DNA modification by cis-diaminoplatinum(II) complexes with aminonitroxide ligands

Article abstract of DOI:10.1023/A:1019686323902

The mixed-ligand complexes cis-pt^II[R(CH₂) _nNH₂](NH₃)X₂, where R = 2,2,6,6-tetramethyl-1-oxylpiperidin-4-yl, X₂ = ClI or Cl ₂, n = 1 or 2, and binuclear complexes trans-3,4-bis[cis-ammine(iodochloro or dichloro)platinum(II)amino]-2,2,6,6-tetramethylpiperidin-1-oxyls were synthesized. The reactivity of the aminonitroxide complexes toward DNA, the destabilizing effect of the adducts on DNA structure, and the distribution of the Pt adducts along the DNA duplex were studied. The platination activity of the complexes is affected by the natures of both the leaving ligands X and the carrying amino ligands. The decrease in the platination activity of the complexes with an increase in the amino ligand sizes is probably caused by steric hindrance. The complexes that effectively platinate isolated DNA and cause a moderate destabilizition of DNA duplex possess high antitumor activities.

Full text of DOI:10.1023/A:1019686323902

1058
Russian Chemical Bulletin, International Edition, Vol. 51, No. 6, pp. 1058—1064, June, 2002
Chemistry of Natural Compounds,
Bioorganic, and Biomolecular Chemistry
DNA modification by cisꢀdiaminoplatinum(II)
complexes with aminonitroxide ligands
ꢀ
V. D. Sen´, A. V. Shugalii, and A. V. Kulikov
Institute of Problems of Chemical Physics, Russian Academy of Sciences,
142432 Chernogolovka, Moscow Region, Russian Federation.
Fax: +7 (096) 514 3244. Eꢀmаil: senvd@icр.аc.ru
The mixedꢀligand complexes cisꢀPtII[R(CH2)nNH2](NH3)X2, where R = 2,2,6,6ꢀtetraꢀ
methylꢀ1ꢀoxylpiperidinꢀ4ꢀyl, X2 = ClI or Cl2, n = 1 or 2, and binuclear complexes
transꢀ3,4ꢀbis[cisꢀammine(iodochloro or dichloro)platinum(II)amino]ꢀ2,2,6,6ꢀtetramethylꢀ
piperidinꢀ1ꢀoxyls were synthesized. The reactivity of the aminonitroxide complexes toward
DNA, the destabilizing effect of the adducts on DNA structure, and the distribution of the Pt
adducts along the DNA duplex were studied. The platination activity of the complexes is
affected by the natures of both the leaving ligands X and the carrying amino ligands. The
decrease in the platination activity of the complexes with an increase in the amino ligand
sizes is probably caused by steric hindrance. The complexes that effectively platinate isolated
DNA and cause a moderate destabilizition of DNA duplex possess high antitumor activities.
Key words: platinum(II) complexes, nitroxides, spin labels, DNA, platination, ESR specꢀ
troscopy, Cisplatin.
The key process in the inhibition of tumor growth by
cisꢀdiaminoplatinum(II) complexes is platination of DNA,
which violates its replication.1 The platination of DNA
has been studied in detail by various methods, mainly,
using Cisplatin.2 Recently, we synthesized complexes 1,
2, 30 (i.e., n = 0) and 4 with aminonitroxide ligands (L)
and showed that some of them compare well with
Cisplatin in antitumor activity3—5 and are less toxic, preꢀ
sumably, due to the antioxidant properties of nitroxide
radicals.5 Since these complexes are paramagnetic, DNA
platination can be studied by ESR, as we have done
previously4,6 for complexes 1 and 4. The relationship
between the structure and the antitumor activity of
diaminoplatinum complexes is rather complicated and
depends on many factors.7 However, it is clear that a
necessary condition is efficient binding of the complex
to DNA in a cell.
The ligands X in diamino PtII complexes tend to be
removed rather easily upon hydrolysis and/or replaceꢀ
ment by nucleophilic groups of biomolecules. For this
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 6, pp. 972—978, June, 2002.
1066ꢀ5285/02/5106ꢀ1058 $27.00 © 2002 Plenum Publishing Corporation

Platinum(II) complexes with aminonitroxides
Russ.Chem.Bull., Int.Ed., Vol. 51, No. 6, June, 2002
1059
bility of the nitroxide radicals in the LPtII—DNA adꢀ
ducts, and the distances between them. The degree of
DNA destabilization due to the formation of adducts by
complexes of various structures was studied by UV specꢀ
troscopy. These data could be useful for elucidating the
relationship between the structure and activity of platiꢀ
num complexes. In order to extend the structural series
of the complexes, new mixedꢀligand complexes 31, 32
(i.e., n = 1, 2), and binuclear complex 5 were prepared.
Experimental
The platinum content in the complexes was determined by
atomicꢀabsorption spectroscopy using an AASꢀ3 spectrometer
with an error of 3 rel. %. IR spectra were recorded in the
400—4000 cm–1 range on a Specord IRꢀ75 spectrometer in
mineral oil, and ESR spectra were run at room temperature
and at 77 К on an SE/X 2544 instrument with a microwave
power of 0.3 mW and 0.32ꢀmT modulations.
The initial 4ꢀaminomethylꢀ, 4ꢀ(βꢀaminoethyl)ꢀ2,2,6,6ꢀ
tetramethylpiperidinꢀ1ꢀoxyls (6a,b) and transꢀ3,4ꢀdiaminoꢀ
2,2,6,6ꢀtetramethylpiperidinꢀ1ꢀoxyl (7) were prepared by preꢀ
viously described procedures.9—11
cisꢀAmmine(4ꢀaminomethylꢀ2,2,6,6ꢀtetramethylpiperidinꢀ1ꢀ
oxyl)iodochloroplatinum(II) (31a). A solution of NaI (0.72 g,
4.8 mmol) in 0.4 mL of Н2О and a solution of 4ꢀaminomethylꢀ
2,2,6,6ꢀtetramethylpiperidinꢀ1ꢀoxyl (0.62 g, 3.4 mmol) in
0.4 mL of Н2О were added successively with stirring at ∼20 °C
to a solution of Nа[Pt(NH3)Cl3] in 10 mL H2O prepared from
Cisplatin (1.00 g, 3.3 mmol).12 The reaction mixture was stirred
for 2 h at ∼20 °C and kept in a refrigerator (∼10 °C) for 10 h.
The resulting precipitate was filtered off, thoroughly washed
with cold water (4×2 mL), and dried in air to give 1.24 g (67%)
of complex 31a, which was used in the synthesis of 31b. To
prepare an analytical grade specimen 31a, the synthesis was
carried out with K[Pt(NH3)Cl3]•0.5Н2О isolated intermediꢀ
ately.12 Complex 31a represents small yellow crystals, m.p.
155—158 °C (decomp.). Found (%): C, 21.6; H, 4.41; N, 7.43;
Pt, 34.1. C10H24ClIN3OPt. Calculated (%): C, 21.47; H, 4.33;
N, 7.51; Pt, 34.88. IR, ν/cm–1: 1555, 1605, 3170, 3225, 3255
(NH2, NH3).
a: X = Cl, X´ = I; b: X = X´ = Cl; c: X = X´ = NO3;
d: X + X´ =
; e: X + X´ =
cisꢀAmmine(4ꢀaminomethylꢀ2,2,6,6ꢀtetramethylpiperidinꢀ
1ꢀoxyl)dichloroplatinum(II) (31b). Silver nitrate (220 mg,
1.29 mmol) was added to a suspension of complex 31a (0.402 g,
0.72 mmol) in 10 mL of water, and the mixture was stirred with
a magnetic stirrer for 12 h in the dark. The precipitated silver
halides were thoroughly separated by centrifuging and filtering
through a dense glass filter to give a solution of orangeꢀcolored
dinitrato complex 31c, which did not contain AgNO3.5 A soluꢀ
tion of KCl (0.30 g) in 1 mL of water was added with stirring to
the solution of 31c and the mixture was allowed to stand for
12 h at ∼20 °C. The reaction solution was concentrated to
∼3 mL and the orange crystals thus formed were filtered
off, washed with cold water and EtOH, and dried in vacuo
to give 195 mg (58%) of complex 31b, m.p. 172—174 °C
(decomp.). Found (%): C, 25.9; H, 5.37; N, 8.83; Pt, 40.9.
C10H24Cl2N3OPt. Calculated (%): C, 25.69; H, 5.18; N, 8.99;
Pt, 41.74. IR, ν/cm–1: 1557, 1630, 3178, 3230 (NH2, NH3).
cisꢀAmmine[4ꢀ(βꢀaminoethyl)ꢀ2,2,6,6ꢀtetramethylpiperidinꢀ
1ꢀoxyl]iodochloroplatinum(II) (32a) was prepared similarly to
reason, these ligands are called leaving groups. The rate
constants for hydrolysis of ligands X can differ6 by a
factor of 105 and, as a consequence, the biological efꢀ
fects of complexes differing only by ligands X are disꢀ
similar. The amino ligands remain bound to platinum
throughout the whole metabolism route until they reach
the primary (regarding the antitumor activity) target, viz.,
DNA molecule; thetefore, they are referred to as carryꢀ
ing ligands.7,8
The purpose of this work was to study the binding of
new complexes to DNA using ESR and UV spectroscopy
and, taking into account the previous results,4,6 to sumꢀ
marize data on the influence of structures of aminoꢀ
nitroxide ligands on the efficiency of DNA platination
and the properties of the resulting LPtII—DNA adducts
(L is an aminonitroxide). Using ESR spectroscopy, one
can determine the degree of DNA platination, the moꢀ

1060
Russ.Chem.Bull., Int.Ed., Vol. 51, No. 6, June, 2002
Sen´ et al.
31a, yield 74%, m.p. 145—148 °C (decomp.). Found (%):
without isolation, successively with NaI and aminoꢀ
nitroxides 6a and 6b, respectively. Binuclear complex 5a
was prepared in a good yield and with good purity in a
similar way from the isolated K[Pt(NH3)Cl3]•0.5Н2О 12
and 0.5 equiv. of radical 7. The syntheses of intermediꢀ
ate dinitrato complexes 31c and 5c and dichloro comꢀ
plexes 31b and 5b were carried out as described previꢀ
ously.5
C, 23.3; H, 4.69; N, 7.53; Pt, 33.2. C11H26ClIN3OPt. Calcuꢀ
lated (%): C, 23.03; H, 4.57; N, 7.33; Pt, 34.02. IR, ν/cm–1
:
1560, 1609, 1627, 3190, 3245 (NH2, NH3).
transꢀ3,4ꢀBis(cisꢀammineiodochloroplatinum(II)amino)ꢀ
2,2,6,6ꢀtetramethylpiperidinꢀ1ꢀoxyl (5a). A solution (0.42 g,
2.55 mmol) of KI in 3 mL of water was added with stirring at
∼20 °C to a solution of K[Pt(NH3)Cl3]•0.5Н2О (313 mg,
0.85 mmol) in 4 mL of H2O, and then a solution of transꢀ3,4ꢀdiꢀ
aminoꢀ2,2,6,6ꢀtetramethylpiperidinꢀ1ꢀoxyl (85 mg, 0.45 mmol)
in 2 mL of water was added dropwise over a period of 1 h. The
reaction mixture was allowed to stand for 3 h at ∼20 °C, the
precipitated crystals were filtered off, washed successively with
water, EtOH, and Et2O, and dried in vacuo to give 290 mg
(73%) of complex 5a as a redꢀbrown finely crystalline material,
which decomposes (grows dark) without melting at a temperaꢀ
ture of ≥200 °C. Found (%): C, 11.3; H, 2.89; N, 7.53;
Pt, 41.2. C9H26Cl2I2N5OPt2. Calculated (%): C, 11.56; H, 2.80;
N, 7.49; Pt, 41.72. IR, ν/cm–1: 1550, 1608, 3162, 3215
(NH2, NH3).
transꢀ3,4ꢀBis(cisꢀamminedichloroplatinum(II)amino)ꢀ2,2,6,6ꢀ
tetramethylpiperidinꢀ1ꢀoxyl (5b). Silver nitrate (242 mg,
1.42 mmol) was added to a triturated suspension of complex 5a
(350 mg, 0.37 mmol) in 8 mL of water and the mixture was
stirred with a magnetic stirrer for 16 h in the dark. The precipiꢀ
tated silver halides were thoroughly separated by centrifuging
and filtering through a dense glass filter to give a solution of
orangeꢀcolored complex 5c, which did not contain AgNO3.
A solution of KCl (0.23 g, 3 mmol) in 1 mL of water was added
with stirring to a solution of complex 5c and the mixture was
allowed to stand for 12 h at ∼20 °C. The orange crystals
were filtered off, washed with cold water and EtOH, and
dried in vacuo to give 110 mg (40%) of complex 5b, which
decomposes (grows dark) without melting at a temperaꢀ
ture of ≥210 °C. Found (%): C, 14.8; H, 3.59; N, 9.53;
Pt, 50.6. C9H26Cl4N5OPt2. Calculated (%): C, 14.37; H, 3.48;
N, 9.31; Pt, 51.86. IR, ν/cm–1: 1556, 1627, 3165, 3238, 3265
(NH2, NH3).
Platination of DNA. Platination was carried out using fragꢀ
mented DNA from the bovine spleen (Reakhim, Russia) with a
mean molecular mass of ~1.65•106 (∼2500 nucleotide pairs).
The DNA concentration was determined from the optical denꢀ
sity of the sample at λ = 260 nm, using the extinction coeffiꢀ
cient ε = 6.6•103 L mol–1 cm–1. The platination of DNA was
studied for sufficiently waterꢀsoluble complexes (≥0.3 mg mL–1).
The platination was carried out in a 0.01 М solution of NaHCO3
for 24 h at 37 °C; the initial rin = [complex]/[nucleotide] molar
ratios ranged from 0.001 to 0.3. After separation of the unreacted
complex,4,6 the degree of modification r = [bound LPtII]/[nucleꢀ
otide] was determined by ESR on the basis of the >N—O•
content or by atomicꢀabsorption spectroscopy on the basis of
the platinum content. The results of both methods coincided to
within experimental errors (5—20%).
Complexes 3na,b and 5a,b are crystalline materials,
which decompose on melting. Iodineꢀcontaining comꢀ
plexes 31a, 32a, and 5a are poorly soluble in water
(≤0.1 mg mL–1); the solubility of dichloro complexes
31b and 5b is ∼0.4 mg mL–1, and 31c, 32c, and 5c cannot
be isolated from aqueous solutions even at a concentraꢀ
tion of ∼40 mg mL–1
.
The new complexes were identified as cisꢀisomers
judging from stronger transꢀeffect of the halide ligands
upon inclusion of the Nꢀligand in the PtII coordination
sphere13 and by analogy with complex 2b, for which the
cisꢀstructure was established by Xꢀray diffraction analyꢀ
sis.5 Complexes 31a, 32a, and 5a are, apparently, isomer
mixtures regarding the positions of ligands Х and Х´.
Complexes 4 and 5, prepared from diamino ligand 7,
contain transꢀarranged amino groups, as indicated by
the Xꢀray diffraction data for one of the bisꢀsulfꢀ
amide diastereomers, prepared by the reaction of 7 with
(1S)ꢀ(+)ꢀcamphorsulfonyl chloride.14 The bands at
3160—3265 and 1550—1630 cm–1 in the IR spectra of
31a,b, 32a, and 5a,b correspond to the stretching and
bending vibrations, respectively, of the NH2 and NH3
groups. The ESR spectra of dilute aqueous solutions of
the complexes comprise three lines, which is consistent
with the monoradical structures of these compounds.
The HFC constants at nitrogen (aN = 1.69 mT) and the
gꢀfactors for complexes 3 and 5 (2.0056) are equal and are
typical of aqueous solutions of sixꢀmembered nitroxides.
One more piece of evidence for the binuclear structure
of 5b is DNA stabilization upon the formation of adꢀ
ducts with 5b (see below), unlike the situation with comꢀ
plexes 1—4.
The relationship between the structure of complexes
and their activity in adduct formation with DNA. Accordꢀ
ing to presentꢀday data,2 the antitumor activity of diamino
PtII complexes is due to the bidentate binding of platiꢀ
num to DNA through the formal displacement of the
ligands X mainly by the N(7) atoms of purine bases. In
the case of mononuclear complexes (for example,
Cisplatin, complexes 1—4), up to 80—90% of the adꢀ
ducts are 1,2ꢀintrastrand crossꢀlinks either between neighꢀ
boring guanines or between adenine and guanine in ∼2 : 1
ratio. The rest are 1,3ꢀintrastrand crossꢀlinks (∼10%),
monodentate adducts (∼3%), and interstrand crossꢀlinks
(∼1%). The data for model oligonucleotides obtained by
various methods show2 that the 1,2ꢀcrossꢀlinking of neighꢀ
boring guanines induces bending of the helix by 40–80°
The melting point (Tm) and the melting range (∆T ) for the
initial and platinated DNA were found as described previously.6
Results and Discussion
Synthesis and structure of complexes. Complexes 31a
and 32a were synthesized from the Cisplatin via the inꢀ
termediate Nа[Pt(NH3)Cl3] complex,12 which was treated

Platinum(II) complexes with aminonitroxides
Russ.Chem.Bull., Int.Ed., Vol. 51, No. 6, June, 2002
1061
toward the major groove and deteriorates the coplanarity
of the guanine rings, so that the angle between the planes
of the platinated bases is ∼50°. In the case of binuclear
complexes such as 5, which also exhibit high antituꢀ
mor activities, the fraction of interꢀstrand crossꢀlinks
reaches15 ∼70%.
The effect of the structure of PtII nitroxide comꢀ
plexes on their activity toward the formation of DNA
adducts was studied for the initial ratio of the reactants
rin = 0.1 (see Experimental). The degrees of modificaꢀ
tion r obtained for various complexes correspond to the
xꢀcoordinates of the points in Fig. 1. The yꢀcoordinates
of these points show the specific destabilization δTm of
the DNA duplex, which corresponds to the decrease in
the DNA melting point (in °C) induced by the formaꢀ
tion of one adduct per every 100 nucleotides. This value
was calculated using the formula
butanedicarboxylato complexes is reflected in the deꢀ
crease in the number of DNA adducts formed. The
platinating activity of 4c can decrease if the reacꢀ
tion mixture contains anions capable of forming PtII
complexes stable to hydrolysis. Thus the number of
DNA adducts formed by complex 4c in a citrate buffer
(point 4c (c.b.), see Fig. 1) is about half that formed in a
bicarbonate buffer (point 4c), all other factors being
the same.
When the carrying amino ligands are rather large,
they exert a crucial influence on the platinating activity
of the complexes. Apparently, both the volume and the
linear size of the ligands are significant, because the
bulky biradical complex 1c and complex 32c, in which
the amino group is separated from the piperidine ring by
an ethylene bridge are characterized by close r values,
and these values are 10—15 times lower than those for
complex 4c.
δTm = (Tm´ – Tm)/100r,
The plot presented in Fig. 1 indicates that the DNA
adducts formed by the most inert complexes are characꢀ
terized by the highest specific destabilization δTm of the
DNA duplex, whereas the least destabilization is induced
by the adducts formed by reactive complexes 4c,d. Comꢀ
plexes 2b and 30b, which show a high antitumor activꢀ
ity,5 platinate DNA only 1.5—2 times less efficient than
Cisplatin, and their adducts, despite the large bulk of the
ligands, destabilize DNA to an even smaller extent than
the Cisplatin adducts.
The degree of DNA destabilization can also depend
on the orientation of the amino ligands in the adducts
relative to the axis of the DNA helix. It is clear that
the orientations are different for mono and diamino
ligands. In addition, racemic complexes 2, 4, and 5
are expected to produce two diastereomeric adducts
each. In the two diastereomeric adducts formed by the
PtII(transꢀDACH)Х2 complexes (DACH is 1,2ꢀdiaminoꢀ
cyclohexane), the cyclohexane rings are arranged at right
angle to the axis of the DNA helix; in the opinion of the
authors of previous publications,17,18 this creates the lowꢀ
est steric interaction. It follows from Fig. 1 that the
adducts of complexes 4 having a similar structure cause
the least pronounced destabilization of the DNA molꢀ
ecule. Much greater distortions of DNA are induced by
the diastereomer of the PtII(cisꢀDACH) adduct in which
the cyclohexane ring is arranged along the axis of the
DNA helix.18 Since complexes 1—3 form adducts in
which the rather bulky amino ligands are arranged along
the DNA major groove,19 the observed higher destabilizꢀ
ing influence of these adducts (see Fig. 1) can be reasonꢀ
ably attributed to the enhanced steric hindrance.
where Tm and Tm´ are the melting points of the initial
and platinated DNA, respectively (the melting curves
were published previously6). The r value depends on the
natures of both the leaving ligand Х and the carrying
amino ligand. For complexes containing identical carryꢀ
ing ligands, the difference between platination activities
is determined by the rate of hydrolysis of the leaving
ligand Х. However, there exists evidence that complexes
stable against hydrolysis, such as 1,1ꢀcyclobutaneꢀ
dicarboxylates of PtII, can react with DNA, bypassing
the intermediate step of hydrolysis.6,16 Easily hydrolyzꢀ
able complex 4c, which is converted almost completely
into a reactive aqua complex during dissolution,6 exhibꢀ
its the highest platinating activity. The retardation of
hydrolysis on passing to dichloro, oxalatoꢀ, or 1,1ꢀcycloꢀ
δTm/°C
0.5
5b
0
4c
2b
–0.5
–1.0
–1.5
–2.0
4d
4c (c.b.)
4e
30b
30c
A
31b
32c
1c
0.02
0.04
0.06
0.08
r
Fig. 1. Relationship between the platinating activity (r) and
specific destabilization of the DNA duplex (δTm) for various
platinum complexes. The platination of DNA was carried out
in 0.01 М NaHCO3 over a period of 24 h at 37 °C at the initial
molar ratio rin = 0.1. Point 4c (c.b.) corresponds to platination
in a 0.01 М citrate buffer, pH 7, A is Cisplatin.
Unlike complexes 1—4, adducts formed by binuclear
complex 5b induce an increase in the DNA melting point
(see Fig. 1). This does not mean that these adducts do
not disturb the DNA structure. The effect is due to the
predominant (up to 70%)15 formation of interstrand crossꢀ

1062
Russ.Chem.Bull., Int.Ed., Vol. 51, No. 6, June, 2002
Sen´ et al.
d1/d
0.8
links. For point 5b in Fig. 1 (r = 0.032), provided that
two nucleotides are crossꢀlinked, the length of the DNA
fragment is 2500 pairs of nucleotides, and the proportion
of the interstrand crossꢀlinks is 70%, one can calculate
the average number of crossꢀlinks per DNA fragment,
which would be 0.032•2500•0.7•2/2 = 56. Thus covaꢀ
lent crossꢀlinking of DNA strands prevents thermal deꢀ
struction of the duplex.
2b
30b
30c
31b
32c
4c
4d
4e
5a
0.7
0.6
0.5
The positions of the LPtII fragments in DNA. The
differences in the reactivity of complexes caused by difꢀ
ferent rates of displacement of the ligands X and by the
influence of the amino ligand bulk can also be maniꢀ
fested as different selectivities of DNA platination. In the
case of selective binding, one should expect nonuniform
distribution of the adducts along the DNA strand. In
terms of the helix—coil transition theory in doubleꢀ
stranded polynucleotides,20 slight destabilization of the
polymeric DNA is expected when local defects of the
secondary structure are concentrated in particular secꢀ
tions of the molecule, whereas a uniform distribution of
the defects would entail a more pronounced destabilizing
effect. It is reasonable to assume that the adducts formed
from lowꢀreactivity complexes would be arranged more
uniformly due to steric restrictions and, therefore, they
would cause a more pronounced DNA destabilization.
Yet another factor that may influence the degree of DNA
destabilization is orientation of the amino ligands in adꢀ
ducts relative to the axis of the DNA helix. The ESR
spectra of a DNA with a high degree of modification
(r > 0.05) recorded at 77 К show a pattern typical of
closely spaced nitroxide groups that undergo magnetic
dipoleꢀdipole coupling. The ESR spectral pattern should
depend on the positions of the adducts in DNA. The
shape of the ESR line of nitroxides observed at 77 К can
be characterized by the parameter d1/d, which is the ratio
of the sum of the amplitudes of the terminal peaks to the
amplitude of the central peak.21 This parameter is rather
sensitive to magnetic interaction. In the case of two inꢀ
teracting radicals, the distance L between them can be
determined from the d1/d value using the relation21
0.05
0.10
0.15
0.20
r
Fig. 2. ESR parameter d1/d vs. degree of modification r for
various complexes. The dots are experimental data and the
curve is the theoretical dependence obtained by the Monte
Carlo method for b = 0.6 nm (see the text). The errors in
determination of the d1/d parameter do not exceed 0.02. The
ESR spectra were recorded at 77 К for magnetic field modulaꢀ
tion of 0.32 mT and a microwave power of 0.3 mW.
The next step was to determine the probability (pn) that
no adducts are present between a distinguished (zero)
nucleotide and an nth nucleotide pair. The d1/d value
was calculated using the relation
,
(2)
where Ln is the distance between the nitroxide groups of
two adducts attached to the distinguished nucleotide pair
and to the nth nucleotide pair. Equation (2) is derived by
averaging the d1/d values found by relation (1) over the
distances L. The Ln distances were calculated from the
Cartesian coordinates of the nitroxide groups assuming
that they are located at distance b from the axis of the
DNA molecule (Fig. 3). If the coordinate system is oriꢀ
ented in such a way that the Z axis is directed along the
DNA axis and the nitroxide group of the distinguished
adduct is located in the X axis, the coordinates of the
nitroxide groups are bcosϕ, bsinϕ, and h. For the distinꢀ
guished adduct ϕ = 0, h = 0; for the adduct attached to
the nth nucleotide of the same strand as the distinguished
adduct, ϕ = 36n (deg), h = 0.34n (nm), while for the
complementary strand, ϕ = 143 + 36n, h = 0.34n (nm).
The ϕ and h values corresponding to the Вꢀform of DNA
were used. It is known for this form that the distance
between the neighboring nucleotide pairs is 0.34 nm and
L = 0.93 + 0.077/[d1/d – (d1/d)0] (nm),
(1)
where the subscript "0" refers to the case of zero interacꢀ
tion between the nitroxides. The (d1/d)0 value is deterꢀ
mined by measuring the d1/d values for low modification
degrees. In our case, (d1/d)0 = 0.45.
The d1/d values for DNA modified by various comꢀ
plexes are shown in Fig. 2 as points for different modifiꢀ
cation degrees r. The continuous line represents the theoꢀ
retical dependence of d1/d on r calculated by the Monte
Carlo method in terms of a model assuming nonselective
(random) arrangement of adducts in the DNA molecule.
The addition of 2rN complexes to a DNA molecule comꢀ
prising N nucleotide pairs was simulated on a computer.

Platinum(II) complexes with aminonitroxides
Russ.Chem.Bull., Int.Ed., Vol. 51, No. 6, June, 2002
1063
X
b
L
Z
0
Fig. 3. Probable arrangement of the platinum nitroxide fragments for binding to neighboring bases within one DNA strand.
a full convolution contains 10 nucleotide pairs. The angle
between the positions of nucleosides from different chains
(with the same coordinate Z ) amounts to 143°. This
angle is found as 360•1.15/(1.15 + 1.75) = 143, where
1.15 and 1.75 nm are the widths of the minor and major
grooves, respectively, between the DNA phosphate
chains.22 It was assumed that, due to steric restrictions,
the adducts in one strand are separated by at least one
nucleotide. The probabilities pn can also be found using
the relation pn = 0.5(1 – r)n–1r/0.88; the coefficient 0.88
is the sum of all pn values; it differs from unity because
for an isolated strand, p1 = 0 due to steric restrictions.
For the simplified model considered, the theoretical deꢀ
pendence of d1/d on r (the curve in Fig. 2) describes
adequately the data for most of the complexes with
b = 0.6 nm. Taking into account the Xꢀray diffraction
data5 for 2b, the distance between the N(7) atom of
guanine and the nitroxide group of the adduct is ∼1 nm.
The shorter theoretical b value may be due to the fact
that the angle between nitroxide and the axis of the
DNA helix differs from 90° and/or to structure distorꢀ
tions, resulting in proximity of the neighboring nitroxide
groups. The results for complex 4с do not fit the theoꢀ
retical dependence of d1/d on r (see Fig. 2). The data for
highly reactive 4с could be interpreted by assuming bindꢀ
ing to peripheral (e.g., phosphate) groups of DNA. Howꢀ
ever, this interpretation does not seem convincing in
view of the expected lability of such adducts.
In the adducts of complexes 31 and 32, the distance
between the Pt atom and the piperidinoxyl molecule
increases by one or two methylene groups with respect to
that in the adducts formed by complexes 30. According
to ESR, this results in a sharp increase in the radical
mobility (Fig. 4). Apparently, this is associated with the
partial exit of the piperidine ring from the relatively shalꢀ
low major DNA groove and with an increase in its rotaꢀ
tional mobility.
Thus, the study of new complexes 1—5 made it posꢀ
sible to elucidate the dependence of the efficiency of
their covalent binding to DNA on the nature of leaving
groups Х and the carrying aminonitroxide ligands and to
identify the dependence of the degree of destabilization
of the modified DNA on the structure of the adducts.
Complexes 2, 30, and 4 are comparable with Cisplatin
regarding the efficiency of binding to DNA and the deꢀ
gree of DNA destabilization. The presence of two bulky
amino ligands (as in complexes 1) or an increase in the
bulk of one ligand (as in complexes 32) decreases the
τ•109/s–1
6
4
2
The coincidence of the theoretical and experimental
dependences of d1/d on r for most of the complexes
attests to the absence of selective platination of DNA for
r ≥ 0.05. The insufficient sensitivity of the ESR method
precludes determination of the distances between the
radicals and, hence, the possible selectivity of DNA
platination for the low degrees of modification attainꢀ
able in vivo.
0
1
2
n
Fig. 4. Rotational correlation time for piperidinoxyls (τ) in the
DNA modified with complexes 30, 31, and 32, vs. the length of
the (CH2)n bridge.

1064
Russ.Chem.Bull., Int.Ed., Vol. 51, No. 6, June, 2002
Sen´ et al.
9. A. B. Shapiro, L. S. Bogach, V. M. Chumakov, A. A.
Kropacheva, V. I. Suskina, and E. G. Rozantsev, Izv. Akad.
Nauk SSSR, Ser. Khim., 1975, 2077 [Bull. Acad. Sci. USSR,
Div. Chem. Sci., 1975, 24 (Engl. Transl.)].
10. R. I. Zhdanov, N. G. Kapitanova, and E. G. Rozantsev,
Izv. Akad. Nauk SSSR, Ser. Khim., 1980, 364 [Bull. Acad.
Sci. USSR, Div. Chem. Sci., 1980, 29 (Engl. Transl.)].
11. V. D. Sen´, Izv. Akad. Nauk SSSR, Ser. Khim., 1989, 2094
[Bull. Acad. Sci. USSR, Div. Chem. Sci., 1989, 38, 1928
(Engl. Transl.)].
12. C. M. Giandomenico, M. J. Abrams, B. A. Murrer, J. F.
Vollano, M. I. Rheinheimer, S. B. Wyer, G. E. Bossard,
and J. D. Higgins, Inorg. Chem., 1995, 34, 1015.
13. F. Cotton and G. Wilkinson, Advanced Inorganic Chemisꢀ
try, Wiley, New York, 1966, 1.
platinating activity, apparently, due to steric hindrance.
The destabilizing effect measured based on the DNA
melting point increases with an increase in the linear
dimensions and the volume of the amino ligands. Acꢀ
cording to ESR data, when r ≥ 0.05, the adducts are
uniformly distributed along the DNA strand, which does
not rule out the presence of selectivity at lower degrees
of platination. With allowance for published data,3—5 it
can be concluded from the biological activities of comꢀ
plexes 1—4 that high antitumor activity is characteristic
of complexes that efficiently platinate an isolated DNA
and, simultaneously, entail a moderate destabilization of
the DNA duplex.
14. V. D. Sen´, V. V. Tkachev, and L. O. Atovmyan, Izv. Akad.
Nauk, Ser. Khim., 1993, 367 [Russ. Chem. Bull., 1993, 42,
328 (Engl. Transl.)].
References
15. R. Zaludova, A. Zakovska, J. Kasparkova, Z. Balcarova,
V. Kleinwachter, O. Vrana, N. Farrell, and V. Brabec, Eur.
J. Biochem., 1997, 246, 508.
16. U. Frey, J. D. Ranford, and P. J. Sadler, Inorg. Chem.,
1993, 32, 1333.
17. K. Inagaki, C. Ninomoya, and Y. Kidani, Chem. Lett.,
1986, 233.
18. V. Boudny, O. Vrana, F. Gaucheron, V. Kleinwachter,
M. Leng, and V. Brabec, Nucleic Acids Res., 1992, 20, 267.
19. S. U. Dunham, C. J. Turner, and S. J. Lippard, J. Am.
Chem. Soc., 1998, 120, 5395.
20. Yu. S. Lazurkin, M. D. FrankꢀKamenetskii, and E. N.
Trifonov, Biopolym., 1970, 9, 1253.
21. V. N. Parmon, A. I. Kokorin, and G. M. Zhidomirov,
Stabil´nye biradikaly [Stable Biradicals], Nauka, Moscow,
1980, 240 pp. (in Russian).
1. M. J. Bloemink and J. Reedijk, in Metal Ions in Biological
Systems, Eds. H. Sigel and A. Sigel, M. Dekker, New York,
1996, 32, 641.
2. E. R. Jamieson and S. J. Lippard, Chem. Rev., 1999,
99, 2467.
3. V. D. Sen´, V. A. Golubev, L. M. Volkova, and N. P.
Konovalova, J. Inorg. Biochem., 1996, 64, 69.
4. V. D. Sen´, A. V. Kulikov, A. V. Shugalii, and N. P.
Konovalova, Izv. Akad. Nauk, Ser. Khim., 1998, 1640 [Russ.
Chem. Bull., 1998, 47, 1598 (Engl. Transl.)].
5. V. D. Sen´, N. A. Rukina, V. V. Tkachev, A. V. Pis´menskii,
L. M. Volkova, S. A. Goncharova, T. A. Raevskaya, A. G.
Tikhomirov, L. B. Gorbacheva, and N. P. Konovalova, Izv.
Akad. Nauk, Ser. Khim., 2000, 1624 [Russ. Chem. Bull., Int.
Ed., 2000, 49, 1613].
22. W. Saenger, Principles of Nucleic Acid Structure, Springer,
New York, 1984.
6. A. V. Shugalii, A. V. Kulikov, M. V. Lichina, V. A. Golubev,
and V. D. Sen´, J. Inorg. Biochem., 1998, 69, 67.
7. T. V. Hambley, Coord. Chem. Rev., 1997, 166, 181.
8. S. G. Chaney, Int. J. Oncology, 1995, 6, 1291.
Received November 20, 2001;
in revised form March 20, 2002