Chem. Res. Toxicol. 2000, 13, 489-500
489
Pure-Silica Zeolites (Porosils) as Model Solids for the Evaluation of the Physicochemical Features Determining Silica Toxicity to Macrophages Ivana Fenoglio,† Antonietta Croce,‡ Francesco Di Renzo,§ Roberta Tiozzo,‡ and Bice Fubini*,† Dipartimento di Chimica Inorganica, Chimica Fisica e Chimica dei Materiali, Facolta` di Farmacia, Universita` di Torino, via Pietro Giuria 7, 10125 Torino, Italy, Dipartimento di Scienze Biomediche, Sezione di Patologia Generale, Facolta` di Medicina, Universita` di Modena, Via Campi 287, 41100 Modena, Italy, and Laboratoire de Mate´ riaux Catalytiques et Catalyse en Chimie Organique, UMR 5618 CNRS, ENSCM Montpellier, France Received September 30, 1999
The interaction between inhaled particles and alveolar macrophages plays a key role in silicarelated diseases. It has been previously shown [Fubini, B., et al. (1999) Chem. Res. Toxicol. 12, 737-745] that a monocyte-macrophage cell line (J774) may be employed in the evaluation of the degree of cytotoxicity to alveolar macrophages of various silica dusts. In this paper, pure-silica zeolites (porosils) in microcrystalline form have been employed as “model solids” in an effort to show which physicochemical properties of the silica particle are playing a major role in the toxicity to macrophages. The samples employed covered four different porosil crystal structures (MFI, FAU, TON, and MTT) and also include a synthetic rodlike cristobalite (CRISrd). When compared at equal weight, the samples cover a wide range of cytotoxicity from inert to toxic as unheated mineral cristobalite [Fubini, B., et al. (1999) Chem. Res. Toxicol. 12, 737745]. Mild grinding did not affect cytotoxicity. Calcined (open pores) and uncalcined (pore filled with template) TON exhibited the same cytotoxicity, indicating that only the outer surface is implied. The hydrophobic and/or hydrophilic character of TON, evaluated by adsorption calorimetry, is close to what has been previously found for silicalite and is consistent with a hydrophilic outer surface and hydrophobic pore walls. The potential for generating hydroxyl radicals from hydrogen peroxide varies among the various porosils that have been studied. A model is proposed for the correlation between inhibition of growth on proliferating cells and physicochemical properties varying from one to the other sample. The extent of external surface and the aspect ratio were related to the intensity of the cytotoxic effect, while the level of radical release was not. This suggests, on one hand, that comparison of toxicity among various dusts should be made at equal particle surface and, on the other, that in the model studied, free radical release does not play a crucial role in the primary event of toxicity to alveolar macrophages.
Introduction Silicosis, one of the most ancient occupational diseases, is caused by the inhalation of crystalline silica particles in respirable size. The International Agency for Research on Cancer declared in 1997 that quartz and cristobalite, two of the most common silica polymorphs, are potentially carcinogenic to humans (1). Furthermore, the development of some autoimmune diseases is often associated with the inhalation of silica dusts (2). Despite massive experimental work in the field, reported in the IARC Monograph (1) and in some recent reviews (3-6), the mechanism of action at the molecular level is still poorly understood. The various sources of silica dusts appear to have different pathogenic potential, related either to the kind of silica polymorph or to external factors associated with the silica particle, namely, im* Corresponding author. Telephone: 011/6707566. Fax: 011/ 6707855. E-mail:
[email protected]. † Universita ` di Torino. ‡ Universita ` di Modena. § UMR 5618 CNRS.
purities (1). The variability of “quartz hazard” and consequent difficulties in risk assessment have been discussed in recent papers (7, 8). The pathogenic mechanism is still undefined at the molecular level mainly because silica, as a particulate toxicant, is not eliminated or metabolized, but acts at different stages. Different surface functionalities are implied in the subsequent stages. Long-term diseases arise from unaltered silica particles active for extremely long periods of time in the body. Because of the substantial covalency of the siliconoxygen bond, silica exists in nature (9) and may be prepared in a large variety of forms. In an international conference, silica characteristics and uses of the various forms have been recently discussed (10). Silicas exhibit different surface properties (5, 6, 9), and some of them have been found to elicit different biological responses (1, 5, 11, 12). The major forms of silica are reported in Table 1 which also reports indications on their pathogenicity when available. In some cases, their toxicity is still under debate or never properly tested. The natural
10.1021/tx990169u CCC: $19.00 © 2000 American Chemical Society Published on Web 05/16/2000
490
Chem. Res. Toxicol., Vol. 13, No. 6, 2000
Fenoglio et al.
Table 1. Existing Silica Polymorphs and Pathogenicitya crystalline natural
artificial
quartz tridymite cristobalite coesite stishovite silica-based zeolites
pathogenicity + + + ( never tested
amorphous
pathogenicity
biogenic origin (diatomaceous earth, rice husks) volcanic origin (vitreous silica)
( (
chemically prepared silicas (precipitated and pyrogenic silica) ground silica glass
(
a Polymorphs are classed as active (+), not active (-), or both ((), and the subject is debated with respect to their fibrogenic and/or carcinogenic activity.
crystalline forms are clearly the most pathogenic ones, but some amorphous forms are also suspect. The toxicity of the artificial crystalline forms has never been tested, likely because massive exposition to these materials, mainly used as adsorbents and molecular sieves (13), never took place. Crystalline polymorphs are usually present in nature as macrocrystals. They may occur in respirable size either when ground or when amorphous silica dusts are submitted to thermal treatments. The latter case comprises biogenic cristobalite, originated from diatomaceous earth, or quartz particles, generated in fly ashes from amorphous silica microparticles. The origin of the dust markedly defines the state of the particle surface. Depending upon their origin, particles of the same polymorph may exist with different surface properties and pathogenic potential (5, 7, 8, 14). To unravel such a complicated matter, we propose to examine the biological response to silicas, differing in one (or few) property from each other, to evaluate the effect of that very property on the response elicited. In a previous paper, the J774 monocyte-macrophage cell line was employed to establish a relationship between hydrophilicity and cytotoxicity of variously heated cristobalite (14). The effect of the dusts on cell survival and lactate dehydrogenase (LDH)1 release was measured on confluent and proliferating J774 cultures. The results were compared with what was obtained with a different cell line (AE6) and with rat alveolar macrophages. All tests classed the samples in the same order of toxicity. In this paper, we report data on the cytotoxicity to the monocyte-macrophage cell line (J774) elicited by some artificial crystalline silicas (porosils) with similar properties but differing in micromorphology, size, and surface area. Cytotoxicity per se does not imply fibrogenicity (15, 16) or carcinogenicity. The alveolar macrophage, however, plays a key role in silica-related diseases (1, 7, 15) either by clearing the particle out of the lung or by causings following a continuous cycle of ingestion, cell activation, death, and recruitmentsa status of chronic inflammation with release of oxidants, cytokines, and proteases in the surrounding medium and in proximity to target cells. The link between particle-macrophage interaction and development of silica-related diseases is illustrated in Figure 1. The cytotoxicity of the silica particle to macrophages may determine the intensity of inflammation 1 Abbreviations: LDH, lactate dehydrogenase; DMEM, Dulbecco’s modified Eagle’s medium; Ads I and II, first and second adsorption runs, respectively; FAU, faujasite; TON, theta-1; MTT, ZSM-23; MFI, ZSM-5; SEM, scanning electron microscopy; FD, framework density; BET, Braunauer, Emmet, and Teller; ROS, reactive oxygen species; RNS, reactive nitrogen species; PMN, polymorphonuclear leukocyte; DMPO, 5,5-dimethyl-1-pirroline N-oxide.
and the time of residence of the particle within the macrophages and in the lung (5-8). In this respect, it will be relevant to the overall pathogenic process. Indeed, the two substances employed to inhibit silica fibrogenicity, aluminum salts and PVPNO (6, 17-19), have been reported to lower silica-induced cytotoxicity (20, 21). The general term “porosils” encompass various forms of zeolites prepared via template-assisted hydrothermal synthesis from pure silica. Opposite to common zeolites, they consist of pure SiO4 tetrahedra linked to each other having been crystallized in the absence of aluminum (22). They differ from each other in crystal structure and spatial arrangement of tetrahedral units. In comparison to the dusts obtained with natural polymorphs, porosils offer the following characteristics. (i) They are prepared in micron-size form so that they do not need any processingsgrinding, heating, etc.sto be obtained in respirable size. (ii) They exhibit a surface made up of natural crystal faces. (iii) They may be prepared in different forms and sizes. (iv) They may be prepared with a high level of purity, never attained with mineral specimens, unless the minerals are purified with complex procedures. We have therefore chosen porosils as “model solids” for silica toxicity. In this paper, we report on the effect of size, form, and surface area on cytotoxicity, which is relevant both to the understanding of the physicochemical properties determining the cytotoxicity of a given material and to the choice of the appropriate parameter in in vitro and in vivo dose-response studies, when comparing the effect of different kinds of particulate samples.
Experimental Procedures Materials. For the preparation of the various porous silicas, the following materials have been employed: colloidal silica Ludox HS-40 from Dupont (95% SiO2, 0.02% Al, and 0.46% Na), pyrogenic silica from Serva (99% SiO2,
