Copper(II) complexes with substituted thiosemicarbazones of α-ketoglutaric acid: Synthesis, X-ray structures, DNA binding studies, and nuclease and biological activity

Article abstract of DOI:10.1021/ic049883b

New α-ketoglutaric acid thiosemicarbazone (H₃ct) derivatives and their copper complexes were synthesized and characterized by analytical and spectroscopic (IR and NMR) methods. For two of the ligands, Me-H₃ct and Allyl-H₃ct, and for a complex, [Cu(Me-Hct)(OH₂)]_n·2nH₂O, the X-ray structures were also determined. In the latter the copper atom shows a 4 + 1 pyramidal coordination, a water oxygen appears in the apical position, and three of the basal positions are occupied by the SNO tridentate ligand and the fourth by a carboxylic oxygen of an adjacent molecule that gives rise to a polymeric chain. DNA binding constants were determined, and studies of thermal denaturation profiles and nuclease activity were also performed. Tests in vitro on human leukemia cell line U937 were carried out on cell growth inhibition, cell cycle, and apoptosis induction.

Full text of DOI:10.1021/ic049883b

Inorg. Chem. 2004, 43, 7170−7179
Copper(II) Complexes with Substituted Thiosemicarbazones of
-Ketoglutaric Acid: Synthesis, X-ray Structures, DNA Binding Studies,
and Nuclease and Biological Activity
r
Monica Baldini,† Marisa Belicchi-Ferrari,*,† Franco Bisceglie,† Pier Paolo Dall’Aglio,‡ Giorgio Pelosi,†
Silvana Pinelli,‡ and Pieralberto Tarasconi†
Dipartimento di Chimica Generale ed Inorganica, Chimica Analitica, Chimica Fisica, and
Dipartimento di Clinica Medica, Nefrologia e Scienza della PreVenzione, UniVersita` degli Studi di
Parma, 43100 Parma, Italy
Received January 28, 2004
New
characterized by analytical and spectroscopic (IR and NMR) methods. For two of the ligands, Me-H3ct and Allyl-
H3ct, and for a complex, [Cu(Me-Hct)(OH2)]n 2nH2O, the X-ray structures were also determined. In the latter the
copper atom shows a 4 1 pyramidal coordination, a water oxygen appears in the apical position, and three of
R-ketoglutaric acid thiosemicarbazone (H3ct) derivatives and their copper complexes were synthesized and
‚
+
the basal positions are occupied by the SNO tridentate ligand and the fourth by a carboxylic oxygen of an adjacent
molecule that gives rise to a polymeric chain. DNA binding constants were determined, and studies of thermal
denaturation profiles and nuclease activity were also performed. Tests in vitro on human leukemia cell line U937
were carried out on cell growth inhibition, cell cycle, and apoptosis induction.
Introduction
doxal,10,11 5-formyluracil12-14) affect the complicated mech-
anisms that lead to a leukemic transformation by inhibiting
replication and by triggering apoptotic processes and thus
they act as potentially important antitumor substances. The
debate on the lability of coordinate bonds to copper(II) is
well-known in the community of bioinorganic chemists
especially as regards copper-based drugs. Nevertheless the
formation constants of these thiosemicarbazones com-
plexes15,16 are large enough to render most of the molecules
that can be found in a culture medium ineffective in
competing for complexation. This phenomenon is therefore
much limited in the case of SNO chelate systems such as
Thiosemicarbazones and their derivatives have raised
considerable interest in chemistry, pharmacology, and biol-
ogy thanks to their antibacterial, antiviral, and antitumor
activity.1-9 In particular our previous studies revealed that
copper complexes with aromatic thiosemicarbazones (pyri-
* Author to whom correspondence should be addressed. E-mail: marisa.
ferrari@unipr.it. Tel.: +39-0521-905-420. Fax: +39-0521-905-557.
† Dipartimento di Chimica Generale ed Inorganica, Chimica Analitica,
Chimica Fisica.
‡ Dipartimento di Clinica Medica, Nefrologia e Scienza della Prevenzione.
(1) Afrasiabi, Z.; Sinn, E.; Padhye, S.; Dutta, S.; Padhye, S.; Newton, C.;
Anson, C. E.; Powell, A. K. J. Inorg. Biochem. 2003, 95, 306-314.
(2) Byrnes, R. W.; Mohan, M.; Antholine, W. E.; Xu, R. X.; Petering, D.
H. Biochemistry 1990, 29, 7046-7053.
(9) Condit, R. C.; Easterly, R.; Pacha, R. F.; Fathi, Z.; Meis, R. J. Virology
1991, 185, 857-861.
(3) Kolocouris, A.; Dimas, K.; Pannecouque, C.; Witvrouw, M.; Foscolos,
G.-B.; Stamatiou, G.; Fytas, G.; Zoidis, G.; Kolocouris, N.; Andrei,
G.; Snoeck, R.; De Clercq, E. Bioorg. Med. Chem. 2002, 12, 723-
727.
(10) Belicchi Ferrari, M.; Gasparri Fava, G.; Tarasconi, P.; Albertini, R.;
Pinelli, S.; Starcich, R. J. Inorg. Biochem. 1994, 53, 13-25.
(11) Belicchi Ferrari, M.; Bisceglie, F.; Leporati, E.; Pelosi, G.; Tarasconi,
P.; Pinelli, S. Bull. Chem. Soc. Jpn. 2002, 75, 781-788.
(12) Belicchi Ferrari, M.; Gasparri Fava, G.; Leporati, E.; Pelosi, G.; Rossi,
R.; Tarasconi, P.; Albertini, R.; Bonati, A.; Lunghi, P.; Pinelli, S. J.
Inorg. Biochem. 1998, 70, 145-154.
(4) Hall, I. H.; Chen, S. Y.; Barners, B. J.; West, D. X. Met.-Based Drugs
1999, 6, 143-147.
(5) Kelly, P. F.; Slawin, A. M. Z.; Soriano-Rama, A. J. Chem. Soc., Dalton
Trans. 1996, 53-60.
(6) Klayman, D. L.; Bartosovich, J. F.; Griffin, T. S.; Mason, C. J.; Scovill,
J. P. J. Med. Chem. 1979, 22, 855-862.
(13) Belicchi Ferrari, M.; Bisceglie, F.; Pelosi, G.; Tarasconi, P.; Albertini,
R.; Bonati, A.; Lunghi, P.; Pinelli, S. J. Inorg. Biochem. 2001, 83,
169-179.
(7) Bermejo, E.; Carballo, R.; Castineiras, A.; Dominguez, R.; Liberta,
A. E.; Maichle-Mossmer, C.; West, D. X. Z. Naturforsch. 1999, B54,
777-787.
(14) Baldini, M.; Belicchi-Ferrari, M.; Bisceglie, F.; Pelosi, G.; Pinelli,
S.; Tarasconi, P. Inorg. Chem. 2003, 42, 2049-2055.
(15) Borges, R. H. U.; Oaniago, E.; Beraldo, H. J. Inorg. Biochem. 1997,
65, 267-275.
(8) Lakovidou, Z.; Papageorgiou, A.; Demertzis, M. A.; Mioglou, E.;
Mourelatos, D.; Kotsis, A.; Yadav, P. N.; Kovala-Demertzi, D.
Anticancer Drugs 2001, 12, 65-70.
(16) Offiong, O. Transition Met. Chem. 1998, 23, 553-555.
7170 Inorganic Chemistry, Vol. 43, No. 22, 2004
10.1021/ic049883b CCC: $27.50
© 2004 American Chemical Society
Published on Web 10/05/2004

Cu(II) r-Ketoglutaric Acid Thiosemicarbazones
Chart 1
apoptosis induction) on human leukemia cell line U937 were
performed, together with a study of their cell cycle.
Experimental Section
Physical Measurements. Elemental analyses (C, H, N, S) were
performed with a Carlo Erba model EA 1108 automatic analyzer.
IR spectra (4000-400 cm-1) for KBr disks were recorded on a
Nicolet 5PC FT. Chemical ionization-mass (m/z, 70 eV) fragmenta-
tion patterns were obtained, using samples dissolved in methanol,
with a Finnegan 1020 6c mass spectrometer equipped with a
quadrupole mass selector MATSSQ 710. Melting points were
determined with a Gallenkamp instrument. UV measurements were
performed on a Perkin-Elmer UV/vis Lambda 25 spectrometer with
quartz cuvettes. 1H NMR spectra were obtained at room temperature
with a Brucker AMX-300 MHz spectrometer, and chemical shifts
are given in units of δ relative to TMS as an internal reference,
using CD3OD as the solvent.
these with thiosemicarbazones. Given the order of magnitude
of the formation constants, Cu-thiosemicarbazone species
can be assumed present at high percentage even in the
presence of strong competitors, whereas Cu-Cl, Cu-OH2,
and Cu-O(OOCR) bonds quickly dissociate and allow
bonding formation with other ligands (for instance biomol-
ecules). We focused then our attention on aliphatic thiosemi-
carbazones derived from natural aldehydes (which are
potentially SN bidentate) and their nickel and copper
complexes.17 More recently our attention was devoted to
copper complexes of the R-ketoglutaric acid thiosemicarba-
zone (H3ct), an aliphatic ligand with many potential donor
atoms and a variety of possible conformations which
determine a versatile chelating behavior. In particular,
coordination geometry, ligand behavior, and biological
activity were compared for two copper complexes [Cu(H2-
ct)Cl]n[{Cu(H2ct)Cl}2] and [Cu(Hct)]n‚3nH2O which exert
proliferation inhibition through an apoptosis mechanism on
cell line U937.18 To extend further our knowledge of the
chemical and biological behavior of R-ketoglutaric acid
thiosemicarbazones and to widen our comprehension of the
mechanism of action and of the possible biological targets,
we synthesized condensation compounds of the R-ketoglu-
taric acid with seven N-substituted thiosemicarbazides pos-
sessing aliphatic and aromatic groups (see Chart 1) that were
subsequently used for preparing copper complexes. The
choice on the substituents was made to find out which part
of the molecule is determinant in endowing the complex of
activity and also in the light of the considerations reported
by Hambley and co-workers19 that the stability of copper-
thiosemicarbazone species is increased in the case the
external surface of the complex molecule is hydrophobic.
Materials. 4-Methyl-3-thiosemicarbazide, 97% (Aldrich), 4-eth-
yl-3-thiosemicarbazide, 97% (Aldrich), 4-allyl-3-thiosemicarbazide,
97% (Aldrich), 2-methyl-3-thiosemicarbazide, 97% (Aldrich), 2,4-
dimethyl-3-thiosemicarbazide, 98% (Lancaster), 4-phenyl-3-thio-
semicarbazide, 99% (Janssen), 4-(3-methylphenyl)-3-thiosemicar-
bazide, 97% (Aldrich), and CuCl2‚2H2O (Carlo Erba) were com-
mercially available and used without further purification. Calf
thymus (CT) DNA was obtained from Serva and used as received.
TRIS and TRIS-acetate-EDTA 10× buffers and agarose were
purchased from Sigma. pBR322 DNA plasmid (Roche), DirectLoad
Wide-Range DNA marker, ethidium bromide, and gel loading
solution type I 6× concentrate were from Sigma.
Preparation of the Ligands. The general procedure used to
prepare the ligands can be summarized as follows: to a solution
(the solvent differs from case to case depending on the substance
and is reported below) of substituted thiosemicarbazide, obtained
by magnetic stirring and slight heating, an equimolar amount of
R-ketoglutaric acid, dissolved in the same solvent, is added
dropwise. After a while, a white powder is formed, isolated by
filtration, and analyzed. The purity of the products dissolved in
the minimum quantity of CH3OH was checked by TLC analysis
using 1:1 (v/v) THF/acetone as eluting mixture and revealed by
means of a UV lamp at 254 nm showing the spots of the pure
products. Purity was also confirmed by the melting points. Ligands
were recrystallized from CH3OH solutions, obtained as crystals,
and two of them were characterized by X-ray analysis. Crystalline
products presented the same analytical details of the powders. The
reaction is reported in Scheme 1.
Synthesis of Me-H3ct (1) (R1 ) Me, R2 ) H). A 0.26 g (2.5
mmol) amount of 4-methyl-3-thiosemicarbazide was solubilized in
10 mL of H2O. An equimolar amount (0.36 g) of R-ketoglutaric
acid was dissolved in 7 mL of H2O, added dropwise to the solution
of the thiosemicarbazide, and stirred for 30 min. Anal. Calcd for
C7H11N3O4S: C, 36.01; H, 4.28; N, 18.00; S, 13.72. Found: C,
35.42; H, 4.85; N, 17.82; S, 14.42. IR (KBr disks, cm-1): 3581
(w) ν(OH); 3337 (s) and 3191 (m) ν(NH); 2946 (w) and 2916 (w)
ν(CH); 1718 (s) and 1700 (s) ν(CdO); 1600 (w) and 1560 (m)
ν(CdN); 869 (w) ν(CdS). UV-visible (solvent MeOH, nm): 206
(n f σ*); 256 and 307 (n f π*). 1H NMR (CD3OD, ppm): 2.65(m)
-CH2CH2COOH; 2.87(m) -CH2CH2COOH; 3.16 (d) of methyl
substituent; 8.40 (d) NH aminic; 10.75 (d) NH hydrazinic. Mass
spectrum (m/z): 262 (M+ + Et); 234 (MH+); 215 (M+ - H2O);
188 (MH+ - CO2 or M+ - COOH). Mp: 174 °C. Yield ) 72%.
Synthesis of Et-H3ct.H2O (2) (R1 ) Et, R2 ) H). A 0.3 g (2.5
mmol) amount of 4-ethyl-3-thiosemicarbazide was solubilized in
Ligands Me-H3ct (1) and Allyl-H3ct (3) and complex [Cu-
(Me-Hct)(OH2)]n‚2nH2O (9) were also characterized by X-ray
diffractometry. With the aim to ascertain the entity and the
mode of interaction that active molecules exert toward
specific biological targets, the binding interactions of the new
synthesized complexes with calf thymus DNA were inves-
tigated by absorption titration in the UV-visible region and
thermal denaturation (melting temperature). Nuclease activity
study was also performed to verify if oxidative breakings
take place. To confirm the cytotoxicity of this class of
compounds, tests in vitro (cell proliferation inhibition and
(17) Belicchi Ferrari, M.; Bisceglie, F.; Pelosi, G.; Sassi, M.; Tarasconi,
P.; Cornia, M.; Capacchi, S.; Albertini, R.; Pinelli, S. J. Inorg.
Biochem. 2002, 90, 113-126.
(18) Belicchi Ferrari, M.; Bisceglie, F.; Gasparri-Fava, G.; Pelosi, G.;
Tarasconi, P.; Albertini, R.; Pinelli, S. J. Inorg. Biochem. 2002, 89,
36-44.
(19) Weder, J. E.; Dillon, C. T.; Hambley, T. W.; Kennedy, B. J.; Lay, P.
A.; Biffin, J. R.; Regtop, H. L.; Davies, N. M. Coord. Chem. ReV.
2002, 232, 95-126.
Inorganic Chemistry, Vol. 43, No. 22, 2004 7171

Baldini et al.
Scheme 1
10 mL of H2O. An equimolar amount (0.36 g) of R-ketoglutaric
acid was dissolved in 7 mL of H2O, added dropwise to the solution
of the thiosemicarbazide, and stirred for 30 min. Anal. Calcd for
C8H15N3O5S: C, 36.18; H, 5.65; N, 15.83; S, 12.06. Found: C,
36.57; H, 5.69; N, 15.56; S, 12.65. IR (KBr disks, cm-1): 3436
(m) ν(OH); 3350 (w) and 3236 (w) ν(NH); 2976 (w) and 2936 (w)
ν(CH); 1709 (s) ν(CdO); 1543 (m) ν(CdN); 836 (w) ν(CdS).
UV-visible (solvent MeOH, nm): 228 (π f π*); 272 (n f π*).
1H NMR (CD3OD, ppm): 1.24 (t) CH3 of substituent; 2.65 (m)
CH2CH2COOH; 2.87 (m) -CH2CH2COOH; 3.70 (m) CH2 of
substituent; 8.42 (d) NH aminic; 10.77 (d) NH hydrazinic. Mass
spectrum (m/z): 276 (M+ + Et); 248 (MH+); 230 (MH+ - H2O);
212 (MH+ - H2S); 202 (MH+ - CO2 or M+ - COOH). Mp: 157
°C. Yield ) 73%.
of methyl aminic; 3.50 (s) of methyl hydrazinic; 8.45 (s) NH aminic.
Mass spectrum (m/z): 229 (M+ - H2O); 212 (MH+ - H2S). Mp:
154.5 °C. Yield ) 87%.
Synthesis of Ph-H3ct (6) (R1 ) Ph, R2 ) H). A 0.66 g (4 mmol)
amount of 4-phenyl-3-thiosemicarbazide was solubilized in 40 mL
of EtOH. An equimolar amount (0.58 g) of R-ketoglutaric acid was
dissolved in 20 mL of EtOH, added dropwise to the solution of
the thiosemicarbazide, and stirred for 3 h. Anal. Calcd for
C12H13N3O4S: C, 48.76; H, 4.40; N, 14.22; S, 10.83. Found: C,
48.98; H, 4.32; N, 14.02; S, 11.09. IR (KBr disks, cm-1): 3285
(m) ν(NH); 3084 (w) ν(CH aromatic); 2990 (w) ν(CH aliphatic);
1698 (s) ν(CdO); 1595(s) ν(CdC aromatic); 1541 (s) ν(CdN);
1437 (m) δ(OH); 1296 (m) δ(CH aromatic); 1166 (m) ν(CN); 935
(m) and 747 (m) ν(CdS); 615 (m) fingerprint. UV-visible (solvent
1
Synthesis of Allyl-H3ct (3) (R1 ) Allyl, R2 ) H). A 0.33 g
(2.5 mmol) amount of 4-allyl-3-thiosemicarbazide was solubilized
in 10 mL of H2O. An equimolar amount (0.36 g) of R-ketoglutaric
acid was dissolved in 7 mL of H2O, added dropwise to the solution
of the thiosemicarbazide, and stirred for 2 h. Anal. Calcd for
C9H13N3O4S: C, 41.65; H, 5.01; N, 16.19; S, 12.34. Found: C,
41.89; H, 4.96; N, 15.92; S, 12.91. IR (KBr disks, cm-1): 3361
(w) ν(OH); 3204 (m) ν(NH); 2924 (w) ν(CH); 1710 (s) and 1681
(s) ν(CdO); 1560 (m) ν(CdN); 907 (w) ν(CdS). UV-visible
MeOH, nm): 304 (π f π*). H NMR (CD3OD, ppm): 2.65 (m)
-CH2CH2COOH; 2.87 (m) -CH2CH2COOH; 7.70 (m) aromatic
hydrogens; 8.43 (m) NH aminic; 10.75 (s) NH hydrazinic. Mass
spectrum (m/z): 295 (M+); 278 (MH+ - H2O); 260 (MH+ - H2S);
250 (MH+ - CO2 or M+ - COOH); 217 (M+ - Ph). Mp: 154
°C. Yield ) 98%.
Synthesis of 3-Me-Ph-H3ct (7) (R1 ) 3-Me-Ph, R2 ) H). A
0.18 g (1 mmol) amount of 4-(3-methylphenyl)-3-thiosemicarbazide
was solubilized in 20 mL of MeOH. An equimolar amount (0.14
g) of R-ketoglutaric acid was dissolved in 10 mL of MeOH, added
dropwise to the solution of the thiosemicarbazide, and stirred for 3
h. Anal. Calcd for C13H15N3O4S: C, 50.43; H, 4.85; N, 13.58; S,
10.34. Found: C, 50.25; H, 4.82; N, 13.35; S, 10.45. IR (KBr disks,
cm-1): 3326 (w) ν(OH); 3194 (w) ν(NH); 3019 (w) ν(CH
aromatic); 2922 (w) ν(CH aliphatic); 1710 (s) and 1682(s) ν(Cd
O); 1537(s) ν(CdC aromatic); 1446 (m) δ(OH); 1276 (w) δ(CH
aromatic); 1200 (m) ν(CN); 935 (m) and 789 (w) ν(CdS); 646
(w) fingerprint. UV-visible (solvent MeOH, nm): 206 (n f σ*);
305 (π f π*).1H NMR (CD3OD, ppm): 2.41 (s) CH3Ar; 2.65 (m)
-CH2CH2COOH; 2.87 (m) -CH2CH2COOH; 7.69 (m) aromatic
hydrogens; 8.43 (m) NH aminic; 10.75 (s) NH hydrazinic. Mass
spectrum (m/z): 338 (M+ + Et); 309 (M+); 292 (MH+ - H2O);
274 (MH+ - H2S); 264 (MH+ - CO2 or M+ - COOH). Mp: 161.5
°C. Yield ) 91%.
Preparation of the Complexes. To a solution of the ligand (the
solvent differs from case to case depending on the substance and
is reported below), a solution containing an equimolar amount of
CuCl2‚2H2O dissolved in the same solvent is added dropwise. The
resulting mixture is stirred for 1 h at room temperature. By slow
evaporation of the solvent, a green or blue microcrystalline powder
or crystals are formed. All compounds were recrystallized in several
solvents as CH3OH, C2H5OH, and THF. In all cases crystals were
obtained, but only for 9 were crystals of size suitable for X-ray
diffraction collected.
1
(solvent MeOH, nm): 288 (n f π*). H NMR (CD3OD, ppm):
1.24 (t) CH3 of substituent; 2.64(m) -CH2CH2COOH; 2.86 (m)
-CH2CH2COOH; 4.31 (m) CH2dCHCH2; 5.20 (m) CH2dCHCH2;
5.97 (m) CH2dCHCH2; 8.40 (d) NH aminic; 10.75 (d) NH
hydrazinic. Mass spectrum (m/z): 274 (M+ + Me); 260 (MH+);
241 (M+ - H2O); 224 (MH+ - H2S); 214 (MH+ - CO2 or M+
COOH). Mp: 172.5 °C. Yield ) 41%.
-
Synthesis of 2-Me-H2ct (4) (R1 ) H, R2 ) Me). A 0.15 g (1.4
mmol) amount of 2-methyl-3-thiosemicarbazide was solubilized in
10 mL of THF. An equimolar amount (0.2 g) of R-ketoglutaric
acid was dissolved in 7 mL of THF, added dropwise to the solution
of the thiosemicarbazide, and stirred for 2 h. Anal. Calcd for
C7H11N3O4S: C, 36.01; H, 4.28; N, 18.00; S, 13.72. Found: C,
36.67; H, 4.66; N, 17.82; S, 13.99. IR (KBr disks, cm-1): 3329
(w) ν(OH); 3156 (m) ν(NH); 3030 (w) ν(CH); 1740 (s) and 1676
(s) ν(CdO); 1613 (s) ν(CdN); 814 (w) ν(CdS). UV-visible
1
(solvent MeOH, nm): 213 (n f π*). H NMR (CD3OD, ppm):
2.65 (m) -CH2CH2COOH; 2.87 (m) -CH2CH2COOH; 3.50 (s)
of methyl hydrazinic; 8.42 (s) NH2 aminic. Mass spectrum (m/z):
215 (M+ - H2O); 200 (MH+ - H2S). Mp: 164 °C. Yield ) 76%.
Synthesis of 2,4-Me2-H2ct (5) (R1 ) Me, R2 ) Me). A 0.36 g
(3 mmol) amount of 2,4-dimethyl-3-thiosemicarbazide was solu-
bilized in 20 mL of MeOH. An equimolar amount (0.44 g) of
R-ketoglutaric acid was dissolved in 10 mL of MeOH, added
dropwise to the solution of the thiosemicarbazide, and stirred for 2
h. Anal. Calcd for C8H13N3O4S: C, 38.82; H, 5.25; N, 16.98; S,
12.94. Found: C, 39.51; H, 5.13; N, 16.74; S, 13.16. IR (KBr disks,
cm-1): 3162 (w) ν(OH); 3049 (m) and 2848 (w) ν(NH); 2839 (w)
ν(CH); 1752 (s) and 1653 (s) ν(CdO); 747(w) ν(CdS). UV-visible
Only for obtaining complexes with the ligand Me-H3ct (1) the
pH was adjusted to 7 with NaOH before the addition of the metal
salt solution.
Synthesis of [Cu(Me-H2ct)Cl]‚H2O (8). A 0.32 g (1.4 mmol)
amount of Me-H3ct (1) was solubilized in 13 mL of H2O. Before
the addition of the metal salt solution, the pH value was adjusted
1
(solvent MeOH, nm): 214 (n f π*). H NMR (CD3OD, ppm):
2.65 (m) -CH2CH2COOH; 2.87 (m) -CH2CH2COOH; 3.42 (s)
7172 Inorganic Chemistry, Vol. 43, No. 22, 2004

Cu(II) r-Ketoglutaric Acid Thiosemicarbazones
Table 1. Experimental Data for Crystallographic Analyses
Me-H3ct (1)
C14H24N6O9S2
Allyl-H3ct (3)
[Cu(Me-Hct)(OH2)]n‚2nH2O (9)
formula
Mr
C9H13N3O4S
259.28
C7 H15CuN3O7S
348.81
484.50
cryst system
space group
a (Å)
b (Å)
c (Å)
monoclinic
P21/n
16.209(3)
6.788(3)
20.634(3)
90.0
111.80(2)
90.0
2108(1)
1.541 84
4
monoclinic
P21/n
13.227(3)
5.269(3)
18.927(3)
90.0
91.86(2)
90.0
1318.4(4)
0.710 69
4
triclinic
P1h
8.993(1)
9.749(1)
8.919(1)
111.00(2)
118.68(2)
71.49(2)
630.3(2)
0.710 69
2
R (deg)
â (deg)
γ (deg)
V (Å3)
λ (Å)
Z
Dc (Mg/m3)
F(000)
1.53
1016
28.44
1.31
544
2.53
1.84
358
1.93
µ (cm-1
)
cryst size (mm)
0.3 × 0.5 × 0.3
0.6 × 0.5 × 0.2
0.4 × 0.7 × 0.5
diffractometer
CAD4
Philips
Philips
θ min-max (deg)
3-50
3-30
3-30
dataset
-18/18, 0/7, 0/23
3521, 3128
2865
-18/18, 0/7, 0/26
7690, 3847
1044
-12/12, -12/12, 0/12
3680, 3680
2619
tot. reflcns, unique reflcns
obsd reflcns [I > 2.0σ(I)]
N
reflcns, Nparams
3128, 368
0.059, 0.16, 1.13
1044, 177
0.044, 0.14, 0.74
3680, 179
0.048, 0.14, 1.41
R, wR2, Sa
2
2
2
a R ) Σ||Fo| - |Fc||/Σ|Fo|; wR2 ) {Σ[w(Fo2 - Fc )2]/Σw(Fo )2]}1/2; S ) {Σ[w(Fo2 - Fc )2]/(n - p)}1/2, where n ) number of reflections and p ) number
of parameters refined.
to 7 with NaOH. An equimolar amount (0.25 g) of CuCl2‚2H2O
was then dissolved in 2 mL of H2O, added to the ligand solution,
and stirred for 1 h. Anal. Calcd for C7H12N3O5SCuCl: C, 24.05;
H, 3.43; N, 12.03; S, 9.16. Found: C, 24.38; H, 3.36; N, 11.73; S,
9.05. IR (KBr disks, cm-1): 3438 (m) H2O of crystallization; 3219
(m) ν(NH); 2982 (m) and 2937 (w) ν(CH); 1718 (s) and 1635 (s)
ν(CdO); 1600 (w) ν(CdN); 806 (w) ν(CdS). Mp: 193.5 °C.
Synthesis of [Cu(Me-Hct)(OH2)]n‚2nH2O (9). A 0.32 g (1.4
mmol) amount of Me-H3ct was dissolved in 13 mL of H2O. Before
the addition of the metal salt solution, the pH value was adjusted
to 7 with NaOH. An equimolar amount (0.25 g) of CuCl2‚2H2O
was then dissolved in 2 mL of H2O, added to the ligand solution,
and stirred for 1 h. Anal. Calcd for C7H15N3O7SCu: C, 24.07; H,
4.31; N, 12.03; S, 9.17%. Found: C, 24.43; H, 3.98; N, 12.34; S,
9.41. IR (KBr disks, cm-1): 3329 (s) ν(NH); 2982 (m) and 2937
(w) ν(CH); 1710 (s) and 1558 (s) ν(CdO); 1600 (w) ν(CdN); 806
(w) ν(CdS). Mp: 202 °C.
Synthesis of [Cu(Et-H2ct)Cl]‚H2O (10). A 0.5 g (2 mmol)
amount of Et-H3ct‚H2O was dissolved in 50 mL of MeOH. An
equimolar amount (0.33 g) of CuCl2‚2H2O was solubilized in 3
mL of H2O, added to the ligand solution, and stirred for 1 h. Anal.
Calcd for C8H14N3O5SCuCl: C, 26.42; H, 3.85; N, 11.56; S, 8.81%.
Found: C, 26.79; H, 3.42; N, 11.19; S, 8.63. IR (KBr disks, cm-1):
3412 (w) H2O of crystallization; 3217 (m) ν(NH); 2981 (m) ν(CH);
1723 (m) and 1633 (s) ν(CdO); 1593 (w) ν(CdN); 806 (w) ν(Cd
S). Mp: 168 °C.
mL of H2O, added to the ligand solution, and stirred for 1 h. Anal.
Calcd for C9H15N3O5SCuCl: C, 26.43; H, 4.13; N, 11.56; S, 8.81.
Found: C, 26.91; H, 3.87; N, 11.69; S, 8.55. IR (KBr disks, cm-1):
3440 (m) H2O of crystallization; 3308 (s) ν(NH); 3213 (w), 3114
(w) and 2912 (vw) ν(CH); 1701 (m) and 1616 (s) ν(CdO); 1563
(m) ν(CdN); 819 (w) ν(CdS). Mp: 190 °C.
Synthesis of Cu(Ph-H2ct)Cl (13). A 0.22 g (0.7 mmol) amount
of Ph-H3ct was solubilized in 25 mL of MeOH. An equimolar
amount (0.12 g) of CuCl2‚2H2O was dissolved in 5 mL of H2O,
added to the ligand solution, and stirred for 1 h. Anal. Calcd for
C12H12N3O4SCuCl: C, 36.61; H, 3.05; N, 10.68; S, 8.13. Found:
C, 37.13; H, 3.41; N, 10.32; S, 8.47. IR (KBr disks, cm-1): 3295
(w) ν(NH); 3064 (vw) and 2953 (vw) ν(CH); 1725 (m) and 1599
(m) ν(CdO); 1493 (m) ν(CdN); 755 (w) ν(CdS). Mp: 172 °C.
Synthesis of Cu(3-Me-Ph-H2ct)Cl (14). A 0.047 g (0.15 mmol)
amount of 3-Me-Ph-H3ct was dissolved in 15 mL of MeOH. An
equimolar amount (0.025 g) of CuCl2‚2H2O was dissolved in 5
mL of H2O, added to the ligand solution, and stirred for 1 h. Anal.
Calcd for C13H15N3O4SCuCl: C, 38.29; H, 3.68; N, 10.31; S, 7.85.
Found: C, 38.45; H, 3.51; N, 10.03; S, 8.15. IR (KBr disks, cm-1):
3418 (w) ν(NH); 3011 (vw) and 2945 (vw) ν(CH); 1719 (m) and
1610 (s) ν(CdO); 1557 (s) ν(CdN); 1492 (s) δ(OH); 1365 (s)
δ(CH); 778 (m) ν(CdS). Mp: 180 °C.
Crystallography. Crystal data and details of structure refine-
ments are given in Table 1. All intensity data were collected at
room temperature on a Philips diffractometer with Mo KR radiation
(λ ) 0.71069 Å) for ligand 3 and complex 9, while for ligand 1
the intensity data were carried out on an Enraf-Nonius CAD-4
instrument using graphite-monochromatized Cu KR radiation (λ
) 1.54184 Å). The intensities were measured following a modified
version20 of the method of profile analysis by Lehmann and Larsen21
and were corrected for Lorentz and polarization effects. Calculations
Synthesis of Cu(Allyl-H2ct)Cl‚H2O (11). A 0.36 g (1.4 mmol)
amount of Allyl-H3ct was dissolved in 10 mL of MeOH. An
equimolar amount (0.23 g) of CuCl2‚2H2O was solubilized in 5
mL of MeOH, added to the ligand solution, and stirred for 1 h.
Anal. Calcd for C9H14N3O5SCuCl: C, 28.77; H, 3.73; N, 11.19; S,
8.53. Found: C, 29.15; H, 3.39; N, 10.68; S, 8.81. IR (KBr disks,
cm-1): 3436 (m) H2O of crystallization; 3215 (m) ν(NH); 2994
(m) and 2928 (m) ν(CH); 1724 (w) and 1637 (s) ν(CdO); 1580
(m) ν(CdN); 819 (w) ν(CdS). Mp: 143 °C.
(20) Belletti, D.; Cantoni, A.; Pasquinelli, G. Gestione on line di diffrat-
tometro a cristallo singolo Siemens AED con sistema IBM PS 2/30;
Internal Report 1/88; Centro di studio per la Strutturistica diffratto-
metrica del CNR, 1988.
Synthesis of Cu(2,4-Me2-Hct)Cl‚H2O (12). A 0.05 g (0.2 mmol)
amount of 2,4-Me2-H2ct was dissolved in 15 mL of H2O. An
equimolar amount (0.035 g) of CuCl2‚2H2O was dissolved in 3
(21) Lehman, M. S.; Larsen, F. K. Acta Crystallogr. Sect. A 1974, 30, 580-
585.
Inorganic Chemistry, Vol. 43, No. 22, 2004 7173

Baldini et al.
were performed using the WinGX 1.64 program.22 The structure
for all compounds was solved by direct methods using SIR9223 for
ligand 1 and SIR9724 for the other compounds. Successive Fourier
syntheses allowed the assignment of the atoms to the electron
density peaks.
(ꢀA - ꢀF) vs [DNA] and is expressed as M-1. The above-mentioned
equation, originally used to calculate the binding constants for
hydrophobic derivatives, is now broadly used to investigate a wide
variety of metal complexes containing phenanthroline and its
derivatives and is usually adopted to obtain binding constants values
from metal complexes with different ligands.14,32,33 Fixed amounts
of the ligands and of the complexes were dissolved in CH3OH
because the high solubility of the compounds in this solvent allows
one to prepare concentrated solutions and therefore to utilize
reduced volumes (5 µL) for titrations. It was also verified that the
CH3OH percentage (0.7%) added to the DNA solution did not
interfere with the nucleic acid; in fact the 260 nm absorption band
is not subject to modifications in intensity and position. Concen-
trated solutions of NaCl, Tris-HCl (pH 7.2) buffer, and DNA were
prepared. Calculated amounts of the above-mentioned stock solu-
tions were brought to final concentration values of 50 mM NaCl,
5 mM Tris-HCl, and increasing amounts of DNA over a range of
DNA concentrations from 10-5 to 10-3 M and were added to the
5 µL solution of the considered compound to maintain the final
volume of the solutions fixed at 700 µL. The compounds were
titrated at room temperature. The changes in absorbance of DNA
of an intra ligand (IL) band upon each addition were monitored at
the maximum wavelengths 284, 281, 288, 290, 290, and 320 nm
for 1, 2, 6, 8, 10, and 13, respectively. For compound 10 the titration
was monitored also at 629 nm, where the metal d-d transition was
observed. The titrations were carried out for a bp/mol ratio r in the
range 0.01-8. Melting measurements were carried out in the above-
mentioned solutions. DNA (45 µM) was then treated with our
compounds at a mol/bp ratio of r ) 0.01, and each sample was
incubated for 24 h at room temperature. Samples were continuously
heated with 1° min-1 rate of temperature increase while monitoring
the absorbance change at 260 nm. The investigated interval of
temperature ranged from 50 to 90 °C. Upon reaching 90 °C, samples
were cooled back to 50 °C to allow the renaturation process. Values
for the melting temperature (Tm) and for the melting interval (∆T)
were determined according to the reported procedures.34 Differential
melting curves were obtained by numerical differentiation of
experimental melting curves.
For compounds 1, 3, and 9 refinements were carried out by full-
matrix least-squares cycles (SHELXL-9725). Anisotropic thermal
parameters were used for non-hydrogen atoms except in ligand 3
for the carbon atoms of the allyl substituent on the aminic nitrogen
which present a certain degree of disorder and in complex 9 for a
crystallization water molecule O3W that is distributed on two
positions (O3W1 and O3W2). For ligand 1 the hydrogen atoms
were located on a difference map, except those of the water
molecules O5 and O5A, which lie on a symmetry center, and those
of the carboxylic groups. In compound 3 all hydrogen atoms of
the aliphatic chain, except those of disordered allyl group, are placed
in calculated positions. Only the hydrogen atoms of methylene
carbon C7 and those of two hydroxylic oxygen were located on a
difference map and refined. In complex 9 only the hydrogen atoms
of coordinated water molecule were located on a difference map
and isotropically refined. All ligand hydrogen atoms were placed
in calculated positions and not refined.
Analytical expressions of neutral atom scattering factors were
employed according to ref 26. Molecular geometry calculations
were performed using the PARST27 computer program and the
structure drawings obtained with the ORTEPIII28 and PLATON29
programs.
DNA Interaction Studies. DNA samples were dissolved in
aqueous solution 50 mM NaCl, 5 mM Tris, pH 7.2. A solution of
CT-DNA (ca. 10-5 M in base pair, [bp]) was prepared, and in this
buffer it gave a UV absorbance ratio at 260 and 280 nm (A260
/
A280) of ca. 1.8, indicating that the CT-DNA was sufficiently protein
free. The concentration of the nucleic acid solutions was determined
by UV absorbance at 260 nm after 1:100 dilution. The extinction
coefficient ꢀ260 was taken as 13 100 M-1 cm-1 as reported in the
literature.30 Stock solutions were stored at 4 °C and used after no
more than 4 days. Binding constants for the interaction of the
compounds with the nucleic acid were determined as already
described.31 The intrinsic binding constant Kb for the interaction
of the compounds with CT-DNA has been calculated by an
Nuclease Activity: Materials and Method. The DNA cleavage
was carried out on double stranded plasmid DNA pBR 322 by
electrophoresis. Buffers were prepared using sterile distilled water.
Solutions composed by 1.6 µL of plasmid (40 µg/mL), 4 µL of
studied compound dissolved in DMSO (0.1 mM), and 8.4 µL Tris
buffer solution (10 mM) were incubated for 1 h at 37 °C. This was
followed by addition of gel loading solution, and each sample was
loaded directly into different wells on a 1.5% agarose gel for
analysis by electrophoresis at 75 V for 3 h. The gel was stained
with ethidium bromide and photographed under UV light in a
transilluminator.
absorption spectral titration data using the equation 1/∆ꢀap
)
1/(∆ꢀKbD) + 1/∆ꢀ, where ∆ꢀap ) |ꢀA - ꢀF|, ∆ꢀ ) |ꢀB - ꢀF|, D )
[DNA], and ꢀA , ꢀB, and ꢀF are respectively the apparent, bound,
and free extinctions coefficient of the compound. Kb is given by
the ratio of the slope to intercept when it is reported in plot [DNA]/
(22) Farrugia, L. J. WinGX 1.64. J. Appl. Crystallogr. 1999, A 32, 837-
838.
(23) Altomare, A.; Cascarano, G.; Giacovazzo, C.; Guagliardi, A. SIR 92.
J. Appl. Crystallogr. 1993, 26, 343-350.
(24) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G.; Giacovazzo,
C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.; Spagna, R. SIR
97. J. Appl. Crystallogr. 1999, 32, 115-122.
(25) Sheldrick, G. SHELXL 97 A Program for Structure Refinement;
University of Goettingen: Goettingen, Germany, 1997.
(26) International Tables for X-ray Crystallography, 4th ed.; Kynoch Press:
Birmingham, U.K., 1974.
(27) Nardelli, M. PARST95. An update to PARST: a system of Fortran
routines for calculating molecular structure parameters from the results
of crystal structure analyses. J. Appl. Crystallogr. 1995, 28, 659.
(28) Johnson, C. K.; Burnett, M. N. ORTEPIII; Report ORNL-6895; Oak
Ridge National Laboratory: Oak Ridge, TN, 1996.
(29) Spek, A. L. PLATON. A Multipurpose Crystallographic Tool; Utrecht
University: Utrecht, The Netherlands, 1999.
Biological Data: Materials and Methods. The compounds were
evaluated for their possible biological activity in DMSO solutions.
Different concentration ranges for each of the compounds had to
be used because they present different degrees of solubility. At the
same time for each experiment the biological effect of DMSO, used
as a blank, was determined for each concentration.
Human leukemic cell line U937 from a patient with histiocytic
lymphoma has been used. This line was grown in RPMI-1640
(32) Liu, J.; Zhang, T.; Lu, T.; Qu, L.; Zhou, H.; Zhang, Q.; Ji, L. J. Inorg.
Biochem. 2002, 91, 269-276.
(33) Mudasir; Yoshioka, N.; Inoue, H. J. Inorg. Biochem. 1999, 77, 239-
247.
(34) Messori, L.; Casini, A.; Vullo, D.; Haroutiunian, S. G.; Dalian, E. B.;
Orioli, P. Inorg. Chim. Acta 2000, 303, 283-286.
(30) Reichmann, M. E.; Rice, S. A.; Thomas, C. A.; Doty, P. J. Am. Chem.
Soc. 1954, 76, 3047-3051.
(31) Wolfe, A.; Shimer, G. H., Jr.; Meehan, T. Biochem. 1987, 26, 6392-
6396.
7174 Inorganic Chemistry, Vol. 43, No. 22, 2004

Cu(II) r-Ketoglutaric Acid Thiosemicarbazones
Table 2. Bond Distances (Å) for Me-H3ct (1)a and Allyl-H3ct (3)
Me-H3ct (1)
bond
unlabeled
A labeled
Allyl-H3ct (3)
S1-C1
O1-C3
O2-C3
O3-C6
O4-C6
N1-C1
N1-C7
N2-N3
N2-C1
N3-C2
C2-C3
C2-C4
C4-C5
C5-C6
C7-C8
C8-C9
1.680(4)
1.260(6)
1.253(5)
1.275(6)
1.238(5)
1.327(6)
1.459(6)
1.356(4)
1.363(5)
1.283(5)
1.512(6)
1.525(7)
1.512(6)
1.521(8)
1.676(4)
1.264(6)
1.259(5)
1.327(4)
1.200(5)
1.330(6)
1.456(6)
1.361(4)
1.368(5)
1.284(5)
1.498(5)
1.508(6)
1.517(6)
1.506(7)
1.670(3)
1.319(4)
1.218(4)
1.328(5)
1.197(4)
1.317(4)
1.461(5)
1.347(3)
1.381(4)
1.288(4)
1.490(4)
1.498(4)
1.521(4)
1.490(4)
1.460(7)
1.30(2)
a In the third column are reported the data relative to the atoms labeled
with the letter A.
Figure 1. ORTEP drawing of two independent molecules of ligand 1.
medium added with antibiotics (penicillin 100 U/mL, streptomycin
100 µg/mL), 200 mM L-glutamine, and 10% fetal bovine serum
(FBS) and kept at 37 °C in humid atmosphere at 5% CO2.
Cell Growth. The cells were seeded in a 96-well tray at a
concentration of 2 × 105/mL in culture medium and treated with
the different compounds. The effect on the cell proliferation process
has been determined for each compound to establish the maximum
concentration that can be used without appreciable toxic effects.
The ranges of the concentrations used were 10-100 µg/mL for
compounds 1 and 2 and 5-40 µg/mL for 8-10. Cell mortality,
evaluated on the fourth day by the trypan blue exclusion method
and determined by using a hemocytometer, never exceeded 5%.
Cells-Cycle Analysis. Cells seeded at 2 × 105/mL exposed to
various compounds were analyzed after 24, 48, and 72 h, were
washed three times with PBS, and were resuspended in 1 mL of
propidium iodide (PI) staining solution (50 µg/mL PI, 100 U/mL
RNase, 0.1% Igepal CA-630 in 0.1% sodium citrate) at RT (room
temperature) for 45 min in the dark. The cells were acquired on a
FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA),
and the cell-cycle distribution was analyzed using Flow-Jo software.
Apoptosis Assay. The cell lines were seeded at 8 × 105 mL-1
in the presence of the above-mentioned compounds and employed
in the apoptosis assay using the agarose gel 2% electrophoresis.
To these cells (2 × 106) (washed with PBS at 2000 rpm for 10
min at 4 °C) was added 20 µL of a solution of 10 µM EDTA, 50
mM Tris HCl at pH 8.0, and 0.5% (w/v) sodium laurylsarkosinate.
The pellet was subsequently redissolved and added of 2.5 µL of
Proteinase K (4 mg/mL) to a final concentration of 0.5 µg/mL.
After 1 h at 50 °C 2.5 µL of Rnase A (2 mg/mL) was added and
then incubated for 1 h at 50 °C. A gel electrophoresis carried out
with the whole sample with a voltage of 80 V allowed one to
observe the typical apoptotic DNA fragmentation.
The oxygen atoms O1 and O1A present a cis conformation
with respect to the N3 and N3A atoms, in a similar way as
found for the R-ketoglutaric acid thiosemicarbazone.35 The
conformation of the two molecules is mainly determined by
the intramolecular hydrogen bonds N1-H‚‚‚N3 and N1A-
H‚‚‚N3A, but also C-H‚‚‚O interactions are present that
differ from one molecule to the other, involving respectively
the C5 and C4A carbon atoms (C5-H‚‚‚O1A ) 3.390(5)
Å, C4A-H‚‚‚O1 ) 3.206(4) Å) as a consequence of the
different orientation of the two terminal -CH2CH2COOH
chains.
The orientation of these two groups is defined by the
torsion angles respectively for the two molecules as fol-
lows: C2-C4-C5-C6 -176.7(4)° (-antiperiplanar) and
C2A-C4A-C5A-C6A 175.3(3)° (antiperiplanar); C4-
C5-C6-O3 -174.9(4)° (-antiperiplanar) and C4A-C5A-
C6A-O3A 8.9(5)° (synperiplanar); C3-C2-C4-C5 84.4(5)°
(synclinal)
and
C3A-C2A-C4A-C5A
87.5(4)°
(synclinal). The molecule is not planar, but the angles
between the normal to the planes of the thiourea group and
that of the carboxyl group next to the hydrazine nitrogen
are fairly small: 3.3(1) and 9.4(1)°, respectively, for the two
independent molecules. The plane of the terminal carboxylic
group is nearly perpendicular to the mean plane of the
remaining side of the molecule (the dihedral angles formed
with the thiourea plane are 85.5(1) and 81.1(1)°, respectively,
for the two molecules). A major distortion is observed in
the hydrazinic chain C1-N2-N3-C2, while in the other
molecule the same chain is planar. A comparison between
the bond distances in the two molecules shows that they are
quite similar, both presenting the thiosemicarbazonic chain
in the thionic form (Table 2) as previously found in the
R-ketoglutaric acid thiosemicarbazone.32 The only remarkable
difference concerns the carboxylic group of the C6A carbon
atom, which shows the distances C6A-O4A (carbonylic)
) 1.200(5) Å and C6A-O3A (hydroxylic) ) 1.327(4) Å
typical of localized bonds, as already noticed for both
carboxylic groups of H3ct,35 while in the other three
Results and Discussion
Crystal and Molecular Structures. In the structure of
Me-H3ct (1) two nonequivalent ligand molecules are present,
together with two water molecules lying on symmetry centers
at 1/2, -1/2, 0 and 1, -1/2, 0. In Figure 1 the structure of the
two crystallographically independent ligand molecules is
reported. The main difference between the two molecules is
in the orientation of the -CH2CH2COOH chains. This is
highlighted by the two different torsion angles: C4-C5-
C6-O3 -174.9(4)° and C4A-C5A-C6A-O3A 8.9(5)°.
(35) Slouka, J. Pharmazie 1960, 15, 317-319.
Inorganic Chemistry, Vol. 43, No. 22, 2004 7175

Baldini et al.
Figure 2. ORTEP drawing of ligand 3.
Figure 3. Polymeric chain of coordination polyhedra [Cu(Me-Hct)(OH2)]n
in complex 9.
Chart 2
Table 3. Bond Distances for [Cu(Me-Hct)(OH2)]n·2nH2O (9)
Cu1-S1
Cu1-O1
Cu1-O1W
Cu1-N3
Cu-O3(I)
S1-C1
2.305(2)
1.994(5)
2.306(4)
1.964(4)
1.920(3)
1.713(6)
1.365(5)
1.360(9)
1.285(8)
C5-C6
O4-C6
O1-C3
O2-C3
O3-C6
N1-C1
N1-C7
C2-C3
C4-C5
1.508(8)
1.240(6)
1.276(8)
1.232(9)
1.265(6)
1.317(6)
1.453(10)
1.521(6)
1.528(6)
N2-N3
functional groups the C-O distances are delocalized (Table
2). The packing is determined, besides the interactions
between the two independent molecules, by an extended
network of hydrogen bonds between the two water mol-
ecules, the carboxylic groups, and the aminic nitrogen atom.
The ribbons so formed develop along the x axis at 0 and 1/2
of z. These ribbons are then joined along the y axis direction
by bonds O2‚‚‚O2A (x, y + 1, z) particularly short (2.672(4)
Å) and form sheets parallel to the (001) plane linked by the
hydrogen bonds N2-H‚‚‚O4 (3/2 - x, y - 1/2, -1/2 - z) )
3.099(6) Å and N2A-H‚‚‚O4A (3/2 - x, y + 1/2, 1/2 - z) )
2.928(5) Å. The distance O2‚‚‚O2A is very short probably
because of the presence of a strong hydrogen bond even
though hydrogen atoms were not experimentally located for
the wide delocalization in the two carboxylic groups.
An ORTEP plot of ligand 3 is shown in Figure 2. The
molecule is formed by three fragments: the R-ketoglutaric
moiety; the thiosemicarbazone chain; the allylic group. The
molecular core composed of the thiosemicarbazonic chain
and the carboxylic group next to the imine nitrogen is almost
planar and the planes A, B, and C (see Chart 2) form dihedral
angles of 4.0(2) and 7.1(1)° between them.
This planar conformation is determined by a strong
intramolecular hydrogen bond between the aminic nitrogen
atom N1 and the imine N3 (N1-H‚‚‚N3 ) 2.594(3) Å). The
plane of terminal carboxylic group (D) is nearly perpen-
dicular to the core of the molecule forming an angle of
87.2(1)° with the mean plane as found in ligand 1. The
relative orientation of two carboxylic groups allows hydrogen
bonds partly responsible for the molecular packing in the
structure.
N2-C1
N3-C2
are arranged in cavities allowing them to rotate around the
simple C-C bond and result statistically distributed in four
positions, a feature frequently found in compounds containing
the allyl fragment.
In complex 9 the copper atom shows a 4 + 1 pyramidal
coordination with the apical position occupied by a water
oxygen and the ligand behaving as a SNO tridentate in the
basal plane. The fourth basal position is occupied by an O3
carboxylate oxygen of an adjacent molecule translated along
the x axis, and polymeric chains are thus formed (Figure 3).
The ligand is doubly deprotonated and acts as a bridge
through a carboxylic group between two copper centers. The
basal plane shows a tetrahedral distortion, and the copper
atom lies 0.14 Å out of the mean plane toward the O1W
atom. The axial bond Cu-O1W is nearly perpendicular to
the basal plane [angle formed with the normal to the plane
is 9.2(1)°]. The coordination geometry around the copper
atom can be described using the angular structural parameter
τ,36 and its value of 0.0065 allows us to confirm a square-
pyramidal arrangement.
The ligand is planar in the portion involved in metal
coordination, while the terminal chain with the second
carboxylic group, which bridges the adjacent molecule, is
rotated of 78.8(6)° as in the corresponding free ligand 1.
The Cu-S1-C1-N2-N3 chelation ring presents a φ2 of
-161(4)° corresponding to a twist-envelope conformation,
while the Cu-N3-C2-C3-O1 is characterized by a φ2 of
-26(3)°, i.e., showing a twist conformation.37
The orientation of the terminal -CH2CH2COOH chain
with respect to the remaining side of the molecule is defined
by the torsion angles C2-C4-C5-C6 -175.0(2)° (-anti-
periplanar) and C4-C5-C6-O3 -173.5(3)° (-antiperipla-
nar). A comparison of the bond distances with those of ligand
1 and those of the unsubstituted H3ct35 shows that also this
ligand, 3, is in the thione form (Table 2); in both carboxylic
groups the C-O distances are very different from each other
as observed in H3ct. The allyl groups face one another and
The Cu-S and Cu-N coordination distances (Table 3)
conform to the values found in the literature and also Cu-
O1W (apical). The difference between the Cu-O (carbox-
ylate) bond lengths could be ascribed to the constraints
determined by the chelation effect.
(36) Addison, A. W.; Rao, T. N.; Reedijk, J.; van Rijn, J.; Verschoor, G.
C. J. Chem. Soc., Dalton Trans. 1984, 1349-1355.
(37) Cremer, D.; Pople, J. A. J. Am. Chem. Soc. 1975, 97, 1354.
7176 Inorganic Chemistry, Vol. 43, No. 22, 2004

Cu(II) r-Ketoglutaric Acid Thiosemicarbazones
Hypochromism and red shift (bathochromism) in electronic
absorption spectra of DNA bound to different compounds
is generally assigned to intercalation, involving a strong
stacking interaction between the aromatic chromophores and
the base pairs of DNA. In this work, for ligands 1, 2, and 6
these features were observed with 0 < [DNA]/[ligand] e 1
ratios, while with [DNA]/[ligand] > 1 ratios the batho-
chromism was conserved but the absorbance values increased
as can be observed in Figure 5 (left) where, as an example,
is reported the absorption spectra of 1.
Figure 4. Plot of [CT-DNA]/(ꢀA - ꢀF) vs [CT-DNA] for absorption
titration with CT-DNA in 50 mM NaCl, 5 mM Tris buffer, pH 7.2 at 25
°C: [, 2; 9, 10.
This behavior can be related to the amount of DNA: at
the first DNA addition the disappearance of the typical ligand
band and the appearance of an absorption band (263 nm)
assignable to the DNA-ligand adduct formation are easily
observable. At 284 nm, where the compound has its
maximum, no absorbance additivity due to the presence of
non interacting species was observed. Further DNA additions
confirmed this trend.
Hydrogen bonds between the apical O1W and the carbon-
ylic O2 of the centrosymmetry-related molecule induce the
formation of a double polymeric chain of metal complexes
which are kept together by two crystallization water mol-
ecules.
As regards the examined complexes (8, 10, and 13) the
band of the nonbound complex disappeared probably because
of the bathochromism in the same wavelength range covered
by the nucleic acid absorbance. This means that these
complexes interact with DNA differently from the ligands
as can be seen from the different spectroscopic behavior. In
Figure 5 (right-hand side) is reported the case of [Cu(Et-
H2ct)Cl]‚H2O (10). With the used ratios (r ) 1, 2, 4, 6, 8),
the DNA amount is the major responsible of the absorption
in the examined region. To make clear if the absorbance
value was due to the compound-DNA interaction and not
to DNA alone, the titration was also carried out with lower
DNA amounts using the following different ratios: r ) 0.2,
0.4, 0.6, 0.8, and 1, where [complex] ) 40 µM. In this case
the DNA absorption did not interfere with the adduct DNA-
complex absorption. The obtained Kb value with lower r does
not differ significantly from that obtained with higher r, and
this confirms our hypothesis (Figure 6).
To further confirm that the added DNA does not interfere
with our titration, the experiment was also carried out in the
visible region, where only the copper complex absorbs. The
ratios used for the experiment were r ) 0.01, 0.02, 0.05,
0.08, 0.1, and 0.2 and [complex] ) 2 mM. The Kb value
found was once again the same. This shows that Kb is
independent from the chosen absorption band and from the
considered concentrations.
Absorption Spectral Features of DNA Binding and
Thermal Denaturation. To evaluate the behavior of these
compounds toward DNA we chose among the ligands
Me-H3ct (1) and Ph-H3ct (6) (since they show the largest
difference in polarity and sterical hindrance of the sub-
stituents) to evaluate if these variables were effective in
influencing DNA-compound interactions. Among the
complexes, [Cu(Me-H2ct)Cl]‚H2O (8) and Cu(Ph-H2ct)Cl
(13) were chosen to study if the presence of the metal
interferes with the processes leading to the binding to the
nucleic acid in comparison to their parent free ligands. Also
[Cu(Et-H2ct)Cl]‚H2O (10), with the ethyl-substituted ligand,
was taken into account because from our biological assays
this compound resulted to be the most active as regards
proliferation inhibition and was therefore subsequently
included in the DNA interaction studies together with its free
ligand Et-H3ct (2).
The binding constants obtained for ligands 1, 2, and 6 were
(1.4 ( 4) × 103, (1.6 ( 5) × 104, and (2.8 ( 5) × 103 M-1,
respectively. The binding constants for complexes 8, 10, and
13 were (4.5 ( 5) × 103, (2.1 ( 5) × 104, and (9.3 ( 6) ×
103 M-1, respectively (each experiment was performed in
triplicate). As an example, in Figure 4 the plot of [DNA]/
(ꢀA - ꢀF) vs [DNA] for absorption titration of ligand (2)
and its copper complex (10) is reported, useful to obtain Kb
from the ratio of the slope to intercept.
The difference with respect to the classical absorption
spectral behavior is due to the possibility for our compounds
to have many hydrogen-bonding interactions because of the
presence of two carboxylic groups and NH functions, in
agreement with refs 32 and 14; in addition, the electronic
spectra of the adduct with DNA suggests the presence of a
new molecular species responsible for the interactions.
Kb values and spectral behavior observed for these
compounds are usually found when aspecific interactions
with DNA are present,14,39 but these data do not exclude an
intercalative behavior.
Our experimental Kb values were lower than those
observed for classical intercalators (ethidium-DNA, 1.4 ×
106 M-1 in 25 mM Tris-HCl/40 mM NaCl buffer, pH ) 7.9
(ref 38), 3.0 × 106 M-1 in 5 mM Tris-HCl/50 mM NaCl
buffer, pH ) 7.2 (this work)), indicating that the compounds
bind DNA with less affinity, as already observed in the
literature for copper(II) complexes with macrocyclic ligands32
and with heterocyclic thiosemicarbazones.14 Moreover since
upon complexation no significant increase of Kb was
observed, we can infer that the binding mode to DNA is
mainly due to the ligand.
(39) Pasternack, R. F.; Gibbs, E. J.; Villafranca, J. J. Biochemistry 1983,
22, 2406-2412.
(38) Le Pecq, J. B.; Paoletti, C. J. Mol. Biol. 1967, 27, 87-106.
Inorganic Chemistry, Vol. 43, No. 22, 2004 7177

Baldini et al.
Figure 5. Variation in IL absorption spectra of compounds Me-H3ct (1) (40 µM, left) and [Cu(Et-H2ct)Cl]‚H2O (10) (40 µM, right), with increasing
amount of CT-DNA. The ratio r (r ) [CT-DNA]/[complex]) varied between 0 (bold line) and 8 in 5 mM Tris-HCl buffer (pH ) 7.2), 50 mM NaCl, at 25
°C. The arrows indicate the directions of absorbance changes as a function of r.
of the biological effects of our compounds with the cells
treated with DMSO alone used as control, ligands 1 and 2
at their maximum reachable concentration in DMSO inhib-
ited cell proliferations of 13% and 30%, respectively, and
copper complexes 8 and 9 showed an inhibition activity of
35%, while for complex 10 it reached 50%. Noteworthy is
that the complexes exert higher biological activity at lower
concentration value with respect to those of the ligands
(Figure 8).
Among all the studied compounds, [Cu(Et-H2ct)Cl]‚H2O
(10) showed the best activity in inhibiting cell proliferation.
For this reason it seemed useful to investigate which phase
of the cell cycle was directly implicated in the biological
behavior of this complex.
Figure 6. Variation in IL absorption spectra of compounds [Cu(Et-H2-
ct)Cl]‚H2O (10) (40 µM), with increasing amount of CT-DNA. The ratio r
(r ) [CT-DNA]/[complex]) varied between 0 (bold line) and 1 in 5 mM
Tris-HCl buffer (pH ) 7.2), 50 mM NaCl, at 25 °C. The arrow indicates
the directions of absorbance changes as a function of r.
A 48 h treatment on leukemic cell line U937 showed a
slow of the cell cycle progression with a 28% reduction of
the DNA synthesis phase (phase S) and with an increase of
cell population in the G1 and G2/M phases. The percentage
distribution of U937 cells in the various phases of the cell
cycle is as follows: in the presence of DMSO 0.4% G1 )
46.4%, S ) 43.8%, and G2 ) 10.4%, and in the presence
of compound [Cu(Et-H2ct)Cl]‚H2O (10) G1 ) 53.5%, S )
31.5%, and G2 ) 15.4%.
To verify this hypothesis, thermal denaturation profiles
of DNA have been studied for compound [Cu(Et-H2ct)Cl]‚
H2O (10). The melting temperature (Tm) was compared with
that of native CT-DNA Tm (70 °C), and the thermal
denaturation profiles and the differential melting curves in
the presence of compound [Cu(Et-H2ct)Cl]‚H2O (10) are
reported in Figure 7 (left- and right-hand side, respectively).
At the studied concentration, complex [Cu(Et-H2ct)Cl]‚
H2O (10), which among these compounds shows the highest
Kb and the lowest IC50 values, shows a good effect on the
melting temperature, suggesting its capability to stabilize the
DNA duplex because it significantly increases Tm by 8.5 °C.34
This experiment confirms the presence of intercalation as
one could expect, particularly from square planar copper
complexes.
Nuclease Activity. A gel electrophoresis experiment using
plasmid pBR322 was performed with all our compounds.
At millimolar concentrations all tested compounds appear
not to affect pBR322 plasmid (data not shown). Under these
experimental conditions there is therefore no production of
hydroxyl (OH‚) radicals that cleave the sugar phosphate
backbone. It can be supposed that our ligands stabilize the
higher oxidation state of the copper center, which conse-
quently does not exert its redox activity.
It was also verified that after a 48 h treatment period there
is no significant apoptosis induction by [Cu(Et-H2ct)Cl]‚H2O
(10).
Conclusions
From the observations reported in this paper a series of
conclusions can be drawn both from a chemical and from a
biological viewpoint.
Looking at the chemical details, we can point out that,
among all the substituents introduced on the aminic nitrogen,
the ethyl derivatives showed the largest activity and this had
already been observed in other aliphatic thiosemicarba-
zones.40 By comparison of the biological activity of these
compounds with analogous complexes with the nonsubsti-
tuted ligand,18 it can be highlighted that substituents decrease
cell proliferation inhibition, except for the case of [Cu(Et-
H2ct)Cl]‚H2O (10) that, on the other hand, follows a different
pathway from apoptosis induction. The presence of the
Effects on Cell Proliferation and Cell Cycle Study.
From toxicity assays carried out on the tested cell line, the
maximum percentage of the DMSO tolerated must not
exceed 0.4-0.5%. The effects of both ligands 1 and 2 and
of the three complexes 8-10 on U937 cell line proliferation
on the fourth day are reported in Figure 8. By comparison
(40) Belicchi Ferrari, M.; Bisceglie, F.; Pelosi, G.; Tarasconi, P.; Albertini,
R.; Pinelli, S. J. Inorg. Biochem. 2001, 87, 137-147.
7178 Inorganic Chemistry, Vol. 43, No. 22, 2004

Cu(II) r-Ketoglutaric Acid Thiosemicarbazones
Figure 7. Thermal denaturation profiles (left) and differential melting curves (right) of calf thymus DNA before (a) and after addition of [Cu(Et-H2ct)-
Cl]‚H2O (10) at r ) 0.01 (b) where r ) [complex]/[DNA]. DNA concentration: 4.5 × 10-5 M. Buffer: 50 mM NaCl, 5 mM Tris-HCl.
S because of the impossibility for the cell to cross the damage
checkpoint. From the cell cycle analysis we notice instead
an accumulation in phase G2 and a depletion of the cells in
phase S that is unexpected in the hypothesis of a ribonucle-
otide reductase inhibition.
Another suggested way of action of these copper com-
pounds is related to their potential capability of generating
OH radicals and causing oxidation stress with irreversible
damage to DNA with subsequent induction of apoptosis.
With our electrophoresis experiments also this mechanism
can be excluded since we observe only a tiny amount of
Figure 8. U937 cell proliferation inhibition for compounds 1 (Me-H3ct;
nicked plasmid and none in its linear form even with
429 µM), 2 (Et-H3ct; 377 µM), 8 ([Cu(Me-H2ct)Cl]‚H2O; 86 µM), 9 ([Cu-
relatively high concentrations of the complexes.
(Me-Hct)(OH2)]n‚2nH2O; 91 µM), and 10 ([Cu(Et-H2ct)Cl]‚H2O; 24 µM).
As regards their biological activity, we can therefore
conclude that these compounds interfere with the cell cycle
at different levels but the mechanisms that lie behind their
behavior are still unclear.
thiosemicarbazone free aminic group seems to be determining
in inducing apoptosis.17,18
Thiosemicarbazones and related molecules have been
suggested to act as iron chelators, and their mechanism of
action has been assigned to their ability to interfere with
DNA synthesis by sequestering iron from ribonucleotide
reductase and thus preventing the production of deoxyribo-
nucleotides.41 If this was the case, we would observe first
of all a larger activity of ligands if compared with their metal
complexes and moreover an accumulation of cells in phase
Acknowledgment. This work was supported by the
Ministero dell’Istruzione Universita` e Ricerca of Italy
(MIUR, cofin. 2001053898).
Supporting Information Available: Crystallographic data in
CIF format. This material is available free of charge via the Internet
at http://pubs.acs.org.
(41) Richardson, D. R.; Milnes, K. Blood 1997, 89, 3025-3038.
IC049883B
Inorganic Chemistry, Vol. 43, No. 22, 2004 7179