Synthesis of tritium-labelled 3′-azido-3′-deoxythymidine

Article abstract of DOI:10.1002/jlcr.707

New approaches to the synthesis of 3′-azido-3′-deoxythymidine labelled with tritium in the heterocyclic base have been developed. With this aim, enzymatic transribosylation with [³H]thymine using the enzyme preparation from rat liver and a three-step chemical synthesis with use of the tritium labelled precursor were studies. The enzyme preparation did not catalyse the transfer of the 3′-azido-3′-deoxyribosyl fragment to the [³H]thymine residue. 5′-O-Benzoyl-2,3′-anhydrothymidine was taken as a precursor for the tritium labelling by the chemical methods. The resulting [³H]3′-azido-3′-deoxythymidine was obtained with a specific radioactivity of 18.3 Ci/mmol, the tritium is located in the C-6 position of the thymine residue. Copyright

Full text of DOI:10.1002/jlcr.707

JOURNAL OF LABELLED COMPOUNDS AND RADIOPHARMACEUTICALS
J Label Compd Radiopharm 2003; 46: 669–676.
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jlcr.707
Research Article
Synthesis of tritium-labelled 30-azido-30-
deoxythymidine
G.V. Sidorov1,*, Yu.B. Zverkov1, N.F. Myasoedov1, M.V. Jasko2 and
Yu.S. Skoblov2
1 Institute of Molecular Genetics, Russian Academy of Sciences, Kurchatov Sq. 2
123182 Moscow, Russia
2 Institute of Molecular Biology, Russian Academy of Sciences, Vavilov St. 32,
119991 Moscow, Russia
Summary
New approaches to the synthesis of 30-azido-30-deoxythymidine labelled with
tritium in the heterocyclic base have been developed. With this aim, enzymatic
transribosylation with [3H]thymine using the enzyme preparation from rat liver
and a three-step chemical synthesis with use of the tritium labelled precursor
were studies. The enzyme preparation did not catalyse the transfer of the 30-
azido-30-deoxyribosyl fragment to the [3H]thymine residue. 50-O-Benzoyl-2,30-
anhydrothymidine was taken as a precursor for the tritium labelling by the
chemical methods. The resulting [3H]30-azido-30-deoxythymidine was obtained
with a specific radioactivity of 18.3 Ci/mmol, the tritium is located in the C-6
position of the thymine residue. Copyright # 2003 John Wiley & Sons, Ltd.
Key Words: tritium-labelling; solid and liquid state catalytic hydrogen
exchange; 30-azido-30-deoxy-[6-3H]thymidine
Introduction
The mechanisms of action, metabolic pathways, and cellular uptake of
30-azido-30-deoxythymidine have been intensively studied in recent
*Correspondence to: G.V. Sidorov, Institute of Molecular Genetics, Russian Academy of Sciences,
Kurchatov Sq. 2, 123182, Moscow, Russia. E-mail: sidgv@img.ras.ru
Copyright # 2003 John Wiley & Sons, Ltd.
Received 18 November 2002
Revised 24 February 2003
Accepted 26 February 2003

670
G.V. SIDOROV ET AL.
years. These studies require 30-azido-30-deoxythymidine labelled in
different parts of the molecule and at high specific radioactivity
(Amol>10
Ci/mmol).
The synthesis of tritium labelled 30-azido-30-deoxythymidine has been
achieved by several methods. For example, [50-3H]30-azido-30-deoxythy-
midine (14 Ci/mmol) was synthesized by the reduction of the
corresponding 30-azido-50-oxo-30,50-dideoxythymidine with tritium la-
belled sodium borohydride.1 For the preparation of [methyl-3H]30-
azido-30-deoxythymidine (35.9 mCi/mmol), a four-step procedure has
been used.2 The starting [methyl-3H]thymidine was protected at position
50 by a trityl group, transformed to the corresponding [methyl-3H]50-O-
trityl-2,30-anhydrothymidine, treated with sodium azide, and depro-
tected.
Tritium labelled 30-azido-30-deoxythymidine can also be synthesized
using the reaction of solid-state catalytic hydrogenation (SSCH).3,4 It
was shown,5 that catalytic reactions of 30-azido-30-deoxythymidine with
gaseous tritium in solution result mainly in the formation of the
corresponding 30-aminonucleoside. This process also took place in
SSCH reactions to yield 20–70% of 30-amino-30-deoxythymidine. It has
been shown that the specific radioactivity of the 30-azido-30-deoxythy-
midine obtained in the reaction of solid state catalytic isotope exchange
with gaseous tritium and in the reaction of isotope exchange with
tritiated water did not exceed 0.5 Ci/mmol. About 90% of the tritium
was localized in the thymine part of the molecule.5
The present study is concerned with the development of new
approaches to the effective synthesis of tritium labelled 30-azido-30-
deoxythymidine, in which the tritium is localized in the heterocyclic
base. Two routes were chosen for the study:
1. Enzymatic synthesis of 30-azido-30-deoxythymidine by transdeox-
yribosylation reaction of thymine using an enzyme preparation isolated
from rat liver6:
½3Hꢀthymine þ 30 ꢁ azido ꢁ 30 ꢁ deoxythymidine $
½3Hꢀ30 ꢁ azido ꢁ 30 ꢁ deoxythymidine þ thymine
2. Chemical synthesis of [3H]30-azido-30-deoxythymidine similar to the
method described in Zaitseva et al.7 50-O-Benzoyl-2,30-anhydrothymi-
dine (I) was chosen as a precursor for the tritium labelling (Figure 1).
We could not find any published data on the enzyme transfer of the
30-azido-30-deoxyribosyl residue from one base to another. Therefore,
Copyright # 2003 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2003; 46: 669–676

THE SYNTHESIS OF TRITIUM
671
Figure 1. The synthesis of tritium labelled 30-azido-30-deoxythymidine. (I) – 50-O-
benzoyl-2,30-anhydrothymidine, (II) – [6-3H]50-O-benzoyl-2,30-anhydrothymidine,
(III) – [6-3H]50-O-benzoyl-30-azido-30-deoxythymidine, (IV) – [6-3H]30-azido-30-
deoxythymidine
we tried to find out whether this reaction can take place and, if the result
is positive, optimize the conditions.
The chemical synthesis performed by us was based on the isotopic
exchange of 50-O-benzoyl-2,30-anhydrothymidine (I) with gaseous
tritium both in solution and in the solid phase. Further studies were
concerned with the optimization of transformation conditions for
tritium labelled 50-O-benzoyl-20,30-anhydrothymidine (II) to tritium
labelled 30-azido-30-deoxythymidine taking into consideration the
microquantities of labelled compounds involved.
Results and discussion
The use of transribosylation enzyme for the preparation of labelled
compounds has been developed into a routine procedure.6 However, in
most cases it has involved the synthesis of natural nucleosides, but for
various nucleoside analogues the procedure method needs to be
optimized.
The results of the transdeoxyribosylation reaction of thymine with the
enzymatic preparation from rat liver are shown in Table 1. The accuracy
Copyright # 2003 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2003; 46: 669–676

672
G.V. SIDOROV ET AL.
Table 1 Some characteristics of the transdeoxyribosilation reaction of [3H]thy-
mine (48 Ci/mmol)
Donor of ribose residue
Molar ratio Amol of target
thymine/donor compound (Ci/mmol)
Conversion
(%)
30-Azido-30-deoxythymidine
5-Hydroxymethyl-30-azido-20,
30-dideoxyuridine
1:4
1:6
0.0014
0.0007
0.012
0.005
20-Deoxyuridine
1:4
1:4
48
7.8
81.0
65.0
Thymidine
Table 2 Characteristics of the SSCH reaction of 50-O-benzoyl-2,30-anhydrothy-
midine using tritium–protium (1:250) mixture
Solvent
Catalyst
T(8C) Duration 50-O-benzoyl-2,30-anhydrothymidine
Ci/mol
Yield (%)
160
5%PdO/Al2O3 170 30 min
180
20
21
22
47
31
15
No
5%PdO/Al2O3 25
5%Pd/BaSO4 25
20 Hrs
95
25
33
37
NH4OH 0.1 N
5%Pd/CaCO3 25
10%Pd/CaCO3 25
75
64
28
30
of the method allowed us to detect less than 0.001% of the maximal
enzymatic activity. The data show that the enzymatic preparation is
sensitive to the structure of the sugar residue, and in the case of 30-azido-
30-deoxyribosyl its transfer to another base did not proceed.
The first stage of the chemical synthesis was the preparation of tritium
labelled 50-O-benzoyl-2,30-anhydrothymidine (II, Figure 1) by the SSCH
reaction in the presence of gaseous tritium (Table 2). Evidently, an
increase in the reaction temperature caused a considerable decrease in
the product yield, whereas Amol remained practically constant and did
not exceed 21 Ci/mol. The Amol of the preparation (II) obtained by the
reaction of catalytic isotope exchange with gaseous tritium in solution
was considerably higher and reached a maximum of 95 Ci/mol (or about
20 Ci/mmol for 95% tritium, Table 2).
The next step was the optimization of the synthesis of [3H]50-O-
benzoyl-30-azido-30-deoxythymidine (III). The kinetic curves for the
azidation of [3H]50-O-benzoyl-2,30-anhydrothymidine (II) with sodium
Copyright # 2003 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2003; 46: 669–676

THE SYNTHESIS OF TRITIUM
673
azide are presented in Figures 2 and 3. Equilibrium had been reached
after 15 h and a maximum yield of 36% was achieved when the molar
ratio of NaN3/compound (II) was 16:1 (Figure 3).
In the last stage [3H]50-O-benzoyl-30-azido-30-deoxythymidine
(III) was treated with 5 M ammonia solution to remove the benzoyl
group. After heating at 608C for 1 h, the reaction was complete to give
(IV); the specific radioactivity of the product had decreased by 24%
which may be because some of the tritium was located in the benzoyl
group.
On the basis of the above described model experiments, we performed
a preparative synthesis of tritium labelled 30-azido-30-deoxythymidine
(IV). Tritium was introduced by the reaction of catalytic isotopic
exchange with gaseous tritium in solution. The resulting (IV) was
prepared with a specific radioactivity of 18.3 Ci/mmol and radio-
chemical purity of more than 99%. The chemical yield was 13.5%
starting from 50-O-benzoyl-2,30-anhydrothymidine (I). The product was
stored in ethanol-water (1:1 v/v) solution at ꢁ208C, at a radioactive
concentration of 1 mCi/ml.
Comparison of the 1H and 3H NMR spectra for compound (IV) show
that the singlet at d 7.63 in the latter spectrum is associated with the H-6
position of the thymine base.
To conclude, unlike the enzymatic reaction of transdeoxyribosylation
of 30-azido-30-deoxythymidine, the chemical approach allowed the
Figure 2. Yield of 50-O-benzoyl-30-azido-30-deoxythymidine in a reaction mixture
(tenfold surplus NaN3) as a function of time
Copyright # 2003 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2003; 46: 669–676

674
G.V. SIDOROV ET AL.
Figure 3. Yield of 50-O-benzoyl-30-azido-30-deoxythymidine as a function of
NaN3/compound (II) ratio (reaction time 3 h)
synthesis of [6-3H]30-azido-30-deoxythymidine at a high specific radio-
activity. The radiochemical purity of the target compound exceeded
99% (by TLC analysis). The radioactive yield was 6.7%.
Experimental
The catalysts used were 5% PdO/Al2O3 (RSS ‘Applied Chemistry’,
St. Petersburg), 5% Pd/BaSO4 (Aldrich) and 10% Pd/CaCO3 (Fluka).
50-O-Benzoyl-2,30-anhydrothymidine (I) was prepared according to
Kimuva et al.8 UV absorption spectra were obtained using a Specord
M-40 spectrophotometer. The NMR spectra were obtained using D2O
as solvent and a AMX III-400 (Bruker) spectrometer with the 400 MHz
operating frequency for 1H and 426 MHz for 3H NMR. Sample
radioactivity was measured on a RLC-20 scintillation counter using
EcoLM TM scintillation liquid (ICN). Reactions involving gaseous
tritium were performed on an installation suitable for manipulations
with gaseous tritium.9 TLC was carried out on Silufol plates (Czech
Republic) in chloroform-ethanol (95:5 v/v) solvent systems. The Rf
values for reference 50-O-benzoyl-30-azido-30-deoxythymidine (I) and 30-
azido-30-deoxythymidine were 0.35 and 0.15, respectively. For the
analysis of 30-azido-30-deoxythymidine, plates with PEI-cellulose
Copyright # 2003 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2003; 46: 669–676

THE SYNTHESIS OF TRITIUM
675
(Merck), water solvent (Rf 0.63) were also used. HPLC was conducted
using a 13-mm (10 ꢂ 250 mm) Nucleosil C18 (Macherey-Nagel) column.
The mobile phase for isolation of 50-O-benzoyl-2,30-anhydrothymidine
was 60% methanol in water, elution rate 1.5 ml/min, retention time
13.1 min; for isolation of 30-azido-30-deoxythymidine: 15% acetonitrile
in 0.1 mol/l TEAB, pH 7.3, elution rate 1.5 ml/min, retention time
19.5 min.
Enzymatic transdeoxyribosylation reactions were carried out with
trans-N-glycosidase (EC 2.4.2.6) isolated from rat liver as previously
described.6 The reaction mixture contained 1 mmol of thymine, a donor
of a ribosyl residue (see Table 1) 4–6 mmol; MgCl2 14 mmol; potassium
phosphate buffer pH 7.0 – 400 mmol; enzyme 800 ml (with the activity of
37.5 nmol minꢁ1 mlꢁ1); total volume was 4.0 ml. The reaction mixture
was incubated at 378C for 2 h. After the incubation, the protein was
denatured by heating at 1008C for 5 min and then separated by
centrifugation.
Synthesis of tritium labelled 50-O-benzoyl-2,30-anhydrothymidine:
3.3 mg of 50-O-benzoyl-2,30-anhydrothymidine was placed in a reaction
glass vial, and 400 ml of 0.1 N ammonia in 50% water ethanol and 50 mg
of 5% PdO/Al2O3 catalyst were added. The vial was then frozen in
liquid nitrogen, evacuated, and filled with gaseous tritium up to pressure
of 400 mm Hg. Upon defrosting, the reaction mixture was stirred at
room temperature for 21 h. After the reaction was complete, tritium was
removed from the vial and the catalyst separated by centrifugation.
Labile tritium was removed by evaporation with 10 ml of 50% ethanol.
Isolation of the product achieved by HPCL. As a result, 110 mCi of
product was obtained with a specific radioactivity of 24.6 Ci/mmol and
45% yield.
Synthesis of tritium labelled 50-O-benzoyl-30-azido-30-deoxythymidine:
1.5 mg of [3H]50-O-benzoyl-2,30-anhydrothymidine was dissolved in
0.9 ml of absolute dimethylformamide, then 5 mg of NaN3 was added
and the solution inserted in a glass vial that was then sealed. The vial
was thermostated at 1008C for 15 h.
To extract [3H]5’-O-benzoyl-30-azido-30-deoxythymidine, the reaction
mixture was placed in a tube, to which a 0.1 M solution of NaHCO3
(pH 9.9 to precipitate the excess of NaN3) and 2 ml of distilled
chloroform were added. The solution was centrifuged, and the product
allowed to precipitate over 15 min and the choloroform layer separated
with a pipette. This operation was repeated two times to achieve the
maximum extraction. The chloroform extractions were combined and
Copyright # 2003 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2003; 46: 669–676

676
G.V. SIDOROV ET AL.
evaporated in vacuum with ethanol for complete removal of traces of
the solvent. The radioactivity yield of the generated compound,
estimated by TLC, was 36%.
Removal of benzoyl protection and isolation of tritium labelled 30-azido-
30-deoxythymidine: The dry remainder after extraction was dissolved in
2 ml of 5 N ammonia solution. The solution was then placed in a glass
vial which was sealed and thermostated at 608C for 1 h. After
completion of the reaction the contents of the vial were evaporated in
vacuum, then 5 ml of ethanol was added to the remainder and it was
evaporated again. HPLC was used for the isolation of [3H]30-azido-30-
deoxythymidine from the reaction mixture. The result was 25 mCi of
tritium labelled 30-azido-30-deoxythymidine with a specific radioactivity
of 18.3 Ci/mmol and 13.5% chemical yield. The radiochemical purity of
the product, estimated by thin layer chromatography in the above
mentioned systems, was over 99%. [6-3H]30-azido-30-deoxythymidine
was stored at ꢁ208C as a solution of 1 mCi/ml in 1:1 (v/v) water–
ethanol. 1H-NMR (D2O, d, ppm; J, Hz): 7.65s (0.8H, H-6), 6.22t (1H, J
6.7, H-10), 4.39m (1H, H-30), 4.04m (1H, H-40), 3.85m (2H, H-50), 2.55m
3
(2H, H-20), 1.92s (3H, 5-CH3). H-NMR (D2O, d, ppm): 7.65s (3H-6).
References
1. Hill JA, Freeman GA. J Label Compd Radiopharm 1988; 25: 277–280.
2. Aggarwal SK, Shalinsky DR, Argawal KC. J Label Compd Radiopoharm
1988; 25: 1055–1060.
3. Sidorov GV, Myasoedov NF. Bioorganicheskaya Khimia 1993; 19:
1220–1225.
4. Sidorov GV, Myasoedov NF, Yasko MV, J Label Compd Radiopoharm
1994; 34: 339–351.
5. Sidorov GV, Zverkov YuB, Myasoedov NF. Radiokhimia 2003; 3, in press.
6. Ivankova EK, Sidorov GV, Myasoedov NF. Radiokhimia 1981; 23:
296–300.
7. Zaitseva VE, Dyatkina NB, Krayevsky AA, Skaptsova NV, Turina OV,
Gnuchev NV, Gottikh BP, Azhayev AV. Bioorganicheskaya Khimia 1984;
10: 670–680.
8. Kimura J, Fujisawa Y, Yoshizawa T, Fukuda K, Mitsunobu O. Bull.
Chem. Soc. Jpn 1979; 52: 1191–1196
9. Mikhailov KS, Lavrov OV, Myasoedov NF. Organicheskie soedinenia
mechennye padioacrivnymi izotopami [Organic compounds labelled by
radioactive isotopes]. 1977, 253–258.
Copyright # 2003 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2003; 46: 669–676