INTRODUCTION

Iron coordination compounds have been studied extensively because of what they reveal about molecular structure and chemical bonding, as well as because of the unusual chemical nature and specific physical and biological properties [1–5]. Moreover, iron coordination chemistry is playing a major role in the development of new iron chelators for clinical application [6, 7].

Iron is an essential element for humans. It is transported and stored by specific proteins being its absorption/excretion ratio finely regulated by a complex system of specific protein. However, iron is associated with toxic effects: it can damage tissues by catalyzing the conversion of H2O2 into toxic OH• radicals that attack cellular membranes, protein and DNA [8–10]. Fenton reagents and the oxidized products are recognized as important initiators and mediators of cell death in a variety of organisms and pathological situations [11–13]. A proper chelation plays a central role to prevents iron participation to Fenton reaction. The main points are that ferric complexes should not be reduced under biological conditions and that the ferrous complex can’t transfer an electron to a peroxide. In the case of ascorbate and in physiologic conditions, this implies that metal complexes with electrode potentials larger than +0.9 V (H2O2/OH•, H2O, pH 7) or smaller than +0.1 V (ascorbyl/monohydroascorbate, pH 7) cannot catalyze the redox cycling [14].

Chelators in current medicinal use include natural compounds derived from microorganisms (siderophores) and synthetic chelators [15]. Deferoxamine, Deferiprone and Deferasirox are the most important specific U.S.FDA-approved iron chelators. Despite the current availability of different and promising molecules the search for new chelators is crucial to minimize the undesired side effects of the actual drugs [16].

The hydroxypyrone core is present in many natural organic compounds widely investigated for medicinal applications [17, 18]. In this context, we synthetized the new iron chelator inspired to maltol, 3-hydroxy-2-(5'-hydroxypentyl)-4H-pyran-4-one, with the aim to obtain a tridentate chelator which stabilize the ferric state and result in redox-inactive complex suitable for the treatment of Fe-overload diseases.

EXPERIMENTAL

Melting points were measured with a Kofler apparatus and are uncorrected. The 1H and 13C NMR spectra were recorded for solution in CDCl3 with a Bruker AVANCE 400 instrument operating at 400.13 MHz for 1H and 100.13 MHz for 13C. ESI-MS spectra were recorded with a LCQ-DECA Thermo Finnigan instrument. TLC was performed on aluminum precoated 4 × 10 cm silica gel plates with fluorescent indicator (Fluka). Column chromatography was carried out on Silica gel (J.T. Baker, 0.060–0.200 mm).

The synthesis was carried out without purification of the intermediates until the final product. The crude reaction was checked by 1H NMR and ESI-MS.

3-Hydroxy-2-(5'-hydroxypentyl)-4H-pyran-4-one (5-H). 6-(Benzyloxy)-1-(furan-2'yl)-hexan-1-ol (1) was prepared according to Ramanathan and Levine [19] from 2-furyllithium (10 mmol) and 6-benzyloxyhexan-1-al (10 mmol) obtained from 6-benzyloxyhesan-1-ol oxidized in according to Leanna et al. [20]. ESI-MS: m/z 273(M-1). 2-(5-(Benzyloxy)pentyl)-4-bromo-6-methoxytetrahydro-2H-pyran-3-ol (2) was prepared according to Weeks et al. [21]. 6-(Benzyloxy)-1-(furan-2'yl)-hexan-1-ol and Br2 (in stoichiometric ratio 1/2) were separately but contemporaneously added drop by drop to a solution of methanol/water at the temperature of –20°C. The reaction mixture was allowed to reach room temperature and then refluxing for 3 h. The residue treated with Na2CO3 (s.s) and extracted with ether (2 × 50 mL). The organic layer was dried over Na2SO4, the solvent evaporated in vacuo. ESI-MS: m/z 386 (M – 1). 2-(5-(Benzyloxy)pentyl)-4-bromo-6-methoxytetrahydro-2H-pyran-3-ol (2) was refluxing in CH3COOH as solvent for 2 h to obtain 2-(5'-(benzyloxy)pentyl)-3-hydroxy-4H-pyran-4-one (3). ESI-MS: m/z 287 (M – 1). Compound 3 was refluxing in CH3COOH and HBr (in excess) to remove benzyl protection with the simultaneous formation of the acetyl ester (4). ESI-MS: m/z 239 (M – 1). The solvent was evaporated under vacuum and the crude (solution of methanol/water 1/1) was treated with HCl to get the final product 3-hydroxy-2-(5'-hydroxypentyl)-4H-pyran-4-one (5-H). Yield, 50%. White solid, mp 99–101°C. 1H NMR (400 MHz, CDCl3): δ (ppm). 1.39–1.45 (m, 2H, 2'CH2 or 3'CH2), 1.53–1.62 (m, 2H, 3'CH2 or 2'CH2), 1.63–1.72 (m, 2H, 4'CH2), 2.71 (t, 2H, 1'CH2, J = 7.6Hz), 3.64 (t, 2H, 5'CH2, J = 6.4 Hz), 6.40 (d, 1H, 5CH, J = 5.2Hz), 7.71 (d, 1H, 6CH, J = 5.2 Hz). 13C NMR (200 MHz, CDCl3): δ (ppm) 25.22, 26.26, 28.08, 32.27, 62.59, 112.87, 142.89, 152.19, 154.29, 173.13. ESI-MS: m/z 197 (M – 1). UV-vis: (H2O, pH 7.4) λmax = 276 nm, (H2O, pH 9.0) λmax = 321 nm. IR (KBr): ν(C=O) 1608 cm–1.

Cyclic voltammetric studies were performed by using a three-electrode cell having a glassy carbon working electrode. A platinum-spiral was used as counter-electrode and the aqueous saturated calomel reference electrode (SCE) was mounted with a Luggin capillary. A BAS 100A electrochemical analyzer was used as a polarizing unit. Potential values are referred to the SCE.

The supporting electrolyte was NaClO4 (0.2 M) and the solutions were purged with nitrogen prior to measurement. Conversion to values vs. NHE was obtained upon adding +0.24 V to the relevant SCE values.

JOB test. Different ratios of iron(III) ad 5-H were prepares while keeping the total concentration of metal plus ligand at 0.6 mM MOPS buffer with 50% methanol (v/v) was used to maintain pH at 7.4. The solution in the UV cuvette was mixed by pipet several times and then sealed. All experiments were carried out at 25°C.

RESULTS AND DISCUSSION

With the aim to obtain a biocompatible iron-chelator inspired to maltol, we synthetized 3-hydroxy-2-(5'-hydroxypentyl)-4H-pyran- 4-one (5-H) (Scheme 1).

Scheme 1 . Preparation of compound 5-H.

To study the reaction of 5-H with iron, we recorded electronic spectra of solution of the complex formed at concentration 1 mM of Fe(NO3)3 and a fourfold of ligand (4 mM) for pH 3.0, 5.0 and 7.4. Ferric complexes were prepared by mixing these reagents: a red-purple colour developed at once. At pH 3.0, the spectrum was characterized by two absorption with λ = 410 and 470 nm (Fig. 1). By varying the pH from 3.0 to 7.4, the intensity of absorption with λ = 410 nm increased.

Fig. 1.
figure 1

UV-vis spectra of a mixture of aqueous solutions of Fe(NO3)3 · 9H2O and 5-H (molar ratio 1 : 4) at different pH.

Full size image

The shape of the curves recalled electronic spectra of a mixture of aqueous solutions of iron(III) and maltol [22]. By comparing the spectra we assumed that the two bands may be due to dimeric and trimeric specie whose ratio was function of pH. This hypotesis was confirmed by ESI/MS spectrum in which the monocharged dimer [(Fe(5)2]+ was at m/z 450 (M + 1) while trimeric specie [Fe(5-H)(5)2]+ was at m/z 648 (M + 1). At pH 3.0 the ratio dimer : trimer is about 5, while at pH 7.4 the ratio was reversed.

Predominance of trimeric specie at pH 7.4 was also been confirmed by Job’s test. It was carried out by preparing 12 solutions covering the whole range of molar fractions for 5-H and Fe(III), keeping the total concentration constant (0.6 mM). The UV absorbance at 410 nm for each solution was measured. A normalized scale was used. It was obtained when experimental absorbances were divided by maximum absorbance term, designated ABS-MAX (Fig. 2) [23]. Thus, Job’s plot generated a curve, where two straight lines drawn through initial and final points generated a crossed point, X = 0.76. The stoichiometry of the complex was then calculated based on Eq. (1), giving m = 3 and n = 1.

$$m{\mathbf{5}}{\text{-}}{\mathbf{H}} + n{\text{F}}{{{\text{e}}}^{{3 + }}} \to {{\left[ {{\text{F}}{{{\text{e}}}_{n}}{{{\left( {\mathbf{5}} \right)}}_{m}}} \right]}^{{(3n - m)}}} + m{{{\text{H}}}^{ + }}.$$
((1))
Fig. 2.
figure 2

Proposed structure and Job plot of Fe(III) and 5-H mixtures monitored at 410 nm, pH 7.4.

Full size image

Efficiency of an iron chelator is related to its ability to promote or inhibit various reactions involved in reactive oxygen species generation (ROS), particularly in physiologic conditions. Therefore, we performed a pH dependent electrochemical characterization. Cyclic voltammogram of an acid solution of Fe(NO3)3 showed a quasireversible Fe(III)/Fe(II) reduction with Epc = +0.26 V and Epa = +0.60 V (Fig. 3, curve a).

Fig. 3.
figure 3

Cyclic voltammograms of an aqueous solution containing (a) Fe(NO3)3, 1 × 10–3 mol dm–3, pH 2.0; after addition of 5-H (4 × 10–3 mol dm–3) at pH (b) 3.0, (c) 5.0 and (d) 7.4. Glassy carbon working electrode; scan rate 0.2 V s–1.

Full size image

As consequence of the addition of 5-H at the Fe(III) containing solution, the pH decreased to 3.0 and the Fe(III)/Fe(II) cathodic peak shifted at –0.17 V (Fig. 3, curve b). Adding NaOH when pH was 5.0, the Fe(III)/Fe(II) reduction occurred at more negative potential and assumed a more character of electrochemical reversibility (E°' = –0.53, ΔE = 98 mV at 100 mV s–1). At the same time, a minor (asterisked) peak appeared. At low scan rates (0.005 V/s) this later peak disappeared and the ipa/ipc ratio was 1 for the primary reduction. This imply the occurrence of a reversible chemical reaction following the electron transfer.

Since electrochemical experiments may entail local pH changes at the electrode surface, e.g. due to protonation/deprotonation of the ligand after reduction of its Fe(III) complex, the electrochemical behavior of solution contained Fe(III) and 5-H was checked on buffered media (Veronal) at pH 7.4. In addition, under such conditions, a partially chemically reversible one-electron reduction at E°' = –0.54 V (–0.52 V for iron-maltol solution in the same conditions) is recorded. As discussed previously, changes in the pH range 5.0 < pH < 7.4 did not alter the chemical composition of the {Fe(III)-5} system, consequently we recorded similar voltammograms.

To avoid redox cycling in vivo, it is critical that iron(III) complex is fully formed and not reducible. It has been demonstrated that under physiological conditions, involving μM concentrations of iron and chelator (maltol, thiomaltol, deferiprone), electrode potentials become more positive with decreasing iron concentration [24, 25]. Therefore, even if slightly higher value is expected for electrode potential of [Fe(5)3], it is surely lower than that for the ascorbyl/monohydrogenascorbate couple (+0.1 V, pH 7.0). Consequently, chelation of Fe(III) with 5-H could prevents iron redox cycling.

CONCLUSIONS

In conclusion, the alcoholic chain could lead to increase of the hydrophilic character of 5-H to respect maltol without alter iron binding ability of 3-hydroxypyran-4-one core. An obvious reason is that oxygen remains protonated at physiologic pH, but also an unfavorable entropic contribution inhibit the ligand acting as tridentate. Indeed, a large degree of conformational freedom is associated with the alcoholic chain placed at ring position 2 in the free ligand 5-H, which would be lost upon chelation with iron. Although 5-H not acts as tridentate ligand, it appears an interesting potential candidate as iron-chelating agent because result in iron redox-inactive complex at pH 7.4.