Nucleotide Delivery from cycloSaligenyl-3′-azido-3′-deoxythymidine Monophosphates (cycloSal-AZTMP)

Article abstract of DOI:10.1002/(SICI)1099-0690(199805)1998:5<837::AID-EJOC837>3.0.CO;2-7

The application of our cycloSaligenyl- (cycloSal) pronucleotide concept to the approved anti-HIV dideoxynucleoside 3′-azido-3′-deoxythymidine AZT (1) is reported. This pronucleotide concept has been designed to deliver the corresponding 3′-azido-3′-deoxythymidine monophosphate AZTMP (2) by selective chemical hydrolysis from the lipophilic precursors cycloSal-AZTMP 4a-h. All derivatives 4a-h were synthesized using differently substituted salicyl alcohols 7a-h as starting materials. In hydrolysis studies, compounds 4 decomposed selectively releasing AZTMP (2) and the salicyl alcohols 7 following the designed tandem reaction. Furthermore, due to the electronic properties introduced by substituents, the half-lives of the triesters 4 could be ajusted over a wide range. Phosphotriesters 4 exhibited considerable biological activity in HIV-1 and HIV-2 infected wild-type human T-lymphocyte (CEM/O) cells, whereas, contrary to our expectations, nearly all activity was lost in HIV-2 infected thymidine-kinase-deficient CEM cells.

Full text of DOI:10.1002/(SICI)1099-0690(199805)1998:5<837::AID-EJOC837>3.0.CO;2-7

FULL PAPER
Nucleotide Delivery from cycloSaligenyl-3Ј-azido-3Ј-deoxythymidine
Monophosphates (cycloSal-AZTMP)[1]
Chris Meier*a, Eric De Clercqb and Jan Balzarinib
Institut für Organische Chemie der Julius-Maximilians-Universität Würzburga,
Am Hubland, D-97074 Würzburg, Germany
Fax: ϩ49 (0)931/888-4606
E-mail: meier@chemie.uni-wuerzburg.de
Rega Institute for Medical Research, Katholieke Universiteit Leuvenb,
B-3000 Leuven, Belgium
Received September 16, 1997
Keywords: AZT / cycloSal-pro-nucleotide / Prodrugs / Nucleotide delivery / HIV chemotherapy
The application of our cycloSaligenyl- (cycloSal) pronucle-
otide concept to the approved anti-HIV dideoxynucleoside
compounds 4 decomposed selectively releasing AZTMP (2)
and the salicyl alcohols 7 following the designed tandem re-
3Ј-azido-3Ј-deoxythymidine AZT (1) is reported. This pro- action. Furthermore, due to the electronic properties intro-
nucleotide concept has been designed to deliver the corre- duced by substituents, the half-lives of the triesters 4 could
sponding
3Ј-azido-3Ј-deoxythymidine
monophosphate
be ajusted over a wide range. Phosphotriesters 4 exhibited
AZTMP (2) by selective chemical hydrolysis from the lipo-
considerable biological activity in HIV-1 and HIV-2 infected
philic precursors cycloSal-AZTMP 4a؊h. All derivatives wild-type human T-lymphocyte (CEM/O) cells, whereas,
4a؊h were synthesized using differently substituted salicyl
alcohols 7a؊h as starting materials. In hydrolysis studies,
contrary to our expectations, nearly all activity was lost in
HIV-2 infected thymidine-kinase-deficient CEM cells.
Introduction
point mutations within the HIV-1 reverse transcriptase (RT)
gene of the virus and, on the other hand, to the low suscep-
tibility of the resistant cells to the drug due to a decreased
activity of the phosphorylating enzyme thymidine-kinase
(TK)[9]. Consequently, the direct application of the nucleo-
tide 3Ј-azido-3Ј-deoxythymidine monophosphate (AZTMP,
2) would circumvent the metabolization-limiting enzyme
TK, leading to TK-bypass (Scheme 1). However, due to the
high polarity of AZTMP (2), this compound is not able to
penetrate biological membranes and hence circumvent high
intracellular concentrations of this phosphorylated deriva-
tive. Furthermore, 2 is rapidly dephosphorylated in blood
serum. Thus, a key goal in trying to improve the therapeutic
potential of AZT and other nucleoside analogues could be
the delivery of the corresponding nucleotide from neutral,
lipophilic prodrugs[10] of the general type 3 (Pronucleotide
Approach[11]; Scheme 1).
To achieve this goal, generally, uncharged nucleotide tri-
esters are used as membrane-permeable nucleotide precur-
sors[12][13]. The major differences between the approaches
are the delivery mechanisms of the nucleotides. While al-
most all approaches based on chemical hydrolysis reported
to date have proved to be unable to deliver the nucleotide
selectively and consequently serve only as nucleoside “de-
pots”, the concepts based on enzymatic activation as in the
bis-POM phosphotriesters[14], the bis-SATE phosphotri-
Nucleoside analogs are among the most effective
classes of antiviral agents investigated so far for the treat-
ment of HIV infections[2]. Since the discovery of 3Ј-azido-
3Ј-deoxythymidine (AZT, Zidovudine, Retrovir , 1)[3] as
the first drug for the treatment of AIDS, considerable ef-
forts have been made to develop more active and less toxic
inhibitors of the critical virus-associated enzymes. Most of
these nucleoside analogues are 2Ј,3Ј-dideoxynucleosides like
AZT (1). The general mode of action of these drugs is the
inhibition of HIV reverse transcriptase[4][5][6] or incorpo-
ration into the growing DNA chain, which results in DNA
chain termination[4]. However, the anti-HIV dideoxynucleo-
sides are not active as such. After membrane penetration,
intracellular conversion of the nucleoside analogues to their
5Ј-mono-, 5Ј-di- or 5Ј-triphosphates is essential for the ex-
pression of the biological activity. This biotransformation
is brought about by host cell kinases in three steps via the
nucleoside mono- and diphosphates[7]. Although dideoxy-
nucleosides offer great promise for the treatment of HIV
infections, several shortcomings limit realization of their
full therapeutic efficacy. One problem is the limited ability
of certain dideoxynucleosides to undergo biotransformation
into the triphosphates[8] which is at least as important as
the affinity of these compounds for the target enzyme.
Furthermore, the resistance of the human immunodefici- esters[15] and the aryloxyphosphoramidates[16] demon-
ency virus to the clinically used antiviral deoxynucleoside strated the successful intracellular delivery of free nucleot-
AZT (1) is, on the one hand, directly related to multiple ides from highly lipophilic precursors.
Eur. J. Org. Chem. 1998, 837Ϫ846
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1434Ϫ193X/98/0505Ϫ0837 $ 17.50ϩ.50/0
837

C. Meier, E. De Clercq, J. Balzarini
FULL PAPER
Scheme 1. Principle of an AZTMP cycloSal-pro-nucleotide dehydes 5 or salicylic acids 6 in 75Ϫ90% yield. 3,5-Dimeth-
approach
ylsalicylaldehyde 5h was prepared by ortho-formylation of
the corresponding phenol using dichloromethyl methyl
ether in the presence of TiCl4. The reaction sequence to
yield the cycloSal-phosphotriesters 4 is summarized in
Scheme 3.
Scheme 3. Synthetic pathway to the cycloSal-AZTMP triesters
4a؊h. Reaction conditions: a) NaBH4 for 5 (LiAlH4 for
6); b) PCl3, pyridine, Et2O, Ϫ10°C, 2 h; c) AZT 1,
DIPEA, CH3CN, 0°C, 20 min.; d) TBHP, CH3CN, r.t.,
30 min.
As part of our ongoing program to develop efficient pro-
nucleotide systems[17], we have reported on the design and
synthesis of an entirely new pro-nucleotide approach based
on a pH-driven selective chemical hydrolysis: the cyclosali-
genyl nucleoside monophosphates (cycloSal-NMP)[18]. This
new concept has been successfully introduced with the anti-
HIV nucleosides ddT[19] and d4T[20] for the TK-bypass as
well as with ddA[21] for the adenosine deaminase-bypass[22]
.
Here, we report on the synthesis and biological evalu-
ation of cycloSal-3Ј-azido-3Ј-deoxythymidine monophos-
phates 4a؊h (cycloSal-AZTMP, Scheme 2) as potential neu-
tral prodrugs AZTMP (2) of the approved drug AZT (1).
Scheme 2. Structures of AZT (1), AZTMP (2) and the general
structure the cycloSal-AZTMP phosphotriesters 4
After chromatographic purification (60Ϫ85% yield), the
cycloSal-AZTMPs 4a؊h were characterized by means of
1H-, 13C-, and 31P-NMR, UV spectroscopy, as well as by
electrospray mass spectrometry (ESI, negative mode). As
expected, the phosphotriesters displayed two closely spaced
signals in the 31P-NMR spectra (1:1 mixtures; δ ϭ Ϫ8.2 to
Ϫ9.8), corresponding to the presence of diastereomers that
result from mixed stereochemistry at the phosphorus center.
The purity of the cycloSal-AZTMPs 4a؊h was checked by
analytical HPLC. For the biological assays, small amounts
of the cycloSal-AZTMP triesters were additionally purified
by semi-preparative HPLC using acetonitrile/water eluents
in order to ensure complete elimination of all traces of AZT
(1). Even using semi-preparative HPLC, we were unable to
separate the diastereomers. After HPLC purification, the
phosphotriesters 4 were obtained by lyophilization of the
fractions.
Furthermore, we report on their lipophilic properties
(partition coefficients; Pa values), on hydrolysis studies in
aqueous buffers at different pH values, as well as on hy-
drolysis studies under more biologically adapted conditions
using RPMI-1640 culture medium containing 10% heat-in-
activated fetal calf serum (FCS) at 37°C.
Determination of the Partition Coefficients (Pa Values):
The partition coefficients (Pa values) of the cycloSal-
AZTMPs 4a؊h were determined in 1-octanol/sodium phos-
phate buffer (pH 6.8). Triesters 4a؊h were partitioned be-
Results and Discussion
The title compounds 4a؊h were synthesized using reac- tween 1-octanol and phosphate buffer for 5 min. and after
tive phosphorus(III) chemistry. The salicyl alcohols 7 were centrifugation the relative concentration in each solvent
obtained by reduction of the corresponding salicylal- phase was determined by reversed-phase HPLC analysis.
838
Eur. J. Org. Chem. 1998, 837Ϫ846

Nucleotide Delivery from cycloSaligenyl-3Ј-azido-3Ј-deoxythymidine Monophosphates
FULL PAPER
This very simple method of partition coefficient determi- Table 1. Hydrolysis in different aqueous buffers and Pa values of
cycloSal-AZTMPs 4aϪh
nation allowed a qualitative estimation of the lipophilicities
of compounds 4a؊h[23]. Unlike most nucleosides, AZT 1
Hydrolysis (t1/2
)
Hydrolysis (t1/2
)
Pa
can enter mammalian cells by passive, non-facilitated dif-
4
in aqueous buffers at 37°C
in CM at 37°C[e] value[g]
fusion[24]. Consequently, the determined Pa value of 1 and
pH 4.6[a] pH 6.9[b] pH 7.3[c] pH 8.9[d] without with
(1.06[25]) was used as a reference for the potential ability of
1
FCS[f]
(h)
FCS[f]
(h)
(h)
(h)
(h)
(h)
compounds 4 to diffuse passively through membranes. As
can be seen in Table 1, the measured Pa values were six- to
sixty-fold higher (Pa ϭ 6.5Ϫ62.2) as compared to the value 4b
for AZT (1).
4a
1.8
6.7
1.5
6.4
0.2
0.1
n.d.[h] n.d.[h]
6.5
41.5
12.5
13.2
7.3
0.7
3.7
0.25
0.83
0.9
0.3
2.3
3.9
0.8
4.1
4.9
9.4
n.a.[i]
0.3
2.3
4.0
0.8
4.1
4.8
9.3
n.a.[i]
4c
4d
4e
4f
4g
4h
1
25.0 24.5
28.7 28.3
4.1
Kinetic Studies: The nucleotide delivery system described
here has been designed to selectively release the nucleotide
and the two-component masking group by a controlled,
chemically-induced tandem reaction involving successive,
coupled cleavage of the phenyl and the benzyl ester groups
of the cycloSal-phosphotriester. Within this concept, the
coupling of the two ester hydrolyses is very important be-
cause it has previously been reported that the chemical[26]
and sometimes even the enzymatic hydrolysis[27] of phos-
phodiesters may be extremely difficult. In our cycloSal pro-
nucleotide concept, the lower stability of the phenyl as com-
pared to the benzyl phosphate ester allows us to discrimi-
nate between these two phosphate ester bonds in the initial
step yielding a benzylphosphodiester from of the cycloSal-
phosphotriester. As a consequence, the initial hydrolysis of
the phenyl ester causes an important activation of the re-
maining masking group and induces a spontaneous cleav-
age to yield the nucleotide[17]. The hydrolysis pathway has
10.1
9.5
1.6
0.5
29.0 28.3
69.9 68.5
102.1 98.2
4.1
1.0
30.7
26.9
62.2
10.7
14.7
n.a.[i]
1.4
2.1
n.a.[i]
n.a.[i]
n.a.[i]
1.09
[a]
[b]
50 mmol acetate buffer, pH 4.6. Ϫ 50 mmol TRIS buffer, pH
[c]
[d]
6.9. Ϫ 50 mmol phosphate buffer, pH 7.3. Ϫ 50 mmol borate
[e]
[f]
buffer, pH 8.9. Ϫ CM: RPMI-1640 culture medium. Ϫ FCS:
[g]
[h]
fetal calf serum. Ϫ Pa: partition coefficient. Ϫ
Not determi-
[i]
ned. Ϫ Not available.
strong influence of the substituent electronic effect is dem-
onstrated by the 3-methoxy and the 3-methyl groups: in
contrast to the corresponding 5-substituted derivatives,
where the methoxy derivative is equally stable as compared
to the methyl derivative in all buffers, the situation is differ-
ent for the 3-substituted phosphotriesters. The reason for
this is that in the 5-position the mesomeric electronic effects
are more important than the inductive effect, whereas in
the 3-position the inductive effects of the methoxy group
overcompensates the mesomeric effect, with the conse-
quence of a considerable increase of susceptibility of this
compound towards hydrolysis. These examples clearly dem-
onstrate that we have a tool at our disposal whereby the
hydrolysis kinetics of the phosphotriesters 4 can be con-
trolled by varying the substituents on the aryl ring.
Moreover, all the results are in full agreement with the
designed hydrolysis pathway of a selective cleavage of the
phenyl ester to give 2-hydroxybenzylphosphodiester 10 (step
a) and a subsequently induced spontaneous cleavage of 10,
presumably via a CϪO bond cleavage following the Grob
fragmentation mechanism, followed by water quenching[28]
of the remaining masking group, leaving AZTMP (2) and
salicyl alcohols 7 (step b) (Scheme 4).
been verified by NMR spectroscopy[20]
.
In order to verify the selective delivery of AZTMP (2),
hydrolysis experiments of the cycloSal-AZTMPs 4a؊h were
carried out in different aqueous buffers and RPMI-1640
culture medium with or without heat-inactivated fetal calf
serum. First, the triesters 4 were hydrolyzed in 50 m phos-
phate buffer, pH 7.3, at 37°C as a model of the physiologi-
cal milieu. Furthermore, the pH dependence of the hydroly-
sis was studied by carrying out experiments in borate buffer
(50 m, pH 8.9), TRIS buffer (50 m, pH 6.9) and in ace-
tate buffer (50 m, pH 4.6). The degradation of 4a؊h was
followed by means of reversed-phase HPLC analysis. The
half-lives were determined by quantifying the decreasing
peaks of 4 vs. time and are summarized in Table 1.
Most importantly, cycloSal-triesters 4a؊g were degraded
following pseudo-first-order kinetics yielding only AZTMP
2 and the salicyl alcohols in all the studied media and the
results confirmed the anticipated pH dependence of the hy-
drolysis. At pH < 7, only the 3,5-dimethyl-substituted de-
rivative 4h additionally yielded some (30%) of the ЉwrongЉ
phenyl AZT-phosphodiester, which is stable to further hy-
drolysis. As can be seen in Table 1, for a given substituent
the stability of the phosphotriesters 4a؊h do not differ
greatly under mild acidic conditions ( cf. the hydrolyses at
pH 4.6 vs. pH 6.9) but are still considerably varying for the
different substituents, whereas the half-lives decrease mark-
edly at more basic pH values (hydrolysis at pH 8.9). More
strikingly, we observed a clear correlation between the elec-
tronic effects of the various salicyl alcohol substituents and
the hydrolysis half-lives in different aqueous buffers. The
Scheme 4. The hydrolysis pathway of cycloSal-AZTMP phospho-
triesters 4a؊h
Eur. J. Org. Chem. 1998, 837Ϫ846
839

C. Meier, E. De Clercq, J. Balzarini
FULL PAPER
Furthermore, an interesting change in the rate-determin-
From the results summarized in Table 1, the donor-sub-
ing step of this tandem reaction was observed: whereas we stituted compounds of the cycloSal-AZTMPs 4d, f؊h in
detected the intermediate 2-hydroxybenzylphosphodiester particular, which exhibit half-lives of 4Ϫ14 hours in phos-
10 in the hydrolysis of the 5-nitro- and the 5-chloro-substi- phate buffer at pH 7.3, seem to be interesting candidates
tuted cycloSal-AZTMP derivatives 4a and 4b (Figure 1), we for biological evaluation.
were unable to detect this intermediate in any reaction in-
Further hydrolysis studies were carried out in RPMI-
volving the donor-substituted phosphotriesters (Figure 2). 1640 culture medium with and without 10% heat-inacti-
As indicated in Figure 1, the formation of AZTMP from vated fetal calf serum (FCS) at 37°C. This medium is nor-
5-chloro-cycloSal-AZTMP (4b) shows a sigmoidal increase, mally used for the cultivation of human T-lymphocytes used
whereas it shows an exponential increase in the hydrolysis in in vitro anti-HIV tests[29]. The hydrolysis products in
of 5-methyl-cycloSal-AZTMP 4f. The same was observed RPMI-1640 culture medium were again exclusively
for the salicylic alcohol. This observation points to the AZTMP (2) and the diols 7a؊h, but the half-lives were
transformation of the phosphodiester 10b to AZTMP 2 be- slightly shorter than in phosphate buffer. The shorter half-
ing the rate-limiting step in the former case and is in ac- lives in the RPMI-1640 medium are certainly an effect of its
cordance with the conversion of the phosphotriester 4f to more basic pH value (pH 7.8). In the RPMI-1640 medium
the phosphodiester 10f being the rate-limiting step in the supplemented with 10% FCS, we also detected the dephos-
latter case. This behavior can again be explained by the in- phorylation of AZTMP into AZT by residual phosphatases
fluence of the different substituents: the more electron-do- in the serum. On the basis of the measured half-lives, the
nating the substituent, the faster is the degradation of the donor-substituted derivatives of 4 should be sufficiently
intermediate 2-hydroxybenzylphosphodiester 10.
stable to serve as intracellular “depots” of AZTMP (2). Fi-
nally, it should be mentioned that so far we have not found
any evidence for a contribution from an enzymatic degra-
dation of the cycloSal-phosphotriesters in the RPMI-1640/
FCS hydrolyses. Thus, under the conditions used in these
studies, the degradation of our pro-nucleotides seems to be
purely chemical. Nevertheless, additional hydrolysis studies
in sera, in plasma and in cell extracts will be the subject
of further studies in our laboratory and the results will be
reported in due course.
Figure 1. Hydrolysis of 5-chloro-cycloSal-AZTMP (4b) in phos-
phate buffer, pH 7.3
Antiviral Evaluation: The cycloSal-d4TMP and cycloSal-
ddAMP phosphotriesters have already demonstrated the ef-
fectiveness of our new pro-nucleotide system in inhibiting
HIV-1 and HIV-2 replication in rapidly dividing human T-
lymphoblastic leukaemia cells (CEM/O cells)[20][21]. More-
over, particularly striking was the complete retention of bio-
logical activity of donor-substituted cycloSal-d4TMP tri-
esters in thymidine kinase-deficient CEM cells, as well as
the structure-bioactivity correlation: The stronger the elec-
tron-donating effect of the substituent, the better was the
antiviral activity against HIV-1 and HIV-2 infected wild-
type CEM/O cells and in mutant thymidine-kinase-deficient
CEM/TKϪ cells[20]. These results confirmed the highly
selective intracellular delivery of d4TMP and ddAMP and
demonstrated efficient TK- and ADA-bypasses, respec-
tively.
Figure 2. Hydrolysis of 5-methyl-cycloSal-AZTMP (4f) in phos-
phate buffer, pH 7.3
For these reasons, the antiviral activities of the cycloSal-
AZTMPs 4 were also evaluated against HIV-1 and HIV-2
infected CEM/O cells and HIV-2 infected CEM/TKϪ cells
(Table 2). The test compounds were proven to be free of the
parent nucleoside 1 by means of analytical HPLC.
The parent nucleoside AZT (1) proved to be extremely
active against HIV-1 and HIV-2 induced cytopathicity in
CEM/O cells (EC50: 0.006 µ). As anticipated, AZT (1) vir-
tually loses its antiviral potential when evaluated against
HIV-2 in CEM/TKϪ cells, owing to the absence of its bioac-
tivating enzyme thymidine-kinase. As can be seen in Table
2, all cycloSal-AZTMPs 4aϪh proved to be as active as
AZT (1) against HIV-1 or HIV-2 induced cytopathicity in
840
Eur. J. Org. Chem. 1998, 837Ϫ846

Nucleotide Delivery from cycloSaligenyl-3Ј-azido-3Ј-deoxythymidine Monophosphates
FULL PAPER
Table 2. Antiviral evaluation of the inhibitory effect of HIV-1 and half-lives of hydrolysis could be readily controlled by varia-
HIV-2 replication of cycloSal-AZTMPs 4aϪh in CEM/O cells and
tion of the substitution on the aromatic ring. All com-
pounds exhibited considerably higher partition coefficients
in thymidine kinase-deficient CEM/TKϪ cell
than the parent nucleoside AZT (1). Surprisingly, and in
Antiviral Activity
Cytotoxicity
4
and
1
EC50 (µM)[a]
CC50 (µM)[b]
contrast to the corresponding derivatives of the nucleoside
analog d4T, the biological activity of 4aϪh found in the
wild-type cells was not retained in TK-deficient cells.
Further work is currently in progress aimed at elucidating
the underlying reasons for this loss of activity, as this is an
important question with respect to the exploration of the
pro-nucleotide concept.
X
CEM/O
HIV-1
CEM/O
HIV-2
CEM/TKϪ
HIV-2
4a
4b
4c
4d
4e
4f
4g
4h
1
5-NO2 0.008
5-Cl
5-H
0.02
>100
>40
>20
30
79
34
0.005
0.004
0.006
0.005
0.021
0.009
0.006
0.013
0.017
0.005
70
5-OCH3 0.006
3-OCH3 0.009
5-CH3 0.005
3-CH3 0.006
3,5-CH3 0.007
66
>100
21
>100
79
We gratefully acknowledge support by the Adolf-Messer-Stif-
tung, the Deutsche Forschungsgemeinschaft, the Fonds der Che-
mischen Industrie and the Human Capital Mobility and Biomedical
Research Programme of the European Commission. We thank Mrs.
Ann Absillis for excellent technical help.
15
7
40
41
Ϫ
0.006
>100
>100
[a]
[b]
50% effective concentration. Ϫ 50% cytotoxic concentration.
the wild-type cell line. Nevertheless, no correlation of the
antiviral activity with either the hydrolytic stability in the
aqueous buffers and in RPMI-1640 culture medium, or
with the lipophilicity of 4aϪh (Pa values) was observed. To
our surprise, and despite the pronounced antiviral activity
in the wild-type cell line, the biological activity was almost
completely lost in the HIV-2 infected CEM/TKϪ cells. Only
Experimental Section
General: Chromatography: Chromatotron (Harrison Research
7924), silica gel 60PF (Merck, “gipshaltig”), UV detection at 254
nm. Ϫ Column chromatography: Merck silica gel 60 (40Ϫ60 µm).
Ϫ TLC: Analytical thin-layer chromatography was performed on
Merck precoated aluminum plates 60 F254 with a 0.2 mm layer of
silica gel containing a fluorescent indicator; sugar-containing com-
the donor-substituted triesters 4fϪ4h exhibited some antivi- pounds were visualized with the sugar spray reagent (0.5 ml 4-
ral activity in the TKϪ cells (5 to 14-times more active than
methoxybenzaldehyde, 9 ml ethanol, 0.5 ml concentrated sulfuric
acid and 0.1 ml glacial acetic acid) by heating with a hot air blower
or a hot plate. Ϫ HPLC: (Merck-Hitachi) semi-preparative HPLC:
LiChroCART 250-10 containing LiChrospher 100 RP-18 (10 µm);
analytical HPLC: LiChroCART 250-4 with LiChrospher 100 RP-
18 (5 µm) standard gradient: 18Ϫ100% CH3CN in water (0Ϫ16
min), 100% CH3CN (16Ϫ19 min), 18% CH3CN (19Ϫ31 min), flow
AZT (1)). These results are in complete contrast to the ob-
served antiviral activity of the related d4TMP and ddAMP
derivatives studied previously[20][21]. As yet, we have no ex-
planation for the failure of the cycloSal-AZTMP derivatives
in the CEM/TKϪ anti-HIV test. On the other hand, it
should be mentioned that comparable effects were observed
rate 0.5 ml/min., UV detection at 269 nm. Ϫ Distillation: Büchi
in the phosphoramidate approach of McGuigan et al.[16a]
and the bis-SATE approach of Imbach et al.[15a]: the AZT-
Kugelrohr apparatus. Ϫ NMR spectra were recorded using: 1H
NMR: Bruker AMX 400 at 400 MHz (tetramethylsilane or DMSO
phosphoramidates and the bis-SATE-AZTMP phospho- as internal standard). Ϫ 13C NMR: Bruker AM 250 at 63 MHz
(CDCl3 or DMSO as internal standard). Ϫ 31P NMR: Bruker
triesters also showed a marked decrease (100-fold and 10-
fold, respectively) in biological activity on going from the
wild-type cells to the TK-deficient cells. As with our system,
1
AMX 400 at 162 MHz (H3PO4 as external standard). All H- and
13C-NMR chemical shifts (δ) are quoted in parts per million (ppm)
downfield from tetramethylsilane. The 31P-NMR chemical shifts
both of the approaches have been proven to give good re-
are quoted in ppm using H3PO4 as external reference. The spectra
sults with d4T[15a][16a] and ddA[15c][16e]. Consequently, it
were recorded at room temperature. Ϫ UV/Vis spectroscopy:
seems that there is a particular metabolic limitation that
Varian Cary 1E. Ϫ IR spectroscopy: Perkin-Elmer 1600 Series FT-
IR. Ϫ Elemental analyses: Heraeus CHN-O Rapid. The resulting
lyophilized triesters 4 did not give acceptable microanalytical data,
probably due to incomplete combustion of the compounds, but
prevents the retention of activity in the TKϪ cells. Another
possibility that should be considered is a different cellular
uptake of the cycloSal-d4TMP and cycloSal-ddAMP phos-
photriesters as compared to the corresponding cycloSal- were found to be pure by rigorous HPLC analysis (three different
gradients), high-field multinuclear NMR spectroscopy, and electro-
spray mass spectrometry. Ϫ Mass spectrometry: electrospray MS:
Fisons VG platform II, negative mode, acetonitrile as solvent;
FAB-MS: VG Analytical ZAB 2 SEQ.
AZTMP derivatives. Furthermore, it has been reported that
AZTMP (2) accumulates at a sufficiently high level for re-
adily measurable levels of the nucleotide to diffuse back
into the extracellular medium. We are currently investigat-
ing whether the activity can be recovered in AZT-resistant
cells.
General Procedure for the Synthesis of the Cyclic Salicylchloro-
phosphanes 8aϪh: The cyclic chlorophosphanes were prepared ac-
cording to a procedure published by Wissner et al.[31]. Briefly, the
salicyl alcohols 7, which have been prepared by reduction from the
corresponding salicylaldehydes or the salicylic acids using NaBH4
or LiAlH4 in multi-gram scales (ca. 10 g), were dissolved in dry
diethyl ether at Ϫ10°C. Then, 1.2 equiv. of phosphorus trichloride
was added. To this mixture, a solution of pyridine in diethyl ether
was added dropwise over a period of 30 min. After stirring for 1
Conclusion
In summary, as with the previously reported cycloSal-
d4TMP and the cycloSal-ddAMP triesters, hydrolysis stud-
ies clearly show that our prodrug approach is suitable for
the delivery of AZTMP (2) from cycloSal-AZTMPs 4aϪh
by a non-enzymatic activation at physiological pH. The h, the reaction mixture was allowed to warm to room temperature
Eur. J. Org. Chem. 1998, 837Ϫ846 841

C. Meier, E. De Clercq, J. Balzarini
FULL PAPER
and stirring was continued for further 2 h. The mixture was 13C NMR ([D6]DMSO): δ ϭ 152.17 (Caryl), 147.77 (Caryl), 129.55
stored for about 12 h at 0°C and the precipitated pyridinium hydro- (Caryl), 115.05 (Caryl), 112.76 (Caryl), 112.11 (Caryl), 58.25 (Cbenzyl),
chloride was filtered off under argon using a Schlenk flask. The
solvent was evaporated under reduced pressure and the residue was
subjected to Kugelrohr distillation in high vacuum. The chloro-
phosphanes 8 were isolated as colorless liquids, which solidified on
standing at Ϫ20°C. Compounds 8 were too unstable to allow the
measurement of infrared spectra or analysis of their elemental com-
positions.
55.25 (OCH3). Ϫ IR (KBr): ν˜ ϭ 3442 cmϪ1, 3183, 1615, 1515,
1480, 1458, 1436, 1384, 1315, 1255, 1229, 1208, 1164, 1040, 1000,
999, 984, 930, 864, 818, 757, 701. Ϫ C8H10O3 (154.2): calcd. C
62.21, H 6.53; found C 62.44, H 6.68.
5-Methoxysalicylchlorophosphane (8d): Yield: 50% (colorless oil).
1
Ϫ B.p. 160Ϫ175°C (2.5 mbar). Ϫ H NMR (CDCl3): δ ϭ 6.80 (d,
1 H, 3JHH ϭ 8.3 Hz, H-3aryl), 6.71 (dd, 1 H, 3JHH ϭ 8.9 Hz, 4JHH ϭ
4
5-Nitrosalicyl Alcohol (7a): Yield: 45% (yellow solid). Ϫ M.p.
129°C. Ϫ Rf (TLC) ϭ 0.50 (CH2Cl2/CH3OH, 9:1); 0.44 (ethyl ace-
2.9 Hz, H-4aryl), 6.38 (d, 1 H, JHH ϭ 2.9 Hz, H-6aryl), 5.34 (dd, 1
H, 3JHH ϭ 14.3 Hz, 3JHP ϭ 2.3 Hz, Hbenzyl), 4.97 (dd, 1 H, 3JHH ϭ
tate/CH3OH, 7:3). Ϫ 1H NMR ([D6]DMSO): δ ϭ 11.07 (s, br, 1 14.3 Hz, 3JHP ϭ 9.6 Hz, Hbenzyl), 3.67 (s, 3 H, OCH3). Ϫ 13C NMR
4
H, OHaryl), 8.21 (td, 1 H, JHH ϭ 2.9 Hz, 4JHH ϭ 0.8 Hz, H-6aryl),
(CDCl3): δ ϭ 155.79 (Caryl), 122.02 (Caryl), 120.37 (Caryl), 114.82
3
4
8.02 (dd, 1 H, JHH ϭ 8.9 Hz, JHH ϭ 2.9 Hz, H-4aryl), 6.94 (d, 1 (Caryl), 110.30 (Caryl), 107.85 (Caryl), 61.17 (Cbenzyl), 55.67 (OCH3).
3
H, JHH ϭ 8.9 Hz, H-3aryl), 5.34 (s, br, 1 H, OHbenzyl), 4.51 (s, 2
H, Hbenzyl). Ϫ 13C NMR ([D6]DMSO): δ ϭ 160.69 (C-2aryl), 139.85
(C-5aryl), 130.51 (C-1aryl), 124.20 (C-4aryl), 122.89 (C-6aryl), 114.86
Ϫ
31P NMR (CDCl3): δ ϭ 139.38.
3-Methoxysalicyl Alcohol (7e): Yield: 65% (colorless solid). Ϫ
M.p. 79°C. Ϫ Rf (TLC) ϭ 0.54 (CH2Cl2/CH3OH, 9:1); 0.51 (ethyl
(C-3aryl), 57.62 (Cbenzyl). Ϫ IR (KBr): ν ϭ 3459 cmϪ1, 3077, 1592,
˜
1
acetate/CH3OH, 7:3). Ϫ H NMR ([D6]DMSO): δ ϭ 8.49 (s, br, 1
1487, 1334, 815, 757, 701. Ϫ C7H7NO4 (169.3): calcd. C 49.62, H
4.13; found C 49.57, H 4.09.
3
4
H, OHaryl), 6.91 (dd, 1 H, JHH ϭ 7.5 Hz, JHH ϭ 1.7 Hz, Haryl),
3
4
6.83 (dd, 1 H, JHH ϭ 8.0 Hz, JHH ϭ 1.7 Hz, Haryl), 6.74 (t, 1 H,
5-Nitrosalicylchlorophosphane (8a): Yield: 45% (yellow oil). Ϫ
3JHH ϭ 7.8 Hz, Haryl), 4.92 (s, br, 1 H, OHbenzyl), 4.48 (s, 2 H,
1
B.p. 155Ϫ163°C (2.5 mbar). Ϫ H NMR (CDCl3): δ ϭ 8.17 (dd, 1 Hbenzyl), 3.77 (s, 3 H, OCH3). Ϫ 13C NMR ([D6]DMSO): δ ϭ
3
4
4
H, JHH ϭ 8.0 Hz, JHH ϭ 0.9 Hz, H-4aryl), 7.98 (d, 1 H, JHH
ϭ
146.95 (C-3aryl), 142.94 (C-2aryl), 129.04 (C-6aryl), 119.39 (C-1aryl),
3
0.9 Hz, H-6aryl), 7.13 (d, 1 H, JHH ϭ 7.9 Hz, H-3aryl), 5.55 (dd, 1 118.45 (C-5aryl), 110.18 (C-4aryl), 58.15 (Cbenzyl), 55.78 (OCH3). Ϫ
H, 2JHH ϭ 14.2 Hz, 3JHP ϭ 2.2 Hz, Hbenzyl), 5.14 (dd, 1 H, 3JHH ϭ IR (KBr): ν ϭ 3442 cmϪ1, 3183, 1615, 1515, 1480, 1458, 1384,
˜
14.2 Hz, 3JHP ϭ 9.6 Hz, Hbenzyl). Ϫ 31P NMR (CDCl3): δ ϭ 140.01.
1315, 1255, 1229, 1208, 1164, 1040, 984, 930, 757. Ϫ C8H10O3
(154.2): calcd. C 62.33, H 6.54; found C 62.58, H 6.66.
5-Chlorosalicyl Alcohol (7b): Yield: 73% (colorless plates). Ϫ
M.p. 91°C. Ϫ Rf (TLC) ϭ 0.60 (CH2Cl2/CH3OH, 9:1); 0.52 (ethyl
3-Methoxysalicylchlorophosphane (8e): Yield: 65% (colorless oil).
1
1
acetate/CH3OH, 7:3). Ϫ H NMR ([D6]DMSO): δ ϭ 9.61 (s, br, 1 Ϫ B.p. 169Ϫ180°C (2.5 mbar). Ϫ H NMR (CDCl3): δ ϭ 7.03 (t,
4
3
3
H, OHaryl), 7.27 (d, 1 H, JHH ϭ 2.7 Hz, H-6aryl), 7.06 (dd, 1 H, 1 H, JHH ϭ 8.0 Hz, H-5aryl), 6.85 (d, 1 H, JHH ϭ 8.2 Hz, H-
4
3
3
3JHH ϭ 8.5 Hz, JHH ϭ 2.7 Hz, H-4aryl), 6.77 (d, 1 H, JHH ϭ 8.5 4aryl), 6.56 (d, 1 H, JHH ϭ 7.8 Hz, H-6aryl), 5.42 (dd, 1 H, 2JHH ϭ
3
3
3
Hz, H-3aryl), 5.09 (t, 1 H, JHH ϭ 5.5 Hz, OHbenzyl), 4.46 (d, 2 H, 14.2 Hz, JHP ϭ 2.2 Hz, Hbenzyl), 4.99 (dd, 1 H, JHH ϭ 14.2 Hz,
3JHH ϭ 5.0 Hz, Hbenzyl). Ϫ 13C NMR ([D6]DMSO): δ ϭ 152.84 3JHP ϭ 9.6 Hz, Hbenzyl), 3.87 (s, 3 H, OCH3). Ϫ 13C NMR (CDCl3):
(C-2aryl), 131.03 (C-1aryl), 126.72 (C-4aryl), 126.54 (C-6aryl), 122.41 δ ϭ 149.8 (Caryl), 136.15 (Caryl), 124.1 (Caryl), 122.7 (Caryl), 117.6
(C-5aryl), 116.00 (C-3aryl), 57.71 (C-benzyl). Ϫ IR (KBr): ν˜ ϭ 3442 (Caryl), 111.9 (Caryl), 61.07 (Cbenzyl), 56.54 (OCH3). Ϫ 31P NMR
cmϪ1, 3183, 1615, 1515, 1480, 1458, 1436, 1384, 1315, 1255, 1229,
1208, 1164, 1040, 1000, 999, 984, 930, 864, 818, 757, 701. Ϫ
C7H7ClO2 (158.6): calcd. C 52.99, H 4.41; found C 52.85, H 4.36.
(CDCl3): δ ϭ 140.36.
5-Methylsalicyl Alcohol (7f): Yield: 68% (colorless solid). Ϫ M.p.
104.5°C. Ϫ Rf (TLC) ϭ 0.62 (CH2Cl2/CH3OH, 9:1); 0.56 (ethyl
1
5-Chlorosalicylchlorophosphane (8b): Yield: 85% (pale yellow oil). acetate/CH3OH, 7:3). Ϫ H NMR ([D6]DMSO): δ ϭ 9.03 (s, br, 1
Ϫ B.p. 171Ϫ180°C (2.5 mbar). Ϫ 1H NMR (CDCl3): δ ϭ 7.21 (dd, H, OHaryl), 7.09 (d, 1 H, JHH ϭ 1.6 Hz, H-6aryl), 6.83 (dd, 1 H,
4
1 H, 3JHH ϭ 8.7 Hz, 4JHH ϭ 2.5 Hz, H-4aryl), 6.97 (d, 1 H, 4JHH ϭ
3JHH ϭ 8.1 Hz, JHH ϭ 1.7 Hz, H-4aryl), 6.65 (d, 1 H, JHH ϭ 8.1
4
3
3
2.5 Hz, H-6aryl), 6.91 (d, 1 H, JHH ϭ 8.7 Hz, H-3aryl), 5.40 (dd, 1 Hz, H-3aryl), 4.46 (s, 2 H, Hbenzyl), 3.51 (s, br, 1 H, OHbenzyl), 2.18
H, 2JHH ϭ 14.4 Hz, 3JHP ϭ 2.3 Hz, Hbenzyl), 4.98 (dd, 1 H, 3JHH ϭ (s, 3 H, CH3). Ϫ 13C NMR (CDCl3): δ ϭ 151.87 (C-2aryl), 128.17
14.4 Hz, 3JHP ϭ 9.6 Hz, Hbenzyl). Ϫ 13C NMR (CDCl3): δ ϭ 144.68 (C-5aryl), 127.92 (C-4aryl), 127.58 (C-6aryl), 126.93 (C-1aryl), 114.42
Ϫ1
˜
(C-3aryl), 58.32 (Cbenzyl), 20.33 (CH3). Ϫ IR (KBr): ν ϭ 3320 cm ,
(C-2aryl), 130.34 (C-1aryl), 129.31 (C-5aryl), 125.67 (C-6aryl), 122.79
(C-4aryl), 120.98 (C-3aryl), 65.95 (C-benzyl). Ϫ 31P NMR (CDCl3): 1595, 1469, 1438, 1343, 1212, 1085, 995, 778, 772, 679, 611. Ϫ
δ ϭ 139.83.
C8H10O2 (138.2): calcd. C 69.55, H 7.29; found C 69.29, H 7.22.
Salicylchlorophosphane (8c): Yield: 65% (colorless oil). Ϫ B.p.
160Ϫ175°C (2.5 mbar). Ϫ H NMR (CDCl3): δ ϭ 7.16 (dt, 1 H, B.p. 167Ϫ182°C (2.5 mbar). Ϫ H NMR (CDCl3): δ ϭ 7.20 (d, 1
5-Methylsalicylchlorophosphane (8f): Yield: 75% (colorless oil). Ϫ
1
1
4
3
3
3
3JHH ϭ 8.3 Hz, JHH ϭ 0.9 Hz, Haryl), 7.00 (dt, 1 H, JHH ϭ 7.6 H, JHH ϭ 8.0 Hz, H-4aryl), 7.01 (d, 1 H, JHH ϭ 8.0 Hz, H-3aryl),
4
3
3
Hz, JHH ϭ 1.2 Hz, Haryl), 6.88 (2 d, 2 H, JHH ϭ 8.3 Hz, Haryl), 6.92 (s, 1 H, H-6aryl), 5.55 (d, 1 H, JHP ϭ 14.3 Hz, Hbenzyl), 5.12
3
3
3
5.50 (dd, 1 H, JHH ϭ 14.3 Hz, JHP ϭ 2.3 Hz, Hbenzyl), 4.92 (dd,
(d, 1 H, JHP ϭ 14.1 Hz, Hbenzyl), 2.45 (s, 3 H, CH3aryl). Ϫ 13C
1 H, JHH ϭ 14.3 Hz, JHP ϭ 9.6 Hz, Hbenzyl). Ϫ 13C NMR NMR (CDCl3): δ ϭ 143.71 (C-2aryl), 133.58 (C-5aryl), 129.74 (C-
3
3
(CDCl3): δ ϭ 129.11 (Caryl), 125.66 (Caryl), 123.86 (Caryl), 121.46
1aryl), 125.94 (C-4aryl), 120.96 (C-6aryl), 119.10 (C-3aryl), 61.07
(Caryl), 121.26 (Caryl), 119.28 (Caryl), 60.95 (Cbenzyl). Ϫ 31P NMR (Cbenzyl), 20.65 (CH3). Ϫ 31P NMR (CDCl3): δ ϭ 139.87.
(CDCl3): δ ϭ 140.67.
3-Methylsalicyl Alcohol (7g): Yield: 80% (colorless solid). Ϫ M.p.
5-Methoxysalicyl Alcohol (7d): Yield: 85% (colorless solid). Ϫ
72.5°C. Ϫ Rf (TLC) ϭ 0.69 (CH2Cl2/CH3OH, 9:1); 0.62 (ethyl acet-
1
M.p. 79°C. Ϫ Rf (TLC) ϭ 0.55 (CH2Cl2/CH3OH, 9:1); 0.49 (ethyl ate/CH3OH, 7:3). Ϫ H NMR ([D6]DMSO): δ ϭ 8.38 (s, br, 1 H,
1
3
3
acetate/CH3OH, 7:3). Ϫ H NMR ([D6]DMSO): δ ϭ 8.80 (s, br, 1 OHaryl), 7.04 (d, 1 H, JHH ϭ 7.6 Hz, Haryl), 6.95 (d, 1 H, JHH
H, OHaryl), 6.88 (d, 1 H, Haryl), 6.57Ϫ6.69 (m, 2 H, Haryl), 4.95 (s,
7.6 Hz, Haryl), 6.71 (t, 1 H, 3JHH ϭ 7.6 Hz, Haryl), 5.27 (s, br, 1 H,
br, 1 H, OHbenzyl), 4.45 (s, 2 H, Hbenzyl), 3.65 (s, 3 H, OCH3). Ϫ OHbenzyl), 4.56 (s, 2 H, Hbenzyl), 2.15 (s, 3 H, CH3). Ϫ 13C NMR
ϭ
842
Eur. J. Org. Chem. 1998, 837Ϫ846

Nucleotide Delivery from cycloSaligenyl-3Ј-azido-3Ј-deoxythymidine Monophosphates
FULL PAPER
2
3
3
([D6]DMSO): δ ϭ 152.57 (Caryl), 129.01 (Caryl), 127.95 (Caryl),
H, JHH ϭ 11.4 Hz, JHH ϭ 4.6 Hz, JHP ϭ 2.6 Hz, H-5Љ), 4.36
2
3
3
125.01 (Caryl), 124.07 (Caryl), 118.97 (Caryl), 59.87 (Cbenzyl), 16.14 (dd, 1 H, JHH ϭ 11.4 Hz, JHH ϭ 5.2 Hz, JHP ϭ 7.5 Hz, H-5Ј),
(CH3). Ϫ IR (KBr): ν˜ ϭ 3320 cmϪ1, 1595, 1469, 1438, 1343, 1212, 3.98 (m, 1 H, H-4Ј), 2.45 (m, 1 H, H-2Љ), 2.33 (m, 1 H, H-2Ј), 1.74
1085, 995, 778, 772, 679, 611. Ϫ C8H10O2 (138.2): calcd. C 69.40,
H 7.28; found C 69.64, H 7.37.
(2 d, 3 H, 4JHH ϭ 1.1 Hz, CH3-thymine). Ϫ 13C NMR ([D6]DMSO):
δ ϭ 163.99 (C-4), 154.1 (2 d, C-2aryl), 150.7 (C-2), 143.8 (C-5aryl),
136.7 (C-6), 125.4 (C-4aryl), 122.7, 122.8 (C-6aryl), 122.2 (2 d, C-
1aryl), 119.8 (2 d, C-3aryl), 110.31, 110.28 (C-5), 83.34, 83.31 (C-1Ј),
81.35, 81.28 (C-4Ј), 68.2 (2 d, Cbenzyl), 67.3 (2 d, C-5Ј), 59.8, 60.0
(C-3Ј), 35.9, 35.8 (C-2Ј), 12.2, 12.3 (CH3-thymine). Ϫ 31P NMR
([D6]DMSO): δ ϭ Ϫ9.49 and Ϫ9.70 [2 s, diastereomeric mixture].
Ϫ UV (acetonitrile): λmax (ε) ϭ 265 nm (10100), λmin (ε) ϭ 235
nm (4100). Ϫ IR (KBr): ν˜ ϭ 3580 cmϪ1, 3190, 3010, 2136, 1686,
1560, 1489, 1458, 1379, 1350, 1294, 1189, 1104, 998, 946, 917. Ϫ
MS (m/z): 478 (M Ϫ 1). Ϫ HPLC tR: 20.24 min.
3-Methylsalicylchlorophosphane (8g): Yield: 70% (colorless oil).
Ϫ B.p. 165Ϫ178°C (2.5 mbar). Ϫ 1H NMR (CDCl3): δ ϭ 7.10 (dd,
1 H, 3JHH ϭ 7.5 Hz, 4JHH ϭ 0.7 Hz, H-6aryl), 6.96 (t, 1 H, 3JHH ϭ
4
7.6 Hz, H-5aryl), 6.77 (dd, 1 H, 3JHH ϭ 7.6 Hz, JHH ϭ 0.6 Hz, H-
4aryl), 5.40 (dd, 1 H, 2JHH ϭ 14.1 Hz, 3JHP ϭ 2.6 Hz, Hbenzyl), 4.98
3
3
(dd, 1 H, JHH ϭ 14.1 Hz, JHP ϭ 9.4 Hz, Hbenzyl), 2.22 (s, 3 H,
CH3). Ϫ 13C NMR (CDCl3): δ ϭ 130.48 (Caryl), 123.35 (Caryl),
123.19 (Caryl), 123.17 (Caryl), 61.26 (Cbenzyl), 15.52 (CH3). Ϫ 31P
NMR (CDCl3): δ ϭ 141.45.
3,5-Dimethylsalicyl Alcohol (7h): Yield: 57% (colorless crystalls).
Ϫ M.p. 52.2°C. Ϫ Rf (TLC) ϭ 0.60 (CH2Cl2/CH3OH, 9:1); 0.52
(ethyl acetate/CH3OH, 7:3). Ϫ 1H NMR ([D6]DMSO): δ ϭ 8.15 (s,
br, 1 H, OHaryl), 6.84 (s, 1 H, H-6aryl), 6.75 (s, 1 H, H-4aryl), 5.23
(s, br, 1 H, OHbenzyl), 4.53 (s, 2 H, Hbenzyl), 2.15 (s, 3 H, CH3), 2.10
(s, 3 H, CH3). Ϫ 13C NMR ([D6]DMSO): δ ϭ 150.27 (C-2aryl),
129.52 (C-4aryl), 127.74 (C-5aryl), 127.26 (C-1aryl), 125.50 (C-6aryl),
123.84 (C-3aryl), 59.93 (Cbenzyl), 20.14 (CH3), 16.12 (CH3). Ϫ IR
cyclo-(5-Chlorosaligenyl)-5Ј-O-(3Ј-azido-3Ј-deoxythymidinyl)-
phosphate (4b): 0.167 g of 5-chlorosalicylchlorophosphane (8b) was
used. Ϫ Yield: 0.145 g (84%). Ϫ Rf (TLC) ϭ 0.44 (CH2Cl2/CH3OH,
9:1); 0.65 (ethyl acetate/CH3OH, 8:2). Ϫ 1H NMR ([D6]DMSO):
δ ϭ 11.33 (s, 1 H, N-H), 7.45 (m, 1 H, H-6thymine), 7.42 (m, 2 H,
3
H-4aryl and H-6), 7.17 (d, 1 H, JHH ϭ 9.1 Hz, H-3aryl), 6.10, 6.09
3
2
(2 t, 1 H, JHH ϭ 6.6 Hz, 2ϫ H-1Ј), 5.52 (dd, 1 H, JHH ϭ 14.5
Hz, 3JHH ϭ 5.8 Hz, 3JHP ϭ 17.9 Hz, Hbenzyl), 5.56 (dd, 1 H, 2JHH ϭ
(KBr):ν ϭ 3394 cmϪ1, 2920, 1469, 1486, 1213, 1085, 995, 853, 772,
14.5 Hz, 3JHH ϭ 4.9 Hz, JHP ϭ 9.3 Hz, Hbenzyl), 4.45 (m, 1 H, H-
3
˜
3Ј), 4.38 (dd, 1 H, 2JHH ϭ 11.6 Hz, 3JHH ϭ 5.1 Hz, 3JHP ϭ 2.4 Hz,
679, 611. Ϫ C9H12O2 (152.2): calcd. C 71.03, H 7.95; found C
71.26, H 7.84.
2
3
3
H-5Љ), 4.32 (dd, 1 H, JHH ϭ 11.5 Hz, JHH ϭ 5.2 Hz, JHP ϭ 7.6
Hz, H-5Ј), 3.98 (m, 1 H, H-4Ј), 2.41 (m, 1 H, H-2Љ), 2.32 (m, 1 H,
3,5-Dimethylsalicylchlorophosphane (8h): Yield: 75% (colorless
oil). Ϫ B.p. 167Ϫ182°C (2.5 mbar). Ϫ 1H NMR (CDCl3): δ ϭ 6.93
(s, 1 H, H-4aryl), 6.60 (s, 1 H, H-6aryl), 5.37 (br, 1 H, Hbenzyl), 4.97
(br, 1 H, Hbenzyl), 2.26 (s, 3 H, CH3), 2.20 (s, 3 H, CH3). Ϫ 13C
NMR (CDCl3): δ ϭ 142.34 (C-2aryl), 132.92 (C-5aryl), 132.81 (C-
1aryl), 131.18 (C-3aryl), 129.42 (C-4aryl), 123.39 (C-6aryl), 61.28
(Cbenzyl), 20.61 (5-CH3), 15.34 (3-CH3). Ϫ 31P NMR (CDCl3):
δ ϭ 141.58.
H-2Ј), 1.75 (2 d, 3 H, JHH ϭ 1.0 Hz, CH3-thymine). Ϫ 13C NMR
4
([D6]DMSO): δ ϭ 163.61 (C-4), 150.3 (C-2), 148.23, 148.16 (2 d,
C-2aryl), 135.9 (C-6), 129.5 (C-6aryl), 128.3 (C-5aryl), 125.99 (C-4aryl),
123.00 (2 d, C-1aryl), 122.95, 122.84 (2 s, C-1aryl), 120.14, 120.00 (2
s, C-3aryl), 110.31, 110.28 (C-5), 83.80, 83.31 (C-1Ј), 81.02, 80.92
(C-4Ј), 68.11, 68.00 (2 d, Cbenzyl), 67.03 (m, C-5Ј), 59.73, 59.52 (C-
3Ј), 35.56, 35.46 (C-2Ј), 12.1, 12.0 (CH3-thymine). Ϫ 31P NMR
([D6]DMSO): δ ϭ Ϫ9.07 and Ϫ9.18 [2 s, diastereomeric mixture].
Ϫ UV (acetonitrile): λmax (ε) ϭ 266 nm (9800), λmin (ε) ϭ 234 nm
(4400). Ϫ IR (KBr): ν˜ ϭ 3480 cmϪ1, 3210, 3020, 2130, 1686, 1489,
1458, 1379, 1294, 1189, 1104, 1020, 998, 946. Ϫ MS (m/z): 468.5/
470.5 (M Ϫ 1, 35Cl/M Ϫ 1, 37Cl; ratio 3:1). Ϫ HPLC tR: 21.31 min.
General Procedure for the Synthesis of the cycloSal-AZTMP
Phosphotriesters 4a؊h: The reactions were performed under anhy-
drous conditions in an argon atmosphere. To a solution of AZT
(1) (0.10 g, 0.37 mmol) in 6.0 ml CH3CN, cooled to 0°C in ice/
water, was added DIPEA (0.097 g, 0.75 mmol). Then, the appropri-
ate chlorophosphane 8 (0.75 mmol) was added over a period of 5
min. and the solutions were stirred for a further 15 min. to ensure
complete reaction (TLC analysis). For oxidation of the intermedi-
ate cyclic phosphites, t-butyl hydroperoxide (0.072 g, 0.80 mmol; as
a 5.0-6.0 M solution in decane commercially available from Ald-
rich) was added to the reaction mixtures at 0°C. After stirring for
30 min. and warming to room temperature, the solvent was re-
moved under reduced pressure. The residues were purified twice by
chromatography on silica gel plates on a chromatotron, first using
a gradient of CH3OH in ethyl acetate and then a gradient of
CH3OH in CH2Cl2, to yield the title compounds 4. The compounds
were isolated as mixtures of diastereomers as a result of the mixed
stereochemistry at the phosphorus atom.
cyclo-Saligenyl-5Ј-O-(3Ј-azido-3Ј-deoxythymidinyl)phosphate
(4c): 0.141 g of salicylchlorophosphane (8c) was used. Ϫ Yield:
0.112 g (69%). Ϫ Rf (TLC) ϭ 0.36 (CH2Cl2/CH3OH, 9:1); 0.51
1
(ethyl acetate/CH3OH, 8:2). Ϫ H NMR ([D6]DMSO): δ ϭ 11.36
4
(s, 1 H, N-H), 7.45, 7.43 (2 d, 1 H, JHH ϭ 1.0 Hz, H-6aryl), 7.37
3
4
(tt, 1 H, JHH ϭ 7.3 Hz, JHH ϭ 1.0 Hz, H-4aryl), 7.27 (d, 1 H,
3JHH ϭ 7.6 Hz, H-6thymine), 7.19 (tt, 1 H, JHH ϭ 7.5 Hz, JHH
ϭ
3
4
3
1.0 Hz, H-5aryl), 7.12 (d, 1 H, JHH ϭ 7.8 Hz, H-3aryl), 6.10 (t, 1
H, 3JHH ϭ 6.7 Hz, H-1Ј), 5.51 (ddd, 1 H, 2JHH ϭ 14.7 Hz, JHH ϭ
3
3
2
5.3 Hz, JHP ϭ 17.3 Hz, Hbenzyl), 5.46 (ddd, 1 H, JHH ϭ 14.7 Hz,
3JHH ϭ 5.1 Hz, JHP ϭ 9.6 Hz, Hbenzyl), 4.45 (m, 1 H, H-3Ј), 4.38
3
2
3
(dd, 1 H, JHH ϭ 11.1 Hz, JHH ϭ 7.06 Hz, 3JHP ϭ 4.6 Hz, H-5Љ),
2
3
3
4.31 (dd, 1 H, JHH ϭ 11.5 Hz, JHH ϭ 5.15 Hz, JHP ϭ 9.6 Hz,
cyclo-(5-Nitrosaligenyl)-5Ј-O-(3Ј-azido-3Ј-deoxythymidinyl)- H-5Ј), 3.98 (m, 1 H, H-4Ј), 2.38 (m, 1 H, H-2Љ), 2.32 (m, 1 H, H-
phosphate (4a): 0.175 g of 5-nitrosalicylchlorophosphane (8a) was
2Ј), 1.75, 1.72 (2 d, 3 H, 4JHH ϭ 0.9 Hz, CH3-thymine). Ϫ 13C NMR
3
used. Ϫ Yield: 0.125 g (70%). Ϫ Rf (TLC) ϭ 0.38 (CH2Cl2/CH3OH, ([D6]DMSO): δ ϭ 163.61 (C-4), 150.3 (C-2), 145.50 (2 s, JCP
ϭ
9:1); 0.55 (ethyl acetate/CH3OH, 8:2). Ϫ 1H NMR ([D6]DMSO): 7.2 Hz, C-2aryl), 135.9 (C-6), 126.2 (C-6aryl), 124.6 (C-5aryl), 129.92
4
3
3
δ ϭ 11.30, 11.29 (2 s, 1 H, N-H), 8.30 (d, 1 H, JHH ϭ 2.7 Hz, H-
(C-4aryl), 121.11 (2 s, JCP ϭ 9.9 Hz, C-1aryl), 118.31 (2 s, JCP ϭ
3
4
5
6thymine), 8.24 (d, 1 H, JHH ϭ 9.2 Hz, JHH ϭ 2.7 Hz, JHP ϭ 1.1 8.8 Hz, C-3aryl), 110.31, 110.28 (C-5), 83.9 (C-1Ј), 81.0 (C-4Ј),
4
Hz, H-4aryl), 7.42 (2 d, 1 H, JHH ϭ 1.1 Hz, H-6), 7.38 (dd, 1 H, 68.11, 68.60 (2 d, Cbenzyl), 68.2 (m, C-5Ј), 59.7 (C-3Ј), 35.56, 35.6
3JHH ϭ 9.1 Hz, 5JHH ϭ 1.3 Hz, H-3aryl), 6.09 (2 t, 1 H, 3JHH ϭ 6.5
(C-2Ј), 12.0 (CH3-thymine). Ϫ 31P NMR ([D6]DMSO): δ ϭ Ϫ8.59
2
3
Hz, 2ϫ H-1Ј), 5.66 (dd, 1 H, JHH ϭ 14.9 Hz, JHH ϭ 6.5 Hz, and Ϫ8.62 [2ϫ s, diastereomeric mixture]. Ϫ UV (acetonitrile):
3JHP ϭ 17.6 Hz, Hbenzyl), 5.56 (dd, 1 H, JHH ϭ 14.9 Hz, JHH
ϭ
λmax (ε) ϭ 265 nm (9800), λmin (ε) ϭ 234 nm (4600). Ϫ IR (KBr):
2
3
5.0 Hz, JHP ϭ 9.6 Hz, Hbenzyl), 4.45 (m, 1 H, H-3Ј), 4.43 (dd, 1 ν˜ ϭ 3520 cmϪ1, 3230, 3020, 2140, 1684, 1489, 1458, 1379, 1290,
3
Eur. J. Org. Chem. 1998, 837Ϫ846
843

C. Meier, E. De Clercq, J. Balzarini
FULL PAPER
3
1191, 1106, 1020, 998, 946. Ϫ MS (m/z): 434.5 (M Ϫ 1). Ϫ HPLC 3JHP ϭ 17.6 Hz, Hbenzyl), 5.40 (ddd, 1 H, 2JHH ϭ 14.3 Hz, JHH
ϭ
3
tR: 19.03 min.
5.1 Hz, JHP ϭ 9.2 Hz, Hbenzyl), 4.45 (m, 1 H, H-3Ј), 4.36 (dd, 1
2
3
3
H, JHH ϭ 11.5 Hz, JHH ϭ 3.5 Hz, JHP ϭ 4.6 Hz, H-5Љ), 4.31
cyclo-(5-Methoxysaligenyl)-5Ј-O-(3Ј-azido-3Ј-deoxythymidinyl)-
phosphate (4d): 0.164 g of 5-methoxysalicylchlorophosphane (8d)
was used. Ϫ Yield: 0.125 g (73%). Ϫ Rf (TLC) ϭ 0.36 (CH2Cl2/
CH3OH, 9:1); 0.51 (ethyl acetate/CH3OH, 8:2). Ϫ 1H NMR
([D6]DMSO): δ ϭ 11.35 (s, 1 H, N-H), 7.42, 7.44 (2 d, 1 H, 4JHH ϭ
1.1 Hz, H-6thymine), 7.07 (d, 1 H, 3JHH ϭ 9.0 Hz, H-4aryl), 6.90 (m,
2
3
3
(dd, 1 H, JHH ϭ 11.5 Hz, JHH ϭ 5.0 Hz, JHP ϭ 7.5 Hz, H-5Ј),
3.98 (m, 1 H, H-4Ј), 2.37 (m, 1 H, H-2Љ), 2.32 (m, 1 H, H-2Ј), 2.26
(s, 3 H, CH3), 1.75, 1.71 (2 d, 3 H, 4JHH ϭ 1.1 Hz, CH3-thymine). Ϫ
13C NMR ([D6]DMSO): δ ϭ 163.62 (C-4), 150.33 (C-2), 147.35,
3
147.22 (2 s, JCP ϭ 7.2 Hz, C-2aryl), 135.87, 135.80 (C-6), 133.70
(C-5aryl), 130.08 (C-4aryl), 126.10 (C-6aryl), 120.61, 120.46 (2 s,
3JCP ϭ 9.9 Hz, C-1aryl), 117.94, 117.81 (2 s, 3JCP ϭ 8.8 Hz, C-3aryl),
109.93 (C-5), 83.77, 83.70 (C-1Ј), 81.10, 80.99 (C-4Ј), 68.58, 68.47,
(d, Cbenzyl), 66.95, 66.87 (m, C-5Ј), 59.75, 59.57 (C-3Ј), 35.60, 35.47
(C-2Ј), 20.13 (CH3-aryl), 12.02, 11.97 (CH3-thymine). Ϫ 31P NMR
([D6]DMSO): δ ϭ Ϫ8.52 and Ϫ8.55 [2 s, diastereomeric mixture].
Ϫ UV (acetonitrile): λmax (ε) ϭ 265 nm (11200), λmin (ε) ϭ 236 nm
(3500). Ϫ IR (KBr): ν˜ ϭ 3580 cmϪ1, 3250, 3037, 2126, 1680, 1489,
1458, 1379, 1297, 1189, 1114, 1020, 998, 946. Ϫ MS (m/z): 448.5
(M Ϫ 1). Ϫ HPLC tR: 20.77 min.
4
1 H, H-3aryl), 6.86 (d, 1 H, JHH ϭ 3.0 Hz, H-6aryl), 6.10 (t, 1 H,
3JHH ϭ 6.7 Hz, H-1Ј), 5.44 (ddd, 1 H, JHH ϭ 14.5 Hz, JHH
ϭ
2
3
3
2
6.1 Hz, JHP ϭ 16.9 Hz, Hbenzyl), 5.46 (ddd, 1 H, JHH ϭ 14.5 Hz,
3JHH ϭ 4.5 Hz, JHP ϭ 9.7 Hz, Hbenzyl), 4.44 (m, 1 H, H-3Ј), 4.36
3
2
3
(dd, 1 H, JHH ϭ 11.1 Hz, JHH ϭ 7.06 Hz, 3JHP ϭ 4.6 Hz, H-5Љ),
2
3
3
4.31 (dd, 1 H, JHH ϭ 11.5 Hz, JHH ϭ 5.15 Hz, JHP ϭ 9.6 Hz,
H-5Ј), 3.98 (m, 1 H, H-4Ј), 3.72 (2 s, 3 H, OCH3), 2.39 (m, 1 H,
4
H-2Љ), 2.31 (m, 1 H, H-2Ј), 1.75, 1.70 (2 d, 3 H, JHH ϭ 1.1 Hz,
CH3-thymine). Ϫ 13C NMR ([D6]DMSO): δ ϭ 163.63 (C-4), 155.64
3
(C-5aryl), 150.35 (C-2), 143.20, 142.93 (2 s, JCP ϭ 7.2 Hz, C-2aryl),
135.87 (C-6), 129.92 (C-4aryl), 121.81, 121.65 (2 s, JCP ϭ 9.9 Hz,
C-1aryl), 119.11, 118.97 (2 s, JCP ϭ 8.8 Hz, C-3aryl), 115.11 (C-
3
cyclo-(3-Methylsaligenyl)-5Ј-O-(3Ј-azido-3Ј-deoxythymidinyl)-
phosphate (4g): 0.152 g of 3-methylsalicylchlorophosphane (8g) was
used. Ϫ Yield: 0.142 g (86%). Ϫ Rf (TLC) ϭ 0.58 (CH2Cl2/CH3OH,
9:1); 0.66 (ethyl acetate/CH3OH, 8:2). Ϫ 1H NMR ([D6]DMSO):
3
6aryl), 110.65, 109.94 (C-5), 83.81, 83.70 (C-1Ј), 81.15, 81.00 (C-4Ј),
68.63, 68.55 (2 d, Cbenzyl), 66.87 (m, C-5Ј), 59.83, 59.60 (C-3Ј),
55.56 (OCH3), 35.60, 35.48 (C-2Ј), 12.04, 12.00 (CH3-thymine). Ϫ
31P NMR ([D6]DMSO): δ ϭ Ϫ8.41 and Ϫ8.45 [2 s, diastereomeric
mixture]. Ϫ UV (acetonitrile): λmax (ε) ϭ 267 nm (10200), λmin
(ε) ϭ 232 nm (4000). Ϫ IR (KBr): ν˜ ϭ 3620 cmϪ1, 3250, 3030,
2140, 1686, 1497, 1466, 1363, 1287, 1266, 1199, 1101, 1024, 994,
948, 915. Ϫ MS (m/z): 464.5 (M Ϫ 1). Ϫ HPLC tR: 19.75 min.
4
δ ϭ 11.36 (s, 1 H, N-H), 7.43 (d, 1 H, JHH ϭ 1.0 Hz, H-6thymine),
7.24 (m, 1 H, H-4aryl), 7.08 (m, 2 H, H-6 and 5aryl), 7.00 (d, 1 H,
4JHH ϭ 8.3 Hz, H-4aryl), 6.10 (t, 1 H, JHH ϭ 6.3 Hz, H-1Ј), 5.50
3
2
3
3
(ddd, 1 H, JHH ϭ 14.4 Hz, JHH ϭ 3.7 Hz, JHP ϭ 17.3 Hz,
Hbenzyl), 5.43 (ddd, 1 H, JHH ϭ 14.4 Hz, JHH ϭ 5.1 Hz, JHP
9.2 Hz, Hbenzyl), 4.44 (m, 1 H, H-3Ј), 4.38 (dd, 1 H, JHH ϭ 11.5
Hz, JHH ϭ 3.5 Hz, JHP ϭ 4.6 Hz, H-5Љ), 4.31 (dd, 1 H, JHH
2
3
3
ϭ
2
3
3
2
ϭ
3
3
cyclo-(3-Methoxysaligenyl)-5Ј-O-(3Ј-azido-3Ј-deoxythymidinyl)-
phosphate (4e): 0.164 g of 3-methoxysalicylchlorophosphane (8e)
was used. Ϫ Yield: 0.109 g (63%). Ϫ Rf (TLC) ϭ 0.52 (CH2Cl2/
CH3OH, 9:1); 0.61 (ethyl acetate/CH3OH, 8:2). Ϫ 1H NMR
([D6]DMSO): δ ϭ 11.33 (s, 1 H, N-H), 7.44, 7.40 (2 d, 1 H, 4JHH ϭ
1.2 Hz, H-6thymine), 7.12 (2 t, 1 H, 4JHH ϭ 8.1 Hz, H-5aryl), 7.09 (d,
1 H, 3JHH ϭ 8.3 Hz, H-4aryl), 6.81 (2 d, 1 H, H-6aryl), 6.10 (t, 1 H,
11.5 Hz, JHH ϭ 5.0 Hz, JHP ϭ 7.5 Hz, H-5Ј), 3.97 (m, 1 H, H-
4Ј), 2.37 (m, 1 H, H-2Љ), 2.32 (m, 1 H, H-2Ј), 2.21, 2.20 (2 s, 3 H,
CH3), 1.73, 1.72 (2 d, 3 H, JHH ϭ 1.0 Hz, CH3-thymine). Ϫ 13C
4
NMR ([D6]DMSO): δ ϭ 163.63 (C-4), 150.35 (C-2), 147.95, 147.84
3
(2 s, JCP ϭ 7.2 Hz, C-2aryl), 135.90 (C-6), 130.89 (C-4aryl), 126.91,
126.78 (2 s, 3JCP ϭ 8.8 Hz, C-3aryl), 123.98, 123.57 (C-5aryl), 121.01
3
(C-6aryl), 120.61, 120.46 (2 s, JCP ϭ 9.9 Hz, C-1aryl), 109.95 (C-5),
3JHH ϭ 6.6 Hz, H-1Ј), 5.49 (ddd, 1 H, JHH ϭ 14.5 Hz, JHH
ϭ
2
3
83.78, 83.68 (C-1Ј), 81.12, 81.00 (C-4Ј), 68.58, 68.50, (d, Cbenzyl),
67.04 (m, C-5Ј), 59.71, 59.63 (C-3Ј), 35.58, 35.49 (C-2Ј), 14.84
(CH3-aryl), 12.02 (CH3-thymine). Ϫ 31P NMR ([D6]DMSO): δ ϭ
Ϫ7.92 and Ϫ7.99 [2 s, diastereomeric mixture]. Ϫ UV (acetonitrile):
λmax (ε) ϭ 266 nm (10500), λmin (ε) ϭ 237 nm (4500). Ϫ IR (KBr):
ν˜ ϭ 3380 cmϪ1, 3240, 3020, 2133, 1686, 1489, 1458, 1379, 1291,
1268, 1191, 1106, 1020, 998, 946, 843, 809, 760, 721, 656. Ϫ MS
(m/z): 448.4 (M Ϫ 1). Ϫ HPLC tR: 20.31 min.
3
2
6.7 Hz, JHP ϭ 17.7 Hz, Hbenzyl), 5.43 (ddd, 1 H, JHH ϭ 14.4 Hz,
3JHH ϭ 4.7 Hz, JHP ϭ 9.8 Hz, Hbenzyl), 4.44 (m, 1 H, H-3Ј), 4.36
3
2
3
3
(dd, 1 H, JHH ϭ 11.4 Hz, JHH ϭ 3.5 Hz, JHP ϭ 7.0 Hz, H-5Љ),
4.31 (dd, 1 H, 2JHH ϭ 11.5 Hz, JHH ϭ 4.8 Hz, 3JHP ϭ 7.6 Hz, H-
3
5Ј), 3.98 (m, 1 H, H-4Ј), 3.81 (2 s, 3 H, OCH3), 2.37 (m, 1 H, H-
4
2Љ), 2.33 (m, 1 H, H-2Ј), 1.74, 1.70 (2 d, 3 H, JHH ϭ 1.1 Hz,
CH3-thymine). Ϫ 13C NMR ([D6]DMSO): δ ϭ 163.62 (C-4), 150.36
3
(C-2), 148.29, 148.18 (2 s, JCP ϭ 7.2 Hz, C-2aryl), 138.44 (C-3aryl),
135.85, 135.72 (C-6), 124.61 (C-4aryl), 121.95, 121.88 (2 s, JCP
cyclo-(3,5-Dimethylsaligenyl)-5Ј-O-(3Ј-azido-3Ј-deoxy-
thymidinyl)phosphate (4h): 0.163 g of 3,5-dimethylsalicylchloropho-
sphane (8h) was used. Ϫ Yield: 0.148 g (86%). Ϫ Rf (TLC) ϭ 0.62
(CH2Cl2/CH3OH, 9:1); 0.69 (ethyl acetate/CH3OH, 8:2). Ϫ 1H
NMR ([D6]DMSO): δ ϭ 11.31, 11.30 (2 s, 1 H, N-H), 7.42 (d, 1
3
ϭ
9.9 Hz, C-1aryl), 119.11, 118.97 (2 s, 3JCP ϭ 8.8 Hz, C-3aryl), 116.95
(C-6aryl), 112.58 (C-5aryl), 109.98, 109.93 (C-5), 83.66 (C-1Ј), 81.12,
81.00 (C-4Ј), 68.47, (d, Cbenzyl), 67.06 (m, C-5Ј), 59.75, 59.58 (C-
3Ј), 55.85 (OCH3), 35.65, 35.45 (C-2Ј), 11.94 (CH3-thymine). Ϫ 31P
NMR ([D6]DMSO): δ ϭ Ϫ8.12 and Ϫ8.16 [2 s, diastereomeric mix-
ture]. Ϫ UV (acetonitrile): λmax (ε) ϭ 266 nm (9200), λmin (ε) ϭ
233 nm (4200). Ϫ IR (KBr): ν˜ ϭ 3480 cmϪ1, 3210, 3020, 2130,
1686, 1489, 1458, 1379, 1294, 1189, 1104, 1020, 998, 946. Ϫ MS
(m/z): 464.4 (M Ϫ 1). Ϫ HPLC tR: 18.87 min.
4
H, JH,H ϭ 1.0 Hz, H-6thymine), 7.03 (s, 1 H, H-4aryl), 6.86 (s, 1 H,
H-6aryl), 6.09 (t, 1 H, 3JHH ϭ 6.6 Hz, H-1Ј), 5.42 (ddd, 1 H, 2JHH ϭ
3
3
14.5 Hz, JHH ϭ 3.7 Hz, JHP ϭ 17.5 Hz, Hbenzyl), 5.37 (ddd, 1 H,
2JHH ϭ 14.4 Hz, 3JHH ϭ 5.1 Hz, JHP ϭ 9.3 Hz, Hbenzyl), 4.44 (m,
3
1 H, H-3Ј), 4.38 (dd, 1 H, 2JHH ϭ 11.5 Hz, 3JHH ϭ 3.5 Hz, 3JHP ϭ
2
3
4.6 Hz, H-5Љ), 4.27 (dd, 1 H, JHH ϭ 11.5 Hz, JHH ϭ 5.0 Hz,
cyclo-(5-Methylsaligenyl)-5Ј-O-(3Ј-azido-3Ј-deoxythymidinyl)-
phosphate (4f): 0.152 g of 5-methylsalicylchlorophosphane (8f) was
3JHP ϭ 7.5 Hz, H-5Ј), 3.97 (m, 1 H, H-4Ј), 2.38 (m, 1 H, H-2Љ),
2.31 (m, 1 H, H-2Ј), 2.22 (s, 3 H, CH3), 2.167, 2.161 (2 s, 3 H,
used. Ϫ Yield: 0.135 g (81%). Ϫ Rf (TLC) ϭ 0.58 (CH2Cl2/CH3OH, CH3), 1.74, 1.72 (2 d, 3 H, JH,H ϭ 1.0 Hz, CH3-thymine). Ϫ 13C
4
9:1); 0.66 (ethyl acetate/CH3OH, 8:2). Ϫ 1H NMR ([D6]DMSO):
NMR ([D6]DMSO): δ ϭ 163.62 (C-4), 150.35 (C-2), 145.95, 145.84
4
3
δ ϭ 11.27 (s, 1 H, N-H), 7.43, 7.41 (2 d, 1 H, JHH ϭ 1.1 Hz, H- (2 s, JCP ϭ 7.2 Hz, C-2aryl), 135.89 (C-6), 130.87 (C-4aryl), 126.91,
6thymine), 7.15 (d, 1 H, JHH ϭ 8.4 Hz, H-3aryl), 7.05 (s, 1 H, H- 126.78 (2 s, 3JCP ϭ 8.8 Hz, C-3aryl), 123.98, 123.57 (C-5aryl), 121.01
4
4
3
3
6aryl), 7.00 (d, 1 H, JHH ϭ 8.3 Hz, H-4aryl), 6.09 (t, 1 H, JHH
6.6 Hz, H-1Ј), 5.46 (ddd, 1 H, JHH ϭ 14.2 Hz, JHH ϭ 3.7 Hz,
ϭ
(C-6aryl), 120.61, 120.46 (2 s, JCP ϭ 9.9 Hz, C-1aryl), 109.95 (C-5),
2
3
83.78, 83.68 (C-1Ј), 81.11, 81.02 (C-4Ј), 68.56, 68.48, (d, Cbenzyl),
Eur. J. Org. Chem. 1998, 837Ϫ846
844

Nucleotide Delivery from cycloSaligenyl-3Ј-azido-3Ј-deoxythymidine Monophosphates
FULL PAPER
[1]
A preliminary account of this work was presented at the 9th
International Conference on Antiviral Research, Fukushima,
Japan, May 19Ϫ24, 1996; Congress abstract: C. Meier, M. Lo-
rey, E. De Clercq, J. Balzarini, Antiviral Res. 1996, 28, A29, and
at the 4th Swedish-German Symposium, Reisensburg, Ger-
many, November 23Ϫ26, 1996; Congress abstract: C. Meier, E.
De Clercq, J. Balzarini, Nucleosides & Nucleotides 1997, 16,
793Ϫ796.
67.04 (m, C-5Ј), 59.73, 59.66 (C-3Ј), 35.58, 35.49 (C-2Ј), 20.56
(CH3-aryl), 14.84 (CH3-aryl), 12.02 (CH3-thymine). Ϫ 31P NMR
([D6]DMSO): δ ϭ Ϫ7.86 and Ϫ7.96 [2 s, diastereomeric mixture].
Ϫ UV (acetonitrile): λmax (ε) ϭ 266 nm (10500), λmin (ε) ϭ 237 nm
(4500). Ϫ IR (KBr): ν˜ ϭ 3580 cmϪ1, 3240, 3120, 2126, 1686, 1489,
1466, 1364, 1287, 1265, 1199, 1100, 1024, 994, 948, 914. Ϫ MS
(m/z): 462.4 (M Ϫ 1). Ϫ HPLC tR: 21.89 min.
[2]
[3]
[2a] E. De Clercq, J. Med. Chem. 1995, 38, 2491Ϫ2517. Ϫ [2b] E.
De Clercq, AIDS Res. Human Retroviruses 1992, 8, 119Ϫ134.
[2c]
[2d]
Determination of the Partition Coefficients: Pa values were deter-
mined as follows: A sample of one of the compounds 4a؊h was
dissolved in 1.0 ml of 1-octanol. To this solution, 1.0 ml of aqueous
phosphate buffer solution (10 mmol, pH 6.8) was added. After mix-
ing the phases extensively for 10 min. (vortex) and separation by
centrifugation (3 min. at 9000 rpm), aliquots of each phase were
analyzed by analytical HPLC [Merck Lichrocart column filled with
Lichrospher 100 reversed-phase silica gel RP-18 (5 µm), gradient:
18Ϫ100% CH3CN in water (0Ϫ16 min.), 100% CH3CN (16Ϫ19
min.), 18% CH3CN (19Ϫ31 min), flow rate 0.5 ml/min., UV detec-
tion at 269 nm]. By quantification of the peak areas, the Pa values
were obtained from the quotient.
Ϫ
J. Balzarini, Pharm. World Sci. 1994, 16, 113Ϫ126. Ϫ
P. Herdewijn, J. Balzarini, E. De Clercq, R. Pauwels, M. Baba,
S. Broder, H. Vanderhaeghe, J. Med. Chem. 1987, 30,
1270Ϫ1278.
P. A. Furman, J. A. Fyfe, M. H. St. Clair, K. Weinhold, J. L.
Rideout, G. A. Freeman, S. N. Lehrmann, D. P. Bolognesi, S.
Broder, Proc. Natl. Acad. Sci., USA 1986, 83, 8333Ϫ8337.
O. Heidenreich, F. Eckstein, Nucleosides & Nucleotides 1991,
10, 535Ϫ536.
Z. Hao, D. A. Cooney, N. R. Hartman, C.-F. Perno, A. Frid-
land, A. L. DeVico, M. G. Sarngadharan, S. Broder, D. G.
Johns, Mol. Pharmacol. 1988, 34, 431Ϫ435.
Z. Hao, D. A. Cooney, D. Farquhar, C. F. Perno, K. Zhang, R.
Masood, Y. Wilson, N. R. Hartman, J. Balzarini, D. G. Johns,
Mol. Pharmacol. 1990, 37, 157Ϫ163.
P. Herdewijn, J. Balzarini, E. De Clercq, in: Advances in Antivi-
ral Drug Design (Ed. E. De Clercq), 1993, Vol. 1, 233Ϫ318.
J. Balzarini, P. Herdewijn, E. De Clercq, J. Biol. Chem. 1989,
264, 6127Ϫ6133.
[4]
[5]
[6]
[7]
[8]
Kinetic Data: a) All aqueous buffers: 1.0 ml aqueous buffer (50
m acetate buffer, pH 4.6; 50 m TRIS buffer, pH 6.9; 50 m
phosphate buffer, pH 7.3 and 50 m borate buffer, pH 8.9) was
equilibrated at 37°C and mixed with 0.5 ml of a solution of 1.0 mg
of compound 4a؊h in 1.2 ml H2O (1.9 m), also equilibrated at
37°C (final concentration of the hydrolysis mixture: 0.63 m). To
obtain the kinetic data, aliquots (100 µl) were removed from the
hydrolysis mixture, the reaction was terminated by the addition of
5.0 µl acetic acid, and subsequently analyzed by analytical HPLC
[Merck Lichrocart column filled with Lichrospher 100 reversed-
phase silica gel RP-18 (5 µm), gradient: 18Ϫ100% CH3CN in tetra-
butylammonium phosphate, pH 3.8 (0Ϫ16 min.), 100% CH3CN
(16Ϫ20 min.), 18% CH3CN (20Ϫ31 min.), 0.5 ml/min. flow rate,
UV detection at 269 nm]. The degradation of 4 was followed by
quantification of the peak areas in the HPLC chromatograms. The
rate constants k were determined from the slopes of the logarithmic
degradation curves of the title compounds. The half-lives (t1/2) were
calculated from the rate constants k.
[9] [9a]
J. Cinatl Jr, B. Gröschel, R. Zehner, J. Cinatl, C. Pe´rigaud,
G. Gosselin, J.-L. Imbach, H. W. Doerr, Antiviral Chem. &
[9b]
Chemother. 1997, 8, 343Ϫ352. Ϫ
F. Dianzani, G. Antonelli,
O. Turriziani, E. Riva, E. Swimeoni, C. Signoretti, S. Strosselli,
M. Cianfriglia, AIDS Res. Human Retroviruses 1994, 10,
1471Ϫ1478.
[10] [10a]
[10b]
V. J. Stella, J. Med. Chem. 1980, 23, 1275Ϫ1285. Ϫ
A.
A. Sinkula, S. H. Yalkowsky, J. Pharm. Sci. 1980, 64, 181Ϫ210.
[11]
[12]
´
C. Perigaud, J.-L. Giradet, G. Gosselin, J.-L. Imbach, in: Ad-
vances in Antiviral Drug Design. (Ed. E. De Clercq), 1996, Vol.
2, 147Ϫ172.
R. J. Jones, N. Bischofberger, Antiviral Res. 1995, 27, 1Ϫ17 and
cited literature.
[13] [13a]
Y. Henin, C. Gouyette, O. Schwartz, J.-C. Debouzy, J.-M.
Neumann, T. Huynh-Dinh, J. Med. Chem. 1991, 34,
[13b]
1830Ϫ1837. Ϫ
C. Meier, J.-M. Neuman, F. Andre´, Y.
Henin, T. Huynh-Dinh, J. Org. Chem. 1992, 57, 7300Ϫ7308. Ϫ
[13c]
C. McGuigan, C. Nickson, J. Petrick, A. Karpas, FEBS
Lett. 1992, 310, 171Ϫ174. Ϫ [13d] C. McGuigan, C. Nickson, T.
J. OЈConnor, D. Kinchington, Antiviral Res. 1992, 17, 197Ϫ212.
b) RPMI-1640 (Roswell Park Memorial Institute) culture me-
dium containing 10% of heat-inactivated fetal calf serum (FCS):
this medium was used instead of aqueous buffers and the data were
collected in the same way.
[13e]
Ϫ
W. Thomson, D. Nicholls, W. J. Irwin, J. S. Al-Musha-
dani, S. Freeman, A. Karpas, J. Petrik, N. Mahmood, A. J. Hay,
[13f]
J. Chem. Soc., Perkin Trans. 1 1993, 1239Ϫ1245. Ϫ
A. D.
Briggs, M. Camplo, S. Freeman, J. Lundström, B. G. Pring,
Tetrahedron, 1996, 47, 14937Ϫ14950.
[14] [14a]
D. Farquhar, S. Khan, D. N. Srivastva, P. P. Saunders, J.
Antiretroviral Evaluation: Human immunodeficiency virus type 1
[HIV-1 (HTLV-IIIB)] was obtained from persistently HIV-infected
H9 cells as described previously[32]. Virus stocks were prepared
from the supernatants of HIV-1-infected MT-4 cells. HIV-2 (ROD)
was provided by Dr. L. Montagnier (Pasteur Institute, Paris,
France). CEM/0 cells were obtained from the American Tissue Cul-
ture Collection (Rockville, MD), and CEM/TKϪ cells were a gift
from Prof. S. Eriksson and Prof. A. Karlsson (Karolinska Institute,
Stockholm, Sweden). CEM cells were infected with HIV as de-
scribed previously[33]. Briefly, 4 ϫ 105 cells/ml were infected with
HIV-1 or HIV-2 at 100 CCID50 (50% cell culture infective dose)/ml
of cell suspension. Then, 100 µl of the infected cell suspension was
transferred to a 96-well microtitre plate and mixed with 100 µl of
appropriate diluted solutions of the test compounds. After 4 days,
giant cell formation was recorded microscopically in the HIV-in-
fected cell cultures. The 50% effective concentration (EC50) and the
50% cytotoxic concentration (CC50) of the test compounds were
defined as the compound concentrations required to inhibit virus-
induced cytopathicity by 50% or to reduce the number of viable
cells in mock-infected cell cultures by 50%, respectively.
[14b]
Med. Chem. 1994, 37, 3902Ϫ 3909. Ϫ
D. Farquhar, S.
Khan, M. C. Wilkerson, B. S. Anderson, Tetrahedron Lett.
[ 14c]
1995, 36, 655Ϫ658.Ϫ
J. K. Sastry, P. N. Nehete, S. Khan,
B. J. Nowak, W. Plunkett, R. B. Arlinghaus, D. Farquhar, Mol.
Pharmacol. 1992, 41, 441Ϫ445.
[15] [15a]
I. Lefebvre, C. Pe´rigaud, A. Pompon, A.-M. Aubertin, J.-
L. Girardet, A. Kirn, G. Gosselin, J.-L. Imbach, J. Med. Chem.
1995, 38, 3941Ϫ3950. Ϫ [15b] J.-L. Girardet, C. Pe´rigaud, A.-M.
Aubertin, G. Gosselin, A. Kirn, J.-L. Imbach, Bioorg. Med.
[15c]
Chem. Lett. 1995, 5, 2981Ϫ2984. Ϫ
C. Pe´rigaud, A.-M.
Aubertin, S. Benzaria, H. Pelicano, J.-L. Girardet, G. Maury,
G. Gosselin, A. Kirn, J.-L. Imbach, Biochem. Pharmacol. 1994,
48, 11Ϫ14. Ϫ [15d] F. Puech, A. Pompon, I. Lefebvre, G. Gosse-
lin, J.-L. Imbach, Bioorg. Med. Chem. Lett. 1992, 2, 603Ϫ606.
[15e]
Ϫ
F. Puech, G. Gosselin, I. Lefebvre, A. Pompon, A.-M.
Aubertin, A. Kirn, J.-L. Imbach, Antiviral Res. 1993, 22,
155Ϫ174.
[16] [16a]
C. McGuigan, R. N. Pathirana, J. Balzarini, E. De Clercq,
[16b]
J. Med. Chem. 1993, 36, 1048Ϫ1052. Ϫ
C. McGuigan, D.
Cahard, H. M. Sheeka, E. De Clercq, J. Balzarini, J. Med.
[16c]
Chem. 1996, 39, 1748Ϫ1753. Ϫ
J. Balzarini, A. Karlsson,
S. Aquaro, C.-F. Perno, D. Cahard, L. Naesens, E. De Clercq,
C. McGuigan, Proc. Natl. Acad. Sci. USA 1996, 93,
7295Ϫ7299. Ϫ [16d] J. Balzarini, H. Egberink, K. Hartmann, D.
Eur. J. Org. Chem. 1998, 837Ϫ846
845

C. Meier, E. De Clercq, J. Balzarini
FULL PAPER
[24]
[25]
Cahard, T. Vahlenkamp, E. Thormar, E. De Clercq, C. McGui-
gan, Mol. Pharmacol. 1996, 50, 1207Ϫ1213. Ϫ [16e] J. Balzarini,
J. Kruining, O. Wedgwood, C. Pannecouque, S. Aquaro, C. F.
Perno, L. Naesens, M. Witvrouw, R. Heijtink, E. De Clercq, C.
McGuigan, FEBS Lett. 1997, 410, 324Ϫ328.
T. P. Zimmermann, W. B. Mahony, K. L. Prus, J. Biol. Chem.
1987, 262, 5748Ϫ5754.
Using a somewhat different methodology, a Pa value of 0.96
was determined for AZT (1): J. Balzarini, J. M. Cools, E. De
Clercq, Biochem. Biophys. Res. Commun. 1989, 158, 413Ϫ422.
It is well known in the literature that phosphodiesters are ex-
tremely stable towards chemical hydrolysis at physiological pH
values: C. A. Bunton, M. M. Mhala, K. G. Oldham, C. A.
Vernon, J. Chem. Soc. 1960, 3293Ϫ3301. Even the ЉreactiveЉ
diphenylphosphate showed a half-life of 180 years(!) at pH 7.0
and 100°C: A. J. Kirby, M. Younas, J. Chem. Soc. (B) 1970,
510Ϫ513.
[26]
[17] [17a]
C. Meier, L. W. Habel, J. Balzarini, E. De Clercq, Liebigs
[17b]
Ann. 1995, 2203Ϫ2208. Ϫ
C. Meier, Angew. Chem. 1993,
105, 1854Ϫ1856; Angew. Chem., Int. Ed. Engl. 1993, 32,
[17c]
1704Ϫ1706. Ϫ
Clercq, Liebigs Ann. 1995, 2195Ϫ2202. Ϫ
Habel, W. Laux, J. Balzarini, E. De Clercq, Nucleosides & Nu-
C. Meier, L. W. Habel, J. Balzarini, E. De
[17d]
C. Meier, L. W.
L. W. Habel, C. Meier, Nu-
[17e]
cleotides 1995, 14, 759Ϫ762. Ϫ
[27]
Enzymatically catalyzed approaches have also shown some limi-
tations in the hydrolyses of non-natural products: G. Valette,
A. Pompon, J.-L. Girardet, L. Cappellacci, P. Franchetti, M.
cleosides & Nucleotides 1997, 16, 1311Ϫ1314.
[18] [18a]
C. Meier, Angew. Chem. 1996, 108, 77Ϫ79; Angew. Chem.,
[18b]
Int. Ed. Engl. 1996, 35, 70Ϫ72. Ϫ
C. Meier, T. Knispel, M.
[18c]
´
Grifantini, P. La Colla, A. G. Loi, C. Perigaud, G. Gosselin, J.-
Lorey, J. Balzarini, Int. Antiviral News 1997, 5,183Ϫ185. Ϫ
C. Meier, T. Knispel, M. Lorey, J. Balzarini, GIT Laborfach-
zeitschrift, 1997, 41, 965Ϫ967.
C. Meier, University of Würzburg, 1997, unpublished results.
C. Meier, M. Lorey, E. De Clercq, J. Balzarini, Bioorg. Med.
Chem. Lett. 1997, 7, 99Ϫ104.
L. Imbach, J. Med. Chem. 1996, 39, 1981Ϫ1990.
[28] [28a]
W. Thomson, D. Nicholls, A. G. Mitchell, J. A. Corner,
[19]
[20]
W. J. Irwin, S. Freeman, J. Chem. Soc., Perkin Trans. 1 1993,
[28b]
2303Ϫ2308. Ϫ
A. G. Mitchell, D. Nicholls, W. J. Irwin, S.
Freeman, J. Chem. Soc., Perkin Trans. 2 1992, 1145Ϫ1150.
C.-F. Perno, R. Yarchoan, J. Balzarini, A. Bergamini, G. Mi-
lanese, R. Pauwels, E. De Clercq, G. Rocci, R. Calio, Antiviral
Res. 1992, 17, 289Ϫ304.
[29]
[30]
[21]
[22]
C. Meier, T. Knispel, E. De Clercq, J. Balzarini, Bioorg. Med.
Chem. Lett. 1997, 7, 1577Ϫ1582.
Beside the anti-HIV active dideoxynucleosides mentioned here,
the cycloSal-concept has also been applied to the anti-tumor
agent 5-fluoro-2Ј-deoxyuridine:
A. Fridland, M.C. Connelly, R. Ashmun, Mol. Pharmcol. 1995,
49, 829Ϫ837.
[31]
[32]
A. Wissner, M. L. Carroll, J. Med. Chem. 1992, 35, 1650Ϫ166.
M. Popovic, M. G. Sarngadharan, E. Read, R. C. Gallo, Sci-
ence 1984, 224, 497Ϫ500.
[22a]C. Meier, M. Lorey, E. De Clercq, J. Balzarini, Nucleosides &
Nucleotides 1997, 16, 1303Ϫ1306.
[22b]M. Lorey, C. Meier, E. De Clercq, J. Balzarini, Nucleosides &
Nucleotides 1997, 16, 789Ϫ792.
[33]
J. Balzarini, L. Naesens, J. Slachmuylders, H. Niphuis, I. Rosen-
berg, A. Holy, H. Schellekens, E. De Clercq, AIDS 1991, 5,
21Ϫ28.
[23]
S. Rim, K. L. Audus, R. T. Borchardt, Int. J. Pharm. 1986,
32, 79Ϫ85.
[97276]
846
Eur. J. Org. Chem. 1998, 837Ϫ846