Does Vitreous Silica Contradict the Toxicity of the Crystalline Silica

Jan 19, 2010 - associated with VS, whether the paradigm of crystallinity as a prerequisite factor for silica toxicity still holds. Here, we report a s...
0 downloads 0 Views 1MB Size
620

Chem. Res. Toxicol. 2010, 23, 620–629

Does Vitreous Silica Contradict the Toxicity of the Crystalline Silica Paradigm? Mara Ghiazza,†,‡,§ Manuela Polimeni,†,§,| Ivana Fenoglio,†,‡ Elena Gazzano,†,| Dario Ghigo,†,| and Bice Fubini*,†,‡ Interdepartmental Center “G. Scansetti” for Studies on Asbestos and Other Toxic Particulates and Department of Chemistry IFM and Interdipartmental Centre for Nanostructured Interfaces and Surfaces, UniVersita` degli Studi di Torino, Via Pietro Giuria 7, 10125 Torino, Italy, and Department of Genetics, Biology and Biochemistry, UniVersita` degli Studi di Torino, Via Santena 5/bis, 10126 Torino, Italy ReceiVed October 9, 2009

“Vitreous silica” is a particular form of amorphous silica, much neglected in experimental studies on silica toxicity. In spite of the incorrect term “quartz glass”, often employed, this material is fully amorphous. When reduced in powdered form by grinding, the particulate appears most close to workplace quartz dust but, opposite to quartz, is not crystalline. As silicosis and lung cancer are also found among workers exposed to “quartz glass”, the question arises of whether crystallinity is the prerequisite feature that makes a silica dust toxic. We compare here the behavior of comminuted quartz, vitreous silica, and monodispersed silica spheres, as it concerns surface reactivity and cellular responses involved in the accepted mechanisms of silica toxicity. Care was taken to choose samples of extreme purity, to avoid any effect due to trace contaminants. Quartz and vitreous silica, opposite to silica spheres, show irregular particles with sharp edges, stable surface radicals, and sustained release of HO• radicals via a Fentonlike mechanism. The evolution of the heat of adsorption of water as a function of coverage shows with quartz and vitreous silica a similar pattern of strong hydrophilic sites, nearly absent on the other silica specimen. When tested on a macrophage cell line (MH-S), vitreous silica and pure quartz, but not the monodispersed silica spheres, showed a remarkable potency in cytotoxicity, nitric oxide synthase activation and release of nitrite, and tumor necrosis factor-R production, suggesting a common behavior in inducing an oxidative stress. All of the above features appear to indicate that crystallinity might not be a necessary prerequisite to make a silica particle toxic. 1. Introduction Silicosis, one of the most ancient occupational diseases, has been associated for decades with exposure to dusts generated by some crystalline silica polymorphs. Also, lung cancer and autoimmune pathologies appear, so far, only related to exposure to some crystalline forms of silica (1–3). However, in the latter case, a role of amorphous silica has not been completely ruled out. The differences between crystalline and amorphous specimens were also highlighted in experimental studies. When comparing the effects of crystalline and amorphous silica dusts in vitro and in vivo, the amorphous specimens were much less active than the crystalline ones, mostly biologically inert (4). They showed some cytotoxic effects on cells (5, 6), but in vivo, the inflammatory response, which was sustained with crystalline polymorphs, was transient with the amorphous forms (7–9). It has to be pointed out, however, that the amorphous specimens employed were always chemically prepared silica powders, made up of very fine smooth particles, obtained by precipitation * To whom correspondence should be addressed. Tel: +39 011 670.75.66. Fax: +39 011 670 0.75.77. E-mail: [email protected]. † Interdepartmental Center “G. Scansetti” for Studies on Asbestos and other Toxic Particulates. ‡ Department of Chemistry IFM and Interdipartmental Centre for Nanostructured Interfaces and Surfaces, Universita` degli Studi di Torino. § These authors contributed equally to this work. | Department of Genetics, Biology and Biochemistry, Universita` degli Studi di Torino.

from aqueous solutions with a great variety of procedures or by combustion in air of silicon compounds (pyrogenic silica). We consider here “vitreous silica” (VS)1 or “silica glass” or “quartz glass”, a particular form of amorphous silica, much neglected in experimental studies on silica toxicity. In spite of the term quartz, erroneously employed, this material is fully amorphous from the standpoint of crystallinity. Quartz glass is the common product of rapid solidification of any molten silica. It is noteworthy that the dust generated by VS shares, with most quartz dusts, the form of the particles and the procedure whereby it is obtained in particulate form. Exposure to VS dusts in glass factories is not uncommon in the workplace, particularly among glass blowers, who daily handle silica solidified from the melt. Several cases of silicosis and lung cancer have been registered among workers involved in such activities (10–14), but epidemiological studies on this particular case are somewhat lacking. A question arises on whether health effects could be due not only to the quartz powders occasionally handled in these circumstances but also to the heavy exposure to silica glass particles.

1 Abbreviations: XRD, X-ray diffraction; XPS, X-ray photoelectron spectroscopy; SEM, scanning electron microscopy; EPRS, electron paramagnetic resonance spectroscopy; DMPO, 5,5-dimethyl-pirroline-N-oxide; LDH, lactate dehydrogenase; NO, nitric oxide; NOS, nitric oxide synthase; Qz, pure quartz; VS, vitreous silica; MSS, monodispersed silica spheres; MH-S, murine alveolar macrophages; FITC, fluorescein isothiocyanate; PI, propidium iodide; TNF-R, tumor necrosis factor-R; IL-1β, interleukin-1β.

10.1021/tx900369x  2010 American Chemical Society Published on Web 01/19/2010

Potential Toxicity of Vitreous Silica

To gain information on the hazard associated with exposure to VS dust, we have prepared and/or characterized three “model” silica samples: • one pure quartz (Qz) • one extremely pure silica glass (VS) • one monodispersed synthetic amorphous silica, made up of spheres (MSS), whose diameter is close to the size of the smallest particles obtained by grinding VS thus differs in one single characteristic from the other two. It shares a common origin (grinding) with quartz and the amorphous nature with the synthetic form. This set of samples appears the most appropriate to investigate, beside the hazard associated with VS, whether the paradigm of crystallinity as a prerequisite factor for silica toxicity still holds. Here, we report a study on some physicochemical properties and cellular responses involved in crystalline silica toxicity (1, 15–22).

2. Experimental Procedures 2.1. Silica Samples. VS was obtained by grinding in a ball mill (agate jar) a very pure silica glass (Suprasil) produced for optical applications. The grinding process was prolonged for 3 h, to have a sample with the size, micromorphology, and surface area close to those typical of commercial quartz dusts (23). Qz was obtained by grinding in a ball mill (agate jar) a very pure natural quartz crystal from Madagascar for 12 h. A longer grinding process than for VS was required to obtain the size distribution and specific surface area close to what obtained for VS, as a Qz macrocrystal is much harder to grind. The amorphous silica (Ångstro¨m sphere) made up by MSS was purchased from Fiber Optic Center Inc. (New Bedford, MA). All of the original materials were free of any trace contaminant to avoid any effect not arising from the intrinsic nature of the materials. The grinding process was performed in an agate jar to keep the samples free from acquired impurities, as may occur with steel or hard metal (WC/Co) jars. 2.2. Chemical Reagents. When not otherwise specified, reagents were from Sigma-Aldrich srl (Milan, Italy). 2.3. X-ray Diffraction (XRD). XRD spectra of the powdered materials were collected on a Philips PW1830, with θ-2θ geometry and CoKR radiation, in the (10-85°) 2θ range, with a step width 2θ ) 0.02 and time per step ) 2 s. The patterns obtained were compared with those reported in the J.C.PDS (Joint Committee of Powder Diffraction Standard) archives. 2.4. Specific Surface Area. The surface area was evaluated by means of the Braunauer, Emmet, and Teller (BET) method using a ASAP 2020 apparatus (Micromeritics, United States). Because of the relatively low specific surfaces, Kr adsorption at -196 °C was employed. 2.5. Scanning Electron Microscopy (SEM). The samples were examined by means of SEM in the secondary electron imaging mode (Stereo Scan 420 Leica) at two different magnifications to evaluate in detail micromorphology and the presence of smaller particles. Rough information on the polydispersity of silica particles was obtained from these pictures. 2.6. Electron Paramagnetic Resonance Spectroscopy (EPRS). The EPR spectra of the powdered silicas were recorded in vacuo at 77 K on a Bruker EMX spectrometer operating in the X-band mode (9.5 GHz) following a technique reported in previous studies (24). The spectra were recorded with the following instrument settings: scan range, 400 G; receiver gain, 1 × 104; microwave power, 10 mW; modulation amplitude, 1G; scan time, 80 s; and three scans. 2.7. Free Radical Detection. Free radical release was monitored by means of EPR spectroscopy (Miniscope 100 EPR spectrometer, Magnettech) using the spin trapping technique [5,5-dimethylpirroline-N-oxide (DMPO) as the spin trapping molecule] with a procedure largely described in previous studies (25, 26).

Chem. Res. Toxicol., Vol. 23, No. 3, 2010 621

2.7.1. Release of HO• Radicals from Hydrogen Peroxide. A 150 mg amount of each powder was suspended in 500 µL of buffered solution (0.5 M potassium phosphate buffer, pH 7.4) and 250 µL of 0.15 M DMPO. The reaction was initiated by adding hydrogen peroxide (500 µL of 0.20 M solution in distilled water) to the silica particles suspension, and the radical yield was progressively measured on an aliquot of 50 µL of the suspension up to 1 h. All experiments were repeated at least twice. 2.7.2. Release of COO• Radicals from the Formate Ion. A 75 mg amount of each powder was suspended in 250 µL of 0.15 M DMPO and 250 µL of formate ion (1.0 M solution in 0.2 M potassium phosphate buffer, pH 7.4). The reaction was initiated by adding a target molecule to the silica particles suspension, and the radical yield was progressively measured on an aliquot of 50 µL of the suspension up to 1 h. All experiments were repeated at least twice. 2.8. Adsorption Calorimetry. The distribution of hydrophilic and hydrophobic surface sites was evaluated by means of adsorption calorimetry by making the powdered silicas come into contact with water vapor following a technique extensively described in previous studies (27–29). The heat of adsorption of water vapor was measured on a Tian-Calvet microcalorimeter (Setaram) connected to a volumetric apparatus, which allowed simultaneous measurement of the adsorbed amount (uptake, na), heat released (Q), and equilibrium pressure (p) for small increments of water vapor in contact with the silica samples. Doses of the adsorptive were subsequently admitted onto the sample, and the pressure was continuously monitored by means of a 0-100 Torr transducer gauge (Baratron MKS). To have a surface free of contaminants adsorbed from the environment, the samples were outgassed in the calorimetric cells for 2 h at 150 °C and subsequently located in the calorimetric vessel without exposure to the atmosphere. The temperature of the calorimeter was maintained at 30 °C throughout the adsorption experiment. After the first adsorption, water vapor was removed, and a second adsorption run was performed. Differences between the results of the two runs indicate the occurrence of an irreversible uptake of water. 2.9. Cells. Murine alveolar macrophages (MH-S), provided by Istituto Zooprofilattico Sperimentale “Bruno Ubertini” (Brescia, Italy), were cultured in 35 mm diameter Petri dishes in RPMI1640 + 10% FBS up to 90% confluence and then incubated in the same culture medium for 24 h in the absence or presence of silica before the assays. The protein content of the cell monolayers was assessed with the BCA kit from Pierce (Rockford, IL). 2.10. Lactate Dehydrogenase (LDH) Activity. The cytotoxic effect of the silica powders was measured as the leakage of LDH activity from the cell into the extracellular medium (30, 31). After each incubation, the extracellular medium was collected and centrifuged at 13000g for 30 min. The cells were washed with fresh medium, detached with trypsin/ethylenediaminetetraacetic acid (EDTA) (0.05/0.02% v/v), washed with phosphate-buffered saline (PBS), resuspended in 1 mL of TRAP (82.3 mM triethanolamine, pH 7.6), and sonicated on ice with two 10 s bursts. Aliquots of cell lysate (5 µL) and extracellular medium (50 µL) were diluted with TRAP and supplemented with 0.5 mM sodium pyruvate and 0.25 mM NADH (final volume of the mix, 300 µL) to start the reaction. The reaction was followed for 10 min, measuring the absorbance at 340 nm (37 °C) with a Synergy HT microplate reader (Bio-Tek Instruments, Winooski, VT). Each kinetic reaction was linear throughout the time of measurement. Both intracellular and extracellular enzyme activities were expressed as µmol of NADH oxidized/min/dish, and then, the extracellular LDH activity (LDH out) was calculated as a percentage of the total (intracellular + extracellular) LDH activity (LDH tot) in the dish. 2.11. Detection of Apoptosis. The induction of apoptosis or necrosis was performed as previously described (32). After a 24 h incubation in the absence or presence of silica powders, the cells were washed with PBS, detached with trypsin/EDTA, added to floating cells previously collected from the supernatant, and resuspended at 3 × 105 cells/0.5 mL of binding buffer [containing 10 mM N-(2-hydroxyethyl)piperazine-N′-ethanesulfonic acid (Hepes),

622

Chem. Res. Toxicol., Vol. 23, No. 3, 2010

Ghiazza et al.

Table 1. Physicochemical Characteristics of the Silica Samples sample

surface area (m2/g)

VS Qz MSS

4.1 ( 0.011 5.0 ( 0.015 4.4 ( 0.0185

a

particle size (µm) a

80% of particles ranging from 0.5 to 5 80% of particles ranging from 0.5 to 5a 0.7-1.4b

crystallinity XRD

surface topology SEM

amorphous crystalline amorphous

indented irregular indented irregular smooth

atoms at the surface XPS Si, O Si, O ND

Measured by counting particles in the SEM picture. Qualitative determination. b As declared by the supplier.

2.5 mM CaCl2, and 140 mM NaCl, pH 7.4]. These cell suspensions were incubated for 10 min with 25 µg of propidium iodide (PI, Calbiochem-Novabiochem Corp., La Jolla, CA) and 5 µL of annexin V-fluorescein isothiocyanate (FITC) (0.05 mg/mL). Cells were washed, and fluorescence was measured at 488 (excitation)-530 nm (emission) for annexin V or 536 (excitation)-617 nm (emission) for PI detection using a Synergy HT microplate reader. Results were expressed as average percentage values obtained from triplicate experiments versus controls. 2.12. Nitric Oxide (NO) Synthesis. After a 24 h incubation in the absence or presence of silica samples, the culture supernatant was removed and tested for nitrite, which is a stable derivative of NO, using the Griess method as previously described (33). The nitrite amount was corrected for the content of cell proteins, and results were expressed as nmol/mg cellular protein. 2.13. NO Synthase (NOS) Activity. After a 24 h incubation under the experimental conditions described in the Results, cells were detached by trypsin/EDTA, washed with PBS, resuspended in 0.3 mL of Hepes/EDTA/dithiothreitol (DTT) buffer (20 mM Hepes, 0.5 mM EDTA, and 1 mM DTT, pH 7.2), and sonicated on crushed ice with two 10 s bursts. The NOS activity was measured on 100 µg of cell lysates with the Ultrasensitive Colorimetric Assay for Nitric Oxide Synthase kit (Oxford Biomedical Research, Oxford, MI). This method employed a NADPH recycling system that permitted NOS to catalyze NO production at a constant level for many hours. The stable NO degradation product nitrite accumulated during this period was determined using the Griess reagent. Results were expressed as nmol nitrite/min/mg cellular protein. 2.14. Measurement of Tumor Necrosis Factor-r (TNF-r) and Interleukin-1β (IL-1β). After a 24 h incubation with silica samples, the extracellular medium was collected and centrifuged at 13000g for 30 min. The supernatant concentration of each cytokine was determined by using the Conventional ELISA kits from Bender MedSystems (Vienna, Austria), following the manufacturer’s instructions. The absorbance was measured at 450 nm with a Synergy HT microplate reader. The cytokines amount was corrected for the content of cell proteins, and results were expressed as pg/mL/mg cellular protein. 2.15. Statistical Analysis. All data in text and figures are provided as means ( standard errors of the mean (SEMs). The results were analyzed by a one-way analysis of variance (ANOVA) and Tukey’s test (software: SPSS 11.0 for Windows, SPSS Inc., Chicago, IL). p < 0.05 was considered significant.

3. Results 3.1. Physicochemical Characteristics of the Silica Samples. The major physicochemical characteristics of the three silica powders are summarized in Table 1. The size of MSS was chosen to match with the smaller particles of VS and Qz. Nevertheless, the surface area is lower than VS and Qz. This may be ascribed both to the smoothness of their surface and to the roughness at the nanometric level of the particles obtained by grinding. 3.1.1. Crystallinity. The XRD patterns of the three samples are compared in Figure 1. As expected, only Qz exhibits the typical XRD spectrum of the R-crystalline polymorph. The sharp peaks indicate that the grinding process has not significantly modified the ordered crystal structure. The spectrum of the vitreous and monodispersed silicas only exhibits the halo centered at 20° (2θ scale), typical of amorphous silica forms (34).

Figure 1. XRD spectra of VS, Qz, and MSS in the 10-100 2θ range.

3.1.2. Surface Purity. VS and Qz have been tested by means of the X-ray photoelectron spectroscopy (XPS) technique to check the purity of the samples and to confirm that no surface impurities were acquired during the grinding process. The XPS spectra [M-Probe Instrument (SSI) equipped with a monochromatic Al KR source (1486.6 eV)] of VS and Qz consist of only two lines at 103.5 (Si2p) and 532.6 eV (O1s), indicating that both samples are composed at the surface of Si and O atoms only (Claudia Bianchi, University of Milan, personal communication). MSS have a purity >99.9%, as declared from the supplier. 3.1.3. Particle Micromorphology and Polydispersity. SEM images of the three silica samples are reported in Figure 2 at different magnifications. VS (Figure 2A,B) and Qz (Figure 2C,D) have a similar, very heterogeneous, distribution of particle size, which ranges from 10 to below 1 µm, with a predominant presence of particles ranging from 0.5 to 5 µm (data not shown). Most particles exhibit irregular shapes, very sharp edges, and acute spikes, a morphology characteristic of silica dusts obtained by grinding. It was not possible to evaluate the polydispersity of particles by using an image analyzer or dynamic light scattering (DLS), because in both samples there was evidence of several small particles sticking to the bigger ones, likely due to the presence of surface charges generated during the grinding process. Monodispersed silica, reported in Figure 1E,F, is composed by perfectly spherical particles ranging from 0.7 to 1.4 µm in diameter, very closely packed in a nearly regular way, according to the supplier’s declaration. 3.2. Surface Properties Related to Silica Toxicity. 3.2.1. Dangling Bonds/Surface Radicals Generated by Grinding. The surface radicals generated at the cleaved crystal planes on quartz have been extensively described by some of us in previous studies (19, 24, 35, 36). The signal of commercial quartz dusts is usually a complex signal because of the superimposition of silicon and oxygen radicals and other paramagnetic species arising from trace contaminants (37), while the signal of pure silica is only due to silicon and oxygen species. Accordingly, as shown in Figure 3A, VS and Qz only exhibit the signals corresponding to silicon and oxygen-centered radicals

Potential Toxicity of Vitreous Silica

Chem. Res. Toxicol., Vol. 23, No. 3, 2010 623

Figure 2. SEM micrographs of (A and B) VS, (C and D) Qz, and (E and F) MSS at two different magnifications (1000× and 2000×).

Figure 3. (A) EPR spectra of the powdered silica particles recorded at 77 K. Instrumental settings: sweep width, 400 G; receiver gain, 104; modulation amplitude, 10 mW; scan time, 80 s; three scans; center field of 3030 G; and power, 10 mW. (B) Free radical-generating surface centers and their reaction with atmospheric oxygen. (44) Adapted with permission from Fubini et al. ( 2001) J. Environ. Pathol., Toxicl. Oncol. 20, 87-100.

(at g values of 2.06 and 2.04, respectively), confirming what was previously reported (15, 24). The spectrum of VS is somehow even more intense than that of Qz, either because more dangling bonds were generated during the grinding process or because of a difference in decay kinetics (38).

Silicon/oxygen radicals and surface charges (not detectable with EPR spectroscopy) are generated during the grinding process by homolytic and heterolytic cleavage (Figure 3). These reactive species in part rapidly decay (38) but also tend to react with atmospheric oxygen following the scheme in Figure 3B

624

Chem. Res. Toxicol., Vol. 23, No. 3, 2010

Ghiazza et al.

Figure 4. HO• generation by VS, Qz, and MSS from hydrogen peroxide via Fenton reaction after 60 min of incubation with the spin trap DMPO (0.05 M) and hydrogen peroxide (78 mM).

(19), yielding radical active species at the surface, somehow stable with time. Surface radicals are thus present at the surface of VS even in a larger amount than on Qz but are absent at the surface of MSS. All spectra do not exhibit other paramagnetic centers, further confirming the purity of the samples. 3.2.2. Generation of Free Radicals. All samples were tested by means of the spin trapping technique for their potential to generate HO• radicals from hydrogen peroxide (Fenton-like activity) and carbon-centered radicals following the homolytic cleavage of a hydrogen-carbon bond. The ESR spectra of the DMPO/HO• adduct obtained after contact of the three samples with a solution of hydrogen peroxide are compared in Figure 4. VS and Qz are able to generate hydroxyl radicals via a Fenton-like reaction, while MSS are completely inactive. None of the three samples cleaved the C-H bond of the formate ion employed as a target molecule (data not reported), in agreement with previous results. Such activity was in fact found to be absent in pure specimens but present only in most commercial quartz dusts, thus relatable to surface contaminants (26, 39). 3.2.3. Energy of Interaction with Water and Hydrophilicity. To examine in detail whether Qz and VS differ in surface reactivity, we have examined the characteristics of their reaction with water vapor by means of adsorption microcalorimetry, a technique enabling the detection of hydrophilic/ hydrophobic sites and their distributions at the surface. The curves in Figure 5, which represent the interaction energy as a function of the equilibrium pressure of water vapor, indicate the abundance of hydrophilic and hydrophobic sites (energy of interaction higher or lower than 44 kJ/mol, heat of liquefaction of water). Figure 5 illustrates the evolution of the interaction energy, with the progressive coverage of the surface by water molecules (differential heat vs equilibrium pressure). Each experimental point corresponds to a dose of either the first adsorption run (full points) or the successive readsorption following desorption (empty points). Water is adsorbed mostly reversibly on all samples with no substantial difference between the first and the second adsorption run. The massive irreversible uptake of water by some commercial quartz dusts reported in a previous study was in fact mostly ascribed to contaminants (23). Accordingly, no irreversible water adsorption takes place on pure silica dusts. VS and Qz exhibit the same evolution of the interaction energy, typical of very heterogeneous adsorption sites (Figure 5). A fraction of hydrophilic sites (corresponding to the adsorption of 5.5 µmol/m2), fill in at low coverage with decreasing heat of adsorption (from 90 to 44 kJ/mol), followed by adsorption on hydrophobic patches (