Synthesis of S6-(2,4-dinitrophenyl)-6-thioguanosine phosphoramidite and its incorporation into oligoribonucleotides

Article abstract of DOI:10.1016/S0960-894X(03)00716-9

The preparation of N²-phenoxylacetyl-S ⁶-(2,4-dinitrophenyl)-6-thioguanosine phosphoramidite and its subsequent incorporation into oligoribonucleotides is described. The identity of the oligonucleotides was confirmed by UV spectrophotometry and nucleoside composition analysis.

Full text of DOI:10.1016/S0960-894X(03)00716-9

Bioorganic & Medicinal Chemistry Letters 13 (2003) 3141–3144
Synthesis of S6-(2,4-Dinitrophenyl)-6-thioguanosine
Phosphoramidite and Its Incorporation into Oligoribonucleotides
Qinguo Zheng,* Yang Wang and Eric Lattmann
School of Life & Health Sciences, Aston University, Aston Triangle, Birmingham B4 7ET, UK
Received 5 May 2003; revised 17 June 2003; accepted 8 July 2003
Abstract—The preparation of N2-phenoxylacetyl-S6-(2,4-dinitrophenyl)-6-thioguanosine phosphoramidite and its subsequent
incorporation into oligoribonucleotides is described. The identity of the oligonucleotides was confirmed by UV spectrophotometry
and nucleoside composition analysis.
# 2003 Elsevier Ltd. All rights reserved.
Oligonucleotides containing a thiol or a thione function
on the base have been very useful tools in areas such as
molecular biology and cancer research (for a latest
review, see ref 1). They are also useful intermediates for
the synthesis of other modified oligonucleotides by post-
synthetic modification.2 Photochemical cross-linking is
such as U small nuclear RNA (U snRNA),5 rRNA6 and
hammerhead ribozymes.7 In addition, it was reported that
these photo-cross-linkable oligonucleotides were very
useful probes for in situ hybridization assays8 and for
investigation of RNA-protein interactions.9 The use of
these oligomers for studying the mechanism of hammer-
head ribozyme cleavage has also been documented.10,11
a
proven method for probing RNA-protein or
RNA-RNA interaction at the atomic level as cross-
linking can only take place between the molecules that
are at the interface of these bio-molecules.3 4-Thiopyr-
imidine and 6-thiopurine nucleosides possess several
desirable properties for such approach. The sulphur
atom is only slightly larger than oxygen, but it otherwise
chemically resembles oxygen, hence the introduction of
these thionucleosides into oligonucleotides should not
appreciably disturb the interaction between these mol-
ecules. Furthermore, these nucleosides are photoactive
at long wavelength UV light (330–360 nm), which is well
away from the usual absorption maxima of proteins
(280 nm) and nucleic acids (260 nm), thus cross-linking
can be carried out at a wavelength which there is no
appreciable detrimental effect on proteins and nucleic
acids. Finally, the cross-linking is site-specific as it can
only occur between the photoactivable thiocarbonyl
function and the target molecules. The photo-cross-
linking approach using oligoribonucleotides bearing
these thiobases has made it possible to probe the
mechanism of hammerhead ribozymes4 and to build the
plausible three-dimensional models of biomolecules
In our effort of developing methods for chemical syn-
thesis of oligonucleotides containing thio-bases, we have
reported the synthetic incorporation of 4-thiothymine,12
6-mercaptopurine13 and 6-thioguanine14 into DNA, and
exploited the use of these oligonucleotides for studying
DNA-protein interactions by photochemical cross-link-
ing approach.15 Recently, we reported a method for the
introduction of structural diversity into the thiocarbo-
nyl group of 6-thioguanine within support-bound, fully
protected oligodeoxyribonucleotides via ‘on-column’
conjugation.2a As part of our continuing effort to
develop methods for thio-modified oligonucleotide
synthesis, we herein describe the synthesis of S6-(2,4-
dinitrophenyl)-6-thioguanosine phosphoramidite and its
subsequent incorporation into oligoribonucleotides.16
In comparison with oligodeoxyribonucleotides, less
work has been reported with oligoribonucleotides con-
taining thiobases, mainly because RNA synthesis is
more difficult than DNA. To date there has been one
report detailing the chemical incorporation of 6-thio-
guanosine into oligoribonucleotides, using a cyanoethyl
group for the protection of the thiocarbonyl function.11
The synthesis of the phosphoramidite involved eight
synthetic procedures starting with guanosine. More
recently, enzymatic ligation was applied to incorporate
*Corresponding author. Tel.: +44-121-359-3611x4167; fax: +44-121-
359-0733; e-mail: q.zheng@aston.ac.uk
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00716-9

3142
Q. Zheng et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3141–3144
6-thioguanosine into oligoribonucleotides using its 30,50-
bisphosphate derivative.17
6-thioguanosine, as removal of N2-phenoxylacetyl in
20-deoxyguanosine can be achieved under mild condi-
tions.21 It was introduced into the exocyclic amino
group of 2, using a one-pot, three-step transient protec-
tion protocol22 yielding N2-phenoxylacetyl-S6-(2,4-di-
nitrophenyl)-6-thioguanosine (3). After tritylation by
the standard method,23 compound 4 was reacted with
tert-butyldimethylsilyl chloride (TBDMS-Cl), produc-
ing compound 5 and its 30-TBDMS isomer and
20,30-diTBDMS 6-thioguanosine derivative, which on
purification by flash chromatography yielded pure
compound 5. Our initial attempt of protecting the
20-OH using the triisopropylsilyl group following a
recommended procedure24 resulted in no success.
Treatment of compound 5 with 2-cyanoethyl-N,N-di-
isopropyl phosphonamidic chloride for 12 h at room
temperature24 produced the desired phosphoramidite 6.
The structures of these compounds were confirmed by
NMR and high resolution mass spectrometry.25
Our synthetic route for the preparation of 6-thioguano-
sine phosphoramidite (6) is shown in Scheme 1. To
simplify the synthetic route commercially available
6-thioguanosine (1) was chosen as the starting material.
To protect the thiocarbonyl group 1 was reacted with
2,4-dinitrofluorobenzene in the presence of triethyl-
amine to give compound 2. Although 1 was not pro-
tected, the reaction was highly chemoselective, most
likely due to the fact that the thio-function is more
nucleophilic than the 20-, 30- and 50-hydroxyl group and
the 2-amino function of 6-thioguanosine. We2a,13,14
have previously used the 2,4-dinitrophenyl group for the
protection of the thiocarbonyl function of deoxy-thio-
nucleosides. Various other protective groups have also
been used for the thio-function in the automated solid-
phase synthesis of both DNA and RNA,1 examples
including the pivaloyloxymethyl group18,19 and the
2-cyanoethyl group.10,11 Adams et al.11 used the latter
for the protection of the thio-function of 6-thioguano-
sine. Removal of the S-cyanoethyl group needed 5 h
DBU treatment in acetonitrile, and thorough washing
resin after deprotection to remove any trace of DBU
was essential to prevent degradation of the thiocarbonyl
group during desilylation.11 We chose the 2,4-dini-
trophenyl (DNP) group since previous studies2a,13,14
indicated that DNP was stable during automated oligo-
nucleotide synthesis and easily removable with mercapto-
ethanol under very mild alkaline conditions. More
significantly, its removal did not cleave the oligomers
from the resin and effect other protecting groups within
the oligomers, which provided a means of carrying out
postsynthetic modification on the thio-function ‘on-
column’.2a
The stability of S6-(2,4-dinitrophenyl)-thioguanosine
phosphoramidite (6) to the reagents used in phosphor-
amidite chemistry had been tested before the monomer
was used for RNA synthesis. Compound 3 was dis-
solved in dichloroacetic acid/dichloroethane (deblock-
ing reagent); and compound 5 in acetic anhydride/
lutidine/THF (capping reagent A), N-methylimidazole
in THF (capping reagent B), and iodine in THF/pyri-
dine/water (oxidation reagent). By monitoring the
changes in these solutions with TLC, it was found that
compound 5 was stable towards the reagents for at least
24 h at room temperature, and compound 3 was stable
towards the deblocking solution for at least 3 h. It was
also observed that, in agreement with our previous
observations,2a the S6-(2,4-dinitrophenyl) group could
be readily removed within 30 min by 10% mercapto-
ethanol in CH3CN containing 1% of Et3N. These
results indicated that monomer 6 would be stable during
the synthesis, and the protecting group (2,4-dinitro-
phenyl) easily removable after synthesis.
Adams et al.11 used the benzoyl group for the protection
of the 2-amino group of 6-thioguanosine, which was
previously used for the protection of the 2-amino func-
tion of 20-deoxy-6-thioguanosine.20 We chose the phe-
noxylacetyl for the protection of the 2-amino group of
The phosphoramidite (6) was incorporated into a 5-mer
(50-GCGSH AU-30) and a 12-mer (50-UAC CAGSH
UGA GCU-30) oligoribonucleotides, together with four
natural nucleoside phosphoramidites (from Glen
Research) protected with base-labile groups on the exo-
cyclic amino functions (phenoxylacetyl for adenosine,
i-propylphenoxyacetyl for guanosine and acetyl for
cytidine). Standard automated procedures were used
except the coupling time for 6 was extended to 30 min.
Based on trityl assay, the overall yields of the 5-mer and
the 12-mer were 92.7 and 71%, respectively (average
stepwise yields 98.5 and 97.2%). Coupling of 6 was
similar to the standard phosphoramidites. Deprotection
of the 2,4-dinitrophenyl was achieved with 10% of
mercaptoethanol in acetonitrile containing 1% of tri-
ethylamine, and the excessive deprotection reagent was
removed and the resin-attached oligomers washed with
acetonitrile by filtration. Cleavage from the support and
removal of the protecting groups on the exocyclic amino
functions and O-cyanoethyl groups in phosphate esters
were accomplished with saturated ammonia solution in
ethanol (freshly prepared) at 25 ꢀC for 8 h. After
Scheme 1. The synthetic route for the thioguanosine phosphoramidite
(6). Reaction conditions were as follows: (a) dinitrofluorobenzene/
Et3N; (b) (CH3)3SiCl/pyridine; (c) phenoxyacetyl chloride/pyridine;
(d) NH3H2O; (e) MMTCl/pyridine; (f) t-butyldimethylsilyl chloride/
AgNO3/pyridine; (g) i-Pr2NP(Cl)OCH2CH2CN/i-Pr2Net.

Q. Zheng et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3141–3144
3143
removal of the ammonia solution, the oligomers were
11
of these oligomers was fully confirmed by UV spectro-
metry, nucleoside composition analysis, and mass spec-
trometry. Use of easily removable protection groups for
the thiol and 2-amino functions minimised the possible
damage to the alkaline-labile thiol group during the
postsynthetic procedures. It is worth pointing out that
as the S6-(2,4-dinitrophenyl) group in 6-thiodeoxy-
guanosine can be removed under mild conditions with-
.
treated with Et3N 3HF/DMSO (1:1) to remove the
20-O-TBDMS groups. Using tetrabutylammonium fluor-
ide (TBAF) in THF, a standard procedure for removing
the silyl group,24 resulted in a product which showed
multiple small peaks in HPLC analysis and loss of the
identity peak with the thiol function (340–350 nm) in
the UV spectrum. The 20-O-TBDMS deprotected oligo-
mers were then purified with Nen-sorb column (Fig. 1a).
Further purification with HPLC resulted in pure oli-
goribonucleotides as a single peak (Fig. 1b). The UV
spectrum of the HPLC purified 5-mer (50-GCGSH
AU-30) showed peaks at 260 and 346 nm (Fig. 2a). The
peak at 346 nm is characteristic of the thiocarbonyl
chromophore (Fig. 2b). Nucleoside composition analy-
sis26 of the purified oligoribonucleotides showed a
nucleoside ratio consistent with that expected, as shown
in Figure 3a for (50-GCGSH AU-30). Peaks on the HPLC
trace were identified by retention time comparison with
the authentic samples. No adenosine was observed due
to the deamination of adenosine by adenosine deami-
nase. This has also been previously observed by oth-
ers.24 The adenosine deaminase is
a
common
contaminant in the enzyme preparations. When mon-
itored at 345 nm only one peak was observed with the
retention time corresponding to an authentic sample of
6-thioguanosine (Fig. 3b). The identity of the oligomers
was also determined by ESI mass spectrometry.27
In conclusion, S6-(2,4-dinitrophenyl)-6-thioguanosine
phosphoramidite was prepared by a reliable and short
synthetic procedure. Its incorporation into the oligo-
nucleotides proceeded with good yield, and the integrity
Figure 2. UV spectrum of the oligoribonucleotide 50-GCGSH AU-30
showing peaks at (a) 260 and 346 nm; and (b) that of 6-thioguanosine
showing a major peak at 343 nm.
Figure 3. Nucleoside composition analysis26 of 50-GCGSH AU-30. The
nucleosides generated from digestion with snake venom phos-
phordiesterase and alkaline phosphatase were separated by HPLC with
monitoring at (a) 260 nm; and at (b) 345 nm. The identity of the peaks
was confirmed by retention time comparison with the authentic samples.
Inosine was resulted from the deamination of adenosine by adenosine
deaminase, which is a contaminant in the enzyme preparations.
Figure 1. The reverse phase HPLC chromatograms of the oligo-
ribonucleotide 50-GCGSH AU-30. (a) After Nen-sorb column purifica-
tion; (b) after further purification with HPLC. C18 columns were used
with 0.1M triethylammonia acetate, pH 6.5, and acetonitrile gradients
of 2–20% over 25 min at a flow rate of 1mL/min, recorded at 260 nm.

3144
Q. Zheng et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3141–3144
out affecting other functional and protecting groups and
removing the oligomers from the resin,2a phosphor-
amidite 6 could be compatible with introducing struc-
tural diversity at multiple sites in RNA via ‘on-column’
conjugation. This is currently being investigated.
20. Waters, T. R.; Connolly, B. A. Nucleosides Nucleotides
1992, 11, 985.
21. Schulhof, J. C.; Molko, D.; Teoule, R. Nucleic Acids Res.
1987, 15, 397.
22. Jones, R. A. In Oligonucleotide Synthesis, Gait, M. J., Ed.;
Oxford University Press: Oxford, New York, Tokyo, 1984;
p 23.
23. Wu, T.; Ogilvie, K. K.; Pon, R. T. Nucleic Acids Res.
1989, 17, 3501.
Acknowledgements
24. Damha, M. J.; Ogilvie, K. K. In Protocols for Oligo-
nucleotides and Analogues, Agrawal, S., Ed.; Human: Totowa,
NJ, 1993; p 81.
We wish to thank The Royal Society for their financial
support. We would also like to thank Dr. Y.-Z. Xu of
Open University for useful advice and use of the DNA/
RNA synthesiser in his laboratory, and Dr. Peter Ash-
ton of Birmingham University for obtaining high-reso-
lution mass spectra.
25. 1H NMR and HRMS data: 2: 1H NMR spectrum
(DMSO-d6): 3.47–3.64 (m, 2H, 50-H, 500-H), 3.89 (m, 1H, 40-
H), 4.10 (m, 1H, 30-H), 4.47 (m, 1H, 20-H), 5.07 (t, 1H, 50-OH),
5.20 (d, 1H, 30-OH), 5.50 (t, 1H, 20-OH), 5.80 (d, 1H, 10-H),
6.77 (s, 2H, 2-NH2, ex), 7.96 (d, 1H, 6-H of 2,4-dinitrophenyl),
8.32 (s, 1H, 8-H), 8.40 (d, 1H, 5-H of 2,4-dinitrophenyl), 8.88
(s, 1H, 3-H of 2,4-dinitrophenyl). HRMS: calcd for
C16H15N7O8S [M+Na]+ 488.0601, found 488.0592. 3: 1H
NMR spectrum (DMSO-d6): 3.34–3.58 (m, 2H, 50-H, 500-H),
3.94 (m, 1H, 40-H), 4.18 (m, 1H, 30-H), 4.59 (m, 1H, 20-H), 4.79
(s, 2H, –CH2– of PhOCH2CO), 5.94 (d, 1H, 10-H), 6.80–6.92
(m, 3H, Ar), 7.25 (m, 2H, Ar), 8.21(d, 1H, 6-H of 2,4-dini-
trophenyl), 8.37 (d, 1H, 5-H of 2,4-dinitrophenyl), 8.72 (s, 1H,
8-H), 8.87 (s, 1H, 3-H of 2,4-dinitrophenyl), 10.86 (s, 1H, 2-
NH, ex). HRMS: calcd for C24H21N7O10S [M+Na]+
References and Notes
1. Chambert, S.; Decout, J.-L. Org. Prep. Proced. Int. 2002,
34, 27.
2. (a) Zheng, Q.; Wang, Y.; Lattmann, E. Tetrahedron 2003,
59, 1925. (b) Wang, Y.; Zheng, Q. J. Pharm. Pharmacol. 2000,
52 (S), 98. (c) Xu, Y.-Z. Bioorg. Med. Chem. Lett. 1998, 8,
1839. (d) Coleman, R. S.; Arthur, J. C.; McCary, J. L. Tetra-
hedron 1997, 53, 11191. (e) Meyer, K. L.; Hanna, M. M. Bio-
conjugate Chem. 1996, 7, 401.
1
622.0968, found 622.0975. 4: H NMR spectrum (DMSO-d6):
3.13–3.16 (m, 2H, 50-H, 500-H), 3.35 (m, 1H, 40-H), 3.69 (s, 3H,
CH3O–), 4.05 (m, 1H, 30-H), 4.29 (m, 1H, 20-H), 4.73 (s, 2H,
–CH2– of PhOCH2CO), 6.01(d, 1H, 1 0-H), 6.75–6.96 (m, 4H,
Ar), 7.15–7.29 (m, 15H, Ar), 8.15 (d, 1H, 6-H of 2,4-dini-
trophenyl), 8.46 (d, 1H, 5-H of 2,4-dinitrophenyl), 8.61 (s, 1H,
8-H), 8.89 (s, 1H, 3-H of 2,4-dinitrophenyl), 10.84 (s, 1H, 2-
NH, ex). HRMS: calcd for C44H37N7O11S [M+Na]+
894.2169, found 894.2150. 5: 1H NMR spectrum (CDCl3):
ꢁ0.11 (s, 3H, CH3–Si), 0.02 (s, 3H, CH3–Si), 0.84 (s, 9H, t-Bu-
Si), 3.45 (m, 2H, 50-H, 500-H), 3.76 (m, 3H, CH3O–), 4.25 (m,
1H, 40-H), 4.43 (m, 1H, 30-H), 4.53 (s, 2H, –CH2– of
PhOCH2CO), 4.90 (m, 1H, 20-H), 6.04 (d, 1H, 10-H), 6.77–6.87
(m, 4H, Ar), 7.18–7.38 (m, 15H, Ar), 8.21 (s, 1H, 2-NH, ex),
8.37 (m, 2H, 5-H and 6-H of 2,4-dinitrophenyl), 8.75 (s, 1H, 8-
H), 8.96 (s, 1H, 3-H of 2,4-dinitrophenyl). HRMS: calcd for
C50H51N7O11SSi [M+Na]+ 1008.3034, found 1008.3063. 6:
31P NMR spectrum (CDCl3): 153.79, 152.82. HRMS: calcd for
C59H69N7O12SSiP [M+Na]+ 1186.1707, found 1186.1723.
26. In a typical nucleoside composition analysis, 1O.D. of
oligoribonucleotide was dissolved in 1.0 mL of a solution
containing 0.2 mM ZnCl2, 16 mM MgCl2, 250 mM Tris–HCl
pH 6.0, 0.2 unit of snake venom phosphodiesterase (Sigma),
and 4 units of calf-intestinal alkaline phosphatase (Sigma),
and the mixture was incubated at 37 ꢀC for 8 h. The digested
sample was then analysed with reverse-phase HPLC (C18 col-
umn), which was eluted with buffer A (0.1M TEAA/aceto-
nitrile, 98:2) and buffer B (0.1M TEAA/acetonitrile, 20:80). A
linear gradient was formed from 0 to 20% of the buffer B over
25 min. Peaks on the HPLC trace were identified by retention
time comparison with the authentic samples. Retention times
3. Hanna, M. M. Methods Enzymol. 1996, 274, 403.
4. Wang, L. X.; Ruffner, D. E. Nucleic Acids Res. 1997, 25,
4355.
5. Yu, Y.; Stertz, J. A. Proc. Natl. Acad. Sci. U.S.A. 1997, 74,
6030.
6. Baravov, P. V.; Gurvich, O. L.; Bogdanov, A. A.; Brima-
combe, R.; Dontsova, O. A. RNA 1998, 4, 658.
7. Laugaa, P.; Woisard, A.; Fourrey, J.-L.; Favre, A. Life Sci.
1995, 318, 307.
8. Huan, B.; Van, Atta R.; Cheng, P.; Wood, M. L.;
Zychlinsky, E.; Albagli, D. Biotechniques 2000, 28, 254.
9. McGregor, A.; Vaman Rao, M.; Duckworth, G.; Stockley,
P. G.; Connolly, B. A. Nucleic Acids Res. 1996, 24, 3173.
10. Adams, C. J.; Murray, J. B.; Arnold, J. R. P.; Stockley,
P. G. Tetrahedron Lett. 1994, 35, 765.
11. Adams, C. J.; Murray, J. B.; Farrow, M. A.; Arnold,
J. R. P.; Stockley, P. G. Tetrahedron Lett. 1995, 36, 5421.
12. Xu, Y.-Z.; Zheng, Q.; Swann, P. F. J. Org. Chem. 1992,
57, 3839.
13. Xu, Y.-Z.; Zheng, Q.; Swann, P. F. Tetrahedron Lett.
1992, 33, 5837.
14. Xu, Y.-Z.; Zheng, Q.; Swann, P. F. Tetrahedron 1992, 48,
1729.
15. Zheng, Q.; Xu, Y.-Z.; Swann, P. F. Nucleosides Nucleo-
tides 1997, 16, 1799.
16. Preliminary results were presented in XV International
Round Table Nucleosides, Nucleotides and Nucleic Acids, 10–
14 September 2002, Leuven, Belgium.
17. Kadokura, M.; Wada, T.; Seio, K.; Sekine, M. J. Org.
Chem. 2000, 65, 5104.
18. Clivio, P.; Fourrey, J.-L.; Faver, A. J. Chem. Soc., Perkin
Trans. 1 1993, 2585.
19. Clivio, P.; Fourrey, J.-L.; Gasche, J. Tetrhadron Lett.
1992, 33, 65.
were: C: 4.50 min, U: 5.63 min, I: 8.26 min, G: 8.85 min, GSH
10.50 min, and A: 13.12 min.
:
27. ESI-MS: calcd for (50-GCGSH AU-30) [MꢁH]ꢁ 1583.23,
found 1583.02; calcd for (50-UAC CAGSH UGA GCU-30)
[MꢁH]ꢁ 3808.51, found 3808.16.