Gallic acid grafting modulates the oxidative potential of ferrimagnetic bioactive glass-ceramic SC-45

Colloids and Surfaces B: Biointerfaces 148 (2016) 592–599

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb

Gallic acid grafting modulates the oxidative potential of ferrimagnetic bioactive glass-ceramic SC-45 Ingrid Corazzari a,b,∗ , Maura Tomatis a,b , Francesco Turci a,b , Sara Ferraris c , Elisa Bertone c , Enrico Prenesti a , Enrica Vernè c a

University of Torino, Department of Chemistry, Via Pietro Giuria 7, 10125 Torino, Italy “G. Scansetti” Interdepartmental Centre for Studies on Asbestos and other Toxic Particulates, Via Pietro Giuria 7, 10125 Torino, Italy c Politecnico di Torino, Department of Applied Science and Technology, Institute of Materials Physics and Engineering, Corso Duca degli Abruzzi, 24, 10129, Torino, Italy b

a r t i c l e

i n f o

Article history: Received 5 May 2016 Received in revised form 22 July 2016 Accepted 23 September 2016 Available online 23 September 2016 Keywords: Ferrimagnetic glass-ceramic Gallic acid Free radical release Lipoperoxidation Bioactivity Iron reduction

a b s t r a c t Magnetite-containing glass-ceramics are promising bio-materials for replacing bone tissue after tumour resection. Thanks to their ferrimagnetic properties, they generate heat when subjected to an alternated magnetic field. In virtue of this they can be employed for the hyperthermic treatment of cancer. Moreover, grafting anti-cancer drugs onto their surface produces specific anti-neoplastic activity in these biomaterials. Gallic acid (GA) exhibits antiproliferative activity which renders it a promising candidate for anticancer applications. In the present paper, the reactivity of ferrimagnetic glass-ceramic SC-45 grafted with GA (SC-45+GA) was studied in terms of ROS release, rupture of the C–H bond of the formate molecule and Fenton reactivity by EPR/spin trapping in acellular systems. The ability of these materials to cause lipid peroxidation was assessed by UV–vis/TBA assay employing linoleic acid as a model of membrane lipid. The results, compared to those obtained with SC-45, showed that GA grafting (i) significantly enhanced the Fenton reactivity and (ii) restored the former reactivity of SC-45 towards both the C–H bond and linoleic acid which had been completely suppressed by prolonged contact with water. Fe2+ centres at the surface are probably implicated. GA, acting as a pro-oxidant, reduces Fe3+ to Fe2+ by maintaining a supply of Fe2+ at the surface of SC-45+GA. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Bioactive glasses and glass-ceramics are versatile materials for bone-tissue replacement. The properties of such materials, depending upon their composition, can be tailored to specific needs. By varying the glass composition (mainly SiO2 and P2 O5 content), it is possible to obtain quite stable as well as highly bioactive glasses and even fully resorbable materials [1–4]. Moreover, the introduction of different metal oxides allows the material reactivity to be further modulated and the release of ions with specific biological activities, including the stimulation of osteoblast proliferation, bone formation and mineralization, and antibacterial activity, to be induced [3–6]. Thanks to their particular structure, specific ions (e.g. Ag+ for

∗ Corresponding author at: University of Torino, Department of Chemistry, Via Pietro Giuria 7, 10125 Torino, Italy. E-mail addresses: [email protected] (I. Corazzari), [email protected] (M. Tomatis), [email protected] (F. Turci), [email protected] (S. Ferraris), [email protected] (E. Bertone), [email protected] (E. Prenesti), [email protected] (E. Vernè). http://dx.doi.org/10.1016/j.colsurfb.2016.09.034 0927-7765/© 2016 Elsevier B.V. All rights reserved.

antibacterial properties) can be put on the surface by means of an ion exchange process [4,7]. Besides their applications in producing scaffolds, bone fillers and coatings, bioactive glasses and glass-ceramics have been exploited to repair bone defects after tumour resection [8]. In particular ferrimagnetic glass-ceramics find their principal scope in contributing to bone healing after cancer removal. Ferrimagnetic properties are assured by the presence of magnetite crystals (Fe3 O4 ) embedded in the glass matrix [9]. Magnetite crystals nucleation is obtained by introducing iron oxides into the glass composition [9–13]. It is well known that under an alternating magnetic field, magnetic materials can generate heat by hysteresis loss. An increasing interest on hyperthermia has been shown by the scientific community due to its promising medical applications. Hyperthermia can be considered as a specific anticancer treatment, since cancer cells tolerate heat less efficiently than healthy cells when the temperature of the treated tissue reaches 44 ◦ C [14]. Moreover several authors are developing drug-delivery systems where the hyperthermic effect, assured by the presence of magnetic particles, can be exploited to induce the release of specific drugs embedded in a

I. Corazzari et al. / Colloids and Surfaces B: Biointerfaces 148 (2016) 592–599

polymeric matrix [15,16]. When the ferrimagnetic glass-ceramic is subjected to an alternating magnetic field, heat is generated causing a local rise of the temperature (42–45 ◦ C). The induced hyperthermia can damage directly tumour cells or make them more susceptible to radio/chemotherapy. For this reason, it can be exploited as a complementary therapy in the treatment of bone cancer besides surgery, radio- and/or chemo-therapy. In the case of bone implants after tumour resection, this local overheating can act preferentially against new bone-tumour formation after the surgical removal in case of relapse. Moreover, a recent study by some of the authors refers a positive influence of the mild magnetic field induced by magnetite crystals on cell viability [17], an effect that is not provided by superparamagnetic nanoparticles, which do not retain any magnetism after removal of the external magnetic field. After implantation, bioactive glasses and glass-ceramics interact with bone cells, bio-fluids and biomolecules, causing ion-exchange reactions culminating with the precipitation of calcium phosphates and the crystallization of hydroxyapatite on the biomaterial surface [1]. The formation of a hydroxyapatite layer induces the chemical bonding between the artificial surface and the bone tissue and thus it represents a beneficial outcome of the interaction with biofluids. For this particular application, in addition to the general features required of bone substitute materials (i.e. biocompatibility, bioactivity, bio-stability, and bio-degradability) [18], a specific anti-neoplastic activity is desirable. The exploitation of bioactive glasses and glass-ceramics functionalized with chemotherapics (e.g. doxorubicin, cis-platinum complexes or dexamethasone) for the local treatment of bone cancer results in an increased efficiency of the therapy minimizing the undesired effects due to the systemic administration of anti-neoplastic drugs [19,20]. Grafting bio-active molecules onto the surface of ferrimagnetic glass-ceramics may be an interesting strategy to obtain specific therapeutic tools which combine hyperthermia with the anti-cancer properties of bioactive molecules. Some studies documented selective toxicity of polyphenols towards cancer cells [21] and their ability to stimulate healthy osteoblasts [22]. Moreover, polyphenols exert, on one hand, a protective role on healthy cells and, on the other, a sensitizing effect on tumour ones rendering them more susceptible to antineoplastic drugs [23,24]. These molecules exhibit also a certain thermal stability which preserves their activity in the temperature range of cement hardening and hyperthermia treatment [25]. In previous works, some of the authors modified the surface of the well characterized ferrimagnetic glass-ceramic SC-45 with gallic acid (3,4,5-trihydroxybenzoic acid, GA) [26]. GA is a phenolic acid widely distributed in many different families of higher plants. In nature it exists both in free state, and as a precursor of polyphenols. It was chosen because, besides its recognized antioxidant activity, it shows some specific anti-tumour properties. In particular, the literature reports that GA can induce apoptosis of the tumour cells via different mechanisms depending upon the cell type, but all these mechanisms are likely driven by ROS generation [27]. Moreover, some authors have observed that GA strongly inhibits cancer cell migration and metastasis by suppressing the expression of some metalloproteinases [28]. The presence of GA could modify the surface reactivity of a magnetite-containing biomaterial acting as a reducing agent for iron (III). As observed with toxic particulates including lunar dust simulants and volcanic ashes [29,30] the presence of redox active centres (most likely iron) accounts for the oxidative potential of these particulates. Analogously, SC-45 could show a similar reactivity mediated by iron centres at the surface of the magnetite crystals. Moreover, the interaction of redox active metal ions and some antioxidant molecules enhances the reactivity mediated by iron [29,31]. In the present article the chemical interactions of SC-45 modified with GA and some models of biologically relevant molecules, mimicking the environment this biomaterial may be in contact with after implantation, has been considered.

593

Table 1 Chemical composition (as oxides%) of SC-45 glass-ceramic.

wt.%

Fe2 O3

FeO

SiO2

CaO

Na2 O

P2 O5

31

14

24.7

13.5

13.5

3.3

In particular, the anti-oxidant/pro-oxidant potential of the bare glass-ceramic compared to the gallic acid-grafted one has been investigated for the first time. Since SC-45 is a ferrimagnetic glassceramic with heating ability under alternating magnetic field, it can be applied for bone substitution after cancer removal. GA grafting on its surface can potentiate its activity in cancer treatment. In this context surface reactivity was investigated because it can play a crucial role in exerting toxic outcomes with a dual effect representing a useful tool which can be exploited to kill tumour cells, but also a possible interference with the process of bone regeneration. 2. Material and methods 2.1. Reagents Where not otherwise specified, all the reagents were from Sigma Chemicals (Sigma-Aldrich St. Louis, MO, USA), and all the solutions employed were prepared with Milli-Q ultrapure water system (Merck Millipore, Darmstadt, Germany). 2.2. Material preparation SC-45 glass-ceramic was prepared by traditional melt and quenching technique and fully characterized in previous works [13,20,32]. Its composition (wt%) is reported in Table 1. The obtained glass-ceramic was ball milled and sieved up to a grain size of 20 ␮m. The powder surface was functionalized with gallic acid via direct grafting to the surface exposed OH groups without any spacer, as described in Ferraris et al. [26]. Briefly SC-45 powder was suspended in ultrapure water (0.10 g powder in 10 ml water) for one week at 37 ◦ C in order to expose –OH groups at the surface, as described in Vernè et al. [20]. Hydroxyl groups are fundamental for the direct grafting of gallic acid molecules on the glass-ceramic surface. After incubation, the powder was separated from the supernatant with the use of a magnet, and dried under a laminar flow cabinet (FASTER CYTOSAFE) in order to avoid surface contaminations. The dried powder (hereafter named SC-45-pretreated) was suspended in an aqueous solution 1 mg/ml of GA (0.1 g powder in 5 ml of GA solution) for 3 h at 37 ◦ C. After incubation, the sample was separated from the supernatant by the use of a magnet, gently washed twice in ultrapure water and dried under the laminar flow cabinet. The GA functionalized sample is hereafter named SC-45+GA. A thorough physico-chemical characterization of the SC-45+GA was object of a previous study [26]. 2.3. Bioactivity In order to evaluate in vitro the bioactivity of SC-45 samples before and after gallic acid grafting, the powders were soaked in Simulated Body Fluid (SBF) [33] for 14 and 28 days (0.1 g powder in 25 ml SBF) at 37 ◦ C in an incubator. Solution refresh was performed every two days in order to mimic the physiological turnover of body fluids. At each refresh the solution pH was measured. SBF was prepared according to the protocol proposed by Kokubo [34]. At the end of the soaking period samples were gently washed in ultrapure water and dried under a laminar flow cabinet. In order to investigate surface reactivity and hydroxyapatite precipitation, samples were deposited onto conductive carbon tape, fixed to alu-

594

I. Corazzari et al. / Colloids and Surfaces B: Biointerfaces 148 (2016) 592–599

minium stub and covered with a thin chromium layer (5–10 nm) for FESEM observation (FESEM, SUPRATM 40, Zeiss). Also, pellets of SC-45-pretreated and SC-45+GA before and after SBF soaking were prepared with KBr (2 mg glass-ceramic powder and 198 mg KBr) by uniaxial pressing (4 t, 10 s) and analyzed by means of Fourier Transformed Infrared Spectroscopy (FTIR Alpha, Bruker Optics, Ettlingen, Germany). 2.4. Durability in simulated body fluid (SBF) Solubility of the samples in SBF was assessed by means of ICPAES (Inductively Coupled Plasma – Atomic Emission Spectroscopy) measuring the kinetics of Si release. SC-45, SC-45-pretreated and SC-45+GA (3 mg/ml) were dunked in SBF and incubated at 37 ◦ C for 2 weeks and 1 month in static leaching condition. After incubation the suspensions were centrifuged and the supernatant filtered through cellulose acetate filters (cut off 0.20 ␮m) to remove any residual particle. Aqueous HNO3 solution (3.00 M) was added to the clear solution in order to prevent precipitation phenomena. The concentration of Si released was measured by ICP-AES on a Varian Liberty 100 ICP atomic emission spectrometer. 2.5. Free radical generation study The release of radical species was monitored by EPR (Electron Paramagnetic Resonance) spectroscopy (Miniscope 100 X-band EPR spectrometer, Magnettech, Germany) by means of the spin-trapping technique with 5,5-dimethyl-1-pyrroline-N-oxide (DMPO, Alexis-Biochemicals, Lausen, Switzerland) employed as spin trap molecule. 2.5.1. Generation of oxygen centred radicals SC-45, SC-45-pretreated and SC-45+GA (10 mg/ml) were suspended in a DMPO (30 mM) buffered aqueous solution (potassium phosphate buffer, KPB, 0.1 M, pH 7.4) and kept under continuous stirring. The release of oxygen-centred radicals (HO• , HOO• and • O −) was evaluated after 10, 30 and 60 min of incubation. Blanks 2 were performed in parallel in the absence of powder. As a positive control for ROS generation, a TiO2 (Aeroxide® , Evonik Industries, Essen, Germany) suspension was irradiated with a simulated solar light under the same experimental conditions [35]. The irradiation of TiO2 suspension was performed with a 500 W mercury/xenon lamp (Oriel instruments) equipped with an IR water filter to prevent the overheating of the suspension. Simulated solar light was obtained by applying a 400 nm cut-off filter. This filter let to pass about 5% of UV light in the UV A region. The experiments were performed in triplicate. 2.5.2. Surface-driven Fenton reactivity (target molecule H2 O2 ) 10 mg of the powders were suspended in 1.25 ml of an aqueous buffered solution (KPB 0.2 M, pH 7.4) containing DMPO (0.034 M) and H2 O2 (0.080 M). After 10, 30 and 60 min of incubation under continuous stirring, the EPR spectra were recorded on the suspension. Blanks were performed in parallel in the absence of powder. All the experiments were performed in triplicate. 2.5.3. Generation of carbon centred radicals (target molecule formate ion) 10 mg of the powders were suspended in 1 ml of an aqueous buffered solution (KPB 0.25 M, pH 7.4) of sodium formate (1.0 M) and DMPO (0.085 M). After 10, 30 and 60 min of incubation under continuous stirring, the EPR spectra were recorded on the suspension. Blanks were performed in parallel in the absence of powder. All the experiments were performed in triplicate.

2.6. Oxidative degradation of linoleic acid (TBA assay) The potential of inducing lipid peroxidation was assessed by means of thiobarbituric acid (TBA) assay using linoleic acid as a model of polyunsaturated fatty acids in the cell membrane. The main lipoperoxidation product malonyldialdehyde (MDA) forms with TBA a coloured complex. The assay is based on the reactivity of MDA – a colourless molecule – with TBA to produce a pink adduct which absorbs at 535 nm. The powder was suspended (15 mg/ml) in a buffered (KPB 5 mM, pH 7.4) dispersion of linoleic acid (1 mM) containing 2.5 wt.% ethanol. The suspension was continuously stirred for 1 h. The lipid peroxidation was stopped by adding 0.1 ml of an ethanol solution of butyl hydroxyl toluene (BHT, 0.2 wt.%) to the suspension. The powder was removed by centrifuging (20,000 g for 30 min). A solution of TBA (0.034 M) containing HCl (0.25 M) and trichloroacetic acid (0.92 M) was added to the supernatant (2:1 v:v), and the resulting solution was heated at 100 ◦ C for 1 h. After cooling in an ice bath, 3 ml 1-butanol was added to extract the coloured complex. The absorbance at 535 nm was measured on the organic phase by means of a UV–vis spectrophotometer (Uvikon, Kontron Instruments, Inc., Everett, MA, USA). Blanks were performed in parallel in the absence of powder. The experiments were performed in triplicate.

3. Results and discussion 3.1. Bioactivity Hydroxyapatite nucleation upon contact with simulated body fluid (SBF) was assessed as an index of bioactivity on both nongrafted and GA-grafted glass ceramic. The evidence of GA grafting on this sample had been documented in a previous paper [26] by means of both X-ray photoelectron spectroscopy (XPS) and Folin & Ciocalteu test, to measure the surface elemental composition and the specific polyphenols redox activity, respectively. To evidence the formation of hydroxyapatite, electron microscopy and infrared spectroscopy were used and FESEM images and FTIR spectra of the different samples before and after 14 and 28 days of soaking with SBF are reported in Fig. 1. The deposition of particles with the typical morphology of hydroxyapatite could be evidenced after SBF soaking. The EDS analysis evidenced that the crystals were constituted mainly by Ca and P. The Ca/P ratio ranged between 1.40 and 1.77, and was consistent with precipitated hydroxyapatite [36]. The FTIR spectrum of the SC-45-pretreated in the 400–1300 cm−1 region (Fig. 1, spectrum a) evidenced the characteristic absorption bands of the glass-ceramic constituents [36–38]: Si–O–Si bending (470 cm−1 ); P–O bending (565 cm−1 ); Fe–O vibrations (560–600 cm−1 ); Si–O–Si asymmetric vibrations (1000–1050 cm−1 ); P–O stretching (1035 cm−1 ); Si–OH stretching (1082 cm−1 ). In all spectra, the intensity of the Si–O–Si stretching mode at 1035 cm−1 was normalized to 1 A.U. to take into account the effect of different local density of the KBr pellet. Spectra were baseline corrected and superimposed to highlight the spectral variations. FTIR data confirm that the GA grafting did not induce a significant structural alteration of the SC-45 pretreated sample (spectra a ≈ b; a’ ≈ b’; a” ≈ b”). This result is consistent with the small amount of GA grafted on the outermost surface layer of the sample and with the intense signals related to the bulk Si–O and Fe–O bonds. Nevertheless, precipitated hydroxyapatite can be roughly quantified [33] by investigating the spectral region between 400 and 1300 cm−1 . The signals in the 550–650 cm−1 region and at 1035 cm−1 , ascribed to PO4 3− bending and stretching, have been suggested as sensitive to the hydroxyapatite precipitation on bioactive glass surfaces [36,38]. The relative intensity of the

I. Corazzari et al. / Colloids and Surfaces B: Biointerfaces 148 (2016) 592–599

595

Fig. 1. Hydroxyapatite nucleation upon contact with BSF. FESEM images and FTIR spectra of SC-45 pretreated and SC-45+GA before and after different soaking time in SBF (rows). Arrows indicate Ca-P rich precipitates. FTIR spectra of SC-45-pretreated and SC-45+GA before (spectra a, b) and after 14 days (a and b ) and 28 days (a and b ) SBF soaking are reported in the 400–1300 cm−1 region. The strong and convoluted band between 1200 and 800 cm−1 is mainly due to the Si–O–Si stretching of the silica matrix superimposed with the stretching modes of distorted PO4 3− of hydroxyapatite. The band at 565 cm−1 relates to the triply degenerate O–P–O mode and the absorption at 470 cm−1 is due to Si–O–Si bending mode.

Table 2 Relative intensity of the O–P–O bending mode at 565 cm−1 (I565 ) with respect to Si–O–Si mode at 470 cm−1 (I470 ) evaluated for SC-45-pretreated and SC-45+GA before and after 14 and 28 days soaking time in SBF.

No soaking 14 days SBF 28 days SBF

SC-45 pretreated SC-45+GA SC-45-pretreated SC-45+GA SC-45-pretreated SC-45+GA

I565

I470

I565 /I470

0.009 0.010 0.011 0.012 0.012 0.013

0.009 0.009 0.008 0.009 0.008 0.009

1.01 1.15 1.37 1.31 1.54 1.43

O–P–O bending mode at 565 cm−1 (I565 ) with respect to Si-–O-–Si mode at 470 cm−1 (I470 ) were employed to evaluate the formation of hydroxyapatite on SC-45 pretreated and SC-45+GA before and after 14 and 28 days soaking time in SBF (Table 2). The O–P–O bending mode is observable with SC-45 pretreated and SC-45+GA before soaking. Analogously to what previously observed by Stanciu et al. [36] with Bioglass® , the presence of such signal also before soaking was attributed to the presence of P–O bonds in the bulk of SC-45 which derives from its compositional features. No significant differences were evidenced between SC-45 preatreated and SC-45+GA, at each soaking time. A significant increase of hydroxyapatite was conversely observed for increasing SBF soaking times. These results evidenced that grafting GA did not cause any bulk modification nor quenched the bioactivity of the glass ceramic.

3.2. Durability in SBF In order to assess the effect of the grafting-procedure steps (namely pre-treatment in water, and functionalization with GA) on the durability of SC-45, the concentration of Si (␮g/ml) released from SC-45, SC-45-pretreated and SC-45+GA after 2 weeks and 1 month of incubation (T = 37 ◦ C) in SBF was measured by means of ICP-AES. The results are reported in Fig. 2. After 2 weeks of incubation (white bars), the amount of Si detected in the supernatant was 7.77 ± 0.24 ␮g/ml with SC-45, 7.22 ± 0.50 ␮g/ml with SC-45pretreated, and 4.44 ±0.53 ␮g/ml with SC-45+GA. This result shows that the solubility of SC-45 and SC-45-pretreated did not differ significantly. On the other hand the amount of Si detected with SC-45+GA was slightly, but significantly, lower. After 1 month of incubation (grey bars), the amount of Si detected in the supernatant was 9.95 ± 0.97 ␮g/ml with SC-45, 10.9 ± 0.33 ␮g/ml with SC-45pretreated and 9.15 ± 0.72 ␮g/ml with SC-45 +GA. In this case all the three samples exhibited a very similar behaviour. These results indicate that the durability of such glass-ceramic is only minimally influenced by the pretreatment with water (SC45-pretreated) and the grafting with GA (SC-45+GA). 3.3. Free radical-mediated reactivity The effect of grafting GA on the reactivity of SC-45 was investigated employing a set of in vitro cell-free tests usually employed in the study of particle reactivity. Beside the ability of buffered sus-

596

I. Corazzari et al. / Colloids and Surfaces B: Biointerfaces 148 (2016) 592–599

Fig. 2. Durability in simulated body fluid. Si concentration (␮g/ml) detected by ICP/AES after 2 weeks (white bars) and 1 month (gray bars) of incubation of SC-45, SC-45pretreated and SC-45+GA suspensions (3 mg/ml) in SBF at 37 ◦ C in dynamic leaching condition.

pensions of SC-45, SC-45-pretreated and SC-45+GA to induce the release of oxygen-centred radicals, the free radical mediated reactivity towards some model of biologically relevant molecules (i.e. H2 O2 , formate ion and linoleic acid) was also assessed. 3.3.1. Generation of oxygen centred radicals Reactive oxygen species (ROS) are involved in the toxicity mechanism of several particles [39]. In particular, the involvement of oxygen-centred radicals in inducing oxidative stress toward cells and tissue is well assessed. The ability of the samples to generate HO• , HOO• or • O2 − in a buffered suspension was investigated by means of EPR coupled with spin trapping technique. The samples (10 mg/ml) were suspended in a buffered solution (potassium phosphate buffer, KPB, pH 7.4) in the presence of DMPO as spin trapping agent. A well-known experiment able to generate HO• from TiO2 suspension after irradiation (␭ > 400 nm) was employed as positive control. The spectra recorded on all the samples studied evidence that none of the powders was able to generate oxygencentred radicals in solution (Fig. 3). 3.3.2. Radical reactivity towards biologically-relevant molecules In the first test, H2 O2 was allowed to interact with iron reactive centres at the surface of the samples (Fenton reactivity) and the generation of HO• following the reaction (Eq. (1)) was monitored. Fe2+ surf + H2 O2 → Fe3+ surf + OH− + HO• The release of HO•

(1)

radicals in the presence of H2 O2 mimics the reaction that might take place when a particle is exposed to an oxidative environment such as the lysosomal fluid of macrophages and polymorphonucleated cells, which are involved in bone implantassociated acute and chronic inflammation. H2 O2 is also produced in large amount in several tumour cells [40]. Moreover, also the osteoclasts are able to produce H2 O2 which on one hand plays a role in the differentiation process [41] and on the other hand can stimulate bone matrix resorption allowing bone remodeling [42]. Using DMPO as spin trapping agent, the capability of SC-45+ GA (15 mg/ml) to generate HO• was assessed and compared to that of SC-45-pretreated and SC-45. The spectra recorded evidence that all the samples were able to generate HO• , via a Fenton-like mechanism, at each time point. The same experiment was carried out in the absence of the powder (blank) and no signal was detected.

2000 AU

ctrl +

SC-45

SC-45-pretrea ted

SC-45 + GA

326

328

330

332

334

336

338

340

342

field, mT Fig. 3. ROS release in buffered suspension of SC-45, SC-45-pretreated and SC-45+GA in the presence of DMPO. EPR spectra were collected after 30 min of incubation. Positive control experiment (Ctrl+) was carried out in the same conditions with TiO2 after irradiation with a solar lamp (␭ > 400 nm).

Representative EPR spectra of [DMPO–HO]• adduct recorded after 60 min of incubation of the three samples with H2 O2 are reported in Fig. 4A. In order to assess the kinetics of HO• yield, the signal recorded at 10, 30 and 60 min of incubation were double-integrated and the intensity reported as the mean values (± SE) of three separated experiments (Fig. 4B). The amount of HO• generated from SC-45 did not significantly differ from that of SC-45-pretreated and remained constant over time. This result indicates that SC-45 reactivity was not modified by the contact with water. At the opposite, a clear time-dependent increase in the intensity of the EPR signal was observed with SC-45+GA. Moreover, SC-45+GA (Fig. 4B) was significantly more reactive than both SC-45 and SC-45-pretreated at 30 and 60 min of incubation, indicating that the anchoring of GA onto the surface modified the reactivity of SC-45 towards H2 O2 . This enhanced reactivity is probably related to the reduction of part of Fe3+ to Fe2+ . This hypothesis is corroborated by the literature which reports that GA and Fe3+ can form complexes which in turn decomposes to Fe2+ and semiquinone [43].

I. Corazzari et al. / Colloids and Surfaces B: Biointerfaces 148 (2016) 592–599

A

150

597

B

140

2000 A.U.

10 min 30 min 60 min

130

[DMPO-HO]° intensity (A.U.)

SC-45+ GA

SC-45-pretreated

SC-45

120 110 100 90 80 70 60 50 40 30 20

Blank

326

328

330 332 334 336 338

10 0

340

SC-45

SC-45-pretreated

SC-45+GA

Field (mT) Fig. 4. Fenton-like reactivity measured on buffered suspensions of SC-45+GA, SC-45-pretreated and SC-45 in the presence of H2 O2 and DMPO: (A) representative EPR spectra collected after 60 min of incubation, (B) Average spectra signal intensity from double integration of the EPR spectra recorded at 10, 30 and 60 min of incubation.

A

2000 A.U.

B

2000 A.U.

C

60'

60'

60'

30'

30'

30'

10'

10'

10'

2000 A.U.

326 328 330 332 334 336 338 340 326 328 330 332 334 336 338 340 326 328 330 332 334 336 338 340

field, mT

field, mT

field, mT

Fig. 5. EPR spectra recorded on SC-45 (A), SC-45-pretreated (B) and SC-45+GA (C) buffered suspensions (10 mg/ml) in the presence of sodium formate as target molecule and DMPO as spin trapping agent after 10, 30 and 60 min of incubation.

The ability of GA to enhance the amount of Fe (II) on the surface of the bioactive glass has been shown by thermal gravimetric analysis in the paper of Ferraris et al. [26]. In that article, the gain of weight measured by heating SC-45, and SC-45+GA under oxygen was attributed to the oxidation of Fe2+ to Fe3+ . The gain of weight observed with SC-45+GA was significantly higher than that observed with SC-45. This result indicated that the amount of Fe2+ formerly present on the sample was increased by the treatment with gallic acid in virtue of its reducing action. 3.3.3. Homolytic cleavage of the C–H bond of formate molecule The second radical reaction investigated follows the cleavage of the C–H bond in the formate ion which can be taken as a model reaction occurring with several molecules of biological interest

such as peptides, proteins and lipids. In the presence of a reactive dust, a direct surface-assisted homolytic cleavage possibly occurs on redox-reactive transition metal ions (such as iron ion) exposed at the particle surface. Such reaction yields the formation of a carboxyl radical (• CO2 − ) which is stabilized by resonance and can react with DMPO forming the [DMPO–CO2 ]•− paramagnetic adduct characterized by a specific EPR spectrum. In Fig. 5 the EPR spectra recorded after 10, 30 and 60 min on buffered suspensions of SC-45 (Fig. 5A), SC-45-pretreated (Fig. 5B) and SC-45+GA (Fig. 5C) are shown. Blank experiments in the absence of the powder were also carried out and no EPR signal was detected (data not shown for brevity). SC-45 was able to cleave C–H bond of formate anion at each time point with a kinetics increasing with time. Conversely, SC-45-pretreated was inactive at each time point tested. This result suggests that one

598

I. Corazzari et al. / Colloids and Surfaces B: Biointerfaces 148 (2016) 592–599

Fe3+ to Fe2+ . These results are similar to what observed with the EPR experiments employing formate ion (Fig. 6) and indicate that Fe2+ ions are involved in the lipid peroxidation.

0.25

Absorbance@535nm

0.20

4. Conclusions 0.15

0.10

0.05

0.00

blank

SC-45

SC-45-pretreated SC-45+GA

Fig. 6. Peroxidation of linoleic acid assessed measuring the malondialdehyde (MDA) production on the supernatant of SC-45, SC-45-pretreated and SC-45+GA (15 mg/ml) in a buffered suspension of linoleic acid after 1 h of incubation. The same experiment was carried out in the absence of the powder (blank).

week of treatment with pure water modified the surface decreasing the reactivity of the bioglass towards formate molecule, probably due to the oxidation of surface Fe2+ . Opposite to HO• generation, that may be catalyzed both by Fe2+ and Fe3+ , the homolytic rupture of the C–H bond in the formate anion is mainly triggered by iron in the reduced state (Fe2+ ), as shown by studies on toxic particulates [31,44]. The same experiment carried out with SC-45+GA evidenced that the bio-glass ceramic was inactive after 10 min of incubation. After 30 min the spectrum showed a weak signal due to [DMPO–CO2 ]•− adduct. The signal intensity increased with time, as evidenced by the EPR spectrum recorded after 60 min, whose intensity was similar to that observed with SC-45. The grafting with GA restores the former reactivity of SC-45 probably by reducing the iron centres previously oxidized by the prolonged contact with water. Accordingly, previous studies on iron-containing minerals (e.g. asbestos and lunar dust) showed that the reduction of Fe3+ to Fe2+ increases the ability to cleave the C–H bond of formate anion as well as to induce lipoperoxidation [29,44,45]. 3.4. Lipid peroxidation The potential of SC-45+GA to induce radical-mediated oxidative degradation of biomolecules was also assessed by evaluating its potency to induce cell-free lipid peroxidation. Linoleic acid was used as a model molecule for polyunsaturated fatty acid which naturally composes the cell membranes. Thiobarbituric acid (TBA) assay was used to quantify malonyldialdehyde as a product of linoleic acid peroxidation. SC-45, SC-45-pretreated, and SC-45+GA (15 mg/ml) were suspended in a buffered suspension of linoleic acid and the oxidative potential was measured after 1 h of incubation. In order to assess the entity of lipid peroxidation due to autoxidation reaction the same experiment was carried out on a buffered suspension of linoleic acid without powder (blank). The collected data, which are reported in Fig. 6, evidence that all the three samples were able to cause linoleic acid degradation, but with some differences in the amount of MDA produced. SC45 produced the larger amount of MDA. With SC-45-pretreated the amount of MDA detected was lower than with SC-45 indicating that the prolonged contact with water decreased the oxidative activity of the sample towards linoleic acid. With SC-45+GA the amount of MDA detected was significantly higher than that measured with SC-45 pretreated, indicating that the grafting of GA induces a partial reactivation of the surface sites, probably due to the reduction of

Surface functionalization of a ferrimagnetic glass-ceramic (SC45) with gallic acid (GA) has been performed. The effect on the bioactivity, solubility in simulated body fluids, and antioxidant/pro-oxidant potential of the material obtained on different molecular models of biologically relevant molecules has been investigated. Grafting GA did not influence the bioactivity of SC-45 and only minimally modified solubility in physiological solution. The main effect was observed to be on the redox activity. The ability of SC-45 to catalyze HO• radical release in the presence of H2 O2 was significantly increased by GA grafted onto the surface. Moreover, grafting GA restored the former reactivity of SC-45 towards both the formate ion and linoleic acid, which had been suppressed upon prolonged contact of the glass-ceramic with water during the grafting procedure. In both cases, GA acts as a pro-oxidant, probably reducing Fe3+ surf to Fe2+ surf . This phenomenon could represent the chemical basis for the development of a new generation of biomaterials with therapeutic properties. Authors’ disclosure statement The authors declare that they have no competing interests. Authors’ contribution I.C. participated in the conception, the experimental design, and the coordination of the study, carried out part of the experiments and contributed to the interpretation of the results, and the writing of the manuscript; M.T. participated in the conception of the study, the interpretation of the results, and the writing of the manuscript; F.T. participated in the interpretation of the results, and the writing of the manuscript; S.F. carried out the preparation of the samples, and contributed to the assessment of bioactivity, the interpretation of the results and the writing of the manuscript; E.B. contributed in the assessment of bioactivity, the interpretation of the results, and the writing of the manuscript; E.P. contributed to plan the equilibrium-based experiments, to interpret the data in a chemical model for the redox reactivity and to write the manuscript; E.V. participated in the conception and the coordination of the study, contributed to the experimental design, the interpretation of the results and the writing of the manuscript. All the Authors read and approved the final manuscript. Acknowledgments The authors are grateful to Professor Bice Fubini for her contribution in the conceivement of this study. The authors are also grateful to Ms. Micaela Orsi for participating to the experiments during her master’s degree thesis. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. References [1] W.P. Cao, L.L. Hench, Bioactive materials, Ceram. Int. 22 (1996) 493–507. [2] C. Vitale-Brovarone, G. Novajra, D. Milanese, J. Lousteau, J.C. Knowles, Novel phosphate glasses with different amounts of TiO2 for biomedical applications Dissolution tests and proof of concept of fibre drawing, Mater. Sci. Eng. C: Mater. 31 (2011) 434–442. [3] L.L. Hench, N. Roki, M.B. Fenn, Bioactive glasses: importance of structure and properties in bone regeneration, J. Mol. Struct. 1073 (2014) 24–30.

I. Corazzari et al. / Colloids and Surfaces B: Biointerfaces 148 (2016) 592–599 [4] S. Ferraris, M. Miola, Surface Modification of Bioactive Glasses, Bentham Sciences, e-Book, 2012. [5] S.M. Rabiee, N. Nazparvar, M. Azizian, D. Vashaee, L. Tayebi, Effect of ion substitution on properties of bioactive glasses: a review, Ceram. Int. 41 (2015) 7241–7251. [6] M. Miola, C.V. Brovarone, G. Maina, F. Rossi, L. Bergandi, D. Ghigo, S. Saracino, M. Maggiora, R.A. Canuto, G. Muzio, E. Vernè, In vitro study of manganese-doped bioactive glasses for bone regeneration, Mater. Sci. Eng. C: Mater. 38 (2014) 107–118. [7] E. Vernè, S. Ferraris, M. Miola, G. Fucale, G. Maina, G. Martinasso, R.A. Canuto, S. Di Nunzio, C. Vitale-Brovarone, Synthesis and characterisation of bioactive and antibacterial glass-ceramic part 1-microstructure, properties and biological behaviour, Adv. Appl. Ceram. 107 (2008) 234–244. [8] Y. Li, Y.Z. Liu, T. Long, X.B. Yu, T.T. Tang, K.R. Dai, B. Tian, Y.P. Guo, Z.A. Zhu, Mesoporous bioactive glass as a drug delivery system: fabrication, bactericidal properties and biocompatibility, J. Mater. Sci: Mater. Med. 24 (2013) 1951–1961. [9] T. Leventouri, A.C. Kis, J.R. Thompson, I.M. Anderson, Structure, microstructure, and magnetism in ferrimagnetic bioceramics, Biomaterials 26 (2005) 4924–4931. [10] Y. Ebisawa, F. Miyaji, T. Kokubo, K. Ohura, T. Nakamura, Bioactivity of ferrimagnetic glass-ceramics in the system FeO-Fe2 O3 -CaO-SiO2 , Biomaterials 18 (1997) 1277–1284. [11] D. Arcos, R.P. del Real, M. Vallet-Regi, A novel bioactive and magnetic biphasic material, Biomaterials 23 (2002) 2151–2158. [12] S.H. Oh, S.Y. Choi, Y.K. Lee, K.N. Kim, Research on annihilation of cancer cells by glass-ceramics for cancer treatment with external magnetic field. I. Preparation and cytotoxicity, J. Biomed. Mater. Res. 54 (2001) 360–365. [13] O. Bretcanu, E. Vernè, M. Coisson, P. Tiberto, P. Allia, Magnetic properties of the ferrimagnetic glass-ceramics for hyperthermia, J. Magn. Magn. Mater. 305 (2006) 529–533. [14] K. Riehemann, S.W. Schneider, T.A. Luger, B. Godin, M. Ferrari, H. Fuchs, Nanomedicine-challenge and perspectives, Angew. Chem. Int. Ed. 48 (2009) 872–897. [15] A. Tishin, Y. Spichkin, V. Zverev, P. Egolf, A review and new perspectives for the magnetocaloric effect: new materials and local heating and cooling inside the human body, Int. J. Refrig. 68 (2016) 177–186. [16] J. Chomoucka, J. Drbohlavova, D. Huska, V. Adam, R. Kizek, J. Hubalek, Magnetic nanoparticles and targeted drug delivering, Pharmacol. Res. 62 (2010) 144–149. [17] E. Vernè, M. Bruno, M. Miola, G. Maina, C. Bianco, A. Cochis, L. Rimondini, Composite bone cements loaded with a bioactive and ferrimagnetic glass-ceramic: leaching, bioactivity and cytocompatibility, Mater. Sci. Eng.: C 53 (2015) 95–103. [18] G. Kaur, O.P. Pandey, K. Singh, D. Homa, B. Scott, G. Pickrell, A review of bioactive glasses: their structure, properties, fabrication, and apatite formation, J. Biomed. Mater. Res. A 102 (2014) 254–274. [19] M. Zhu, J.H. Zhang, C.L. Tao, X. He, Y.F. Zhu, Design of mesoporous bioactive glass/hydroxyapatite composites for controllable co-delivery of chemotherapeutic drugs and proteins, Mater. Lett. 115 (2014) 194–197. [20] E. Vernè, M. Miola, S. Ferraris, C.L. Bianchi, A. Naldoni, G. Maina, O. Bretcanu, Surface activation of a ferrimagnetic glass-ceramic for antineoplastic drugs grafting, Adv. Eng. Mater. 12 (2010) B309–B319. [21] B.R. You, H.J. Moon, Y.H. Han, W.H. Park, Gallic acid inhibits the growth of HeLa cervical cancer cells via apoptosis and/or necrosis, Food Chem. Toxicol. 48 (2010) 1334–1340. [22] C.R. Dosier, C.P. Erdman, J.H. Park, Z. Schwartz, B.D. Boyan, R.E. Guldberg, Resveratrol effect on osteogenic differentiation of rat and human adipose derived stem cells in a 3-D culture environment, J. Mech. Behav. Biomed. Mater. 11 (2012) 112–122. [23] H. Piotrowska, M. Kucinska, M. Murias, Biological activity of piceatannol: leaving the shadow of resveratrol, Mutat. Res. Rev. Mutat. Res. 750 (2012) 60–82. [24] G. Oktem, A. Uysal, O. Oral, E.D. Sezer, M. Olukman, A. Erol, S.A. Akgur, A. Bilir, Resveratrol attenuates doxorubicin-induced cellular damage by modulating nitric oxide and apoptosis, Exp. Toxicol. Pathol. 64 (2012) 471–479. [25] E. Roldán, C. Sánchez-Moreno, B. de Ancos, M.P. Cano, Characterisation of onion (Allium cepa L.) by-products as food ingredients with antioxidant and antibrowning properties, Food Chem. 108 (2008) 907–916.

599

[26] S. Ferraris, X. Zhang, E. Prenesti, I. Corazzari, F. Turci, M. Tomatis, E. Vernè, Gallic acid grafting to a ferrimagnetic bioactive glass-ceramic, J. Non-Cryst. Solids 432 (2016) 167–175. [27] C. Locatelli, F.B. Filippin-Monteiro, T.B. Creczynski-Pasa, Alkyl esters of gallic acid as anticancer agents: a review, Eur. J. Med. Chem. 60 (2013) 233–239. [28] C.H. Yoon, S.J. Chung, S.W. Lee, Y.B. Park, S.K. Lee, M.C. Park, Gallic acid a natural polyphenolic acid, induces apoptosis and inhibits proinflammatory gene expressions in rheumatoid arthritis fibroblast-like synoviocytes, Joint Bone Spine 80 (2013) 274–279. [29] F. Turci, I. Corazzari, G. Alberto, G. Martra, B. Fubini, Free-radical chemistry as a means to evaluate lunar dust health hazard in view of future missions to the moon, Astrobiology 15 (2015) 371–380. [30] C.J. Horwell, G.W. Stannett, D. Andronico, A. Bertagnini, I. Fenoglio, B. Fubini, J.S. Le Blond, B.J. Williamson, A physico-chemical assessment of the health hazard of Mt. Vesuvius volcanic ash, J. Volcanol. Geotherm. Res. 191 (2010) 222–232. [31] M. Tomatis, L. Prandi, S. Bodoardo, B. Fubini, Loss of surface reactivity upon heating amphibole asbestos, Langmuir 18 (2002) 4345–4350. [32] O. Bretcanu, E. Vernè, M. Coisson, P. Tiberto, P. Allia, Temperature effect on the magnetic properties of the coprecipitation derived ferrimagnetic glass-ceramics, J. Magn. Magn. Mater. 300 (2006) 412–417. [33] A.L.B. Macon, T.B. Kim, E.M. Valliant, K. Goetschius, R.K. Brow, D.E. Day, A. Hoppe, A.R. Boccaccini, I.Y. Kim, C. Ohtsuki, T. Kokubo, A. Osaka, M. Vallet-Regi, D. Arcos, L. Fraile, A.J. Salinas, A.V. Teixeira, Y. Vueva, R.M. Almeida, M. Miola, C. Vitale-Brovarone, E. Vernè, W. Holand, J.R. Jones, A unified in vitro evaluation for apatite-forming ability of bioactive glasses and their variants, J. Mater. Sci.—Mater. Med. 26 (2015). [34] T. Kokubo, H. Takadama, How useful is SBF in predicting in vivo bone bioactivity? Biomaterials 27 (2006) 2907–2915. [35] I. Corazzari, S. Livraghi, S. Ferrero, E. Giamello, B. Fubini, I. Fenoglio, Inactivation of TiO2 nano-powders for the preparation of photo-stable sunscreens via carbon-based surface modification, J. Mater. Chem. 22 (2012) 19105–19112. [36] G. Stanciu, I. Sandulescu, B. Savu, S. Stanciu, K. Paraskevopoulos, X. Chatzistavrou, E. Kontonasaki, P. Koidis, Investigation of the hydroxyapatite growth on bioactive glass surface, J. Biomed. Pharm. Eng. 1 (2007) 34–39. [37] R.K. Singh, T.H. Kim, K.D. Patel, J.C. Knowles, H.W. Kim, Biocompatible magnetite nanoparticles with varying silica-coating layer for use in biomedicine: physicochemical and magnetic properties, and cellular compatibility, J. Biomed. Mater. Res. A 100A (2012) 1734–1742. [38] A. Balamurugan, G. Sockalingum, J. Michel, J. Faure, V. Banchet, L. Wortham, S. Bouthors, D. Laurent-Maquin, G. Balossier, Synthesis and characterisation of sol gel derived bioactive glass for biomedical applications, Mater. Lett. 60 (2006) 3752–3757. [39] B. Fubini, I. Fenoglio, Toxic potential of mineral dusts, Elements 3 (2007) 407–414. [40] T.P. Szatrowski, C.F. Nathan, Production of large amounts of hydrogen-peroxide by human tumour-cells, Cancer Res. 51 (1991) 794–798. [41] N. Mody, F. Parhami, T.A. Sarafian, L.L. Demer, Oxidative stress modulates osteoblastic differentiation of vascular and bone cells, Free Radical Biol. Med. 31 (2001) 509–519. [42] A.G. Darden, W.L. Ries, W.C. Wolf, R.M. Rodriguiz, L.L. Key, Osteoclastic superoxide production and bone resorption: stimulation and inhibition by modulators of NADPH oxidase, J. Bone Miner. Res. 11 (1996) 671–675. [43] M.J. Hynes, M.O. Coinceanainn, The kinetics and mechanisms of the reaction of iron(III) with gallic acid, gallic acid methyl ester and catechin, J. Inorg. Biochem. 85 (2001) 131–142. [44] M. Turci, I.G. Tomatis, N. Lesci, The iron-related molecular toxicity mechanism of synthetic asbestos nanofibres: a model study for high-aspect-ratio nanoparticles, Chem.—Eur. J. 17 (2011) 350–358. [45] M. Tomatis, F. Turci, R. Ceschino, C. Riganti, E. Gazzano, G. Martra, D. Ghigo, B. Fubini, High aspect ratio materials: role of surface chemistry vs. length in the historical long and short amosite asbestos fibers, Inhal. Toxicol. 22 (2010) 984–998.