Raman spectroscopy a promising technique for investigations of metallothioneins

Abstract

Metallobiomolecules are highly elaborated coordination complexes, and their fundamental metal–ligand interactions are critical components of metalloprotein folding, assembly, stability, electrochemistry, and catalytic function. Herein, we have described the benefits in using Raman spectroscopy to define the metal-ion binding properties of MTs toward metal ions such as Zn(II) and Cd(II). In particular, this vibrational technique can shed light on the secondary structures eventually present in MTs and the ligands involved in metal coordination. The oxidation state of Cys residues and their participation in the metal chelation can be clearly defined, as well as the eventual involvement of His residues. With regards to exogenous metal ligands such as sulfide anions, their presence can be identified by some marker bands whose intensity is linearly correlated with sulfide/metal molar ratio. Finally, Raman can be also an useful tool for providing information on the favourite sites of the radical attack and radical-induced modification in protein folding. In conclusion, many advantages such as the capability of defining local regions in large complexes and detecting several structural features at the same time, the ability in supporting mechanisms, as well as the requirement of low sample amount, make to propose Raman spectroscopy, in coupling with analytical techniques such as atomic emission spectroscopy, gas chromatography, and circular dichroism, as one of the most promising experimental strategies in the research on structure–activity relationships in MTs.

---

Armida Torreggiani

Armida Torreggiani received her degree in Chemistry in 1992 and her PhD in Biochemistry in 1997 at the University of Bologna (Italy). After several fellowships from the National Research Council, University of Bologna and Brown University in Providence, Rhode Island (USA), she became a researcher in 2001. She is member of the scientific committee of European Conference on the Spectroscopy of Biological Molecules. Her research interests are centred in chemical biology, with particular reference to characterisation of protein structure by Raman spectroscopy and evaluation of free radical-damages on proteins. She is author of over 80 papers on international journals.

Anna Tinti

Anna Tinti received her degree in Chemistry in 1981 and PhD on Biocompatibility of materials in Orthopaedics and Dentistry in 1986 at the University of Bologna. Since 2001 she is researcher to the school of Medicine. Her main research interests involved the structure of apatite-based biomaterials and the degradation of polymer-apatite composites for tissue engineering; the crystallinity and oxidation of explanted polyethylene acetabular cups; the characterisation of secondary and tertiary structure of self-assembling oligopeptides and the characterisation of secondary structure, metal and binding properties of proteins. She has co-authored over 120 research publications, in the biomedical and protein fields.

---

1. Introduction

The exploration of protein structures provides the backbone for today's research in molecular biology and biochemistry. Although X-ray crystallography and NMR spectroscopy are two accurate methods which are helping in expanding the molecular view of the biological world, techniques are needed that facilitate the transition in our understanding from structure to function. Raman spectroscopy is one of these.

Raman spectroscopy involves analysing the scattered photons from a laser beam focused into the sample. A small percentage of the scattered photons exchanges energy with the vibrational energy levels (or, crudely, the “vibrations”) of the molecules in solution. Thus, by analyzing the scattered photons information on the vibrational motions of atoms in molecules is obtained. These motions are a function of molecular conformation, of the distribution of electrons in the chemical bonds, and of the molecular environment. Thus, the interpretation of Raman spectrum provides information on all these factors and very detailed information on local sites in a much larger macromolecule complex can be obtained.¹

Raman spectroscopy is beginning to fulfil its potential to contribute to structural biology because the developments in commercial Raman instrumentation have permitted to remove the three roadblocks that impeded its application to biological systems. These were low sensitivity, interference from fluorescence background, and problems with data interpretation. Sensitivity has increased several orders of magnitude with the corresponding decrease in concentration requirements because of advances in optical filters and photon detectors.² Fluorescence interference is now minimized by using deep-red excitation in the 800–1070 nm range, made possible by the advent of photon detectors with high efficiency in this region.³ Problems with interpreting Raman spectra have receded with the availability of “friendly” software packages⁴ and ever increasing computational power that enable us to calculate, ab initio, the Raman spectra of mid-sized molecules (of the size of many ligands or co-factors found at biological sites).

Thus, Raman spectroscopy has become a versatile tool in protein science and its applications in characterising protein in biotechnology , pharmaceutical production and food industry have been developed.

Raman is extremely useful in detecting bonding changes. The structural changes resulting from these can be among only a few atoms and the overall protein structure.^5,6 In fact, this technique can be successfully used for the determination of protein secondary structure, the identification of metal coordination sites, hydrogen bonding, oxidation state of cysteine residues, local environments of aromatic residues (i.e.Tyr, Trp, His), protein–ligand and protein–DNA interactions, etc. Thus, it is emerging as an indispensable tool for detecting minute structural changes^5,6 and, in that sense, it complements techniques such as X-ray crystallography that provide the “big picture.”

2. Applications to metallothioneins

Metallothioneins (MTs) is a collective name for a superfamily of low molecular weight (6–7 kDa) metalloproteins. They typically have about 60 amino acid residues, one-third of which are cysteine. For example, rat liver metallothionein (MT) and human MT-2 have 61 residues, 20 of which are Cys. Not infrequently, proteins from a single organism occur as isoforms, sometimes differing by only a single amino acid residue. No other protein type has such a high Cys content. These residues bind a variable number of metal ions, with loadings of 4–12 metal ions per molecule rather common. Metal content depends on the organism and the extent of heavy metal exposure. While a number of functions have been ascribed to MTs, those which are frequently associated with these proteins are homeostatic regulation of essential metals such as zinc and copper, and detoxification by sequestration of physiologically adverse metals such as cadmium and mercury.^7,8 For example, cadmium induces the biosynthesis of MTs.

Additional suggested functions are, among others, protection against reactive oxygen species (ROS), adaptation to stress, antiapoptotic effects and regulation of neuronal outgrowth.^9–11

As would be expected, apo-MT (apo means free metal-bound) binds practically any transition metal ion, but their interactions with an array of soft metals including Zn(II), Cd(II), Hg(II), Cu(I), Ag(I), Au(I), and Pt(II) have been the most thoroughly examined.

Petering et al. have effectively summarized the functional aspects of MT related to metal ion homeostasis and sequestration of toxic metals: “Because of this unusual kinetic lability as well as the thermodynamic stability (owing to soft acid –soft base interactions) of the MT species that are formed, MT acts as a sink for the binding of a variety of essential and toxic metal ions which enter cells.”¹²

MTs are ubiquitously distributed among all living organisms such as Eukaryota (animals, plants, fungi and protozoa) and some Prokaryota (Cyanobacteria).¹³ Most of the chemical, structural, genetic and physiological information available nowadays for MTs has its origin in the exhaustive analysis of the mammalian isoforms, particularly MT1 and MT2,^7,8 but unfortunately, the lack of homology among most MTs precludes any direct extrapolation of their structural and functional features. For this reason, the direct characterization of at least one member of each kind of MT subfamily is necessary to render a general overview of how these peptides emerged, evolved and adapted to diverse functional conditions in the different taxa.

Overall, the general structural properties that characterize all MTs are the formation of metal–thiolate clusters involving terminal and bridging cysteinyl thiolate groups.⁸ For example, in Cd₅Zn₂-MT-2 from rat liver, the seven metals are bound in two individual subdomanis, the β or N-terminal domain consisting of residues 1–30, among which are 9 Cys, and the α or C-terminal domain including residues 33–61, of which 11 are Cys.¹⁴ The metal sites of the rat liver protein divide into two polynuclear aggregates and all seven coordination sites are tetrahedral, as would be expected for four-coordinate Zn(II) and Cd(II); the 20 Cys residues are fully utilized in metal binding.

Recently, it has been shown that metal coordination to the MT polypeptides constitute a more complex scenario that the simplistic consideration of metal–thiolate bonds contributed by the MT cysteine residues. Particularly, the participation of other ligands in the metal coordination sphere of MTs has been shown for other amino acid side chains (His, Asp, Glu)^15–19 and for exogenous inorganic ligands, such as sulfide or chloride ions.^20,21 Consequently, a deeper knowledge of the metal coordination sites can be of benefit in establishing structure–activity relationships in MTs.

Besides, the presence of secondary structure elements can play a role in the determination of MT functional properties. Up to now data on secondary structures in MTs are quite sparse and it is generally assumed that secondary structure elements are poor or functionally insignificant in MTs, although they play a definite role in other metalloprotein functionality, like zinc fingers.²² However, the few data reported on MT secondary structure elements support their relevance for MT functionality, significantly in prokaryote SmtA¹⁵ and mammalian MT3.²³

The huge amount of knowledge gathered especially for mammalian MT structures, but also for other vertebrate, fungal, and prokaryote MTs, contrasts with the scarcity of structural data on invertebrate MTs.²⁴ Three-dimensional data are only available for the Cd-complexes of crustacean^25,26 and echinoderm²⁷isoforms. X-ray crystal or NMR structure determination is a time- and material-consuming process, and therefore all efforts to enlarge the range of alternative methodologies to get insights into the structural features of the metal-MT complexes have to be strengthened.

Raman spectroscopy has been used for protein investigation for more than three decades, during which specific band assignments, signatures of secondary structure and Raman markers of side chain environments have been established.^28–31 The assignments are based on model compounds (e.g.amino acids or short peptides) and are universal for the whole class of molecules (e.g.amide bands of proteins are assigned to polypeptide backbone modes). Despite the well-known potentialities of this techniques, to our knowledge it has been scarcely used in MT structural studies until now. Thus, a major goal of this review is to provide researchers with enough information to determine whether the Raman technique could provide structural insights into their MT systems. In fact, this spectroscopy can be very useful to approach unambiguously two basic structural points in MTs: the participation of ligands to the metal-coordination sphere (i.e.Cys, His, sulfide and chloride ions, etc.) and the presence of secondary structure elements.

Thus, in this review we will describe the capacities of the Raman technique on the characterization of the metal-binding properties and secondary and tertiary structure of some Zn- and Cd-complexed MTs, synthesized by recombinant expression in metal-supplemented bacterial cultures (E. coli). The analysed metal-metallothionein (M^II-MTs) complexes are representative of different MT families, enclosing the well-known vertebrate and echinoderm, and the poorly understood nematode, diptera, molluscan, protozoa and plant MTs (Table 1).^32–37

Table 1 Description of the MT isoforms, including their total amino acid content, number of Cys and aromatic amino acid residues, analyzed by Raman spectroscopy and considered in this work. Arom aa = aromatic amino acid residues; Synthesis = heterologous synthesis in E. coli, followed by purification

MT	Origin and organism		Amino acids	Cys	Arom. aa	Synthesis
CeMT2	Isoform 2 of the nematode Caenorhabditis elegans. It contains a C-terminal His residue.^32		63	18	1 His 1 Tyr	Zn-suppl. cultures
MeMT	Isoform MT-10-IV of the mussel (bivalve mollusc) Mytilus edulis.^33 It has a high sequence similarity with the vertebrate MTs.		73	21	0	Zn-suppl. cultures
SpMTA	Isoform MTA of the sea urchin (echinoderm) Strongylocentrotus purpuratus.^27 Its Cys residues are arranged in patterns similar to those of mammalian MT1.		64	20	1 Phe	Zn-suppl. cultures
MT1	Isoform 1 of the mammal Mus musculus (mouse, vertebrate).^27,34 Vertebrate MTs are considered the paradigm of MTs. It forms two metallic clusters of 4 and 3 divalent metals.		61	20	0	Zn-suppl. cultures
MtnB	Isoform of the fly (diptera) Drosophila melanogaster.^35 It has an exceptionally short sequence		43	12	0	Zn-suppl. cultures
QsMT	Isoform 2 plant MT from Quercus suber. It contains a His residue in its Cys-devoid spacer region.^36		77	14	1 His 2 Phe	Zn- and Cd-suppl. cultures
TpyMT1	Isoform MT1 of Tetraymena pyriformis (ciliate protozoa). It is considerably divergent from the vertebrate paradigm.^37		107	31	0	Zn- and Cd-suppl. cultures

---

The first step to characterise the M^II-MTs complexes is to quantify the metal content. The considered samples contained variable amounts of metal and sulfide ions, quantitatively evaluated by acid Inductively-Coupled Plasma Atomic Emission Spectroscopy (acid ICP-AES) and by gas chromatography (GC-FPD), respectively (Table 2).

Table 2 Quantification of the metal and sulfide content of the corresponding recombinant M^II-MT preparations. The M^II/MT and S²⁻/MT molar ratios were calculated by acid ICP-AES and GC-FPD measurements, respectively.

Samples	M^II/MT Molar Ratio^a	Sulfide (S^2−)/MT Molar Ratio
a An error of ±0.1 has to be considered. b Ref. 38; c Ref. 37; d Ref. 17.
Zn-MtnB	4.2	0.8^b
Zn-SpMTA	5.8	0.7^b
Zn-MeMT	7.6	1.0^b
Zn-MT1	7.1	0.8^b
Zn-CeMT2	5.6	0.6^b
Zn-TpyMT	9.0	1.1^c
Cd-TpyMT	10.7	1.2^c
Zn-QsMT	3.5	1.0^d
Cd-QsMT	5.6	3.0^d

---

To obtain samples suitable for the Raman analysis avoiding the spectral interference of the buffer, a dialysis-lyophilization protocol has been implemented prior to spectroscopic measures,^17,38 analogously to what has been done by other authors that have studied MTs.³⁹Lyophilization is frequently used by biochemists in order to store proteins in a powder form instead of keeping them in solution. However, it is known that some proteins subjected to this process retain full biological activities, whereas others lose activity. Recent advances in X-ray crystallography enable identification of the three-dimensional structure of proteins in crystals but one always wonders whether the structure of a molecule in the solid-state is identical to the one in aqueous solution. Raman spectroscopy enables study of proteins also in aqueous solution due to the low intensity of the O–H vibration in Raman spectra. However, very concentrated solutions must be used (at least 10⁻² M) to study aqueous solutions of biological molecules, since the sensitivity of the method is not high when excitation in the NIR region is used. However, this excitation is necessary to alleviate undesired backgrounds due to either fluorescent parts of the macromolecules or to unavoidable impurities.

With regards to our M^II-MTs complex, we checked the effect of the lyophilization treatment on one MT (QsMT, from plant) by comparing the Raman spectra obtained in powder form and aqueous solution. Since significant differences were not found under our conditions and the availability of the heterologously synthetized samples was restricted, we chose to perform the measurements on lyophilised samples for obtaining a reasonable signal to noise ratio. In addition, in all cases the maintenance of the global fold of M^II-MTs after lyophilization was verified by checking their mean metal content and their chiroptical features by acid ICP-AES and circular dichroism (CD) measurements (after resolubilisation of the lyophilized samples). Changes in these parameters would have indicated degradation of the samples during the lyophilization process. In conclusion, this procedure can be used without sensitively affecting the MT folding.

2.1 Metal coordination sites

MTs generally coordinate metal ions exclusively via the thiolates of cysteine residues. For examples in rabbit MT1 the entire metal content is distributed into two clusters consisting of four and three metal ions⁴⁰ and each metal ions is coordinated tetrahedrally to Cys.⁴¹

However, first indications of potentially metal-coordinating histidine residues in MTs emerged in the 80s,^42–44 although the small amounts of native protein that could be isolated precluded detailed studies at that time. In the past two decades, a plethora of MT sequences have become available, and it is now clear that histidine residues are a frequent occurrence in MTs from diverse species.^15,16

Recently, it has been shown that also exogenous ligands such as inorganic ions (i.e.sulfide or chloride ions) can participate in the coordination sphere of metals in MTs.^{20,21,35,45,46}

This section will be focused on the information that Raman spectroscopy can give about the metal-coordination sites of MTs.

Cysteine . Cys residues contain the thiol groups (–SH) which can coordinate a wide diversity of metals in its deprotonated form (–S⁻). The binding of Cys residues to a metal ion can generate a metallic cluster with diverse coordination geometries depending on the bound metal. In the constitution of such complexes, the sulfur atom of one Cys can bind to one or two metals: in the first case, Cys is indicated as terminal, whereas in the second as bridging. It has been extensively described that both kinds of M–Cys bonds are present in M^II-MT complexes.⁴⁷

The oxidation of Cys (R–SH) can result in the liberation of the bound metal and the formation of Cystine (R–S–S–R).⁴⁸ Thus, in MTs Cys can be present in three main states: free, coordinated to a metal, or oxidised with the setting up of a disulfide bridge (Cystine). All three states of Cys can be evidenced by Raman spectroscopy in the spectral region typical of sulfur-containing moiety vibrations.⁴⁹

With regards to free Cys, their presence can be evidenced by the –SH stretching band appearing in the 2500–2700 cm⁻¹ Raman region, where no other group displays bands; the position of this band is also indicative of the presence and strength of hydrogen bonds.⁵⁰ Since this band has not been found in all the examined MTs, it can be concluded that these M^II-MT species do not contain free Cys residues, as has already been observed for some MTs.^17,37

The 500–550 cm⁻¹ Raman region yields information on disulfide bridges, allowing a qualitative evaluation of the overall MT oxidation degree (Fig. 1). In all the MTs considered in this review only a weak contribution of the disulfide vibration was found in the Raman spectra, indicating that almost all the cysteinic sulfurs are involved in metal coordination. However, the peak heights of the disulfide bands obtained after normalizing the spectra on the peak intensity of the Raman band at about 1450 cm⁻¹ (CH₂ in-plane deformation) showed that in some cases the presence of cystine was not negligible.^37,38 For example, the higher intensity of the Raman S–S bands (at 512 and 523 cm⁻¹) of Zn^II-SpMTA in relation to those of Zn^II-MT1 (Fig. 1d and b, respectively), having the identical number of Cys residues in the sequence, informs about the larger amount of cystine in the former. This result was in concordance with the lower mean metal content of the SpMTA vs. the MT1 preparation (5.8 and 7.1 Zn/MT, respectively—see Table 2).³⁸

Fig. 1 Raman Spectra of (a) human serum albumin , (b) mammalian Zn-MT1, (c) mollusc Zn-MeMT, and (d) echinoderm Zn–SpMTA in the 800–250 cm⁻¹ region where the bands due to the stretching vibrations of disulfide bridges (νS–S and νC–S) and Zn–Cys bonds (νZn–S and νC–S–Zn) are detectable.

Not all the disulfide bonds detected in the samples prefer to take the most common conformation having the lowest potential energy, that is, gauche–gauche–gauche conformation. For example, the doublet at 512 and 523 cm⁻¹, visible in the Raman spectrum of Zn^II-SpMTA (Fig. 1d), suggested the presence of at least two different configurations of the C_β–S–S–C_β′ disulfides bridges (gauche–gauche–gauche and gauche–gauche–trans, respectively). The S–S conformation mixture was also confirmed by the broad bands in the 700-660 cm⁻¹ region, due to νC–S modes of cystine (Fig. 1).

As regards the metal–thiolate bonds, the involvement of cysteinic sulfur in metal coordination can be visualized in the <500 cm⁻¹ Raman spectral region through several bands attributable to the metal-S stretching modes (Fig. 1).^51–55 In fact, these bands, typical of the M–Cys clusters, are not present in the Raman spectra of free metal-bound proteins such as human serum albumin (for example, Fig. 1a).

In addition, the formation of metal clusters with different geometries,^56,57 leads to a greater number of M–S stretching bands visible, as well as their broadening (Fig. 1), in agreement with the coexistence of different M^II-MT species in the samples, evidenced by other techniques such as Electrospray Ionization Mass Spectrometry. In fact, the formation of more than one species (i.e. Zn₄MtnB, Zn₄S₁MtnB, and Zn₄S₂MtnB) was a constant for all the MTs assayed,³⁸ referring both to species of different M^II–to-MT stoichiometries and also to the coexistence of sulfide-containing and sulfide-devoid complexes.

Contributions of the bridging and terminal sulfur ligands to the MT Raman spectra have been qualitatively identified:¹⁷ the highest wavenumber bands (395–430 cm⁻¹) are essentially due to metal-S bridging vibrations, whereas the lowest wavenumber modes (250–370 cm⁻¹) are contributed by both S-terminal and S-bridging ligands (Fig. 1).^17,38 These assignments were done on the basis of many similarities between the sulfur-related Raman region of MTs and some ferredoxins. Ferredoxins are iron-sulfur proteins having sulfur-atom bridging ligands. So far three well-defined types of [Fe–S] centres in such proteins have been found. One is a binuclear cluster, as shown in Fig. 2A; this represents the [Fe–S] clusters of i.e. plant-type ferrodoxins. The second type of [Fe–S] cluster is found in rebredoxins and is a tetrahedron shape. The third type can be found in bacterial ferredoxin and it has a cubane-type shape in which the iron and sulfur atoms occupy alternating corners (Fig. 2A). Ferredoxins with binuclear clusters show Raman bands at 270, 302 and 343 cm⁻¹ due to Fe–S_bridging stretching and at 378 and 417 cm⁻¹ related to Fe–S_terminal stretching.^58,59 In addition, the presence of bands at fixed frequencies, i.e. at about 290 cm⁻¹, indicates a cubane-type cluster,⁶⁰ whereas binuclear centres are indicated by Raman bands at 282, 327, 340, 367, 395, and 426 cm⁻¹ in [2Fe–2S] ferredoxin.⁵⁶ On the basis of the similarity of the latter bands with the Raman bands of the Zn- and Cd-QsMT spectra, the formation of a binuclear Zn–S centre as well as a cubane-type cluster for Cd^II ions was suggested in these MT complexes (Fig. 2B).¹⁷

Fig. 2 Structure of the metallic clusters present in (A) iron-sulfur proteins: binuclear and cubane-type clusters present in [2Fe–2S] and [4Fe–4S] ferrodoxins. (B) Proposed clusters for plant MTs (Zn-QsMT and Cd-QsMT), in addition to the tetrahedral metal sites.

Also the C–S stretching bands at ca. 765 and ca. 780 cm⁻¹, arising from M–Cys bonds and the backbone vibrations (amide V),⁶¹ are specific of metal-MT complexes (Fig. 1). These bands have been found at 767 and 784 cm⁻¹ in M^II-MT1 complex, closely corresponding to the predicted bands for a type III β-turn (or one turn of a 3₁₀-helix).⁶² These two components are peculiar of metal-bound MTs and become one band ca. 750 cm⁻¹ in metal-free MT form (apo).⁶¹

This band was visible in all the MTs assayed and is very useful as a marker of Cys involvement in metal binding.

Histidine . The imidazole ring of His residue can exist in two tautomeric forms: the N_τ-protonated (or N_ε–H, tautomer I) and N_π-protonated (N_δ–H, tautomer II) (Fig. 3).

Fig. 3 The tautomeric equilibrium of the imidazole moiety of His and the metal-coordinated tautomeric forms. M = metal.

In metalloproteinsHis is by far the most common metal-binding residue, whereas in MTs its involvement in metal binding has been found quite recently.^42–44 It prefers to bind through N_ε rather than N_δ. It is rather surprising since in free His at neutral pH, the N_ε is protonated and N_δ is deprotonated (pK(Im) = 6.1). However, both N atoms in the imidazole ring are almost equivalent (as a result of its aromaticity) but the presence of metal may change the above protonation equilibrium. N_ε is sterically more accessible and upon approaching of the metal, the hydrogen may moves from N_ε to N_δ.

Binding of a metal ion to one of the two N atoms of the imidazole ring limits the protonation site to the other N atom (Fig. 3); thus, the existence of the tautomeric equilibrium makes it difficult to predict the coordination mode of the imidazole moiety to the metal.

Raman spectroscopy is an useful tool for analysing the tautomeric equilibrium since some His bands change in wavenumber depending on whether the imidazole ring takes the tautomeric form I or II. Generally, the attention is focused on the His marker band due to the C₄=C₅ stretching mode. If the aromatic residues are absent or present in a low percentage in the protein sequence, as happens in MTs, it is possible to identify this weak C₄=C₅ stretching band, whose frequency is strongly dependent by the tautomeric form of His and its involvement in metal ion chelation.^53,63–68

Free His gives rise to a band at about 1570 or 1585 cm⁻¹ depending on the tautomeric form adopted by its imidazole ring (I or II, respectively) and the binding of metal ion pushes up the C=C stretching wavenumbers by 10–20 cm⁻¹.^63–67,69 In addition, this marker band shifts towards lower wavenumber (at about 1555 cm⁻¹) when the anionic form of the His imidazole ring binds two different metal ions, making a metal-Im⁻-metal bridge.

Among the analysed MTs, plant QsMT and C. elegans CeMT2 polypeptides contain one His residue and, respectively, two Phe or one Tyr (Table 1). Thus, the eventual His involvement in the metal binding in Zn-CeMT, Zn-QsMT and Cd-QsMT complexes was evaluated by carrying out the curve fitting analysis of the 1630–1565 cm⁻¹ spectral range that allows to distinguish also the contribution of overlapped weak bands generated by His and Tyr (or Phe) residues (Fig. 4).^38,70

Fig. 4 Raman Spectrum of Zn-CeMT2 in the 1710–1500 cm⁻¹ range and curve fitting analysis of the Raman region where the bands sensitive to free and metal-bound His are present. The components due to free His and M^II-His (N_t-M) are visible at about 1570 and 1600 cm⁻¹, respectively.

The Raman spectra exhibited up to six components, two or three attributable to Tyr or Phe residues and the others to coordinated or free His (Table 3). The component at 1573 ± 1 cm⁻¹ is due to free His as tautomer I, that is the most stable form, whereas the other two components at 1579 and 1597 ± 1 cm⁻¹ to the coordinated His as tautomer II and I, respectively. By considering the integrated intensity of these bands, it was concluded that His residues are almost fully coordinated through the N_τ-imidazole nitrogen to Cd^II ions in QsMT (95%) and Zn^II ions in CeMT2 (≈80%), whereas it does not take part to the metal chelation in Zn-QsMT.^38,70 In fact, in the latter His is mainly present as free tautomer I (90%). These conclusions were in concordance with CD data and the findings reported for the participation of His tautomer II (N_ε–M) in metal binding in metalloproteins such as hemoglobin and Clostridium pasteurianumiron hydrogenases.^71,72

Table 3 Wavenumbers (cm⁻¹) and assignments of the components found in the 1625–1565 cm⁻¹ Raman spectral range after fitting procedure. The percentages of the different tautomeric forms of His are reported in parenthesis

Samples	Wavenumber (cm^−1)
	Free taut I^a	Coord taut I	Phe or Tyr	Coord taut II^a	Phe or Tyr	Tyr
a I and II standing for His tautomer I (Nτ–H) and II (Nπ–H).
Zn-CeMT2	1572 (7%)	1579 (5%)	1587	1596 (88%)	1607	1616
Cd-QsMT	1573 (5%)	—	1588	1598 (95%)	1608	—
Zn-QsMT	1574 (90%)	—	1589	1596 (10%)	1607	—

---

Non-protein metal ligands: sulfide and chloride ions. Non-proteic ligands, such as sulfide and chloride anions, contribute in some cases to the structure and stability of the metal-MT species, although this contribution may substantially differ depending on the MT polypeptide and the metal ions considered.^{20,21,35,45,46}

Sulfide ions (S²⁻) play key functional roles in metalloenzymes like ferredoxins characterised by the presence of polymetallic systems (iron-sulfur clusters) containing iron ions with variable oxidation state.⁷³ Biological [Fe–S] clusters are characterized by the presence of multiple iron ions bridged by sulfide ions and coordinated to the protein, generally viaCys residues (see above). The most common types of [Fe–S] clusters, found in the widest variety of proteins and enzymes, are the [2Fe–2S] and [4Fe–4S] clusters, which contain the indicated number of iron and S²⁻ ions and are typically bound to the protein by four Cys (Fig. 2A). These clusters have diverse roles in biology, acting as catalytic centers, structural elements, and sensors in regulating gene expression.⁷⁴

For a long time the functional role of S²⁻ ions was considered to be limited to their enzymatic redox properties; however, the discovering of crystallites changed this vision. Crystallites are semi-crystal particles constituted by metal ions and phytochelatins, enzymatically-synthesized polymers of Glu and Cys γ-linked to a terminal Gly residue, produced in higher plants and some fungi upon exposure to heavy metals.⁷⁵ These structures typically include acid-labile S²⁻ ligands, that contribute to the formation of crystallites and can increase the metal detoxification potential of the MT complexes.^76–78

MTs are proteins similar to phytochelatins in many ways, including the high number of Cys residues in the protein and the fact that both are responsible for the detoxification of heavy metals. Only recently the presence of sulfide ions has been discovered also in in vivo-synthesized M^II-MT species,^20,46,79 probably since most of the data referring to MT structure available to date comes from non-biological synthesis of M^II-MT complexes.⁸⁰ The presence of the acid-labile S²⁻ ligands has been determined both qualitatively and quantitatively by analytical (GC-FPD), spectroscopic (CD), and spectrometric (ESI-MS) techniques and the features of the recovered Zn^II- and Cd^II-MT complexes correlate well with those reported for plants and yeast phytochelatins, therefore bridging the behaviour gap between both types of metal-binding molecules.

Acid-labile S²⁻ ions were detected in all the recombinant M^II-MT preparations analyzed in this review; their amount depends on MT and the coordinated metal (Table 2) but, generally, their presence does not increase the chelating potential of MTs.

The involvement of S²⁻ ions in metal coordination can be evidenced by Raman spectroscopy since the M–S_b-M bond vibrations (S_b standing for bridging sulfur), gives rise to a broad band in the 400–440 cm⁻¹ Raman region (Fig. 1 and inset of Fig. 5).^17,38

Fig. 5 Linear correlation between the sulfide/zinc molar ratio and the sum of the intensity of the Raman bands in the 410–430 cm⁻¹ region. The spectra were normalised on the ∼1450 cm⁻¹ band (CH def.), which is not sensitively affected by structural changes. Inset: Raman spectral region of five Zn^II-MT complexes where the metal-S²⁻ stretching vibrations give a relevant contribution.

In addition, the intensity of this band reflects the content of S²⁻ ions present in the sample. In fact, by plotting the sum of the peak intensity of two main band components (at ≈418 and ≈425 cm⁻¹) as a function of the sulfide/zinc molar ratio, found by flame photometric detector gas chromatography (GC-FPD),²⁰ a very good linear correlation was found for several Zn^II-MTs complexes (Fig. 5).³⁸ Although the presence of both bridging sulfide anions and bridging cysteines has not allowed the definitive assignments of the band components, the linear relationship between the Raman intensity of the two components and the relative S²⁻ content, clarifies the relevance of the contribution of the metal-S²⁻ vibrations to these bands. Consequently, this Raman region can be considered as a qualitative and semi-quantitative marker of the presence of S²⁻ ligands in metal-MT complexes.

With regards to chloride anions, the MT capacity of establishing this association, or not, may be of crucial biological relevance since chloride (Cl⁻) has been related to ATP-MT1 interaction.⁸¹NMR studies on mammalian metallothionein MT2 showed that Cl⁻ ions are able to participate to metal binding.⁸¹ This evidence has been later confirmed also by CD studies on MTs from three different families^20,35 In these complexes, particular CD absorptions partially due to Cl⁻ has been observed; however, these absorptions resulted not to be always indicative of the Cl⁻ participation, and strictly, the Cl⁻ ions were not detectable by mass spectrometry analysis.^21,35 In the case of Zn^II-CeMT2, for example, a positive signal visible at ≈230 nm could be due both to the Cl⁻–M interaction and to the His participation in metal-coordination.

Raman spectroscopy can be an helpful tool also for evaluating the MT capacity to involve these anions in metal coordination. In fact, the Zn–Cl⁻ stretching vibration gives rise to a band at about 290 cm⁻¹, that was visible in two of the Zn-MT preparations analysed, Zn^II-MtnB and Zn^II-MeMT (Fig. 1), and at about 220 cm⁻¹ in Cd^II-TpyMT, thus supporting the participation of this non-proteic ligand to the stabilisation of the metal clusters.^37,38 These Cl⁻ ions probably replace H₂O or OH⁻ in the M^II coordination sphere. Interestingly, Cl⁻ participation in Zn(II)-MT aggregates is not detectable by mass spectrometry³⁵ and until now, information about their participation in metal-MT complexes has only been achieved by CD analysis, thus enlarging the possibilities of the use of Raman spectroscopy for this purpose. Unfortunately, it is not possible to obtain unquestionable information about Cl⁻ ligands if there is interference in this region (≈295 cm⁻¹) from other bands such as the band produced by the Zn–N (imidazole) stretching vibration (i.e. in Zn^II-CeMT2).⁸²

2.2 Side chain amino acid residues

Aromatic amino acids . Almost all the MTs considered are completely devoid of aromatic residues. Only three MTs contains one or two aromatic residues: Phe residue is present in the sequence of SpMTA and QsMT, Tyr in CeMT2, and His in CeMT2 and QsMT (Table 1). In the Raman spectra of these MTs the typical bands of these aromatic residues can be observed. For example, the aromatic ring of Phe gives rise to characteristic bands as that observed at ∼1005 cm⁻¹ (ring breathing), always quite intense (Fig. 6);⁷⁰ this band is indicative of the presence of Phe residues but it is not sensitive to the amino acid environment. Other bands typical of aromatic amino acids lie in the 1620–1570 cm⁻¹ region (see Table 3) and also at lower wavenumbers (Fig. 6).^38,70

Fig. 6 Raman spectrum of the echinoderm Zn–SpMTA metallothionein in the 1740–250 cm⁻¹ range. ν = stretching vibration; def. = in-plane deformation vibration.

As regards Tyr residues, the phenolic moiety generally exhibits an intense Raman doublet in the spectral interval 820–860 cm⁻¹, whose intensity ratio is known as a good indicator of hydrogen bonding degree of the phenoxyl –OH group.^31,83 As an example, the doublet due to Tyr residues of human serum albumin is shown in Fig. 7A. In the Raman spectrum of Zn^II-CeMT2 the Tyr residue, present in the central region of the sequence, gave rise to an anomalous singlet (≈860 cm⁻¹) rather than a doublet in this spectral region (Fig. 7B).³⁸ This finding indicates the non Hydrogen-bonded state of the Tyr phenoxyl group,⁸⁴ which is consistent with the location of Tyr within a highly hydrophobic, tightly packed region. This result is relevant in elucidation of the CeMT2 tertiary structure. In fact, it has been assumed that CeMT2 folds into a bi-dominial structure with 9 Cys residues in each domain when it binds divalent ions.⁸⁵ Thus, the Tyr residue, situated at the edge of one of the two putative domains, would anyhow be buried in a folded metal cluster, rather than solvent-exposed in the interdominial hinge.^84,85

Fig. 7 Raman spectra of (A) human serum albumin (HSA), chosen as example of protein containing Tyr residues, and (B) Zn-CeMT2. Tyr generally gives rise to the 860–830 cm⁻¹ doublet, marker of H-bonding state of the –OH group. This does not happen in the case of Zn-CeMT2, where an anomalous singlet is visible.

With regards to His, the marker bands of this residues has been examined in the above section.

Carboxylate groups. Carboxylate groups from Asp or Glu side chains or carboxy-terminal edges of proteins can take part in metal coordination^86,87 and play key roles as supporting ligands in a diverse array of metalloprotein active sites.⁸⁸ Various roles in metal binding and selectivity of metalloproteins have been assigned to Asp/Glu side chains, based mainly on their charge rather than on their identity. For example, due to the negative charge of the carboxylate group, interactions with the metal cation in a buried protein cavity are not only thermodynamically favourable but also generally more favourable than those of other neutral ligands; thus, Asp/Glu of MTs are thought to be mainly responsible for sequestering the metal cation from physiological fluids.

Acidic residues are quite abundant in MT sequences and are often situated near to metal-coordinating Cys although COO⁻ groups seem not to play a role in metal coordination in MTs. In fact, carboxylate ligands are usually involved in Fe coordination, but they are quite unusual in Zn or Cd coordination.

The carboxyl group exhibits the COO⁻ symmetric stretch vibration in the 1400–1430 cm⁻¹ region, giving rise to bands of medium intensity in comparison with those due to the CH deformations (at about 1450 cm⁻¹) (Fig. 6 and 7).^69,89 By analysing the second derivative spectra of the MTs considered, in order to better distinguish also weak adjacent peaks, two or three components, attributable to the COO⁻ vibrations, were detected in the spectra of Zn^II-SpMTA, Zn^II-CeMT2, Zn^II-MeMT, and Cd^II-TpyMT1. In particular, the presence of the lower wavenumber component at about 1405 cm⁻¹ suggests the formation of stronger bonds in these species. Thus, the COO⁻ groups of MT lateral chains play a role in the MT structure stabilization depending on the metal complex. This participation is not a structural feature restricted to one MT family, but a more general behaviour of MTs, since the spectral results concerning the carboxylate groups were obtained for MTs from organisms of different MT-families.

2.3 Secondary structure analysis

Although secondary structure elements, such as beta-sheet, beta-turn or alpha helix, have a high functional importance in metalloproteins like zinc fingers,⁵³ in MTs for a long time it has been generally assumed that secondary structure elements are poor or functionally insignificant.^90,91 Most of the literature reviews dealing with MT structure were focused mainly on the primary and tertiary structure, conceding scarce interest to the secondary structure. Nowadays, some authors claim for a crucial participation of secondary structure elements in the functional properties of MTs^92–94 and proofs of their existence and functional importance in MTs are growing. The few literature data indicate in many cases the importance of secondary structure elements at several levels for MT functionality.^{15,27,95–99} In prokaryote SmtA and mammalian MT3 metallothioneins, the MT regions presenting secondary structure elements play key roles in the physiological functionality.^15,23,97,100

Additionally, it has been recently demonstrated that in plant QsMT metallothionein, the loss of a Cys-devoid region-with an attributed beta-sheet conformation¹⁷—implies a decrease of its in vivo Cd- and Cu-detoxification abilities.^36,46

Raman spectroscopy can shed light on this aspect of MTs structure since it can quantify the contribution of one distinct secondary structure motifs to the overall protein structure, although it has been seldom used on MTs.

The first Raman results were obtained on rabbit liver MT-1 in its metal-free and two of its metal-bound form (Zn^II and Cd^II).⁶¹ They unambiguously evidenced the presence of β-turns in metal-bound MT-1; in contrast, the apo-protein has a predominantly unordered conformation. This result was confirmed successively by studying the same MT in the Tris HCl buffer solution by Raman and IR spectroscopy.¹⁰¹ Also the secondary structure of porcine brain Cu₄Zn₃-MT3 and Cd₅Zn₂MT1, and Cd₅Zn₂ MT2 were investigated by Raman. These M^II-MTs resulted to contain prevalently β-turns and random coil with only about 10% of β-extended chain.³⁹

The best studied Raman band of proteins is the amide I band which is centered between 1670 and 1674 cm⁻¹ in the Raman spectra of all the M^II-MT complexes analysed (for example Fig. 6 and 8).^17,37,38 This band arises from the stretch vibration of the peptide C=O group; since the C=O group is differently involved in secondary structure elements via hydrogen bonds to the peptide NH group, the band position due to the peptide backbone vibrations is correlated with different protein conformations. Thus, the amide I band envelopes a multitude of single bands at different wavenumbers which can be resolved in many components attributable to different secondary structures (Table 4).^102–104

Fig. 8 Raman spectra of Zn-TpyMT1 (a) and Cd-TpyMT1 (b) in the 1730–1150 cm⁻¹ range.

Table 4 Wavenumbers of the principal Raman bands characteristic of secondary protein structure.

Conformation	Raman bands/cm^−1
	Amide I	Amide III	Other bands
alpha-helix	1650–1657	1270–1290	890–945
beta-sheets	1668–1675	1227–1240	1020–1060
beta-turns	1639–1649		725–765
		1290–1320	595 (type I)
	1673–1700		644 (type II)
			670 (type III)
Random	1660–1665	1240–1256	—

---

Important for the determination of the secondary structure of proteins is also the amide III band, in the 1240–1300 cm⁻¹ range, resulting from coupled C–N stretching and N–H bending motions Raman active (Fig. 8 and Table 4). Besides the amide I and III bands, other bands can be used to further confirm the presence of secondary structure motifs.^102,105 The 890–945 cm⁻¹ band is characteristic of α-helical conformation, the 1020–1060 cm⁻¹ band appears in the presence of β-sheets, and 725–770 cm⁻¹ as well as 595–670 cm⁻¹ are visible if different types of β-turns are formed.^49,106

The evaluations of secondary structures from Raman amide I band are usually good for the proteins containing mostly α-helix and β-sheet structures but no turns.^107,108 Due to the overlap of the amide components, especially when turns and other structures could scatter in the same region or nearby region, it is judicious to analyse not only the amide I but also other modes. Since the secondary structure predictions,¹⁰⁹ X-ray¹¹⁰ and NMR^27,111 studies on MTs have reported the large presence of β-turns, in this case we have token advantage of this possibility.

According with the theoretical predictions reported for the normal modes of β-turns,¹⁰⁶ intense amide III bands above 1290 cm⁻¹ have been observed in all the M^II-MT spectra (as example Fig. 8), as well as component bands at about 765, 745 and 725 cm⁻¹ (in the second derivative spectra), which closely correspond to the predicted vibrational modes (Amide V) of β-turn peptides.⁶²

A contribution from non-ordered (random) conformations to the overall folding of the M^II-MT complexes was also deduced by the appearance of a broad amide III component at about 1250 cm⁻¹ in all the spectra (Fig. 8), although β-sheet structures can also contribute, since generally give rise to one band in the near spectral region (Table 4).⁴⁹ Finally, an eventual contribution from α-helical conformation was considered when a weak skeletal stretching mode at ≈940 cm⁻¹ appeared,⁴⁹ in particular in the Raman spectra of SpMTA and MtnB (Fig. 6). On the contrary, in the case of Zn-MT1 the contribution from α-helix was excluded on the basis of previous studies.^61,62

Taking into account the previous considerations, the semi-quantification of the secondary structure was generally performed by using the curve fitting analysis of the Raman amide I bands. This method is able to estimate secondary structure composition of proteins from the relative area of the individual components assigned to specific structural entity. An important assumption implicit in such an approach is that the effective intrinsic intensities of the bands corresponding to different conformations are very similar. As a consequence, the curve fitting procedure cannot give the absolute content of a structure. However, it can be useful to probe differences in protein folding or to evaluate conformational changes induced by external factors such as metal binding.

A realistic identification of the peak component number and position was carried out by the fourth-derivative spectra and these band positions were used as the initial guess for the curve fitting of the original spectra.¹⁰⁸ Thus, every component of our Raman spectra was assigned to one secondary structure conformation on the basis both of previous reports and the supporting evidence from other bands sensitive to the protein structure (see above).³⁸

The Raman spectra of the analyzed M^II-MTs exhibited a maximum of nine components resulting from the contribution of the different secondary structure elements (Table 5).³⁸ The fitted spectra had R² > 0.999 and a root mean square (RMS) of 0.00001; the associated error to the percentages was of 2–4%. It's worth pointing out that 3₁₀-helicies, reported to be present in some MTs,^27,112 are repeats of one of the three β-turn type (type III). Thus, we have included the contribution from this conformation in the β-turn content.

Table 5 Wavenumbers (cm⁻¹) and assignments of the Amide I Raman band components obtained after the fitting procedure of some M^II-MT complexes. The secondary structure percentages were calculated from the individually assigned bands and their fraction of the total intensity

M^II-MT	Wavenumbers	Assignments	Contents (%)
a The components at 1657 and 1671 cm^−1 were assigned to β-turns on the basis of previous studies which have excluded the presence of α-helices and β-sheets in this MT.^61,109
Zn-CeMT2	1669, 1676	β-sheets	≤9
	1642, 1648, 1676, 1685, 1693	β-turns	66–75
	1661	Random	25
Zn-MtnB	1655	α-helix	4
	1668, 1675	β-sheets	≤2
	1639, 1649, 1675, 1681, 1689, 1695	β-turns	77–79
	1660	Random	17
Zn-MeMT	1670,1677	β-sheets	≤13
	1637, 1650, 1677, 1685, 1693	β-turns	68–81
	1662	Random	19
Zn-MT1 ^a	1642, 1657, 1671, 1679, 1686, 1694	β-turns	75
	1661	Random	25
Zn-SpMTA	1654	α-helix	5
	1667,1675	β-sheets	≤6
	1640, 1647, 1675, 1682, 1688, 1696	β-turns	72–78
	1661	Random	17
Zn-TpyMT1	1655	α-helix	4
	1670,1674	β-sheets	30
	1639, 1648, 1685, 1692	β-turns	44
	1662	Random	22
Cd-TpyMT1	1655	α-helix	9
	1668,1673	β-sheets	22
	1640, 1647, 1679, 1685, 1693	β-turns	50
	1661	Random	19
Zn-QsMT	1654	α-helix	1
	1670	β-sheets	61
	1640, 1648, 1685, 1692, 1698	β-turns	24
	1662	Random	14
Cd-QsMT	1672	β-sheets	64
	1644, 1680, 1688, 1697	β-turns	22
	1661	Random	14

---

The curve fitting results have indicated that the main contribution to the secondary structures of the analysed MT isoforms generally arises from β-turn (up to ∼80%) and unordered segments (Table 5). Elements of regular secondary structure have been mainly found in metal complexes of TpyMT1 and QsMT. In particular, β-sheet structure has been found to contribute in a relevant way to the overall folding of Zn^II- and Cd^II-QsMT.¹⁷ This contribution of β-sheets is fairly rare among MTs. β-sheets are probably located in the spacer region of QsMT, as suggested by the Chou and Fasman prediction method,¹¹³ and it is in concordance with the observation that other secondary structures such as α-helices are mostly restricted to amino acid sequences devoid of Cys, and thus very rare in prokaryotic and animal MTs.^15,23

Thus, the curve fitting results were in agreement with the generally accepted conclusions suggesting a high contribution of β-turns to the secondary structure of the metal-MT complexes and a low contribution from α-helix and extended β-sheet segments.^101,109 In addition, our conclusions on Zn(II)-MT1, consisting largely of β-turns, were in good agreement with those drawn on vibrational and NMR spectra of rabbit liver MTs reconstituted in vitro.^61,112

3. Applications to metallothioneins under radical stress

Thanks to the extreme sensitivity to molecular interactions and structure, Raman spectroscopy can be also useful to prove or refuse models (i.e. in enzyme or receptor mechanisms), to improve models, and to suggest mechanisms. Recently, Raman spectroscopy has been used in our lab for evaluating the effects of radical exposure on sulfur-containing proteins, including MTs. An interesting aspect of MTs is their ability to react with radical species which are continuously generated in cells by physiological processes and pathological conditions. In this contest there is copious evidence for an MTs contribution to stress resistance. In fact, many studies have supported a MT function in control of oxidative stress, i.e. by probing the efficiency of some M^II-MTs in acting as scavengers^113–116 or the dramatical increase in MT expression in response to tissue injury, inflammation, and tumour.^117,118 Thus, the ability of MTs to act as an antioxidant may provide a survival advantage at a time of major infection and inflammation, conditions where a large increase in radical-mediated tissue damages take place.

Although all amino acids are susceptible to modifications caused by free radical exposure, some of them are more sensitive. In particular, aromatic and sulfur-containing residues generally undergo free radical-induced modifications.^119–121

Recently, it has been shown that reducing species like ˙H and e⁻_aq are able to produce a desulfurisation reaction involving sulfur-containing residues.^122–124 The result of this reaction corresponds to a mutation of the natural sequence: the reductive attack of ˙H atom on Met causes the conversion of Met into α-aminobutyric acid (Aba), whereas Cys residues are mutated to Ala.¹²⁵ This desulfurisation is accompanied by the release of diffusible sulfur-centred radicals that can reach membrane bilayer and gives rise to another damage at the level of the fatty acid residues of cellular membranes (Scheme 1). Since MTs contain both sulfur-containing amino acids (Met and Cys), as well as S²⁻ anions, all capable to generate sulfur-centred radicals and promote membrane lipid transformation, the M^II-MTs complexes have been considered a very interesting case to study in the contest of tandem protein-lipid damage.

Scheme 1 Sulfur-containing moieties like Met, Cys or S²⁻ ligands in MTs are modified by attack of H˙ and/or e⁻_aq with the formation of diffusible sulfur-centred radicals, such as CH₃S˙ or S˙⁻, able to migrate into the lipid bilayer; these sulfur-centred species can induce cis–trans isomerization of unsaturated fatty acid residues.

Thus, this section will be focused on the use of Raman spectroscopy for elucidating the effect of free radical exposure on the overall MT structure. In particular, the degradation of two M^II-MT complexes from a plant MT of the cork oak (Quercus suber) has been studied in detail under conditions of reductive radical stress, in order to assess the main sites of the attack and modification in MT folding.^70,126

Free radical generation, that mimics the conditions of an endogenous radical stress, has been obtained by using γ-radiolysis of aqueous solutions, that allows to select the reacting radical species by changing the appropriate conditions of irradiation.^70,121

Ar-flushed aqueous solutions of M^II-QsMT containing 0.2 M tBuOH were irradiated at different doses (under these experimental conditions the only reactive radical species are ˙H and e⁻_aq), lyophilised, and analysed by Raman spectroscopy (Fig. 9).^70,126Among the amino acid residues present in QsMT, Cys resulted to be among the most sensitive residues towards radical attack. In fact, exposure of the two M^II-MT species to reductive stress led to an intensity decrease of the metal–thiolate stretching bands, particularly evident in the Zn-QsMT spectrum (i.e., 310 and 320 cm⁻¹), suggesting a partial de-construction of tetrahedral metal clusters (Fig. 9). Also a slight intensity increase in the bands in the 410–440 cm⁻¹ region indicates a higher number of bridging Cys than in the two native structures.

Fig. 9 Raman spectra of (A) Zn-QsMT and (B) Cd-QsMT, in the 800–240 cm⁻¹ region of Ar-flushed aqueous solutions containing 0.2 M tBuOH (a) before and (b) after 100 Gy irradiation. H˙ and/or e⁻_aq are the radical species formed under these experimental conditions.

Conversely, only in Zn-QsMT the formation of new disulfide bridges took place, as indicated by the significant intensity increase in the S–S stretching bands (520–500 cm⁻¹ region) (Fig. 9A). Thus, a partial oxidation of the protein takes place upon radical stress exposure as a consequence of the partial deconstruction of the Cys-Zn clusters.^70,126 This leads to the release of some chelated Zn(II) ions, as suggested both by the intensity decrease of the 760–780 cm⁻¹ bands, peculiar of the metal-bound MTs, and the intensity increase of the ca. 750 cm⁻¹ component, typical of the apo-MTs (Fig. 9A).⁶¹ Similar behaviour has been observed for rabbit Zn-MT and Zn/Cd-MT undergone to oxidative stress.^127,128

Since Cd(II) is known to be bound more tightly than Zn(II),¹⁰³ therefore it comes off more difficultly during stress conditions, thus explaining both the lesser formation of disulfides in our Cd-containing MT, as well as the absence of the component due to the apo-MT. This result agrees with no founding of free cadmium from rabbit Zn/CdMT under oxidative stress conditions.¹⁰²

The oxidation of some Cys residues induces the necessity of new ligands such as His for stabilizing the Zn-QsMT complex. Indeed, the curve fitting analysis of the 1630–1565 cm⁻¹ region revealed that His is completely involved in the Zn^II binding after 100 Gy irradiation (10% and 100% coordinated His before and after irradiation, respectively). Moreover, the presence of two components at 1580 and 1600 cm⁻¹, due to the metal-bound His, indicates that both tautomeric imidazolic forms take part in the metal binding, although the tautomer II is the preferential one (i.e. ≈ 80%).⁷⁰ Thus, in Zn-QsMT both the formation of new S–S bonds and the involvement of His in metal coordination indicates the occurrence of a significant structural rearrangement as a consequence of the radical attack.

On the contrary, when the protein binds Cd(II) ions the radical attack occurs towards both the Cys-metal clusters and Met residues that are not involved in metal binding. In fact, the 727 cm⁻¹ band due to Met changed its spectral features after a 100 Gy dose exposure (Fig. 9B); on the contrary, only a very weak intensity decrease of the 725 cm⁻¹ band was observed in Zn-QsMT for the same conditions.

The different sensitivity to reductive attack displayed by the Met residues of the two metallated QsMT forms is due to the different folding of the native protein when binding Zn^II or Cd^II.^129,130 Hence, the ˙H atom attack towards Met results partially limited by the polypeptide folding in the Zn-QsMT complexes, and then other moieties become the preferential sites of the attack (i.e., metal thiolate and/or metal-sulfide bonds). Since the reductive radical attack towards Met residues can yield diffusible sulfur radicals able to induce damages in cellular membranes, this result is related to the lower capability of Zn-QsMT to induce cis–trans isomerisation of lipid membrane than Cd-QsMT.¹²⁶

Analogously to Met, Phe residues are significantly attacked by reductive species only when QsMT is bound to Cd^II. In fact, irradiation exposure of the Cd-QsMT complexes induced a splitting of the 625 cm⁻¹ band due to Phe side chain (∼630 and ∼620 cm⁻¹), whereas no significant spectral changes in this band were visible in the Zn-QsMT system (Fig. 9).¹²⁶

With regards to modifications in MT folding, the exposure of the two M^II-QsMT complexes to reductive stress (H˙ and e⁻_aq) lead to slightly different conformational changes depending on the metal bound.⁷⁰ In the case of Zn-QsMT the reductive radical attack increases the β-turn conformation content, whereas in Cd-QsMT it causes a slight decrease in the content of β-sheet structure and an increase in the random coil percentage.

4. Conclusions

Interactions of metals with biomolecules belong to one of the most studied fields in inorganic biochemistry. The areas of application of new knowledge or information may range from medicinal chemistry¹³⁰ through “classical” organometallic chemistry to environment protection (metal-binding biomass ).^131,132

This review explores the possibilities of Raman spectroscopy in providing structural insights, semiquantitative for α-helix and β-sheets, into M^II-MTs systems. In fact, the recent technical devices have increased the achievable signal-to-noise ratio in this vibrational spectroscopy, allowing it to emerge as a helpful tool for affording a detailed picture of local regions of large complexes. Despite the potentialities of this technique, to our knowledge it has been scarcely used in MT studies until now.

Raman has proven to be very useful in identifying metal coordination sites (i.e.Cys and His) of Zn- and Cd-complexed MTs, representative of different MT families, enclosing the well-known vertebrate and echinoderm, and the poorly understood nematode, diptera, molluscan, protozoa and plant MTs. In addition, this technique is useful for evidencing the secondary structure motifs eventually present in MTs, as well as the participation of extra-protein ligands, such as chloride and sulfide ions, in the M^II-MT coordination environment.

In conclusion, Raman spectroscopy, in coupling with analytical techniques, such as acid ICP-AES, GC-FPD, CD, etc., can be a very useful experimental strategy in the research on structure–activity relationships in MTs.

Finally, will Raman spectroscopy become an important tool in many biochemistry laboratories as for example CD now is? Perhaps it will not be as widespread, but it may become an indispensable tool for scientists interested in the chemistry of many classes of small molecule–big molecule interactions. The tide is running in its favour, the cost of instrumentation is falling, experiments are becoming easier to undertake, and the results are becoming easier to interpret.

Acknowledgements

We are grateful to the research groups of Prof. Silvia Atrian, Dr Jordi Domenech (Dep. Genètica, Universitat de Barcelona, Spain), and Prof. Merce’ Capdevila (Dep. Química, Universitat Autònoma de Barcelona, Spain) for the heterologous synthesis of the M^II-MTs aggregates, their purification and their initial characterisation by analytic techniques.

References

1. P. R. Carey, in Biochemical Applications of Raman and Resonance Raman Spectroscopies, Academic Press, New York, 1982.
2. M. Kim, H. Owen and P.R. Carey, Appl. Spectrosc., 1993, 47, 1780–1783.
3. J. Dong, D. Dinakarpandian and P. R. Carey, Appl. Spectrosc., 1998, 52, 1117–1122.
4. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, M. C. Strain, J. C. Burant, E. Stratman, G. A. Petersson, J. A. Montgomery, V. G. Zakrzewski, T. Keith, K. Raghavachari, M. A. Al-Laham, J. V. Ortiz, J. B. Foresman, J. Cioslowski, B. B. Stefanov, A. Nanayakkara, M. Challacombe, C. Y. Peng, P. Y. Ayala, W. Chen, M. W. Wong, J. L. Andres, E. S. Replogle, R. Gomperts, R. L. Martin, D. J. Fox, J. S. Binkley, D. J. Defrees, J. Baker, J. P. Stewart, C. Gonzalez, M. Head-Gordon, P. M. W. Gill, B. G. Johnson and J. A. Pople, GAUSSIAN 95 Development Version (Revision D.2)., Gaussian, Inc., Pittsburgh, PA, 1996.
5. R. Tuma, H. Tsuruta, J. M. Benevides, P. E. Prevelige Jr. and G. J. Thomas Jr., Biochemistry, 2001, 40, 665–674.
6. R. Callender and H. Deng, Annu. Rev. Biophys. Biomol. Struct., 1994, 23, 215–245.
7. J. H. R. Kägi, Methods Enzymol., 1991, 205, 613–626.
8. J. H. R. Kägi, in Metallothionein III, Biological Roles and Implications, ed. K. T. Suzuki, Birkhauser Verlag, Basel, 1993, pp. 29–55.
9. M. Vašák and D. W. Hasler, Curr. Opin. Chem. Biol., 2000, 4, 177–183.
10. M. Vašák, D. W. Hasler and P. Faller, J. Inorg. Biochem., 2000, 79, 7–10.
11. J. Hidalgo, M. Aschner, P. Zatta and M. Vasak, Brain Res. Bull., 2001, 55, 133–145.
12. D. H. Petering, S. Krezoski, P. Chen, A. Pattanaik and C. F. Shaw III, Metallothioneins, ed. M. J. Stillman, C. F. Shaw III and K. T. Suzuki, VCH Publishers, New York, 1992, ch. 7.
13. P. A. Binz and J. H. R. Kägi, http://www.bioc.uzh.ch/mtpage/MT.html, 2001.
14. A. H. Robbins and C. D. Stout, Metallothioneins, ed. M. J. Stillman, C. F. Shaw III and K. T. Suzuki, VCH Publishers, New York, 1992, ch. 3.
15. C. A. Blindauer, M. D. Harrison, J. A. Parkinson, A. K. Robinson, J. S. Cavet, N. J. Robinson and P. J. Sadler, Proc. Natl. Acad. Sci. U. S. A., 2001, 98, 9593–9598.
16. L. Villarreal, L. Tio, M. Capdevila and S. Atrian, FEBS J., 2006, 273, 523–535.
17. J. Domènech, A. Tinti, M. Capdevila, S. Atrian and A. Torreggiani, Biopolymers, 2007, 86, 240–248.
18. O. I. Leszczyszyn, R. Schmid and C. A. Blindauer, Proteins: Struct., Funct., Bioinf., 2007, 68, 922–35.
19. C. A. Blindauer, J. Inorg. Biochem., 2008, 102, 507–21.
20. M. Capdevila, J. Domenech, A. Pagani, L. Tio, L. Villarreal and S. Atrian, Angew. Chem., Int. Ed., 2005, 44, 4618–4622.
21. L. Villarreal, L. Tio, S. Atrian and M. Capdevila, Arch. Biochem. Biophys., 2005, 435, 331–335.
22. B. L. Vallee and K. H. Falchuk, Physiol Rev., 1993, 73, 79–118.
23. H. Wang, Q. Zhang, B. Cai, H. Li, H. K. Sze, Z. X. Huang, H. M. Wu and H. Sun, FEBS Lett., 2006, 580, 795–800.
24. P. Gonzalez-Duarte, in Comprehensive Coordination Chemistry II, ed. J. McCleverty and T. J. Meyer, Elsevier, Amsterdam, 2003, pp. 213–228.
25. Z. Zhu, E. F. DeRose, G. P. Mullen, D. H. Petering and C. F. Shaw III, Biochemistry, 1994, 33, 8858–8865.
26. S. S. Narula, M. Brouwer, Y. Hua and I. M. Armitage, Biochemistry, 1995, 34, 620–631.
27. R. Riek, B. Prêcheur, Y. Wang, E. A. Mackay, G. Wider, P. Güntert, A. Liu, J. H. R. Kägi and K. Wüthrich, J. Mol. Biol., 1999, 291, 417–428.
28. J. Bandekar, Biochim. Biophys. Acta, 1992, 1120, 123–143.
29. T. Miura and G. J. Thomas Jr., in Subcellular Biochemistry, ed. B. B. Biswas and S. Roy, Plenum Press, New York, 1995, p. 55.
30. G. J. Thomas Jr., Biopolymers, 2002, 67, 214–225.
31. R. Tuma, J. Raman Spectrosc., 2005, 36, 307–319.
32. Y. Wang, E. A. Mackay, M. Kurasaki and J. H. R. Kägi, Eur. J. Biochem., 1994, 225, 449–457.
33. E. A. Mackay, J. Overnell, B. Dunbar, I. Davidson, P. E. Hunziker and J. H. R. Kägi, Eur. J. Biochem., 1993, 218, 183–194.
34. N. Cols, N. Romero-Isart, M. Capdevila, B. Oliva, P. Gonzalez-Duarte, R. Gonzalez-Duarte and S. Atrian, J. Inorg. Biochem., 1997, 68, 157–166.
35. J. Domenech, O. Palacios, L. Villarreal, P. Gonzalez-Duarte, M. Capdevila and S. Atrian, FEBS Lett., 2003, 533, 72–78.
36. J. Domènech, G. Mir, G. Huguet, M. Capdevila, M. Molinas and S. Atrian, Biochimie, 2006, 88, 583–593.
37. J. Domènech, R. Bofill, A. Tinti, A. Torreggiani, S. Atrian and M. Capdevila, Biochim. Biophys. Acta, 2008, 1784, 693–704.
38. A. Torreggiani, J. Domènech, S. Atrian, M. Capdevila and A. Tinti, Biopolymers, 2008, 89, 1114–1124.
39. Y. B. Shi, J.-L. Fang, X.-Y. Liu, L. Du and W.-X. Tang, Biopolymers, 2002, 65, 81–88.
40. J. D. Otvos and I. M. Armitage, Proc. Natl. Acad. Sci. U. S. A., 1980, 77, 7094–7098.
41. M. Vašák and J. H. R. Kägi, Met. Ions. Biol. Syst, 1983, 15, 213–273.
42. D. P. Higham, P. J. Sadler and M. D. Scawen, Science, 1984, 225, 1043–1046.
43. B. Lane, R. Kajioka and T. Kennedy, Biochem. Cell Biol., 1987, 65, 1001–1005.
44. R. W. Olafson, W. D. McCubbin and C. M. Kay, Biochem. J, 1988, 251, 691–699.
45. W. Maret, G. Heffron, H. A. Hill, D. Djuricic, L. J. Jiang and B. L. Vallee, Biochemistry, 2002, 41, 1689–1694.
46. J. Domènech, R. Orihuela, G. Mir, M. Molinas, S. Atrian and M. Capdevila, J. Biol. Inorg. Chem., 2007, 12, 867–82.
47. W. F. Furey, A. H. Robbins, L. L. Clancy, D. R. Winge, B. C. Wang and C. D. Stout, Experientia Suppl., 1987, 52, 139–148.
48. M. V. Kumari, M. Hiramatsu and M. Ebadi, Free Radical Res., 1998, 29, 93–101.
49. A. T. Tu, in Spectroscopy of Biological Systems, ed. R. J. H. Clark and R. E. Hester, Wiley, Chichester, 1986, vol. 13, pp. 47–112.
50. S. W. Raso, P. L. Clark, C. Haase-Pettingell, J. King and G. J. Thomas, J. Mol. Biol., 2001, 307, 899–911.
51. S. Han, R. R. Czernuszewicz, T. Kimura, M. W. W. Adams and T. G. Spiro, J. Am. Chem. Soc., 1989, 111, 3505–3511.
52. A. I. Boldyrev and J. Simons, Mol. Phys., 1997, 92, 365–379.
53. T. Miura, T. Satoh and H. Takeuchi, Biochim. Biophys. Acta, 1998, 1384, 171–179.
54. M. A. Quail, P. Jordan, J. M. Grogan, J. N. Butt, M. Lutz, A. J. Thomson, S. C. Andrews and J. R. Guest, Biochem. Biophys. Res. Commun., 1996, 229, 635–642.
55. J. B. Broderick, T. F. Henshaw, K. Wojtuszewski, S. R. Smith, M. R. Trojan, R. McGhan, A. Kopf, M. Kibbey and W. E. Broderick, Biochem. Biophys. Res. Commun., 2000, 269, 451–456.
56. F. J. Rotsaert, J. D. Pikus, B. G. Fox, J. L. Markley and J. Sanders-Loehr, J. Biol. Inorg. Chem., 2003, 8, 318–326.
57. M. Vargek, X. Zhao, Z. Lai, G. L. McLendon and T. G. Spiro, Inorg. Chem., 1999, 38, 1372–1373.
58. S.-P. W. Tang, T. G. Spiro, C. Antanaitis, T. H. Moss, R. H. Holm, T. Herskovitz and L. E. Mortensen, Biochem. Biophys. Res. Commun., 1975, 62, 1–6.
59. J. B. Broderick, T. F. Henshaw, K. Wojtuszewski, S. R. Smith, M. R. Trojan, R. McGhan, A. Kopf, M. Kibbey and W. E. Broderick, Biochem. Biophys. Res. Commun., 2000, 269, 451–456.
60. T. Løver, G. A. Bowmaker, J. M. Seakins and R.P. Cooney, Chem. Mater., 1997, 9, 967–975.
61. J. Pande, C. Pande, D. Gilg, M. Vasak, R. Callender and J. H. R. Kägi, Biochemistry, 1986, 25, 5526–5532.
62. P. Lagant, G. Vergoten, G. Fleury and M. H. Loucheux-Lefebvre, Eur. J. Biochem., 1984, 139, 149–154.
63. A. Torreggiani, A. Degli Esposti, M. Tamba, G. Marconi and G. Fini, J. Raman Spectrosc., 2006, 37, 291–298.
64. H. Takeuchi, Biopolymers, 2003, 72, 305–317.
65. A. Torreggiani, M. Tamba, S. Bonora and G. Fini, Biopolymers, 2003, 72, 290–298.
66. I. Ashikawa and K. Itoh, Biopolymers, 1979, 18, 1859.
67. M. Tanokura, M. Tasumi and T. Myazawa, Biopolymers, 1976, 15, 393.
68. H. Takeuchi, N. Watanabe, Y. Satoh and I. Harada, J. Raman Spectrosc., 1989, 20, 233–237.
69. A. Torreggiani, S. Bonora and G. Fini, Biopolymers, 2000, 57, 352–364.
70. A. Torreggiani, J. Domenech and A. Tinti, J. Raman Spectrosc., 2009, 40, 1687–1693.
71. M. W. Adams, Biochim. Biophys. Acta, Bioenerg., 1990, 1020, 115–145.
72. J. W. Peters, W. N. Lanzilotta, B. J. Lemon and L. C. Seefeldt, Science, 1998, 282, 1853–1858.
73. D. C. Johnson, D. R. Dean, A. D. Smith and M. K. Johnson, Annu. Rev. Biochem., 2005, 74, 247–281.
74. J. B. Broderick, Nat. Chem. Biol., 2007, 3, 243–244.
75. C. S. Cobbett and P. B. Goldsbrough, Annu. Rev. Plant Biol., 2002, 53, 159–182.
76. Y. Hayashi, C. W. Nakagawa and A. Murasugi, Environ. Health Perspect., 1986, 65, 13–19.
77. D. Winge, C. T. Dameron and R. K. Mehra, In: Metallothioneins, ed. M. J. Stillman and C. F. Shaw, 3rd, KTS, New York VCH, 1992, pp. 257–270.
78. J. L. Hall, J. Exp. Bot., 2002, 53, 1–11.
79. R. Bofill, M. Capdevila and S. Atrian, Metallomics, 2009, 1, 229–234.
80. P. González-Duarte, in Comprehensive Coord. Chem II, ed. J. A. McCleverty and T. J. Meyer, Amsterdam, Elsevier-Pergamon, 2003, vol. 8, pp. 213–228.
81. W. Maret, G. Heffron, H. A. O. Hill, D. Djuricic, L.-J. Jiang and B. L. Vallee, Biochemistry, 2002, 41, 1689–1694.
82. K. Hasegawa, T. Ono and T. Nogochi, J. Phys. Chem., 2002, 106, 3377–3390.
83. M. N. Siamwiza, R. C. Lord, M. C. Chen, T. Takamatsu, I. Harada, H. Matsuura and T. Shimanouchi, Biochemistry, 1975, 14, 4870–4876.
84. Z. Arp, D. Autrey, J. Laane, S. A. Overman and G. J. Tomas, Biochemistry, 2001, 40, 2522–2529.
85. C. You, E. A. Mackay, P. M. Gehrig, P. E. Hunziker and J. H. Kagi, Arch. Biochem. Biophys., 1999, 372, 44–52.
86. K. Degtyarenko, Bioinformatics, 2000, 16, 851–864.
87. D. S. Auld, BioMetals, 2001, 14, 271–313.
88. R. H. Holm, P. Kennepohl and E. I. Solomon, Chem. Rev., 1996, 96, 2239–2314.
89. A. Torreggiani, C. Fagnano and G. Fini, J. Raman Spectrosc., 1997, 28, 23–27.
90. P. A. Cobine, R. T. McKay, K. Zangger, C. T. Dameron and I. M. Armitage, Eur. J. Biochem., 2004, 271, 4213–4221.
91. V. Calderone, B. Dolderer, H. J. Hartmann, H. Echner, C. Luchinat, C. Del Bianco, S. Mangani and U. Weser, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 51–56.
92. K. E. Rigby and M. J. Stillman, Biochem. Biophys. Res. Commun., 2004, 325, 1271–1278.
93. L. Vergani, M. Grattarola, C. Borghi, F. Pondero and A. Viarengo, FEBS J., 2005, 272, 6014–6023.
94. K. E. Rigby, J. Chan, J. Mackie and M. J. Stillman, Proteins: Struct., Funct., Bioinf., 2006, 62, 159–172.
95. G. Wagner, M. H. Frey, D. Neuhaus, E. Worgotter, W. Braun, M. Vasak, J. H. Kagi and K. Wuthrich, Experientia Suppl., 1987, 52, 149–157.
96. I. Bertini, H. J. Hartmann, T. Klein, G. Liu, C. Luchinat and U. Weser, Eur. J. Biochem., 2000, 267, 1008–1018.
97. G. Öz, K. Zangger and I. M. Armitage, Biochemistry, 2001, 40, 11433–11441.
98. A. Muñoz, F. H. Forsterling, C. F. Shaw and D. H. Petering, JBIC, J. Biol. Inorg. Chem., 2002, 7, 713–724.
99. R. Scudiero, P. A. Temessi and E. Parisi, Gene, 2005, 345, 21–26.
100. W. J. Cook, S. R. Kar, K. B. Taylor and L. M. Hall, J. Mol. Biol., 1998, 275, 337–346.
101. H. Qing, Z. Baolin, S. Fenling and T. Zuyi, Spectrosc. Lett., 1995, 28, 829–838.
102. A. Tu, in Raman Spectroscopy in Biology. Principles and Application, Wiley-Interscience, New York, 1982, p. 448.
103. F. S. Parker, in Application of Infrared, Raman and Resonance Raman Spectroscopy in Biochemistry, Plenum Press, New York, 1983, p. 550.
104. A. Torreggiani, G. Bottura and G. Fini, J. Raman Spectrosc., 2000, 31, 445–450.
105. A. Torreggiani and G. Fini, J. Raman Spectrosc., 1999, 30, 295–300.
106. S. Krimm and J. Bandekar, Biopolymers, 1980, 19, 1–29.
107. R. W. Williams, Methods Enzymol., 1986, 130, 311–331.
108. A. Torreggiani and G. Fini, J. Raman Spectrosc., 1998, 29, 229–236.
109. Y. Boulanger, C. M. Goodman, C. P. Forte, S. W. Fesik and I. M. Armitage, Proc. Natl. Acad. Sci. U. S. A., 1983, 80, 1501–1505.
110. W. F. Furey, A. H. Robins, L. L. Clancey, D. R. Winge, B. C. Wang and C. D. Stout, Science, 1986, 231, 704–710.
111. W. Braun, G. Wagner, E. Wörgötter, M. Vasák, J. H. R. Kägi and K. Wüthrich, J. Mol. Biol., 1986, 187, 125–129.
112. G. Wagner, D. Neuhaus, E. Wörgötter, M. Vasak, J. H. R. Kägi and K. Wüthrich, J. Mol. Biol., 1986, 187, 131–135.
113. P. Y. Chou and G. D. Fasman, Annu. Rev. Biochem., 1978, 47, 251–276.
114. K. Felix, E. Lengfelder, H. J. Hartmann and U. Weser, Biochim. Biophys Acta, 1993, 1203, 104–108.
115. C. T. Aravindakumar, J. Ceulemans and M. De Ley, Biochem. J., 1999, 344, 253–258.
116. H. J. Hartmann and U. Weser, BioMetals, 2000, 13, 153–156.
117. D. X. Deng, S. Chakrabarti, M. P. Waalkes and M. G. Cherian, Histopathology, 1998, 32, 340–347.
118. G. W. Wang, Z. Zhou, J. B. Klein and Y. J. Kang, Am. J. Physiol. Heart Circ. Physiol., 2001, 280, H2292–H2299.
119. W. M. Garrison, Radiat. Res. Rev., 1972, 3, 305–326.
120. M. J. Davies, S. Fu, H. Wang and R. T. Dean, Free Radical Biol. Med., 1999, 27, 1151–1163.
121. A. Torreggiani, in Handboook of Free Radicals: Formation, Types and Effects, ed. D. Kozyrev and V. Slutsky, NovaScience Publishers Inc, Hauppauge NY, 2009, ch. 11, in press.
122. A. Torreggiani, M. Tamba, I. Manco, M. R. Faraone-Mennella, C. Ferreri and C. Chatgilialoglu, J. Mol. Struct., 2005, 767, 744–747.
123. A. Torreggiani, M. Tamba, I. Manco, M. R. Faraone-Mennella, C. Ferreri and C. Chatgilialoglu, Biopolymers, 2006, 81, 39–50.
124. A. Torreggiani, Spectroscopy, 2008, 22, 279–286.
125. C. Ferreri, C. Chatgilialoglu, A. Torreggiani, A. M. Salzano, G. Tenzone and A. Scaloni, J. Proteome Res., 2008, 7, 2007–2015.
126. A. Torreggiani, J. Domènech, R. Orihuela, C. Ferreri, S. Atrian, M. Capdevila and C. Chatgilialoglu, Chem.–Eur. J., 2009, 15, 6015–6024.
127. L. P. Thomas, G. J. Bachiwski and A. W. Girotti, Biochim. Biophys. Acta, 1986, 884, 448–461.
128. X. Fang, J. Wu and G. Wei, Radiat. Res., 1994, 138, 165–170.
129. G. Mir, J. Domenech, G. Huguet, W. J. Guo, P. Goldsbrough, S. Atrian and M. Molinas, J. Exp. Bot., 2004, 55, 2483–2493.
130. J. Domènech, R. Orihuela, G. Mir, M. Molinas, S. Atrian and M. Capdevila, J. Biol. Inorg. Chem., 2007, 12, 867–882.
131. Handbook of Metal–Ligand Interactions in Biological Fluids, ed. G. Berthon, Marcel Dekker, New York, 1995, vol. 2.
132. C. Sousa, A. Cebolla and V. de Lorenzo, Nat. Biotechnol., 1996, 14, 1017–1020.

---