Journal of Inorganic Biochemistry 92 (2002) 171–176www.elsevier.com/ locate/ jinorgbio

57 IIT he dynamics of Fe nuclei in Fe –DNA andII[Fe (1-methyl-2-mercaptoimidazole) ]–DNA condensates2

*Arturo Silvestri, Giuseppe Ruisi , Maria Assunta Girasolo`Dipartimento di Chimica Inorganica e Analitica Stanislao Cannizzaro, Universita di Palermo, Viale delle Scienze–Parco d’ Orleans II,

I-90128 Palermo, Italy

Received 23 April 2002; received in revised form 25 July 2002; accepted 23 August 2002

Dedicated to Professor Renato Barbieri on his 72nd Birthday

Abstract

IIAlcoholic solutions of FeCl and Fe (Hmmi) Cl (Hmmi51-methyl-2-mercaptoimidazole) induce calf thymus DNA condensation2 2 2II 57 ¨from aqueous solutions buffered at pH 7.4. A 1:1 Fe –(DNA monomer) stoichiometry is assumed. The Fe Mossbauer hyperfine

II IIparameters suggest an octahedral coordination environment, severely distorted, in both Fe –(DNA monomer) and [Fe (Hmmi) ]–(DNA2

monomer) condensates. The dynamic properties of iron nuclei in freeze-dried samples were investigated by means of variable temperature57 2¨Fe Mossbauer spectroscopy. Mean square displacements,kx l(T ), were calculated, such as the effective vibrating mass and the

2¨Mossbauer lattice temperature of the solids.kx l increases linearly with the temperature in the whole temperature range explored; theabsolute values are typical for lattice or solid-state vibrations. Very similar values for the effective vibrating masses were extracted,

¨suggesting comparable covalency of the bonding interaction between the metal atom and its ligands, while the Mossbauer latticeII IItemperatures show a softening of the lattice for [Fe (Hmmi) ]–(DNA monomer) with respect to Fe –(DNA monomer) condensate.2

2002 Elsevier Science Inc. All rights reserved.

¨Keywords: DNA; Iron(II); Dynamics; Mossbauer

1 . Introduction generally observed for cations with charge lower than 31

[4], unless cooperatively induced by alcohol [5].Metal ions and complexes interact with nucleic acids in We have previously reported on the temperature-depen-

¨a variety of ways, both strong covalent interactions and dent Mossbauer spectroscopy study of the dynamics ofweak noncovalent complexes being observed [1]. Soft DNA condensates obtained by reaction of aqueous (pHmetal ions may coordinate to nucleophilic centres on the 7.4) calf thymus DNA with ethanol solutions of alkyl-bases, while the harder transition metal ions are able to tin(IV) chlorides (Alk SnCl ,n51–3) [6], and FeCln 42n 3

coordinate to phosphate oxygen atoms. Less common are [7]. As far as the organotin(IV) derivatives are concerned,covalent interactions with the sugar moiety. Noncovalent bonding of tin to phosphodiester oxygen was assumed,

IVinteractions include intercalative stacking of a metal polymeric species at tin being formed in MeSn conden-IVcomplex and hydrogen bonding of coordinated ligands [1]. sate and in lyophilised samples of Me Sn condensate. In2

IIIIron(II,III) is known to interact with both oxygen (from Fe –DNA condensates bonding of iron(III) with nitrogenphosphate, H O or bases) and nitrogen (from bases) [2]. bases was also considered; the mean square displacement2

2III kx l of the iron nuclei as a function of the temperatureBesides, Fe forms stable complexes with DNA, andshowed a trend analogous to that reported for iron nuclei incauses DNA condensation by charge neutralization of the

2myoglobin, with a dramatic increase ofkx l at highphosphodiester groups [3], while no DNA condensation istemperatures as a consequence of the activation of proteinspecific movement modes [8–12]. In the present work we

57 ¨report on the temperature-dependent Fe Mossbauer spec-troscopy study of the dynamics of iron nuclei in conden-*Corresponding author. Fax:139-91-427-584.

E-mail address: gruisi@unipa.it(G. Ruisi). sates obtained by interaction of FeCl (from ethanol2

0162-0134/02/$ – see front matter 2002 Elsevier Science Inc. All rights reserved.PI I : S0162-0134( 02 )00566-4

172 A. Silvestri et al. / Journal of Inorganic Biochemistry 92 (2002) 171–176

Fig. 1. Thiol(I) and thione(II) forms of 1-methyl-2-mercaptoimidazole.

IIsolution) and the complex Fe (1-methyl-2-mercap-toimidazole) Cl (from methanol solution) with aqueous2 2

IIcalf thymus DNA at pH 7.4. 1-Methyl-2-mercap- Fig. 2. Iron(II) content in Fe –DNA condensate as a function of theIItoimidazole, whose thiol(I) and thione(II) tautomeric molar ratio Fe –(DNA monomer).

forms are shown in Fig. 1, belongs to a class ofN,Sligands pharmacologically very active, being itself (knownas methimazole) widely used in the treatment of thyroid DNA condensates, condensation experiments were per-

IIdisorders [13]. formed, varying the Fe –(DNA monomer) ratio (r) from0.75 to 3.45, followed by UV analysis of the supernatant

II IIliquor for the Fe content. The graphmmol Fe in thecondensate versusr shows a break point forr51 as

II2 . Experimental expected for a 1:1 stoichiometry Fe –(DNA monomer)(see Fig. 2 and Section 3). Iron(II) concentration in the

The products employed in the present study were supernatant liquor was determined by measuring theIIobtained from Carlo Erba (Milan, Italy) and Sigma–Al- absorbance of the complex Fe (1,10-phenanthroline) at3

21 21drich (Milan, Italy) and used without further purification 508 nm (́ 511100 M cm ) [14]. Measurements wereexcept methanol, which was distilled over magnesium. performed with a Varian CARY (1E) spectrophotometer.

IIFe (Hmmi) Cl was obtained by adding, dropwise and ¨The Mossbauer spectrometer, the data reduction method2 2

under constant stirring, a solution of iron(II) chloride (1 and thermal control have been described earlier [7]. Themmol) in ethanol (10 ml) to Hmmi (1-methyl-2-mercap- isomer shifts are reported with respect to the centroid of antoimidazole; 2 mmol) in ethanol (10 ml). A yellow a-Fe absorber spectrum at room temperature. Temperaturemicrocrystalline precipitate formed slowly. The reaction control is considered to be accurate within60.5 degrees.mixture was left for 2 h at room temperature, then the solid Experimental errors of the reported parameters reflect awas filtered off, washed with cold ethanol and vacuum 6s confidence level.dried over silica gel. (Found: C, 27.13; H, 2.57; N, 15.47;S, 18.89; Cl, 19.45. FeC H N S Cl requires C, 27.06;8 12 4 2 2

H, 3.41; N, 15.78; S, 18.06; Cl, 19.97%.) M.p. 162– 3 . Results and discussion1648C.

II IIThe condensates Fe –(DNA monomer) were obtained Fe (Hmmi) Cl was characterized by means of IR and2 257by adding a solution of iron(II) chloride in dry, de- ¨Fe Mossbauer spectroscopy. Relevant IR frequencies

oxygenated ethanol to calf thymus DNA (|13 mM, in [3114mw, n(N–H); 1470m, n(C=N); 1281m, d(N–H);monomer units) in aqueous Tris (Tris5 775s,p(N–H) out of plane; 495 mw,n(C=S)] are fullytris(hydroxymethyl)-aminomethane) 1 mM, pH 7.4, in the consistent with those reported for the analogous zinc and

IImolar ratio Fe –(DNA monomer) 1:1. The final mixture cadmium complexes, in which the metal was assumed tocontained|10% EtOH; the solution was shaken for few be tetrahedrally coordinated by sulphur, in the thione form,minutes until formation of a condensate took place. Pellets and chlorine ligand atoms [15]. Besides,n(Fe–Cl) stretch-

21were recovered by centrifugation and freeze-dried with a ing frequencies are found at 337 and 287 cm (medium21Heto (Denmark) lyophiliser. The same procedure was intensity bands), while a weak absorption at 222 cm

IIemployed to obtain lyophilised [Fe (Hmmi) ]–DNA con- could be assigned ton(Fe–S) [16]. A tetrahedral structure2II 57densates, being the complex Fe (Hmmi) Cl dissolved in with S atoms coordinated to iron is also supported by Fe2 2

¨methanol owing to his poor solubility in ethanol. Mossbauer parameters (see Table 1), quite comparable, forIIIn order to determine the stoichiometry of the Fe – example, to those of thioacetamide and thiourea deriva-

A. Silvestri et al. / Journal of Inorganic Biochemistry 92 (2002) 171–176 173

Table 157 II II II¨Fe Mossbauer data for the complex Fe (1-methyl-2-mercaptoimidazole) Cl and the condensates Fe –(DNA monomer) and [Fe (1-methyl-2-2 2

mercaptoimidazole) ]–(DNA monomer)2

a b c dCompound T d D G21 21 21(K) (mm s ) (mm s ) (mm s )

II eFe (Hmmi) Cl 77.3 0.99 3.33 0.522 2

RT 0.85 3.05 0.27IIFe (Hmmi) Cl , 0.015 M 77.3 1.35 3.32 0.392 2

in MeOHIIFe –(DNA monomer) (gel) 77.3 1.35 3.36 0.36

0.29 0.42 0.44II[Fe (Hmmi) ]–(DNA monomer) 77.3 1.31 2.99 0.682

(gel) 0.31 0.40 0.73IIFe –(DNA monomer) (freeze- 77.3 1.28 2.87 0.64

dried) 0.36 0.53 0.65280 1.21 2.59 0.75

0.36 0.59 0.69II[Fe (Hmmi) ]–(DNA monomer) 77.3 1.16 2.97 0.802

(freeze-dried) 0.36 0.53 0.60280 1.10 2.63 0.91

0.31 0.43 0.82a Hmmi51-methyl-2-mercaptoimidazole.b Isomer shift relative to RTa-iron.c Nuclear quadrupole splitting.d Full width at half height of the resonant peaks.e Solid state absorber.

¨tives FeL X [17] where coordination of iron(II) through Mossbauer parameters for the condensates obtained in such2 2

sulphur atoms was proposed on the basis of infrared experiments indicates that the same iron species are21spectra. The high quadrupole splitting value (3.33 mm s formed, irrespective ofr value. The same 1:1 stoichiome-

IIat 77.3 K) and its small temperature dependence (D53.05 try was assumed for the condensate [Fe (Hmmi) ]–DNA.221 ¨mm s at RT) reveal a remarkable distortion of the The Mossbauer spectra of all the condensates, in both

structure due to the steric hindrance of the ligand as well gelled and freeze-dried phases (Table 1 and Fig. 3) areas to the difference in covalent character of Fe–L and characterized by a wide doublet, whose isomer shift is

IIFe–Cl bonds. The complex is appreciably soluble in typical of high spin Fe compounds, and a low intensitymethanol and in water where it dissociates extensively as doublet with small isomer shift and quadrupole splitting

24evidenced by its molar conductivity; for a 5310 M values indicating the presence of ferric ions. The presence21 21 2 II IIIaqueous solutionL 5246.8 V mol cm at 228C of both Fe and Fe ions is supported by the similarity ofM

21 21 2(L 543.0V mol cm for tetraphenylarsonium chlo- the spectra with those of DNA condensates containingM

ride in the same conditions). The high value of the iron(II) and iron(III) ions, reported by Tsitskishvili [19].¨Mossbauer parameter isomer shift,d, of a 0.015 M Partial oxidation of iron(II), resulting from bonding to

ˇmethanol solution, frozen by rapid immersion into liquid DNA, was already observed by Greguskova et al. [20]. Thenitrogen (see Table 1), shows that iron(II) expands its relative amount of ferric iron can be estimated from the

¨coordination number [18] due to solvation. relative area of the Mossbauer spectrum at 77.3 K,Gel phases (condensates) were obtained by interaction of assuming that the recoil-free fractions for iron(II) and

a calf thymus DNA solution in aqueous 0.1 mM Tris, pH iron(III) nuclei assume comparable values at low tempera-II7.4, with iron(II) chloride and Fe (Hmmi) Cl solutions ture. Iron(III) content so calculated is about 20 and 18%,2 2

II IIin ethanol and methanol, respectively. As far as the Fe – respectively, in the condensates Fe –(DNA monomer) andIIDNA condensates are concerned, the stoichiometry was [Fe (Hmmi) ]–(DNA monomer).2

established performing condensation experiments with The high value of the isomer shift for the main doubletvariable amounts of FeCl added to a solution of calf suggests octahedral coordination of iron(II) in all samples,2

thymus DNA. The analysis of the supernatant liquor for whether gelled or freeze-dried [18]; the small temperatureIIthe Fe content gave an estimate of the condensate dependence in the quadrupole splittings of freeze-dried

IIcomposition. The graphmmol Fe in the condensate samples indicates that distortional splittings in the t2gIIversusr, the ratiommol Fe tommol (DNA monomer) in levels are of considerable size [21]. Minor differences in

II ¨the reaction mixture (Fig. 2), shows that the Fe content in the Mossbauer hyperfine parameters are related to thethe condensate increases linearly withr, attaining a nearly nature of the ligands and/or to the distortion degree.

II IIsteady value forr51, which suggests a 1:1 Fe –(DNA The dynamics of iron nuclei in the condensates Fe –57 IImonomer) stoichiometry. The constancy of the Fe DNA and [Fe (Hmmi) ]–DNA was investigated by means2

174 A. Silvestri et al. / Journal of Inorganic Biochemistry 92 (2002) 171–176

2 IIFig. 4. Mean square displacement of iron nuclei,kx l(T ), for Fe –(DNAIImonomer) (n) and [Fe (Hmmi) ]–(DNA monomer) (s) condensates as2

a function of temperature.

mass: at high temperatures, the difference in the isomershifts for a sample at different temperatures is given by Eq.(2) [22]:

2 3kB]]Dd 5 DT (2)M ? ceff

wherek is the Boltzmann constant,M is the effectiveB eff

vibrating mass andc is the speed of light in vacuum. Eq.(2) can be rearranged to give

21dd22 ]F GM 5 2 4.1602310 (3)57 II eff¨Fig. 3. Fe Mossbauer spectra of Fe –(DNA monomer) and dTII[Fe (Hmmi) ]–(DNA monomer) condensates, freeze-dried, at 77.3 K.2 so that the effective vibrating mass can be easily calcu-

lated.57 ¨of variable temperature Fe Mossbauer spectroscopy, ¨The Mossbauer lattice temperature,Q , can be esti-M

analysing the temperature dependence of both the areamated from the temperature dependence of both the recoil-under the resonant peaks and the isomer shifts in thefree fraction, which is related to the area under thetemperature range 77.3–280 K. Mean square displace-resonant curve, and the isomer shift. The temperature

2ments,kx l(T ), of the iron nuclei were determined from the dependence of the recoil-free fraction is given by [22]:recoil-free fraction:

23Ed ln A g]] ]]]]5 2 (4)2 2 2 2f 5exp(2k kx l) (1) dT M c k Qa eff B M

57 Eq. (4) can be combined with Eq. (3) to give:˚with k 52p /l andl50.86 A for Fe, which is obtained]]]¨from the absorption area of the Mossbauer spectra. The dd /dT2 ]]]procedure outlined by Parak and Reinisch in Ref. [8] was Q 54.32033 10 (5)M œd ln A /dT2followed. kx l(T ) values were not estimated for the

iron(III) site as the low intensity iron(III) doublet heavily The lattice dynamic data are summarized in Table 2.¨overlaps one side of iron(II) doublet and this hinders a Mossbauer absorption spectroscopy was proved to be an

precise evaluation of its hyperfine parameters. The func- useful tool in the study of the dynamics of iron-containing2tions kx l(T ) for the iron nuclei in the systems here proteins [8–12]. It allows the determination of the mean

2investigated are shown in Fig. 4. Error bars reflect the square displacementkx l of the iron caused by all modes ofstandard deviations, which are 0.0004 27 and 0.0004 14 for motions coupled to the iron with a characteristic time of

II II 28 27Fe –(DNA monomer) and [Fe (Hmmi) ]–(DNA mono- 10 to 10 s. The time sensitivity is determined by the257¨mer), respectively. lifetimet of the Mossbauer 14.4 keV level of the FeN

The temperature dependence of the isomer shifts,d, nucleus.analysed in terms of the Debye model for the second-order In proteins, the three-dimensional architecture of theDoppler shift, allows to estimate the effective vibrating molecules is held together by weak forces, such as

A. Silvestri et al. / Journal of Inorganic Biochemistry 92 (2002) 171–176 175

IIITable 2 studies effected on Fe –(DNA monomer) condensates257 II¨Lattice dynamic Fe Mossbauer data for Fe –(DNA monomer) and both in gel and freeze-dried phase [7]. The trends observedII[Fe (Hmmi) ]–(DNA monomer) condensates freeze-dried 22 for the functionskx l(T ) are analogous to those determinedIIFe –(DNA monomer) (21 points, 77.3#T#280 K) for iron proteins, with aT of about 150 K: iron(III)–DNAc2 25kx l(T ) 5 (5.6860.15)310 T R50.9932 condensates seem to behave as iron proteins as far as the24

d(T ) 5 1.31160.005–(3.6560.18)310 T R50.977623 dynamics of iron nuclei is concerned. Data practicallyln A(T ) 5 22.46760.015–(2.9560.08)310 T R50.9934

21 coincided for gelled and freeze-dried specimens suggestingM 5 11466 g moleff

Q 5 15268 K that in these systems the presence of water (in the gelledM

with respect to freeze-dried samples) has practically noII

III[Fe (Hmmi) ]–(DNA monomer) (15 points, 77.3#T#280 K)2 influence on the Fe nuclei dynamics. On the other hand,2 24kx l(T ) 5 (1.0560.02)310 T R50.998224 no characteristic temperature was previously observed ford(T ) 5 1.19460.006–(3.5360.38)310 T R50.9304

23 R Sn(DNA monomer) and related complexes [23,24]. Itln A(T ) 5 22.22060.016–(5.4760.09)310 T R50.9982 2 221M 5 118613 g mol was assumed that while in the case of R Sn(DNAeff n

Q 5 110612 KM monomer) (n52,3) only phosphate–tin bonds occurs,42nIII IIIfor Fe –DNA systems variable bonding of Fe with

phosphates and nitrogen bases would occur, the latterpossibly through metal coordinated solvent molecules [7].hydrogen bridges, hydrophobic interaction, van der Waals

IIIIn contrast with the findings concerning Fe –DNAforces, and Coulomb interaction, which make segmental2 IIsystems, the functionskx l(T ) for Fe –(DNA monomer)motions possible. The conformation of the biomolecule as

IIand [Fe (Hmmi) ]–(DNA monomer) condensates aredetermined by X-ray diffraction is an ensemble average 2

linear in the temperature range explored, only harmonicwhich contains molecules with structures distributedvibrations being detected. This may be tentatively inter-around the average structure. Each individual molecule ispreted in terms of less extensive interactions betweenin one conformational substate. One biomolecule caniron(II) and DNA, with respect to iron(III), with phosphatefluctuate between different conformational substates if thegroups alone probably involved in the bonds, as inthermal energy is large enough to overcome the energy

57 organotin derivatives. We do not exclude however that thebarrier between these states. If a Fe atom is incorporatedlack of any ‘protein dynamics’ in iron(II)–DNA conden-into a biomolecule this atom can be used as a marker forsates may be a consequence of the freeze-drying process,fluctuations.which would be more effective in the suppression of anyMean square displacements calculated as a function ofconformational transition than in the case of myoglobinthe temperature for myoglobin, as well as for some model[10] or Fe(III)–DNA condensates.systems, exhibit different behaviour in the low and high

2 The temperature dependence of the isomer shifts intemperature regions. Below about 200 K,kx l increaseII IIFe –(DNA monomer) and [Fe (Hmmi) ]–(DNA mono-linearly with temperature, the absolute values being typical 2

mer) is shown in Fig. 5. The standard deviations are 0.005for solid-state vibrations. Above about 200 K a drastic2 02 and 0.009 23, respectively, as shown by the error barsincrease of kx l values with temperature suggests the

in the figure.activation of new modes of motion. The large displace-The effective vibrating masses, extracted from thements observed at high temperature are believed to be

21slopes dIS /dT, are 11466 and 118613 gmol , respec-protein specific and are due to the structure dynamics ofII IItively, for Fe –(DNA monomer) and [Fe (Hmmi) ]–the protein system. Only above the dynamic transition 2

21(DNA monomer). The difference from 57 gmol , thetemperature,T , can the protein fulfil its physiologicalc

atomic weight of a ‘bare’ iron atom, reflect the covalencyfunction. Quasi diffusive motions are obviously necessaryof the bonding interaction between the metal atom and itsfor conformational transitions; moreover, when quasiligands [22,25]. On the other hand, the effective vibratingdiffusive motions are present, channels can open allowingmasses of both the condensates practically coincide, andsmall molecules to diffuse into the protein [12].this would suggest that the interaction between the metalThe mobility of a protein is strongly correlated with theatom and its nearest neighbor environment be qualitativelymobility of the surrounding water which acts as a plasti-

¨similar. Nevertheless, the Mossbauer lattice temperatures,cizer for the protein. It is generally recognized that proteinsII

Q , differ somewhat, being that of [Fe (Hmmi) ]–(DNAwith a hydrophilic surface do not exhibit the normal M 2IIdynamics in the dry state. It is important to note, however, monomer) some 28% lower with respect to Fe –(DNA

¨that a dry sample of myoglobin investigated by Mossbauer monomer). The lower value forQ , and the consequentlyM2spectroscopy showed specific dynamics clearly visible, larger temperature dependence of the functionkx l(T ), is

even if strongly reduced [10]. Moreover, the protein- indicative of a ‘softening’ of the lattice presumably due to2specific increase in thekx l value occurs at about the same a larger density of low frequency vibrational motions in

II IItemperature in dry myoglobin and in myoglobin crystals the [Fe (Hmmi) ]–(DNA monomer) with respect to Fe –2

[10]. (DNA monomer), to which the iron(II)–Hmmi bonds57 ¨Recently, we reported on Fe Mossbauer spectroscopy probably contribute.

176 A. Silvestri et al. / Journal of Inorganic Biochemistry 92 (2002) 171–176

[4] Y. Yamasaki, K. Yoshikawa, J. Am. Chem. Soc. 119 (1997) 10573–10578.

[5] N.A. Kasyanenko, A.V. Zanina, O.V. Nazarova, E.F. Panarin, Lang-muir 15 (1999) 7912–7917.

[6] A. Silvestri, G. Ruisi, R. Barbieri, Hyperfine Interact. 126 (2000)43–46, and Refs. therein.

[7] A. Trotta, A. Barbieri Paulsen, A. Silvestri, G. Ruisi, M.A. Girasolo,R. Barbieri, J. Inorg. Biochem. 88 (2002) 14–18.

[8] F. Parak, L. Reinisch, Methods Enzymol. 131 (1986) 568–607.[9] F. Parak, E.W. Knapp, Proc. Natl. Acad. Sci. USA 81 (1984)

7088–7092.[10] F. Parak, M. Fischer, E. Graffweg, H. Formanek, in: E. Clementi, S.

Chin (Eds.), Structure and Dynamics of Nucleic Acids, Protein andMembranes, Plenum Press, New York, 1986, pp. 139–148, Chapter24.

[11] F. Parak, Comments Mol. Cell. Biophys. 4 (1987) 265–280.[12] F.G. Parak, G.U. Nienhaus, Chem. Phys. Chem. 3 (2002) 249–254.[13] U. Bandyopadhyay, K. Biswas, R.K. Banerjee, Toxicol. Lett. 128

(2002) 117–127.[14] G. Charlot (Ed.), Chimie Analytique Quantitative, Vol. 2, Masson et

C.ie, Paris, 1974.[15] R. Shunmugam, D.N. Sathyanarayana, J. Coord. Chem. 12 (1983)

151–156.[16] I.P. Evans, G. Wilkinson, J. Chem. Soc., Dalton Trans. (1974)IIFig. 5. Isomer shift, d, for Fe –(DNA monomer) (n) and

946–951.II[Fe (Hmmi) ]–(DNA monomer) (s) condensates as a function of2 [17] T. Birchall, M.F. Morris, Can. J. Chem. 50 (1972) 211–216.temperature.

¨[18] A.G. Maddock, in: Mossbauer Spectroscopy—Principles and Appli-cations of the Techniques, Horwood, Chichester, 1997, pp. 104–140, Chapter 5.A cknowledgements

[19] K.F. Tsitskishvili, Biofizika 28 (1983) 928–930.ˇˇ ´ ´ ´[20] M. Greguskova, J. Cirak, J. Novotny, I. Cernohorsky, in: 5th Int.This work was supported by MURST, Ministero de-

¨Conf. On Mossbauer Spectroscopy, Bratislawa, Czechoslovakia`ll’Universita e della Ricerca Scientifica e Tecnologica, (1973); Proceedings, Part 2, 1975, pp. 415–419.

Roma. [21] T. Birchall, M.F. Morris, Can. J. Chem. 50 (1972) 201–210.¨[22] R.H. Herber, in: R.H. Herber (Ed.), Chemical Mossbauer Spec-

troscopy, Plenum, New York, 1984, pp. 199–216, Chapter 7.[23] R. Barbieri, G. Ruisi, A. Silvestri, A.M. Giuliani, A. Barbieri, G.R eferences

Spina, F. Pieralli, F. Del Giallo, J. Chem. Soc., Dalton Trans. (1995)467–475.

[1] K. Barton, in: I. Bertini, H.B. Gray, S.J. Lippard, J.S. Valentine [24] R. Barbieri, F. Huber, L. Pellerito, G. Ruisi, A. Silvestri, in: P.J.(Eds.), Bioinorganic Chemistry, University Science Books, Smith (Ed.), Chemistry of Tin, Blackie, London, 1998, pp. 496–540,Sausalito, CA, 1994, pp. 455–503, Chapter 8. Chapter 14.

[2] C. Bertoncini, R. Meneghini, D.Z. Cruz, M.C. Martins Alves, H. [25] G.J. Long, D. Hautot, F. Grandjean, D.T. Morelli, G.P. Meisner,Tolentino, J. Synchrotron Rad. 6 (1999) 417–418. Phys. Rev. B 62 (2000) 6829–6831.

[3] N. Kasyanenko, N. Arikainen, E. Frisman, Biophys. Chem. 70(1998) 93–100.