Intramolecular hydrogen bonding of nucleobases

Article abstract of DOI:10.1016/S0040-4039(02)00483-5

The intramolecular hydrogen bonding of a series of linked uracil, thymine and adenine derivatives is described. The nucleoside bases have been linked by semi-flexible spacers and the chemical shifts and ΔδNH/ΔT values for the nucleoside NH protons confirm

Full text of DOI:10.1016/S0040-4039(02)00483-5

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 3157–3160
Intramolecular hydrogen bonding of nucleobases
Geoffrey T. Crisp* and Yu-Lin Jiang
Department of Chemistry, Adelaide University, Adelaide 5005, South Australia, Australia
Received 8 October 2001; revised 26 February 2002; accepted 8 March 2002
Abstract—The intramolecular hydrogen bonding of a series of linked uracil, thymine and adenine derivatives is described. The
nucleoside bases have been linked by semi-flexible spacers and the chemical shifts and DlNH/DT values for the nucleoside NH
protons confirm the strong intramolecular hydrogen bonding motif. © 2002 Elsevier Science Ltd. All rights reserved.
The design of artificial DNA-like complementary
purine–pyrimidine base pairs has enabled a better
understanding of the relationship between structural
variations and complex stability.1–3 The design of lig-
ands and receptors that will preferentially bind a spe-
cific DNA motif requires a detailed understanding of
the hydrogen bonding ability of purine and pyrimidine
bases, including normal and reversed Watson–Crick as
well as Hoogsteen bonding (Fig. 1).4 Artificial systems
that lack some form of conformational rigidity and
exhibit poor solubility often display weak binding for
dimerization.1,5 In addition to the traditional duplex
formation associated with AT(U) or CG complemen-
tary base pairs, the formation of localized triplet struc-
tures has been recognized as an important contributor
for the binding of oligonucleotides to double stranded
DNA.5–7 We report in this paper the preparation and
hydrogen bonding properties of a series of nucleobases
containing spacers with both a rigid ethynylbenzene
and a flexible alkyl chain.
The preparation of receptors 4–7 commenced with the
alkylation of 1,10-decanediol with 3-iodobenzylbro-
mide in the presence of sodium hydride to give 1 in
30% yield (Scheme 1). Subsequent coupling of 1 with
trimethylsilylacetylene in the presence of Pd(PPh3)4
and CuI8 gave 2 in 99% yield, and desilylation with
Bu4NF yielded 3 in 89% yield. Palladium-catalyzed
coupling of an excess of 3 with 3%,5%-diacetoxy-5-iodo-
deoxyuridine gave a mixture of bis-coupled product 4
in 23%9 and mono-coupled product 5 in 39% yield.
Similarly, palladium catalyzed coupling of an excess of
H
N H
H
N H
O
H N
O
CH3
O
H N
O
R
N
N
N
N
N
N
N
R
R
N
R
N
N
CH3
Watson-Crick
Reversed Watson-Crick
3
with 2%,3%,5%-triacetoxy-8-bromoadenosine gave a
R
O
CH3
mixture of bis-coupled product 6 in 14% and mono-
coupled product 7 in 27% yield. The mono-coupled
product 5 was subsequently coupled with 2%,3%,5%-tri-
acetoxy-8-bromoadenosine to give 8 in 17% yield.
N
CH3
H
O
N
H
O
H
N H
N
CH3
H
N
O
H
N H
N
N
N
R
N
N
N
O
R
N
N
R
N
O
The preparation of DNA base pair triplet models
commenced with the alkylation of thymine with vari-
ous alkynyliodides in the presence of potassium car-
bonate as described in Scheme 2 and the alkylation of
adenine with 3,5-diiodobenzylbromide in aqueous
NaOH in the presence of Aliquat 336 to give 10
(Scheme 3).10 The palladium catalyzed coupling of 10
with alkynylthymines 9a–f then gave 11a–f11 (Scheme
3).
N
R
Hoogsteen
TAT Base-Triplet
Figure 1. Hydrogen bonding in nucleoside bases.
* Corresponding author. Fax: +61
geoffrey.crisp@adelaide.edu.au
8
83034358; e-mail:
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)00483-5

3158
G. T. Crisp, Y.-L. Jiang / Tetrahedron Letters 43 (2002) 3157–3160
R
R
I
1.
2.
SiMe3
Pd(PPh3)4 / CuI
PPh3 / NEt3 / DMA
H
2
I
O
NaH
THF/DMA
30%
O
(CH2)10
O
(CH2)10
Bu4NF.xH2O
87%
Br
+
O
HO(CH2)10OH
I
2 R = SiMe3
3 R = H
1
R1
4 23% R1 = R2 = A
5 39% R1 = A, R2 = H
6 14% R1 = R2 = B
7 27% R1 = B, R2 = H
8 17% R1 = A, R2 = B
Pd(PPh3)4 / CuI
PPh3 / NEt3 / DMA
O
(CH2)10
O
3
or
A
B
I
Br
R2
O
NH2
N
I
H
O
Pd(PPh3)4 / CuI
N
N
N
Br
PPh3 / NEt3 / DMA
N
AcO
8
AcO
5
N
O
B
Br
O
17%
OAc
AcO OAc
Br
A
I
B
Scheme 1. Preparation of DNA base pair models 4–8.
O
O
standard. The chemical shift for the adenine NH2 pro-
tons for bis-adenine 6 was 5.89 ppm, indicative of a
weak intramolecular hydrogen bond between the two
adenine units. The DlNH/DT values for the NH pro-
tons were obtained by variable temperature NMR
experiments from −60 to 60°C. Compound 8 displayed
an extraordinarily high value of 42.6 ppb/K for the NH
proton of the uracil component, much larger than the
corresponding value of 23.2 ppb/K for 4. Furthermore,
the DlNH/DT value for the amino group of the
adenine residue in 8 was 21.1 ppb/K, while the corre-
sponding value for compound 6 was only 9.6 ppb/K,
confirming the presence of an efficient hydrogen bond-
ing arrangement involving the complementary base
pairs.
H
O
CH3
(CH2)nI
K2CO3, DMSO, rt
H
O
CH3
H
N
N
N
N
H
(CH2)n
H
9a n = 4 60%
9b n = 5 62%
9c n = 6 70%
9d n = 7 56%
9e n = 8 65%
9f n = 9 74%
Scheme 2. Preparation of alkynylthymines 9a–f.
Fig. 2 shows the downfield region of the 1H NMR
spectra for compounds 11a–f at 2 mM in CDCl3. The
1H NMR signals for the thymine H-6 protons and the
protons of the central phenyl ring appear almost at the
same position in compounds 11a–f. The chemical shifts
for the adenine H-8 protons vary with the length of the
alkyl spacer between the uracil rings, whilst the signal
for the adenine H-2 protons appeared relatively con-
stant. The alternating changes in the chemical shifts for
the thymine H-6 and the adenine H-8 protons from 11a
through to 11f indicates that when the alkyl chain
contains an odd number of carbons the intramolecular
hydrogen bonding increases and that when the alkyl
chain contains an even number of carbons the
intramolecular hydrogen bonding decreases.
Compound 5 was not capable of forming intramolecu-
lar hydrogen bonds as it contains only one nucleobase,
so the 1H NMR chemical shift for the uracil NH
proton at 8.05 ppm could be used as a standard. The
chemical shift for the NH protons in bis-2%deoxyuridine
derivative 4 was 8.86 ppm (1 mM at room temperature
in CDCl3) under conditions known to minimize inter-
molecular hydrogen bonding,12 indicating that the two
uracil residues were capable of intramolecular hydrogen
bonding. Compound 8, containing both a purine and a
pyrimidine residue, displayed a NH resonance for uracil
at 9.65 ppm, again reflecting a strong intramolecular
hydrogen bond. The chemical shift for the NH2 protons
at 5.66 ppm for mono-adenine derivative 7, under
similar conditions, could also be used as an internal

G. T. Crisp, Y.-L. Jiang / Tetrahedron Letters 43 (2002) 3157–3160
3159
O
NH2
N
NH2
N
Ht
CH3
N
N
N
N
N
N
I
N
H8a
O
N
H6t
H2a
(CH2)n
10
I
9a-f
Pd(PPh3)4 / ZnCl2
NaI / piperidine
Ht
O
N
(CH2)n
N
O
CH3
H6t
11a n = 4 58%
11b n = 5 62%
11c n = 6 61%
11d n = 7 69%
11e n = 8 66%
11f n = 9 67%
Scheme 3. Preparation of DNA base pair triplet models 11a–f.
Figure 2. 1H NMR spectra of 11a–f at 2 mM in CDCl3.
(0.248 g, 0.5 mmol) and PPh3 (0.026 g, 0.10 mmol) in
NEt3/DMA (20 ml; 1:5) under N2, were added Pd(PPh3)4
(0.058 g, 0.05 mmol), CuI (0.019 g, 0.10 mmol) and 3
(0.320 g, 0.8 mmol), respectively. The dark red solution
was heated at 80°C for 2 h. The reaction solvents were
removed in vacuo. The residue was purified on silica gel
(MeOH:acetone:CH2Cl2; 1:5:70) to give 4 (0.060 g) and 5
(0.112 g), in an overall yield of 57%. For 4 1H NMR
(CDCl3, 300 MHz): l (ppm) 1.19 (m, 16H, CH2); 1.61 (m,
4H, CH2); 2.13 (s, 6H, CH3); 2.19 (s, 6H, CH3); 3.45 (t,
4H, J=6.6 Hz, CH2); 4.32, 4.39 (m, 4H); 4.39 (s, 4H,
CH2); 5.28 (m, 2H); 6.35 (m, 2H); 7.32 (m, 8H, Ar); 7.90
(s, 2H); 8.92 (s, 2H, NH). Calcd for [M+H]+
C54H63N4O16: 1023.4238. Found: 1023.4190.
References
1. Sessler, J. L.; Wang, R. J. Org. Chem. 1998, 63, 4079–
4091.
2. Sessler, J. L.; Wang, R. Angew. Chem., Int. Ed. 1998, 37,
1726–1729.
3. Schall, O. F.; Gokel, G. W. J. Org. Chem. 1996, 61,
1449–1458.
4. Nielsen, P. E. Chem. Eur. J. 1997, 3, 505–508.
5. Kuroda, Y.; Lintuluoto, J. M.; Ogoshi, H. J. Chem. Soc.,
Perkin Trans. 2 1997, 333–339.
6. Zimmerman, S. C.; Schmitt, P. J. Am. Chem. Soc. 1995,
117, 10769–10770.
7. Cho, Y. L.; Jeong, K.-S. Tetrahedron Lett. 1997, 38,
8337–8340.
8. Brandsma, L.; van den Hevel, H. G. M.; Verkruijse, H.
D. Synth. Commun. 1990, 20, 1889–1892.
9. 1,10-(Decandiol)-di{3-[5-(3%,5%-di-O-acetyl-2%-deoxy uri-
dine)ethynyl]benzyl} ether (4)
10. Shinkai, I.; Zwan, M. C. V.; Harter, F. W.; Reamer, R.
A.; Tull, R. J.; Weinstock, L. M. J. Heterocycl. Chem.
1981, 18, 197–198.
11. 9-{3,5-Di[(11-undecyn-1-yl)-1-thyminyl)]-1-benzyl} adenine
(11f)
To
a
solution of 3%,5%-diacetoxy-5-iododeoxyuridine
To a solution of 1-(3,5-diiodo-1-benzyl)adenine (0.108 g,

3160
G. T. Crisp, Y.-L. Jiang / Tetrahedron Letters 43 (2002) 3157–3160
0.25 mmol) and PPh3 (0.026 g, 0.10 mmol) in pipe-
MHz): l (ppm) 1.33, 1.54, 1.55, 1.67, 1.89 (m, CH2);
1.93 (s, 6H, CH3); 2.37 (t, 4H, J=6.6 Hz, CH2); 3.69
(t, 4H, J=7.8 Hz, CH2); 5.27 (s, 4H, CH2); 6.70 (s,
2H, NH2); 6.97 (s, 2H, CH, thymine); 7.32, 7.33 (m,
4H, Ar); 8.06 (s, 1H, H8, adenine); 8.46 (s, 1H, H2,
adenine); 10.94 (s, 2H, NH, thymine). Calcd for [M]+
C44H55N9O4: 773.4377. Found: 773.4373.
ridine (5 mL) under N2, were added Pd(PPh3)4 (0.058
g, 0.05 mmol), ZnCl2 (0.0136 g, 0.10 mmol), NaI (0.015
g, 1.0 mmol) and N-9-undecynylthymine (0.143 g, 0.50
mmol), respectively. The dark red solution was heated
at reflux (110°C) for 1 h. The solvent was removed in
vacuo and the residue was purified on silica gel
(MeOH:CH2Cl2; 10:90) to give 11f as a white solid
(0.129 g, 67%). 1H NMR (CDCl3, 2 mM, rt, 600
12. Gellman, S. H.; Dado, G. P.; Liang, G.-B.; Adams, B.
R. J. Am. Chem. Soc. 1991, 113, 1164–1173.