Coordination behavior of sulfathiazole. Crystal structure of dichloro-disulfathiazole ethanol Cu(II) complex. Superoxide dismutase activity

ELSEVIER

Coordination Behavior of Sulfathiazole. Crystal Structure of [Cukulfathiazole) (py) 3Cl] Superoxide Dismutase Activity J. Casanova, G. Alzuet, J. Borrk, J. Latorre, M. Sanati, and S. Garcia-Granda JC, GA, JB, JL, MS. Departamento de Q&mica Inotghica, Facultad de Farmacia, Universidad de Valencia, Valencia, Spain.-SG. Departamento de Q&mica Fisica y Anal&a, Universidad de Oviedo, Oviedo, Spain

ABSTRACT The preparation, spectroscopic, magnetic properties, and crystal structure of [Cu(stzXpy),Cl] (stzstands for the deprotonated form of sulfathiazole, 4-amino-N-2-thiazolylbencenosulfonamide) are reported. Crystals are orthorhombic, space group Pbca, with cell constants a = 15.834(21, b = 17.512(4), and c = 18.79(2) A, and Z = 8. The structure was solved and refined to R = 0.041 (R, = 0.040). The structure consists of mononuclear units linked via hydrogen bonds to form the tridimensional pyramid. The geometry of CuN,N*NCl chromophore is distorted square-pyramid. The superoxide-dismutase mimetic activity of the compound is measured and compared with those of the SOD enzyme, the free drug, and other related sulfathiazole complexes.

INTRODUCTION Reactive oxygen radicals have been postulated as playing an important role in a wide variety of pathological processes [l-4]. In conditions such as &hernia, artheriosclerosis, irradiation, and inflammation, the dismutation of superoxide anion radicals is important in order to avoid serious damage on living organisms. One of the enzymes involved in the endogenous protection against oxygen radical toxicity is superoxide dismutase (SOD). It has often been reported that

Address correspondence to: Prof. J. Borrb, Departamento de Quimica Inorganica, Farmacia, Universidad de Valencia, 46100-Bujassot, Valencia, Spain. Journal of Inorganic Biochemistry, 0 1995 Elsevier Science Inc.,

60, 219-230

655 Avenue of the Americas, NY, NY 10010

Facultad de

(1995)

0162-0134/95/$9.50 SSDI 0162-0134(95)00022-G

220

J. Casanova et al.

copper complexes can exhibit SOD activity [5-81. Furthermore, some patents of Cu(I1) compounds with antioxidant activity have been registered 19-111. A possible mechanism of action of these complexes involves the scavenging of superoxide anions [12]. A limited steric hindrance to the approach of the superoxide anion is considered an essential requirement for the successful binding of the superoxide radical [13]. In previous papers, we have tested the superoxide dismutase-like activity of the sulfathiazole copper complexes [14]. We have observed that there is a relation between this activity and the degree of geometrical distortion of the coordination polyhedron. Effectively, the Cu(Hstz),CH,(OH)Cl, presents double superoxide dismutase activity compared to the Cu(Hstz),CH,CH,(OH)Cl, as a consequence of its higher geometrical distortion. We report here the synthesis, spectroscopic studies, and superoxide dismutase activity of the [Cu(stzXpy),Cl] complex.

EXPERIMENTAL Materials Copper chloride dihydrate and sulfathiazole were without purification. Analytical data (C,H,N) were model 2400 of the Departamento de Quimica Salamanca. The copper content was determined troscopy.

of reagent grade and used obtained in a Perkin Elmer Inorganica, Universidad de by atomic absorption spec-

Synthesis CuCl,H,O [0.17 g, 1 mmol] was added to a solution of sulfathiazole [0.5 g, 2 mmoll in methanol [lo0 mL]. Then, to the resulting yellowish mixture, 0.8 mL ( = 10 mmol) of pyridine was also added. Following the addition of pyridine, the color of the solution turned into green-blue, and immediately, a blue precipitate was obtained. The solid was filtered out and the filtrate was left standing at room temperature. In a few days, prismatic green-blue crystals were obtained from this filtrate. Physical Techniques IR spectra were recorded on a Perkin-Elmer 843 instrument. Samples were prepared by using the KBr technique. Vis-UV spectra were recorded with a Perkin-Elmer Lambda 15 spectrophotometer. EPR spectra of polycrystalline samples were recorded at X-band frequencies with a Bruker ER 200D spectrometer equipped with a liquid nitrogen continuous flow cryostat. Magnetic susceptibility measurements were carried out in the 80-300 K temperature range with a fully automatized AZTEC DSM8 pendulum-type susceptometer. Mercury tetrakis(thyocianato)cobaltate(II) was used as a susceptibility standard. Corrections for the diamagnetism of the complex were estimated from Pascal’s constants. Electrochemical

Measurements

The electrochemical experiments were carried out in a three-electrode cell; the working and auxiliary electrode were platinum, and the reference electrode was

SUPEROXIDE

DISMUTASE

ACTIVITY

OF Cu(SULFATHIAZOLE1

221

a saturated calomel electrode (SCE) electrically connected to the solution via a “salt bridge” containing a saturated solution of supporting electrolyte and the solvent. Cyclic voltammograms were obtained with a PAR potentiostat 273A and recorded with a Riken-Denshi F-35 x-y recorder. Crystallographic

Data Collection

A blue-green crystal, size 0.15 x 0.20 X 0.1 mm. Throughout the experiment, MoKa radiation was used with a graphite crystal monochromator on an Enraf-Nonius CAD-4 single-crystal diffractometer (A = 0.71071 A). The unit-cell dimensions were determined from the angular setting of 25 reflections with 13 between 15” and 25”. The space group was determined to be Pbca from the structure determination. The intensity data of 4566 reflections, in hkl range (O,O,O)-(18,20,22) and 8 limits (1 < 0 < 251, were measured using the w - 28 scan technique and a variable scan rate with a maximum scan time of 60 set per reflection. The intensity of the primary beam was checked throughout the data collection by monitoring three standard reflections every 60 min. The final drift correction factors were between 0.98 and 1.03. On all reflections, profile analysis was performed [15, 161. Some double measured reflections were averaged, Rint = z(I + (I))/zI = 0.02, resulting in 4566 “unique” reflections of which only 2138 were “observed” with I > 3~ (I). Lorentz and polarization corrections were applied and the data were reduced to IFa] values. The structure was solved by Patterson using the program SHELXS 86 [17] and expanded by DIRDIF 1181. Isotropic least-squares refinement, using SHELX76 [19, 201, converged to R = 0.08. At this stage, additional empirical absorption correction factors were applied using DIFABS [21]. The maximum and minimum absorption correction factors were, respectively, 1.45 and 0.67. Hydrogen atoms were geometrically placed. During the final stages of the refinement, the positional parameters and the anisotropic thermal parameters of the non-H atoms were refined. The hydrogen atoms were isotropically refined with a common thermal parameter. The final conventional agreement factors were R = 0.041 and wR = 0.040 for the 2138 “observed” reflections and 384 variables. The function minimized was Cw(F,, F,>*, w = l.O/[ cr2(F,,) + 0.0003 F,2] with a(F,,) from counting statistics. The maximum shift to e.s.d. ratio in the last full-matrix least-squares cycle was 0.013. The final difference Fourier map showed no peaks higher than 0.5 e.Am3 not deeper than -0.42 e.Ae3. At omit scattering factors were taken from the International Tables forX-Ray Crystallography [22]. Geometrical calculations were made with PARTS [23]. The crystallographic plots were made with EUCLID 1241. All calculations were made at the MicroWAXat the Scientific Computer Center of the University of Oviedo. Crystallographic data are given in Table 1. Superoxide Dismutase Assay Superoxide dismutase activity of the metal complex was determined according to Oberley and Spitz [25] with minor modifications. Xanthine (1.5 X 10m4 M) and xanthine oxidase in 50 mM potassium phosphate buffer, pH 7.8, were used to generate a reproducible and constant flux of superoxide anions. An amount of xanthine oxidase giving in the control assays an A56,, rate of O.OS/min was used. Reduction of NBT (5.6 x lop5 M) was used as an indicator of superoxide anion

222

J Casanova et al.

TABLE

1. Crystal and Crystallographic Data Collection for the Cu(stz)Py$l Complex

FoMluia Formula weight Crystal system Space group

CuC,,H,,N,S,O,CI 594.63 Orthorombic Pbca

a (A)

15.834(Z)

b (A)

17.512(4)

c (AI V (A’) Crystal Dimensions (mm) Z Dx(Mg.m-‘) m (cm-‘) No. observed reflections No. variables R wR

18.790(2) 6210(5) 0.15 x 0.20 x 0.10 8 1.52 11.30 5079 384 0.041 0.040

production and followed spectrophotometrically at 560 nm. The metal complexes were dissolved in 50 nM Tris-HCI buffer, pH 7.8, and added to the assay in a mixture in a volume representing one-tenth of the total. The percentage inhibition of NBT reduction was used as a measure of SOD activity of the complexes. Xanthine, xanthine oxidase, NBT, and superoxide dismutase (Bovine erytrocyte) were from Sigma Chemical Co. The inhibition of xanthine oxidase by the complexes was calculated measuring uric acid formation from xanthine at 310 nm in the presence of the complex concentrations used. The inhibition percentage of enzyme activity was subtracted from that of the NBT reduction.

RESULTS AND DISCUSSION An ORTEP drawing representation including the atomic numbering scheme is given in Figure 1. Bond distances and angles are listed in Table 2. The structure consists of monomeric units linked by hydrogen bonds. The environment of the copper(B) ion is (4 + 1). Four short bonds of c.a. 2.0 A are formed with the N(l), N(4), and N(5) from the pyridine molecules and the N(2) from the thiazole ring of the deprotonated ligand. The axial position is occupied by the chlorine atom at a longer distance, 2.50 A. The (4 + 1) coordination mode is compatible with two idealized geometries: square pyramidal and trigonal bipyramidal. The structural parameter 7 (7 defined as (a - p)/60, where p is the largest and (Y is the next largest metal ligand bond angles) is applicable to five coordinate structures as an index of the degree of trigonality; for perfect tetragonal-pyramidal and trigonal-bipyramidal geometries, the value of Q- is zero and unity, respectively 1261.For the complex a, 7 of 0.19 indicates that the geometry is distorted square pyramidal. According to this stereochemistry, the equatorial atoms are not coplanar, but pairwise distorted, so that the N(1) and N(5) of the trans pyridine molecules are raise by 0.1167 and 0.0982 A, respec-

SUPEROXIDE

FIGURE 1.

DISMUTASE

ORTEP

ACTIVITY

OF Cu(SULFATHIAZOLE)

223

drawing of the [Cu(stz)Py,Cl] complex.

tively above the mean basal plane, while the N(2) of the thiazole ring ligand and the N(4) of the other pyridine molecule are located 0.0900 and 0.0945 A, respectively, above the plane. As usual, the copper00 ion is raised by 0.2505 8, from the mean plane towards the apical Cl atom [27, 281. The Cu-N, bond distances are in the range found for pyridine complexes [29, 301. The Cu-Nthiazole are comparable to the Zn-Nthiazole in the Zn(stz),H,O complex [31]. It must be pointed out that the behavior of the sulfathiazole is as a monodentate ligand, while in the other crystal structure of the metal-sulfathiazole complex where the stz- is monodeprotonated, the behavior is as bidentate ligand, through the Nthiazoleand Naminoatoms. As a consequence, we can observe three types of behavior of the sulfathiazole: in neutral form, it acts as monoden-

224

J. Casanova et al.

TABLE

2.

Selected Bond Lengths (& and Angles (“) with ESDs in Parentheses 2.500(l) 2.05X4) 1.439(4) 1.751(6) 1.335(7) 1.333(7) 1.336(6) 1.374(8) 1.380(g) 1.373(9) 1.365(9) 1.378(8) 1.371(9) 1.369(g)

cm-CL1 CUl-N4 Sl-01 s2-c20 Nl-Cl5 N3-C20 N5-C51 Cll-Cl2 c14-Cl5 C42-C43 C51-C52 c54-c55 C62-C63 C65-C66 Nl-CUl-CL1 NZCUl-Nl NCCUl-Nl NS-CUl-CL1 N5-CUl-N2 Cll-Nl-CUl C21-N2-CUl C45-N4-CUl C55-N5-CUl

CUl-Nl CUl-N5 Sl-02 s2-c22 N2-C20 N4-C41 N5-C55 C12-Cl3 c21-c22 c43-a4 C52-C53 C61-C62 C63-C64

2.037(4) 2.042(4) 1.443(4) 1.721(6) 1.345(6) 1.338(7) 1.352(6) 1.382(8) 1.340(8) 1.362(9) 1.373(9) 1.387(7) 1.377(9)

NZ-CUl-CL1 NCCUl-CL1 N4-CUl-N2 NS-CUl-Nl N5-CUl-N4 ClS-Nl-CUl C41-N4-CUl C51-N5-CUl

98.60) 91.2(2) 87.4(2) 101.5(l) 89.6(2) 121.3(4) 127.7(4) 121.5(4) 121.5(4)

CUl-N2 Sl-N3 Sl-C61 Nl-Cl1 N2-C21 N4-C45 N6-C64 c13-Cl4 C41-C42 (x-c45 c53-c54 C61-C66 C64-C65

2.008(4) 1.591(4) 1.775(6) 1.328(7) 1.382(7) 1.3447) 1.373(8) 1.348(9) 1.360(8) 1.372(8) 1.363(9) 1.378(8) 1.394(8)

95.m 93.7(l) 171.2(2) 159.7(2) 88.7(2) 121.1(4) 122.1(4) 120.5(5)

tate through the Namino atom [14, 321; when deprotonated, it can act as a monodentate and in bidentate form [31]. Table 3 compares the significant bond distances of the free ligand and in the title complex and in other sulfathiazole complexes. Two remarkable aspects can be deduced from the table: a) There are no significant changes in the thiazole ring distances, in spite of the ligand acting as a neutral form or deprotonated and binding or not binding the metal ion. This is probably due to the delocalization of the electronic density in the ring.

TABLE

3.

Significant Bond Lengths of the Hstz in the Different Complexes

Bond Cu-Nthiazole Cu-Namino Cu-Nm ino Zn-N,

C-Nthiazole

C-Namino

1.342-1.380 1.334-1.405 1.335-1370

1.378 1.447 1.418

1.332-1.385

1.446

1.342-1.357

1.368

ino

Zn-Nthiazale

‘1 = [Cuktzxpy),Cl], 2 = [Cu(Hstz),(EtOH)Cl,], 3 = [Cu(Hstz),(MeOH)Cl,], 4 = [Zn(stz),] 5 = Hstz.

. 2H,O,

SUPEROXIDE

DISMUTASE

ACTIVITY

OF Cu(SULFATHIAZOLE)

b) The C-Nanilino bond length is increased through the Nanilinoatom.

when the metal ion is linked

The mononuclear entities are held together b CI(l)....H-N(6) hydrogen bonds at distances of 3.322 and 3.075 %, respectively. IR Spectroscopic

225

an N(6)-N....O

Data

A comparative study of the IR spectral data of the complex with those of the uncomplexed ligand and the previously reported complexes may give positive information regarding the binding sites of the ligand molecule. The medium bands at 3440 and 3350 cm-‘, corresponding to the v(N-H), are shifted to higher energies with respect to those in the free ligand (3320 and 3280 cm-’ 1. Although similar changes have been assigned to the coordination of the NH, group with the metal ion, in those complexes in which this moiety is involved in the coordination [33], the lack of interaction of this group in the present compound seems to suggest that such changes also arise from the hydrogen bonding and/or packing effects, such as is observed from the decrease of the N-H bond length in the compound compared with the uncoordinated ligand [N-Hav, 0.87 in the ligand; 0.80 in the compound]. Evidence of the Nthiazoleinteraction arises from the red shifts of the band assigned in the free sulfa drug to the thiazole ring vibrations (from 1550 cm-’ in the sulfathiazole to 1500 cm-’ in the compound) together with a marked intensity reduction. As expected, the bands at 1310, 1150, and 560 cm-’ attributed to an SO, remain unchanged with respect to those of the ligand. The characteristic bands of the coordinated pyridine molecules appear at 3080, 1450, 1080, 650, and 440 cm-’ [34, 351. Electronic Spectrum The reflectance spectrum of the title compound shows the d-d transitions as a very broad band centered around 14,810 cm- ‘. This band falls in the range of those usually reported for five-coordinate copper complexes [36] and for the recently published five-coordinate complexes of sulfathiazole. In fact, the whole spectrum closely resembles that found for both sulfathiazole copper complexes, except for the fact that the d-d band is shifted to higher energy in the present compound (14,100 and 13,990 cm-’ Cu-EtOH and Cu-MeOH complexes, respectively). If we compare the T value for the three complexes, 0.19, 0.28, and 0.36, the shift of the d-d band to lower energy is according to the higher distortion of the square pyramidal to bipyramidal trigonal geometry, in agreement with Hathaway [37]. EPR Spectrum and Magnetic Moment The room-temperature polycrystalline powder EPR spectrum is typical of a monomeric square pyramidal copper complex with a dx2-y2 ground state (g]] = 2.22 and g I = 2.07). This axial spectrum confirms the lower distortion of the square pyramidal geometry of the complex with respect to that found in the related sulfathiazole complexes Cu-EtOH and Cu-MeOH which have rhombic EPR spectra. The magnetic moment of the copper complex is 1.87 BM, in agreement with the EPR and the electronic spectra.

226

J. Casanova et al.

Electrochemistry In order to obtain experimental data that can explain the superoxide dismutase activity of the entitled copper complexes, we have tested the electrochemical behavior of these compounds in different solvents by electrochemical techniques like cyclic voltammetry. Complex Cu(stz)py,Cl and the related Cu(Hstz),(MeOH)Cl, and Cu(Hstz),(EtOH)Cl, were studied in DMSO solution (NBu,PF, as supporting electrolyte), which provides a wide work range of potential, as well as under similar conditions to those used in the measurement of dismutase-mimetic activity (water and HCl-tris buffer, which also was the supporting electrolyte). All the complexes show, by cyclic voltammetry in DMSO at platinum disk electrode, two well-defined one-electron reduction peaks A, and A, (see Figure 2). In contrast, if the sweep is reversed after the A, peak is scanned, only one oxidation peak A!:, is detected. The redox couple A,/Fi, (Table 4) seems to Al

I 12.5 )r A

A’, +o.s

A’2

0.0

-0.5

-1.0

-1.5

qv) vs. S.C.E

FIGURE 2. Cyclic voltagrams for [Cu(stz)Py,Cl], Cu(Hstz),(MeOH)Cl,, Cu(Hstz),(EtOH)Cl, complexes in DMSO.

and

SUPEROXIDE

TABLE

DISMUTASE

4. Electrochemical

Compound Cu(Hstz)z(EtOH)CI, Cu(Hstz),(MeOH)Cl, Cu(stz)py,C1

ACTIVITY

OF Cu(SULFATHIAZOLE)

227

Data for stz Complexes

-4,

AE, = E&A,) - E&A’,)

ipGl!,)/ipC4,)

A,

A’,

0.18 0.19 0.16

0.17 0.18 0.17

1.0 1.0

- 1.22 - 1.31 - 1.15

0.23 0.23 0.23

Potential values in volts vs. SCE scan rate = 0.1 V. se1 and measured in DMSO as solvent with NBu,PF, as supporting electrolyte and Pt as working electrode.

not be totally reversible; the high value of AE,(AE, = E,(A,) - Ep(A’,)) about 0.20 V suggests the possibility of a structural change or a chemical reaction coupled to the electrochemical reduction. However, the ratio ip(A’,)/ip(A,) is very close to 1.0, indicating a relative stability for the intermediate species, except for the Cu(stz)py,Cl complex. The second reduction process A, is totally irreversible. Taking into account the above-mentioned results, we can propose that the first reduction peak corresponds to the reduction process Cu(I1) + le + Cu(I), whereas the second reduction peak corresponds to the reduction process Cu(1) + le + Cu”; the total irreversibility of this process is,probably due to the dissociation of ligands and the formation of metallic copper. A’, and A, peaks must be assigned to the oxidation processes Cu” - le + Cu(1) and Cu(1) - le + Cu(II), respectively. For complex Cu(stz)py,Cl, peak A, is observed with very low intensity probably due to the small stability of the intermediate Cu(1) species. However, clearly detected in the reverse sweep is the oxidation peak A, assignable to the oxidation process Cu” - le + Cu(1). This peak is detected even if the A, is not scanned. We can tentatively interpret this result assuming a rapid disproportionation reaction of Cu(1) to give Cu” and C&I) species. The above-mentioned complexes were also studied by cyclic voltammetry at platinum disk electrode, under similar conditions to those used in the measure of the dismutase-mimetic activity (water and HCl-tris buffer, which also was the supporting electrolyte). However, under this condition, the electrochemical information obtained is limited by the narrower electrochemical range of water ( - 0.80 V vs. SCE electrode was the cathodic limit) as a consequence of the reduction of the protonic species. Under this condition, only the first reduction process of about 0.0 V is now detected, which we assign tentatively to the reduction to Cu(1) species. Superoxide Dismutation Activity The superoxide dismutase activity of the title complex was assayed by their ability to inhibit the reduction of nitroblue tetrazolium [25]. Figure 3 shows the % inhibition with increase in concentration of the copper complex. Table 5 shows the data of SOD activity of the copper sulfathiazole complexes. From the table, we can appreciate that the title complex shows about two times more activity than the Cu-MeOH and five times more than the Cu-EtOH complex, and 200 times less activity than the SOD. These values are consistent with those obtained for other complexes published in the literature [38-401. The mechanism believed to be operating in the metalloproteins involves one-electron reduction of a metal ion by superoxide followed by reoxidation of the reduced metal ion by a second superoxide anion. Metal complexes that can

228

J. Casanova et al.

60

0 0

2

4

6

6

10

Concentration( c M)

FIGURE 3. Inhibition of NBT reduction in the presence of the title compound. Each point represents the mean f standard deviation of triplicate determinations.

undergo such redox cycling are likely to function as superoxide scavengers. Recently, it was proposed that only flexible Cu(II)/Cu(I) chelates are genuine active site analogs of Cu/Zn-superoxide dismutase [41]. The distorted geometry of copper in the SOD enzyme also changes from distorted square pyramidal (for Cu(II)) to distorted tetrahedral (for Cu(1)) during catalysis [42]. Another factor that must be taken into account is the fact exchange of the axial molecules, generally water or other solvents. Finally, a possible correlation between the strength of equatorial field and the superoxide activity was suggested [43]. If we observe the value for the three copper complexes of sulfathiazole reported in Table 3, the higher distortion of the square pyramidal geometry is in the order Cu(II)-MeOH > Cu(II)-EtOH > Cu(II)-Py, while the order of superoxide-dismutase activity is Cu(II)-Py > Cu(II)-MeOH > Cu(II)-EtOH. As a consequence, the distortion may be one of the factors, but clearly is not the only important one. Bhirud and Srivastava [38] observed that the copper-aspirinate adduct of pyridine shows higher activity than the copper-aspirinate and the adduct with Imidazol, 4-picoline, and 1-Methyl-Imidazol. We consider that the presence of three pyridines in the title complex can increase the superoxide-dismutase activity in spite of the smaller distortion of the coordination polyhedron

TABLE

5. Superoxide

Cu(stz)py,Cl Cu(Hstz),(MeOH)Cl, Cu(Hstz),(EtOH)CI, Superoxide dismutase

Dismutase-Mimetic

Activity

IC;c,(PM)

-log IC,,

1.310

5.88

2.510 5.170 0.006

5.60 5.28 8.21

a ICs,, is defined as the concentration of complex or enzyme which produces 50% inhibition of NBT reduction. A molecular weight of 31,200 was considered for calculating the enzyme concentration.

SUPEROXIDE

DISMUTASE

ACTIVITY

OF Cu(SULFATHIAZOLE)

229

of the copper(R) ion. The potentials obtained under similar experimental conditions do not suggest important reactivity differences. SUPPLEMENTARY

MATERIAL

Final position and thermal parameters, anisotropic thermal parameters, hydrogen-atom parameters, distances and angles involving hydrogen atoms, leastsquares planes data, principal torsion angles, and a list of structure amplitudes of the complex have been deposited. We thank Dr. J. Timoneda (University of Valencia) for his help in the superoxide dismutase test. We greatly appreciate financial support from the Spanish CICYT (FAR 91-197). JC thanks the Generalitat Valenciana (Spain) for a doctoral fellowship.

REFERENCES 1. 2. 3. 4. 5. 6. 7.

P. A. Cerutti, Science 227, 375 (1985). S. J. Weiss, Acta Physiol. &and. 548, 9 (1986). C. Cross, Ann. Inter. Med. 107, 526 (1986). F. J. Burdon, V. Gill, and C. Rice-Evans, Free Rad. Res. Comms. 11, 65 (1990). G. Czapski, S. Goldstein, and D. Meyerstein, Free Rad. Res. Comms. 4, 221 (1988). G. Czapski and S. Goldstein, Free Rad. Res. Comms. 1, 157 (1986). K. L. Huber, R. Sridhar, E. H. Griffith, E. L. Amma, and J. Roberts, Biochim.

Biophys. Acta 915, 267 (1987). 8. J. R. J. Sorenson, J. Med. Chem. 27, 1747 (1984). 9. L. R. Pickart, Brit. U.K. Pat. Appl. GB. 2.213.060, Dec. 1987.

10. 11. 12. 13.

U. Weser and R. Meisel, German Patent Appl. P3912642, 1990. U. Weser, H. J. Hartmann, and R. Meisel, German Patent Appl. P3920826.5, 1991. J. R. J. Sorenson, Chem. in Brit. 16, 1110 (1984). N. Boden, M. C. Holmes, and P. F. Knowles, Biochem. Biophys. Res. Commun. 57, 845 (1975).

14. J. Casanova, G. Alzuet, J. Borrh, J. Timoneda, S. Garcia-Granda, and Y. CandanoGonzalez, J. Inorg. Biochem. (1994). 15. M. S. Lehman and F. K. Larsen, Acta Ctyst. A30, 580 (1974). 16. D. F. Grant and E. J. Gabe, J. Appl. Ctyst. 11, 114 (1978). 17. G. M. Sheldrick, SHELX86, in Crystallographic Computing 3, G. M. Sheldrick, C. Kruger, and R. Goddard, Eds., Clarendon Press, Oxford, 1985. 18. P. T. Beurkens, G. Admiraal, G. Beurkens, W. P. Bosman, S. Garcia-Granda, R. 0. Gould, J. M. M. Smits, and C. Smykalla, The DIRDIF Program System, Technical Report of the Crystallographic Laboratory, University of Nimegen, The Netherlands, 1992. 19. G. M. Sheldrick, SHELX76, Program for Crystal Structure Determination, University of Cambridge, 1976. 20. J. F. Van der Maelen, Ph.D. Thesis, Univ. of Oviedo, Spain. 21. N. Walker and D. Stuart, Acta Cryst. A39, 158 (1983). 22. International Tables for X-Ray Crystallography. Vol. IV, Kynoch Press, Birmingham, 1974. Present distributor: Kluwer Academic Publishers, Dordrecht. 23. M. Nardelli, Computer Chem. 7, 95 (1983). 24. A. L. Speck, The EUCLID Package, in Computational Crystallography, D. Sayre, Ed., Clarendon Press, Oxford, 1982.

230

J.Casanova et al.

25. L. W. Oberley and D. R. Spitz, in Handbook of Methods for Oxygen Radicals Research, R. A. Greenwald, Ed., 217, CRC Press, Boca Raton, FL, 1986. 26. W. Addison, T. N. Rao, J. Reedijk, J. van Rijn, and G. C. Verschoor, J. Chem. Sot. Dalton Trans. 1349 (1984).

21. B. J. Hathaway, Struct. Bonding (Berlin) 57, 55 (1984). 28. S. P. Sudhakara Rao, H. Manohar, K. Aoki, and H. Yamazaki, J. Chem. Sot. Dalton Trans. 1009 (1987).

29. M. Graf, B. Greaves, and H. Stoeckli-Evans, Inorg. Chim. Acta 204, 239 (1993). 30. S. S. Tandon, L. Chen, L. C. Thompson, S. P. Connors, and J. N. Bridson, Znorg. Chim. Acta 213, 289 (1993).

31. J. Casanova, G. Alzuet, S. Ferrer, J. Borrh, S. Garcia-Granda, and E. Perez-Carreiio, J. Znorg. Biochem. 51, 689 (1993). Zeit jiir 32. J. Casanova, G. Alzuet, J. Borras, J. M. Amigo, and T. Debaerdemaeker, Krist. 209, 271 (1993).

33. A. Bult and H. B. Klasen, J. Pharm. Sci. 67, 284 (1978). 34. N. S. Gill, R. H. Nuttall, D. E. Scaife, and D. W. A. Sharp, J. Znorg, Nucl. Chem. 18, 79 (1961). 35. J. H. Clark and C. S. Williams, Znorg. Chem. 4, 350 (1965). 36. J. Casanova, G. Alzuet, J. Borrbs, L. David, and D. Gatteschi, Znorg. Chim. Acta 211, 183 (1993). 37. B. J. Hathaway, Copper, in Comprehensiue Coordination Chemistry. G. Wilkinson, R. D. Gillard, and J. A. Cleverty, Eds., Pergamon Press, New York, 1987, Vol. 5. 38. R. G. Bhirud and T. Srivastava, Znorg. Chim. Acta 173, 121 (1990). 39. R. G. Bhirud and T. Srivastava, Inorg. Chim. Acta 179, 125 (1991). 40. L. L. Costanzo, G. D. Guidi, S. Giuffrida, E. Rizarelli, and G. Vecchio, J. Inorg. Biochem. 50, 273 (1993).

41. M. Linss and U. Weser, Znorg. Chim. Acta 138, 163 (1987). 42. J. S. Richardson, D. C. Richardson, J. A. Trainer, and E. D. Geltzoff, (London) 306,284 (1983). 43. C. Amar, E. Wilkas, and J. Foos, 1. Znorg. Biochem. 17, 313 (1982). Received October 4, 1994; accepted January 9, 1995

Nature