product. On the other against AChE (data not shown ). It was interesting to
hand, the triester 2d gave 15 % of the corresponding diester note that we observed a clear correlation of the inhibitory
[10]
5.
A second hydrolysis pathway was taking place in ad- potency against BChE and the substitution pattern in the
dition to the formation of d4TMP. This unwanted pathway cycloSal part. While triesters 2a, 2b, 2f, and particularly 2h
gained importance if the hydrolysis half-lives increased. The showed considerable BChE inhibition (IC50(BChE) ϭ 0.8 µ,
hydrolysis of 3,5-di-(tert-butyl)-cycloSal-d4TMP (2g) 1.2 µ, 1.4 µ, and 0.6 µ, respectively), double meth-
showed the formation of 34 % of diester 5 and a chemical ylation in positions 3 and 5 of the cycloSal ring of triester
half-life of 73 h. Consequently, high chemical stability and/ 2d produced a pronounced decrease in the inhibitory effect
or the formation of considerable amounts of diester 5 re- (IC50(BChE) ϭ 46 µ). With increasing steric bulk of the
Ϫ
sulted in losses of antiviral activity in the CEM/TK cell substituents, as in compound 2g (tert-butyl), no inhibition
assay (Table 1). Additionally, some cycloSal triesters (2aϪc, of BChE was observed at all (IC50(BChE) Ͼ 50 µ; Table 1).
[
11]
2e, and 2f) were found to be inhibitors of human BChE,
There are therefore distinct differences in the properties
a serine hydrolase present in different human tissues and of the described cycloSal-d4TMP triesters 2. Our aim in the
Eur. J. Org. Chem. 2003, 4786Ϫ4791
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4787

C. Ducho, S. Wendicke, U. Görbig, J. Balzarini, C. Meier
FULL PAPER
study presented here was to combine all positive properties by reduction of the corresponding salicylic acid. These diols
into one masking group. We therefore prepared 3,5-di-(tert- were treated with PCl and subsequently with d4T accord-
3
[
4a]
butyl)-6-fluoro-cycloSal-d4TMP (2i). As shown above, the ing to our previously reported procedure to give the tar-
presence of two tert-butyl groups produces high stability get cycloSal triesters 2h and 2i in 47 % and 59 % yields,
in the triester. Moreover, the tert-butyl groups block the respectively. As reported previously, triesters 2h and 2i were
interaction with BChE. The fluorine atom in position 6 was obtained as 1:1 mixtures (crude material) of diastereomers.
introduced in order to achieve two goals. Firstly, we knew The structures were unambiguously confirmed by different
from previous studies on 6-chloro-cycloSal-d4TMP that the NMR techniques and mass spectrometry.
presence of an acceptor group in this position prevents the
formation of the unwanted phenyl phosphate diester 5. The
Chemical hydrolysis of the triester 2i in 25 m phosphate
formation of 5 is due to a heterolytic bond cleavage of the buffer at pH 7.3 showed a half-life of 6.2 h. This half-life
benzyl-CϪO bond, affording a benzyl cation intermedi- (t1/2) is similar to the values for the 5-(tert-butyl)- (2f) and
ate.[ Thus, although two tert-butyl groups are present, hy- 5-methyl-cycloSal-d4TMP triester (2c) and significantly
drolysis of 2i should result in a highly selective delivery of lower than that of 3-(tert-butyl)-cycloSal-d4TMP (2e)
d4TMP. Secondly, we had observed in the same study that (Table 1). However, the introduction of the fluorine atom in
the presence of a halogen atom in position 6 reduces the position 6 of the cycloSal ring produced a pronounced de-
chemical stability of the cycloSal triester (Table 1). Because crease in the hydrolytic stability (t1/2 ϭ 1.1 h) in relation to
of this effect, we expected a considerable decrease in the the reference cycloSal derivative 2a (t ϭ 4.4 h; Table 1).
10]
1/2
31
hydrolysis half-life for the new triester 2i in relation to the In P NMR studies in imidazole/HCl buffer, pH 7.3 after
tert-butyl triesters (2e, 2f) and the di-(tert-butyl) triester 2g. six weeks, only the formation of d4TMP 3 was observed for
The target triester 2i was prepared by starting from com- both triesters (2h, 2i), which shows that the introduction of
mercially available 3-fluorophenol (6, Scheme 2). An acid- the fluorine atom completely avoids the formation of the
catalyzed alkylation with isobutene was carried out first, phenyl phosphate diesters 5 (Scheme 1). In contrast, the
according to a modified literature procedure,[ to give 4,6- triesters 2aϪg also gave varying amounts of the phenyl
12]
(tert-butyl)-3-fluorophenol (7) in 70 % yield. Treatment of phosphate diesters 5 (Table 1). Next, incubation studies
this material with a basic formaldehyde solution resulted in with human AChE showed that neither 2h nor 2i was as
the formation of the corresponding salicyl alcohol 4i in 71 inhibitory to the enzyme as the triesters reported before.
[
13]
%
yield. For comparison, 6-fluorosalicyl alcohol (4h) was Moreover, incubation with human serum confirmed that
also prepared (86 % yield). The diol was readily accessible the new triester 2i has almost no inhibitory potency towards
BChE (IC50(BChE) ϭ 48 µ). Interestingly, the 6-fluoro-sub-
stituted derivative 2h was even more inhibitory
(IC50(BChE) ϭ 0.4 µ) than the unsubstituted reference com-
pound 2a (IC50(BChE) ϭ 0.8 µ), thus confirming the domi-
nating effect of the two tert-butyl groups. Finally, in antivi-
ral tests with HIV-1- and HIV-2-infected wild-type CEM/O
cells, triester 2i exhibited pronounced antiviral potency,
even better than the activity of the parent 1. Most import-
antly, the antiviral activity was fully retained in HIV-2-in-
Ϫ
fected CEM/TK cells, so triester 2i achieved the TK-by-
pass. This retention of activity has also been observed for
other cycloSal-d4TMP triesters (2aϪe). However, some of
those triesters exhibited considerable inhibitory effects
towards BChE, while triester 2i was non-inhibitory
(Table 1). Surprisingly, although the half-life of the 6-
fluoro-triester 2h was significantly lower, 2h also showed
considerable retention of activity in the TK-deficient cells.
It may be noted that, except for 6-fluoro-cycloSal-d4TMP
(2h), all other compounds shown in Table 1 had virtually
identical 50 % cytostatic concentrations (CC50 values), ir-
respective of antiviral potency and degree of inhibition of
BChE. The CC50 value is defined as the compound concen-
trations required to reduce the number of viable cells by 50
%. It is important to mention that the cytotoxicity of tries-
ter 2h is identical to, while that of triester 2i is only slightly
higher than that of the parent nucleoside 1. Keeping the
retention of the antiviral activity in the thymidine kinase-
deficient cells in mind, these results clearly show the advan-
tage of the cycloSal-pronucelotide approach.
Scheme 2
788
4
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 4786Ϫ4791

A Pronucleotide with a Highly Optimized Masking Group
FULL PAPER
In
conclusion,
3,5-di-(tert-butyl)-6-fluoro-cycloSal- tion was warmed to room temperature, and stirring was continued
for 1 h. Subsequently, tert-butyl hydroperoxide (3.3 equiv., solution
d4TMP (2i) has for the first time fulfilled all the expected
properties and combines four desired properties of the
cycloSal approach. The masking group disclosed here is
therefore the best masking group so far and will now be
incorporated into various other nucleotide analogues. Work
towards this is currently underway in our laboratories.
in n-decane) was added at Ϫ20 °C. After the mixture had warmed
to room temperature and been stirred for 1 h, the solvent was re-
moved under reduced pressure. The resulting residue was purified
by preparative TLC (Chromatotron, 1. ethyl acetate/methanol, 9:1;
2. dichloromethane/methanol gradient). Lyophilization yielded the
products as colorless foams. Diastereomeric ratios were obtained
3
1
from the P NMR spectra.
6
-Fluoro-cycloSal-d4T Monophosphate (2h): The crude chloro-
Experimental Section
phosphite was synthesized from 4h (0.50 g, 2.8 mmol), phosphor-
ous() chloride (0.47 g, 3.4 mmol), and pyridine (0.51 g, 6.4 mmol)
in diethyl ether (20 mL) as described in the General Procedure
General Remarks: NMR spectra were recorded with Bruker AMX
1
4
00 and Bruker DRX 500 Fourier transform spectrometers. All H
(
yield 0.32 g). CycloSal d4T monophosphate 2h was synthesized
from d4T (1, 163 mg, 0.725 mmol), diisopropylethylamine (187 mg,
.45 mmol), chlorophosphite (300 mg, 1.45 mmol), and tert-butyl
13
and C NMR chemical shifts (δ) are quoted in ppm and calibrated
31
on solvent signals. The P NMR chemical shifts (proton decou-
pled) are quoted in ppm relative to H PO as the external reference.
The 19F NMR chemical shifts (proton coupled) are quoted in ppm
relative to CFCl as the external reference. The spectra were re-
1
3
4
hydroperoxide (5.5  solution in n-decane, 0.45 mL, 2.5 mmol) in
acetonitrile (15 mL). The product was obtained as a colorless foam:
yield 47 % (141 mg), mixture of two diastereomers, ratio 1.0:1.0.
R
3
corded at room temperature. Electron impact mass spectra were
measured with a VG Analytical VG/70Ϫ250S spectrometer (double
focussing). FAB high-resolution (HR) mass spectra were recorded
with a VG Analytical 70Ϫ250S spectrometer by use of an MCA
method and polyethylene glycol as support. Merck precoated 60
1
f
ϭ 0.55 (dichloromethane/methanol, 9:1). H NMR (500 MHz,
[
D
6
]DMSO): δ ϭ 1.67 (s, 1 ϫ 3 H, CH
H, CH -thymine), 4.28Ϫ4.42 (m, 2 ϫ 2 H, 2 ϫ H5Ј), 4.95Ϫ5.01
m, 2 ϫ 1 H, 2 ϫ H4Ј), 5.44Ϫ5.65 (m, 2 ϫ 2 H, 2 ϫ H-benzyl),
3
-thymine) , 1.72 (s, 1 ϫ 3
3
(
3
3
6
J
.04 (dd, JH,H ϭ 6.3, JH,H ϭ 6.3 Hz, 2 ϫ 1 H, 2 ϫ H2Ј), 6.39 (d,
F
254 plates with a 0.2 mm layer of silica gel were used for thin layer
3
3
H,H ϭ 5.7 Hz, 1 ϫ 1 H, 1 ϫ H3Ј), 6.44 (d, JH,H ϭ 5.7 Hz, 1 ϫ
chromatography (TLC). All preparative TLCs were performed with
a chromatotron (Harrison Research, Model 7924T) on glass plates
coated with 1-mm or 2-mm layers of Merck 60 PF254 silica gel
containing a fluorescent indicator. For column chromatography,
Merck silica gel 60, 230Ϫ400 mesh was used. Analytical HPLC was
performed on a MerckϪHitachi HPLC system (-7000) equipped
with a LiChroCART 125Ϫ3 column containing reversed-phase sil-
ica gel Lichrospher 100 RP 18 (5 µm) (Merck, Darmstadt, Ger-
many). The lyophilized products 2h and 2i did not give useful
microanalytical data, most probably due to incomplete combustion
of the compounds or to the presence of varying amounts of water,
but were found to be pure by rigorous HPLC analysis (gradient of
3
1
8
8
H, 1 ϫ H3Ј), 6.79Ϫ6.82 (m, 2 ϫ 1 H, 2 ϫ H1Ј), 7.01 (dd, JH,H
ϭ
3
3
.6, JH,H ϭ 8.6 Hz, 2 ϫ 1 H, 2 ϫ H3-aryl), 7.12 (ddd, JH,H
.8, JH,H ϭ 8.8, JH,F ϭ 5.7 Hz, 2 ϫ 1 H, 2 ϫ H4-aryl), 7.19Ϫ7.22
ϭ
3
4
3
3
(m, 2 ϫ 1 H, 2 ϫ H6-thymine), 7.44 (ddd, JH,H ϭ 14.7, JH,F ϭ
4
14.7, JH,H ϭ 7.7 Hz, 2 ϫ 1 H, 2 ϫ H5-aryl), 11.35 (s, 1 ϫ 1 H, 1
1
3
ϫ NH), 11.37 (s, 1 ϫ 1 H, 1 ϫ NH) ppm. C NMR (101 MHz,
]DMSO): δ ϭ 11.86 (1 ϫ CH -thymine), 11.90 (1 ϫ CH -thy-
mine), 63.70 (dd, JC,P ϭ 2.0, JC,F ϭ 5.1 Hz, 1 ϫ C-benzyl), 63.91
[D
6
3
3
2
3
2
3
2
(dd, JC,P ϭ 2.0, JC,F ϭ 5.1 Hz, 1 ϫ C-benzyl), 84.00 (d, JC,P ϭ
2
2
.6 Hz, 1 ϫ C5Ј), 84.10 (d, JC,P ϭ 2.6 Hz, 1 ϫ C5Ј), 89.28 (2 ϫ
C1Ј), 109.64 (1 ϫ C5-thymine), 109.69 (1 ϫ C5-thymine), 111.29
4
4
(d, JC,F ϭ 2.0 Hz, 1 ϫ C3-aryl), 111.49 (d, JC,F ϭ 2.0 Hz, 1 ϫ
5Ϫ100 % acetonitrile in water over 25 min, flow 0.5 mL/min). All
3
C3-aryl), 114.34 (d,
JC,F ϭ 3.6 Hz, 1 ϫ C4-aryl), 114.43 (d,
reactions were carried out under dry nitrogen except for the syn-
thesis of 4i and 7. Solvents used in the inert gas syntheses were
commercially available dry solvents stored under argon and over
molecular sieves (Fluka). Diethyl ether was dried with sodium/
benzophenone and distilled under nitrogen.
3
J
C,F ϭ 3.1 Hz, 1 ϫ C4-aryl), 127.31 (1 ϫ C2Ј), 127.34 (1 ϫ C2Ј),
2
2
1
1
30.76 (d,
J
C,F
C,F ϭ 10.7 Hz, 1 ϫ C1-aryl), 130.87 (d, J ϭ
0.2 Hz, 1 ϫ C1-aryl), 132.69 (1 ϫ C3Ј), 132.76 (1 ϫ C3Ј), 135.70
(
2 ϫ C6-thymine), 150.18 (1 ϫ C2-thymine), 150.21 (1 ϫ C2-thy-
mine), 150.69 (1 ϫ C2-aryl), 150.72 (1 ϫ C2-aryl), 157.70 (d,
1
J
C,F ϭ 249.2 Hz, 2 ϫ C6-aryl), 163.73 (1 ϫ C4-thymine), 164.05
General Procedure for the Synthesis of the cycloSal Phosphate Tri-
esters 2: A solution of the salicyl alcohol derivative (1.0 equiv.) in
dry diethyl ether was cooled to Ϫ20 °C. After addition of freshly
distilled phosphorous() chloride (1.2 equiv.) and stirring at Ϫ20
1
9
(
6
1 ϫ C4-thymine) ppm. F NMR (471 MHz, [D ]DMSO): δ ϭ
3
4
3
Ϫ116.97 (dd, JH,F ϭ 9.7, JH,F ϭ 5.2 Hz), Ϫ117.00 (dd, JH,F
ϭ
.7, JH,F ϭ 5.2 Hz) ppm. 31P NMR (202 MHz, [D
4
9
6
]DMSO): δ ϭ
Ϫ9.57, Ϫ9.72 ppm. HRMS (FAB) calcd. for C17
H
2 7
16FN O P
°C for 5Ϫ10 min , a solution of dry pyridine (2.3 equiv.) in dry
ϩ
[MH ] 411.0757, found 411.0767. HPLC: t
R
ϭ 13.7 min.
diethyl ether was added at the same temperature over a 2Ϫ3 h per-
iod. After completion of the addition, the reaction mixture was
warmed to room temperature and stirred for 1Ϫ2 h. It was kept at
Ϫ20 °C overnight for complete precipitation of pyridinium chlo-
ride. Filtration under nitrogen and evaporation of the filtrate under
reduced pressure afforded the phosphitylating agents (saligenyl
chlorophosphites) as crude products suitable to be directly used for
the synthesis of the cycloSal phosphate triesters without further
purification.
3,5-Di-(tert-Butyl)-6-fluoro-cycloSal-d4T Monophosphate (2i): The
crude chlorophosphite was synthesized from 4i (1.01 g, 3.97 mmol),
phosphorous() chloride (0.66 g, 4.8 mmol), and pyridine (0.73 g,
9
.2 mmol) in diethyl ether (30 mL) as described in the General Pro-
cedure (yield: 1.11 g). Triester 2i was synthesized from d4T (1,
20 mg, 0.535 mmol), diisopropylethylamine (140 mg, 1.08 mmol),
chlorophosphite (340 mg, 1.08 mmol), and tert-butyl hydroperoxide
5.5  solution in n-decane, 0.33 mL, 1.8 mmol) in acetonitrile
1
(
The general synthesis of cycloSal d4T monophosphates has been
(12 mL). The product was obtained as a colorless foam: Yield 59
[
3,5,13]
1
published before.
Diisopropylethylamine (DIPEA, 2.0 equiv.) % (164 mg), mixture of two diastereomers, ratio 0.8:1.0. H NMR
was added to a solution of the nucleoside analogue (1.0 equiv.) in
dry acetonitrile. The resulting solution was cooled to Ϫ20 °C and
the appropriate chlorophosphite (2.0 equiv.) was added. The solu-
(500 MHz, [D
(s, 2 ϫ 9 H, 2 ϫ CH
(s, 2 ϫ 3 H, 2 ϫ CH
6
]DMSO): δ ϭ 1.32 (s, 1 ϫ 9 H, 1 ϫ CH
-tBu), 1.35 (s, 1 ϫ 9 H, 1 ϫ CH
-thymine), 4.35Ϫ4.37 (m, 2 ϫ 2 H, 2 ϫ H5Ј),
3
-tBu), 1.34
-tBu), 1.61
3
3
3
Eur. J. Org. Chem. 2003, 4786Ϫ4791
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4789

C. Ducho, S. Wendicke, U. Görbig, J. Balzarini, C. Meier
FULL PAPER
4
2
3
.98Ϫ5.02 (m, 2 ϫ 1 H, 2 ϫ H4Ј), 5.41 (d JH,H ϭ 15.1 Hz, 1 ϫ 1 (s, 1 H, phenol-OH), 6.39 (d, JH,F ϭ 13.0 Hz, 1 H, H2), 7.19 (d,
2
4
13
H, 1 ϫ H-benzyl), 5.47 (d,
J
H,H ϭ 13.9 Hz, 1 ϫ 1 Hz, 1 ϫ H-
J
H,F ϭ 9.7 Hz, 1 H, H5) ppm. C NMR (101 MHz, CDCl
3
): δ ϭ
2
3
4
benzyl), 5.53 (dd, JH,H ϭ 15.1, JH,P ϭ 7.6 Hz, 1 ϫ 1 H, 1 ϫ H-
29.78 (CH
3
-tBu), 30.19 (d,
J
C,F ϭ 3.1 Hz, CH
C,F ϭ 3.1 Hz, C-tBu), 34.39 (C-tBu), 104.90 (d, JC,F ϭ 27.0 Hz,
benzyl), 6.03Ϫ6.08 (m, 2 ϫ 1 H, 2 ϫ H2Ј), 6.42Ϫ6.45 (m, 2 ϫ 1 C2), 125.39 (d, JC,F ϭ 7.1 Hz, C5), 128.08 (d, JC,F ϭ 11.2 Hz,
3
2
-tBu), 33.88 (d,
2
3
3
benzyl), 5.56 (dd, JH,H ϭ 13.9, JH,P ϭ 8.2 Hz, 1 ϫ 1 H, 1 ϫ H-
J
3
2
4
3
H, 2 ϫ H3Ј), 6.81Ϫ6.85 (m, 2 ϫ 1 H, 2 ϫ H1Ј), 7.22 (s, 1 ϫ 1 H, C4), 130.83 (d, JC,F ϭ 3.6 Hz, C6), 152.72 (d, JC,F ϭ 10.2 Hz,
4
C1), 159.86 (d, JC,F ϭ 246.7 Hz, C3) ppm. 19F NMR (471 MHz,
1
1
ϫ H6-thymine), 7.23 (q,
J
H,H ϭ 1.3 Hz, 1 ϫ 1 H, 1 ϫ H6-
4
): δ ϭ Ϫ113.34 (dd, 3
J
H,F ϭ 13.0, 4JH,F ϭ 9.7 Hz) ppm.
thymine), 7.26 (d, JH,F ϭ 9.5 Hz, 2 ϫ 1 H, 2 ϫ H4-aryl), 11.36 (s, CDCl
3
ϫ 1 H, 1 ϫ NH), 11.37 (s, 1 ϫ 1 H, 1 ϫ NH) ppm. 1 C NMR HRMS (FAB) calcd. for C14
3
H
21FO (M) 224.1576, found 224.1559.
1
(
6 3
101 MHz, [D ]DMSO): δ ϭ 12.07 (2 ϫ CH -thymine), 29.83 (1 ϫ
-tBu), 29.99 (d, 4
3,5-Di-(tert-Butyl)-6-fluorosalicyl
Alcohol (4i): Compound
7
CH
CH
3
-tBu), 29.86 (1 ϫ CH
3
JC,F ϭ 3.1 Hz, 2 ϫ
-tBu), 34.30 (d, 3
(5.55 g, 24.7 mmol) and sodium hydroxide (1.18 g, 27.5 mmol) were
dissolved in methanol (15 mL). Aqueous formaldehyde solution (37
%, 5.9 mL, 79 mmol formaldehyde) was added to this solution, and
the resulting mixture was stirred at room temperature for 3 days.
The reaction was monitored by TLC (dichloromethane/methanol,
9:1) and finally quenched by the addition of water and contd. hy-
drochloric acid (resulting pH 4Ϫ5). The aqueous mixture was ex-
tracted five times with dichloromethane. The combined organic lay-
3
JC,F ϭ 3.1 Hz, 2 ϫ C-tBu), 34.62 (1 ϫ C-
3
tBu), 34.67 (1 ϫ C-tBu), 63.86 (2 ϫ C-benzyl), 69.34 (d, JC,P
ϭ
.1 Hz, 1 ϫ C5Ј), 69.45 (d, 3
C,P ϭ 3.0 Hz, 1 ϫ C4Ј), 84.51 (d, JC,P ϭ 3.0 Hz, 1 ϫ C4Ј), 89.49
6
JC,P ϭ 6.1 Hz, 1 ϫ C5Ј), 84.44 (d,
3
3
J
(
1 ϫ C1Ј), 89.66 (1 ϫ C1Ј), 109.98 (1 ϫ C5-thymine), 110.07 (1 ϫ
2
3
C5-thymine), 111.27 (dd, JC,F ϭ 9.2, JC,P ϭ 4.1 Hz, 1 ϫ C1-aryl),
2
3
1
11.47 (dd, JC,F ϭ 9.2, JC,P ϭ 4.1 Hz, 1 ϫ C1-aryl), 125.21 (d,
3
J
C,F ϭ 7.1 Hz, 2 ϫ C4-aryl), 127.71 (1 ϫ C2Ј), 127.75 (1 ϫ C2Ј),
31.57 (1 ϫ C3-aryl), 131.67 (d, 4
2 4
ers were dried with Na SO and the solvent was removed in vacuo.
The resulting crude product was recrystallized from petroleum
ether to yield 4.49 g (71 %) of the title compound as a colorless
1
1
JC,F ϭ 4.1 Hz, 1 ϫ C3-aryl),
33.08 (1 ϫ C3Ј), 133.16 (1 ϫ C3Ј), 133.70 (2 ϫ C5-aryl), 136.05
(2 ϫ C6-thymine), 147.32 (2 ϫ C2-aryl), 151.08 (2 ϫ C2-thymine),
1
55.11 (d, 1
JC,F ϭ 248.2 Hz, 2 ϫ C6-aryl), 164.05 (1 ϫ C4-thy-
solid. R
(400 MHz, CDCl
1.41 (s, 9 H, CH
f
ϭ 0.74 (dichloromethane/methanol, 9:1). H NMR
1
4
19
3 3
): δ ϭ 1.34 (d, JH,F ϭ 1.0 Hz, 9 H, CH -tBu),
mine), 164.10 (1 ϫ C4-thymine) ppm. F NMR (471 MHz,
[D ]DMSO): δ ϭ Ϫ117.37 (d, JH,F ϭ 9.5 Hz), Ϫ117.50 (d, JH,F ϭ
6
9
Ϫ9.07 ppm. HRMS (FAB) calcd. for C26
5
4
4
4
3
-tBu), 5.02 (s, 2 H, H-benzyl), 7.14 (d, JH,F
ϭ
9.9 Hz, 1 H, H4), 7.92 (s, 1 H, phenol-OH) ppm. 13C NMR
.5 Hz) ppm. 1P NMR (202 MHz, [D
3
6
]DMSO): δ ϭ Ϫ8.71,
4
ϩ
(101 MHz, CDCl
CH
3
): δ ϭ 29.95 (CH
-tBu), 34.28 (d, JC,F ϭ 3.1 Hz, C-tBu), 34.95 (C-tBu), 58.13
3
-tBu), 30.48 (d, JC,F ϭ 3.1 Hz,
2 7
H32FN O P [MH ]
3
3
3
R
23.2009, found 523.2092. HPLC: t ϭ 20.2 min.
2
(
d, JC,F ϭ 10.2 Hz, C-benzyl), 112.89 (d, JC,F ϭ 17.3 Hz, C1),
3
2
1
1
1
24.64 (d, JC,F ϭ 8.1 Hz, C4), 126.94 (d, JC,F ϭ 12.2 Hz, C5),
6
3
7
-Fluorosalicyl Alcohol (4h): 6-Fluorosalicylic acid (0.55 g,
.5 mmol) was added dropwise to a suspension of LiAlH (0.27 g,
.0 mmol) in THF (20 mL) over 30 min. The reaction mixture was
4
J
C,F ϭ 3.1 Hz, C3), 154.67 (d, 3JC,F ϭ 5.1 Hz, C2),
31.88 (d,
56.86 (d,
4
1
19
JC,F ϭ 252.2 Hz, C6) ppm. F NMR (471 MHz,
4
CDCl
3
): δ ϭ Ϫ119.74 (d,
JH,F ϭ 9.9 Hz) ppm. HRMS (FAB)
stirred at room temperature for 2 h, then heated under reflux for
h. The reaction was quenched with 2  HCl, and the product was
extracted with diethyl ether. The organic layer was washed twice
with water and dried with MgSO , and the solvent was removed
calcd. for C15
H23FO
2
(M) 254.1682, found 254.1681.
1
Hydrolysis Studies on the cycloSal Phosphate Triesters: Hydrolysis
studies of cycloSal nucleotides (phosphate buffer, pH 7.3) by HPLC
analysis have been described previously.
4
under reduced pressure. Purification of the residue by chromatogra-
phy on silica gel with a gradient of methanol in dichloromethane
yielded 0.43 g (86 %) of the title compound as a colorless solid.
R
CDCl
[13]
3
1P NMR Hydrolysis Studies on the cycloSal Phosphate Triesters:
1
31
ϭ 0.52 (dichloromethane/methanol, 9:1). H NMR (400 MHz,
P NMR hydrolysis studies of cycloSal nucleotides 2aϪg have been
f
described previously.[
13]
3
3
): δ ϭ 2.48 (s, 1 H, benzyl-OH), 4.99 (s, 2 H, H-benzyl), 6.59
3
4
(
ddd, JH,H ϭ 9.5, JH,H ϭ 8.3, JH,F ϭ 1.1 Hz, 1 H, H4), 6.66 (d,
Cholinesterase Assay: For the modified cholinesterase assay used
3
3
H,H ϭ 8.3, 3
J
H,H ϭ 8.3 Hz, 1 H, H3), 7.12 (ddd,
J
J
H,F ϭ 8.3,
for inhibition studies of human BChE (serum) see ref.[
11b]
.
4JH,H ϭ 6.6 Hz, 1 H, H5), 7.83 (br., 1 H, phenol-OH) ppm.
NMR (101 MHz, CDCl
1
1
1
ppm. F NMR (471 MHz, CDCl
1
1
13
C
3
3
): δ ϭ 51.46 (d, JC,F ϭ 5.1 Hz, C-benzyl),
2
4
05.49 (d,
J
C,F ϭ 22.4 Hz, C5), 111.29 (d,
15.40 (d, JC,F ϭ 18.3 Hz, C4), 128.95 (d, JC,F ϭ 11.2 Hz, C1),
JC,F ϭ 2.0 Hz, C3),
Acknowledgments
3
2
3
1
Financial support by the Deutsche Forschungsgemeinschaft, Ger-
many, the Fonds der Chemischen Industrie, Germany and the
European Commission Ren e´ Descartes Prize 2001 is gratefully
acknowledged. C.D. is grateful to a Ph.D. fellowship from the
Fonds der Chemischen Industrie/BMBF, Germany.
57.27 (d, JC,F ϭ 8.1 Hz, C2), 162.15 (d, JC,F ϭ 243.1 Hz, C6)
19
): δ ϭ Ϫ119.45 (dd, 3
H,F
J ϭ
3
0.0, 4
JH,F ϭ 10.0 Hz) ppm. MS (EI) calcd. for C H FO M ϭ
7 7 2
ϩ
42, found m/z ϭ 142 (M , 54 %), 124 (76), 96 (100), 77 (5), 51 (5).
,6-Di-(tert-Butyl)-3-fluorophenol (7): Isobutene gas was bubbled
4
[
[
1]
2]
through 3-fluorophenol (6) (4.10 g, 36.6 mmol) for 5 min at 40 °C.
Afterwards, concd. sulfuric acid (0.34 g, 3.5 mmol) was added and
the inflow of isobutene was continued for 1 h. After the addition
of water, dichloromethane was added and the phases were sepa-
rated. The organic phase was neutralized with satd. sodium hydro-
gencarbonate solution and dried with Na SO . The solvent was
2 4
removed in vacuo and the resulting residue was purified by column
chromatography (petroleum ether/dichloromethane, 4:1), yielding
J. Balzarini, Pharm. World Sci. 1994, 16, 113Ϫ126.
J. Balzarini, P. Herdewijn, E. De Clercq, J. Biol. Chem. 1989,
264, 6127Ϫ6133.
[
[
3]
C. Meier, Synlett 1998, 233Ϫ242, C. R. Wagner, V. V. Iyer, E.
J. McIntee, Med. Res. Rev. 2000, 20, 417Ϫ451.
4] [4a]
[4b]
C. Meier, Mini Rev. in Med. Chem. 2002, 2, 219Ϫ234.
J.
Balzarini, S. Aquaro, T. Knispel, C. Rampazzo, V. Bianchi, C.-
F. Perno, E. De Clercq, C. Meier, Mol. Pharmacol 2000, 58,
9
28Ϫ935.
5
(
.73 g (70 %) of the title compound as a colorless solid. R
f
ϭ 0.14
petroleum ether/dichloromethane, 4:1). H NMR (500 MHz,
CDCl ): δ ϭ 1.37 (s, 9 H, CH -tBu), 1.42 (s, 9 H, CH -tBu), 4.76
[5] [5a]
C. Meier, M. Lorey, E. De Clercq, J. Balzarini, J. Med.
1
[5b]
Chem. 1998, 41, 1417Ϫ1427.
Clercq, J. Balzarini, Bioorg. Med. Chem. Lett. 1997, 7, 99Ϫ104.
C. Meier, M. Lorey, E. De
3
3
3
4790
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 4786Ϫ4791

A Pronucleotide with a Highly Optimized Masking Group
FULL PAPER
C. Meier, J. Renze, C. Ducho, J. Balzarini, Current Topics in
Med. Chem. 2002, 2, 1111Ϫ1121.
[
6] [6a]
[10]
C. Meier, A. Lomp, A. Meerbach, P. Wutzler, J. Med. Chem.
6b]
2
002, 45, 5157Ϫ5172, [ C. Meier, A. Lomp, A. Meerbach, P.
Wutzler, ChemBioChem 2001, 283Ϫ285.
[11] [11a]
C. Ducho, J. Balzarini, C. Meier, Nucleosides, Nucleot-
[
7] [7a]
C. Meier, L. Habel, F. Haller-Meier, A. Lomp, M. Herder-
[11b]
ides & Nucleic Acids 2003, in press,
C. Meier, C. Ducho,
ich, R. Klöcking, A. Meerbach, P. Wutzler, Antiviral Chem. &
U. Görbig, R. Esnouf, E. De Clercq, J. Balzarini, J. Med.
Chem. 2002, submitted.
B. Rickborn, D. A. May, A. A. Thelen, J. Org. Chem. 1964,
[7b]
Chemotherapy 1998, 9, 389Ϫ402,
A. Meerbach, R.
Klöcking, C. Meier, A. Lomp, B. Helbig, P. Wutzler, Antiviral
Res. 2000, 45, 69Ϫ77.
[
[
12]
13]
29, 91Ϫ93.
[
[
8]
9]
C. Meier, T. Knispel, E. De Clercq, J. Balzarini, J. Med. Chem.
C. Ducho, J. Balzarini, L. Naesens, E. De Clercq, C. Meier,
Antiviral Chem. & Chemother. 2002, 13, 129Ϫ141.
1
999, 42, 1604Ϫ1614.
C. Meier, T. Knispel, V. E. Marquez, M. A. Siddiqui, E. De
Clercq, J. Balzarini, J. Med. Chem. 1999, 42, 1615Ϫ1624.
Received August 28, 2003
Eur. J. Org. Chem. 2003, 4786Ϫ4791
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4791