Inverting the charges of natural nucleobase quartets: A planar platinum-purine quartet with pronounced sulfate affinity

Article abstract of DOI:10.1002/anie.200502878

(Chemical Equation Presented) Variations on nature: Coordinative metal-ligand bonds instead of cyclic hydrogen bonds, an anion (sulfate) in the center rather than a cation, and an overall positive charge instead of a negative one are the major differences between the artificial platinum-purine quartet 2 (left) and the naturally occurring guanine quartet G₄ (right).

Full text of DOI:10.1002/anie.200502878

Angewandte
Chemie
Nucleobase–Metal Complexes
the same time we report on the corresponding molecular
triangle, the analogy of which to natural purine triplets is less
obvious. As we have shown in a number of cases,[5] simulta-
neously N1- and N7-coordinated purine bases can behave as
bridging ligands forming 908 angles. Complexes with linear
coordinating metal ions M such as trans-(NH3)2PtII, Ag+ or
Hg2+ lead to diverse architectures containing right angles, for
example, rectangles of the type cyclo-(N1-M-N1,N7-M-N7)2
(Scheme 2).
DOI: 10.1002/anie.200502878
Inverting the Charges of Natural Nucleobase
Quartets: A Planar Platinum–Purine Quartet with
Pronounced Sulfate Affinity**
Michael Roitzsch and Bernhard Lippert*
The formation of four-stranded nucleic acid structures like
those in DNA and RNA is closely connected with the
existence of guanine tetrads (G4). In G4 four purine bases are
interconnected by eight hydrogen bonds, and the structure is
further stabilized by a central cation.[1] The negative charges
of the phosphate groups are located at the periphery of this
motif. Similar tetrads consisting of uracil[2] or thymine[3] are
based essentially on the same principle. The significance of
tetrastranded DNA concerning its function in the telomeres
as well as in regulatory regions of the DNA is presently the
subject of intense investigations.[4]
Here we describe a non-natural quartet of 9-methylpur-
ine, which differs from the natural nucleobase quartets in that
cationic trans-(NH3)2PtII moieties are located at the periphery
of the quartet and an anion is trapped in the center
(Scheme 1). The hydrogen bonds in the natural-base quartets
Scheme 2. Possible architectures of purine–metal complexes.
Using unsubstituted 9-methylpurine (Pu), we have now
succeeded in preparing a molecular square with the coordi-
nation sequence cyclo-(N1-Pt-N7)4. This compound, all-trans-
[{(NH3)2Pt(m-N1-Pu-N7)}4]8+ (2), forms spontaneously from
the 1:1 complex trans-[(NH3)2Pt(Pu-N7)(H2O)]2+ (1). Obser-
1
vation of the product formation by H NMR spectroscopy
shows that several molecules of 1 condense to form short
chains, which undergo complete reaction at room temper-
ature over five days to yield two distinct cyclic complexes
(Scheme 3). Besides the molecular square 2, a molecular
triangle of composition all-trans-[{(NH3)2Pt(m-N1-Pu-N7)}3]6+
(3) is formed. In both compounds, the Pu ligands are
symmetrically equivalent and hence in the 1H NMR spectrum
only a single set of signals can be observed for the respective
species. According to signal integration, the formation of 3 is
preferred over formation of 2 (2/3 = 0.6:1). The situation
changes when the reaction is carried out in the presence of
sulfate, which acts as a template. In 100 mm Na2SO4,
formation of 2 is preferred, and the 2/3 ratio increases to
2.5:1. In addition, the presence of sulfate anions speeds up the
reaction, such that product formation is complete within three
days at room temperature.
Scheme 1. G=guanine, Pu=9-methylpurine, M2+ =trans-[(NH3)2PtII]2+
.
are replaced by coordinative Pt–nucleobase bonds in the
artificial quartet, and the overall negative charge of the
natural quartets is changed to positive in the artificial one. At
Crystallization of 2 and 3 proved to be difficult, and only
very small crystals could be obtained, which have the
composition 2-(ClO4)8·6H2O (2a) and 3-(PF6)6·6H2O (3a),
respectively.[6] The cations of 2a and 3a are shown in Figure 1.
In both compounds the purine ligands and the PtII atoms are
essentially coplanar, and the ammine ligands are perpendic-
[*] Dr. M. Roitzsch, Prof. B. Lippert
Fachbereich Chemie, Universität Dortmund
44221 Dortmund (Germany)
À
ular to this plane. Pt N distances are normal. In 2a, the
distances between the PtII atoms on opposing sides of the
Fax: (+49)231-755-3797
E-mail: bernhard.lippert@uni-dortmund.de
À
À
square are 9.03(1) Š (Pt1 Pt1a) and 9.09(1) Š (Pt2 Pt2a).
À
The square has outer diagonals of 17.07(4) Š (C9a C9aa) and
[**] Support by the Deutsche Forschungsgemeinschaft is gratefully
acknowledged.
À
17.24(6) Š (C9b C9ba). The diagonals in the center are
Angew. Chem. Int. Ed. 2006, 45, 147 –150
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
147

Communications
À
À
6.08 Š (H6a H6aa) and 6.24 Š (H6b H6ba). The
N1-Pt-N7 angles deviate slightly from the ideal
angle of 1808 and measure 176.8(5)8 (N1b-Pt1-
N7a) and 177.4(5)8 (N1aa-Pt2-N7b), respectively.
As a consequence, the angle between the PtII atoms
coordinated to the N1 and N7 sites of a purine
ligand likewise deviate from the ideal 908–88.18 in
purine a and 85.18 in purine b. In the crystal
structure, the cations are packed on top of each
other in a slightly staggered manner with a stacking
distance of 7.5 Š. Some of the perchlorate anions,
which display pronounced positional disorder, are
located between the cations yet appear not to be
firmly fixed to these.
In 3a, the distances between the PtII atoms are
in the range of 5.91(1)–6.00(1) Š, while the dis-
tances between the C9 atoms are between 12.67(2)
and 12.73(3) Š. The N1-Pt-N7 angles in 3a deviate
markedly from the ideal 1808 and are in the range
of 168.8(5)–170.8(4)8. The angles between the PtII
atoms coordinated to the N1 and N7 sites of a
purine ligand deviate strongly from 908 and range
between 67.88 and 70.88.
Scheme 3.
Of the six hexafluorophosphate anions in 3a,
one is located above and another below the center
of each cation, and the C3 axis of the anion points towards the
cation. The other face of the PF6À anion is oriented toward a
purine ligand of a parallel cation. This leads to a helical
arrangement of the cations in the structure, making it chiral
(Figure 2). The distance between the planes of the cations is
7.6 Š.
Both compounds show a pronounced affinity for sulfate
anions. The binding of SO42À was monitored by concentration-
dependent 1H NMR spectroscopy, where the concentration of
À
Figure 2. Section of the packing pattern of 3a. The PF6 anions are
Figure 1. Cations of 2a and 3a with atomic numbering.
layered between the cations, thereby producing a helix.
148
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 147 –150

Angewandte
Chemie
the anion was varied, while the concentration of the
respective cyclic compound was kept constant. In
[D6]DMSO, 2 shows a new signal set upon addition of
nBu4NHSO4. The resonances of the H6 protons and those of
the ammine ligands are strongly shifted towards low field,
whereas the other signals of the purine ligands are scarcely
influenced (Figure 3). In D2O, the H6 signals shift downfield
with increasing concentrations of Na2SO4 (Figure 4). The
influence on the other resonances of the purine ligands is only
very weak, and the resonances of the protons of the ammine
ligands cannot be observed because of isotope exchange with
the solvent.
The concentration dependence of the H6 signals can be
used to determine the association constants.[7] For the binding
of sulfate in D2O, we have found association constants of
(7.2 Æ 1.2) ” 104 LmolÀ1 for 2 and (9.9 Æ 0.6) ” 103 LmolÀ1 for
3. These values indicate very strong sulfate binding in water
and are in the range known for very strong anion receptors
such as aza crown ethers and cryptands.[8,9] We are currently
performing experiments to investigate the selectivity of 2 and
3 for different anions. These studies will be accompanied by
DFT calculations in order to elucidate the mode of anion
binding in more detail. Preliminary results suggest that the
anion is trapped by 2 by means of multiple hydrogen bonds,
with the S atom residing in the center of the cavity. In addition
to the general aspect of sulfate binding, the templating effect
of this anion during the formation of the square 2 is
noteworthy, as is the topical issue of factors influencing
equilibria of metallosupramolecular assemblies.[10] Finally, 2
represents the first example of a true purine square with four
metal entities at the periphery and alternating N1,N7
coordination. Previously described cases of metal-containing
purine quartets displayed rectangular shapes[5a,e] as a conse-
quence of pairwise N7,N7 and N1,N1 metal binding. Another
point of interest is whether these cationic nucleobase squares
are capable of interacting with natural tetrastranded nucleic
acid structures.[11]
Figure 3. 1H NMR spectra of 2 recorded in [D6]DMSO. The uppermost
spectrum was recorded in the absence of sulfate ions, the sÀpectra
Experimental Section
below with increasing amounts of nBu4NHSO4: a) no HSO4
,
Preparation of trans-[(NH3)2Pt(9MePu-N7)Cl]ClO4 (1a): A solution
of Pu[12] (134 mg, 1.00 mmol) in 5 mL of 2.4m HClO4 was added at
once to trans-[(NH3)2Pt(OH)Cl]·H2O[13] (300 mg, 1.00 mmol). The
mixture was stirred for 5 min, then the yellowish precipitate was
filtered off and washed with ca. 20 mL of 1m NaClO4 and 2 mL of
H2O. The product was dried under vacuum. Yield: 275 mg, 55.2%;
white powder. C,H,N analysis calcd (%): C 14.5, H 2.4, N 16.9; found:
C 14.2, H 2.4, N 16.7. 1H NMR: (D2O, pD 2.4): d = 9.56 (s, H6), 9.12 (s,
H2), 9.09 (s, H8), 4.06 ppm (s, CH3).
À
b) 0.1 equiv HSO4À, c) 0.5 equiv HSO4À, d) 1.0 equiv HSO4
,
e) 3.0 equiv HSO4À. Upon addition of sulfate a host–guest complex
forms, which displays a new set of signals indicating the slow
exchange of the sulfate ions. While the signals assigned to H6 and to
the protons of the ammine ligands are strongly low-field shifted, the
other resonances are only weakly affected. This is consistent with the
view that the sulfate is trapped in the center of the molecular square
and that the NH3 ligands are involved in hydrogen bonding with
sulfate.
Preparation of 2a and trans-[{(NH3)2Pt(m-N1-Pu-N7)}3]-
(ClO4)6·3H2O (3b): A solution of 1a (199 mg, 0.400 mmol) in
15 mL of H2O was treated with 1 mL of 1m HNO3 and AgNO3
(67.9 mg, 0.400 mmol), and the mixture was stirred for 3 d at 458C
in the dark. Precipitated AgCl was removed by filtration, and the
solution was concentrated at reduced pressure to a volume of 5 mL.
The solution was cooled to 48C for 12 h, and the crude 2a that
precipitated was isolated by filtration. Addition of 800 mL of 1m
NaClO4 to the remaining solution precipitated crude 3b, which was
isolated by filtration. The crude products were washed twice with
250 mL of H2O and dried under vacuum. The products obtained were
dissolved in 40 mL (2a) and 20 mL of H2O (3b), respectively, and
then precipitated by addition of 1m NaClO4, filtered off, and dried
under vacuum. Yields of isolated products were 38 mg (16%) of 2a
and 39 mg (17%) of 3b. C,H,N analysis calcd (%): 2a: C 12.4, H 2.4,
N 14.5; found: C 12.1, H 2.5, N 14.8; 3b: C 12.2, H 2.6, N 14.3; found:
1
C 12.1, H 2.5, N 14.3. H NMR: (2, D2O, pD 2.0): d = 10.47 (s, H6),
9.81 (s, H2), 9.60 (s, H8), 4.22 ppm (s, CH3); (3, D2O, pD 2.0): d =
11.44 (s, H6), 9.49 (s, H2), 9.37 (s, H8), 4.18 ppm (s, CH3). The
hexafluorophosphate salt 3a was prepared by precipitation from a
solution of 3b with 0.2m solution of KPF6 in H2O. 3a was
characterized by X-ray crystallography.
Figure 4. Change of the chemical shifts Dd of the H6 signals of 2 (&)
~
and 3 ( ) in D2O with increasing amounts of Na2SO4. In both cases
the perchlorate salts (2a and 3b) where used, since perchlorate ions
appear to be only weakly bound to the cations. The solid lines where
obtained by a least-squares fitting procedure,[7] which yielded the
association constants discussed in the text. Concentrations of the host
molecules were 0.40 mm 2 and 0.47 mm 3.
Received: August 12, 2005
Published online: November 21, 2005
Angew. Chem. Int. Ed. 2006, 45, 147 –150
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
149

Communications
[7] H. Sigel, K. H. Scheller, V. M. Rheinberger, B. E. Fischer, J.
Chem. Soc. Dalton Trans. 1980, 7, 1022.
[8] D. Grell, E. Grell, P. Bugnon, B. Dietrich, J. M. Lehn, J. Therm.
Anal. Calorim. 2004, 77, 483.
[9] J. L. Sessler, E. Katayev, G. D. Pantos, Y. A. Ustynyuk, Chem.
Commun. 2004, 1276.
[10] See, e.g.: a) A. Sautter, D. G. Schmid, G. Jung, F. Würthner, J.
Am. Chem. Soc. 2001, 123, 5424; b) M. Schweiger, S. R. Seidel,
A. M. Arif, P. J. Stang, Inorg. Chem. 2002, 41, 2556; c) M. Ferrer,
M. Mounir, O. Rossell, E. Ruiz, M. A. Maestro, Inorg. Chem.
2003, 42, 5890.
[11] E. Freisinger, I. B. Rother, M. S. Lüth, B. Lippert, Proc. Natl.
Acad. Sci. USA 2003, 100, 3748.
[12] N. C. Gonnella, J. D. Roberts, J. Am. Chem. Soc. 1982, 104, 3162.
[13] J. Arpalahti, R. Sillanpää, M. Mikola, J. Chem. Soc. Dalton
Trans. 1994, 1499.
Keywords: base quartets · guanine · nucleobases · platinum ·
purine
.
[1] J. T. Davis, Angew. Chem. 2004, 116, 684; Angew. Chem. Int. Ed.
2004, 43, 668, and references therein.
[2] a) C. J. Cheong, P. B. Moore, Biochemistry 1992, 31, 8406; b) J.
Deng, Y. Xiong, M. Sundaralingam, Proc. Natl. Acad. Sci. USA
2001, 98, 13665; c) B. C. Pan, Y. Xiong, K. Shi, J. P. Deng, M.
Sundaralingam, Structure 2003, 11, 815.
[3] a) P. K. Patel, R. V. Hosur, Nucleic Acids Res. 1999, 27, 2457;
b) C. Cµceres, G. Wright, C. Gouyette, G. Parkinson, J. A.
Subirana, Nucleic Acids Res. 2004, 32, 1097.
[4] a) J. Seenisamy, S. Bashyam, V. Gokhale, H. Vankayalapati, D.
Sun, A. Siddiqui-Jain, N. Streiner, K. Shin-Ya, E. White, W. D.
Wilson, L. H. Hurley, J. Am. Chem. Soc. 2005, 127, 2944; b) A. T.
Phan, V. Kuryavyi, H. Y. Gaw, D. J. Patel, Nat. Chem. Biol. 2005,
1, 167.
[5] See, e.g.: a) I. B. Rother, M. Willermann, B. Lippert, Supramol.
Chem. 2002, 14, 189; b) M. Roitzsch, I. B. Rother, M. Willer-
mann, A. Erxleben, B. Costisella, B. Lippert, Inorg. Chem. 2002,
41, 5946; c) M. S. Lüth, M. Willermann, B. Lippert, Chem.
Commun. 2001, 2058; d) M. S. Lüth, E. Freisinger, B. Lippert,
Chem. Eur. J. 2001, 7, 2104; e) M. S. Lüth, E. Freisinger, F.
GlahØ, B. Lippert, Inorg. Chem. 1998, 37, 5044.
[6] Crystal structure data for 2a: C24 H60 N24 O38 Cl8 Pt4, Mr =
2256.81, colorless crystals, monoclinic C2/c, a = 20.825(4), b =
18.584(4), c = 16.142(3) Š, a = g = 908, b = 98.67(3)8, V=
6176(2) Š3, Z = 2, 1calcd = 2.419 MgmÀ3, m = 9.485 mmÀ1, T=
150 K. 44983 data (7069 unique, 3675 observed, Rint = 0.146,
3.08 < q < 27.52 8) were collected. Final R = 0.084 (wR2 = 0.144,
GoF = 1.104). A major disorder of three of the four crystallo-
graphically independent perchlorate anions over two or three
positions, respectively, made it necessary to use numerous
constraints to fix the geometry of these ions. As a consequence,
the R values are comparably poor and the assignment of crystal
water molecules was not possible; the amount of crystal water
given in the empirical formula therefore is based on elemental
analysis data. Crystal structure data for 3a: C18 H48 N18 O6 F36 P6
Pt3, Mr = 2067.71, colorless crystals, hexagonal P61, a =
15.085(2), b = 15.085(2), c = 42.280(9) Š, a = b = 908, g = 1208,
V= 8332(2) Š3, Z = 6, 1calcd = 2.420 MgmÀ3, m = 7.881 mmÀ1, T=
150 K. 29176 data (9153 unique, 7118 observed, Rint = 0.068,
3.06 < q < 27.498) were collected. Final R = 0.046 (wR2 = 0.088,
GoF = 1.027). The reflections of both crystals were collected on
a Nonius KappaCCD diffractometer using graphite-monochro-
mated MoKa radiation (l = 0.71069 Š) 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 EVALCCD (Bruker AXS Inc.,
Madison, WI, 2000) program. Both structures were solved by
direct methods and refined by full-matrix least-squares methods
based on F2 using the SHELXTL-PLUS (G. M. Sheldrick,
Siemens Analytical X-ray Instruments, Inc., Madison, WI, 1990)
and SHELXL-97 (G. M. Sheldrick, SHELXL-97, Universität
Göttingen, Göttingen, Germany) programs. With the exception
of the disordered perchlorate atoms in 2a and the C6b atom, two
À
fluorine atoms of a disordered PF6 anion, and the nitrogen
atoms of the ammine ligands in 3a, all non-hydrogen atoms in
the structures were refined anisotropically. The hydrogen atoms
were included in geometrically calculated positions and refined
with isotropic displacement parameters. CCDC-280955 and
280954 (2a and 3a) 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.
150
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 147 –150