JOURNAL OF

www.elsevier.com/locate/jinorgbio

Journal of Inorganic Biochemistry 99 (2005) 1205–1219

InorganicBiochemistry

Copper(II) complexes of 1,10-phenanthroline-derived ligands:Studies on DNA binding properties and nuclease activity

Tomoya Hirohama a, Yuko Kuranuki a, Eriko Ebina a,Takashi Sugizaki a, Hidekazu Arii a, Makoto Chikira a,*,Pitchumony Tamil Selvi b, Mallayan Palaniandavar b,*

a Department of Applied Chemistry, Chuo University, 1-13-27 Kasuga, Bunkyo-ku, Tokyo 112-8551, Japanb Department of Chemistry, Bharathidasan University, Tiruchirappalli 620 024, India

Received 25 October 2004; received in revised form 18 February 2005; accepted 19 February 2005Available online 25 March 2005

Abstract

A series of copper(II) complexes of the type [Cu(L)]2+, where L = N,N 0–dialkyl-1,10-phenanthroline-2,9-dimethanamine andR = methyl (L1), n-propyl (L2), isopropyl (L3), sec-butyl (L4), or tert-butyl (L5) group, have been synthesized. The interactionof the complexes with DNA has been studied by DNA fiber electron paramagnetic resonance (EPR) spectroscopy, emission, viscos-ity and electrochemical measurements and agarose gel electrophoresis. In the X-ray crystal structure of [Cu(HL2)Cl2]NO3, cop-per(II) is coordinated to two ring nitrogens and one of the two secondary amine nitrogens of the side chains and two chlorideions as well and the coordination geometry is best described as trigonal bipyramidal distorted square based pyramidal (TBDSBP).Electronic and EPR spectral studies reveal that all the complexes in aqueous solution around pH 7 possess CuN3O2 rather thanCuN4O chromophore with one of the alkylamino side chain not involved in coordination. The structures of the complexes in aque-ous solution around pH 7 change from distorted tetragonal to trigonal bipyramidal as the size of the alkyl group is increased. Theobserved changes in the physicochemical features of the complexes on binding to DNA suggest that the complexes, except[Cu(L5)]2+, bind to DNA with partial intercalation of the derivatised phen ring in between the DNA base pairs. Electrochemicalstudies reveal that the complexes prefer to bind to DNA in Cu(II) rather than Cu(I) oxidation state. Interestingly, [Cu(L5)]2+ showsthe highest DNA cleavage activity among all the present copper(II) complexes suggesting that the bulky N-tert-butyl group plays animportant role in modifying the coordination environment around the copper(II) center, the Cu(II)/Cu(I) redox potential and hencethe formation of activated oxidant responsible for the cleavage. These results were compared with those for bis(1,10-phenanthro-line)copper(II), [Cu(phen)2]

2+.� 2005 Elsevier Inc. All rights reserved.

Keywords: Copper(II); 1,10-Phenanthroline-derived ligands; DNA fiber; Absorption and emission spectroscopy; Electron paramagnetic resonance;Electrochemistry; Oxidative DNA cleavage

1. Introduction

Studies of small molecules, which react at specificsites along a DNA strand as reactive models forprotein–nucleic acid interactions, provide routes toward

0162-0134/$ - see front matter � 2005 Elsevier Inc. All rights reserved.

doi:10.1016/j.jinorgbio.2005.02.020

* Corresponding author. Tel.: +81 3 3817 1897; fax: +81 3 38171895.

E-mail address: chikira@chem.chuo-u.ac.jp (M. Chikira).

rational development of chemotherapeutic agents [1],sensitive chemical probes for DNA structure in solution[2], and tools for the molecular biologist to dissect genet-ic systems [3]. Especially, a number of metal complexesof a variety of ligands have been studied in view of theirpossibility to lead to advanced functional materials,tuning the redox potentials, affinity towards DNA,and specificity for the DNA base sequence recognition[4–7]. Transition metal complexes with their versatile

1206 T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219

structures, redox behavior and physicochemical proper-ties are found to be useful as highly sensitive diagnosticagents and the metal or ligands in the complexes can bevaried in an easily controlled way to facilitate individualapplication. Over the past decades the potential applica-tion of polypyridyl transition metal complexes as non-radioactive nucleic acid probes and DNA cleavingagents that can recognize and cleave DNA has stimu-lated vigorous research [5]. Thus, copper complexes con-taining 1,10-phenanthroline (phen) have receivedconsiderable current interests in nucleic acids chemistrydue to their various applications following the discoveryof the �chemical nuclease� activity of bis(phen)copper(I)complex (Fig. 1(a)) in the presence of hydrogen peroxideand a reducing agent by Sigman et al. [5]. The metal-che-lating properties of phen ligand and its derivatives havebeen utilized in a range of analytical reagents as well asfor the development of biological probes [9]. Among thecopper–phen based synthetic nucleases, bis(phen)coppercomplexes are directed toward the oxidative cleavage ofDNA in the presence of a reducing agent [8,10]. Thereare also some reports of copper–phen complexes cleav-ing DNA hydrolytically [11]. Several copper complexescontaining phen moiety have been synthesized and stud-ied as sequence specific DNA binding or cleaving re-agents [12–16].

An understanding of the modes of binding of metalcomplexes to DNA is required to illustrate the principlesgoverning the DNA recognition by such functional mol-ecules, that is, the factors that decide the affinity andspecificity of the complexes for DNA base sequence.Cationic complexes have been found to both intercalateinto DNA and bind non-covalently in a surface-boundgroove-bound fashion [17]. To assess the mode ofDNA binding our laboratories have been employingseveral spectroscopic, electrochemical and other tech-niques. In one of our laboratories the binding structureand dynamic behavior of paramagnetic metal complexeson orientated DNA fibers have been successfully studiedfor the past several years [18–22]. In the course of suchinvestigations, we have reported that [Cu(phen)]2+ andthe ternary complexes [Cu(phen)(AA)]n+, where AA

N N HN

R R

NH

NN

R : alkyl group

(a) (b)

Fig. 1. (a) 1,10-Phenanthroline (phen), (b) N,N 0-dialkyl-1,10-phenan-throline-2,9-dimethanamine. R = methyl (1); n-propyl (2); isopropyl(3); sec-butyl (4); tert-butyl group (5). Numbers in the parenthesescorrespond to the copper(II) complexes, respectively.

stands for an L-amino acid, bind to DNA both interca-latively and non-intercalatively [21]. It is also revealedthat methyl substituents at 5- and 6-positions of thephen ring (5,6-dimethyl-1,10-phenanthroline, 5,6-dmp)promoted whereas those at 2- and 9-positions (2,9-di-methyl-1,10-phenanthroline, 2,9-dmp) interfered withthe intercalative DNA binding of the complex [23]. Infact, [Cu(2,9-dmp)]2+ fails to cleave DNA. One of ourresearch groups has shown that the [Cu(5,6-dmp)]2+ in-duces B to Z conformational change on binding to calfthymus DNA (CT-DNA) [24].

Recently, Lin and co-workers [25] have prepared the1:1 copper(II) complexes of certain N,N 0-dialkyl-1,10-phenanthroline-2,9-dimethanamine (Fig. 1(b)) and stud-ied the thermodynamic and kinetic properties of thecomplex-DNA binding. They showed that the com-plexes interact with CT-DNA by both intercalativeand covalent binding and that there are at least twosteps in the binding process. The Cu(II) complexes ofthese chelating 4N ligands are expected to be very stableand hence suitable for interaction with DNA. Further,the alkyl group substitution at the secondary amine moi-ety would affect the affinity of the complexes for DNAby raising the basicity of the donor nitrogen atomsand by introducing steric hindrance for coordination.Therefore, it is interesting to investigate systematicallythe effect of varying the coordination structure bychanging the N-alkyl groups of the 4N ligands (Fig. 1)on the mode and extent of interaction of the variousCu(II) complexes with DNA. In this investigation, wefocus our attention on the structure of the complexesbound to DNA and the oxidative DNA cleavage withthese complexes as well. The interaction has been inves-tigated using a host of physical methods like electronicabsorption spectroscopy, DNA fiber electron paramag-netic resonance (EPR) spectroscopy, viscosity measure-ments, competitive ethidium bromide studies andelectrochemical methods. The cleavage reactions havebeen assayed using agarose gel electrophoresis and thecleavage efficiency of the complexes illustrated usingthe Cu(II)/Cu(I) redox potential. Furthermore, these re-sults are compared with those of bis-1,10-phenanthro-line copper(II) complex, [Cu(phen)2]

2+, which has longbeen known as one of the potent oxidative nucleases,but the binding structure of [Cu(phen)2]

2+ bound toDNA has not been fully characterized [26–28].

2. Experimental

2.1. Materials

All the N,N 0-dialkyl-1,10-phenanthroline-2,9-dime-thanamine ligands (L1–L5) were prepared according tothe literature procedure [29] and their identity checkedusing their 1H NMR spectra. Reagents and solvents

Table 1Crystallographic data for [Cu(HL2)Cl2](NO3) (2a)

Empirical formula C20H27Cl2CuN5O3

Formula weight 519.92T (K) 123Wavelength (A) 0.71070Crystal size (mm) 0.35 · 0.20 · 0.20Crystal system MonoclinicSpace group P21/c (#14)Unit cell dimensions

a (A) 9.623(2)b (A) 8.294(2)c (A) 27.921(7)b (�) 91.007(3)

V (A3) 2228.2(9)Z 4Dcalc (g cm�3) 1.550Absorption coefficient (cm�1) 12.52F(000) 1076.00Number of registered reflections 34422Number of unique data (Rint) 5425 (0.078)Number of refined parameters 307Final R indices [I > 2r(I)] 2482R indicesa (all data) R1 = 0.037, wR2 = 0.073

a R1 =P

iFo|�|Fci/P

|Fo|; wR2 ¼ fP

ðwðF 2o � F 2

cÞ2Þ=

PwðF 2

oÞ2g1=2.

T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219 1207

were of commercially available reagent quality and wereused without further purification. 1,10-phenanthrolinemonohydrate, selenium dioxide and superoxidedismutase from bovine erythrocytes were obtained fromNacalai Tesque Inc. Copper(II) nitrate trihydrate, 2,9-dimethyl-1,10-phenanthroline dihydrate, methylamine(40% in methanol solution), n-propylamine, isopropyl-amine, sec-butylamine, and tert-butylamine were pur-chased from Tokyo Kasei Kogyo Co., Ltd. Salmontestes and calf thymus DNA, and catalase from bovineliver were purchased from Sigma Chemical Co. ThepBR322 DNA was obtained from TAKARA ShuzoCo., Ltd. The concentration of DNA per nucleotidewas determined by absorption spectroscopy using themolar absorption coefficient (6600 M�1 cm�1) at 260nm [30]. Water purified using a Milli-Q system was usedfor all the present studies.

2.2. Synthesis of complexes

2.2.1. Cu(HL)(NO3)3 (1–5)The nitrate complexes were synthesized according to

the method for perchlorate complexes [25]. The corre-sponding ligand was dissolved in ethanol and mixedwith a solution of Cu(NO3)2 Æ 3H2O in ethanol. Thereaction mixture was stirred at 50–60 �C for 1 h. Theresulting precipitate was collected and dried in vacuo.

Anal. Calc. for Cu(HL1)(NO3)3 Æ EtOH Æ 1.5H2O: C,36.64; H, 4.78; N, 16.62. Found: C, 36.45; H, 3.57; N,16.64%; Cu(HL2)(NO3)3 Æ 0.25EtOH Æ H2O: C, 42.85;H, 5.51; N, 15.55. Found: C, 43.05; H, 5.03; N,15.19%; Cu(HL3)(NO3)3 Æ 0.75EtOH: C, 42.50; H,5.23; N, 16.14. Found: C, 42.87; H, 4.85; N, 15.84%;Cu(HL4)(NO3)3 Æ 0.75EtOH: C, 42.50; H, 5.23; N,16.14. Found: C, 42.31; H, 4.87; N, 16.10%; Cu(HL5)-(NO3)3 Æ 0.5EtOH Æ 0.5H2O: C, 43.63; H, 5.57; N,15.49. Found: C, 43.45; H, 5.27; N, 15.04%.

2.2.2. [Cu(HL2)Cl2]NO3 (2a)The hydrochloride salt of L2 was dissolved in metha-

nol and mixed with a solution of Cu(NO3)2 Æ 3H2O inmethanol. The reaction mixture was stirred at 50–60�C for 1 h. The resulting precipitate was collected andredissolved in methanol. Slow vapor diffusion of diethylether into the methanol solution of the complex gave agreen prismatic single crystal of [Cu(HL2)Cl2]NO3 suit-able for X-ray diffraction.

2.3. X-ray diffraction studies of [Cu(HL2)Cl2]NO3 (2a)

The single crystal was mounted on a glass fiber andall measurements were made on a Rigaku MercuryCCD area detector with graphite monochromated MoKa radiation. The details of measurement conditionsand crystal data are listed in Table 1. The intensity datawere collected at a temperature of �150 ± 1 �C using the

x–2h scan technique to a maximum 2h value of 55.0�. Atotal of 744 oscillation images were collected. Of the34422 reflections that were collected, 5425 were unique(Rint = 0.078); equivalent reflections were merged. Datawere collected and processed using the CrystalClear pro-gram (Rigaku). The linear absorption coefficient, l, forMo Ka is 12.52 cm�1. The data were corrected for Lor-entz and polarization effects. The structures were solvedby a direct methods (SIR-92) [31] and expanded usingFourier techniques [32]. The non-hydrogen atoms wererefined anisotropically. Hydrogen atoms were refinedusing the riding model. The final cycle of full-matrixleast-squares refinement on F2 was based on 2482 ob-served reflections, and 307 variable parameters. Reliabil-ity factors are defined as R1 =

PiFo|�|Fci/

P|Fc| and

wR ¼ ½P

wðF 2o � F 2

cÞ2=P

wðF 2oÞ

2�1=2. The final R1 andwR values were 0.037 and 0.073, respectively. The max-imum and minimum peaks on the final difference Fou-rier map corresponded to 0.54 and �0.48 e A�3,respectively. Neutral atom scattering factors were takenfrom Cromer and Waber [33]. All other calculationswere performed using the CrystalStructure [34,35] crys-tallographic software package.

2.4. Preparation of DNA pellet and fibers

DNA-fibers were prepared as previously described[22]. In this experiment, a required amount of the com-plex solution was added very slowly to DNA solutions(20 mM aq. NaCl) under stirring to make the molar ra-tio (r) = [base-pair]/[Cu(II) complex] = 10.

1208 T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219

The DNA-pellets were obtained by ultracentrifuga-tion of the mixed solution and the DNA fibers were fab-ricated from the pellet. The amount of the coppercomplexes bound to DNA was estimated by determiningthe concentration of the unbound copper complexes inthe upper supernatant solution obtained afterultracentrifugation.

2.5. EPR measurements

The X-band EPR spectra of the complexes at roomtemperature and �150 �C were measured on a JEOLJES-2X spectrometer with 100 kHz field modulation of0.5 mT and a microwave power of 10 mW. The magneticfield was calibrated with an NMR field meter EFM-2000(ECHO Electronics Ltd.). The frequency of the micro-wave was measured on Anritu MF2412A microwavefrequency counter.

The EPR spectra of frozen aqueous solutions weremeasured after adding a small amount of ethyleneglycolto the solutions. The A-form DNA-fibers placed on thequartz fiber holder were shielded in an EPR tube with asmall piece of wet cotton wool. Though it took severaldays until the EPR line shapes became stationary, thisarrangement of the holder made it easy to measure theEPR spectra at various stages of the transformation ofthe fibers from A- to B-form at various temperatures.The definitions of the angles and the relative orienta-tions of the fiber axis, magnetic field, and the g-tensoraxes are shown in Fig. 2. The EPR spectra were mea-sured at different angles of U between the fiber axisand the static magnetic field.

For spin trapping experiments, a solution containingcomplex 5 (0.1 mM), hydrogen peroxide H2O2 (1 mM)and spin trap agent 5,5-dimethyl-1-pyrroline N-oxide(DMPO) (100 mM) in phosphate buffer (50 mM) were

Φ g//

θ DNA-fiber axis

B : Magnetic field

g

Zf

Yf

Xf

Fig. 2. Coordinate systems: B, static magnetic field; (Xf,Yf,Zf), DNA-fiber axes; (gi,g^), g-tensor axes; U, the angle between B and Zf; h, theangle between gi and Zf axes.

prepared. After the addition, the solution was incubatedat 37 �C for 30 min and the EPR spectrum was mea-sured at room temperature.

2.6. Other physical methods

Following physical measurements were carried out in5 mM tris(hydroxylmethyl)aminomethane (Tris)–HCl/50 mM NaCl buffer, pH 7.1. Absorption spectra wererecorded with a Cary 300 BIO UV–Vis using cuvetteof 1 cm path length.

Emission intensity measurements were carried outusing JASCO FP 6500 spectrofluorometer. The tris buf-fer was used as a blank to make preliminary adjust-ments. The excitation wavelength was fixed and theemission range was adjusted before measurements.DNA was pretreated with ethidium bromide (EthBr)in the ratio [NP]/[EthBr] = 1 for 30 min at 27 �C(NP = nucleotide phosphate). The metal complexeswere then added to this mixture and their effect on theemission intensity was measured.

Viscometric measurements were performed with aSchott Gerate AVS 310 Automated Viscometer. The vis-cometer was thermostated at 25 �C in a constant temper-ature bath. The concentration of DNA was 500 lM inNP and the flow times were measured with a manuallyoperated timer.

All voltammetric experiments were performed in asingle compartment cell with a three-electrode configu-ration on a EG&G PAR 273 potentiostat equipped witha PIV computer. The working electrode was a glassycarbon disk (0.384 cm2) and the reference electrodewas Ag/AgCl. A platinum electrode was used as thecounter electrode. The supporting electrolyte was5mM Tris–HCl/50 mMNaCl buffer at pH 7.1. Solutionswere deoxygenated by purging with nitrogen gas for 15min prior to measurements; during measurements astream of N2 gas was passed over the solution. All theexperiments were carried out at 25 ± 0.2 �C maintainedby a Haake D8-G circulating bath.

2.7. DNA cleavage experiments

Reaction solutions were prepared by mixing 2 lL of50 lM Cu(II) complex solution (pH 7.4), 2 lL of 50lM super coiled pBR322 DNA, 2 lL of 100 mM N-[2-hydroxyethyl]piperazine-N 0-[2-ethanesulfonic acid](Hepes), followed by dilution with sterilized water to atotal volume of 20 lL. The reaction of DNA cleavagewere initiated by adding 2 lL of 1 mM H2O2 and incu-bated at 37 �C. To investigate the nature of the activeoxygen species generated during the course of thesereactions, standard radical scavengers: superoxide dis-mutase (SOD) (3 unit), dimethyl sulfoxide (DMSO)(100 mM), sodium azide (NaN3) (20 mM) and catalase(3 unit) were added, respectively. The reaction was

Table 2Selected bond lengths and angles of [Cu(HL2)Cl2](NO3) 2a

Bond lengths (A)

Cu(1)–N(1) 2.211(4)Cu(1)–N(2) 1.946(3)Cu(1)–N(4) 2.080(4)Cu(1)–Cl(1) 2.251(1)Cu(1)–Cl(2) 2.480(1)

Bond angles (�)N(1)–Cu(1)–N(2) 79.1(1)N(1)–Cu(1)–N(4) 156.0(1)N(1)–Cu(1)–Cl(1) 99.0(1)

T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219 1209

quenched by adding 2 lL of 1 mM quench buffer (100mM EDTA, bromophenol blue, xylene cyanole, su-crose). DNA cleavage was evaluated from the conver-sion of supercoiled DNA to nicked DNA by agarosegel electrophoresis. Cleavage products were separatedon 1% agarose gel in 1· TAE buffer for 1.5 h at 60 Vand stained using 0.5 lg/ml ethidium bromide for 30min. The cleavage activity was evaluated by photoimag-ing densitometry using KODAK 1D software for fluo-rescence intensity of each band by the sum offluorescence intensities for all bands in that lane.

N(1)–Cu(1)–Cl(2) 98.3(1)N(2)–Cu(1)–N(4) 80.2(1)N(2)–Cu(1)–Cl(1) 159.6(1)N(2)–Cu(1)–Cl(2) 102.0(1)N(4)–Cu(1)–Cl(1) 96.3(1)N(4)–Cu(1)–Cl(2) 97.6(1)Cl(1)–Cu(1)–Cl(2) 98.44(4)

3. Results and discussion

3.1. Synthesis of complexes

All the present complexes were isolated by treatingthe ligands with equimolar quantity of copper(II) ni-trate. They were formulated as CuHL(NO3)3based onelemental analysis. The formulation of [Cu(HL2)Cl2]-NO3 (2a) was confirmed by X-ray structure analysis.

3.2. Structural studies of complexes

The ORTEP representation of the structure of[Cu(HL2)Cl2]NO3 (2a) is shown in Fig. 3 and selectedbond lengths and bond angles are gathered in Table 2.The copper atom in the complex is coordinated by thetwo ring nitrogen atoms and one of the two secondaryamine nitrogens of the side chains of the 1,10-phenan-throline-derived ligand and two chloride ions. The othersecondary amine nitrogen of the ligand is protonated andis not coordinated to copper. The coordination geometryaround copper(II) is best described as trigonal bipyrami-dal distorted square based pyramidal (TBDSBP) as indi-cated by the value (0.06) of the trigonal indexs[=(a � b)/60, where a = N(2)–Cu(1)–Cl(1) = 159.6(1)�

Fig. 3. Crystal structure of 2a with atom numbering scheme. Thehydrogen atoms have been omitted for clarity.

and b = N(1)–Cu(1)–N(4) = 156.0(1)�; s is unity for aregular trigonal bipyramidal geometry and is zero for aperfect square pyramidal geometry] [36]. Thus, all thethree nitrogen atoms (N(1), N(2) and N(4)) of the merid-ionally coordinated H(L2)+ ligand and one chloride ion(Cl(1)) occupy the corners of the square plane. The sec-ond chloride ion (Cl(2)) occupies the axial position at dis-tance (Cu–Cl(2), 2.480 A) longer than the equatorial one(Cu–Cl(1), 2.251 A) due to the presence of two electronsin the dz2 orbital. The Cu–Nphen bond (Cu–N(2), 1.946A) is shorter than the Cu–Namine bond (Cu–N(3), 2.080A), which is expected of the sp2 and sp3 hybridisations,respectively, of the phen and amine nitrogens. However,the other Cu–Nphen bond (Cu–N(1), 2.211 A) is longerthan the Cu–Nphen bond (Cu–N(2)) because of the sterichindrance from adjacent alkylaminomethyl substituenton the phen ring. The two coordinated chloride ionsare probably substituted by water molecules in aqueoussolution.

All the complexes except 5 in the frozen solution atpH 7.4 display the axially symmetric EPR spectra (gi:2.265–2.271 and Ai: 0.0160–0.0130 cm�1, see Table 3)(Fig. S1). The electronic spectra of the complexes inaqueous solution are very similar to each other and dis-play a broad low energy ligand field (LF) band (700–741nm) suggesting a square-based coordination geometryaround copper(II). Above results are consistent withthe crystal structure of the L2 complex. It is well knownthat a geometric distortion away from square planargeometry leads to an increase in gi and a decrease inAi values. So the higher gi and lower Ai values observedfor the present complexes suggest the presence of a sig-nificantly distorted tetragonal or a distorted trigonalbipyramidal geometry in solution. On the other hand,5 shows a rhombic EPR spectrum and a LF band withhigher kmax value. Thus, the kmax increases and therhombicity of g and A tensors increase on going from

Table 3EPR spectral data of Cu(II) complexes

Complex Solutiona,d gi(Ai/cm

�1)DNA-pelleta gi(Ai/cm

�1)DNA-fiber (A-form)b

gi (Ai/cm�1)

DNA-fiber (B-form)b

gi (Ai/cm�1)

DNA-fiber (B-form)a

gi (Ai/cm�1)

CuðphenÞ2þ2 2.270 (0.0166) – 2.327 (0.0120) – –1 2.271 (0.0124) 2.264 (0.0143) 2.263 (0.0148) 2.259 (0.0144) 2.261 (0.0155)2 2.269 (0.0130) 2.265 (0.0141) 2.264 (0.0144) 2.258 (0.0140) 2.262 (0.0152)3 2.269 (0.0117) 2.267 (0.0121) 2.270 (0.0134) 2.263 (0.0132) 2.273 (0.0141)4 2.265 (0.0117) 2.270 (0.0116) 2.266 (0.0137) 2.259 (0.0137) 2.264 (0.0147)5c 2.253 (0.0106) 2.284 (0.0126) 2.308 (0.0117)

2.123 2.267 (0.0149) 2.277 (0.0128)2.026 (0.0092) 2.322 (0.0113) 2.334 (0.0139)

a �150 �C.b Room temperature.c Rhombic parameters for frozen solution; parameters for two species in DNA-fiber (B-form).d pH 7.4; for complexes, 2, 3, and 4, the largest g and A values of rhombic parameters are shown.

1210 T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219

1 to 5, revealing an increase in structural distortion froma square-based geometry on changing the N-alkyl sub-stituents from methyl to the more sterically demandingtert-butyl group.

Further, we have shown that square planar Cu(II)complexes of certain tridentate 3N pyridylmethylethy-lenediamine ligands display a LF band in the range645–660 nm and axially symmetric frozen solutionEPR spectra with gi and Ai values around 2.250 and0.0180 cm�1, respectively. Also a planar CuN4 chromo-phore is expected [37] to show a LF band at still lowerwavelength and gi and Ai values around 2.200 and0.0200 cm�1. So the present complex favors a stronglydistorted CuN3O chromophore in solution.

Interestingly, the frozen solution EPR spectra of 1

(Fig. 4(a)) obtained in the pH range 6–10 show thatthe gi (2.260) and Ai (0.0151 cm�1) values of the speciesat pH 6 change to 2.271 and 0.0124 cm�1, respectively,at pH 7 indicating that 1 undergoes deprotonation inthis pH range. An unusual line width dependence for

360310260

B/mT

pH 6.0

pH 7.0

pH 8.0

pH 9.0

pH 10.0

(a)

350300250

B/mT

pH 6.0

pH 7.0

pH 8.0

pH 9.0

pH 10.0

(b)g//:2.25

g//:2.29

g//:2.26

g//:2.27

Fig. 4. pH dependence of the EPR spectra of (a) 1 and (b) 5 in frozensolutions at �150 �C.

M(I) = �3/2, �1/2, etc. in Fig. 4(a) indicates that thespecies with gi = 2.27 may exist a little even at pH 6.0.The frozen solution EPR spectrum of 2 also shows sim-ilar changes. Therefore, it is expected that one of the sec-ondary amine nitrogens in 1 and 2 is protonated and isnot coordinated to copper(II) in solution at pH 6 andthe observed spectral changes at higher pH correspondto the deprotonated secondary amine nitrogen, whichis not coordinated to copper(II) due to its unfavourablegeometric disposition. In fact, it has been already re-ported [38] that in the copper(II) complex of the relateddioxime of 1,10-phenanthroline-2,9-dialdehyde only oneof the two oxime nitrogens is involved in coordinationto copper. Interestingly, the pH dependence of theEPR spectra of 5 indicate that the deprotonation pro-gresses at around pH 8, where two different species areobserved (pH 6, 7: g3 = 2.252, A3 = 0.0101 cm�1; pH10: g3 = 2.287, A3 = 0.0082 cm�1, Fig. 4(b)).

The frozen solution EPR spectrum of [Cu(phen)2]2+

(Fig. S1(a)) shows gi (2.270) > g^, suggesting that thecomplex ion has a tetragonal pyramidal structure inaqueous solution. Also, in the solid state, [Cu(phen)2]

2+

assumes a distorted tetragonal pyramidal or trigonalbipyramidal structure depending on the counter ion[39–41]. Thus, the EPR results suggest that in aqueoussolution a water molecule is coordinated to [Cu-(phen)2]

2+ to form a tetragonal pyramidal geometry.

3.3. DNA fiber EPR spectral study

The EPR spectrum of 1 in DNA pellet exhibits axiallysymmetric line shape with gi = 2.264 (Fig. S2(a)). In theA-form DNA-fibers (Fig. S2(b)), 1 shows U-dependentEPR spectra in which the gi signal becomes the most in-tense at U = 0�. The conformational change of DNA-fiber from A- to B-form does not affect much the EPRline shape both at room and low (�150 �C) temperatures(Fig. 5). As the gi axis is parallel to the normal of aver-age coordination plane, one can reasonably concludethat the plane of the species is oriented parallel to

350300250B/mT

350300250B/mT

Φ=0º

30º

60º

90º

Φ=0º

30º

60º

90º

(a) (b)

Fig. 5. EPR spectra of 1 on B-form DNA-fibers for (a) at roomtemperature and for (b) at �150 �C.

T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219 1211

DNA base pairs. These results indicate that considerableamount of 1 is bound to DNA through intercalativemode. The EPR spectra of 2 and 3 on B-form DNA-fi-bers (Fig. 6(a) and (b)) are similar to those of 1 indicat-ing that the n-propyl or isopropyl groups do notinterfere with the intercalative binding. Similar EPRspectra have been observed for the unsubstituted phencomplex [Cu(phen)]2+ and the ternary complexes, [Cu(-phen)(AA)]n+, where H(AA) stands for amino acid [21].The EPR spectra of 4 (Fig. 6(c)) show line shapes a littledifferent from those of 1, 2, and 3; an enhanced defor-mation of the complex toward trigonal-bipyramidalgeometry might have increased the rhombicity of theg-tensor and/or the amount of non-intercalated species.

The EPR spectra of 5, which has the bulkiest alkylgroup among the present complexes, are quite different

Φ=0º

30º

60º

90º

Φ=0º

30º

60º

90º

350300250B/mT

300250B/mT

(a) (b)

Fig. 6. EPR spectra of: (a) 2, (b) 3 and (c) 4 on

from those of 1–4, which is consistent with the electronicspectral study. In DNA pellet 5 shows a broad EPR lineshape with irregularly spaced hyperfine splitting in the giregion (Fig. S3(a)) indicating that several different spe-cies are bound to DNA. Also the spectra of 5 on A-formDNA-fibers (Fig. S3(b)) change with U, but do not showthe signals observed for the DNA-pellet at the highermagnetic field end (320–340 mT). This is consistent withchanges in the coordination geometry of copper(II) fromtrigonal bipyramidal to tetragonal pyramidal during thefabrication of A-form DNA fibers. The EPR line shapesbecome sharper and the U-dependence becomes moreconspicuous after the conformational change of DNA-fibers from A- to B-form as shown in Fig. 7(a). These re-sults indicate clearly that at least two species of 5 arebound with different g values and similar orientationson the DNA.

Freezing the B-form DNA fibers results in furtherchanges in the EPR line shape; new gi signals character-istic of the species with distorted tetragonal form emergeat lower magnetic field end of the spectra at U = 90�(Fig. 7(b)). These results indicate that 5 is so flexible incoordination geometry that it can take various struc-tures depending on the temperature and the DNA bind-ing mode. It is obvious that such a structural change iscaused by an interplay of DNA and the bulky tert-butylgroups, which interfere with the deep intercalation of thephen moiety only to lead the complex near the surface ofthe double helical DNA. As the orientation of g-tensorof 5 on the molecular frame is not clear, it is difficultto describe the orientation of the phen plane of each spe-cies on the DNA. However, that they bind stereospecif-ically on the DNA.

The spectral characteristics of planar coppercomplex 5 bound to DNA are compared with those of

Φ=0º

30º

60º

90º

350 350300250B/mT

(c)

B-form DNA-fibers at room temperature.

Φ=0º

30º

60º

90º

Φ=0º

30º

60º

90º

350300250B/mT

350300250B/mT

(a) (b)A

BA

B

Fig. 7. Temperature dependence of the EPR spectra of 5 on B-formDNA-fibers for (a) at room temperature and (b) at �150 �C.

1212 T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219

non-planar copper(II) complexes like [Cu(phen)2]2+

bound to DNA to understand further the DNA bindingproperties of the present complexes. Though [Cu(-phen)2]

2+ in frozen solution at pH 7.4 exhibits a typicalEPR signal of a tetragonal pyramidal copper(II) com-plex (gi = 2.270), it exhibits a fairly broad signal cen-tered at g @ 2.13 on DNA pellet (data not shown). TheEPR spectra of the complex bound to A- and B-formDNA fibers change with U as shown in Fig. 8 indicatingthat the complex is stereospecifically orientated on theDNA. In addition, the conformational change of theDNA from A- to B-form is accompanied by a structuralchange of the copper complex. It has been suggestedthat the cuprous complex [Cu(phen)2]

+ cleaves B-formDNA much faster than it does A-form DNA because

Φ=0º

30º

60º

90º

Φ=0º

30º

60º

90º

350300250B/mT

350300250B/mT

(a) (b)

Fig. 8. EPR spectra of [Cu(phen)2]2+ on DNA-fibers at room

temperature: (a) A-form; (b) B-form.

the complex has stronger affinity for the minor grooveof B-form DNA than for that of A-form [42–46]. Thepresent results suggest that the difference in the coordi-nation structures of [Cu(phen)2]

2+ bound to A- and B-forms of DNA may be another factor responsible forthe difference in the rate of the reaction [47]. It shouldbe also noted that the change in the rhombic EPR lineshape with U quite resembles that observed for 5; thegi signals observed at U = 0� (Fig. 8(b)) disappear atU = 90� and the intensity of the peak around 320 mT in-creases considerably with an increase in U. The gi signalat U = 0� disappears again on freezing the B-form DNAfibers at �150 �C (data not shown). It has been observedthat crystallization of water in the B-form fibers at lowtemperatures disrupt the orientation of DNA and de-form the binding site accounting for the inhomogeneousbroadening of the spectra. These changes in the EPRspectra suggest that [Cu(phen)2]

2+ is as flexible as 5

and has a structure in between tetragonal pyramidaland trigonal bipyramidal on DNA.

3.4. Absorption spectral study of DNA binding

The B-form of CT-DNA is a polyanion composed oftwo complementary polymeric subunits hydrogenbonded together in the form of a right-handed double-helix [48]. When the present dicationic complexes inter-act with DNA, it is likely to replace two sodium ionsfrom the compact inner (Stern) layer or the diffuse outerlayer surrounding DNA and interact with the anionicphosphate residues of DNA [49]. The two labile watermolecules/chloride ions in the complex are replaced bya nucleophile on DNA, usually a nitrogenous base suchas guanine N7, leading to a strong covalent bonding ofthe complexes with DNA [50]. But the presence of sterichindrance from the uncoordinated alkylamine side chainwould prevent such a covalent interaction but not thepartial insertion of the phen ring into the DNA basepairs.

Electronic absorption spectroscopy is widely em-ployed to determine the DNA binding affinity of metalcomplexes. The present copper(II) complexes do not ex-hibit any intense d–d or charge transfer band to monitortheir interaction with DNA. So the intense ligand based(p ! p*) absorption band is used to monitor the interac-tion of the complexes with calf thymus DNA. Com-plexes bound to DNA through intercalation, whichinvolves a strong stacking interaction of the planar aro-matic rings of the coordinated ligand with the base pairsof DNA, usually result in hypochromism and redshift ofligand-band or charge transfer bands [51]. All the pres-ent complexes exhibit significant hypochromism(r = 15, 15–27%) on the incremental addition of DNAwith varying redshifts (Fig. 9, Table 4). The observedhypochromism is higher than that observed for[Cu(2,9-dmp)2]

2+ (1.3%) [24] and [Cu(pdto)]2+ (14.4%,

Fig. 9. Absorption spectra of 2 (25 lM) in 5 mM Tris–HCl/50 mMNaCl buffer (pH 7.1) in the absence (a) and in the presence of (b) CT-DNA (250 lM).

Table 4Change in spectral features of the copper(II) complexes on interactionwith CT-DNA (r = 15) in 5 mM Tris–HCl/50 mM NaCl buffer (pH7.1), r = [base-pair]/[ Cu(II) complex]

Complex Change in absorptivity kmax (nm) Dkmax De (%)

1 Hypochromism 277 3 182 Hypochromism 277 2 153 Hypochromism 277 2 204 Hypochromism 276 2 275 Hypochromism 272 2 22

T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219 1213

pdto = bis(pyrid-2-yl)-3,6-dithiaoctane) [52] but similarto that observed for [Cu(bcp)2]

2+ (20%, bcp = 2,9-di-methyl-4,7-diphenyl-1,10-phenanthroline) [24], whichinvolves its phenyl rings in partial intercalation betweenthe DNA base pairs and [Cu(L)Cl2] (25%, L = N,N-di-methyl-N 0-(pyrid-2-ylmethyl)ethylene-diamine), whichhas been proposed to form covalent bond with theDNA base pairs [53]. The lower hypochromism ob-served for the present complexes imply that they donot intercalate very strongly or deeply between theDNA base pairs. Further, as the extent of hypochro-mism is commonly consistent with the strength of inter-calative interaction, it is evident that all the complexesexhibit almost the same DNA binding affinities [54].All these observations reveal that the 2,9-disubstitutedphen ring of the present complexes intercalate lessstrongly. Moreover, the uncoordinated –NH– group ofthe aliphatic amine side chain may be involved in sec-ondary interactions like hydrogen bonding with DNApossessing several hydrogen bonding sites accessibleboth in the minor and major grooves [55]. Similarhydrogen bonding interactions have been proposedfor [Co(NH3)6]

3+ bound to d(CG)3 [56] and for[Ru(NH3)4(diimine)]2+ bound to CT-DNA [57].

Several cupric complexes are known to form coordi-nate bond to DNA bases [53]. In order to explore the

possibility of the present complexes forming a coordi-nate bond with DNA through the displacement of theequatorially coordinated chloride/water molecule, theLF band of the complexes has been titrated with poten-tial small ligands like N-methylimidazole (Meim), thenucleoside cytidine (Cytd) and the nucleotides guano-sine-5 0-monophosphate (GMP) and adenosine-5 0-mono-phosphate (AMP). On the incremental addition ofMeim, the visible bands of all the complexes are signif-icantly blue-shifted (19–43 nm) with increase in absorp-tivity (0–25 M�1 cm�1) suggesting the coordination ofMeim to copper(II). A similar but lower blueshift (2–23 nm) is observed on the incremental addition of Cytdand GMP indicating weaker coordination of N3 atomof Cytd and N7 atom of GMP to copper(II). Interest-ingly, the spectral changes observed for AMP are muchless, which is consistent with the basicity of N7 atom ofAMP, which is lower than that of GMP (Table 5). Thus,these LF spectral studies indicate that the present cop-per(II) complexes exhibit the potential to form a cova-lent bond with N7 of guanine rather than adenineresidues of DNA. However, the DNA melting studies(data not shown) of the copper(II) complexes do not ex-hibit any significant deviation from the melting temper-ature of free DNA revealing that the complexes do notform covalent bonds on interaction with DNA in viewof the steric hindrance from the uncoordinated sidechain –NH–alkyl groups. So we propose that the lowerredshifts observed in the UV spectra are due to the par-tial intercalation of the phen ring, as also suggested byWang et al. [25], followed by the hydrogen bonding ofthe amino-NH group to base nitrogen atom or phos-phate oxygen atom available on the exterior of DNA.Further, this DNA binding model illustrates that the sig-nificant blueshifts (9–72 nm) and hyperchromism (4–66M�1 cm�1) obtained for all the complexes on bindingto CT DNA are due to a change in geometry of CuN3O2

coordination spheres towards more square-based ones,rather than to covalent bond formation. Interestingly,the [Cu(L5)]2+ complex with sterically hindering N-tert-butyl group shows the highest blueshift revealingits enhanced tendency to form a square-based geometryon binding to DNA. This is consistent with the DNA fi-ber EPR spectral results.

3.5. Competitive binding studies

As the copper(II) complexes are non-emissive, com-petitive binding studies with EthBr were carried out togain support for the mode of binding of the complexeswith DNA. The study involves addition of the com-plexes to DNA pretreated with EthBr ([DNA]/[EthBr] = 1) and then measurement of intensity of emis-sion. The observed enhancement in emission intensity ofEthBr bound to DNA is due to intercalation of thefluorophore in between the base pairs of DNA and

Table 5Effect of adding excess CT-DNA, N-methylimidazole (Meim), guanosine-5 0-monophosphate (GMP), adenosine-5 0-monophosphate (AMP) andcytidine (Cytd)a on the LF absorption band

Complex kmax

(nm)e(M�1 cm�1)

CT-DNAb Meimc GMPc AMPc Cytd

Dkmax

(nm)De(%)d

Dkmax

(nm)De(%)d

Dkmax

(nm)De(%)d

Dkmax

(nm)De(%)d

Dkmax

(nm)De(%)d

1 700 97 9 28 43 8 3 13 2 9 6 42 710 98 10 16 19 2 23 20 – 11 4 23 722 89 18 4 20 5 2 18 4 13 10 44 706 96 19 27 28 25 7 77 4 37 7 175 741 99 72e 66 32 – 23 – 9 – 20 68

a In 5 mM Tris–HCl/50 mM NaCl buffer (pH 7.1).b [Cu] = 2 mM; [DNA] = 7.5 mM.c [Cu] = 5 mM; [A]:[Cu] = 3:1 where A is Meim/GMP/AMP/Cytd.d Hypochromism.e [Cu] = 5 mM; [DNA] = 7.5 mM.

1214 T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219

stabilisation of its excited state (Fig. 10). [58]. Additionof all the complexes to CT-DNA incubated with EthBrdecreases the DNA induced enhancement in emission tothe same extent. This suggests that the complexes dis-place DNA-bound EthBr and bind to DNA at the inter-calation sites with almost the same affinity, which isconsistent with the above spectral results suggesting par-tial intercalation of the phen ring.

3.6. Viscosity studies

To explore further the binding mode of the presentcopper(II) complexes viscosity measurements on solu-tions of calf thymus DNA incubated with the complexeswere carried out. Since the relative specific viscosity (g/g0), (g and g0 are the specific viscosities of DNA in thepresence and absence of the complexes, respectively) ofDNA reflects the increase in contour length associatedwith separation of DNA base pairs caused by intercala-tion, a classical intercalator such as EthBr could cause a

Fig. 10. Emission spectra of ethidium bromide in 5 mM Tris–HCl/50mM NaCl buffer (pH 7.1) in the absence (a), in the presence of CT-DNA (b) and CT-DNA + 1 (c).

significant increase in viscosity of DNA solutions. Incontrast, a partial and/or non-classical intercalation ofthe ligand could bend or kink DNA resulting in a de-crease in its effective length with a concomitant increasein its viscosity [59,60]. The plots of relative specific vis-cosities versus 1/R (=[Cu]/[DNA]) are shown in Fig.11. All the present complexes, except 5, show a slight in-crease in viscosity of DNA on increasing the concentra-tion of the complexes. However, the increase in theviscosity was much less compared to that of potentialintercalators like ethidium bromide in the same DNAconcentration range [61]. This observation leads us tosupport the above spectral studies which suggest thatthe complexes interact with DNA via partial intercala-tion between DNA base pairs, which is similar to the

Fig. 11. The effect of 1 (m), 2 (n), 3 (,), 4 (s), and 5 (+) on therelative viscosity of calf thymus DNA in 5 mM Tris–HCl/50 mMNaClbuffer (pH 7.1).

T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219 1215

interaction of [Cu(phen)2]2+ with DNA [23,24]. Interest-

ingly, 5 shows no apparent increase in viscosity of DNA;obviously, the sterically hindering tert-butyl groups inthe complex appear to prevent the partial intercalationof the phen ring to DNA base pairs and hence any in-crease in viscosity. These results suggest the possibilitythat 5 would bind to DNA by external contact (surfacebinding), which is in line with the DNA fiber EPR spec-tral study.

3.7. Redox studies

An electrochemical method has been extremely usefulin probing the nature and mode of DNA binding of me-tal complexes. Cyclic and differential pulse voltammetrictechniques have been employed to study the interactionof the present redox active copper(II) complexes withDNA with a view to further exploring the DNA bindingmodes assessed from the above EPR and viscometricstudies. The cyclic voltammograms of the complexes inthe absence of DNA reveal a non-Nernstian but a fairlyreversible/quasi-reversible one electron redox processinvolving the Cu(II)/Cu(I) couple, as judged from thepeak potential separation of 96–184 mV (59 mV for aone electron transfer process) (Table 6). The linearityof the ipc versus square root of scan rate plot (passingthrough the origin) reveal that the one electron redoxprocess is diffusion controlled. The Cu(II)/Cu(I) redoxpotentials of the present complexes (from �0.107 to�0.281 V) display significant variations depending uponthe bulkiness of the N-alkyl substituents of the ligand:5 > 4 @ 2 > 3 > 1. Thus, the complex 5 possesses thehighest value of Cu(II)/Cu(I) redox potential obviouslydue to the most bulky tert-butyl group on the coordi-nated secondary amine nitrogen atom, which wouldtend to stabilise the Cu(I) oxidation state. On the other

Table 6Electrochemical behavioura of the Copper(II) Complexes of phenanthrolicomplex]

r Epc (V) Epa (V) E1

1 0 �0.318 �0.222 �20 �0.378 �

2 0 �0.256 �0.106 �20 �0.332 �

3 0 �0.200 �0.064 �20 �0.282 �

4 0 �0.192 0.006 �20 �0.250 �

5 0 �0.164 0.006 �20 �0.292 �

a Measured vs. Ag/AgCl electrode; scan rate: 50 mV s�1; supporting electroM.b Differential pulse voltammetry (DPV), scan rate: 1 mV s�1, pulse heightc Based on E1/2 values from DPV measurements.d E1/2 values were determined from the DPV peak potential, Ep, using the r

Epc and Epa in CV experiments and DE is the pulse amplitude (50 mV).

hand, the less bulky but more electron-releasing methylgroup stabilises the higher Cu(II) oxidation state.

All the complexes in the presence of excess DNA(r = 20) display irreversible electrochemical responsesas shown in Fig. 12. The significant reduction in catho-dic peak current on the addition of DNA is due to slowdiffusion of an equilibrium mixture of the free andDNA-bound complexes to the electrode surface. Fur-ther, the observed shifts (26–40 mV) in E1/2 values(DPV) to more negative potentials suggest that bothCu(II) and Cu(I) forms of the present complexes bindto DNA but with Cu(II) displaying higher DNA bind-ing affinity than Cu(I) form. This is illustrated by theratio of the equilibrium constants (K+/K2+) for thebinding of Cu(I) and Cu(II) species to DNA. These val-ues (Table 6) have been estimated from the net shift[62] in E1/2 (DPV) on the addition of DNA assumingreversible electron transfer using the followingequation:

E00b � E00

f ¼ 0:0591 logðKþ=K2þÞ;where E00

b and E00f are the formal potentials of the

Cu(II)/Cu(I) couple in the free and bound forms, respec-tively, and K+ and K2+ the corresponding binding con-stants for the binding of the 1+ and 2+ species toDNA, respectively. The K+/K2+ values of all the com-plexes are less than unity suggesting preferential stabil-ization of Cu(II) form over Cu(I) form on binding toDNA. This is interesting because [Cu(phen)2]

2+ has beenfound to be stabilized preferentially in the Cu(I) form onbinding to DNA (K+/K2+, 1.5). This illustrates that teth-ering the coordinating R–NH–CH2-moiety onto 2-/9-position on phen ring changes the DNA preference fromCu(I) to Cu(II) form. Further, the complexes 1 and 4 areslightly better than others in stabilising Cu(II) overCu(I) form.

ne derivatives on interaction with CT-DNA, r = [base-pair]/[Cu(II)

/2d (DPVb, V) DE1/2 (DPV b, mV) K+/K2+

c

0.2810.307 26 0.40.1650.211 36 0.20.1730.209 36 0.20.1610.187 26 0.40.1070.147 40 0.2

lyte 5 mM Tris–HCl/50 mM NaCl; complex concentration: 2.5 · 10�4

50 mV.

elation E1/2 = Ep + DE/2 where E1/2 is the equivalent of the average of

Fig. 12. Cyclic voltammogram of 2 (0.25 mM) (a) in the absence of and (b) in the presence of CT-DNA (5 mM). Scan rate: 50 mV s�1. Supportingelectrolyte, 5 mM Tris–HCl/50 mM NaCl buffer (pH 7.1).

Fig. 13. Cleavage of pBR 322 DNA by complexes in the presence ofH2O2 for different incubation times (a) 15 min, (b) 30 min at 37 �C.Lane 1 untreated DNA; lane 2 without complex; lane 3 without H2O2;lane 4 CuCl2; lane 5 [Cu(phen)2]; lane 6 complex 1; lane 7 complex 5;lane 8 complex 2; lane 9 complex 3; lane 10 complex 4.

Fig. 14. Cleavage of /X 174 RF I DNA by (a) 1 and (b) 5 in thepresence of H2O2 with different oxidation inhibitors (for 20 min at 37�C). Lane 1 untreated DNA; lane 2 without inhibitor; lane 3 SOD (3unit); lane 4 DMSO (100 mM); lane 5 NaN3 (20 mM); lane 6 catalase(3 unit).

1216 T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219

3.8. DNA cleavage studies

Fig. 13 shows the agarose gel electrophoresis patternof oxidative cleavage of pBR322 plasmid DNA byCu(II) complexes in the presence of H2O2. DNA cleav-age activity of the complexes was assayed from the con-version of supercoiled DNA (Form I) to nicked circularDNA (Form II) or linearized DNA (Form III). For 15min of reaction time, the ability of the complexes to-wards formation of Form II varies as [Cu(phen)2]

2+ >5 > 3 > 4 > 1 > 2, and for 30 min of reaction time, theformation of small amounts of linear fragments (FormIII) were observed for [Cu(phen)2]

2+ and 5 (lanes 5and 7). The oxidative DNA cleavage reactions withCu(II) complexes in the presence of H2O2 and/or reduc-ing reagents are known to proceed via redox cycles be-tween Cu(I) and Cu(II) [45,63,64]. Therefore, the

redox potential is one of the important factors that re-flect the ability of the complexes to cleave DNA in theoxidative process. The cleavage efficiency of 5 beinghigher than complexes 1–4 seems to reflect its highCu(II)/Cu(I) redox potential conferred by the bulky N-tert-butyl group. This is consistent with the interestingobservation that 5 has a coordination structure onDNA similar to that of [Cu(phen)2]

2+ as revealed byDNA fiber EPR. Enhancement in hydrophobicity con-ferred by the tert-butyl groups, which would increasethe mobility of the reactive Cu(II) species on DNA,would also contribute to the higher cleavage activity[57].

It has been known that certain copper(II) complexescleave DNA with H2O2 via Fenton-like mechanisms[65]. In the case of the bis(1,10-phenanthroline), a freelydiffusible hydroxyl radical or non-diffusible copper-oxospecies is assumed to be the active species [5,66–68].To understand the active oxygen species in the reactions,inhibition experiments were carried out with standardscavengers for reactive oxygen intermediates, such asSOD, DMSO, NaN3 and catalase. As Fig. 14 shows,

T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219 1217

all the reactions are inhibited effectively by catalaserevealing the critical role of H2O2 in the reactions.The reagents SOD, DMSO, and NaN3 inhibit thecleavage reaction with 1 to the same extent. The reac-tion with 5 was the most inhibited by DMSO but SODand NaN3 fail to inhibit the cleavage under the presentexperimental condition. These results suggest that thecleavage reaction with 1 involves various oxygen spe-cies whereas that with 5 progresses mainly with hydro-xyl radical. A spin-trapping experiment with DMPOdemonstrates the formation of spin adduct with hydro-xyl radical in the reaction of 5 involving H2O2 (Fig.S4). It should be noted that the reaction with [Cu(phen)2]

2+ and H2O2 in the absence of a reductantsuch as ascorbic acid is remarkably inhibited by SOD(Fig. S5). This suggests that 5 and [Cu(phen)2]

2+ differin the reaction pathway, though both the complexesbind to DNA with similar coordination structure andorientation.

4. Conclusion

The present paper reports a series of 1:1 copper(II)complexes of certain N,N 0-dialkyl-1,10-phenanthroline-2,9-dimethanamine ligands with N-alkyl substituentsvarying from methyl to tert-butyl. The crystal structureof the copper(II) complex of monoprotonated N-propylsubstituted ligand reveals that the tetradentate 4N li-gand actually acts as a tridentate 3N ligand and coordi-nates meridionally to form a square pyramidal CuN3Cl2chromophore. On the basis of electronic and EPR spec-tral studies, the coordination geometries of other com-plexes are suggested to involve a similar CuN3Cl2/CuN3(H2O)2 chromophore with geometries varyingfrom square based to trigonal bipyramidal. The trigonalbipyramidal geometry of the N-tert-butyl complex is un-like the well known [Cu(phen)2Cl] complex with square-based geometry.

The electronic, fiber EPR and emission spectral, vis-cometric and electrochemical studies provide strong evi-dences for the involvement of the complexes in DNAinteraction with partial intercalation of the derivatisedphen ring in between the DNA base pairs. Further,the sterically hindering and non-coordinating N-alkylamine side chain on the phen ring prevents the com-plexes from being involved in covalent binding, whichwould involve the replacement of the coordinated Cl�/H2O. The DNA fiber EPR spectra of the complexes sug-gest that the CuN3X2 ðX ¼ NO�

3 =Cl�=H2OÞ chromoph-

ores undergo changes in geometry from trigonalbipyramidal in solution toward square-pyramidal onDNA. Interestingly, the N-tert-butyl complex, despiteits amount bound to DNA being the least among thepresent complexes, exhibits the highest nuclease activity.Apart from its DNA binding structure, its relatively

high Cu(II)/Cu(I) redox potential conferred upon bythe bulky N-tert-butyl group appears to regulate thecleavage of DNA at specific sites on the DNA polymer.In contrast to this complex, the square pyramidal struc-ture of [Cu(phen)2]

2+ in solution gets deformed on DNAtoward trigonal bipyramidal, which is further modu-lated with change in conformation of DNA and temper-ature. One can reasonably suppose that this structuralflexibility of both [Cu(phen)2]

2+ and the N-tert-butylcomplex on DNA during the CuII/CuI redox cycle playsa very important role in the activation of oxygen species.It is remarkable that the N-tert-butyl complex, in con-trast to [Cu(phen)2]

2+, which binds to DNA in a similarbinding geometry and orientation, differs in the mecha-nism of generating the reactive oxygen species due to aslight change in coordination environment of the com-plexes on DNA.

5. Abbreviations

[Cu(phen)2]2+ bis(1,10-phenanthroline)copper(II)

complex

DMPO 5,5-dimethyl-1-pyrroline N-oxide DMSO dimethyl sulfoxide EDTA ethylenediamine-N,N,N 0,N 0-tetraacetate EPR electron paramagnetic resonance Hepes N-[2-hydroxyethyl]piperazine-N 0-

[2-ethanesulfonic acid]

SOD superoxide dismutase TAE tris acetate EDTA Tris tris(hydroxylmethyl)aminomethane

Acknowledgements

This work was supported by Grant-in-aid for ScienceResearch, No. 14540541 for M.C. from the Japan Min-istry of Education, Science, Sports, and Culture. M.C.also acknowledges the aid from Chuo University Grantfor Special Research and a Grant from the ResearchInstitute of Science and Engineering Chuo University.M.P. acknowledges the support from JSPS InvitationFellowship Program for Research in Japan (ShortTerm).

Appendix A. Supplementary data

EPR spectra of Cu complexes used in this experimentat pH 7.4 (Fig. S1), EPR spectra of 1 on DNA pellet andon A-form DNA fiber (Fig. S2), EPR spectra of 5 onDNA pellet and on A-form DNA fiber (Fig. S3), EPRspectrum of the DMPO spin adduct in the presence of

1218 T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219

5 and H2O2 (Fig. S4), effect of SOD on the cleavage ofDNA with [Cu(phen)2]

2+ (Fig. S5). Supplementary dataassociated with this article can be found, in the onlineversion, at doi:10.1016/j.jinorgbio.2005.02.020.

References

[1] S.J. Lippard, Acc. Chem. Res. 11 (1978) 211–217.[2] J.K. Barton, Science 223 (1986) 727–734.[3] P.B. Dervan, Science 232 (1986) 464–471.[4] M.J. Clarke, F. Zhu, D.R. Frasca, Chem. Rev. 99 (1999) 2511–

2533.[5] D.S. Sigman, A. Mazumder, D.M. Perrin, Chem. Rev. 93 (1993)

2295–2316.[6] S. Routier, N. Cotelle, J.-P. Catteau, J.-L. Bernier, M.J. Waring,

J.-F. Riou, C. Bailly, Bioorg. Med. Chem. 4 (1996) 1185–1196.

[7] D.J. Gravert, J.H. Griffin, Inorg. Chem. 35 (1996) 4837–4847.[8] C.J. Burrows, J.G. Muller, Chem. Rev. 98 (1998) 1109–1151.[9] P.G. Sammes, G. Yahioglu, Chem. Rev. (1994) 327–334.[10] K.J. Humphreys, A.E. Johnson, K.D. Karlin, S.E. Rokita,

J. Biol. Inorg. Chem. 7 (2002) 835–842.[11] D.K. Chand, H.-J. Schneider, A. Bencini, A. Bianchi, C. Giorgi,

S. Ciattini, B. Valtancoli, Chem. Eur. J. 6 (2000) 4001–4008.

[12] P.S. Pendergrast, Y.W. Ebright, R.H. Ebright, Science 265 (1994)959–962.

[13] G.W. Muth, C.M. Thompson, W.E. Hill, Nucleic Acids Res. 27(1999) 1906–1911.

[14] M. Hirai, K. Shinozuka, H. Sawai, S. Ogawa, Bull. Chem. Soc.Jpn. 67 (1994) 1147–1155.

[15] J. Gallagher, C.-H.B. Chen, C.Q. Pan, D.M. Perrin, Y.-M. Cho,D.S. Sigman, Bioconjugate Chem. 7 (1996) 413–420.

[16] M. Pitie, C. Boldron, H. Gornitzka, C. Hemmert, B. Donnadieu,B. Meunier, Eur. J. Inorg. Chem. (2003) 528–540.

[17] J.K. Barton, J.M. Goldberg, C.V. Kumar, N.J. Turro, J. Am.Chem. Soc. 108 (1986) 2081–2088.

[18] M. Chikira, T. Sato, W.E. Antholine, D.H. Peterring, J. Biol.Chem. 266 (1991) 2859–2863.

[19] M. Chikira, T. Iiyama, K. Sakamoto, W.E. Antholine, D.H.Petering, Inorg. Chem. 39 (2000) 1779–1786.

[20] R. Nagane, T. Koshigoe, M. Chikira, E.C. Long, J. Inorg.Biochem. 83 (2000) 17–23.

[21] M. Chikira, Y. Tomizawa, D. Fukita, T. Sugizaki, N. Sugawara,T. Yamazaki, A. Sasano, H. Shindo, M. Palaniandavar, W.E.Antholine, J. Inorg. Biochem. 89 (2002) 163–173.

[22] T. Hirohama, A. Chiba, M. Chikira, Y. Fujii, J. Inorg. Biochem.98 (2004) 371–376.

[23] S. Mahadevan, M. Palaniandavar, Inorg. Chem. 37 (1998) 3927–3934.

[24] S. Mahadevan, M. Palaniandavar, Inorg. Chem. 37 (1998) 693–700.

[25] Z.-M. Wang, H.-K. Lin, Z.-F. Zhou, M. Xu, T.-F. Liu, S.-R.Zhu, Y.-T. Chen, Bioorg. Med. Chem. 9 (2001) 2849–2855.

[26] J.M. Veal, R.L. Rill, Biochemistry 28 (1989) 3243–3250.[27] J.C. Stockert, J. Theor. Biol. 137 (1989) 107–111.[28] J.M. Veal, R.L. Rill, Biochemistry 30 (1991) 1132–1140.[29] G. Zhao, H. Sun, H. Lin, S. Zhu, X. Su, Y. Chen, J. Inorg.

Biochem. 72 (1998) 173–177.[30] M.E. Reichmann, S.A. Rice, C.A. Thomas, P. Doty, J. Am.

Chem. Soc. 76 (1954) 3047–3053.[31] A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, M.

Burla, G. Polidori, M. Camalli, SIR-92, J. Appl. Crystallogr. 27(1994) 435.

[32] P.T. Beurskens, G. Admiraal, G. Beurskens, W.P. Bosman, R. deGelder, R. Israel, J.M.M. Smits, DIRDIF-99: The DIRDIF-99Program System, Technical Report of the Crystallography Lab-oratory, University of Nijmegen, The Netherlands, 1999.

[33] D.T. Cromer, J.T. WaberInternational Tables for X-ray Crystal-lography, vol. IV, The Kynoch Press, Birmingham, 1974, Table2.2A.

[34] CrystalStructure 3.5.1: Crystal Structure Analysis Package, Rig-aku and Rigaku/MSC (2000–2003), 9009 New Trails Dr. TheWoodlands, TX 77381, USA.

[35] D.J. Watkin, C.K. Prout, J.R. Carruthers, P.W. Betteridge,CRYSTALS Issue 10: Chemical Crystallography Laboratory, Oxford,UK, 1996.

[36] A.W. Addison, T.N. Rao, J. Reedijk, J. Rijin, G.C. Verschoor,J. Chem. Soc., Dalton Trans. (1984) 1349–1356.

[37] U. Sakaguchi, A.W. Addison, J. Chem. Soc., Dalton Trans.(1979) 600–608.

[38] A. Angeloff, J.-C. Daran, J. Bernadou, B. Meunier, Eur. J. Inorg.Chem. (2000) 1985–1996.

[39] H. Nakai, Y. Noda, Bull. Chem. Soc. Jpn. 51 (1978) 1386–1390.

[40] D. Boys, C. Escober, Acta Crystallogr., Sect. B 37 (1981) 351–355.

[41] P.C. Healy, L.M. Engelhardt, V.A. Patrick, A.H. White, J. Chem.Soc., Dalton Trans. (1985) 2541–2545.

[42] D.S. Sigman, A. Spassky, S. Rimsky, H. Buc, Biopolymers 24(1985) 183–197.

[43] H.R. Drew, J. Mol. Biol. 176 (1984) 535–557.[44] L.E. Pope, D.S. Sigman, Proc. Natl. Acad. Sci. USA 81 (1984)

3–7.[45] M.D. Kuwabara, C. Yoon, T. Goyne, T.B. Thederahn, D.S.

Sigman, Biochemistry 25 (1986) 7401–7408.[46] J.M. Veal, R.L. Rill, Biochemistry 27 (1988) 1822–1827.[47] F. Schaeffer, S. Rimsky, A. Spassky, J. Mol. Biol. 260 (1996) 523–

539.[48] W. Saenger, Principles of Nucleic Acid Structure, Springer, New

York, 1983, p. 253.[49] G.S. Manning, Q. Rev. Biophys. 11 (1978) 179–246.[50] N. Grover, T.W. Welch, T.A. Fairley, M. Cory, H.H. Thorp,

Inorg. Chem. 33 (1994) 3544–3548.[51] V.A. Bloomfield, D.M. Crothers, I. Tinocco Jr., Physical

Chemistry of Nucleis Acids, Harper & Row, New York, 1974,p. 432.

[52] S. Mahadevan, M. Palaniandavar, Inorg. Chim. Acta 254 (1997)291–302.

[53] A. Raja, V. Rajendiran, P.U. Maheswari, R. Balamurugan, C.A.Kilner, M.A. Halcrow, M. Palaniandavar J. Inorg. Biochem., JIB04-0609, under revision.

[54] S.A. Tysoe, R.J. Morgan, A.D. Baker, T.C. Strekas, J. Phys.Chem. 97 (1993) 1707–1711.

[55] H.L. Chan, Q.L. Liu, B.C. Tzeng, Y.S. You, S.M. Peng, M.Yang, C.M. Che, Inorg. Chem. 41 (2002) 3161–3171.

[56] R.V. Gessner, G.J. Quigley, A.H.J. Wang, G.A. Van der Marel,J.H. Van Boom, A. Rich, Biochemistry 24 (1985) 237–240.

[57] P.U. Maheswari, M. Palaniandavar, J. Inorg. Biochem. 98 (2004)219–230.

[58] J.-B. LePecq, C. Paoletti, J. Mol. Biol. 27 (1967) 87–106.

[59] S. Satyanarayana, J.C. Dabrowiak, J.B. Chaires, Biochemistry 31(1992) 9319–9324.

[60] E.J. Gabbay, R.E. Scofield, C.S. Baxter, J. Am. Chem. Soc. 95(1973) 7850–7857.

[61] I. Haq, P. Lincoln, D. Suh, B. Norden, B.Z. Chowdhry, J.B.Chaires, J. Am. Chem. Soc. 117 (1995) 4788–4796.

[62] A.J. Bard, L.R. Faulkner, Electrochemical Methods, Wiley, NewYork, 1980, p. 194.

T. Hirohama et al. / Journal of Inorganic Biochemistry 99 (2005) 1205–1219 1219

[63] S. Goldstein, G. Czapski, H. Cohen, D. Meyerstein, Inorg. Chem.27 (1988) 4130–4135.

[64] T.B. Thederahn, M.D. Kuwabara, T.A. Larsen, D.S. Sigman,J. Am. Chem. Soc. 111 (1989) 4941–4946.

[65] S. Goldstein, G. Czapski, J. Am. Chem. Soc. 108 (1986) 2244–2250.

[66] D.S. Sigman, D.R. Graham, V. D�Aurora, A.M. Stern, J. Biol.Chem. 254 (1979) 12269–12272.

[67] D.S. Sigman, Acc. Chem. Res. 19 (1986) 180–186.[68] D.R. Graham, L.E. Marshall, K.A. Reich, D.S. Sigman, J. Am.

Chem. Soc. 102 (1980) 5419–5421.