Loss of Hoogsteen pairing ability upon N1 adenine platinum binding.

Article abstract of DOI:10.1021/ic0109602

Chloroform- and Freon-soluble mixed thymine, adenine complexes trans-[Pt(MeNH(2))(2)(ChmT-N3)(ChmA-N1)]NO(3) (2) and trans-[Pt(MeNH(2))(2)(ChmT-N3)(TBDMS-ado-N1)]BF(4) (3) (ChmT = anion of 1-cyclohexylmethylthymine ChmTH, ChmA = 9-cyclohexylmethyladenine, TBDMS-ado = 2',3',5'-tri-tert-butyldimethylsilyladenosine) have been prepared and characterized to study their propensity to undergo Hoogsteen and/or reversed Hoogsteen pairing in solution with free ChmTH and free 3',5'-diacetyl-2'-deoxyuridine, respectively. No Hoogsteen or reversed Hoogsteen pairing between 2 and ChmT takes place in CDCl(3). In Freon, partial H bonding between N1 platinated TBDMS-ado and 3',5'-diacetyl-2'-deoxyuridine as well as its [3-(15)N] labeled analogue is unambiguously observed only below 150 K. Comparison of (1)J ((15)N-(1)H) coupling constants of 3',5'-diacetyl-2'-deoxyuridine involved in Hoogsteen pairing with free and N1 platinated adenine suggests that the interaction is inherently weaker in the case of platinated adenine. To better understand the complete absence of hydrogen bonding between the ChmA ligand in 2 and free ChmTH, ab initio calculations (gas phase, 0 K) have been carried out for Hoogsteen pairs involving adenine (A) and thymine (T), as well as simplified analogues of 2 and T, both in the presence and absence of counteranions. The data strongly suggest that reduction of the effective positive charge of the heavy metal ion Pt(2+) by counterions diminishes interaction energies. With regard to mixtures of 2 and ChmTH in chloroform, this implies that ion pair formation between the cation of 2 and NO(3)(-) may be responsible for the lack of any measurable Hoogsteen pairing in this solvent.

Full text of DOI:10.1021/ic0109602

Inorg. Chem. 2002, 41, 2855−2863
Loss of Hoogsteen Pairing Ability upon N1 Adenine Platinum Binding
,‡
,§
Kathrin S. Schmidt,† Jan Reedijk,* Klaus Weisz,* Eline M. Basilio Janke,§ Judit E. Sponer,|
,|,
,†
Jiri Sponer,* and Bernhard Lippert*
Fachbereich Chemie, UniVersita¨t Dortmund, 44221 Dortmund, Germany, Leiden Institute of
Chemistry, Gorlaeus Laboratories, Leiden UniVersity, 2300 RA Leiden, The Netherlands, Institut
fu¨r Chemie, Freie UniVersita¨t Berlin, 14195 Berlin, Germany, J. HeyroVsky Institute of Physical
Chemistry, Academy of Sciences of the Czech Republic, 18223 Prague, Czech Republic, Institute
of Biophysics, Academy of Sciences of the Czech Republic, 61265 Brno, Czech Republic, and
National Center for Biomolecular Research, 61265 Brno, Czech Republic
Received September 10, 2001
Chloroform- and Freon-soluble mixed thymine, adenine complexes trans-[Pt(MeNH ) (ChmT-N3)(ChmA-N1)]NO
2 2
3
(2) and trans-[Pt(MeNH ) (ChmT-N3)(TBDMS-ado-N1)]BF (3) (ChmT ) anion of 1-cyclohexylmethylthymine ChmTH,
2 2
4
ChmA ) 9-cyclohexylmethyladenine, TBDMS-ado ) 2′,3′,5′-tri-tert-butyldimethylsilyladenosine) have been prepared
and characterized to study their propensity to undergo Hoogsteen and/or reversed Hoogsteen pairing in solution
with free ChmTH and free 3′,5′-diacetyl-2′-deoxyuridine, respectively. No Hoogsteen or reversed Hoogsteen pairing
between 2 and ChmT takes place in CDCl3. In Freon, partial H bonding between N1 platinated TBDMS-ado and
3′,5′-diacetyl-2′-deoxyuridine as well as its [3-15N] labeled analogue is unambiguously observed only below 150 K.
Comparison of 1J (15N−1H) coupling constants of 3′,5′-diacetyl-2′-deoxyuridine involved in Hoogsteen pairing with
free and N1 platinated adenine suggests that the interaction is inherently weaker in the case of platinated adenine.
To better understand the complete absence of hydrogen bonding between the ChmA ligand in 2 and free ChmTH,
ab initio calculations (gas phase, 0 K) have been carried out for Hoogsteen pairs involving adenine (A) and thymine
(T), as well as simplified analogues of 2 and T, both in the presence and absence of counteranions. The data
strongly suggest that reduction of the effective positive charge of the heavy metal ion Pt2+ by counterions diminishes
interaction energies. With regard to mixtures of 2 and ChmTH in chloroform, this implies that ion pair formation
between the cation of 2 and NO - may be responsible for the lack of any measurable Hoogsteen pairing in this
3
solvent.
Introduction
pairing, at least in the solid state.2 Interestingly, gas-phase
calculations provide similar results.3 Only at very low
temperatures (113 K; CDClF2/CDF3 solvent mixture) the
predominance of the Watson-Crick pair for suitably sub-
stituted A and T model bases has recently been established.4
Our interest in the effects of metal ion coordination on the
H bonding pattern of nucleobases5 has recently led us to
demonstrate that Pt(II) binding to N7 of guanine (G)
The complementary nucleobases adenine (A) and thymine
(T) can associate through pairs of H bonds in a number of
ways, viz. in Watson-Crick, reversed Watson-Crick,
Hoogsteen, and reversed Hoogsteen fashion.1 With isolated
bases and unlike in DNA, Hoogsteen and reversed Hoogsteen
pairing appears to be more common than Watson-Crick
* Authors to whom correspondence should be addressed. E-mail:
lippert@pop.Uni-Dortmund.DE (B.L.); reedijk@chem.leidenuniv.nl (J.R.).
† Universita¨t Dortmund.
(2) Voet, D.; Rich, A. Prog. Nucleic Acid Res. Mol. Biol. 1970, 10, 183
and refs. cited.
(3) Sponer, J.; Leszynski, J.; Hobza, P. J. Phys. Chem. 1996, 100, 1965.
(4) Dunger, A.; Limbach, H.-H.; Weisz, K. J. Am. Chem. Soc. 2000, 122,
10109.
(5) See, e.g.: (a) Mu¨ller, J.; Sigel, R. K. O.; Lippert, B. J. Inorg. Biochem.
2000, 79, 261. (b) Sigel, R. K. O.; Freisinger, E.; Lippert, B. Chem.
Commun. 1998, 219. (c) Sigel, R. K. O.; Freisinger, E.; Metzger, S.;
Lippert, B. J. Am. Chem. Soc. 1998, 120, 12000. (d) Lippert, B. J.
Chem. Soc., Dalton Trans. 1997, 3971.
‡ Leiden University.
§ Freie Universita¨t Berlin.
| J. Heyrovsky Institute of Physical Chemistry, Academy of Sciences of
the Czech Republic.
Institute of Biophysics, Academy of Sciences of the Czech Republic,
and National Center for Biomolecular Research.
(1) Saenger, W. Principles of Nucleic Acid Structure; Springer: New York,
1984; pp 121-126.
10.1021/ic0109602 CCC: $22.00 © 2002 American Chemical Society
Published on Web 05/09/2002
Inorganic Chemistry, Vol. 41, No. 11, 2002 2855

Schmidt et al.
reinforces the Watson-Crick pair with cytosine (C) in
DMSO solution.6 Theoretical calculations (0 K, gas phase)
had predicted this feature of metal cations bonded at the
G-N7 site as a consequence of polarization effects.7
Interestingly, no such effects were found with adenine-N7
complexes of divalent cations (Zn2+, Cd2+, Hg2+, Mg2+,
Ca2+, Sr2+, Ba2+) applying high-level quantum mechanical
methods.8 Whether or not this result also holds up in solution
has not been studied as yet. This finding is noteworthy in
the context of A-N7 being the second most important site
of DNA bases for binding of the antitumor agent cisplatin9
and because of the considerable mutagenic potential of the
5′-ApG adduct of cisplatin with DNA.10,11
Of the other possible metal binding sites of adenine, that
is, N1,12 N3,13 and N6,14 the first one is of comparable
significance as N7, at least if the base is not involved in
Watson-Crick base pair formation (vide infra); hence, it is
free, or it may adopt Hoogsteen pairing. The other cases are
not considered here. Here, we report on our studies concern-
ing the H bonding interactions of a chloroform-soluble
platinum complex, with the metal entity fixed at N1, toward
thymine and uracil model nucleobases. Studies of this kind,
employing Pt(II) species, have proved impossible in the past
because of the insolubility of Pt(II) complexes in nonpolar
solvents on one hand, and the absence of H bonding between
A and T in polar solvents such as H2O, DMF, or DMSO on
the other. By using derivatives, carrying large organic entities
attached to the individual components to achieve sufficient
solubility,15 we have been able to circumvent these problems.
Experimental Section
Materials. 9-Cyclohexylmethyladenine (ChmA),16 1-cyclohexyl-
methylthymine (ChmTH),17 9-methyladenine (9-MeA),18 9-ethyl-
adenine (9-EtA),16 1-methylthymine (1-MeTH),19 2′,3′,5′-tri-tert-
butyldimethylsilyladenosine (TBDMS-ado),20 3′,5′-diacetyl-2′-deoxy-
uridine (and the [3-15N]-labeled compound),21 as well as trans-Pt-
22
(MeNH2)2Cl2 were prepared according to literature procedures.
trans-Pt(MeNH2)2(ChmT-N3)Cl (1). The compound was pre-
pared in a slightly modified version of that used for the corre-
sponding 1-MeT complex:23 To 0.502 g (1.531 mmol) of trans-
[Pt(MeNH2)2Cl2] in 30 mL of DMF was added 0.247 g (1.454
mmol) of AgNO3 and the reaction mixture stirred overnight. AgCl
was removed by filtration, and 0.4402 g (1.691 mmol) of the
potassium salt of 1-N-cyclohexylmethylthyminate was added to the
filtrate. After stirring at RT for 3 d, DMF was largely removed in
vacuo. Addition of diethyl ether led to isolation of 0.555 g (71%)
of analytically pure trans-Pt(MeNH2)2(ChmT-N3)Cl (1). Anal.
Calcd for C14H27N4O2ClPt: C, 32.7; H, 5.3; N, 10.9. Found: C,
32.9; H, 5.1; N, 11.0. ESI-MS: m/z 515 (M+). IR (cm-1): 3447.9
(b), 3271.9 (w), 3192.7 (m), 3106.5 (s), 2928.2 (s), 2851.8 (w),
1664.3 (vs), 1633.1 (vs), 1570.5 (vs), 1461.4 (s), 1438.1 (s), 1096.8
(m), 774.1 (m), 585.8 (w), 463.9 (w), 336.5 (w).
trans-[Pt(MeNH2)2(ChmT-N3)(ChmA-N1)]NO3 (2). It was
prepared by adding trans-[Pt(MeNH2)2(ChmT)(DMF)]NO3, ob-
tained in situ by reacting 1 with 0.95 equiv of AgNO3 in DMF
overnight at room temperature, to a solution of ChmA in DMF
and stirring the mixture for 24 h at room temperature. After rotary
evaporation of DMF, the crude product was recrystallized from
methanol/ether (20/1, v/v) to give analytically pure 2 in 30% yield.
Anal. Calcd for C26H44N10O5Pt: C, 40.5; H, 5.8; N, 18.2. Found:
C, 40.3; H, 5.6; N, 18.0. ESI-MS: m/z 709 (M+). IR (cm-1): 3379.1
(b), 3112.0 (s), 2925.9 (s), 2852.5 (m), 1660.3 (vs), 1634.2 (vs),
1557.8 (vs), 1539.3 (vs), 1534.9 (vs), 1307.4 (vs), 1104.8 (s), 827.3
(w), 776.9 (m), 719.0 (m), 652 (w).
(6) (a) Sigel, R. K. O.; Freisinger, E.; Lippert, B. J. Biol. Inorg. Chem.
2000, 5, 287. (b) Sigel, R. K. O.; Lippert, B. Chem. Commun. 1999,
2167.
(7) (a) Sponer, J.; Burda, J. V.; Sabat, M.; Leszczynski, J.; Hobza, P. J.
Phys. Chem. A 1998, 102, 5951. (b) Burda, J. V.; Sponer, J.;
Leszczynski, J.; Hobza, P. J. Phys. Chem. B 1997, 101, 9670. (c)
Anwander, E. H. S.; Probst, M. M.; Rode, B. M. Biopolymers 1990,
29, 757. (d) Hobza, P.; Sandorfy, C. J. Biomol. Struct. Dyn. 1985, 6,
1245. (e) Basch, H.; Krauss, M.; Stevens, W. J. J. Am. Chem. Soc.
1985, 107, 7267.
(8) Sponer, J.; Sabat, M.; Burda, J. V.; Leszczynski, J.; Hobza, P. J. Phys.
Chem. B 1999, 103, 2528.
(9) Fichtinger-Schepman, A. M. J.; van der Veer, J. L.; den Hartog, J. H.
J.; Lohman, P. H. M.; Reedijk, J. Biochemistry 1985, 24, 707.
(10) Burnouf, D.; Gauthier, C.; Chottard, J. C.; Fuchs, R. P. P. Proc. Natl.
Acad. Sci. U.S.A. 1990, 87, 6087.
(11) The insertion of A across the platinated 5′-base A might alternatively
be explained by an effect of the metal on the tautomer equilibrium.
See, e.g.: (a) Lippert, B.; Scho¨llhorn, H.; Thewalt, U. Inorg. Chim.
Acta 1992, 198-200, 723. (b) Burda, J. V.; Sponer, J.; Leszczynski,
J. J. Biol. Inorg. Chem. 2000, 5, 178.
(12) X-ray analysis, e.g.: (a) Schwarz, F.; Lippert, B.; Scho¨llhorn, H.;
Thewalt, U. Inorg. Chim. Acta 1990, 176, 113. (b) Arpalahti, J.; Klika,
K. D.; Sillanpa¨a¨, R.; Kiveka¨s, R. J. Chem. Soc., Dalton Trans. 1998,
1397.
trans-[Pt(MeNH2)2(ChmT-N3)(TBDMS-ado-N1)]BF4 (3). This
compound was prepared by adding trans-[Pt(MeNH2)2(ChmT-N3)-
(DMF)]BF4, obtained from 1 and AgBF4 in DMF, to a solution of
TBDMS-ado in DMF and stirring the mixture for 24 h at room
temperature. The workup of the solution was analogous to that of
2. The yield was 55%. Anal. Calcd for C42H82N9O6Si3BF4: C, 42.9;
H, 7.0; N, 10.7. Found: C, 42.3; H, 6.8; N, 10.7. ESI-MS: m/z
1088 (M+). IR (cm-1): 3273.8 (b), 2930.1 (s), 2857.5 (m), 1652.0
(s), 1567.8 (s), 1463.8 (m), 1253.4 (m), 1069.2 (vs), 836.1 (vs),
777.5 (vs), 670.1 (w), 648.8 (w).
Other reactions mentioned in the text were carried out on a NMR
scale only. N1 and N7 linkage isomers were assigned by means of
pH* dependence (pH* is uncorrected pH meter reading) of the
aromatic protons, and yields were estimated by integration of the
1
respective H NMR resonances.
(13) X-ray analysis, e.g.: Meiser, C.; Song, B.; Freisinger, E.; Peilert, M.;
Sigel, H.; Lippert, B. Chem.sEur. J. 1997, 3, 388.
(16) Nowick, J. S.; Chen, J. S.; Noronha, G. J. Am. Chem. Soc. 1993, 115,
7636.
(17) Browne, D. T.; Eisinger, J.; Leonard, N. J. J. Am. Chem. Soc. 1968,
90, 7302.
(18) Kru¨ger, G. Z. Physiol. Chem. 1893, 118, 153.
(19) Kistenmacher, T.; Rossi, M.; Caradonna, J. P.; Marzilli, L. G. AdV.
Mol. Relax. Interact. Processes 1979, 15, 119.
(20) Ogilvie, K. K. Can. J. Chem. 1973, 51, 3799.
(21) (a) Weisz, K.; Ja¨hnchen, J.; Limbach, H.-H. J. Am. Chem. Soc. 1997,
119, 6436. (b) Dunger, A.; Limbach, H.-H.; Weisz, K. Chem.sEur.
J. 1998, 4, 621.
(14) See, e.g.: (a) Clarke, M. J. J. Am. Chem. Soc. 1978, 100, 5068. (b)
Charland, J. P.; Phan Viet, M. T.; St.-Jacques, M.; Beauchamp, A. L.
J. Am. Chem. Soc. 1985, 107, 8202. (c) Lowe, G.; Vilaivan, T. J.
Chem. Soc., Perkin Trans. 1 1996, 1499. (d) Zamora, F.; Kunsman,
M.; Sabat, M.; Lippert, B. Inorg. Chem. 1997, 36, 1583. (e) Sponer,
J.; Sponer, J. E.; Gorb, L.; Leszczynski, J.; Lippert, B. J. Phys. Chem.
A 1999, 103, 11406. (f) Viljanen, J.; Klika, K. D.; Sillanpa¨a¨, R.;
Arpalahti, J. Inorg. Chem. 1999, 38, 4924. (g) Velders, A. H.; van
der Geest, B.; Kooijman, H.; Spek, A. L.; Haasnoot, J. G.; Reedijk, J.
Eur. J. Inorg. Chem. 2001, 369.
(22) Arpalahti, J.; Lippert, B.; Scho¨llhorn, H.; Thewalt, U. Inorg. Chim.
Acta 1988, 153, 45.
(23) Krizanovic, O.; Sabat, M.; Beyerle-Pfnu¨r, R.; Lippert, B. J. Am. Chem.
Soc. 1993, 115, 5538.
(15) See, e.g., 2,2′:6′,2′′-terpyridine RuII conjugate of T interacting with
2,3′-isopropylidene-adenosine: Constable, E. C.; Fallahpour, R.-A. J.
Chem. Soc., Dalton Trans. 1996, 2389.
2856 Inorganic Chemistry, Vol. 41, No. 11, 2002

Loss of Hoogsteen Pairing Ability
Mass Spectrometry. LC MS analysis was carried out on a PE
Sciex API 165 mass unit connected to a Jasco HPLC system. HPLC
analysis was performed using a Merck LiChrosphere 100 RP18
column (pore size 5 µm, diameter 4 mm, length 250 mm) and 10
mM aqueous ammonium acetate/acetonitrile (gradient: 0-50%
acetonitrile) buffer. Compound 2 (m/z 709 [M+], tr 22 min) was
the only observed product. ESI-MS spectra were recorded on a
Finnegan MAT 900 double-focusing instrument with an electrospray
interface.
metal has been described with the LANL2DZ pseudopotential and
basis set,29 while the 6-31G* valence basis set was employed on
the C, N, O, and H atoms.25 Geometry optimizations were carried
out with relaxation of all inter- and intramolecular parameters.
Interaction energies were computed on the optimized structures and
corrected for the basis set superposition error. Pairwise interaction
energy terms as well as the three body term have been derived
using the procedure given in refs 7a and 30.
NMR Spectroscopy. All NMR spectra except for the 1H-195Pt
HMQC experiment, which was performed on a Bruker 200 MHz
spectrometer, were recorded on a Bruker DPX-300 MHz spec-
trometer using a 5 mm multinucleus probe. A variable temperature
unit was used to keep the temperature stable. 1H NMR spectra were
acquired with normal single-pulse excitation, 45° flip angle, pulse-
recycle time of 6 s, and spectral width of 3.9 kHz consisting of
24K data points. For Fourier transformation, a size of 16K was
used. 1H NMR chemical shifts in ppm were referenced to the solvent
peak. One-dimensional carbon spectra were acquired with normal
single-pulse excitation, 45° flip angle, pulse-recycle time of 3 s,
and spectral widths of 18.8 kHz consisting of 32K data points. For
Fourier transformation, a size of 16K was used. One-dimensional
platinum spectra were acquired with normal single-pulse excitation,
70° flip angle, pulse-recycle time of 0.03 s, and spectral width of
148 kHz consisting of 4K data points. For Fourier transformation,
a size of 16K was used. 195Pt NMR chemical shifts were calibrated
using K2PtCl4 as an external reference at -1614 ppm relative to
K2PtCl6. A sweep width of 148 kHz was used. The NOESY, the
long-range COSY, and the 1H-195Pt HMQC spectra were all
measured in CD3OD.
Low-temperature NMR experiments in a Freon solvent were
performed on a Bruker AMX500 spectrometer. The deuterated
Freon mixture CDClF2/CDF3 was prepared as earlier described24
and handled on a vacuum line which was also used for the sample
preparation. Temperatures were adjusted by a Eurotherm variable
temperature unit to an accuracy of (1.0 °C. 1H chemical shifts in
the Freon mixture were referenced relative to CHClF2 (δH ) 7.13
ppm). A 2D NOE experiment at 128K was recorded with a mixing
time of 60 ms and a pulse-recycle time of 1.2 s. A total of 900
FIDs of 2K complex data points were collected, and the t1 and t2
FIDs zero-filled to give a 2K‚1K data set prior to Fourier
transformation.
Results and Discussion
Metal Binding Sites at Adenine. Established metal
coordination patterns of adenine nucleobases have been
reviewed.31 In the following, only metal binding to two (N1,
N7) of the three (N1, N7, N3) kinetically favored endocyclic
sites of the adenine ring will be discussed. Metal binding to
the deprotonated N6 position, either alone14 or in conjunction
with N1 or N7,31 will not be considered here, although it is
the thermodynamically preferred site, and migration from
N7 to N6,14f as well as from N1 to N6,32 has been observed.
From model chemistry, it is known that the binding behavior
of Pt(II) is influenced by several factors. Thus, at acidic pH,
the most basic site N1 (pKa ∼ 4) is protonated, and
consequently, the metal is directed to N7. In the pH range,
where both N1 and N7 sites are unprotonated (pH > 6),
usually a mixture of linkage isomers is formed. Reasons for
the distribution patterns between the two sites are poorly
understood. The amino group exercises a steric hindrance
for incoming metal electrophiles, which is larger for metals
approaching N1 as compared to N7.33 Similarly, ancillary
ligands already bonded to the metal have an effect on the
distribution between N1 and N7 sites. For example, cis-[Pt-
(NH3)2(1-MeC-N3)(H2O)]2+,34 when reacted with the model
adenine base 9-MeA34 in water, preferentially leads to the
N7 cross-linking product, whereas the corresponding trans
isomer prefers N1 over N7.35 A preference of N1 over N7
has also been reported for reactions of trans-[Pt(NH3)2(1-
MeT-N3)(H2O)]2+ in water,23 and for the dipeptide complex
[Pd(glyhis)(H2O)]+34 in the same solvent.36 On the other
hand, [Pt(dien)(H2O)]2+, although still favoring N1 in the
case of adenosine, distributes evenly between N1 and N7 of
5′-AMP at pH 7-9.37 As demonstrated by Arpalahti et al.,12b
Computational Part. Computations have been carried out at
DFT (density functional theory) level of theory as implemented in
the Gaussian 98 computer code.25 Similarly to our previous
contributions14e,26 published on platinated nucleobase complexes,
we have chosen a combination of Becke’s three-parameter hybrid
functional27 with Lee-Yang-Parr’s exchange functional.28 The
(27) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(28) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. 1998, B37, 785.
(29) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299.
(30) Sˇponer, J.; Burda, J. V.; Leszczynski, J.; Hobza, P. J. Biomol. Struct.
Dyn. 1999, 17, 61.
(31) Lippert, B. Coord. Chem. ReV. 2000, 200-202, 482.
(32) Arpalahti, J.; Klika, K. D. Eur. J. Inorg. Chem. 1999, 1199.
(33) (a) Martin, R. B. Met. Ions Biol. Syst. 1996, 32, 61. (b) Arpalahti, J.;
Lonnberg, H. Inorg. Chim. Acta 1983, 78, 63. (c) Reily, M. D.;
Marzilli, L. G. J. Am. Chem. Soc. 1986, 108, 6785. (d) Arpalahti, J.;
Lehikoinen, P. Inorg. Chim. Acta 1989, 159, 115.
(34) Abbreviations used: 1-MeC ) 1-methylcytosine; 9-MeA ) 9-me-
thyladenine; 9-EtA ) 9-ethyladenine; 1-MeT ) 1-methylthymine
anion; ChmT ) N1-cyclohexylmethylthymine anion; ChmA ) 9-cy-
clohexylmethyladenine; ChmAH ) 9-cyclohexyl-methyladeninium;
ado ) adenosine; T ) thymine (general); A ) adenine (general); glyhis
) glycylhistidine; DMF ) N, N′-dimethylformamide; dien ) dieth-
ylenetriamine.
(24) Golubev, N. S.; Smirnov, S. N.; Gindin, V. A.; Denisov, G. S.;
Benedict, H.; Limbach, H.-H. J. Am. Chem. Soc. 1994, 116, 12055.
(25) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels,
A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone,
V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.;
Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Baboul, A. G.; Stefanov,
B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts,
R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C.
Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.;
Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98; Gaussian,
Inc., Pittsburgh, PA, 1998.
(35) Beyerle-Pfnu¨r, R.; Brown, B.; Faggiani, R.; Lippert B.; Lock, C. J. L.
Inorg. Chem. 1985, 24, 4001.
(26) Sˇponer, J. E.; Glahe´, F.; Leszczynski, J.; Lippert, B.; Sˇponer, J. Inorg.
(36) Wienken, M.; Zangrando, E.; Randaccio, L.; Menzer, S.; Lippert, B.
J. Chem. Soc., Dalton Trans. 1993, 3349.
Chem. 2001, 40, 3269.
Inorganic Chemistry, Vol. 41, No. 11, 2002 2857

Schmidt et al.
Chart 1
high temperature likewise has an effect on the species
distribution, thereby favoring N1.
Formation of Mixed A,T Complexes of Pt(II) in DMF.
Reactions of trans-[Pta2(1-MeT-N3)(H2O)]+ (a ) NH3 or
MeNH2) with 9-MeA in water have previously been de-
scribed.23 They lead to mixtures of 9-MeA-N1 and 9-MeA-
N7 linkage isomers. Representative species have been
characterized by X-ray crystallography.23
Analogous reactions carried out in DMF, hence between
trans-[Pta2(1-MeT-N3)(DMF)]+ and 9-MeA, are hampered
by the insufficient solubility of 9-MeA in this solvent,
yielding mainly the dinuclear µ-9-MeA complex [a2(1-MeT-
N3)Pt(N1-9-MeA-N7)Pta2(1-MeT-N3)]2+. However, reaction
of trans-[Pt(NH3)2(1-MeT-N3)(DMF)]+ with 1 equiv of
9-EtA and 1 equiv of 1-ChmA, respectively, gives the N1
linkage isomer in 80% and 82% yield. The second species
formed is the corresponding N7 linkage isomer in both
instances (20% and 18% yield), respectively.
Figure 1. NOESY spectrum (MeOD) of 2. Cross-peaks are as follows:
a, A-H2/Pt-NH2CH3; b, T-H6/N1-CH2; c, A-H2/Pt-NH2CH3; d: A-H8/N9-
CH2.
A-H2 and A-H8 protons and the identification of the metal
binding sites are accomplished by a combination of NMR
experiments which included 195Pt-1H coupling, 2D NMR
spectra, and pH*-dependent spectra. For example, A-H8 in
2 is identified by its cross-peak with the 9-CH2 resonance
1
of the cyclohexylmethyl substituent in the H-1H NOESY
spectrum (Figure 1), while in a 1H-195Pt HMQC experiment,
3
A-H2 is unambiguously identified by a J coupling of ∼20
The most regioselective platination reactions in DMF are
observed between trans-[Pt(MeNH2)(ChmT-N3)(DMF)]+
and ChmA as well as TBDMS-ado, giving the N1 product
in over 90% yield. Formation of trans-[Pt(MeNH2)2(ChmT-
N3)(DMF)]+ from the corresponding chloro complex 1 by
Cl- abstraction with Ag+ is confirmed by 195Pt NMR
spectroscopy, which reveals a single peak at -2100 ppm,
consistent with an N3O coordination sphere.38 Reactions with
the two adenine derivatives are complete within 25 h at room
temperature. Elemental analysis and ESI mass spectrometry
reveal the compounds to be of 1:1 stoichiometry, and NMR
spectroscopy proves N1 binding (vide infra), indicating a
composition of trans-[Pt(MeNH2)2(ChmT-N3)(ChmA-N1)]-
[NO3] (2) and trans-[Pt(MeNH2)(ChmT-N3)(TBDMS-ado-
N1)][BF4] (3), respectively (Chart 1). Minor components of
these reactions are the 2:1 complexes with [Pt(MeNH2)2-
(ChmT)]+ entities bound to N1 and N7 as confirmed by ESI
Hz with the 195Pt resonance (-2584 ppm). This assignment
is also confirmed by pH*-dependent spectra of the aromatic
protons of 2 in D2O (Supporting Information). The pKa ∼ 1
for protonation of 2 is consistent with a N1 linkage isomer
1
being present.23 Finally, long-range H-13C COSY experi-
ments reveal the expected coupling of A-H2 with A-C2 (1J),
A-C4 (3J), and A-C6 (3J), as well as of A-H8 with A-C8
(1J), A-C4(3J), A-C5 (3J), and A-C9 (3J). The assignment of
the 13C NMR resonances of ChmA is based on literature
data for 9-MeA,39 as well as for [Pt(dien))(ado-N1)]2+.12b
1
The H NMR chemical shifts (D2O, 3 < pH* < 11) of
H2 and H8 of the ChmA ligand in 2 agree well with those
of 9-MeA in trans-[Pt(MeNH2)2(1-MeT-N3)(9-MeA-N1)]+
in the same solvent.23 As compared to neutral 9-MeA, these
resonances are shifted downfield by 0.29 ppm (H8) and 0.95
ppm (H2), respectively. Downfield shifts of these protons
in 2, relative to the free adenine ligand, are also observed in
DMF-d7 (∆δ H2, 0.66; H8, 0.29) and CD3OD (∆δ H2, 0.81;
H8, 0.11). In contrast, in CDCl3, the H8 resonance of 2 is
surprisingly shifted upfield, by δ -0.11, but H2 is again
shifted to lower field (δ 0.32).
1
MS and H NMR. Compound 2 is not found to isomerize
within several days at 70 °C in a MeOH/H2O mixture
(1/1, v/v).
1
NMR Characterization of 2 and 3. H, 13C, and 195Pt
NMR chemical shifts of 2 and 3 as well as of the free adenine
ligands ChmA and TBDMS-ado in different solvents are
provided in the Supporting Information. The assignment of
A-H8 in 3 is identified by its cross-peak with the H1′
proton of the ribose moiety in the 1H-1H NOESY spectrum.
A long-range 1H-13C COSY experiment reveals the expected
coupling of A-H2 with A-C2 (1J), A-C4 (3J), and A-C6 (3J),
(37) Scheller, K.-H.; Scheller-Krattiger, V.; Martin, R. B. J. Am. Chem.
Soc. 1981, 103, 6833.
(38) Pregosin, P. S. In Annual Reports on NMR spectroscopy; Webb, G.
A., Ed.; Academic Press: San Diego, 1986; Vol. 17, p 285.
(39) Chenon, M.-T.; Pugmire, R. J.; Grant, D. M.; Panzica, R. P.; Townsend,
L. B. J. Am. Chem. Soc. 1975, 97, 4627.
2858 Inorganic Chemistry, Vol. 41, No. 11, 2002

Loss of Hoogsteen Pairing Ability
1
Figure 2. Temperature-dependent H NMR spectra of 2 in D2O.
as well as of A-H8 with A-C8 (1J), A-C4(3J), A-C5 (3J),
and A-C9 (3J). The assignment of the 13C NMR resonances
of the TBDMS-ado ligand is again based on literature data
for 9-MeA, as well as for [Pt(dien)(ado-N1)]2+.12b
No measurable self-association of 2 is observed in CDCl3,
1
as evidenced by the concentration independence of its H
NMR resonances in this solvent.
1
Temperature Dependence of H NMR Spectra. Tem-
1
perature-dependent H NMR spectra of 2 were recorded in
D2O (Figure 2), CDCl3, and DMF-d7 (Figure 3). In D2O,
A-H2 occurs as two resonances (1:1) up to ∼318 K,
irrespective of pH*. Above this temperature, coalescence
takes place (Figure 2). This process is fully reversible. A
similar splitting of A-H2 has been reported by some of us23
for trans-[Pt(MeNH2)2(1-MeT-N3)(9-MeA-N1)]+ in D2O. As
in this case, we assign this behavior to hindered rotation of
the nucleobases about the Pt-N bonds. It appears that
hindered rotation depends on the presence of methylamine
ligands at the Pt (the NH3 analogue does not display this
phenomenon23) and the solvent water. The role of water
could be to bridge exocyclic groups of the two trans
positioned bases, as seen in the X-ray crystal structure23 of
trans-[Pt(MeNH2)2(1-MeT-N3)(9-MeA-N1)]ClO4‚3.25 H2O.
While direct H bonding between exocyclic groups of the two
bases is excluded on steric grounds (distance ∼4 Å), the dual
function of a water molecule to act as a H bond acceptor for
N(6)H2 and a H donor for T-O4 and T-O2 is ideal for this
purpose. The rate of rotation at the coalescence point can
be estimated40 as kc ) 2.22∆ν ) 27.3 s-1.
1
Figure 3. Temperature-dependent H NMR spectra of 2 in DMF-d7.
displays coalescence at ∼310 K before splitting into two
components at ∼8.6 and ∼9.2 ppm (280 K). Both resonances
undergo further downfield shifts with decreasing temperature,
but the effect on the lowfield component is more pronounced.
We assign the signal splitting to a freezing of the exocyclic
amino group rotation. (ii) Below 270 K, the highfield
component starts to split further (1:1), while the lowfield
component does so at lower temperature only, namely in
the temperature range <241 K, when the A-H2 signal starts
to split. At 220 K, the A-NH2 resonance consists of four
components of equal intensities, occurring at δ ) 9.72, 9.65,
8.90, and 8.75 (Figure 3). Because splitting of the lowfield
A-NH2 resonance coincides with splitting of A-H2, we assign
this process (below 241 K) to a freezing of the nucleobase
rotation. Consequently, we attribute the A-NH2 components
above δ 9 to the amino proton which is syn to N(1) in the
two possible rotamers, with O4 and O2 of the thymine base
across this proton. (iii) It is not immediately evident why
Consistent with this interpretation, hindered nucleobase
rotation as evident from A-H2 splitting is seen in DMF-d7
at lower temperature only, namely below ∼241 K (vide infra,
Figure 3). In DMF-d7, altogether three dynamic processes
are observed, which involve the N(6)H2 resonance. (i) A
single resonance at ambient temperature (δ = 8.7), this signal
(40) Friebolin, H. Basic One and Two-dimensional NMR Spectroscopy;
VCH: Weinheim, 1993.
Inorganic Chemistry, Vol. 41, No. 11, 2002 2859

Schmidt et al.
observation from the solution NMR studies, that there is
virtually no hydrogen bonding of the Hoogsteen type between
N1 platinated A and free thymine. Thus, we have investigated
molecular interactions in several complexes. Their molecular
structures were first optimized using gradient optimization,
and for the final structures, we have evaluated the interaction
energies characterizing the intrinsic binding strength of the
base pairs in a complete isolation (in vacuo) at 0 K.
Interaction energy of a dimer is defined as the energy
difference between the dimer and the two monomers when
they are separated; negative values of interaction energy
mean attraction.
Interaction (base pairing) energies were first calculated for
a series of models with the following results: (i) adenine,
thymine Hoogsteen pair, -14.1 kcal/mol; (ii) thymine,
thymine self-pair, -11.0 kcal/mol; (iii) Hoogsteen pair
between trans-[a2Pt(A-N1)T]+ and thymine (hereafter re-
ferred to as complex I), -18.8 kcal/mol. These results
suggest that N1 platination of adenine should, in principle,
enhance Hoogsteen pairing with thymine. Metal-induced
strengthening of this base pairing, however, is caused
primarily by long-range electrostatic attraction between the
metal cation and thymine. This can be highlighted by
calculating the so-called three body term. The thymine‚‚‚
Aa2PtT base pairing energy was decomposed into two
pairwise terms (Thymine‚‚‚A and Thymine‚‚‚a2PtT) and the
three-body term. The three-body term shows how A and a2-
PtT subsystems mutually influence their interaction with the
free thymine; thus, the three-body term primarily includes
the polarization effects. For the present complex I, the
calculated polarization (three-body) term is negligible, around
-1.3 kcal/mol.
To further understand the molecular interaction in the
complex studied, we have evaluated the strength of the base
pairing by reduction of the effective charge of the metal entity
by the “addition” of anions. This strongly reduces the
absolute values of interaction energies of thymine with the
rest of the system. Thus, artificial addition of a hydroxide
ion in a nearly axial position to the Pt(II) center (not shown)
leads to an interaction energy of -14.2 kcal/mol. Replace-
ment of an ammonia ligand at the Pt by a hydroxide (which
results in formation of an uncharged complex) further reduces
this value to -12.1 kcal/mol. The largest decrease is
observed, however, if a nitrate counterion is placed in the
vicinity of the exocyclic amino group of adenine (Figure 5).
In the fully optimized structure of this complex, the nitrate
cation displays a Pt‚‚‚N interatomic distance of 3.56 Å. The
nitrate anion forms a close contact with the exocyclic amino
group of adenine as well as with one of the ammonia ligands
of Pt(II) (interatomic H bonding distances are 2.72 and 2.79
Å, respectively). The interaction energy between thymine
and the rest of the complex (i.e., the platinated adenine plus
the nitrate ion) is only -11.6 kcal/mol, close to the value
for thymine, thymine self-pairing and well below the stability
of the nonmetalated Hoogsteen adenine, thymine base pair.
The close contact of the nitrate anion with complex I
overcorrects the base pairing energy gains that would be
caused by the metal cation itself. It is to be noted that the
Figure 4. Concentration-dependent 1H NMR shifts (T-NH, A-NH2) of
an equimolar mixture of 2 and ChmT in CDCl3.
the upfield component of the A-N(6)H2 resonance, observed
in the δ 8.5-9 range, is split already below 270 K.
1
Concentration-Dependent H NMR Spectra of Mix-
tures of 2 and ChmT in CDCl3. To study the extent of
hydrogen bonding between 2 and ChmT, concentration-
1
dependent H NMR measurements of equimolar mixtures
of the two components in CDCl3 at ambient temperature were
carried out (Figure 4). Downfield shifts of other protons with
increasing concentration are not observed, except for the
N(3)H proton resonance of the free ChmT. Specifically, the
shift of the amino resonance of the ChmA ligand of 2 (upfield
shift of 0.09 ppm) is in fact opposite to what would be
expected for intermolecular H bond formation. From the
concentration dependence of the imino proton of the thymine
base, an association constant has been estimated,41 which
lies in the same order of magnitude as for the self-association
of ChmT in this solvent (4.72 ( 0.23 M-1).42 This implies
that the concentration dependence of the N(3)H resonance
of ChmT is primarily due to self-association, rather than to
Hoogsteen pairing with the adenine ligand in 2. This
conclusion was further substantiated by a NOESY experi-
ment (data not shown), which did not reveal any cross-peak
between H8 of ChmA and N(3)H of ChmT. In contrast, the
NOESY spectrum of a 1:1 mixture of ChmA and ChmT in
CDCl3 at 20 °C displays cross-peaks between A-H8 and
T-NH, as well as between A-H2 and T-NH, indicative of
Watson-Crick and Hoogsteen pairing coexisting under these
conditions. The association constant (macro constant for all
possible H bonding patterns) between A and T has previously
been determined by us to be K ) 56.5 ( 9 M-1 using the
same method.42 This value is in good agreement with
literature data obtained from 13C NMR spectroscopy (K )
60 ( 5 and 73 ( 4 M-1)43 and IR spectroscopy (K ) 130
M-1),44 respectively, for slightly different nucleobase deriva-
tives.
Theoretical Calculations. Ab initio calculations were
carried out in order to understand the rather unexpected
(41) Sigel, R. K. O. Unpublished results.
(42) Schmidt, K. S.; Sigel, R. K. O.; Filippov, D. V.; van der Marel, G.
A.; Lippert, B.; Reedijk, J. New J. Chem. 2000, 24, 195.
(43) Iwahashi, H.; Kyogoku, Y. J. Am. Chem. Soc. 1977, 99, 7761.
(44) Kyogoku, Y.; Lord, R. C.; Rich, A. Proc. Natl. Acad. Sci. U.S.A. 1967,
57, 250.
2860 Inorganic Chemistry, Vol. 41, No. 11, 2002

Loss of Hoogsteen Pairing Ability
Figure 5. Becke 3LYP optimized structure of the Hoogsteen pair between
trans-[Pt(NH3)2(T-N3)(A-N1)]+T (T ) thyminate anion, A ) adenine) and
-
neutral thymine in the presence of NO3-. The NO3 is hydrogen bonded
simultaneously to an ammonia ligand of Pt and the NH2 group of adenine.
Although Hoogsteen pairing is possible on steric grounds, the interaction
energy is very much reduced as a consequence of charge neutralization by
the nitrate anion.
position of the nitrate anion does not impede Hoogsteen
pairing with thymine for steric reasons (see Figure 5): The
6-amino group proton syn to N7 is, in principle, still available
for base pairing. The optimized interatomic distance mea-
sured between N6 of adenine and O4 of thymine is the same
(2.98 Å) as in the isolated adenine, thymine Hoogsteen pair.
These results tentatively suggest that ion pair formation
between cationic complex I and the nitrate anion, hence
charge neutralization, is primarily responsible for the non-
existence of Hoogsteen-type H bonds between T and
platinated A in chloroform solution. To further support this
assumption, we have compared the electron densities in the
plane of the N6-H6‚‚‚O4 H-bond of the (i) AT Hoogsteen
pair, (ii) complex I, and (iii) complex I with a nitrate coanion.
It has been found that N1 platination of the adenine results
in an extensive electron density expansion along the N6-
H6‚‚‚O4 H bond (cf. Figure 6a,b), thereby enhancing the
stability of the (Pt-A)T base pair. On the other hand,
addition of a nitrate anion to the platinated complex I leads
to a partial segregation of the electron density on the N6-
H6 and C4-O4 bonds and therefore to a similar profile as
for the nonmetalated adenine, thymine pair (cf. Figure 6a,c).
1
Low-Temperature H NMR Spectra in CDClF2/CDF3
Solvent Mixtures. An alternative method to study hydrogen
bonding interactions is low-temperature NMR spectroscopy.
In a series of studies, the usefulness of deuterated Freon
mixtures as solvents for H bonded associates between
nucleobases has been demonstrated.4,21 Such low-melting
solvents allow measurements in the liquid state down to 90
K and thus permit the observation and characterization of
individual nucleobase associates in slow exchange on the
NMR time scale. Employing specifically 15N-labeled nucleo-
1
bases, additional information such as J scalar couplings
between 15N and the proton in a hydrogen bridge can be
obtained and related to the geometry of the hydrogen bonds.
In the present study, low-temperature 1H NMR spectra of
a 2:1 mixture of 3′,5′-diacetyl-2′-deoxyuridine or its [3-15N]
labeled analogue with the well soluble complex 3 were
recorded in a Freon solvent. Because uridine binding to 3 is
not expected to significantly differ from the thymine base
with a methyl group at the 5-position, the uridine nucleoside
was employed to allow a more rigorous comparison with
low-temperature data on nonplatinated adenosine-uridine
Figure 6. Comparison of calculated in-plane electron densities of (a)
adenine, thymine Hoogsteen pair, (b) complex I, and (c) complex I with
-
NO3 anion included.
association.4 Moreover, as a result of a relatively high
dielectric constant of the medium at these low temperatures,
Inorganic Chemistry, Vol. 41, No. 11, 2002 2861

Schmidt et al.
1
Figure 7. Sections of temperature-dependent H NMR spectra in Freon
showing the imino proton resonances of a 1:2 mixture of 3 and 3′,5′-diacetyl-
2′-deoxyuridine.
a good solvation of the BF4- counterion is expected.45 Thus,
hydrogen-bonding interactions between 3 and 3′,5′-diacetyl-
2′-deoxyuridine are not expected to be influenced by the
presence of the counterion. As seen from Figure 7, various
uridine N(3)H imino resonances in slow exchange appear
in the lowfield portion of the spectrum on lowering the
temperature. The broad upfield shifted signals centered at
about 12 ppm at 133 K correspond to uridine homodimers,21
whereas the downfield shifted resonances Ha-d at 14.41,
14.01, 13.58, and 13.51 ppm must arise from adducts of the
platinated adenosine 3 and uridine. Clearly, and in contrast
to Ha and Hd which broaden out above 133 K, the Hb and
Hc resonances at 14.01 and 13.58 ppm remain in slow
exchange up to 153 K, indicating a higher kinetic stability
of the latter. Integration of the various resonances suggests
that approximately 40% of the uridine is associated with 3,
whereas the remaining 60% are bound in self-associates.
With regard to the fact that the mixture contains two equiv
of 3′,5′-diacetyl-2′-deoxyuridine, this implies that approxi-
mately only 80% of 3 are associated with 3 and that also
free, non-hydrogen-bonded 3′,5′-diacetyl-2′-deoxyuridine is
present.
To further characterize the various coexisting species, a
2D NOE spectrum of the mixture was recorded at 128 K.
As shown in Figure 8, both Hb and Hc imino protons exhibit
NOE contacts to adenine amino and H8 protons, which are
identified by their mutual cross-peaks and cross-peaks to
adenosine sugar protons, respectively (not shown). This
points to their presence in two different Hoogsteen geom-
etries. Whereas no cross-peak to Hd is detected, the Ha imino
proton only exhibits an NOE contact to another signal which,
based on the absence of cross-peaks to sugar protons, is
tentatively assigned to an adenine H2 resonance.
Figure 8. Portion of a 2D NOE spectrum of a mixture of 3 and 3′,5′-
diacetyl-2′-deoxyuridine (1:2) in Freon at 128 K showing imino proton cross-
peaks.
may also be associated with the splitting of the A amino
resonance anti to N(1) occurring below 270 K in DMF (vide
supra). The kinetically more labile Ha and Hd imino protons
only observed at very low temperatures probably represent
complexes with only one hydrogen bond involving adenine
N3, or the 2- or 4-carbonyl group of the ChmT ligand.
Whereas participation of the adenine N3 position as proton
acceptor in a hydrogen bond with the uridine base is
restricted for steric reasons, the hydrogen-bond acceptor
capability of T-O2 and T-O4 is expected to increase upon
platination at N3 and would thus account for the low-field-
shifted imino resonances when compared to the uridine
homodimers.
Employing [3-15N] labeled uridine, a 1J(15N-1H) coupling
constant of 87.6 Hz for both of the two predominant
Hoogsteen complexes has been determined at low temper-
atures. Recently, in a nonplatinated A-U mixture, a value
of 84 Hz has been measured for the A-U Hoogsteen pair
in the same solvent mixture.4 Obviously, in the platinated
system, the imino proton is located closer to the uridine N3
donor in a weaker hydrogen bond. Accordingly, it is also
interesting to note that a H8 resonance in the low-temperature
2D NOE spectrum shows no cross-peak to a U imino proton
and demonstrates the presence of 3 with unoccupied Hoogs-
teen sites even with an excess of uridine.
Relevance and Conclusions
Hydrogen bonding interactions between two nucleobases
can be affected by a metal entity attached to the periphery
of one of the two bases, causing a strengthening or
weakening of H bonding or an alternation of the specificity
of base pairing. Biological consequences may range from
DNA stabilization to DNA destabilization and mispair
formation. Among others, these questions are also relevant
to DNA adducts of cisplatin, for example, to distinguish steric
effects (e.g., DNA distortion) from electrostatic and polariza-
The two heterocomplexes with Hoogsteen bound imino
protons Hb and Hc are likely to arise from normal and reverse
Hoogsteen geometries or from the two rotameric structures,
as a result of hindered rotation about the N-Pt bonds, and
(45) Golubev, N. S.; Shenderovich, I. G.; Smirnov, S. N.; Denisov, G. S.;
Limbach, H.-H. Chem.sEur. J. 1999, 5, 492.
2862 Inorganic Chemistry, Vol. 41, No. 11, 2002

Loss of Hoogsteen Pairing Ability
tion ones (e.g., effect on base pair formation). The general
problem of assessing these questions stems from a frequent
incompatibiliy of complex solubility and solvent allowing
the observation of hydrogen bond formation. For example,
the large majority of synthetic Pt(II) nucleobase complexes
is water-soluble, yet base pairing in water is generally poor
because of competition with the solvent and therefore cannot
be studied experimentally. Only with cationic Pt(II) guanine-
N7 complexes has the Watson-Crick pairing with cytosine
been studied in Me2SO.6 In the present paper, an obvious
way out of this dilemma has been presented, namely the use
of large hydrophobic residues at the nucleobases to achieve
solubility in chloroform and Freon. As examples, adenine
complexes carrying trans-a2Pt(II) entities at the N1 positions
(as well as an anionic thymine base at the fourth coordination
site) have been prepared and their H bonding interactions
with thymine and uridine studied.46 Surprisingly, no H-bond
formation whatsoever is observed between N1-platinated
adenine (2) and free thymine in CDCl3. In Freon, H bonding
between 3 and a uridine derivative is observed at very low
temperature only. No conclusions concerning its existence
In summary, the findings in Freon strongly suggest that
Pt(II) coordination to N1 of A leads to an inherently weaker
Hoogsteen/reversed Hoogsteen pair. Whether the complete
absence of heterobase pairing between the A ligand in 2 and
T in CDCl3 is to be solely explained on this basis remains
questionable. Ab initio calculations suggest that, at least in
the gas phase, Pt(II) coordination to A-N1 enhances pairing,
essentially on the basis of favorable long-range electrostatic
attraction between the Pt(II) cation and T. At the same time,
these calculations reveal that any charge neutralization of
the metal entity drastically reduces interbase association. In
other words, ion pairing, as expected in poorly solvating
chloroform, could effectively suppress interbase hydrogen
bond formation under the present experimental conditions,
despite the fact that the platination itself would rather
stabilize H bonding, as suggested by the gas-phase QM data.
Acknowledgment. This work was supported by a “DAAD
Doktorandenstipendium im Rahmen des gemeinsamen Hoch-
schulsonderprogramms III von Bund und La¨ndern” and
performed in the frame of the BIOMED project BMHA-
CT97-2485 and COST D20. Furthermore, the study has been
supported by the following grants: LN00A016, MSMT CR
(J.S.), VW-Stiftung I/74657 (J.E.S., B.L., J.S.), A4040903,
IGA AS CR (J.E.S.), DFG, and FCI (B.L.) The calculations
were carried out in the Supercomputer Center, Brno.
1
at higher temperatures can be made. J(15N-1H) coupling
of [3-15N] labeled uridine in the adducts with 3 is indicative
of a weaker association as compared to the A-U Hoogsteen
pair in the absence of coordinated Pt(II).
Supporting Information Available: 1H and 13C NMR data (5
tables) and pH* dependence of H2 and H8 protons of adenine in
2. This material is available free of charge via the Internet at
http://pubs.acs.org.
(46) We do not wish to claim that the model compounds are relevant to
the biological chemistry of trans-a2PtCl2, even though A, T Hoogsteen
pairing is possible in (distorted) DNA (See, e.g.: Quigley, G. J.;
Ughetto, G.; van der Marel, G. A.; van Boom, J. H.; Wang, A. H.-J.;
Rich, A. Science 1986, 232, 1255.), and in principle, the N1 site of A
could become metalated.
IC0109602
Inorganic Chemistry, Vol. 41, No. 11, 2002 2863