AZT 5′-cholinephosphate as an anti-HIV agent: The study of biochemical properties and metabolic transformations using its 32P-labelled counterpart

Article abstract of DOI:10.1080/15257770601052257

Biochemical and metabolic transformations of 3′-azido-3′- deoxythymidine 5′-choline phosphate (1) were studied using its ³²P-labelled counterpart for the evaluation of possible reasons for its enhanced anti-HIV activity. An effective synthesis of ³²P- labelled 1 with a specific activity >1,000 Ci/mmol was developed by esterification of ³²P-phosphoric acid with choline in the presence of BrCN followed by the coupling of the resulting choline phosphate with 3′-azido-3′-deoxythymidine (AZT). Chemical and enzymatic stabilities of 1 as well as the dynamics of penetration through HL-60 cell membranes were studied at the concentrations comparable to its antiviral concentrations. The products of intracellular transformations of the studied nucleotide were identified. Copyright Taylor & Francis Group, LLC.

Full text of DOI:10.1080/15257770601052257

This article was downloaded by: [University of Memphis]
On: 07 September 2012, At: 05:46
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Nucleosides, Nucleotides and Nucleic
Acids
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/lncn20
AZT 5′-Cholinephosphate as an
Anti-HIV Agent: The Study of
Biochemical Properties and Metabolic
32
Transformations Using Its P-Labelled
Counterpart
b
a
a
Dmitry V. Yanvarev , Elena A. Shirokova , Maria V. Astapova &
b
Yury S. Skoblov
a
Engelhardt Institute of Molecular Biology, Russian Academy of
Sciences, Moscow, Russian Federation
b
Shemyakin–Ovchinnikov Institute of Bioorganic Chemistry, Russian
Academy of Sciences, Moscow, Russian Federation
Version of record first published: 08 Dec 2006
To cite this article: Dmitry V. Yanvarev, Elena A. Shirokova, Maria V. Astapova & Yury S. Skoblov
(
2007): AZT 5′-Cholinephosphate as an Anti-HIV Agent: The Study of Biochemical Properties and
32
Metabolic Transformations Using Its P-Labelled Counterpart, Nucleosides, Nucleotides and Nucleic
Acids, 26:1, 23-36
To link to this article: http://dx.doi.org/10.1080/15257770601052257
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation
that the contents will be complete or accurate or up to date. The accuracy of any
instructions, formulae, and drug doses should be independently verified with primary
sources. The publisher shall not be liable for any loss, actions, claims, proceedings,

demand, or costs or damages whatsoever or howsoever caused arising directly or
indirectly in connection with or arising out of the use of this material.

Nucleosides, Nucleotides, and Nucleic Acids, 26:23–36, 2007
Copyright ꢀC Taylor & Francis Group, LLC
ISSN: 1525-7770 print / 1532-2335 online
DOI: 10.1080/15257770601052257
ꢀ
AZT 5 -CHOLINEPHOSPHATE AS AN ANTI-HIV AGENT:
THE STUDY OF BIOCHEMICAL PROPERTIES AND METABOLIC
TRANSFORMATIONS USING ITS 32P-LABELLED COUNTERPART
Dmitry V. Yanvarev and Elena A. Shirokova
2
Engelhardt Institute of Molecular
Biology, Russian Academy of Sciences, Moscow, Russian Federation
Maria V. Astapova and Yury S. Skoblov Shemyakin–Ovchinnikov Institute
2
of Bioorganic Chemistry, Russian Academy of Sciences, Moscow, Russian Federation
ꢁ
ꢁ
ꢁ
2
Biochemical and metabolic transformations of 3 -azido-3 -deoxythymidine 5 -choline phosphate
32
(
1) were studied using its P-labelled counterpart for the evaluation of possible reasons for
3
2
its enhanced anti-HIV activity. An effective synthesis of P-labelled 1 with a specific activity
>
1,000 Ci/mmol was developed by esterification of 32P-phosphoric acid with choline in the presence
ꢁ
ꢁ
of BrCN followed by the coupling of the resulting choline phosphate with 3 -azido-3 -deoxythymidine
AZT). Chemical and enzymatic stabilities of 1 as well as the dynamics of penetration through
(
HL-60 cell membranes were studied at the concentrations comparable to its antiviral concentrations.
The products of intracellular transformations of the studied nucleotide were identified.
ꢁ
ꢁ
ꢁ
32
32
Keywords 3 -Azido-3 -deoxythymidine 5 -[ P]-choline phosphate; [ P]orthophos-
3
2
phoric acid; Phosphorylation; Synthesis of [ P]-labeled nucleotides
INTRODUCTION
Synthetic nucleoside analogues are widely used in medical practice for
therapy of viral infections. For many of them, the mechanism of action
involves intracellular phosphorylation into the corresponding triphosphates
that are incorporated into the growing chain of viral DNA and terminate the
DNA synthesis.[ This mechanism implies that the efficacy of inhibition of
viral reproduction is affected by at least several factors, in particular, the
ability of the compound to pass through the cell membrane, the efficacy of
1]
Received January 30, 2006; accepted June 15, 2006.
The work was supported by the Russian Foundation for Basic Research, projects 04-03-32481, 05-04-
4
9500 and 04-04-49617; and the program of Presidium of Russian Academy of Sciences “Molecular and
Cellular Biology.”
Address correspondence to Dmitry V. Yanvarev, Engelhardt Institute of Molecular Biology, Russian
Academy of Sciences, 32 Vavilov st., Moscow, 119991, Russian Federation. E-mail: vsupport@bk.ru
2
3

2
4
D. V. Yanvarev et al.
the triphosphorylation cascade, and the ability of the resulting nucleoside
triphosphate to be recognized by viral DNA polymerases.
One of the trends in the design of novel highly effective antiviral
agents is further modifications of the known synthetic analogues affording
lower toxicity, better penetration into cells, and/or prolonged action. The
nucleoside modified in this way is transformed in vivo (cell or organ-
ism) into the parent nucleoside and affects the virus according to the
ꢁ
ꢁ
ꢁ
ꢁ
known mechanism. 3 -Azido-3 -deoxythymidine 5 -H-phosphonate (AZT 5 -
H-phosphonate, phosphazide, Nicavir) can serve a successful example of
this approach. This drug has been used in anti-HIV therapy since 1999 as a
[2]
less toxic AZT “substituent” and its in vivo precursor.
Earlier we showed that some modified AZT derivatives inhibited
the HIV replication in cell cultures more effectively than the parental
nucleoside.[
3–5]
One of these compounds was AZT 5 -choline phosphate 1:
ꢁ
its antiviral activity measured as IC50 (the concentration, at which 50% of
the virus replication is inhibited) was 0.6 nM versus 17 nM of AZT, whereas
their toxicities were identical.[ The latter is not surprising because, as a
component of many natural compounds, a choline phosphate group must
not contribute to the compound toxicity. Obviously, the anti-HIV activity
of this compound is due to its ability to be intracellularly converted into
6]
ꢁ
AZT 5 -triphosphate, which terminates the synthesis of viral DNA. According
ꢁ
to the chemical structure, compound 1 can donate either/both AZT 5 -
monophosphate or/and AZT. For the understanding of the mechanism, we
ꢁ
synthesized AZT 5 -choline phosphate labelled with phosphorus-32 (1) with
a high specific activity (>1000 Ci/mmol) and studied some of its biochem-
ical and metabolic properties in cell cultures. The use of the radioactive
phosphorus enabled the study of both intracellular transformations of the
phosphate group and alterations in the penetration of the tested compound
through cell membranes at micromolar concentrations and those close to
the concentrations of 90% inhibition of HIV-1 replication (IC ). It also
90
allowed us to compare the dynamics of penetration into cells of the studied
AZT choline phosphate 1 with that of parental AZT.
ꢁ
Herein we report the synthesis of labelled AZT 5 -choline phosphate 1
and present the results of the study of its penetration into HL-60 cells, the
composition of metabolites and potencies of some enzymes to degrade the
studied nucleotide.

ꢁ
32
Synthesis and Metabolism of AZT 5 -[ P]-Cholinephosphate
25
RESULTS
Chemistry
Phosphorylation of an alcoholic hydroxy group with phosphorus oxy-
chloride is the most common chemical way to obtain the correspond-
[7]
ing phosphates (including nucleoside phosphates). The method is re-
producible and allows high yields irrespective of the nucleoside nature.
However, this approach is inapplicable for radiochemical synthesis of
compounds with the high specific activity, since the specific activity of
3
2
POCl as a radioactive starting compound is markedly lower than that
3
32
32
of P-orthophosphoric acid. In addition, P-orthophosphoric acid is sub-
stantially cheaper and more stable than POCl . With all this in mind, we
used carrier-free P-orthophosphoric acid for the synthesis of phosphorus-
labeled compound 1.
32
3
32
Two approaches can be used for the preparation of the target phos-
phate 1 (Scheme 1). The first approach includes phosphorylation of a
nucleoside followed by the coupling of the resulting nucleotide with
32
choline. The second method includes esterification of P-phosphoric acid
with choline followed by the coupling of the resulting choline phosphate
with the nucleoside.
SCHEME 1 Approaches to the synthesis of target phosphate 1.
The scheme resulting in AZT monophosphate 2 at the first stage seemed
to be more convenient, as the intermediate nucleotide is easy to detect and
isolate by HPLC. Chemically, a phosphorus-32-labelled nucleotide with the
specific activity of at least 1000 Ci/mmol can be obtained by interaction of
32
the corresponding nucleoside and H3 PO in the presence of the coupling
4
agent. The nucleoside and the coupling agent are used in this reaction in
large excess (10-fold or more) relative to phosphoric acid. We performed
the synthesis of nucleotide 2 in the presence of BrCN (Scheme 2), which
ꢁ
proved to be more effective for AZT 5 -phosphorylation than commonly
[8]
used coupling agents (DCC, TPS-Cl, and trichloroacetonitrile).
The
◦
reaction proceeded fast (≤1 minute) even at 0 C, and the further incuba-
tion only increased the amount of side products, particularly, symmetrical

2
6
D. V. Yanvarev et al.
ꢁ
SCHEME 2 Synthesis of phosphate 1 via AZT-5 -phosphate.
ꢁ
ꢁꢁ
bis-AZT 5 ,5 -diphoshate. However, acid hydrolysis of the latter allowed an
essential increase in the yield of the target 2: after the 30 minutes incubation
of the reaction mixture with 1M HCl at room temperature, the yield of
32
monophosphate 2 achieved 50% relative to starting H3 PO4.
Monophosphate 2 obtained at the first stage (Scheme 2) must be
isolated and purified prior to the next step, since the choline hydroxy group
is considerably less reactive than that of AZT. After reversed-phase HPLC,
nucleotide 2 was esterified with choline under the conditions described for
the first step to give 25–30% of target 1. However, the total yield of product
1
was only 12–15%.
We used the other route in the hope to increase the product yield using
choline phosphate 3 as an intermediate compound (Scheme 3).
SCHEME 3 Synthesis of phosphate 1 via choline phosphate.
Variations of the reaction conditions (e.g., the reagent ratio, tempera-
ture, or reaction time) only insignificantly influenced on the yield of the
esterification product, which ranged from 25% to 30%. The esterification
32
with the nucleoside was performed without isolating P-choline phosphate
. The coupling of the latter with AZT was very effective and gave 90% of
3
diester 1. Although the total yield did not exceed 25%, it was significantly
higher than that of compound 1 obtained according to Scheme 2. The
specific activity of the product was 1000 Ci/mmol and the radioactive
purity exceeded 95%. This allowed us to study metabolic transformations
of compound 1.
For the identification of the synthesized 1–3 by NMR and UV spectro-
scopies, we synthesized these compounds using the indicator amount of

ꢁ
32
Synthesis and Metabolism of AZT 5 -[ P]-Cholinephosphate
27
radioactivity (specific activity of 0.1 Ci/mmol) and isolated them under the
conditions described in the Experimental section.
Cellular Metabolic Studies
The dynamics of penetration of compound 1 into cells was studied in
HL-60 cell culture. The dynamics of permeation of the radioactivity into
cells is shown in Figure 1.
As is seen in Figure 1, the level of radioactivity was constantly growing
during the 24-hour period and the curve lacked the saturation area. For a
better understanding of this fact, we studied the spectrum of metabolites
32
containing the P isotope in the water-soluble fraction of the cellular lysate.
We showed by HPLC that at any time point within the 0- to 24-hour time
period only 3–6% of the total radioactivity found in the cell was present in
cytoplasm as unchanged diester 1, whereas the rest portion of radioactive
products consisted of ester 3 and orthophosphate in a 4:1 ratio.
Obviously, following such effective intracellular destruction of diester 1
its concentration gradient inside and outside the cell must remain persistent
within the experimental time period. The linearity of the curve in Figure 1
confirms the postulate. As diester 1 is hydrolyzed in the cultural medium
by only 3–4% after 24 hours, the radioactive phosphorus isotope is likely to
pass into cell as unchanged nucleotide 1.
We established the intracellular distribution of metabolites containing
the 32P isotope by successive washing off the cells (after the 24-hour
incubation), cryolysis, extraction of the water-insoluble fraction with a
1
:1 chloroform-methanol mixture similarly to the procedure for the lipid
[4]
isolation, and two-dimensional TLC identification of the products. The
results are shown in the diagram (Figure 2a). The water-soluble fraction
contained 66% of the total intracellular radioactivity, 20% was found in the
lipid one, and 14% of radiolabelled products was detected in the fraction,
which was soluble neither in water nor in the chloroform-methanol mixture
FIGURE 1 Dynamics of penetration of diester 1 into HL-60 cells.

2
8
D. V. Yanvarev et al.
FIGURE 2 (a) Intracellular distribution of the radioactive label after the incubation of HL-60 cells with
32
compound 1. (b) The P distribution in the water-soluble fraction of the cellular lyzate.
(
“unidentified fraction”). The product distribution of the water-soluble
fraction is shown in Figure 2b. As is seen, the portions of choline phosphate
, H PO , and diester 1 totaled to 73%, 21%, and 6%, respectively.
3
3
4
The product composition in the lipid fraction isolated as described
in Folch et al.[ was studied by two-dimensional TLC.
9]
[10]
The major
radioactive metabolic products were phosphatidyl choline and phosphatidyl
ethanolamine (Figure 3). The content of phosphatidic acid was consider-
ably lower.
We also compared the dynamics of concurrent penetration into the
cells of the tested diester 1 and the parent AZT. To this end, we used
3
[2]
32
3
[
6- H]-AZT and detected the products based on the P and H levels.
The results of this experiment are shown in Figure 4.
As is seen in Figure 4, the intracellular amount of [6- H]-AZT was higher
3
3
2
than that of P-labelled product(s) for the first 8 hours. Since diester 1
was quantitatively metabolized to AZT in cells, one can claim that the AZT
amount resulted from diester 1 constitutes 94–97% of the intracellular 32
P
FIGURE 3 The TLC pattern of 32P-labelled metabolites in the lipid fraction. Identification of the
products was performed by comparison of their mobilitites with those of the corresponding unlabelled
authentic reference compounds.

ꢁ
32
Synthesis and Metabolism of AZT 5 -[ P]-Cholinephosphate
29
FIGURE 4 Dynamics of the cellular uptake of diester [32P]-1 and [6- H]-AZT. Both compounds were
added simultaneously, each at the concentration of 4 µM. The presented data are mean values of two
independent experiments made in two parallels.
3
amount. Hence, during the first 8 hours the intracellular AZT concentration
3
from [6- H]-AZT was higher than that formed from the degraded 1. Then
3
the rate of [6- H]-AZT penetration markedly fell, whereas the AZT from
diester 1 was accumulated, and the concentration difference between these
two products achieved approximately 30% after 24 hours. However, it is
noteworthy that the behavior of the penetration curve of compound 1
negligibly changed in the presence of the equivalent amount of AZT after
the 6- to 7-hour incubation.
We also evaluated how the presence of a tenfold excess of AZT affected
the cellular uptake of compound 1 (Figure 5). As is seen in Figure 5,
AZT obviously influenced upon the penetration of 1 into the cells, but this
influence was insignificant and cannot be explained by routine concurrent
interactions.
FIGURE 5 Dynamics of penetration of diester [32P]-1 (1) in the absence of AZT; (2) the presence of
1
µM AZT and (3) 10 µM AZT. Concentration of diester 1 in the cultural medium was 1 µM; specific
activity of 4 Ci/mmol. The data are mean values of three independent experiments.

3
0
D. V. Yanvarev et al.
FIGURE 6 Dynamic curves of diester [32P]-1 at the concentrations of 37 nM and 4 µM. Specific activity
of compound 1 was 4.2 Ci/mmol. The presented data are mean values of two independent experiments
made in two parallels.
The cellular uptake was also studied at the 37 nM concentration of
[6]
diester 1, which is close to its IC90 value (18 nM ) (Figure 6). Unlike the
curve corresponding to the concentration of 4 µM, the curve for the 37 nM
concentration has a saturated area after 1–2 hours of incubation. Thus, a
1
00-fold increase in the intercellular concentration of nucleotide 1 entails
only a 7-fold increase in its intracellular concentration.
Enzymatic Stability
Table 1 demonstrates the data on the chemical and enzymatic stability
of phosphate 1. The half-life (T1/2) of nucleotide 1 in the phosphate buffer
◦
at 37 C in the pH range of 5–8 exceeded 24 hours, whereas in 90% normal
human serum it amounted to 6.3 hours. The only hydrolysis product was
AZT in both cases. Taking into account that the hydrolysis rate of nucleotide
2
was one order of magnitude higher under these experimental conditions
if compared with diester 1 (data not shown), it is not surprising that
monophosphate 2 was not found in the reaction mixture.
The behavior of the enzymes contained in the HL-60 cellular lyzate was
studied as described in Skoblov et al.[ After centrifugation at 14000 g in
2]
100 mM PBS, the supernatant was found to hydrolyze compound 1 nearly
5
times slower than the suspension of the pellet after centrifugation. The
product spectra were identical in both cases.

ꢁ
32
Synthesis and Metabolism of AZT 5 -[ P]-Cholinephosphate
31
TABLE 1 Chemical and enzymatic stability of nucleotide 1
a
Product content at T1/2, %
AZT
(2) Half-life
Phosphate buffer (pH 5–8)
Snake venom preparation
90
50
3
98,5
99
93
99
99
10 >24 h
50 1 h
97 5 min
1,5 24 h
Snake venom phosphodiesterase
ꢁ
Snake venom 5 -nucleotidase
Phospholipase C
1
7
1
1
18 h
Cellular membrane preparation
Water-soluble fraction of the cellular lyzate
Human blood serum
32 min
2.5 h
6.3 h
aThe sum of AZT and (2) is taken as 100%.
We also attempted to determine the enzymes capable of degrading
nucleotide 1. The total enzymatic preparation of snake venom smoothly
hydrolyzed compound 1 to give a mixture of choline, AZT and 2 with T1/2
of about 60 minutes. By this moment AZT and monophosphate 2 were
accumulated in the reaction mixture in a 1:1 ratio. Since monophosphate
2
was digested by snake venom nucleotidase, the hydrolysis products after
the 4-hour incubation were found as a mixture of AZT, choline, and
orthophosphate.
Snake venom phosphodiesterase very effectively hydrolyzed diester 1 to
give choline and nucleotide 2 (with the half-life of 3–5 minutes).
ꢁ
Diester 1 was a poor substrate for 5 -nucleotidase. Its degradation
proceeded 1000 times slower if compared with that of nucleotide 2 and only
2
- to 3-fold exceeded the rate of chemical hydrolysis. This enzymatic activity
was likely due to the residual activity of phosphodiesterase contained in the
preparation as a 0.1% admixture.
Snake venom phospholipase C digested diester 1 nearly 50 times slower
than it degraded phosphatidyl choline and gave choline phosphate 3 and
AZT as the reaction products (data not given).
DISCUSSION
The enzymatic approach reported by Johnson and Walseth[11] is the
most effective method for the preparation of nucleotides labeled with ra-
dioactive phosphorus isotopes and is successfully used for their production.
However, the method has serious limitations due to enzyme specificity
and cannot be employed for the preparation of modified nucleotides
including AZT derivatives. The standard chemical synthesis of 32P-labelled
nucleotides consists of phosphorylation of the excess of the nucleoside
with 32P-orthophosphoric acid in the presence of a coupling agent. With
AZT as an example, we recently compared the efficacy of coupling agents

3
2
D. V. Yanvarev et al.
commonly used in phosphorylation reactions, such as DCC, water-soluble
[
8]
carbodiimide, trichloroacetonitrile, and TPSCl with the efficacy of BrCN
[12]
known as the coupling agent in oligonucleotide ligation.
We showed that
32
the use of BrCN was beneficial for the synthesis of P-AZTMP 2 if compared
not only with carbodiimides and trichloroacetonitrile, but also with TPSCl.
The yield of phosphate 2 in the presence of BrCN was comparable with that
in the presence of TPSCl, but the product isolation was much more difficult
in the latter case. Taking into account this experience, we used BrCN for the
synthesis of diester 2.
Chemical phosphorylation of choline with P-orthophosphoric acid was
considerably less effective than that of AZT. The use of a multifold excess
of choline was inapplicable because of low solubility of choline and its
32
◦
salts in organic solvents, particularly, at the reaction temperatures (0 C).
Therefore, we synthesized diester 1 using Scheme 2.
The dynamics of the cellular uptake of compound 1 was studied in HL-
6
0 cell culture, which was previously used by us for the metabolic studies of
ꢁ
[2]
AZT 5 -H-phosphonate. The curves presenting the penetration of diester
1
at the concentration of 4 µM differ from those for AZT (Figures 2 and
4
): for diester 1, the value of intracellular radioactivity was rising linearly
with the time of the experiment, whereas for AZT, it was rapidly growing
during the first 3 hours, and then the penetration rate evidently reduced
3
(
Figure 4). At any time point after 8 hours the intracellular level of H-
AZT was lower than that after the incubation with diester 1. At the same
time the interdependence of 1 and AZT on the penetration was insignificant
(Figure 5). This effect may be accounted for by the existence of various
routes of penetration for 1 and AZT and/or by competitive passing of other
nucleotides, for example, thymidine, contained in the cultural medium at
higher concentrations than diester 1.
An absolutely different dynamic curve was obtained when the concen-
tration of nucleotide 1 was decreased by approximately 100 times (37 nM).
The saturation area was rapidly achieved, and the level of intracellular
radioactivity was only 7 times lower, whereas the extracellular concentration
declined by 100 times. This contrast to the dynamics of penetration of
the parent AZT allows a suggestion that compound 1 may have some
other channels to cross the cytoplasmic membrane. However, it should be
noted that at the concentration of 37 nM the contribution of extracellular
hydrolysis of diester 1 was higher and achieved 30% after 24 hours. In
addition, the result also may be partially affected by nonspecific adsorption
of radioactive metabolites on the cellular membrane.
The data of enzymatic hydrolysis showed that snake venom phosphodi-
esterase effectively hydrolyzed compound 1 to give choline and phosphate 2.
ꢁ
Like 5 -nucleotidase, spleen phosphodiesterase could not hydrolyze diester
1
3
, whereas phospholipase C digested it to give AZT and choline phosphate
. The comparison of these data with the dynamics and the composition of

ꢁ
32
Synthesis and Metabolism of AZT 5 -[ P]-Cholinephosphate
33
hydrolysis products in HL-60 cells makes it possible to assume that diester
1
is likely to be affected by phosphodiesterase I and then by phosphatase
ꢁ
(
or 5 -nucleotidase) in blood serum, whereas in HL-60 cells it is primarily a
target for phospholipase C bound to the cell membrane.
The analysis of the composition of phosphorus-32-labelled metabolic
products in the HL-60 cellular lyzate showed that compound 1 was degraded
to AZT, which was then phosphorylated to the corresponding phosphate
ꢁ
derivatives. Thus, diester 1 is an AZT rather than AZT 5 -phosphate prodrug.
A higher anti-HIV activity of compound 1 if compared with AZT may be
due to a better cellular uptake rather than specific metabolic transforma-
tions. This statement correlates well with the earlier obtained data on the
metabolism of AZT H-phosphonate.[ The latter phosphonate also was
rapidly hydrolyzed to AZT in HL-60 cells, but its less effective penetration
into cells (one-third of that of AZT) was in the author’s opinion the reason
for a decreased antiviral activity.
2]
EXPERIMENTAL
[32
P]-Orthophosphoric acid with a specific activity of 8000 Ci/mmol
(
296 PBq/mol) was obtained from the Institute of Reactor Materials
3
(Zarechnyi, Russia); [6- H]-AZT was prepared as described in Sidorov et
[14]
al.;
the solvents and reagents were dried directly before use. Cholinium
iodide was from Acros; 98% H PO was from Fluka, AZT was a kind gift
3
4
of AZT Association (Russia). Snake venom (Crotalus attrox), snake venom
ꢁ
phosphodiesterase (Crotalus attrox) and 5 -nucleotidase (Crotalus attrox)
were from Fluka. Phospholipase C (Clostridium perfringens) was from Serva
(Heidelberg, Germany).
HPLC was performed on a Gilson chromatograph (France) supplied
with a UV detector type 315 and a flow-rate Beckman 170 radioactive detec-
tor. Reversed-phase HPLC was performed on a Nucleosil 100 C-18 (5 µm)
column (4 × 150 mm) eluting with buffer A: 50 mM TEAB; buffer B: 75%
ethanol; gradient: 0–5 minute, 0% B; 5–7 minute, 0–12% B; 7–20 minute,
12–20% B; 20–30 minute, 20–50% B; the flow rate was 0.5 ml/minute. Ion-
exchange HPLC was performed on a Nucleosil-NH column (4 × 50 mm)
2
eluting with buffer C: water; buffer D: 500 mM NaH PO : 0–40 minute,
2
4
0
–30% D; 40–50 minute, 100% D; 50–60 minute, 100–0% D; the flow rate
was 0.4 ml/minute. TLC was performed on Kieselgel 60 F254 plates (Merck)
developing in dioxane-aq. ammonia-i-propanol-water 3:3:3:1.
NMR spectra were registered on an AMX III-400 spectrometer (Bruker)
1
with the working frequency of 400 MHz for H and Me Si as an internal
4
31
standard; 162 MHz for P NMR (with phosphorus-proton interaction
decoupling, 85% H PO as an external standard); and 100 MHz for
C NMR. UV absorption spectra were registered on a Specord M-40
3
4
13

3
4
D. V. Yanvarev et al.
spectrophotometer in water. Radioactivity was measured on a LS-counter
SL-4000 Intertechnique (France) by the Cherenkov method. The prod-
ucts were detected by autoradiography and analyzed on an InstantImager
(Packard Instrument Company, USA). Quantitative TLC analysis of reaction
mixtures was performed using the Imager.
ꢁ
ꢁ
ꢁ
ꢁ
3
-Azido-2 ,3 -dideoxythymidine 5 -choline Phosphate (1)
Orthophosphoric acid (10 nmol, 10 µl of 1 mM solution in water) and
32
absolute acetonitrile (40 µl) were added to a solution of 10 mCi [ P]H PO
3
4
in 0.5 mM HCl, and the solvent was evaporated. The residue was dissolved
in 5 M aqueous pyridine (20 µl), and cholinium iodide (1 µmol, 231 µg)
◦
was added. The reaction mixture was kept for 10 minutes at 0 C, a solution
(
6 µl) of 50 mM BrCN (0.3 µmol) in acetonitrile was added, and after 20
◦
minutes at 0 C the solvent was evaporated. The residue was dissolved in 5 M
aqueous pyridine (20 µl), and AZT (1 µmol, 267 µg) and a solution (6 µl)
of 50 mM BrCN (0.3 µmol) in acetonitrile were added. After 10 minutes,
30% aq. ammonia (10 µl) was added, the solvent was evaporated, and the
residue was coevaporated with water (50 µl). The residue was dissolved
in 50 mM TEAB (50 µl) and applied on a reversed-phase HPLC column
eluted in the system given above. The yield of the product was 18–20%,
the radiochemical purity 99%, specific activity 1,000 Ci/mmol. R : 0.6; UV:
f
λ
266.3 nm (8440) (pH 7.0, water), λmax/λminute 3.3; retention time: 18.5
max
1
ꢁ
minute. H-NMR: 7.63d (J 1 Hz, 1H, H-6); 6.20t (J 6.5 Hz, 1H, H-1 ); 4.43 dt
ꢁ
(
J 6.5 and 11 Hz, 1H, H-3 ); 4.25 m (2 H, CH O choline); 4.12–4.04 m (3H,
2
ꢁ
ꢁ
2
H-5 +H-4 ); 3.61 m (2 H, CH N choline); 3.16 s (3H, Me choline) 2.47
2
3
ꢁ
31 13
t (J 6.5 Hz, 2H, 2 H-2 ); 1.86 s (3H, Me-Thy). P-NMR: −0.16 s. C-NMR:
ꢁ
1
8
5
1
66.8, 151.9 (C-2 and C-4); 137.7 (C-6); 111.9 (C-5); 85.4 (C-1 ); 83.1 d (J
.5, C-4 ); 66.3 ts (J
ꢁ
ꢁ
ꢁ
ꢁ
3.9, CH N choline); 65.4 d (J 5.2, C-5 ); 60.5 (C-3 );
NC
2
9.8 d (J 4.9, CH O choline); 54.3 ts (J NC 3.7, Me choline); 36.5 (C-2 );
2.0 (Me-Thy).
2
3
Enzymatic Tests
Unlabelled diester 1 was applied onto a Dowex 50 (H ) column, eluted
+
with 10% ammonia, the solvent was removed in vacuum, and the residue
was dissolved in water up to the concentration of 114 µM and used in the
further experiments as a stock solution.
Hydrolysis with the cellular lyzate. The suspension of HL-60 cell culture
6
(
12·10 cells) in 5 mM Tris-HCl (500 µl, pH 7.5) was subjected to four
freezing-melting cycles using liquid nitrogen and centrifuged for 5 minute
at 14,000 g. A solution of 55 mM MgCl in 500 mM Tris-HCl (5 µl, pH 7.5)
2
was added to the clear lysate (50 µl), the mixture was heated thermostati-
◦
cally at 37 C for 5 minute and 114 µM diester 1 (2 µl, 3.2 Ci/mmol) was

ꢁ
32
Synthesis and Metabolism of AZT 5 -[ P]-Cholinephosphate
35
added. Aliquots (2.5 µl) were taken out, applied onto a TLC silica gel plate,
eluted in the system given above, and analyzed using the Imager.
6
Hydrolysis with membrane preparations. The HL-60 cells (12 × 10 )
cultivated, washed and degraded as described in the section “Cellular
experiments.” The resulting suspension was centrifuged for 5 minute at
1
4,000 g, and the supernatant was removed. The membrane fraction was
◦
washed with PBS cooled to 5 C (2 × 5 ml), suspended in 5 mM Tris-HCl
◦
buffer (200 µl, pH 7.5), heated thermostatically at 37 C for 5 minutes, and
diester 1 (114 µM 4 µl) was added. Aliquots (0.5 µl) were taken in certain
intervals, chromatographed by TLC, and analyzed using the Imager.
Hydrolysis with human blood serum. Human blood serum (100 µl) was
◦
heated at 37 C for 5 minutes and 114 µM diester 1 with a specific activity
of 2.9 Ci/mmol (4 µl) was added. Aliquots (0.5 µl) were taken out and
added to methanol (10 µl). The suspension was centrifuged for 2 minutes at
1
4,000 g, supernatant was concentrated in vacuum to the volume of 4–5 µl,
applied onto a TLC silica gel plate, eluted in the system given above, and
analyzed using the Imager.
Hydrolysis with total snake venom preparation (Crotalus attrox). The prepa-
ration (0.25 mg) was dissolved in 50 mM Tris-HCl (pH 8.0, 200 µl)
◦
containing 10 mM MgCl . The solution was heated at 37 C for 5 minutes
2
and a solution of 114 µM diester 1 (8 µl, 3.2 Ci/mmol) was added. Aliquots
were taken, applied onto a TLC silica gel plate, eluted in the system given
above, and analyzed using the Imager.
Hydrolysis with phosphodiesterase. The enzyme (0.5 U according to the
di(p-nitrophenyl) phosphate test) was dissolved in 50 mM Tris-HCl (pH 8.0,
5
0 µl) containing 10 mM MgCl and 0.11 M NaCl. The solution was heated
2
◦
at 37 C for 5 minutes and a solution of 114 µM diester 1 (2 µl, 2.8 Ci/mmol)
was added. Aliquots were taken, applied onto a TLC silica gel plate, eluted
in the system given above, and analyzed using the Imager.
ꢁ
ꢁ
Hydrolysis with 5 -nucleotidase. The enzyme (1 U according to the 5 -
AMP-dephosphorylation test) was dissolved in 50 mM Tris-HCl (pH 8.0, 50
µl) containing 10 mM MgCl and 0.11 M NaCl. The solution was heated at
2
◦
3
7 C for 5 minutes and a solution of 114 µM diester 1 (2 µl, 3.2 Ci/mmol)
was added. Aliquots (0.5 µl) were taken, applied onto a TLC silica gel plate,
eluted in the system given above, and analyzed using the Imager.
Cellular Experiments
Promyelocytic leukemia cells HL-60 were cultivated in the RPMI–1640
◦
medium containing 2 mM L-Gln and 10% calf serum at 37 C in the
6
atmosphere with 5% CO up to the density of 10 cells/ml. The radioactive
2
1
(4–5 Ci/mmol) was added and after 0, 1, 3, 7, and 24 hours two aliquots
(
3 ml) were taken out in parallel and centrifuged at 180 g for 4 minutes
◦
◦
at 4 C. The cells were washed twice 5 ml with PBS cooled to 5 C and

3
6
D. V. Yanvarev et al.
suspended in 5 mM Tris-HCl buffer (200 µl, pH 7.5). The cells were
degraded by the fourfold freezing-melting procedure in liquid nitrogen,
methanol (400 µl) was added, and the resulting suspension was centrifuged
(
2 minutes at 14,000 g). The supernatant was concentrated in vacuum
◦
at 37 C. The residue was dissolved in TEAB (25 µl) and analyzed by
HPLC.
The lipid composition of the membrane fraction was analyzed by silica
gel TLC with unlabelled authentic samples of the corresponding lipid
components as reference compounds. Phospholipids were detected by the
Vaskovsky and Kostevsky method.[
13]
REFERENCES
ꢁ
1
. Balzarini, J.; Baba, M.; Pauwels, R.; Herdewijn, P.; De Clerq, E. Anti-retrovirus activity of 3 -fluoro-
ꢁ
ꢁ
ꢁ
and 3 -azido-substituted pyrimidine 2 ,3 -dideoxynucleoside analogues. Biochem. Pharmacol 1988, 37,
847–2856.
2
2
. Skoblov, Yu.S.; Karpenko, I.L.; Shirokova, E.A.; Popov, K.E.; Andronova, V.A.; Galegov, G.V.;
ꢁ
Kukhanova, M.K. Intracellular metabolism and pharmacokinetics of 5 -hydrogenphosphonate of
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
3
1
-azido-2 ,3 -dideoxythymidine, a prodrug of 3 -azido-2 ,3 -dideoxythymidine. Antiviral Res. 2004, 63,
07–113.
3. Shipitsyn, A.V.; Zakirova, F.Z.; Belanov, E.F.; Pronyaeva, T.R.; Feduk, N.V.; Kukhanova, M.K.;
Pokrovsky, A.G. Phosphorodiamides as Prodrug for Antiviral Nucleosides. Nucleosides, Nucleotides
&
Nucleic Acids 2003, 22, 963–966.
4
. Shirokova, E.A.; Jasko, M.V.; Khandazhinskaya, A.L.; Ivanov, A.V.; Yanvarev, D.V.; Skoblov, Yu.S.;
Mitkevich, V.A.; Bocharov, E.V.; Pronyaeva, T.R.; Fedyuk, N.V.; Kukhanova, M.K.; Pokrovsky, A.G.
Uncharged AZT and D4 T derivatives of phosphonoformic and phosphonoacetic acids as anti-HIV
pronucleosides. J. Med. Chem. 2004, 47, 3606–3614.
5
6
7
. Shirokova, E.A.; Jasko, M.V.; Khandazhinskaya, A.L.; Yanvarev, D.V.; Skoblov, Yu.S.; Pronayeva, T.R.;
Fedyuk, N.V.; Pokrovsky, A.G.; Kukhanova, M.K. New lipophilic derivatives of AZT and D4 T 5 -
phosphonates. Nucleosides, Nucleotides & Nucleic Acids 2003, 22, 981–985.
ꢁ
. Shipitsyn A.V.; Yasko M.V.; Shirokova E.A.; Kukhanova M.K.; Pronyaeva T.R.; Feduk N.V.; Pokrovsky
ꢁ
ꢁ
ꢁ
A.G.; Gosselin G.; Perigaud, C. 5 -Choline phosphate 3 -azido-3 -deoxythymidine. PCT WO
004/011480 A1, 05.02, 2004.
2
. Turcotte, J.G.; Pivarnik, P.E.; Shirali, S.S.; Singh, H.K.; Sehgal, R.K.; Macbride, D.; Jang, N.I.; Brown,
ꢁ
P.R. Preparative-scale high-performance liquid chromatographic separation and purification of 3 -
azido-3 -deoxythymidine-5 -phosphate. J. Chromatogr 1990, 19, 55–61.
ꢁ
ꢁ
8
. Yanvarev, D.V.; Shirokova, E.A.; Skoblov, Yu.S. The Study of Coupling Agents for Phosphorylation
ꢁ
ꢁ
32
of 3 -Azido-3 -deoxythymidine with [ P]Orthophosphoric Acid. Bioorg. Khim. 2005, 31, 399–403.
. Folch, J.; Lees, M.; Sloane-Stanley, G.H. A simple method for the isolation and purification of total
lipides from animal tissues. J. Biol. Chem. 1957, 226, 497–501.
9
1
0. Skipski, V.P.; Barclay, M.; Reichman, E.S.; Good, J.J. Separation of acidic phospholipids by one-
dimensional thin-layer chromatography. Biochim. Biophys. Acta. 1967, 137, 80–89.
1. Johnson, R.A. The enzymatic preparation of [alpha-32P]ATP, [alpha- P]GTP, [32P]cAMP, and
32
1
32
[
P]cGMP, and their use in the assay of adenylate and guanylate cyclases and cyclic nucleotide
phosphodiesterases. Adv. Cyclic Nucleotide Res. 1979, 10, 135–167.
1
2. Shabarova, Z.A.; Fedorova, O.A.; Dolinnaya, N.G.; Gottikh, M.B. Derivatization and template-guided
ligation of oligodeoxyribonucleotides using cyanogen bromide and N-substituted morpholines. Orig
Life Evol Biosph 1997, 27, 555–566.
13. Vaskovsky, V.E.; Kostevsky, E.Y. Modified spray for the detection of phospholipids on thin-layer
chromatograms. J.Lipid Res. 1968, 9, 396.
14. Sidorov, G.V.; Zverkov, Yu.B.; Myasoedov, N.F.; Jasko, M.V.; Skoblov, Yu.S. Synthesis of tritium-labelled
3
-azido-3-deoxythymidine. J. Labelled Comp. Radiopharm. 2003, 46, 669–676.