The first iron(III) complexes with cyclin-dependent kinase inhibitors: Magnetic, spectroscopic (IR, ES+ MS, NMR, 57Fe M?ssbauer), theoretical, and biological activity studies

Article abstract of DOI:10.1016/j.jinorgbio.2009.12.002

The first Fe^III complexes 1-6 with cyclin-dependent kinase (CDK) inhibitors of the type [Fe(L_n)Cl₃]·nH₂O (n=0 for 1, 1 for 2, 2 for 3-6; L₁-L₆=C2- and phenyl-substituted CDK inhibitors derived from 6-benzylamino-9-isopropylpurine), have been synthesized and characterized by elemental analysis, IR, ⁵⁷Fe Mo?ssbauer, ¹H and ¹³C NMR, and ES+ mass spectroscopies, conductivity and magnetic susceptibility measurements, and thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC). The study revealed that the compounds are mononuclear, tetrahedral high-spin (S=5/2) Fe^III complexes with an admixture of an S=3/2 spin state originating probably from five-coordinated Fe^III ions either connecting with a bidentate coordination mode of the CDK inhibitor ligand or relating to the possibility that one crystal water molecule enters the coordination sphere of the central atom in a portion of molecules of the appropriate complex. Nearly spin-only value of the effective magnetic moment (5.82μ_eff/μ_B) was determined for compound 1 due to absence of crystal water molecule(s) in the structure of the complex. Based on NMR data and DFT calculations, we assume that the appropriate organic ligand is coordinated to the Fe^III ion through the N7 atom of a purine moiety. The cytotoxicity of the complexes was tested in vitro against selected human cancer cell lines (G-361, HOS, K-562 and MCF-7) along with the ability to inhibit the CDK2/cyclinE kinase. The best cytotoxicity (IC₅₀: 4-23μM) and inhibition activity (IC₅₀: 0.02-0.09μM) results have been achieved in the case of complexes 2-4, and complexes 3, 4 and 6, respectively. In addition, the X-ray structure of 2-chloro-6-benzylamino-9-isopropylpurine, i.e. a precursor for the preparation of L₁, L₄ and L₅, is also described.

Full text of DOI:10.1016/j.jinorgbio.2009.12.002

Journal of Inorganic Biochemistry 104 (2010) 405–417
Contents lists available at ScienceDirect
Journal of Inorganic Biochemistry
journal homepage: www.elsevier.com/locate/jinorgbio
The first iron(III) complexes with cyclin-dependent kinase inhibitors: Magnetic,
spectroscopic (IR, ES+ MS, NMR, 57Fe Mössbauer), theoretical, and biological
activity studies
a,
a
a
b
c
a
*
ˇ
ˇ
ˇ
ˇ
ˇ
Zdenek Trávnícek , Igor Popa , Michal Cajan , Radek Zboril , Vladimír Kryštof , Jirí Mikulík
a
ˇ
Department of Inorganic Chemistry, Faculty of Science, Palacky´ University, Tr. 17. listopadu 12, CZ-779 00, Olomouc, Czech Republic
Department of Physical Chemistry and Nanomaterials Research Centre, Faculty of Science, Palacky´ University, Tr. 17. listopadu 12, CZ-779 00, Olomouc, Czech Republic
Laboratory of Growth Regulators, Faculty of Science, Palacky University, Šlechtitelu 11, CZ-783 71, Olomouc, Czech Republic
b
c
ˇ
´
˚
a r t i c l e i n f o
a b s t r a c t
The first FeIII complexes 1–6 with cyclin-dependent kinase (CDK) inhibitors of the type [Fe(Ln)Cl3]ꢀnH2O
(n = 0 for 1, 1 for 2, 2 for 3–6; L1–L6 = C2- and phenyl-substituted CDK inhibitors derived from 6-benzyl-
amino-9-isopropylpurine), have been synthesized and characterized by elemental analysis, IR, 57Fe
Mössbauer, 1H and 13C NMR, and ES+ mass spectroscopies, conductivity and magnetic susceptibility mea-
surements, and thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC). The study
revealed that the compounds are mononuclear, tetrahedral high-spin (S = 5/2) FeIII complexes with an
admixture of an S = 3/2 spin state originating probably from five-coordinated FeIII ions either connecting
with a bidentate coordination mode of the CDK inhibitor ligand or relating to the possibility that one
crystal water molecule enters the coordination sphere of the central atom in a portion of molecules of
Article history:
Received 1 July 2009
Received in revised form 23 November 2009
Accepted 30 November 2009
Available online 5 December 2009
Keywords:
Iron(III) complexes
57Fe Mössbauer spectroscopy
CDK inhibition
In vitro cytotoxicity
DFT calculations
the appropriate complex. Nearly spin-only value of the effective magnetic moment (5.82 leff/lB) was
determined for compound 1 due to absence of crystal water molecule(s) in the structure of the complex.
Based on NMR data and DFT calculations, we assume that the appropriate organic ligand is coordinated to
the FeIII ion through the N7 atom of a purine moiety. The cytotoxicity of the complexes was tested in vitro
against selected human cancer cell lines (G-361, HOS, K-562 and MCF-7) along with the ability to inhibit
the CDK2/cyclinE kinase. The best cytotoxicity (IC50: 4–23
lM) and inhibition activity (IC50: 0.02–
0.09 M) results have been achieved in the case of complexes 2–4, and complexes 3, 4 and 6, respectively.
l
In addition, the X-ray structure of 2-chloro-6-benzylamino-9-isopropylpurine, i.e. a precursor for the
preparation of L1, L4 and L5, is also described.
Ó 2009 Elsevier Inc. All rights reserved.
1. Introduction
R-roscovitine is one of the first selective inhibitors of CDKs, identi-
fied through structure–activity relationships study of 2,6,9-trisub-
Protein kinases are enzymes that catalyze phosphorylation of
suitable protein substrates, which in turn change their conforma-
tion and more importantly also their activity. Timely and spatially
controlled protein kinases therefore participate nearly in all aspect
of cell biology [1]. However, recently published works proved that
mutations and dysregulations of protein kinases should lead to ini-
tiation of many human diseases. Cyclin-dependent kinases (CDKs)
form a subgroup of serine/threonine protein kinases controlling
many cellular events, including a cell cycle division [2]. CDKs are
frequently deregulated in cancer cells, so these enzymes represent
rational drug targets [3]. To date, many pharmacological inhibitors
of CDKs have been identified and the most active of them already
undergo clinical trials as new anticancer drugs [4,5]. One of them,
stituted purines [6,7]. As demonstrated by X-ray crystallography, it
binds specifically to the active site of CDK2 and thus interferes
with its catalytic function [8]. Moreover, R-roscovitine inhibits also
CDK9 that phosphorylates and thus activates C-terminus of RNA
polymerase II. Its inhibition triggers insufficient transcription of
many genes, especially those with short half-lived mRNAs and pro-
teins like oncogenic Hdm2, D and B cyclins, or anti-apoptotic Mcl-1
[9,10], which leads to cell cycle arrest and/or apoptosis of cancer
cells. The one of the most recent works even suggests that CDK9
inhibitors of transcription could also serve as potential drugs
against tumour invasion and metastasis [11]. In July 2006, the chi-
ral form (R) of roscovitine entered a Phase IIb of the clinical trials
under code names of CYC202 and seliciclib (Cyclacel Pharmaceuti-
cals, Inc., www.cyclacel.com). Hitherto, some of 2,6,9-trisubsti-
tuted purines, that are N6-modified by the above described
manner, were used as ligands for the preparation of the cytotoxic
* Corresponding author. Tel.: +420 585 634 944; fax: +420 585 634 954.
ˇ
E-mail address: zdenek.travnicek@upol.cz (Z. Trávnícek).
0162-0134/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.jinorgbio.2009.12.002

ˇ
Z. Trávnícek et al. / Journal of Inorganic Biochemistry 104 (2010) 405–417
406
transition metal complexes [e.g. PtII, PdII, CuII, ZnII] in our previous
papers [12–14].
electronic structure of FeIII ions. Besides, the goal of this study was
to prepare FeIII complexes showing significant in vitro cytotoxicity
against some human cancer cell lines and thus, to obtain com-
pounds having possibility to find the utilization as drug candidates
in anticancer therapy.
Recently, we have published two papers regarding the synthesis
and characterization of FeII and FeIII complexes involving 2,6,9-tri-
substituted purines. The first of them describes the preparation of
the (H+L)2[FeCl5] ionic pair compound [15], where H+L stands for a
protonated form of bohemine, i.e. 2-[(3-hydroxypropyl)amino]-6-
benzylamino-9-isopropylpurine, while the second one [16] deals
with the study of complexes having the composition of
[Fe(L)Cl3]ꢀH2O, where ligands L are N6-benzyladenosine deriva-
tives with variously modified benzyl rings. Moreover, there are
only a few works in the literature describing the preparation and
study of FeII,III complexes with some adenine or purine derivatives.
For instance, FeIII complexes with non-substituted adenine (adH),
2. Results and discussion
2.1. Chemistry and general features
The reaction of FeCl3ꢀ6H2O with the corresponding CDK
inhibitor, Ln (see Schemes 1 and 2) in an ethanolic solution in a
molar ratio 1:1 led to formation of mononuclear complexes of
[Fe(Ln)Cl3]ꢀxH2O (x = 0 for 1, x = 1 for 2, and x = 2 for 3–6). The
prepared compounds are hygroscopic (they need to be stored in
desiccator over P4O10 as a drying medium), well soluble in ethanol,
acetone, N,N0-dimethylformamide (DMF), dimethyl sulfoxide
(DMSO), partially soluble in water, and insoluble in diethylether.
Their colours, analytical data, effective magnetic moment values
and molar conductivities are listed in Table 1, while the most
important infrared spectral data are summarized in Table 2. The
molar conductivity values (1.1–5.4 S cm2 molꢁ1 in methanol and
4.1–16.8 S cm2 molꢁ1 in acetone) clearly showed the non-ionic fea-
ture of the studied complexes 1–6 in the solvents used [22]. No
change in conductivity values was achieved after 24 h.
e.g. {[Cl(H2O)(ad)Fe-l-(adH)-Fe(H2O)Cl2]n [17], [Fe(adH)2Cl3] [18],
[(ClO4)(adH)2Fe-l-(adH)2-Fe(adH)2(ClO4)](ClO4)4 and [Fe(ad)(ClO4)]ꢀ
EtOHꢀ2H2O [19]}, were reported. However, all of these complexes
have been characterized by elemental analyses, room temperature
magnetochemical, spectroscopic (UV–VIS, IR) and conductometric
measurements, and their potential biological activity was not
ever tested. With respect to Fe-complexes containing purine
skeleton in their structure, only three X-ray structures [20,21]
were determined and deposited within the Cambridge Crystallo-
graphic Data Centre up to now, namely [FeII(5-imp)(H2O)5]ꢀ2H2O
(5-imp = inosine-5-monophosphate) [20], [FeII(5-gmp)(H2O)5]ꢀ
3H2O (5-gmp = guanosine-5-monophosphate) [21] and [FeII(6-mp)3
FeIIICl4]Cl (6-mp = 6-mercaptopurine) [21].
2.2. 57Fe Mössbauer spectroscopy and magnetochemistry
Together with all other results, we also bring here a detailed
interpretation of 57Fe Mössbauer spectra, which showed the diver-
sity of geometric arrangements in the vicinity of the central Fe
atoms in the prepared compounds, and which is interrelated with
Hyperfine parameters of room temperature Mössbauer spectra
are summarized in Table 3. As can be clearly seen from the table,
Scheme 1. A representation of 6-benzylamino-9-isopropylpurine derivatives, L1–L6, used as organic ligands for the preparation of the FeIII complexes (1–6).

ˇ
Z. Trávnícek et al. / Journal of Inorganic Biochemistry 104 (2010) 405–417
407
Scheme 2. A general pathway for the synthesis of the FeIII complexes 1–6.
ting (D
EQ = 0.96–1.02 mm sꢁ1), which may be connected with
Table 1
Colour, magnetic and molar conductivity data of the complexes 1–6.
decreasing the symmetry in the vicinity of the central atoms as
compared to the major component. Thus, it may be concluded
Compound
Colour
lefflB
kM (S cm2 molꢁ1
)
that the above-mentioned
DEQ values may be ascribed to a trigo-
a
b
nal–bipyramidal coordination with the 3/2 spin state of iron,
where either one crystal water molecule is probably coordinated
to iron or the corresponding Ln ligand is coordinated as a biden-
tate one. The major (70–75% of spectrum area) component reveals
[Fe(L1)Cl3] (1)
Brownish red
Brown
Brownish orange
Brownish orange
Brownish orange
Brown
5.82
5.20
5.21
5.00
5.33
5.24
5.4
2.5
1.2
1.1
1.5
3.1
16.8
4.4
4.6
5.1
12.0
4.1
[Fe(L2)Cl3]ꢀH2O (2)
[Fe(L3)Cl3]ꢀ2H2O(3)
[Fe(L4)Cl3]ꢀ2H2O (4)
[Fe(L5)Cl3]ꢀ2H2O (5)
[Fe(L6)Cl3]ꢀ2H2O (6)
a relatively low quadrupole splitting (D )
EQ = 0.56–0.58 mm sꢁ1
which may be ascribed to a distorted tetrahedral geometry. To
distinguish between two probable structural types, the
Mössbauer spectrum of representative sample (compound 4)
a
Measured in methanol.
Measured in acetone.
b
was measured at 5 K (Fig. 1). As
a
result, the sextet
(d = 0.57 mm sꢁ1) corresponding to the presence of FeIII species
probably with the coordination number of 5 was observed in
the spectrum together with the singlet having the isomer shift
of 0.41 mm sꢁ1. This value is characteristic for tetrahedrally coor-
dinated FeIII ions [23] and even lower than expected due to the
temperature decreasing. It seems that the higher values of isomer
shifts in room temperature spectra corresponding to tetrahedral
FeIII (d = 0.32–0.35 mm sꢁ1) cations are caused by the presence
of water molecules in an outer coordination sphere, while this ef-
fect is suppressed at low temperatures.
Mössbauer spectra of complexes 2–6 are very similar with two
spectral components differing mainly in quadrupole splitting
parameters, while the isomer shift parameters lie in the narrow
range of 0.32–0.35 mm sꢁ1. The presence of two fitted subspectra
may be associated with the existence of two non-equivalent
structural and spin states of iron within the compounds, although
the values of isomer shift clearly indicate the formal oxidation
state of 3+. The minor component with the spectral area, varying
between 20% and 25%, shows a relatively higher quadrupole split-
Table 2
Selected IR spectral data (cmꢁ1) of the complexes 1–6.
Compound
m(C@N)
m(C@C)
m(Car–O)
m
(Calif–O)
m
(Fe–Cl)
m(Fe–N)
[Fe(L1)Cl3] (1)
1638s
1635s
1637s
1634s
1635s
1638s
1642s
1581m
1452m
1572m
1461m
1589s
1459m
1571m
1453m
1574s
1453m
1576m
1459m
1627s
1535m
1489m
–
–
381s
384s
381s
382s
381s
382s
333w
328w
332w
330w
330w
331w
344w
[Fe(L2)Cl3]ꢀH2O (2)
[Fe(L3)Cl3]ꢀ2H2O (3)
[Fe(L4)Cl3]ꢀ2H2O (4)
[Fe(L5)Cl3]ꢀ2H2O (5)
[Fe(L6)Cl3]ꢀ2H2O (6)
[Fe(L1)Cl3] (1a)a
1214m
1049m
1051m
1051m
1060m
1049m
–
1215m
–
–
1214m
–
403m
411m
a
The values belonging to the optimized structure 1a (see Fig. 4a) and calculated by B3LYP/6–311G*/Wachter’s level.

ˇ
Z. Trávnícek et al. / Journal of Inorganic Biochemistry 104 (2010) 405–417
408
Table 3
57Fe Mössbauer spectral parameters of the complexes 1–6.
Compound
Component
T (K)
d
0.01 (mm sꢁ1
)
D
EQ 0.01 (mm sꢁ1
)
C
0.01 (mm sꢁ1
)
RA 1 (%)
1
Doublet
Singlet
300
5
0.25
0.40
0.30
–
0.40
–
100
100
2
3
4
Doublet
Doublet
300
300
0.34
0.33
0.57
0.99
0.39
0.30
80
20
Doublet
Doublet
300
300
0.35
0.33
0.56
0.99
0.34
0.37
79
21
Doublet
Doublet
Singlet
Sextet
300
300
5
0.35
0.34
0.41
0.57
0.56
0.96
–
0.38
0.33
–
76
24
46
54
5
–0.18a
46.7b
5
6
Doublet
Doublet
300
300
0.35
0.34
0.35
0.97
0.35
0.30
75
24
Doublet
Doublet
300
300
0.34
0.32
0.58
1.02
0.37
0.30
79
20
1ac
1bc
1cc
1dc
Doublet
Doublet
Doublet
Doublet
0
0
0
0
0.364
0.384
0.478
0.450
d – Isomer shift related to metallic iron,
DEQ – quadrupole splitting, C – full width at a half of maximum, RA – percentage of subspectrum area related to iron content.
a
eQ – quadrupole shift (mm sꢁ1).
b
H – hyperfine magnetic field (T).
c
Theoretically calculated values based on optimized geometries of the predicted molecular structures of the complex 1 (for structural details, see the Fig. 4 and Table 4).
The calculations were performed at 0 K and vacuum, thus the value of the temperature Doppler shift (ca. 0.10–0.12 mm sꢁ1 for Fe-complexes) has to be applied to correct the
corresponding values to the values at room temperature.
The interpretation of the room temperature 57Fe Mössbauer
spectrum of the anhydrous complex 1 is not as complicated as it
consists of one doublet with the lower isomer shift
(d = 0.25 mm sꢁ1) compared to complexes 2–6, thus giving an evi-
dence for tetrahedral FeIII compounds with an S = 5/2. This differ-
ence confirms the above-mentioned prediction of the important
influence of the crystal water molecules on the hyperfine parame-
ters [24]. The low temperature spectrum (5 K) of complex 1 yields
the singlet with almost the same d parameter as in the case of the
complex 4, thus proving the suppression of the distortion caused
by water molecule in the vicinity of the central atom.
value for S = 5/2 is 4.377 cm3 K molꢁ1). The Curie constant value
of compound 4 lies between S = 3/2 state (theoretical value
1.876 cm3 K molꢁ1) and S = 5/2 spin states (see above), which indi-
cates the presence of a spin mixture with S = 3/2 state in minority.
Based on the determined Weiss constant values for complexes 1
[h = ꢁ2.7(4) cmꢁ1] and 4 [h = ꢁ5.2(3) cmꢁ1], the presence of a weak
antiferromagnetic interaction has been found in both cases. It may
be connected with variety hydrogen bonds and non-bonding inter-
actions presented within the sample in the solid state. Moreover,
the antiferromagnetic interaction may be also indicated by a sextet
in low-temperature Mössbauer spectrum of complex 4. Finally, we
may not pronounce the univocal conclusion regarding the stereo-
chemistry and electronic structure, but the results obtained from
magnetochemical measurements confirm indirectly our presump-
tion acquired by 57Fe Mössbauer spectroscopy.
The magnetic susceptibility of all complexes have been mea-
sured at room temperature, and the calculated values of effective
magnetic moment {
l
eff = 2.828(vM ꢀ T)1/2
ꢀ
l
B} are as follows: 5.82
lB for 3, 5.00
B for 4, 5.33 leff/lB for 5, and finally 5.24 leff/lB for 6 (see Ta-
leff/lB for compound 1, 5.20 leff/lB for 2, 5.21 leff/
leff/l
ble 1). Accordingly, the vM ꢀ T values are 4.235 cm3 K molꢁ1 (for 1)
and between 3.126 and 3.552 cm3 K molꢁ1 (for compounds 2–6),
respectively. From above-mentioned results it is clear that the cen-
tral FeIII ion in an anhydrous complex 1 is in high-spin state with
five unpaired electrons (spin-only value for FeIII, S = 5/2; 5.92
2.3. 1H and 13C NMR spectroscopies and ES+ mass spectrometry
In an effort to suppress the consequences of paramagnetism of
the complexes on the appearance and interpretation of NMR spec-
tra, low concentration of samples for 1H NMR experiments was
used. However, measuring of these low concentration samples re-
sulted in the spectra, where intensities of signals of deuterated sol-
vents are incomparably higher than intensity of signals of the
compounds themselves in some cases, and thus, some regions of
the spectra were not interpretable. For that reason, various deuter-
ated solvents (DMSO-d6, MeOH-d3, DMF-d7, acetone-d6) were used
for measurements with purpose to suppress these observations
and with the effort to find and interpret as many signals as possi-
ble. The use of different deuterated solvents also proved indirectly
a stability of measured compounds in the solution, because the
corresponding spectra were found to be identical in time depen-
dent experiments. In 1H NMR spectra, the most significant coordi-
leff/lB) [25]. In all other cases (2–6), there is at least one crystal
water molecule, which probably enters the vicinity of central atom
and subsequently increases the coordination number from 4 to 5.
The result of this process is a mixture of complexes with two dif-
ferent geometries (tetrahedron and trigonal–bipyramid) and also
possibly two different spin states (S = 5/2 and S = 3/2).
Temperature dependence of magnetic susceptibility for two
compounds (1 and 4) was also measured within the temperature
range of 80–298 K (see Fig. 2). It has been found, that the effective
magnetic moment value is very smoothly decreased along with
decreasing temperature, specifically from 5.80 leff/lB to 5.70 leff/
lB (for 1) and from 5.36 leff lB to 5.24 leff lB (for 4). The simple
/
/
Curie–Weiss model was used to determination of parameters such
as Curie and Weiss constants. The Curie constant values were
found to be 4.350 cm3 K molꢁ1 (for 1) and 3.717 cm3 K molꢁ1 (for
4). These values support our assumption of the clear high-spin tet-
rahedral arrangement for compound 1 (theoretical Curie constant
nation shifts,
Dd, were found in the case of the C8–H signals
(D
d = 0.29–0.50 ppm;
D
d = dcomplex ꢁ dligand). As for 13C NMR spec-
tra, the most significant coordination shifts were found for the
C8
(D
d = 3.51–3.61 ppm) and C5 atoms
[
D
d = ꢁ2.58ꢁ(ꢁ6.26
ppm)]. In relation to above-mentioned results we assume the

ˇ
Z. Trávnícek et al. / Journal of Inorganic Biochemistry 104 (2010) 405–417
409
Fig. 2. Temperature dependence of the molar magnetic susceptibility (full trian-
gles) and the effective magnetic moment (squares) for complex 1 (up) and complex
4 (down); full lines – fitted data.
hindrance of N1 and N3 sites caused by substitution of purine skel-
eton at C2, C6 and N9 positions.
ES+ mass spectra were measured for complexes 1–6. The
[M+H]+ molecular peak of the complexes, without molecules of
crystal water, appeared at 549 (for 2), 568 (for 3), 525 (for 5) and
568 m/z (for 6). Consecutively, the peaks observed at 359 (for 1),
385 (for 2), 371 (for 3), 355 (for 4), 327 (for 5) and 387 m/z (for
6) clearly indicate the presence of the appropriate organic ligand
Ln in the complexes. The peak located at 513 (for 2), 483 (for 4)
and 454 m/z (for 5) can be assigned to [Fe(Ln)Cl2+H]+ fragment,
whilst peak at 461 (for 3), 445 (for 4), 417 (for 5) and 477 m/z
(for 6) probably belongs to fragment of [Fe(Ln)Cl+H]+. The mole-
cules of organic ligands were also fragmented and the peaks at
137 m/z are connected with [adenine+H]+.
2.4. IR spectra
IR spectra of the complexes 1–6 were measured over the range
of 150–4000 cmꢁ1 (see Table 2). The bands observed between 1634
and 1638 cmꢁ1 may be assigned to the
m(C@N) stretch vibration of
the heterocyclic rings. The same bands appear in the spectra of the
free CDK inhibitors, but near value of 1620 cmꢁ1, which can sup-
port the coordination of the organic ligands through any N-atom
of a purine moiety to FeIII central cation. The bands observed at
Fig. 1. 57Fe Mössbauer spectra of complexes 1 and 4 taken at temperatures of 300
and 5 K, together with the fitted subspectra. Full lines represent the best fit.
1572–1581 and 1420–1450 cmꢁ1 correspond to
m(C@C) vibration
of aromatic rings, while peaks observed in the region of 640–
920 cmꢁ1 are assignable to skeletal vibration of the purine ring
[26]. The maximum near 1050 cmꢁ1 is observable in the spectra
coordination of organic ligand to FeIII ion through nitrogen atom
N7, which is also the most possible coordination site due to steric
of complexes 2–6, and can be assigned to m(Calif–O) vibration of a

ˇ
Z. Trávnícek et al. / Journal of Inorganic Biochemistry 104 (2010) 405–417
410
zene ring [27]. Spectrum of compound 1 showed two maxima at
753 and 658 cmꢁ1 assignable to
m(Calif–Cl) vibration [27]. The
bands of d(CH2) are located in the range of 1482–1496 cmꢁ1 in
the spectra of all compounds. All the above-mentioned vibrational
maxima observed in the infrared spectra clearly proved the pres-
ence of the corresponding Ln ligand in compounds 1–6. The max-
ima, which appeared in spectra of all complexes near 3100 and
3400 cmꢁ1, can be assigned to the
m
(C–H)ar, and
H) vibrations, respectively [27,28]. A very strong band observed
near 382 cmꢁ1 can be clearly assigned to
(Fe–Cl) vibration
[28,29]. There is one more band of weak intensity near
334 cmꢁ1, which based on the literature data [29], may be assigned
to the (Fe–N) stretching vibration.
m(N–H) or m(O–
m
a
m
2.5. Thermal analysis
Fig. 3. TGA and DSC curves of complex 2.
The thermal behaviour of the prepared complexes can be dem-
onstrated on the complex 2 {[Fe(L2)Cl3]ꢀH2O} (Fig. 3). The course of
the thermal decompositions for all compounds are very similar,
however, in the case of 2, the first weight decrease proceeds in
primary alcoholic group [27]. In the spectra of complexes 2, 3, and
6, the maximum about 1214 cmꢁ1 was also found, which can be as-
signed to m(Car–O) vibration of hydroxyl group attached to a ben-
Fig. 4. The geometries of possible structures of complex 1, [Fe(L1)Cl3] (S = 5/2) 1a–1c and [Fe(L1)Cl3(H2O)] (S = 3/2) 1d, optimized on the B3LYP/6-311+G/Wachter‘s level of
theory.

ˇ
Z. Trávnícek et al. / Journal of Inorganic Biochemistry 104 (2010) 405–417
411
the temperature range of 41–140 °C and can be connected with the
elimination of one crystal water molecule (found/calcd.: 2.8/3.0%).
This process is accompanied by an endothermic effect on the DSC
Thus, DG of the complex involving the N3 protonated and N6
coordinated ligand (1b) is by 33.2 kcal molꢁ1 higher as compared
to that of the complex 1a. If the ligand is bidentate coordinated
via the N6 and N7 atoms and the proton is shifted to the N1 po-
sition, then the molecular structure is partially stabilized by weak
curve with the minimum at 93.7 °C
(DH = 54.6 mJ =
10.3 kJ molꢁ1 = 2.5 kcal molꢁ1). Further, the compound exists as
an anhydrous intermediate between temperatures of 141 and
158 °C. The main thermal decomposition of the studied complex
starts at 159 °C and proceeds in two steps without formation of
stable intermediates. The first step can be observed on the TGA
curve in the temperature range of 159–248 °C and is accompanied
by a broad exothermic effect on the DSC curve, with the maximum
interaction between mentioned proton and
P-electron system of
the benzyl moiety. In this case, the geometry of the complex can
be described as a distorted trigonal–bipyramid (structure 1c).
Nevertheless, the
DG energy of such complex is increased by
13.6 kcal molꢁ1 as compared to that of structure 1a. As for the
structure 1d, the coordination number of iron is increased by
the coordination of one water molecule, and thus, the donor
atoms arrangement of the structure has a shape of trigonal–
bipyramid (structure 1d).
at 199.3 °C (
D
H = ꢁ75.3 mJ = ꢁ14.2 kJ molꢁ1 = 3.4 kcal molꢁ1). The
second step, i.e. main weight decrease, of thermal decomposition
appeared in the range of 273–503 °C and is associated with inten-
sive and sharp exothermic effect with the maximum at 386 °C. The
total weight decrease (87.0%) is in a good correspondence with the
calculated value of 86.1% and is clearly connected with decompo-
In quest of interpret infrared spectra of the compounds, the cal-
culation of vibrational frequencies for the complex 1a at the
B3LYP/6-311+G/Wachter‘s level of the theory has been also per-
formed. The calculated infrared spectrum of the complex 1a shows
two strong bands in the far IR region at 403 and 411 cmꢁ1, which
sition of the complex to
a
-Fe2O3 which was proved using 57Fe
Mössbauer spectroscopy (d = 0.37 mm sꢁ1
,
D
EQ ꢁ 0.21 mm sꢁ1
,
hyperfine magnetic filed 51.2 T).
are assignable to the
m(Fe–Cl) vibration. The stretching vibration
of
m
(Fe–N) has been found at 344 cmꢁ1. All these values correspond
2.6. Quantum chemical calculations
well with experimental data (for details, see Table 2). For a com-
parison, the vibrational frequencies of the complexes 1c and 1d
were also calculated. The calculated spectrum of the complex 1c
The complex of [Fe(L1)Cl3] (1) has been chosen as a model sys-
tem for the quantum chemical investigation of the geometry and
electronic structure of iron compounds presented in this work. In
general, there are two different possibilities how the L1 ligand
can be coordinated to FeIII ion: (i) a monodentate coordination
via N1, N3 or N7 atom; (ii) a bidentate coordination via N6 and
N7 atoms. The questions regarding the spin state of the iron(III)
central ion in connection with stereochemistry of the complex
had to be also taken into account during the calculations. Opti-
mized molecular geometries of the complex 1 are shown in
Fig. 4, their important structural data are summarized in Table 4.
Based on the above-mentioned results (Fig. 4, Table 4), and
moreover, taking into account a steric hindrance of some possible
coordination sites of the L1 ligand (both N1 and N3 binding sites
are sterically hindered by adjacent substituents at C2 and N9 posi-
tions – for better view, see the Fig. 4), may be concluded that only
ligand’s coordination via N7 or both N7 and N6 atom should lead to
the complex formation.
showed bands belonging to the
412 cmꢁ1, while maximum attributable to the
has been found at 192 cmꢁ1. Similarly, the
(Fe–Cl) vibrations of
the complex 1d have been found at 378 and 407 cmꢁ1, and the
(Fe–N) one at 199 cmꢁ1
The calculated values of Mössbauer shift, d, was found to be
m
(Fe–Cl) vibration at 383 and
m
(Fe–N) vibration
m
m
.
0.364 mm sꢁ1 for a tetrahedral structure of 1a (S = 5/2) and
0.450 mm sꢁ1 for a trigonal–bipyramidal structure of 1d (S = 3/2).
However, it is necessary to note at this moment that the calcula-
tions were performed at 0 K and vacuum, thus the value of the
temperature Doppler shift (ca. 0.10–0.12 mm sꢁ1 for Fe-com-
plexes) has to be applied to correct the corresponding values to
the values at room temperature. We may conclude after the reali-
zation of these corrections that the calculated values correspond
well with those of experimentally obtained (see Table 3).
The results obtained by magnetic susceptibility measurements
and 57Fe Mössbauer spectroscopy indicated the presence of two
spin states (with S = 5/2 and S = 3/2) within the complexes 2–6,
which may originate from different geometries of the complexes.
The geometry each of the hydrated complexes 2–6 may be influ-
enced by two main manners. The first way may be related with
fact that either one crystal water molecule may enter to the inner
coordination sphere of the central iron atom or the corresponding
organic ligand, Ln, may be coordinated as a bidentate one proba-
bly through the N6 and N7 atoms. On the other hand, the com-
plex 1 can be considered to be magnetically and structurally
‘‘pure”, because nearly the spin-only effective magnetic moment
2.7. Biological activity testing
In vitro antitumour activity of all prepared complexes 1–6 and
the starting compounds L1–L6 and FeCl3ꢀ6H2O have been deter-
mined. The results of biological testing are presented in Table 5.
With the exception of 1, all the prepared complexes showed a sub-
stantial activity in a micromolar range on human malignant mela-
noma (G-361), osteogenic sarcoma (HOS), chronic myelogenous
leukaemia (K-562) and breast adenocarcinoma cell line (MCF-7).
The comparison of IC50 values of the iron complexes 1–6 with
the respective free ligands L1–L6 demonstrates that the complex
formation has no significant impact on the resulting in vitro cyto-
toxicity. From the above results we may draw conclusion that
the ability of the complexes to decrease number of viable cells,
as compared to free ligands L1–L6, is probably caused by the pres-
ence of the organic ligands themselves within the complexes. On
the other hand, the IC50 values obtained on the MCF-7 cell line in
the case of complexes 2 and 3 are even better than those of Cis-
platin and Oxaliplatin (see Table 5), i.e. widely used platinum-
based anticancer drugs. However, it is necessary to note at this
point that partial hydrolysis and dissociation of the iron complexes
should proceed under both in vitro and in vivo conditions. Based on
NMR results (more specifically on the 1H and 13C NMR coordina-
tion shift values) it may by concluded that the appropriate Ln or-
ganic ligand behaves as an N-donor carrier ligand in all the
value (5.82 leff/lB) was determined in this case, and thus, its
geometry can be described as a distorted tetrahedron (based on
57Fe Mössbauer data and DFT geometry optimizations).
As can be seen from the Fig. 4, the monodentate ligand’s coor-
dination to the FeIII ion involving the 5/2 spin state leads to the
formation of stable complex 1a with tetrahedral arrangement of
the coordination sphere (structure 1a), while the monodentate
coordination via the N6 or N7 atom requires, as a first step, to
shift the N6-H proton to any other of non-substituted nitrogen
atoms of the purine moiety (i.e. N1 or N3) – see the structure
1b. However, such rearrangement necessarily leads to the disrup-
tion of the
P-electron delocalization within the purine ring,
which may result in decreasing of the molecular system stability.

ˇ
Z. Trávnícek et al. / Journal of Inorganic Biochemistry 104 (2010) 405–417
412
Table 4
represent a leaving group in in vitro tests. Thus, we believe that un-
der physiological conditions, the hydrolysis and dissociation of the
iron(III) complexes 1–6 should proceed in a much greater rate and
the complexes should be present there in the forms of the [Fe(L-
n)(OH)Cl2], [Fe(Ln)(OH)2Cl] and/or [Fe(Ln)(OH)3] species, and more-
over, a sizable amount of free CDK-inhibitor Ln should also be
present within the system as a consequence of the dissociation
of the complexes. Taking all the above said into consideration, it
may be concluded that just the mentioned processes might be a
cause of the obtained IC50 values for the complexes 1–6, as com-
pared to free ligands L1–L6.
Selected interatomic parameters (Å, °) of the studied complexes calculated on the
B3LYP/6-311+G/Wachter’s level.
1a
1b
1c
1d
1a, exp. 1d, exp.
Fe–Cl1
Fe–Cl2
Fe–Cl3
2.204
2.198
2.209
2.222
2.228
2.217
2.262
2.235
2.226
2.307 2.259a
2.273
2.243
2.392b
Fe–N7
2.074
3.044
2.155
1.955 2.106a
2.182b
2.135b
Fe–N6
Fe–O
–
–
2.092
–
2.536
–
–
2.017
The inhibition of recombinant human cyclin E/Cdk2 kinase by
all the complexes and free ligands was also determined. The re-
sults, summarized in Table 5, proved the ability of all tested com-
plexes to inhibit the activity of this important cell cycle
regulating protein kinase. The most potent complexes 3 and 6
even reached a low submicromolar range of efficiency. To our
best knowledge, the complexes 2–6 have been found to be the
most active cyclin E/CDK2 inhibitors based on transition metal
complexes, and moreover, they were found to be quite compara-
ble to and even better than other CDK inhibitors involving the 6-
benzylaminopurine moiety [9]. It may be concluded that the
coordination of the appropriate L3, L4, L5 and L6 molecules to FeIII
ion led to the formation of complexes for which the determined
inhibitory effect is significantly higher as compared to the free li-
gands. However, it should also be noted here that a partial
hydrolysis and dissociation have to be taken into account in
the case of interpretation of the IC50 values regarding the inhib-
itory activity, similarly as it was mentioned above as for the cyto-
toxicity results.
Cl1–Fe–Cl2 114.06
Cl1–Fe–Cl3 114.37
Cl2–Fe–Cl3 112.67
105.72
119.75
105.95
–
–
–
109.60
103.73
110.68
–
–
–
–
–
103.81
104.25
117.43
89.33
117.46
117.58
84.05
85.06
73.36
162.67
–
120.01
114.28
124.69
95.97
90.89
93.45
–
–
–
–
N7–Fe–Cl1
N7–Fe–Cl2
N7–Fe–Cl3
N6–Fe–Cl1
N6–Fe–Cl2
N6–Fe–Cl3
N6–Fe–N7
Cl1–Fe–O
Cl2–Fe–O
Cl3–Fe–O
N7–Fe–O
104.04
105.06
105.32
–
–
–
–
–
–
–
–
83.50
85.18
91.20
175.10
–
–
–
Td, tetrahedral; TB, trigonal–bipyramidal.
a
The values represent mean values which were determined from known X-ray
structures of complexes bearing Td FeCl3N chromophore and are deposited within
the Cambridge Crystallographic Data Centre (CSD).
The values represent mean values which were determined from known X-ray
structures of complexes bearing TB FeCl3NO chromophore and are deposited within
b
the Cambridge Crystallographic Data Centre (CSD).
2.8. X-ray structure of 2-chloro-6-benzylamino-9-isopropylpurine
Table 5
IC50 values, given together with their standard deviations, assessed by calcein AM
assay of surviving tumour cells and IC50 values for inhibition of Cdk2/cyclin E kinase
(both in lM).
The molecular structure of 2-chloro-6-benzylamino-9-isopro-
pylpurine, i.e. an intermediate for the preparation of L1, L4 and
L5, is shown in Fig. 4. There are two molecules in the crystallo-
graphically independent part. The crystal structure is stabilized
by NaminoꢁHꢀꢀꢀNpurine hydrogen bonds connecting two adjacent
molecules, thus forming centrosymmetric dimers (Fig. 5). Each of
two molecules contains nearly planar benzene (A), pyrimidine (B)
and imidazole (C) ring systems, with maximum deviations from
each plane of 0.005(2) (C14) and 0.003(2) Å (C13A) for rings A,
0.006(2) (C6) and 0.013(2) Å (C4A) for six-membered rings B, and
0.004(2) (C5) and 0.001(2) Å (C4A) for five-membered rings C
[30]. Atoms forming the purine rings (B+C) deviate slightly from
planarity, with the greatest deviation being 0.009(2) and
0.034(2) Å for atom C6, and C5A, respectively. Planes B and C are
nearly coplanar, with a dihedral angle of 0.29(6)°, and 2.79(9)°,
respectively, whilst the dihedral angles between plane A and the
purine ring (B+C) are 84.45(6)°, and 80.50(6)°, respectively. The
Cg1 . . . Cg2, Cg1 . . . Cg3 and Cg2 . . . Cg3 distances are 5.8110(1),
6.7861(2), and 2.0768(1) Å, respectively, where Cg1, Cg2 and Cg3
are the centroids of rings A, B, and C, respectively. The values of
the same centriod-to-centroid distances for the second molecule
are 5.7798(2), 6.8701(2) and 2.0764(1) Å. The torsion angles C6–
N6–C9–C10, C9–N6–C6–C5 and N6–C9–C10–C15, C6–N6–C9–
C10, C9–N6–C6–C5 and N6–C9–C10–C15 are 109.6(2), –178.4(2)
and 148.2(2)° for the first molecule and 108.5(2)°, 171.2(2)° and
ꢁ39.6(3)° for the second one, respectively.
Compound
IC50 (lM)
G-361
HOS
K-562
MCF-7
CDK2/E
[Fe(L1)Cl3] (1)
L1
>50
>100
>50
>100
>50
>100
>50
>100
1.50 0.41
1.05 0.32
0.11 0.02
0.07 0.02
0.05 0.02
0.20 0.11
0.09 0.02
0.20 0.06
0.48 0.08
[Fe(L2)Cl3]ꢀH2O (2)
4
5
1
1
7
8
1
1
12
9
1
5
2
6
7
1
1
2
4
L2
1
[Fe(L3)Cl3]ꢀ2H2O (3) 15
L3 29
[Fe(L4)Cl3]ꢀ2H2O (4) 15
L4 19
[Fe(L5)Cl3]ꢀ2H2O (5) >50
L5
3
10
10
23
20
2
2
4
3
21
21
>50
50
4
3
7
2
2
12
15
43
2
2
8
4
>50
>50
135 18 120 24 118 21 61 16 2.05 0.49
[Fe(L6)Cl3]ꢀ2H2O (6) 33
5
8
41
68
4
5
55
65 11
8
23
12
4
2
0.02 0.01
0.07 0.03
L6
48
FeCl3ꢀ6H2O
Cisplatin
Oxaliplatin
>200
3
7
>200
3
7
>200
5
9
>200
11
18
–
–
–
The tumour cell lines (G-361, human malignant melanoma; HOS, human osteogenic
sarcoma; K-562, human chronic myelogenous leukaemia; MCF-7, human breast
adenocarcinoma) were treated with solution of the tested compounds in the 0.5–
50 lM, 0.5–100 lM, and 0.5–200 lM ranges, respectively, for 72 h.
complexes, i.e. it remains coordinated to the iron(III) cation in a
solution. It should also be mentioned that the NMR experiments
were performed under strictly non-biological conditions, i.e. in sol-
vents such as DMSO-d6, MeOH-d3, DMF-d7, acetone-d6, and thus,
the interpretation given above would not be valid in the case of
experiments running under biological conditions. On the other
hand, it is also highly probable that the three chlorido ligands
It has been found that bond lengths and angles within the two
crystallographically independent molecules of 2-chloro-6-benzyl-
amino-9-isopropylpurine, L, do not differ significantly each other,
and moreover, they are consistent with standard values. For this
reason, they are summarized in Appendix A as the Supplementary
data.

ˇ
Z. Trávnícek et al. / Journal of Inorganic Biochemistry 104 (2010) 405–417
413
Fig. 5. A part of the crystal structure of 2-chloro-6-benzylamino-9-isopropylpurine, i.e. a precursor for the preparation of L1, L4 and L5, showing the formation of
centrosymmetric dimers, NꢁHꢀꢀꢀN hydrogen bonds (dashed lines) and
P
–P
stacking interactions (orange dashed lines). Atoms of the two crystallographically independent
molecules within the unit cell are numbered only.
(N6)H), 7.67 (s, 1H, (C8)H), 7.11 (s, 1H, (C15)H), 6.92 (dd, 1H,
(C13)H, Ja = 8.1 Hz, Jb = 2.1 Hz), 6.69 (d,1H, (C12)H, J = 8.1 Hz),
5.43 (d, 1H, (N2)H, J = 8.1 Hz), 4.53 (s, 1H, (O1)H), 4.58 (s, 2H,
(C9)H), 4.62 (sep, 1H, (C16)H, J = 6.8 Hz), 4.01 (m, 1H, (C19)H),
3.69 (m, 2H, (C20)H), 2.20 (s, 3H, (C23)H), 1.77 (sep, 1H, (C21)Ha,
3. Experimental
3.1. Materials
FeCl3ꢀ6H2O, K2CO3, 2,6-dichlorpurine, i-propylbromide, benzyl-
amine, 2-hydroxy-5-methyl-benzylamine, 3-hydroxy-benzyl-
amine, 3,5-dihydroxy-benzylamine and solvents used were
purchased from Fluka or Aldrich Co. and were used without further
purification.
J = 7.3 Hz), 1.67 (sep, 1H, (C21)Hb, J = 7.3 Hz), 1.53 (d, 6H,
(C17,18)H, J = 6.8 Hz), 1.00 (t, 3H, (C22)H, J = 7.3 Hz). 1H NMR
(DMF-d7, ppm): 10.21 (bs, 1H, (O2)H), 7.58 (s, 1H, (N6)H), 7.82
(s, 1H, (C8)H), 7.12 (d, 1H, (C15)H, J = 2.1 Hz), 6.91 (dd, 1H,
(C13)H, Ja = 8.1 Hz, Jb = 2.1 Hz), 6.75 (d, 1H, (C12)H, J = 8.1 Hz),
5.86 (d, 1H, (N2)H, J = 8.2 Hz), 4.80 (bs, 1H, (O1)H), 4.67 (s, 2H,
(C9)H), 4.62 (sep, 1H, (C16)H, J = 6.8 Hz), 4.01 (m, 1H, (C19)H),
3.66 (m, 2H, (C20)H), 2.18 (s, 3H, (C23)H), 1.77 (sep, 1H, (C21)Ha,
J = 7.3 Hz), 1.65 (sep, 1H, (C21)Hb, J = 7.3 Hz), 1.53 (d, 6H,
(C17,18)H, J = 6.8 Hz), 0.97 (t, 3H, (C22)H, J = 7.3 Hz). 13C NMR
(DMSO-d6, ppm): 159.28 (C2), 154.93 (C6), 153.21 (C11), 151.00
(C4), 135.74 (C8), 129.80 (C14), 128.62 (C15), 127.53 (C13),
126.40 (C10), 115.70 (C12), 114.08 (C5), 63.48 (C20), 54.52 (C19),
46.28 (C16), 24.37 (C21), 22.51 (C17,18), 20.67 (C23), 11.09 (C22).
L3: Yield: ꢂ65%; Anal. Calc. for C19H26N6O2: C, 61.6; H, 7.0; N,
22.7. Found: C, 61.4; H, 7.2; N, 22.5%. 1H NMR (DMSO-d6, ppm):
9.21 (s, 1H, (O2)H), 7.77 (s, 1H, (C8)H), 7.66 (bs, 1H, (N6)H), 7.06
(t, 1H, (C14)H, J = 8.1 Hz), 6.76 (dd, 2H, (C13,15)H, Ja = 7.9 Hz,
Jb = 1.5 Hz), 6.58 (tt, 1H, (C11)H, Ja = 8.2 Hz, Jb = 1.7 Hz), 5.80 (d,
1H, (N2)H, J = 8.2 Hz), 4.59 (bs, 1H, (O1)H), 4.59 (s, 2H, (C9)H),
4.53 (sep, 1H, (C16)H, J = 6.7 Hz), 3.80 (m, 1H, (C19)H), 3.46 (m,
2H, (C20)H), 1.59 (m, 1H, (C21)Ha), 1.47 (dd, 6H, (C17,18)H,
Ja = 6.7 Hz, Jb = 2.2 Hz), 1.43 (m, 1H, (C21)Hb), 0.84 (t, 3H, (C22)H,
J = 7.3 Hz). 1H NMR (MeOH-d3, ppm): 7.77 (s, 1H, (C8)H), 7.12 (t,
1H, (C14)H, J = 7.8 Hz), 6.81 (m, 2H, (C13,15)H), 6.65 (d, 1H,
(C11)H, J = 8.2 Hz), 4.66 (s, 2H, (C9)H), 4.64 (m, 1H, (C16)H, 3.94
(m, 1H, (C19)H, 3.61 (m, 2H, (C20)H), 1.67 (m, 1H, (C21)Ha), 1.51
(m, 1H, (C21)Hb), 1.55 (d, 6H, (C17,18)H, J = 6.7 Hz), 0.95 (t, 3H,
(C22)H, J = 7.3 Hz). 13C NMR (DMSO-d6, ppm): 160.27 (C2), 158.55
(C12), 155.68 (C6), 151.77 (C4), 143.15 (C10), 135.68 (C8), 129.69
(C11), 118.77 (C15), 115.05 (C14), 115.70 (C12), 114.90 (C5),
114.18 (C13), 64.52 (C20), 55.29 (C19), 46.81 (C16), 43.77 (C9)
24.91 (C21), 22.34 (C17,18), 10.99 (C22).
3.2. Preparation of ligands L1–L6
The CDK inhibitors, used as ligands L1–L6 in this study, were
prepared by a slightly modified method described earlier [31]. A
general pathway for the preparation of the free ligands is depicted
in Scheme 3. The purity of the compounds was checked by elemen-
tal analyses, 1H and 13C NMR spectroscopy and thin-layer chroma-
tography (TLC). The splitting of proton resonances in the reported
1H spectra is defined as s = singlet, d = doublet, t = triplet, q = quar-
tet, qui = quintet, sxt = sextet, sep = septet, bs = broad signal,
dd = doublet of doublets, tt = triplet of triplets, m = multiplet.
L1: Yield: ꢂ60%; Anal. Calc. for C18H23N6Cl1: C, 60.3; H, 6.4; N,
23.4. Found: C, 60.0; H, 6.6; N, 23.2%. 1H NMR (DMF-d7, ppm):
7.67 (bs, 1H, (N6)H), 7.79 (s, 1H, (C8)H), 7.45 (d, 2H, (C11,15)H,
J = 7.2 Hz), 7.31 (tt, 2H, (C12,14)H, Ja = 7.5 Hz, Jb = 1.7 Hz), 7.22 (tt,
1H, (C13)H, Ja = 7.3 Hz, Jb = 1.8 Hz), 6.41 (t, 1H, (N2)H, J = 5.8 Hz),
4.80 (s, 2H, (C9)H), 4.63 (sep, 1H, (C16)H, J = 6.8 Hz), 3.72 (t, 2H,
(C21)H, J = 6.6 Hz), 3.50 (qui, 2H, (C19)H, J = 6.6 Hz), 2.07 (qui,
2H, (C20)H, J = 6.6 Hz), 1.53 (d, 6H, (C17,18)H, J = 6.8 Hz). 13C
NMR (DMF-d7, ppm): 160.27 (C2), 155.73 (C6), 151.92 (C4),
135.80 (C8), 141.82 (C10), 128.80 (C12,14), 128.22 (C11,15),
127.19 (C13), 115.06 (C5), 60.12 (C20), 46.82 (C16), 44.11 (C19),
39.64 (C9), 33.70 (C21), 22.40 (C17,18). TLC: one spot.
L2: Yield: ꢂ70%; Anal. Calc. for C20H28N6O2: C, 62.5; H, 7.3; N,
21.9. Found: C, 62.5; H, 7.5; N, 21.9%. 1H NMR (DMSO-d6, ppm):
9.68 (s, 1H, (O2)H), 7.79 (s, 1H, (C8)H), 7.43 (bs, 1H, (N6)H), 6.99
(d, 1H, (C15)H, J = 2.1 Hz), 6.85 (dd, 1H, (C13)H, Ja = 8.1 Hz,
Jb = 2.1 Hz), 6.67 (d, 1H, (C12)H, J = 8.1 Hz), 5.84 (d, 1H, (N2)H,
J = 8.1 Hz), 4.59 (s, 1H, (O1)H), 4.53 (sep, 1H, (C16)H, J = 6.8 Hz),
4.58 (s, 2H, (C9)H), 3.82 (m, 1H, (C19)H), 3.48 (qui, 1H, (C20)Ha,
J = 5.3 Hz), 3.41 (m, 1H, (C20)Hb), 2.14 (s, 3H, (C23)H), 1.62 (sep,
1H, (C21)Ha, J = 7.3 Hz), 1.48 (m, 1H, (C21)Hb), 1.47 (dd, 6H,
(C17,18)H, Ja = 6.8 Hz, Jb = 2.3 Hz), 0.86 (q, 3H, (C22)H, J = 7.3 Hz).
1H NMR (MeOH-d3, ppm): 7.77 (s, 1H, (C8)H), 7.07 (s, 1H,
(C15)H), 6.92 (d, 1H, (C13)H, J = 8.2 Hz), 6.70 (d, 1H, (C12)H,
J = 8.2 Hz), 4.61 (s, 2H, (C9)H), 4.66 (m, 1H, (C16)H), 4.00 (m, 1H,
(C19)H), 3.66 (dd, 2H, (C20)H, Ja = 5.1 Hz, Jb = 2.0 Hz), 2.21 (s, 3H,
(C23)H), 1.74 (sep, 1H, (C21)Ha, J = 7.3 Hz), 1.61 (m, 1H, (C21)Hb),
1.55 (d, 6H, (C17,18)H, J = 6.8 Hz), 1.01 (t, 3H, (C22)H, J = 7.3 Hz).
1H NMR (acetone-d6, ppm): 10.51 (s, 1H (O2)H), 7.24 (bs, 1H,
L4: Yield: ꢂ60%; Anal. Calc. for C19H26N6O1: C, 64.4; H, 7.4; N,
23.7. Found: C, 64.5; H, 7.5; N, 23.4%. 1H NMR (DMSO-d6, ppm):
7.72 (bs,1H, (N6)H), 7.77 (bs, 1H, (C8)H), 7.36 (dd, 2H, (C11,15)H,
Ja = 7.0 Hz, Jb = 1.7 Hz), 7.28 (tt, 2H, (C12,14)H, Ja = 7.3 Hz,
Jb = 1.6 Hz), 7.19 (tt, 1H, (C13-H), Ja = 7.0 Hz, Jb = 1.7 Hz), 5.81 (d,
1H, (N2)H, J = 8.2 Hz), 4.65 (bs, 2H, (C9)H), 4.52 (sep, 1H, (C16-
H), J = 6.9 Hz), 4.52 (d, 1H, (O1)H, J = 5.4 Hz), 3.80 (m, 1H,
(C19)H), 3.46 (sxt, 1H, (C20)Ha, J = 5.4 Hz), 3.37 (sxt, 1H, (C20)Hb,
J = 5.4 Hz), 1.60 (m, 1H, (C21)Ha), 1.46 (dd, 6H, (C17,18)H,
Ja = 6.9 Hz, Jb = 2.1 Hz), 1.44 (m, 1H, (C21)Hb), 0.84 (t, 3H, (C22)H,
J = 7.3 Hz). 1H NMR (MeOH-d3, ppm): 7.76 (s, 1H, (C8)H), 7.66 (s,
1H, (N6)H), 7.36 (dd, 2H, (C11,15)H, Ja = 7.2 Hz, Jb = 1.7 Hz), 7.29

ˇ
Z. Trávnícek et al. / Journal of Inorganic Biochemistry 104 (2010) 405–417
414
ppm): 9.00 (bs, 2H, (O2,3)H), 7.76 (s, 1H, (C8)H), 7.58 (bs, 1H,
(N6)H), 6.20 (d, 2H, (C11,15)H, J = 2.2 Hz), 6.04 (t, 1H, (C13)H,
J = 2.2 Hz), 5.79 (d, 1H, (N2)H, J = 8.2 Hz), 4.54 (s, 1H, (O1)H), 4.54
(sep, 1H, (C16)H, J = 6.8 Hz), 4.54 (s, 2H, (C9)H), 3.81 (m, 1H,
(C19)H), 3.48 (m, 1H, (C20)Ha), 3.39 (m, 1H, (C20)Hb), 1.61 (sep,
1H, (C21)Ha, J = 7.3 Hz), 1.47 (dd, 6H, (C17,18)H, Ja = 6.8 Hz,
Jb = 2.0 Hz), 1.46 (sep, 1H, (C21)Hb, J = 7.3 Hz), 0.86 (t, 3H, (C22)H,
J = 7.3 Hz). 1H NMR (MeOH-d3, ppm): no visible (2H, (O2,3)H),
7.77 (s, 1H, (C8)H), 7.71 (bs, 1H, (N6)H), 6.31 (d, 2H, (C11,15)H,
J = 2.2 Hz), 6.14 (t, 1H, (C13)H, J = 2.2 Hz), no visible (1H, (N2)H),
no visible (1H, (O1)H), 4.61 (sep, 1H, (C16)H, J = 6.8 Hz), 4.60 (s,
2H, (C9)H), 3.95 (m, 1H, (C19)H), 3.62 (m, 2H, (C20)H), 1.69 (sep,
1H, (C21)Ha, J = 7.3 Hz), 1.55 (dd, 6H, (C17,18)H, Ja = 6.8 Hz,
Jb = 1.5 Hz), 1.54 (sep, 1H, (C21)Hb, J = 7.3 Hz), 0.97 (t, 3H, (C22)H,
J = 7.3 Hz). 13C NMR (DMSO-d6, ppm): 159.50 (C2), 158.64
(C12,14), 155.03 (C6), 151.19 (C4), 143.29 (C10), 135.38 (C8),
114.10 (C5), 105.47 (C11,15), 101.14 (C13), 63.53 (C20), 54.53
(C19), 46.15 (C16), 43.07 (C9), 24.36 (C21), 22.56 (C17), 22.47
(C18), 11.14 (C22).
3.3. Synthesis of Fe(III) complexes
A general pathway for the preparation of the complexes is as
follows: totally 1 mmol of FeCl3ꢀH2O (0.27 g) in ethanol (15 mL)
was added to 1 mmol of Ln in ethanol (25 mL) and stirred (or kept
in ultrasonic bath) at the temperature of 65 °C for approximately
3 h. After slow cooling, the reaction mixture was taken to vacuum
desiccator to dry out. After few days, the orange-brown powder
was washed by small amount of diethylether and dried again in
desiccator with P4O10 as a drying medium.
1: Yield: ꢂ60%. Anal. Calc. for C18H23N6Fe1Cl4: C, 41.6; N, 16.2; H,
4.4; Fe, 10.8. Found: C, 41.2; N, 16.0; H, 4.4; Fe, 10.6%. IR (cmꢁ1),
s = strong, m = medium, w = weak: 3400w, 3276w, 3105w,
2935w, 1638s (C@N), 1581m (C@C), 1535m, 1496m, 1452m
(C@C), 1374m, 1351m, 1318w, 1280m, 1208m, 1179w, 1133m,
1080m, 1060m, 1027w, 884m, 763w, 753w, 699w, 658w, 640w,
602m, 484w. ES + MS (m/z): 359 [L1+H]+, 137 [adenine+H]+. 1H
NMR (DMF-d7, ppm): 9.48 (bs, 1H, (N6)H), 8.66 (bs, 1H, (C8)H),
7.53 (s, 2H, (C11,15)H), 7.37 (t, 2H, (C12,14)H, J = 7.2 Hz), 7.30 (t,
1H, (C13)H, J = 7.3 Hz), 5.38 (s, 1H, (N2)H), 4.90 (s, 2H, (C9)H),
4.80 (s, 1H, (C16)H), 3.80 (s, 2H, (C21)H), 3.62 (qui, 2H, (C19)H,
J = 6.6 Hz), 2.11 (s, 2H, (C20)H), 1.62 (s, 6H, (C17,18)H). 13C NMR
(DMF-d7, ppm): 159.78 (C2), 155.18 (C6), 151.08 (C4), 143.24
(C8), 139.64 (C10), 129.54 (C12,14), 127.96 (C11,15), 128.58
(C13), 107.56 (C5), 59.38 (C20), 44.42 (C16), 42.42 (C19), 40.34
(C9), 20.21 (C17,18).
Scheme 3. The three-step synthesis of 6-benzylamino-9-isopropylpurine deriva-
tives, L1–L6, used as organic ligands for the preparation of the FeIII complexes 1–6.
(tt, 2H, (C12,14)H, Ja = 7.5 Hz, Jb = 1.6 Hz), 7.21 (tt, 1H, (C13)H,
Ja = 7.2 Hz, Jb = 1.7 Hz), 5.55 (d, 1H, (N2)H), J = 7.8 Hz), 4.80 (s, 1H,
(O1)H), 4.73 (d, 2H, (C9)H, J = 5.7 Hz), 4.62 (sep, 1H, (C16)H,
J = 6.9 Hz), 3.95 (m, 1H, (C19)H), 3.61 (m, 2H, (C20)H), 1.68 (m,
1H, (C21)Ha), 1.53 (m, 1H, (C21)Hb), 1.53 (dd, 6H, (C17,18)H,
Ja = 6.9 Hz, Jb = 1.7 Hz) 0.95 (t, 3H, (C22)H, J = 7.5 Hz). 1H NMR
(DMF-d7, ppm): 7.72 (bs,1H, (N6)H), 7.77 (s, 1H, C8-H), 7.45 (d,
2H, (C11,15)H, J = 7.2 Hz), 7.30 (tt, 2H, (C12,14)H, Ja = 7.5 Hz,
Jb = 1.6 Hz), 7.22 (tt, 1H, (C13)H, Ja = 7.0 Hz, Jb = 1.7 Hz), 5.86 (d,
1H, (N2)H, J = 8.2 Hz), 4.84 (s, 1H, (O1)H), 4.79 (t, 2H, (C9)H,
J = 5.2 Hz), 4.61 (sep, 1H, (C16)H, J = 6.9 Hz), 3.98 (m, 1H,
(C19)H), 3.66 (m, 1H, (C20)Ha), 3.60 (m, 1H, (C20)Hb), 1.73 (m,
1H, (C21)Ha), 1.52 (d, 6H, (C17,18)H, J = 6.9 Hz), 1.60 (m, 1H,
(C21)Hb), 0.94t, 3H,(C22)H, J = 7.3 Hz). 13C NMR (DMF-d7, ppm):
160.31 (C2), 155.70 (C6), 151.85 (C4), 141.72 (C10), 135.74 (C8),
128.78 (C12,14), 128.24 (C11,15), 127.20 (C13), 114.96 (C5),
64.56 (C20), 55.33 (C19), 46.84 (C16), 43.93 (C9), 24.95 (C21),
22.38 (C17,18), 11.03 (C22).
2: Yield: 45%. Anal. Calc. for C20H30N6O3Fe1Cl3: C, 42.6; N, 14.9;
H, 5.3; Fe, 9.9. Found: C, 42.6; N, 14.6; H, 5.6; Fe, 10.1%. IR (cmꢁ1):
3406w, 2968w, 2934w, 2877w, 1671s, 1635s (C@N), 1572 m
(C@C), 1512m, 1461m (C@C), 1410m, 1373m, 1346w, 1316m,
1261m, 1214m (Car–O), 1149m, 1120m, 1049m (Calif–O), 925m,
886w, 819w, 769w, 717m, 642w, 558w. ES + MS (m/z): 549 [{M–
(H2O)}+H]+, 513 [Fe(L2)Cl2+H]+, 385 [L2+H]+, 137 [adenine+H]+. 1H
NMR (DMSO-d6, ppm): 9.57 (bs, 1H, (O2)H), 8.17 (bs, 1H, (C8)H),
7.69 (bs, 1H, (N6)H), 6.99 (s, 1H, (C15)H), 6.90 (d, 1H, (C13)H,
J = 8.1 Hz), 6.72 (d, 1H, (C12)H, J = 8.1 Hz), 5.11 (bs, 1H, (O1)H),
4.59 (sep, 1H, (C16)H, J = 6.8 Hz), 4.56 (s, 2H, (C9)H), 3.84 (m, 1H,
(C19)H), 3.61 (m, 2H, (C20)H), 2.16 (s, 3H, (C23)H), 1.62 (sep, 1H,
(C21)Ha, J = 7.3 Hz), 1.54 (m, 1H, (C21)Hb), 1.50 (dd, 6H,
(C17,18)H, Ja = 6.8 Hz, Jb = 1.8 Hz), 0.87 (q, 3H, (C22)H, J = 7.3 Hz).
1H NMR (MeOH-d3, ppm): 8.18 (s, 1H, (N6)H), 8.11 (s, 1H, (C8)H),
7.12 (s, 1H, (C15)H), 7.02 (d, 1H, (C13)H, J = 8.2 Hz), 6.77 (d, 1H,
(C12)H, J = 8.2 Hz), 6.39 (s, 1H, (N2)H), 5.55 (s, 1H, (O1)H), 4.69
(s, 2H, (C9)H), 4.06 (m, 1H, (C19)H), 3.73 (s, 2H, (C20)H), 2.26 (s,
3H, (C23)H), 1.77 (sep, 1H, (C21)Ha, J = 7.3 Hz), 1.67 (m, 1H,
(C21)Hb), 1.62 (d, 6H, (C17,18)H, J = 6.8 Hz), 1.04 (t, 3H, (C22)H,
L5: Yield: ꢂ65%; Anal. Calc. for C17H22N6O1: C, 62.5; H, 6.8; N,
25.8. Found: C, 62.8; H, 6.7; N, 25.6%. 1H NMR (DMSO-d6, ppm):
7.77 (bs, 1H, (N6)H), 7.79 (s, 1H, (C8)H), 7.36 (d, 2H, (C11,15)H,
J = 7.7 Hz), 7.28 (tt, 2H, (C12,14)H, Ja = 7.3 Hz, Jb = 2.2 Hz), 7.19 (tt,
1H, (C13)H, Ja = 7.3 Hz, Jb = 2.2 Hz), 6.14 (t, 1H, (N2)H, J = 5.6 Hz),
5.11 (s, 1H, (O1)H), 4.64 (t, 1H, (N6)H, J = 5.3 Hz), 4.54 (sep, 1H,
(C16)H, J = 6.8 Hz), 3.51 (q, 2H, (C20)H, J = 6.0 Hz), 3.32 (q, 2H,
(C19)H, J = 6.0 Hz), 1.46 (d, 6H, C17,18)H, J = 6.8 Hz). 1H NMR
(MeOH-d3, ppm): 7.78 (s, 1H,(C8)H), 7.36 (dd, 2H, (C11,15)H,
Ja = 7.7 Hz, Jb = 1.7 Hz), 7.29 (tt, 2H, (C12,14)H, Ja = 7.3 Hz,
Jb = 2.2 Hz), 7.21 (tt, 1H, (C13)H, Ja = 7.3 Hz, Jb = 2.2 Hz), 4.74 (s,
2H, (C9)H), 4.64 (sep, 1H, (C16)H, J = 6.8 Hz), 3.69 (t, 2H, (C20)H,
J = 5.8 Hz), 3.50 (t, 2H, (C19)H, J = 5.8 Hz), 1.54 (d, 6H, (C17,18)H,
J = 6.8 Hz). 13C NMR (DMSO-d6, ppm): 159.00 (C2), 154.42 (C6),
150.72 (C4), 140.65 (C10), 134.95 (C8), 127.95 (C11,15), 126.33
(C13), 113.61 (C5), 60.32 (C20), 45.52 (C16), 43.62 (C19), 42.66
(C9), 21.96 (C17,18).
L6: Yield: 75%, Anal. Calc. for C19H26N6O3: C, 59.0, H, 6.8, N, 21.8.
Found: C, 58.8, H, 6.8, N, 21.6%. TLC: one spot. 1H NMR (DMSO-d6,

ˇ
Z. Trávnícek et al. / Journal of Inorganic Biochemistry 104 (2010) 405–417
415
J = 7.3 Hz). 1H NMR (acetone-d6, ppm): 8.47 (bs, 1H, (N6)H), 8.12
(bs, 1H, (C8)H), 7.16 (s, 1H, (C15)H), 6.96 (s, 1H, (C13)H), 6.78 (s,
1H, (C12)H), 6.47 (bs, 1H, (N2)H), 5.22 (bs, 1H, (O1)H), 4.80 (s,
2H, (C9)H), 4.68 (m, 1H, (C16)H), 4.09 (m, 1H, (C19)H), 3.76 (s,
2H, (C20)H), 2.21 (s, 3H, (C23)H), 1.81 (m, 1H, (C21)Ha), 1.72 (m,
1H, (C21)Hb), 1.65 (s, 6H, (C17,18)H), 1.03 (s, 3H, (C22)H). 1H
NMR (DMF-d7, ppm): 9.88 (bs, 1H, (O2)H), 8.35 (s, 1H, (N6)H),
8.19 (bs, 1H, (C8)H), 7.16 (s, 1H, (C15)H), 6.96 (d, 1H, (C13)H,
J = 8.1 Hz), 6.80 (d, 1H, (C12)H, J = 8.1 Hz), 5.90 (bs, 1H, (N2)H),
5.25 (bs, 1H, (O1)H), 4.72 (bs, 2H, (C9)H), 4.72 (m, 1H, (C16)H),
4.04 (m, 1H, (C19)H), 3.66 (m, 2H, (C20)H), 2.19 (s, 3H, (C23)H),
1.77 (sep, 1H, (C21)-Ha, J = 7.3 Hz), 1.62 (sep, 1H, (C21)Hb,
J = 7.3 Hz), 1.58 (d, 6H, (C17,18)H, J = 6.8 Hz), 0.96 (t, 3H, (C22)H,
J = 7.3 Hz). 13C NMR (DMSO-d6, ppm): 158.43 (C2), 156.70 (C6),
153.25 (C11), 152.01 (C4), 139.25 (C8), 129.47 (C14), 129.02
(C15), 127.52 (C13), 124.35 (C10), 115.43 (C12), 111.50 (C5),
62.69 (C20), 54.73 (C19), 47.53 (C16), 24.15 (C21), 22.27
(C17,18), 20.72 (C23), 10.91 (C22).
3: Yield: ꢂ50%. Anal. Calc. for C19H30N6O4Fe1Cl3: C, 40.2; N, 14.8;
H, 5.3; Fe, 9.9. Found: C, 40.5; N, 14.9; H, 5.4; Fe, 9.9%. IR (cmꢁ1):
3278w, 3130w, 2968w, 2935w, 2877w, 1667s, 1637s (C@N),
1589s (C@C), 1572m, 1529m, 1482m, 1459m (C@C), 1373m,
1317m, 1282m, 1257m, 1215m (Car–O), 1157m, 1135m, 1051m
(Calif–O), 998m, 920w, 884m, 863w, 781m, 768m, 728m, 694w,
671m, 639w, 602w, 446w. ES + MS (m/z): 568 [M+H]+, 461
[Fe(L3)Cl+H]+, 371 [L3+H]+, 137 [adenine+H]+. 1H NMR (DMSO-d6,
ppm): 9.36 (bs, 1H, (O2)H), 8.20 (bs, 1H, (C8)H), 7.11 (t, 1H,
(C14)H, J = 7.9 Hz), 6.77 (m, 2H, (C13,15)H), 6.64 (d, 1H, (C11)H,
J = 7.9 Hz), 5.14 (bs, 1H, (O1)H), 4.63 (s, 2H, (C9)H), 4.60 (qui, 1H,
(C16)H, J = 6.7 Hz), 3.84 (s, 1H, (C19)H), 3.43 (m, 2H, (C20)H),
1.61 (m, 1H, (C21)Ha), 1.49 (d, 6H, (C17,18)H, J = 6.7 Hz), 1.43 (m,
1H, (C21)-Hb), 0.87 (m, 3H, (C22)H). 1H NMR (MeOH-d3, ppm):
8.15 (bs, 1H, (C8)H), 7.21 (t, 1H, (C14)H, J = 7.8 Hz), 6.87 (m, 2H,
(C13,15)H), 6.76 (d, 1H, (C11)H, J = 8.2 Hz), 4.68 (s, 2H, (C9)H),
4.65 (m, 1H, (C16)H), 4.04 (m, 1H, (C19)H), 3.70 (m, 2H, (C20)H),
1.73 (m, 1H, (C21)Ha), 1.66 (m, 1H, (C21)Hb), 1.63 (d, 6H,
(C17,18)H, J = 6.7 Hz), 0.96 (t, 3H, (C22)H, J = 7.3 Hz).
63.51 (C20), 55.30 (C19), 48.05 (C16), 44.60 (C9), 24.48 (C21),
22.36 (C17,18), 10.82 (C22).
5: Yield: ꢂ40%. Anal. Calc. for C17H26N6O3Fe1Cl3: C, 39.0; N, 16.1;
H, 5.0; Fe, 10.7. Found: C, 39.0; N, 15.7; H, 4.7; Fe, 10.6%. IR (cmꢁ1):
3394w, 3124w, 2977w, 2934w, 2878w, 1635s (C@N), 1574s (C@C),
1526m, 1496m, 1453m (C@C), 1409m, 1373m, 1352m, 1316m,
1277m, 1213m, 1133m, 1060m (Calif–O), 1027m, 882w, 769w,
748m, 726m, 699m, 670w, 638w, 599w, 485w. ES + MS (m/z):
525 [M+H]+, 454 [Fe(L5)Cl2+H]+, 417 [Fe(L5)Cl+H]+, 327 [L5 + H]+,
137 [adenine+H]+. 1H NMR (DMSO-d6, ppm): 8.77 (bs, 1H, (N6)H),
8.18 (bs, 1H, (C8)H), 7.38 (d, 2H, (C11,15)H, J = 7.7 Hz), 7.32 (t,
2H, (C12,14)H, J = 7.3 Hz), 7.25 (t, 1H, (C13)H, J = 7.3 Hz), 6.79 (bs,
1H, (N2)H), 5.19 (bs, 1H, (O1)H), 4.67 (s, 1H, (N6)H), 4.61 (qui,
1H, (C16)H, J = 6.8 Hz), 3.52 (m, 2H, (C20)H), 3.37 (m, 2H,
(C19)H), 1.48 (d, 6H, C17,18)H, J = 6.8 Hz). 1H NMR (MeOH-d3,
ppm): 8.18 (bs, 1H,(C8)H), 8.18 (bs, 1H, (N6)H), 7.44 (d, 2H,
(C11,15)H, J = 7.7 Hz), 7.41 (t, 2H, (C12,14)H, J = 7.3 Hz), 7.34 (t,
1H, (C13)H, J = 7.3 Hz), 4.75 (s, 2H, (C9)H), 4.63 (m, 1H, (C16)H),
3.75 (m, 2H, (C20)H), 3.60 (m, 2H, (C19)H), 1.61 (d, 6H,
(C17,18)H, J = 6.8 Hz).
6: Yield: 50%. Anal. Calc. for C19H30N6O5Fe1Cl3: C, 39.1; N, 14.4;
H, 5.1; Fe, 9.6. Found: C, 39.2; N, 14.0; H, 5.4; Fe, 9.2%. IR (cmꢁ1):
3412w, 3268w, 3121w, 2969w, 2935w, 2878w, 1638s (C@N),
1576m (C@C), 1521m, 1459m (C@C), 1373m, 1350m, 1285m,
1214m (Car–O), 1161m, 1049m (Calif–O), 1013m, 998w, 887w,
843w, 767m, 637m, 600,w 519w. ES + MS (m/z): 568 [{M–
(H2O)}+H]+, 477 [Fe(L6)Cl+H]+, 387 [L6+H]+, 137 [adenine+H]+. 1H
NMR (DMSO-d6, ppm): 9.15 (bs, 2H, (O2,3)H), 8.13 (bs, 1H,
(C8)H), 7.70 (bs, 1H, (N6)H), 6.19 (d, 2H, (C11,15)H, J = 1.5 Hz),
6.08 (s, 1H, (C13)H), 5.01 (bs, 1H, (O1)H), 4.59 (sep, 1H, (C16)H,
J = 6.8 Hz), 4.50 (s, 2H, (C9)H), 3.83 (bs, 1H, (C19)H), 3.61(m, 1H,
(C20)Ha), 3.52 (m, 1H, (C20)Hb), 1.61 (m, 1H, (C21)Ha), 1.53 (m,
1H, (C21)Hb), 1.50 (d, 6H, (C17,18)H, J = 6.8 Hz), 0.87 (m, 3H,
(C22)H). 1H NMR (MeOH-d3, ppm): 8.15 (bs, 1H, (C8)H), 6.34 (d,
2H, (C11,15)H, J = 2.2 Hz), 6.25 (t, 1H, (C13)H, J = 2.2 Hz), 4.66 (bs,
2H, (C9)H), 4.03 (m, 1H, (C19)H), 3.71 (m, 2H, (C20)H), 1.75 (sep,
1H, (C21)Ha, J = 7.3 Hz), 1.65 (m, 1H, (C21)Hb), 1.63 (d, 6H,
(C17,18)H, J = 6.8 Hz), 0.98 (t, 3H, (C22)H, J = 7.3 Hz).
4: Yield: 70%. Anal. Calc. for C19H30N6O3Fe1Cl3: C, 41.4; N, 15.2;
H, 5.4; Fe, 10.1. Found: C, 41.6; N, 15.3; H, 5.4; Fe, 9.8%. IR (cmꢁ1):
3412w, 3274w, 3084w, 3030w, 2966w, 2932w, 2876w, 1671s,
1634s (C@N), 1571m (C@C), 1527m, 1494m, 1453m (C@C),
1373m, 1352m, 1316m, 1285m, 1253m, 1215w, 1133m, 1051m
(Califr–O), 1028m, 967m, 885w, 842w, 769w, 746m, 698m, 671m,
638w, 599w, 485w. ES + MS (m/z): 483 [Fe(L4)Cl2+H]+, 445
[Fe(L4)Cl+H]+, 355 [L4+H]+, 137 [adenine+H]+. 1H NMR (DMSO-d6,
ppm): 9.34 (bs,1H, (N6)H), 8.27 (bs, 1H, (C8)H), 7.38 (d, 2H,
(C11,15)H, J = 7.0 Hz), 7.34 (t, 2H, (C12,14)H, J = 7.3 Hz), 7.26 (t,
1H, (C13-H), J = 7.0 Hz), 5.23 (bs, 1H, (N2)H), 4.72 (bs, 2H, (C9)H),
4.60 (qui, 1H, (C16-H), J = 6.9 Hz), 4.60 (bs, 1H, (O1)H), 3.85 (s,
1H, (C19)H), 1.62 (m, 1H, (C21)Ha), 1.49 (d, 6H, (C17,18)H,
J = 6.8 Hz), 1.45 (m, 1H, (C21)Hb), 0.86 (t, 3H, (C22)H, J = 7.2 Hz).
1H NMR (MeOH-d3, ppm): 8.15 (bs, 1H, (C8)H), 8.04 (bs, 1H,
(N6)H), 7.44 d, 2H, (C11,15)H, J = 7.2 Hz), 7.41 (t, 2H, (C12,14)H,
J = 7.5 Hz), 7.34 (t, 1H, (C13)H, J = 7.2 Hz), 4.78 (bs, 2H, (C9)H),
4.72 (s, 1H, (C16)H), 4.04 (m, 1H, (C19)H), 3.70 (d, 2H, (C20)H,
J = 4.2 Hz), 1.74 (m, 1H, (C21)Ha), 1.67 (m, 1H, (C21)Hb), 1.63 (d,
6H, (C17,18)H, J = 6.7 Hz) 1.02 (t, 3H, (C22)H, J = 7.5 Hz). 1H NMR
(DMF-d7, ppm): 9.25 (bs,1H, (N6)H), 8.30 (bs, 1H, C8-H), 7.51 (d,
2H, (C11,15)H, J = 7.2 Hz), 7.37 (t, 2H, (C12,14)H, J = 7.5 Hz), 7.30
(t, 1H, (C13)H, J = 7.0 Hz), 5.35 (bs, 1H, (N2)H), 4.87 (s, 2H,
(C9)H), 4.74 (qui, 1H, (C16)H, J = 6.9 Hz), 4.03 (m, 1H, (C19)H),
3.79 (m, 1H, (C20)Ha), 3.65 (m, 1H, (C20)Hb), 1.75 (m, 1H,
(C21)Ha), 1.60 (d, 6H, (C17,18)H, J = 6.7 Hz), 1.60 (m, 1H,
(C21)Hb), 0.95t, 3H,(C22)H, J = 7.3 Hz). 13C NMR (DMF-d7, ppm):
159.12 (C2), 156.11 (C6), 153.61 (C4), 143.52 (C10), 139.35 (C8),
128.97 (C12,14), 128.18 (C11,15), 127.70 (C13), 108.70 (C5),
3.4. General methods
Elemental analyses (CHN) were performed on a Flash EA 1112
analyzer (ThermoFinnigan) and the content of Fe was determined
by chelatometric titration with sulphosalicylic acid as an indicator.
Infrared spectra were obtained on a FT-IR Nexus 670 spectrometer
(ThermoNicolet) by means of polyethylene discs or the Nujol tech-
nique (in the region 150–600 cmꢁ1) and the KBr technique (in the
region 400–4000 cmꢁ1). The transmission 57Fe Mössbauer spectra
were collected at 300 and 5 K using a Mössbauer spectrometer in
constant acceleration mode with a 57Co(Rh) source and cryomag-
netic system (Oxford Instruments). The isomer shift values were
calibrated to metallic a
-iron. 1H and 13C NMR spectra of the ligands
and complexes were measured on a Bruker 300 Avance spectrom-
eter with an inverse probe tuned at 300.13 MHz for 1H (concentra-
tions ꢂ0.5 mg mLꢁ1
)
and probe tuned at 75.47 MHz for 13C
(concentrations ꢂ10 mg mLꢁ1). Tetramethylsilane (TMS) was used
as an internal reference standard during 1H and 13C NMR experi-
ments. Conductivity measurements were made with a Cond340i/
SET Conductivity Meter (WTW, Germany) in methanol and acetone
solutions, at 25 °C. The concentration of the complexes in the solu-
tions was 10ꢁ3 M [22]. The magnetic susceptibility measurements
were performed for all the compounds by the Faraday method at
298 K with Hg[Co(SCN)4] as a calibrant. In case of 1 and 4, the tem-
perature dependence of the magnetic susceptibility was measured
over the interval of 298–80 K. The calculated molar susceptibilities
were corrected for diamagnetism using Pascal constants [25]. The

ˇ
416
Z. Trávnícek et al. / Journal of Inorganic Biochemistry 104 (2010) 405–417
Table 6
positive ion electrospray (ES+) mass spectra were recorded using a
flow injection mode on the Waters ZMD 2000 mass spectrometer
with the mass-monitoring interval 40–1300 m/z. The spectra were
collected using 3.0 s cyclical scans and applying the sample cone
voltage 20 or 40 V and capillary voltage +3.0 kV, at the source block
temperature 100 °C, desolvation temperature 250 °C, cone gas flow
rate 50 L hꢁ1 and desolvation gas flow rate 500 L hꢁ1. The mass
spectrometer was directly coupled to a MassLynx data system.
Thermogravimetric analysis (TGA) and differential scanning calo-
rimetry (DSC) were both performed with XP-10 Thermogravimet-
ric Analyzer (THASS, GmbH) in temperature range of 20–1000 °C
(TGA) and 20–400 °C (DSC) with the temperature gradient of
Crystal data and structure refinements for 2-chloro-6-benzylamino-9-isopropylpu-
rine, L.
Empirical formula
Formula weight
Temperature
Wavelength
Crystal system, space group
Unit cell dimensions
C15H16ClN5
301.78
100(2) K
0.71073 Å
Triclinic, P-1
a = 10.0218(2) Å,
a
= 77.189(2)°
b = 11.9984(3) Å, b = 69.640(2)°
c = 13.4265(3) Å,
1471.66(6) Å3
4, 1.362 g cmꢁ3
0.261 mmꢁ1
632
c = 81.5677(19)°
Volume
Z, Calculated density
Absorption coefficient
F(0 0 0)
Crystal size
Theta range for data collection
Limiting indices
Reflections collected/unique
Max. and min. transmission
Refinement method
Data/restraints/parameters
Goodness-of-fit on F2
Final R indices (I > 2
R indices (all data)
Largest difference peak and hole
5 °C minꢁ1
atmosphere.
. All measurements were carried out in the air
0.25 ꢄ 0.25 ꢄ 0.20 mm
2.61–25.00°
ꢁ11 < h < 11, ꢁ13 < k < 12, ꢁ15 < l < 10
11643/4970 (R(int) = 0.0250)
0.9497 and 0.9377
3.5. X-ray crystallography
Full-matrix least-squares on F2
4970/0/383
X-ray data of 2-chloro-6-benzylamino-9-isopropylpurine, L, i.e.
an intermediate for the preparation of L1, L4 and L5, were collected
on a four-circle k-axis diffractometer Xcalibur2 (Oxford Diffraction
Ltd.) equipped with the Sapphire2 CCD detector at 120 K. The Cry-
sAlis software package [32] was used for data collection and reduc-
tion. The structure was solved using direct methods (SHELX,
Sheldrick 2008) [33] and refined anisotropically on F2 using
full-matrix least-squares procedure (SHELX, Sheldrick 2008) [33]
1.027
r(I))
R1 = 0.0360, wR2 = 0.1001
R1 = 0.0554, wR2 = 0.1117
0.277 and ꢁ0.318 e Åꢁ3
and 100% humidity. The cell suspension of approximate density
1.25 ꢄ 105 cells mLꢁ1 was redistributed into 96-well microtitre
plates (Nunc, Denmark). After 12 h of preincubation, the tested
with weight scheme w = 1/[r
2(Fo2) + (0.076P)2 + 0.208P], where
P = (Fo2 + 2Fc2)/3. All hydrogen atoms of L were located from differ-
ence Fourier maps and their parameters were refined using the
riding model with C–H distances of 0.95 and 0.99 Å and N–H
distances of 0.88 Å, and with Uiso(H) = 1.2Ueq (CH, CH2 and NH)
or 1.5Ueq (CH3). Crystal data and structure refinement parameters
for L are given Table 6.
compounds (in the 0.5–50, 0.5–100, and 0.5–200 lM ranges) were
added. Incubation lasted for 72 h. At the end of this period, the cells
were incubated for 1 h with calcein AM and fluorescence of the live
cells was measured at 485/538 nm (ex/em) with a Fluoroskan As-
cent (Labsystems, Finland). IC50 values, the drug concentration
lethal to 50% of tumour cells, were estimated. The presented values
are represented by arithmetic means determined from three
values.
3.6. Quantum chemical calculations
Theoretical calculations using the DFT (B3LYP) approach were
used to verify the assumed structures of studied complexes. All cal-
culations were performed using the Gaussian03 program [34]. Pre-
Human 6xHis-tagged cyclin E/Cdk2 complex was produced in
Sf9 insect cells coinfected with appropriate baculoviral constructs.
The cells were harvested 70 h post infection in lysis buffer (50 mM
Tris pH 7.4, 150 mM NaCl, 5 mM EDTA, 20 mM NaF, 1% Tween 20,
protease inhibitors) for 30 min on ice and the soluble fraction was
recovered by centrifugation at 14,000g for 10 min. The kinase was
*
liminary data were prepared on the HF/6–31 level of theory. Then,
geometries of the obtained structures were fully optimized at the
B3LYP level. Wachter’s original full-electron basis set [35] (con-
tracted as 62111111/3311111/3111) with one set of polarization
functions was used for the iron atom, while for all other atoms
the 6-311G(d) basis set [36] was employed. Harmonic frequency
analysis was used to verify the nature of founded stationary points
as the minima as well as for the calculation of zero-point vibra-
tional energies. Calculated infrared frequencies were not scaled.
´
purified on NiNTA column (Qiagen) according to manufacturers
instructions. To carry out experiments on kinetics under linear
conditions, the final point test system for kinase activity measure-
ment was used. The assay mixture contained 1 mg/mL histone
(Sigma Type III-S), 15
pound in a final volume of 10
l
M ATP, 0.2
L, all in reaction buffer: 50 mM
lCi [c-
33P] ATP and tested com-
l
Mössbauer shifts
d
were calculated according to equation
where calibration constant
Hepes pH 7.4, 10 mM MgCl2, 5 mM EGTA, 10 mM 2-glycerolphos-
phate, 1 mM NaF, 1 mM DTT and protease inhibitors; where
d ¼
a
½
qA0ð0Þ ꢁ qS0ð0Þꢃ;
a
a
(ꢁ0.395 mm sꢁ1 au3) and non-relativistic electron density at the
Fe nucleus of reference compound qS0 (11614.10 auꢁ3) were ob-
tained from previous calculations [37–39]. Electron densities
qA0ð0Þ at the 57Fe nuclei of studied compounds were calculated with
the AIM2000 program [40] that utilizes the wave functions gener-
ated by Gaussian03 program.
Hepes = 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic
acid,
EGTA = ethylene glycol-bis(2-aminoethylether)-N,N,N0,N0-tetraace-
tic acid, and DTT = (dithiothreitol). After 10 min, the incubations
were stopped by adding 5% H3PO4 and spotted on P81 phosphocel-
lulose paper (Whatman). After washing in 5% H3PO4, the kinase
activity was measured by digital imaging analyzer BAS-1800 (Fuji-
film). The kinase activity was expressed as a percentage of maxi-
mum activity, the IC50 values were determined by graphic analysis.
3.7. Biological activity testing
Human malignant melanoma cell line G-361, human osteogenic
sarcoma cell line HOS, human chronic myelogenous leukaemia cell
line K-562 and human breast adenocarcinoma cell line MCF-7 were
used for a cytotoxicity determination of synthesized compounds
by calcein acetoxymethyl (AM) assay, as it was described previ-
ously in the literature [15,26]. The tumour cells were maintained
in plastic tissue culture flasks and grown on Dulbecco’s modified
Eagle’s cell culture medium (DMEM) at 37 °C in 5% CO2 atmosphere
Acknowledgments
Authors thank the Ministry of Education, Youth and Sports of
the
Czech
Republic
(Grant
nos.
MSM6198959218,
1M6198959201) for financial support. Authors also thank to Assoc.
ˇ
ˇ
Prof. Zdenek Šindelár for the temperature dependence of magnetic

ˇ
Z. Trávnícek et al. / Journal of Inorganic Biochemistry 104 (2010) 405–417
417
[18] C.M. Mikulski, S. Cocco, N.D. Franco, N.M. Karayannis, Inorg. Chim. Acta 80
(1983) L23–L25.
[19] A.N. Speca, C.M. Mikulski, F.J. Iaconianni, L.L. Pytlewski, N.M. Karayannis, J.
Inorg. Nucl. Chem. 43 (1981) 2771–2779.
susceptibility measurements, and Mrs. Dita Parobková for her help
with biological testing.
[20] M.V. Capparelli, D.M.L. Goodgame, P.B. Hayman, A.C. Skapski, Inorg. Chim. Acta
125 (1986) L47–L49.
Appendix A. Supplementary material
[21] H.M. Schmalle, E. Gyr, E. Dubler, Acta Crystallogr. C 56 (2000) 957–959.
[22] W.J. Geary, Coord. Chem. Rev. 7 (1971) 81–122.
[23] F. Menil, J. Phys. Chem. Solids 46 (1985) 763–789.
[24] I. Dézsi, P.J. Ouseph, P.M. Thomas, J. Inorg. Nucl. Chem. 36 (1974) 833–836.
[25] O. Kahn, Molecular Magnetism, Wiley VCH Inc., New York, USA, 1993.
Additional crystallographic data are deposited as CCDC 738583
and can be obtained free of charge from the Cambridge Crystallo-
graphic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK;
fax: (+44) 1223–336–033; email: deposit@ccdc.cam.ac.uk, or at
http://www.ccdc.cam.ac.uk. Selected bond lengths and angles, as
well as a view of the molecular structure of the L ligand are depos-
ited. Supplementary data associated with this article can be found,
in the online version, at doi:10.1016/j.jinorgbio.2009.12.002.
ˇ
ˇ
ˇ
ˇ
[26] Z. Trávnícek, M. Malon, M. Biler, M. Hajdúch, P. Broz, K. Dolezal, J. Holub, V.
Kryštof, M. Strnad, Transit. Met. Chem. 25 (2000) 265–269.
[27] C.J. Pouchert, The Aldrich Library if Infrared Spectra, third ed., Aldrich
Chemical Co., Milwaukee, WI, USA, 1981.
[28] G. Socrates, Infrared and Raman Characteristic Group Frequencies (Tables and
Charts), third ed., John Wiley & Sons Ltd., Chichester, England, 2001.
[29] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination
Compounds, fifth ed., John Wiley & Sons Ltd., New York, USA, 1997.
[30] K. Brandenburg, DIAMOND, Release 3.2, Crystal Impact GbR, Bonn, Germany,
2009.
References
[1] P. Cohen, Nat. Cell Biol. 4 (2002) E127–E130.
[2] D.O. Morgan, Nat. Cell Biol. 4 (2002) E127.
[3] P. Cohen, Nat. Rev. Drug Discov. 1 (2002) 309–315.
[4] C. Benson, S. Kaye, P. Workman, M. Garrett, M. Walton, J. de Bono, Brit. J. Cancer
92 (2005) 7–12.
[31] S. Wang, S.J. McClue, J.R. Ferguson, J.D. Hull, S. Stokes, S. Parsons, R. Weswood,
P.M. Fischer, Tetrahedron: Asymmetry 12 (2001) 2891–2894.
[32] CrysAlis CCD and CrysAlis RED, Oxford Diffraction Ltd, Abingdon, Oxfordshire,
England, 2002.
[33] G.M. Sheldrick, Acta Cryst. A 46 (2008) 112–122.
[5] P.M. Fischer, A. Gianella-Borradori, Expert Opin. Invest. Drugs 14 (2005) 457–
477.
[6] Y.T. Chang, N.S. Gray, G.R. Rosania, D.P. Sutherlin, S. Kwon, T.C. Norman, R.
Sarohia, M. Leost, L. Meijer, P.G. Schultz, Chem. Biol. 6 (1999) 361–375.
[7] L. Meijer, A. Borgne, O. Mulner, J.P.J. Chong, J.J. Blow, N. Inagaki, M. Inagaki, J.G.
Delcros, J.P. Moulinoux, Eur. J. Biochem. 243 (1997) 527–536.
[34] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
J.A. Montgomery Jr., T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A.
Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox,
H.P. Hratchian, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E.
Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y.
Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S.
Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K.
Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J.
Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L.
Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M.
Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, C. Gonzalez, J.A.
Pople, Gaussian 03, Revision C.02, Gaussian, Inc., Wallingford, USA, 2004.
[35] A.J.H. Wachters, J. Chem. Phys. 52 (1970) 1033–1036.
ˇ
[8] W.F. De Azevedo, S. Leclerc, L. Meijer, L. Havlícek, M. Strnad, S.H. Kim, Eur. J.
Biochem. 243 (1997) 518–526.
[9] S. Wittmann, P. Bali, S. Donapaty, R. Nimmanapalli, F. Guo, H. Yamaguchi, M.
Huang, R. Jove, H.G. Wang, K. Bhalla, Cancer Res. 63 (2003) 93–99.
[10] D.E. MacCallum, J. Melville, S. Frame, K. Watt, S. Anderson, A. Gianella-
Borradori, D.P. Lane, S.R. Green, Cancer Res. 65 (2005) 5399–5407.
[11] B. Shan, Y. Zhuo, D. Chin, C.A. Morris, G.F. Morris, J.A. Lasky, J. Biol. Chem. 280
(2005) 1103–1111.
ˇ
ˇ
[12] L. Szücová, Z. Trávnícek, M. Zatloukal, I. Popa, Bioorg. Med. Chem. 14 (2006)
479–491.
[36] R. Krishnan, J.S. Binkley, R. Seeger, J.A. Pople, J. Chem. Phys. 72 (1980) 650–
654.
ˇ
[13] Z. Trávnícek, V. Kryštof, M. Šipl, J. Inorg Biochem. 100 (2006) 214–225.
ˇ
´
ˇ
[37] E. Fluck, in: V.I. Goldanskii, R.H. Herber (Eds.), Chemical Applications of
Mössbauer Spectroscopy, Academic Press Inc., New York, 1968, pp. 268–311.
[38] P. Gütlich, J. Ensling, in: A.B.P. Lever, E.I. Salomon (Eds.), Inorganic Electronic
Structure and Spectroscopy, vol. 1, John Wiley and Sons, New York, 1999, pp.
161–212.
[14] P. Štarha, Z. Trávnícek, R. Herchel, I. Popa, P. Suchy, J. Vanco, J. Inorg. Biochem.
103 (2009) 432–440.
ˇ
ˇ
ˇ
ˇ
ˇ
[15] M. Malon, Z. Trávnícek, M. Maryško, R. Zboril, M. Mašlán, J. Marek, K. Dolezal, J.
ˇ
Rolcík, V. Kryštof, M. Strnad, Inorg. Chim. Acta 123 (2001) 119–129.
[16] Z. Trávnícek, J. Mikulík, M. Cajan, R. Zboril, I. Popa, Bioorg. Med. Chem. 16
ˇ
ˇ
ˇ
(2008) 8719–8728.
[39] V.N. Nemykin, R.G. Hadt, Inorg. Chem. 45 (2006) 8297–8307.
[40] F. Biegler-Konig, J. Schonbohm, D. Bayles, J. Comput. Chem. 22 (2001) 545–
559.
[17] C.M. Mikulski, S. Cocco, N.D. Franco, T. Moore, Inorg. Chim. Acta 106 (1985)
89–95.