close

Вход

Забыли?

вход по аккаунту

?

Models of Putative (AH)G(AH)G Nucleobase Quartets.

код для вставкиСкачать
Communications
Purine Quartets
DOI: 10.1002/anie.200500896
Models of Putative (AH)G(AH)G Nucleobase
Quartets**
Pilar Amo-Ochoa, Pablo J. Sanz Miguel, Patrick Lax,
Ins Alonso, Michael Roitzsch, Flix Zamora,* and
Bernhard Lippert*
The existence of homo(nucleobase) quartets such as those of
guanine (G), adenine (A), thymine (T), and uracil (U) in
multistranded nucleic acid structures is well-established, and
the awareness of the potential relevance of such quartets (also
known as tetrads) in biology and medicine is steadily
increasing.[1?3] Guanine quartets (G4) can be formed through
the folding of an individual single-stranded oligonucleotide or
through the association of either four single-stranded oligo[*] Dr. F. Zamora
Departamento de Qumica Inorg"nica
Universidad Aut'noma de Madrid
28049 Madrid (Spain)
Fax:(+34) 91-497-4833
E-mail: felix.zamora@uam.es
Dipl.-Chem. P. J. Sanz Miguel, Dipl.-Chem. P. Lax,
Dipl.-Chem. M. Roitzsch, Prof. B. Lippert
Fachbereich Chemie
Universit?t Dortmund
44221 Dortmund (Germany)
Fax: (+ 49) 231-755-3797
E-mail: bernhard.lippert@uni-dortmund.de
Dr. P. Amo-Ochoa
Departamento de Tecnologa Industrial
Universidad Alfonso X ?El Sabio?
28691 Villanueva de la CaGada, Madrid (Spain)
Dr. I. Alonso
Departamento de Qumica Org"nica
Universidad Aut'noma de Madrid
28049 Madrid (Spain)
[**] We thank the Deutsche Forschungsgemeinschaft, the Fonds der
Chemischen Industrie, UAM (CS13-541A-9-640), and MAT (200405589-C02-02) for financial support, and CJsar Pastor for his
contribution to the solution of the structure of 1.
AH = adeninium ion, G = guanine.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
5670
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 5670 ?5674
Angewandte
Chemie
nucleotides, two hairpins, or two cyclic oligonucleotides. The
exceptional stability of G4 quartets facilitates the formation of
further quartets through stacking. This also applies to mixed
nucleobase quartets such as CGCG,[1] TATA,[2a] and even
GCAT.[4] Surprisingly, the existence of AGAG quartets is still
controversial[5] despite the observation and characterization
of four different AG mispair types in nucleic acid structures.[6]
Herein, we report the X-ray crystal structure analyses of
two closely related quartets, each with two protonated
adenine model nucleobases (9-methyladeninium, 9-MeAH+,
?AH?) and two 6-oxopurine bases (9-ethylguanine, 9-EtGH,
?G? or 9-methylhypoxanthine, 9-MeHxH, ?Hx?). Additionally, ab initio calculations for the model quartet consisting of
two 9-ethyladeninium and two 9-methylguanine bases are
reported which confirm the existence of the quartet observed
in the solid-state structure and at the same time, point to the
existence of a second variant.
[{(9-MeAH)(9-EtGH)}2](BF4)2 (1) and [{(9-MeAH)(9-MeHxH)}2](ClO4)2 (2) were isolated from an ethanol/
water mixture.[7] Figure 1 shows 1 in two perspectives. The
Table 1: Hydrogen-bond lengths [P] between bases in nucleobase
quartets 1 and 2 and in the calculated structure I.
Figure 1. The cationic nucleobase quartet [{(9-MeAH)(9-EtGH)}2]2+ (1)
viewed a) from the top and b) from the side, with atom numbering
scheme and eight intramolecular hydrogen bonds between the bases
shown. The structure of [{(9-MeAH)(9-MeHxH)}2]2+ (2) is virtually
identical (not shown).
Figure 2. Association pattern of base quartet 2 leading to an infinite
tape structure. Structure 1 displays a rather similar arrangement.
Perchlorate anions connect adjacent tapes of base quartets.
structure of 2 is closely related and is provided in the
Supporting Information. Both quartets[8] consist of a central
centrosymmetric 9-methyladeninium dimer with hydrogen
bonds between the AHN7 and AHN6H2 sites. Each of the two 6oxopurine bases are hydrogen bonded through GN7 and GO6
to the AHN1H and AHN6H2 sites of adeninium. The distances
between the nitrogen and oxygen atoms in the hydrogen
bonds NHиииN and NHиииO of 1 and 2 are listed in Table 1.
They are well within the range of distances typically observed
in nucleobase pairing motifs, including Watson?Crick base
pairs, for example.[9] In 1, the hydrogen bonds between the
G AH+ pairs are shorter than those within the central
AH+ AH+ pair. A similar trend is observed in 2. Moreover,
the heteroaromatic AHCH8 proton of the adeninium ions of
Angew. Chem. Int. Ed. 2005, 44, 5670 ?5674
Intramolecular hydrogen bonds
1
I
AH
N7иииAHN6
AH
N6иииGO6
AH
N1иииGN7
AH
C8иииGO6
2.860(4)
2.749(4)
2.833(4)
3.350(4)
3.082
2.875
2.966
4.085[a]
AH
N7иииAHN6
AH
N6иииHxO6
AH
N1иииHxN7
AH
C8иииHxO6
Intermolecular hydrogen bonds
1
G
G
N2иии N3
3.008(4)
?
Hx
Hx
C2иии N3
2
2.842(7)
2.739(7)
2.838(7)
3.318(8)
2
3.284(9)
[a] Average values (max. difference 0.002 P).
both 1 and 2 are involved in weak hydrogen bonds to the
G/Hx
O6 atoms of guanine and hypoxanthine, respectively.
These contacts, although markedly longer than common
NHиииN and NHиииO hydrogen bonds (Table 1), still clearly
qualify for CHиииO hydrogen bonds as observed in other
nucleobase associations.[10]
Both quartets are almost completely flat; the maximal
deviation of individual atoms from the best plane amounts to
0.03 C. The base quartets are linked through pairs of
hydrogen bonds involving the so-called ?sugar edges? of the
6-oxopurines (GN3 and GN2H2 in 1, and HxN3 and HxC2H in 2)
and thus form infinitely long tapes (see Figure 2 and
Supporting Information). The presence of a CHиииN contact
(in 2) is noteworthy. Similar hydrogen-bonding patterns and
lengths are also realized in PtII complexes of these nucleobases.[11] Excluding the long hydrogen-bonding contacts
between AHC8H of 9-MeAH+ and GO6 of the 6-oxopurines
in 1 and 2, the intermolecular separations through AHN3 and
G
N2H2 (1), and AHN3 and HxC2H (2) are clearly longer than
analogous distances observed in other quartet structures,
which supports the consideration of 1 and 2 as true base
quartets.
The packing of individual tapes in 1 and 2 is different. In
both cases the tapes are stacked on top of each other at a
distance of 3.4 C, to produce stacks. The stacks are slightly
inclined (758 in 1, 778 in 2) and separated by the tetrahedral
counterions (BF4 in 1, ClO4 in 2). Whereas adjacent stacks
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5671
Communications
are oriented in a parallel fashion in 1, only every second stack
is parallel in 2. Directly adjacent stacks form angles of 658
with each other in 1 (see Supporting Information). An
analysis of the hydrogen-bonding interactions in 2 reveals
that with the exception of HxC8H, all remaining protons of the
two nucleobases not already engaged in hydrogen-bond
formation between bases are involved in hydrogen bonds
between the quartet and the ClO4 anions (the ClO4 anions
do not display rotational disorder). Specifically, HxN1H and
Hx
C9CH3 as well as AC2H and AC9CH3 of 9-MeAH+ form such
contacts, the lengths of which range between 3.0?3.4 C (see
Supporting Information).
Formation of the base quartets 1 and 2 can be envisaged as
an association of two AH+anti G/Hxsyn mispairs[12] (Figure 3).
There are two principal arrangements (I, II) of dimerization,
which are interrelated by a simple sliding motion of the two
mispairs. In both cases, hydrogen-bonding interactions
between donor (Do) and acceptor (A) sites can form, which
give rise to favorable secondary electrostatic interactions,[13]
even though their distribution is different in the two cases.
Arrangement I is realized in 1 and 2. Arrangement II
combines two of the four possible mispairs between A and
G, namely AH+anti Gsyn[12] and Asyn Ganti,[14] ignoring the fact
that in this second pair, the adenine is likewise protonated.
Both mismatches have been shown to be easily accommodated within double-stranded B-form DNA.
By using the Gaussian suite of programs,[15] energyminimized structures of I and II (+ 2 charged each) were
calculated and fully optimized (RHF/6-31G* level, see
Supporting Information). In all cases, frequency calculations
were carried out to confirm that the geometries represented
energy minima. Results for I and II are respectively shown in
Figure 4 and Figure 5. Arrangement I is slightly less stable
than II by 23.29 kJ mol 1.[16] Hydrogen-bond lengths in the
calculated structure I show trends similar to those found in 1
(Table 1), but absolute values are longer in the calculated
structure. This applies in particular to the central pair of
intermolecular hydrogen bonds between AHN7 and AHN6 sites
Figure 3. Formal relationship between the AH+anti G/Hxsyn mispair (top), base quartets I and II (middle), and a metal-containing base quartet II?
(bottom). Favorable secondary electrostatic interactions in hydrogen bonds are indicated by crossed dashed lines.
5672
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 5670 ?5674
Angewandte
Chemie
Figure 4. Energy-minimized gas-phase structure (Gaussian) of
[{(9-EtAH)(9-MeGH)}2]2+ (I) with views from top (above) and side
(bottom). Calculated hydrogen-bond lengths are listed in Table 1.
units, the four nucleobases are perfectly planar. The situation
in II? stands in sharp contrast to the AGAG quartet, which
contains neutral bases only. As reported by Gu and Leszczynski[5d] and confirmed by our calculations, the AGAG
quartet is V-shaped and far from planar (see Supporting
Information). Consequently this quartet has been considered
unlikely to play a role in biologically relevant base quartets.
Our calculations clearly indicate that protonation of the two
adenine bases in AGAG to give (AH)G(AH)G as realized in
II causes the V-shaped quartet to flatten appreciably
(Figure 5). Furthermore, if the four bases in II are constrained
to coplanarity, the energy penalty is only 4.50 kJ mol 1. This
strongly suggests that II can be incorporated into a fourstranded structure.
It is less clear how structure I fits into a four-stranded
DNA. However, the fact that base sextets can be accommodated into a DNA quadruplex[18] implies that a shift from a
square II to a parallelogram I should not present an
insurmountable obstacle. Moreover, the central symmetric
AH+ AH+ pair corresponds to that realized in the double
helix of poly(AH+) poly(AH+).[19] Stacking of two of these
quartets I is feasible, as shown by molecular modeling with
HyperChem 6.02 (Figure 6).[20] In this model, the four central
adeninium dinucleotides are parallel in alignment and the two
layers are rotated by 258.
Figure 6. Calculated structure (HyperChem 6.02) of the stacked
arrangement of two base quartets (AH)G(AH)G (I) joined by phosphodiester bonds between identical bases. Bases closer to viewer are indicated by larger letters.
Figure 5. Energy-minimized gas-phase structure (Gaussian) of
[{(9-EtAH)(9-MeGH)}2]2+ (II) with views from top (above) and side
(bottom). Calculated hydrogen-bond lengths [P]: AH1N6иииG1O6, 2.877 P;
AH1
N1иииG1N7, 3.013 P; AH1N6иииG2O6, 2.910 P; AH1N7иииG2N1, 3.212 P. The
other hydrogen bonds are either identical or agree within 0.001 P. The
quartet is slightly saddle shaped, with two guanine bases pointing
downward (dihedral angle 12.68), and the two adeninium ions pointing
upward (dihedral angle 20.68).
of the adeninium bases. In the calculations of I and II, the
combinations of the bases used (9-EtAH+ and 9-MeGH) were
applied for a more direct structural comparison with a metalcontaining mixed adenine/hypoxanthine quartet which we
recently characterized by X-ray crystallography (II? in
Figure 3).[17] For II?, in which the acidic AHN1 protons of the
two adeninium ions are replaced by linear [trans-(NH3)2PtII]
Angew. Chem. Int. Ed. 2005, 44, 5670 ?5674
In conclusion, we have shown that unlike the case for the
purine quartet resulting from the dimerization of the neutral
Asyn Ganti pair, the bases that form the purine quartet derived
from two AH+anti Gsyn pairs are completely (structure I) or
almost (structure II) coplanar. The exocyclic GN2H2 groups
are not involved in quartet formation and the hypoxanthine
base behaves analogously. How realistic is the protonation of
a nucleobase, in this case A, at physiological pH? There is no
doubt that protonation equilibria of nucleobases can be
significantly disturbed; consequently nucleobase pKa values
can be shifted if favorable hydrogen bonding stabilizes the
protonated base.[21] The presence of hemiprotonated cytosine
mismatches at physiological pH[1] is yet another example of
this principle.
Received: March 10, 2005
Published online: August 5, 2005
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5673
Communications
.
Keywords: DNA structures и hydrogen bonds и nucleobases и
purine quartets
[1] D. J. Patel, S. Bouaziz, A. Kettani, Y. Wang in Oxford Handbook
of Nucleic Acid Structure (Ed.: S. Neidle), Oxford University
Press, Oxford, 1999, p. 389.
[2] a) G. N. Parkinson, M. P. H. Lee, S. Neidle, Nature 2002, 417,
876; b) P. K. Patel, A. S. R. Koti, R. V. Hosur, Nucleic Acids Res.
1999, 27, 3836; c) B. Pan, Y. Xiong, K. Shi, M. Sundaralingam,
Structure 2003, 11, 825; d) B. Pan, Y. Xiong, K. Shi, J. Deng, M.
Sundaralingam, Structure 2003, 11, 815.
[3] J. T. Davis, Angew. Chem. 2004, 116, 684; Angew. Chem. Int. Ed.
2004, 43, 668.
[4] N. Escaja, J. L. Gelpi, M. Orozco, M. Rico, E. Pedroso? C.
Gonzalez, J. Am. Chem. Soc. 2003, 125, 5654.
[5] a) A. I. H. Murchie, D. M. J. Lilley, EMBO J. 1994, 13, 993; b) D.
Mohanty, M. Bansal, Biophys. J. 1995, 69, 1046; c) M. C. Shiber,
E. H. Braswell, H. Klump, J. R. Fresco, Nucleic Acids Res. 1996,
24, 5004; d) J. Gu, J. Leszczynski, J. Phys. Chem. A 2000, 104,
1898. In ref. [5b] the possibility of adenine protonation is briefly
mentioned, yet is not further elaborated.
[6] W. N. Hunter, T. Brown in Oxford Handbook of Nucleic Acid
Structure (Ed.: S. Neidle), Oxford University Press, Oxford,
1999, p. 389.
[7] 1 was obtained from a water/ethanol mixture (1:1, 15 mL)
solution containing 9-EtGH, 9-MeA (0.21 mmol each) and
Zn(BF4)2 (0.8 mmol) after heating at 80 8C for 24 h, filtration,
and keeping the solution for 2 days at 5 8C. Colorless crystals of 1
were obtained in 63 % yield. Elemental analysis for C, H, and N
was satisfactory. Colorless crystals of 2 were isolated from a
mixture of HClO4 (0.1 mol L 1) in water/ethanol (1:1, 20 mL)
containing 9-MeHxH and 9-MeA (0.15 mmol each) after few
days at room temperature. The composition of 2 was established
by X-ray crystallographic analysis.
[8] Crystal structure data for 1: C26H34N20O2B2F8, Mr = 832.35,
colorless crystals, triclinic P1?, a = 8.4649(2), b = 9.1145(2), c =
12.2988(2) C, a = 95.261(1), b = 105.124(1), g = 97.024(1)8, V =
901.56(3) C3, Z = 2, 1calcd = 2.161 Mg m 3, CuKa radiation, l =
1.54178 C, m = 4.085 mm 1, T = 296 K. 2988 data (2514 unique,
Rint = 0.07, 3.75 < q < 70.318) were collected on a Bruker
SMART 6K CCD area-detector three-circle diffractometer
with a Rigaku Rotating Anode by using narrow frames (0.38 in
w) and were corrected empirically (SADABS: G. M Sheldrick,
Bruker AXS Inc., Madison, WI, 2000) for absorption. The raw
intensity data frames were integrated with the SAINT program
(SAINT + NT Version 6.04, SAX Area Detector Integration
Program, Bruker AXS Inc., Madison, WI, 1997?2001), which
also applied corrections for Lorentz and polarization effects.
Crystal structure data for 2: C48H56N36O20Cl4, Mr = 1595.06,
colorless crystals, monoclinic, space group P21/c, a = 8.804(2),
b = 8.193(2), c = 24.008(6) C, b = 102.85(3)8, V = 1688.4(6) C3,
Z = 1, 1calcd = 1.573 Mg m 3, MoKa radiation, l = 0.71073 C, m =
0.276 mm 1, T = 296 K. 2946 data (891 unique, Rint = 0.05, 2.37 <
q < 26.418) were collected on an Enraf-Nonius Kappa CCD
diffractometer. Data reduction and cell refinement were carried
out by using the programs DENZO and SCALE-PACK (Z.
5674
www.angewandte.org
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
Otwinowski, W. Minor, Methods Enzymol. 1997, 276, 307). Both
structures were solved by standard Patterson methods and
refined by full-matrix least-squares methods based on F2 with the
SHELXTL-PLUS (G. M. Sheldrick, Siemens Analytical X-ray
Instruments, Inc., Madison, WI, 1990) and SHELXL (G. M.
Sheldrick, SHELXL 97, University of GTttingen, Germany,
1997) programs. All non-hydrogen atoms in the structures were
refined anisotropically. The fluoride ligands in BF4 of 1
displayed rotational disorder over five positions. The hydrogen
atoms were either included in geometrically calculated positions
and refined with isotropic displacement parameters according to
the riding model, or found with difference Fourier synthesis, and
refined isotropically. CCDC-265600 and CCDC-265403 (1 and 2)
contain the supplementary crystallographic data for this paper.
These data can be obtained free of charge from the Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_
request/cif.
W. Saenger, Principle of Nucleic Acid Structure, Springer, New
York, 1984, p. 123.
a) M. C. Wahl, S. T. Rao, M. Sundaralingam, Nat. Struct. Biol.
1996, 3, 24; b) B. Auffinger, S. Louise-May, E. Westhof, J. Am.
Chem. Soc. 1996, 118, 1181; c) G. R. Desiraju, T. Steiner, The
Weak Hydrogen Bond, Oxford University Press, Oxford, 1999.
a) R. K. O. Sigel, E. Freisinger, S. Metzger, B. Lippert, J. Am.
Chem. Soc. 1998, 120, 12 000; b) E. Freisinger, I. B. Rother, M. S.
LVth, B. Lippert, Proc. Natl. Acad. Sci. USA 2003, 100, 3748.
X. Gao, D. J. Patel, J. Am. Chem. Soc. 1988, 110, 5178.
a) J. Pranata, S. G. Wierschke, W. L. Jorgensen, J. Am. Chem.
Soc. 1991, 113, 2810; b) J. Sartorius, H.-J. Schneider, Chem. Eur.
J. 1996, 2, 1446.
T. Brown, W. N. Hunter, G. Kreale, O. Kennard, Proc. Natl.
Acad. Sci. USA 1986, 83, 2402.
a) Gaussian 98 (Revision A.7), M. J. Frisch et al., see Supporting
Information; b) Gaussian 03, (Revision C.02), M. J. Frisch et al.,
see Supporting Information.
This value refers to electronic energy only. If zero-point energies
are considered as well, I is less stable than II by only
17.23 kJ mol 1.
M. Roitzsch, B. Lippert, Inorg. Chem. 2004, 43, 5483.
A. Kettani, A. Gorin, A. Majumdar, T. Hermann, E. Skripkin, H.
Zhao, R. Jones, D. J. Patel, J. Mol. Biol. 2000, 297, 627.
A. Rich, D. R. Davies, F. H. C. Crick, J. D. Watson, J. Mol. Biol.
1961, 3, 71.
Hyper Chem Program (Release 6.02), Hypercube, 1115 NW 4th
Street, Gainesville, FL 32601, USA. Following an optimization
of the base quartet by applying the PM3(tm) method, alkyl
groups were removed from the N9 positions and substituted by
deoxyribose units. The layer was then duplicated and bonded to
the first one through phosphate groups (through 5? and 3? ends,
pairs of (AH)2G2 helices, right-handed and antiparallel). External sugar-phosphate chains were selected and optimized by
AMBER. In this situation, the nucleobases were frozen
(AMBER: W. D. Cornell, P. Cieplak, C. I. Bayly, I. R. Gould,
K. Merz, D. M. Ferguson, D. C. Spellmeyer, T. Fox, J. W.
Caldwell, P. A. Kollman, J. Am. Chem. Soc. 1995, 117, 5179).
P. C. Bevilacqua, T. S. Brown, S.-i. Nakano, R. Yajima, Biopolymers 2004, 73, 90, and references therein.
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 5670 ?5674
Документ
Категория
Без категории
Просмотров
1
Размер файла
166 Кб
Теги
model, nucleobases, quartets, putative
1/--страниц
Пожаловаться на содержимое документа