The dynamics of 57Fe nuclei in FeIII–DNA condensates

Journal of Inorganic Biochemistry 88 (2002) 14–18 www.elsevier.com / locate / jinorgbio

The dynamics of

57

Fe nuclei in Fe III –DNA condensates

Alfredo Trotta a , Adriana Barbieri Paulsen b , Arturo Silvestri a , Giuseppe Ruisi a , a a, Maria Assunta Girasolo , Renato Barbieri * a

Dipartimento di Chimica Inorganica, Universita` di Palermo, Viale delle Scienze – Parco d’ Orleans II, I-90128 Palermo, Italy b ¨ Physik, Medizinische Universitat ¨ zu Lubeck ¨ ¨ , Ratzeburger Allee 160, D-23538 Lubeck , Germany Institut f ur Received 15 May 2001; received in revised form 10 July 2001; accepted 20 August 2001

Abstract The dynamics of iron nuclei in the condensates obtained by interaction of Fe III with DNA, Fe III (DNA monomer) 2 , have been 57 ¨ investigated by variable temperature Fe Mossbauer spectroscopy. Studies were effected on gel and freeze-dried samples, obtaining nearly coincident values of the parameters isomer shift and nuclear quadrupole splitting in T ranges 20–260 K. Functions ln(A T /A 77.3 ) vs. T, here employed to investigate the dynamics of Fe nuclei, showed linear trends in the T ranges 20–150 and 150–260 K, respectively, the latter with larger slopes. Data coincided for gelled and freeze-dried specimens. No variation of d or DE parameters occurred at the two T intervals, which suggests constancy of structure and bonding with the temperature changes. Functions kx 2 l(T ) showed trends analogous to the corresponding functions determined for iron proteins, which were attributed to the occurrence of ‘conformational substates’.  2002 Elsevier Science B.V. All rights reserved. Keywords: Dynamics of

57

Fe; Fe III –DNA condensates

1. Introduction The interaction of metal ions with nucleic acids, including Fe II,III with DNA and RNA [1,2], is a very widely investigated field (see for instance Ref. [1]). Iron(III)– DNA complexes in aqueous solution show chelation of the metal ion by phosphate and base moieties [2]; moreover, stoichiometries of one iron to two nucleotides and one iron to 12 nucleotides occur [2]. Studies in the field continue at the present time [3,4], interest also arising from the natural occurrence of metal ions, including iron, in nucleic acids [5–7]. In fact, by XANES spectroscopy, Fe II,III is bound by oxygen (from phosphate, H 2 O or bases) and nitrogen (from bases), the latter causing a distortion of the bonding arrangement in Fe II –DNA [3]. Moreover, it is reported that Fe III forms stable complexes with DNA, being possibly bound to phosphate groups, and causes DNA condensation when the concentration of FeCl 3 , added to a DNA solution, is greater than 1.5310 24 M [4]. *Corresponding author. Present address: Dipartimento di Chimica Inorganica, Universita` di Palermo, Parco d’Orleans II, I-90128 Palermo, Italy. Tel.: 139-091-595322; fax: 139-091-427584. E-mail address: [email protected] (R. Barbieri).

Structure and bonding in iron ions interacted with biological molecules have been extensively investigated by 57 ¨ Fe Mossbauer spectroscopy (see for instance Ref. [8]), including iron–nucleic acid systems [9–12]. Studies in the field began quite early in 1966. Iron(III) in DNA and RNA complexes assumes an octahedral structure with bonds from phosphate groups and nitrogenous bases [9]; the most probable binding sites for Fe III are N 3 and N 7 of adenine and guanine [10]. Both Fe II – and Fe III –DNA have been ¨ investigated by 57 Fe Mossbauer spectroscopy [11,12]. We have previously reported on the interaction of DNA with organotin(IV) compounds with formation of DNA conden¨ sates [13–18]. According to 119 Sn Mossbauer spectroscopy IV IV studies, systems Alk 2 Sn – and Ph 2 Sn –DNA yield R 2 Sn IV(DNA phosphate) 2 with a trans-octahedral C 2 Sn structure, while in R 3 Sn IV(DNA phosphate) a trigonal bipyramidal tin environment occurs; bonding is effected by DNA phosphodiester groups and by water molecules [13– 15]. From tin dynamics by variable temperature 119 Sn ¨ Mossbauer spectroscopy, it has been inferred that R 2 Sn IV moieties bridge DNA phosphodiester groups in toroidal condensates through interstrand bonding, while R 3 Sn IV is appended to the double helix [16,17]. In particular, the dynamics of 119 Sn nuclei in DNA condensates evidenced

0162-0134 / 02 / $ – see front matter  2002 Elsevier Science B.V. All rights reserved. PII: S0162-0134( 01 )00339-7

A. Trotta et al. / Journal of Inorganic Biochemistry 88 (2002) 14 – 18

the occurrence of a linear function mean square displacement of 119 Sn atom, kx 2 l, with temperature [16–18]. In the context of Fe(III)–DNA condensates (this work), 57 Fe ¨ Mossbauer spectroscopy instead evidences a kx 2 l(T ) transition around 150 K, the function increasing with increasing temperature (Fig. 3, this work). This trend strictly corresponds to the structure dynamics of 57 Fe in proteins [19], the latter being a very widely investigated field [19–22], even at the present time [23].

2. Experimental The products employed in the present study were obtained commercially. Fe 2 O 3 enriched in 57 Fe to 95.27%, from here on referred to as 57 Fe 2 O 3 , was supplied by Plachinda ¨ (Lubeck, Germany). Tris was from Sigma (St. Louis, MO, USA), and other reagents and solvents were from C. Erba (Milan, Italy) and Aldrich Italia (Milan, Italy). An ethanol solution of 57 FeCl 3 was prepared by the following procedure: 82.4 mg (|0.5 mmol) 57 Fe 2 O 3 were dissolved in 2 ml of concentrated hydrochloric acid. The resulting solution was dried, and then left for 2 h on an oil bath at |1408C under a constant dry nitrogen flow. The obtained solid was dissolved in dry ethanol. The 57 Fe(III)–DNA condensates were obtained by adding ethanol solutions of 57 FeCl 3 to calf thymus DNA (|13 mmol dm 23 , in monomer units) in 1 mmol dm 23 tris(hydroxymethyl)aminomethane (Tris) in bidistilled water, pH 7.5; |10% EtOH in the final mixture; r (5mmol Fe(III) per mmol DNA monomer)50.4; immediate formation of condensate takes place. Pellets were recovered by centrifugation, and employed as absorber samples without further treatment, or after freeze-drying with Heto (Denmark) lyophilizers. The same procedure was employed to prepare a lyophilized Fe(III)–DNA condensate containing iron with natural isotopic composition. In order to determine the stoichiometry of the Fe(III)– DNA condensates, condensation experiments were performed, varying the Fe(III):DNA monomer ratio (r) from 0.2 to 1.0, followed by UV analysis of the supernatant liquor for the Fe(III) content. The graph mmol Fe(III) in the condensate vs. r shows a break point for r50.5 which allows us assign a 1:2 stoichiometry Fe(III):(DNA monomer). The iron(III) concentration in the supernatant liquor was determined by measuring the absorbance of the complex Fe(III)–Tiron at 667 nm (pH¯2, e 51911.5 cm 21 M 21 ). Measurements were performed with a Varian CARY (1E) spectrophotometer. ¨ The Mossbauer (nuclear gamma resonance) spectra of the condensates containing 57 Fe were recorded in Palermo with conventional spectrometers operating in the transmission mode. The source was 57 Co in rhodium matrix (Ritverc, St. Petersburg, Russia; 10 mCi), moving at room

15

temperature with constant acceleration in a triangular waveform. The variable-temperature spectra, at T $77.3 K, were recorded using an MVTIN 200 sample holder, a CTC 200 temperature controller and an MNC 200 liquid-nitrogen cryostat, from AERE (Harwell, UK), the temperature being measured with a chromel-alumel thermocouple (stabilized to 60.5 K). The driving system was from Halder (Seehausen, Germany), and the NaI(Tl) detector was from Harshaw (De Meern, Holland). Multichannel analysers and the related electronics were from Laben (Milan, Italy; model 4000) and Takes (Bergamo, Italy; model 269). Very narrow windows for the 14.4-KeV 57 Fe g-rays were selected in the analysers. Data reduction was effected by fitting the experimental spectra by Lorentzian lineshapes, using programs based upon iterative non-linear least-squares analysis. Calibration was effected with spectra from 57 Fe, the standard errors being: zero point: 60.001 mm s 21 ; g 0 , 60.0291 mm s 21 ; g 1 , 60.0166 mm s 21 . ¨ Variable temperature Mossbauer spectra for DNA condensate containing iron with natural isotopic composition

57

¨ Fig. 1. Quality of the Fe Mossbauer spectra for condensates [Fe(III)(DNA monomer) 2 ] obtained at 77.3 (a) and 183 K (b) for gelled absorber samples.

A. Trotta et al. / Journal of Inorganic Biochemistry 88 (2002) 14 – 18

16

¨ Physik, Mediziniswere also carried out at the Institut fur ¨ Lubeck, ¨ ¨ che Universitat, Germany, with a Mossbauer spectrometer working in a conventional constant acceleration mode with a g-source of 10 mCi 57 Co / Rh matrix (Amersham, UK). From calibration measurements a standard line width of 0.24 mms 21 was obtained. During ¨ Mossbauer measurements the sample was kept in a helium bath cryostat (MD306, Oxford Instruments, temperature range 1.5–295 K); measurements were executed by applying small fields of 20 mT perpendicular to g-rays with the help of a permanent magnet. The sample temperature in a cryomagnet was controlled with a variable temperature ¨ insert. The Mossbauer spectra were fitted with a locally developed program. Isomer shifts are given relative to a metallic iron (a-Fe) at room temperature. The quality of the spectra is shown in Fig. 1.

3. Results and discussion DNA condensation [24] may be induced by Fe(II, III) ions [4,9–12]. The stoichiometry of the Fe(III)–DNA condensates obtained in the present work has been determined as r50.5 mol Fe(III) per mol of DNA monomer (see Experimental section). This corresponds to a finding by Netto et al. [2] who report the stoichiometry of Fe(III)–DNA in aqueous solutions at pH 7.4 as: (a) one iron to two nucleotides; and (b) one iron to 12 nucleotides. Binding in (a) is interpreted in terms of possible bonds Fe–phosphates and Fe–nitrogen

bases [2], while no hypotheses have been advanced concerning the binding in (b) [2]. ¨ The 57 Fe Mossbauer spectra in Fig. 1 essentially correspond to earlier reports for Fe(III)–DNA condensates [9– 12]. The values of isomer shifts, d (Table 1), indicate the presence of high spin Fe(III) [25,26], analogously to previous studies [9–12]. Binding of Fe(III) to DNA phosphates and bases may be assumed to occur, for example in octahedral structures with four equatorial N→Fe(III) bonds, and axial phosphate–Fe(III) [9]. Two-line spectra, indicating the general occurrence of a limited, although measurable, value of nuclear quadrupole splitting, are detected (Table 1, Fig. 1). The dynamics of 57 Fe nuclei in Fe(III)(DNA monomer) 2 (i.e. the stoichiometry assumed for Fe(III)–DNA condensates) are reported in Table 1 and in Figs. 2 and 3. Data collection and treatment correspond to earlier reports concerning organotin(IV)–DNA condensates [16,17]. Only ‘relative’ values of the recoil-free fraction, f rel a , and consequently of the mean square displacement of 57 Fe nuclei, kx 2 l, are reported here, as their temperature dependence reflects that of ‘absolute’ functions, which is the aspect we privileged at this stage [16,17]. It appears that the dynamic parameters of 57 Fe(III), such as the mean square displacement of 57 Fe nuclei, kx 2 l, show larger changes starting from 150 K (Fig. 3). This behavior strictly corresponds to those reported for iron proteins, the variation of slopes for kx 2 l(T ) functions taking place at 200 K [19–22]. The trend may be interpreted in terms of the occurrence of ‘conformational substates’ (i.e. a number of different structures at the position of iron), which reach the

Table 1 57 ¨ Fe Mossbauer hyperfine parameters and molecular dynamics data for Fe(III)–DNA(calf thymus) condensates Phase

d a av (mms 21 )

DE b av (mms 21 )

G c av (mms 21 )

T range (K)

10 2 d ln(A t /A 77.3 ) / dT d (K 21 )

Gel

0.480 (60.004)

0.720 (60.004)

0.642 (60.004)

77.3–150

20.0191 (0.4425) 20.2779 (0.9813)

150–183

Freeze-dried

0.482 (60.004)

0.720 (60.004)

0.737 (60.003)

77.3–150 150–186

Freeze-dried e

0.474 (60.012)

0.654 (60.006)

0.304 (60.005)

20–110 160–260

20.0217 (0.6106) 20.222 (0.8627) 20.04781 (0.9940) 20.4087 (0.9915)

r5mol Fe(III) / mol DNA monomer50.5. All spectra were fitted with a symmetric doublet. a Isomer shifts with respect to metallic iron at room temperature, averaged over data in the whole temperature range, with standard error. b Nuclear quadrupole splittings, average values (see footnote a). c Full widths at half-height of the resonant peaks, average values (see footnote a). d Slopes of functions of normalized total areas under the resonant peaks (Lorentzian, A 5 (p / 2)eG, e being the percentage resonant effect) vs. T; correlation coefficients in parentheses. e ¨ of Lubeck, ¨ Spectra determined at the Medizinische Universitat in a transverse magnetic field of 20 mT (see Experimental section).

A. Trotta et al. / Journal of Inorganic Biochemistry 88 (2002) 14 – 18

17

Fig. 3. Relative recoil-free fractions, f rel a , and related mean square displacements, kx 2 l, of the 57 Fe nuclei as a function of temperature, for condensates (Fe(III)(DNA monomer) 2 ), freeze-dried (Table 1).

Fig. 2. (a) The function ln(A T /A 77.3 ) vs. T for condensates (Fe(III)(DNA monomer) 2 ), freeze-dried; A5(p / 2)eG is the total Lorentzian area under 57 ¨ the Fe Mossbauer peaks (e, percentage effect; G, full width at halfheight of the resonant peaks). Lines are least-squares fits y 5 ax 1 b (Table 1). (b) The dynamics of condensates (Fe(III)(DNA monomer) 2 ) strictly correspond for gel (D) and lyophilized (freeze-dried) (s) phases, including the transition temperature.

respective transition states and produce structural variations of Fe(III) bonding environment. In short, iron proteins would be characterized by multiple structures for the iron environments, which would contribute to kx 2 l after 200 K, thus giving the experimental with change (increase) of slopes kx 2 l(T ) at 200 K [19–22]. The same could be assumed to occur for Fe(III) (DNA condensates) (i.e. Fe III (DNA monomer) 2 , vide supra) with change of slope at 150 K (Figs. 2 and 3). The latter does not occur for R 2 Sn(DNA monomer) 2 and related complexes [16,17]. It could then be assumed that in the case of R n Sn(DNA monomer) 42n (n52,3) only one structure with phosphate–tin bonds occurs [16,17], while, for Fe III – DNA, structures such as Fe III (DNA monomer) 2 , variable bonding of Fe III with phosphates and nitrogen bases, would also occur [2], the latter possibly through hydrogen bonding by metal coordinated solvent molecules (see for instance Ref. [27]). Besides, Netto’s binding of one iron to

12 nucleotides (vide supra) [2] could be assumed to contribute to the effect. It appears that in the systems Fe(III)(DNA monomer) 2 the possible presence of water (in the gelled with respect to lyophilized samples) has practically no influence on the Fe(III) nuclei dynamics. In fact, data coincide until 150 K, and at higher temperature the gelled adsorber shows a very limited greater slope with respect to the lyophilized sample (Fig. 2b). This behavior corresponds to findings concerning myoglobin, where the presence of H 2 O would be very important for forming and breaking hydrogen bridges at increasing temperature [21]; in fact, H 2 O would act as a ‘plasticizer’, but dry myoglobin also reveals the occurrence of the same protein dynamics [21]. Lastly, it seems opportune to recall that the above discussed trends of mean square displacements of iron nuclei do not depend upon the polymeric nature of proteins and DNA. In fact, the occurrence of a ‘phase transition’, with changes in the slopes of functions ln(A T /A 90 ) vs. T, has been detected also for a series of iron(III) complexes [28–30]. Acknowledgements The contributions by Professor A.X. Trautwein and Dr ¨ Lubeck, ¨ B. Matzanke, Universitat are gratefully acknowledged. This work was supported by M.U.R.S.T., Ministero dell’Universita` e della Ricerca Scientifica e Tecnologica. References [1] A. Sigel, H. Sigel (Eds.), Interactions of Metal Ions With Nucleotides, Nucleic Acids and Their Constituents, Metal Ions in Biological

18

[2] [3] [4] [5] [6]

[7]

[8]

[9] [10] [11] [12] [13] [14] [15] [16]

A. Trotta et al. / Journal of Inorganic Biochemistry 88 (2002) 14 – 18 Systems, Vol. 32, Marcel Dekker, New York, 1996, (see also Vol. 33: Probing of Nucleic Acids by Metal Ion Complexes of Small Molecules). L.E.S. Netto, A.M. Da Costa Ferreira, O. Augusto, Chem. Biol. Interact. 79 (1991) 1–14. C. Bertoncini, R. Meneghini, D.Z. Cruz, M.C. Martins Alves, H. Tolentino, J. Synchrotron Rad. 6 (1999) 417–418. N. Kasyanenko, N. Arikainen, E. Frisman, Biophys. Chem. 70 (1998) 93–100. W.E.C. Wacker, B.L. Vallee, J. Biol. Chem. 234 (1959) 3257–3262. ´ N.E. E.L. Andronikashvili, L.M. Mosulishvili, A.I. Belokobilskiy, Kharabadze, N.I. Shonia, L.S. Desai, G.E. Foley, Biochem. J. 157 (1976) 529–533. T. Wilczok, E. Slotwinska Palugniok, A. Kochanska Dziurowicz, L. Mosulishvili, N. Kharabadze, Radiochem. Radioanal. Lett. 56 (1982) 131–140. ¨ J.G. Stevens et al., Mossbauer Effect Ref. Data J., 1–23 (1978– 2000); data listings Fe-57, 14.40 KeV transition, biological compounds, etc. Y.S. Moshkovskii, A.N. Ilina, R.A. Stukan, V.I. Goldanskii, Biofizika 11 (1966) 524–526, (English version, pp. 596–599). G. Bemsky, J.M. Dartyge, J. Peyre, G. Quintana, Fisica 1 (1970) 83–105. K.F. Tsitskishvili, Biofizika 28 (1983) 928–930, (English version, pp. 984–986). A.A. Kochanska Dziurowicz, T. Wilczok, Z. Dzierzewicz, Stud. Biophys. 122 (1987) 201–207. R. Barbieri, A. Silvestri, J. Inorg. Biochem. 41 (1991) 31–35. R. Barbieri, A. Silvestri, A.M. Giuliani, V. Piro, F. Di Simone, G. Madonia, J. Chem. Soc., Dalton Trans. (1992) 585–590. V. Piro, F. Di Simone, G. Madonia, A. Silvestri, A.M. Giuliani, G. Ruisi, R. Barbieri, Appl. Organometal. Chem. 6 (1992) 537–542. R. Barbieri, G. Ruisi, A. Silvestri, A.M. Giuliani, A. Barbieri, G.

[17]

[18] [19] [20] [21]

[22] [23] [24] [25]

[26]

[27]

[28] [29] [30]

Spina, F. Pieralli, F. Del Giallo, J. Chem. Soc., Dalton Trans. (1995) 467–475. R. Barbieri, F. Huber, L. Pellerito, G. Ruisi, A. Silvestri, in: P.J. Smith (Ed.), Chemistry of Tin, Blackie, London, 1998, pp. 496–540, Chapter 14. R. Barbieri, F. Huber, A. Silvestri, G. Ruisi, M. Rossi, G. Barone, A. Barbieri Paulsen, Appl. Organometal. Chem. 13 (1999) 595–603. F. Parak, L. Reinisch, Methods Enzymol. 131 (1986) 568–607. F. Parak, E.W. Knapp, Proc. Natl. Acad. Sci. USA 81 (1984) 7088–7092. F. Parak, M. Fischer, E. Graffweg, H. Formanek, in: E. Clementi, S. Chin (Eds.), Structure and Dynamics of Nucleic Acids, Proteins and Membranes, Plenum Press, New York, 1986, pp. 139–148, Chapter 24. F. Parak, Comments Mol. Cell. Biophys. 4 (1987) 265–280. C. Zach, C. Keppler, E. Huenges, K. Achterhold, F. Parak, Hyperfine Interact. 126 (2000) 83. V.A. Bloomfield, Curr. Opin. Struct. Biol. 6 (1996) 334. ¨ ¨ P. Gutlich, R. Link, A. Trautwein, Mossbauer Spectroscopy and Transition Metal Chemistry, Springer, Berlin, 1978 (see pp. 19 and 57). ¨ A.G. Maddock, Mossbauer Spectroscopy – Principles and Applications of the Techniques, Horwood Chemical Science Series, Horwood, Chichester, 1997, p. 108. K. Aoki, in: J.M. Lehn (Ed.), Metal Binding Effects On Nucleic Acid Structures, Comprehensive Supramolecular Chemistry, Vol. 5, Pergamon, Oxford, 1995, pp. 249–294. R.H. Herber, K. Temple, I. Manners, M. Buretea, T. Don Tilley, Inorg. Chim. Acta 287 (1999) 152–158. R.H. Herber, Inorg. Chim. Acta 291 (1999) 74–81. R.H. Herber, I. Gattinger, F.H. Kohler, Inorg. Chem. 39 (2000) 851–853.