Synthesis and incorporation into DNA fragments of the artificial nucleobase, 2-amino-8-oxopurine

Article abstract of DOI:10.1016/j.bmcl.2004.12.021

The nucleoside 2-amino-9-(2-deoxy-β-d-ribofuranosyl)-7,8-dihydro-8- oxo-purine (dJ) was obtained in eight steps from 2′-deoxyguanosine. The appropriate protected phosphoramidite was synthesized and incorporated into DNA oligonucleotides. The thermal stability of heteroduplexes containing 2-amino-8-oxopurine (J) was investigated by UV-thermal denaturation experiments. The results obtained can be interpreted by the base pairing schemes involving the two edges of dJ depending on the anti and syn orientations.

Full text of DOI:10.1016/j.bmcl.2004.12.021

Bioorganic & Medicinal Chemistry Letters 15 (2005) 1069–1073
Synthesis and incorporation into DNA fragments of the
artificial nucleobase, 2-amino-8-oxopurine
Claudio Cadena-Amaro,a Muriel Delepierreb and Sylvie Pocheta,*
a
´
Unite de Chimie Organique, URA CNRS 2128, Institut Pasteur, 28, rue du Docteur Roux, 75724 Paris cedex 15, France
b
´ ´
Unite de Resonance Magnetique Nucleaire des Biomolecules, URA CNRS 2185, Institut Pasteur, 28,
´
rue du Docteur Roux, 75724 Paris cedex 15, France
´
´
Received 12 November 2004; revised 7 December 2004; accepted 9 December 2004
Available online 30 December 2004
Abstract—The nucleoside 2-amino-9-(2-deoxy-b-D-ribofuranosyl)-7,8-dihydro-8-oxo-purine (dJ) was obtained in eight steps from
20-deoxyguanosine. The appropriate protected phosphoramidite was synthesized and incorporated into DNA oligonucleotides.
The thermal stability of heteroduplexes containing 2-amino-8-oxopurine (J) was investigated by UV-thermal denaturation experi-
ments. The results obtained can be interpreted by the base pairing schemes involving the two edges of dJ depending on the anti
and syn orientations.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Faithful transmission of hereditary messages by nucleic
acids depends upon the pairing between a large bicyclic
purine and a smaller monocyclic pyrimidine both in the
anti conformation according to a scheme known as the
Watson–Crick geometry.1 However there is no reason
to assume that the requirements for duplex stability
and replication must limit the genetic alphabet to only
two base pairs. Efforts to expand the genetic alphabet
to a third base pair have resulted during the last decades
in the development of several candidates. Works with
modified nucleobases have explored not only hydro-
gen-bonded pairs2–5 but also nonpolar pairs6,7 and
metal-mediated pairs.8 In addition to heteropairs, sev-
eral promising self-pairs have been identified, among
them, purine–purine pairs have been designed.9,10 More
recently, new base pairing motifs in which building
blocks are size-expanded into tricyclic bases have been
reported.11,12
tain purine pairs are indeed genetically encoded, such as
the conserved A(syn):G(anti) pairs present in helices of
ribosomal and transfer RNA secondary structures.14
The formation of other purine pairs is actively avoided
by organisms, the most documented case being the re-
pair process for correcting 8-oxodG(syn):dA(anti) pairs
formed during erroneous replication involving 8-ox-
odGMP.15 The mutagenic properties of 8-oxodG16 re-
flect the stability of the lesion when paired in the syn
conformation with dA(anti). Starting from the 8-oxo-
guanine(syn):adenine(anti) mispairing scheme (Fig. 1a),
we designed an artificial nucleobase, 2-amino-8-oxopu-
rine (noted J), potentially able to form base pairs in
the syn or anti conformation.
As the 8-keto form predominates at physiological pH,
dJ possess a hydrogen bond donor at the 7 position
and a hydrogen bond acceptor at the 8 position, coupled
with the hydrogen bond pattern of 2-amino-purine on
the Watson–Crick side. Thus, J in the anti conformation
presents its 2-aminopurine edge for pairing and can
form two hydrogen bonds with base T according to
Watson–Crick geometry (Fig. 1b). Other pairings with
C and A involve the formation of less favored proton-
ated Watson–Crick or neutral Wobble pyr–pur and
large Wobble pur–pur base pairs. Such hydrogen bonds
between 2-aminopurine and A or C have been supported
by NMR,17,18 however only AP:C base pair is involved
in its mutagenic profile.
Purine nucleosides are inherently ambivalent, capable of
flipping into the syn glycosidic conformation in addition
to the anti conformation and thus of displaying either of
two edges for hydrogen bonding with other bases.13 Cer-
Keywords: Nucleobase; Nucleoside; Conformation; Base pairing;
DNA.
*
Corresponding author. Tel.: +33 01 40 61 33 28; fax: +33 01 45 68 84
04; e-mail: spochet@pasteur.fr
0960-894X/$ - see front matter ꢀ 2004Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.12.021

1070
C. Cadena-Amaro et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1069–1073
Figure 1. Proposed structures of hydrogen-bonded base pairs involving 2-amino-8-oxopurine (J) compared to those of 8-oxoguanine (oxoG).
The absence of any donor or acceptor character at the 6
position changes the pairing of J in the syn conforma-
tion compared to oxoG(syn). However, as the result of
the 7-NH donor and 8-oxo acceptor characters, J(syn)
can pair with itself in the anti configuration with a shape
similar to the oxoG(syn):A(anti) pair (Fig. 1b). Hydro-
gen bonded pairs with T and G result in a displacement
of the purine J in the syn conformation compared to
J(anti):J(syn) pair (Fig. 1c).
The orientation of the aromatic base with respect to the
sugar moiety is determined by the glycosidic torsion
angle v defined as the O-40-C-10-N-9-C-4. A number of
methods have been developed to determine the v values
using NMR data. Although quantitative evaluation of
interproton distances by mean of NOE measurement is
widely used most often, in absence of proton at C8 posi-
tion the values of v are determined from the measure-
ment of three-bond carbon–proton scalar couplings
3JCH across the glycosidic bond and DFT analysis using
3
Nucleoside 1 was synthesized in eight steps from 20-
deoxyguanosine (dG) according to a previous report
with slight modifications.19 We have selected the benzoyl
group for the protection of hydroxyls and exocyclic
amine to improve purification and global yield. The
main steps are outlined in Scheme 1 and nucleoside 1
was isolated in 6% global yield. Nucleoside 1 was fully
characterized by elemental analysis, NMR and mass
spectrometries.20 The UV absorbance spectrum of 1 in
water (pH 7.2) exhibited a maximum at 240 nm (e
8970) and at 309 nm (e 7100), characteristic of 2-amino-
purine derivatives (fluorescence).
Karplus equations.22 The JC4–H1 and 3JC8–H1 of nucleo-
side 1 were determined at 30 mM in D2O (pH 7.2) by
gradient selected 3J-HMBC NMR experiments recorded
at 600 MHz. The experimental values of three-bond car-
0
0
3
bon–proton scalar couplings JCH of 1 were 4.29 Hz for
0
3
3
JC4–H1 and 3.77 Hz for JC8–H1 . Comparison of
0
3
3
these values with the JC4–H1 and JC8–H1 values
0
0
reported for purines22 and taking into account that
both JC4–H1 > 3JC8–H1 and DJ (3JC4–H1 ꢀ JC8–H1 ) is
small23 allowed us to conclude that the nucleoside 1
adopts in solution a preferential syn conformation. As
described for adenine and guanine nucleobases, the
introduction of a bulky group on C8 of 2-aminopurine
moiety results in the increased of the syn population
and should promote the pairing mode of 2-amino-8-
oxopurine (J) in syn conformation.
3
3
0
0
0
0
The conformational analysis of the furanose puckering
of 1 was determined by means of the sums of proton–
proton coupling constants.21 The fraction S-type con-
1
former (pS) of 1 determined by H NMR at 400 MHz
was 65%. This preference indicates a rapid dynamic
equilibrium between N and S states in solution.
For the preparation of the phosphoramidite derivative
5, both the 2-amino and 7-imino functions were

C. Cadena-Amaro et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1069–1073
1071
Scheme 1. Reagents and conditions: (a) 3 N sodium methylate, reflux, 72 h, 77%; (b) benzoyl chloride, pyridine, 1 h, 62%; (c) 2,4,6-tri-iso-
propylbenzenesulfonyl chloride, DIPEA, CH2Cl2, 1.5 h, 64%; (d) anhydrous hydrazine, THF, 6 h; (e) Ag2O, THF-5% H2O, reflux, 2 h, 38%; (f)
thiophenol, triethylamine, DMF, 60 ꢁC, 2.5 h, 80%; (g) 2N NaOH, pyridine/methanol; then 33% aqueous ammonia, 55 ꢁC, 48 h, 63%.
protected with suitable groups. We observed that the
complete removal of isobutyryl and benzoyl groups re-
quired an extensive heating in aqueous ammonia and
was not compatible with the solid-support DNA synthe-
sis. To circumvent this problem, the highly alkali-labile
amino protecting group, the phenoxyacetyl group, was
chosen.24 Our synthetic route for the preparation of
the phosphoramidite 5 is shown in Scheme 2. Reaction
of 1 with an excess of phenoxyacetyl chloride in pyridine
followed by treatment with 0.5 N sodium methylate
Oligonucleotides (17-mers and 22-mers) containing the
modified base J were synthesized on an automated
DNA synthesizer according to standard b-cyanoethyl
phosphoramidite chemistry. The coupling efficiency of
5 was not different from those of the normal phosphor-
amidites as estimated by released DMT. The synthe-
sized oligomers were purified by reverse phase HPLC
at two stages (DMT on, then after detritylation) and
further characterized by HRMS (MALDI-TOF)27 and
nucleoside composition analysis after enzymatic
digestion.
afforded N2-phenoxyacetyl nucleoside
6
in 62%
yield. 50-O-Dimethoxytritylation of 6 gave 7 in 76%
yield. Since the presence of the lactam may interfere with
the solid-phase DNA synthesis, we chose for N-7 protec-
tion the diphenylcarbamoyl group previously used for
the synthesis of the phosphoramidite derivative of 8-ox-
odG.24,25 Treatment of 7 with diphenylcarbamoyl chlo-
ride in pyridine in the presence of N,N-di-iso-
propylethylamine afforded 8 in 61% yield. Finally, 8 re-
acted with 2-cyanoethyl-N,N-di-iso-propylchloro-phos-
phoramidite in CH2Cl2 in the presence of N,N-di-iso-
propylethylamine to afford the phosphoramidite 526 in
77% yield.
The stability of oligonucleotides containing J has been
studied by UV thermal-denaturation experiments using
a system composed of two complementary heptadeca-
mers. Introducing the artificial base in the center (ninth
position) of each strand allowed to measure any pairing
preferences of J and test the effects of local nearest-
neighbour bases (purines or pyrimidines). Such a design
was previously used in our laboratory for studying modi-
fied nucleobases.28,29 The UV thermal denaturating
experiments were carried out at 1 lM duplex concentra-
tion. In all cases, sharp melting profiles were observed
Scheme 2. Reagents and conditions: (a) phenoxyacetyl chloride, pyridine, 30 min; (b) 0.5 N sodium methylate, 30 min, 62%; (c) DMTCl chloride,
pyridine, 2 h, 76%; (d) diphenylcarbamoyl chloride, DIPEA, pyridine, 2 h, 61%; (e) 2-cyanoethyl-N,N-di-iso-propylchlorophosphoramidite, DIPEA,
CH2Cl2, 30 min, 77%.

1072
C. Cadena-Amaro et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1069–1073
indicating co-operativity and reversibility of the dena-
turation/association process (data not shown).
to syn. Insertion of J results in a decreased thermal sta-
bility of all the duplexes compared to A:T and C:G base
pairs (from 5 to 10 ꢁC). However the best Tm terms are
found when J is placed opposite T or itself, as compared
to the other combinations. Our data are in agreement
with the destabilization observed for oxoG(anti):C base
pairs compared to C:G base pairs due to the C8-oxygen.
The results presented here can be interpreted by the base
pairing schemes postulated initially involving the two
edges of J depending on the anti and syn orientations.
Results from the melting temperatures data (Tm) (Table
1) indicate that the central replacement of a canonical
base by J decreased the duplex thermal stability as com-
pared to natural C:G (entry 1) and A:T (entry 2) base
pairs. For example, C:J (entry 3) and G:J (entry 7) pairs
lowered the Tm by 10–11 ꢁC as compared to C:G. A sim-
ilar destabilization is observed when J is placed opposite
A (A:J and J:A pairs, entries 4and 6). No significant
context effect is observed. The most favorable Tm terms
are found when J is placed opposite T (entries 8 and 10)
and opposite itself (entry 11), however the correspond-
ing duplexes are less stable than A:T duplex by 5 ꢁC.
Previous works on oxoG illustrate the difficulty to ex-
pect correlations between the thermodynamics of dis-
ruptions of DNA duplexes with interior lesions and
the biological consequences (preferentially insertion of
nucleotides opposite the lesions). Recognition of the
new base J by DNA polymerases is under investigation
and will be the subject of a separate paper.
We presumed that J(anti) pair with T according to a
Watson–Crick mode similar to 2AP:T base pair. As it
was reported that the thermodynamic stability of
2AP:T in short heteroduplex is only slightly changed
compared to A:T,30 the lowered thermal stability of
J:T and T:J base pairs compared to A:T base pair can
be interpreted by the impact of the 8-oxo function.
These results are consistent with the finding of several
groups studying the influence of oxidative lesion on du-
plex stability. It is known that 8-oxoG in DNA duplex
adopts both anti and syn conformations depending on
the pairing partner (C or A).31–33 Melting studies have
shown that the central 8-oxoG residue reduces duplex
thermal stabilities relative to G duplex when C is the
cross-strand partner. When A is placed opposite 8-oxoG
rather than G, an increase in stability is observed. Thus,
despite having one less hydrogen bond, 8-oxoG(syn):A
base pair is only slightly less stable than 8-oxoG(anti):C
pair.34
Acknowledgments
CONACYT (Mexico) is gratefully acknowledged for a
doctoral scholarship to C.R.C-A. We would like to
thank Professor J. S. Sun (M.N.H.N., Paris) for provid-
ing facility for denaturation experiments.
References and notes
1. Watson, J. D.; Crick, F. H. C. Nature 1953, 171, 737.
2. Piccirilli, J. A.; Krauch, T.; Moroney, S. E.; Benner, S. A.
Nature 1990, 343, 33.
3. Lutz, M. J.; Held, H. A.; Hottiger, M.; Hubscher, U.;
Benner, S. A. Nucleic Acids Res. 1996, 24, 1308.
4. Ohtsuki, T.; Kimoto, M.; Ishikawa, M.; Mitsui, T.; Hirao,
I.; Yokoyama, S. Proc. Natl. Acad. Sci. U.S.A. 2001, 98,
4922.
5. Loakes, D. Nucleic Acids Res. 2001, 29, 2437.
6. Ogawa, A. K.; Wu, Y.; McMinn, D. L.; Liu, J.; Schultz, P.
G.; Romesberg, F. E. J. Am. Chem. Soc. 2000, 122, 3274.
7. Kool, E. T.; Morales, J. C.; Guckian, K. M. Angew.
Chem., Int. Ed. 2000, 39, 991.
Similarly, the relative stability of J:T and J:J duplexes
can be explained by the formation of two hydrogen
bonds base pairs, J(anti):T base pair (according to Wat-
son–Crick mode as 2-AP:T) and J(syn):J(anti) base pair
as illustrated on Figure 1b.
8. Wagenknecht, H. A. Angew. Chem., Int. Ed. 2003, 42,
3204.
9. Henry, A. A.; Romesberg, F. E. Curr. Opin. Chem. Biol.
2003, 7, 727.
10. Henry, A. A.; Olsen, A. G.; Matsuda, S.; Yu, C.;
Geierstanger, B. H.; Romesberg, F. E. J. Am. Chem.
Soc. 2004, 126, 6923.
11. Minakawa, N.; Kojima, N.; Hikishima, S.; Ssaki, T.;
Kiyosue, A.; Atsumi, N.; Ueno, Y.; Matsuda, A. J. Am.
Chem. Soc. 2003, 125, 9970.
12. Liu, H. B.; Gao, J. M.; Maynard, L.; Saito, Y. D.; Kool,
E. T. J. Am. Chem. Soc. 2004, 126, 1102.
13. Donohue, J.; Trueblood, K. N. J. Mol. Biol. 1960, 2, 363.
14. Traub, W.; Sussman, J. L. Nucleic Acids Res. 1982, 19,
2701.
In summary, we described a synthetic scheme for the
preparation of nucleoside dJ and its incorporation into
oligonucleotides using the appropriate phosphoramidite
derivative. Conformational analysis of the free nucleo-
side concludes that the presence of the 8-oxygen
switches the glycosidic orientation preference from anti
Table 1. Melting temperatures (Tm)a of heteroduplexes containing
canonical bases and 2-amino-8-oxopurine (noted J)
5’-ACTTGGCCXCCATTTTG-3’
3’-TGAACCGGYGGTAAAAC-5’
Entry
X:Y
Tm (ꢁC)
Entry
X:Y
Tm (ꢁC)
15. Shibutani, S.; Takeshita, M.; Grollman, A. P. Nature
1991, 349.
1
C:G
C:J
J:C
G:J
J:G
J:J
56
45
46
48
46
50
2
A:T
A:J
J:A
T:J
J:T
55
47
46
50
50
3
4
16. Kamiya, H. Nucleic Acids Res. 2003, 31, 517.
17. Fagan, P. A.; Fabrega, C.; Eritja, R.; Goodman, M. F.;
Wemmer, D. E. Biochemistry 1996, 35, 4026.
18. Sowers, L. C.; Boulard, Y.; Fazakerley, G. V. Biochem-
istry 2000, 39, 7613.
5
6
7
8
10
9
11
a Conditions: 10 lM sodium cacodylate, 0.1 M NaCl at pH 7.2, 1 lM
of each strand.
19. Pochet, S.; Marliere, P. C. R. Acad. Sci. Life Sci. 1996,
319, 1.
`

C. Cadena-Amaro et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1069–1073
1073
20. Compound (1): 1H NMR (DMSO-d6): d 2.00 (m, 1H,
H20), 3.05 (m, 1H, H200), 3.46 (m, 1H, H50), 3.60 (m, 1H,
H500), 3.78 (m, 1H, H40), 4.38 (m, 1H, H30), 4.78 (br s, 1H,
OH50), 5.16 (br s, 1H, OH30), 6.13 (br s, 3H, H10 and
NH2), 7.78 (s, 1H, H6), 10.82 (br s, 1H, N7H). 13C NMR
(DMSO-d6): d 35.73 (C20), 62.69 (C50), 71.65 (C30), 81.21
(C10), 87.77 (C40), 113.58 (C5), 135.43 (C6), 151.02 (C4),
152.71 (C8), 159.09 (C2). HRMS (ESI-TOF): m/z calcu-
lated for C10H13N5O4+Na: 290.0865, found: 290.0872.
H6–R, 1H, H30), 4.75–4.88 (m, 2H, CH2–PhOAc), 6.02–
6.10 (m, 1H, H10), 6.70–6.78 (m, 4H, H arom. DMT),
6.91–6.98 (m, H6–S), 10.61 and 10.62 (2s, 1H, NH–R,
NH–S). 31P NMR (DMSO-d6): d 148.56 and 149.03.
HRMS (ESI-TOF): m/z calculated for C61H63N8O10+Na
1121.4302; found: 1121.4344.
27. HRMS(MALDI-TOF) of synthesized oligonucleotides:
50-ACTTGGCCJCCATTTTG-30 m/z calculated for
C
165H211N57O104P16: 5148.86981, found: 5148.84666. 50-
Anal. Calcd for C10H13N5O4+1/4H O: C, 44.20; H, 4.91;
2
CAAAATGGJGGCCAAGT-30 m/z calculated for
167H207N73O96P16: 5264.92838, found: 5264.90345. 50-
GCATJGTCATAGCTGTTTCCTG-30 m/z calculated for
N, 25.77; found: C, 44.32; H, 5.04; N, 25.37. UV (H2O/
pH 7.6) kmax 240 nm (e 8973) and kmax 309 nm (e 7104).
21. Rinkel, L. J.; Altona, C. J. Biomol. Struct. Dyn. 1987, 4,
621.
22. Munzarova, M. L.; Sklenar, V. J. Am. Chem. Soc. 2003,
125, 3649.
C
C
215H273N76O135P21: 6731.4057, found: 6731.3960. 50-
TGACJGTCATAGCTGTTTCCTG-30 m/z calculated
for C215H273N76O135P21: 6731.4057, found: 6731.4161.
´
28. Pochet, S.; Dugue, L. Nucl. & Nucl. 1995, 14, 1195.
23. Davies, D. B. Prog. NMR Spectrosc. 1978, 12, 135.
24. Varaprasad, C. V.; Bulychev, N.; Grollman, A. P.;
Johnson, F. Tetrahedron Lett. 1996, 37, 9.
29. Beaussire, J.-J.; Pochet, S. Tetrahedron 1998, 54, 13547.
30. Law, S. M.; Eritja, R.; Goodman, M. F.; Breslauer, K. J.
Biochemistry 1996, 35, 12329.
25. Roelen, H.; Saris, C.; Brugghe, H.; van den Elst, H.;
Westra, J.; van der Marel, G.; van Boom, J. Nucleic Acids
Res. 1991, 19, 4361.
31. Kouchakdjian, M.; Bodepudi, V.; Shibutani, S.; Eisen-
berg, M.; Johnson, F.; Grollman, A. P.; Patel, D. J.
Biochemistry 1991, 30, 1403.
26. Compound (5): 1H NMR (DMSO-d6): d 1.05–1.18 (m,
12H, CH3–i-Pr), 2.13–2.27 (m, 1H, H20), 2.55–2.60 (m,
1H, CH2CN), 2.67–2.71 (m, 1H, CH2CN), 2.89 (m, 1H,
H200), 3.11 (m, 1H, H50), 3.24(m, 1H, H5 00), 3.42–3.56 (m,
3H, 2 · CH–i-Pr and CH2O), 3.64–3.68 (m, 1H, CH2O),
3.69 and 3.70 (2s, 6H, OCH3), 3.93 (m, 1H, H40), 4.53–
4.66 (m, 3H, H arom. PhOAc), 7,14–7,35 (m, 21H, H
arom. PhOAc, DMT and DPC), 8.57 and 8.58 (2s, 1H,
32. Oda, Y.; Uesugi, S.; Ikehara, M.; Nishimura, S.; Kawase,
Y.; Ishikawa, H.; Inoue, H.; Ohtsuka, E. Nucleic Acids
Res. 1991, 19, 1407.
33. McAuley-Hecht, K. E.; Leonard, G. A.; Gibson, N. J.;
Thomson, J. B.; Watson, W. P.; Hunte, r. W. N.; Brown,
T. Biochemistry 1994, 33, 10266.
34. Plum, G. E.; Grollman, A. P.; Johnson, F.; Breslauer, K.
J. Biochemistry 1995, 34, 16148.