Document a0ZVwJK4D5B8MQ3K4rO28mn9

FILE NAME: Talc (TALC) DATE: 1992 DOC#: TALC173 DOCUMENT DESCRIPTION: Journal Article - Biological Effects of Inhaled Minerals SluXUr\< y Q j). /francas) . \fcr(, : pp>. ^22-5 "2 ^ 3 . American Mineralogist, Volume 77, pages 225-243, 1992 Biological effects of inhaled minerals G eorge D. G uth rie, Jr. Geology/Geochemistry, MS D469, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, U SA Abstract Numerous studies present data on the biological effects of inhaled minerals, but these data are often disseminated within reports that primarily address the asbestos minerals. Furthermore, these reports are normally published in journals that are unfamiliar to most minerals scientists. This review compiles these data in order to facilitate an understanding of the known biological effects of minerals. An introduction to the types of studies from which the data were drawn is given so that those unfamiliar with such studies can assess the data critically. In general, minerals exhibit a range of biological activities from apparently inactive or slightly active, such as hematite, to highly fibrogenic and carcinogenic, such as fibrous brucite ("nemalite"). The zeolites also exhibit such a range, with some mordenite being slightly active and erionite being highly active; however, erionite is the only zeolite that has been studied extensively. Although several mechanisms have been proposed to explain how minerals induce dis ease, it is still unclear why minerals exhibit a range in biological activity. The diversity of mineral species holds great potential for probing these mechanisms, especially when mineralogical data are integrated with biological data. Unfortunately, many of the studies reporting data on the biological effects of inhaled minerals fail to report detailed mineralogical information; hence, it is difficult at present to interpret the biological activities of minerals in terms of their physical and chemical properties. More collaboration between minerals scientists and health scientists would benefit this area of research by enabling an integration of mineralogical and biological data. Important mineralogical data that are only rarely considered in biological research include exact mineral content of the specimen (i.e., identification and abundance of contaminants), physical and chemical properties of minerals, and surface properties of minerals. I ntroduction Because of their potential to induce a number of lung diseases (e.g., fibrosis, lung cancer, and mesothelioma), the asbestos minerals, fibrous serpentine (chrysotile) and fibrous amphiboles, have been the focus o f numerous ex perimental studies, governmental regulations, and exten sive public concern. Early studies of riebeckite-asbestos miners (Wagner et al., 1960) revealed an association be tween exposure to riebeckite asbestos and mesothelioma, a rare cancer with an extremely high morbidity rate. Sub sequently, a higher than expected incidence of mesothe lioma was found in American asbestos workers who were exposed primarily to chrysotile (Selikoffet al., 1964,1965) but also to amphiboles (Ross, 1984), which occurred more frequently in the lungs of these workers (Langer and No lan, 1989). The widespread use of the asbestos minerals meant that a large population was exposed and poten tially at risk, and so began the proliferation of research on the biological effects of asbestos. Fear of asbestos exposure led to a replacement of the asbestos minerals with substitutes and a skewing of the research on the biological effects of minerals toward the 0003-004X792/0304-0225S02.00 asbestos minerals. Nevertheless, a wide range of fibrous and nonfibrous minerals has been studied to some extent for potential health hazards, although such data are fre quently hidden within reports that focus on asbestos These data are important, however, for several reasons First, all humans are exposed to mineral dusts from both anthropogenic and natural sources. For example, m a 45yr study on the mineral contents of lungs from residents of Tokyo, Shishido et al. (1989) found that by the early 1980s over 80% of Tokyo's residents had been exposed to mineral dusts. Similar observations have been made in a variety of urban environments, and nonoccupational exposure to mineral dusts can also result from living in a rural environment (e.g., Sbastien et al., 1981, 1984) Second, various minerals are used both as replace ments for asbestos and in numerous other commercial products, and the health risks associated with these min erals may differ from those associated with serpentine asbestos or with amphibole asbestos. If minerals are to be used and regulated properly, it is important to assess accurately the risk from exposure to each mineral and not simply assume that all fibrous minerals are equally hazardous. 226 GUTHRIE- EFFECTS OF INHALED MINERALS Third, the mechanisms by which minerals in general (including asbestos) are toxic and carcinogenic can be more readily elucidated if differences in the toxicity and car cinogenicity of various minerals can be related to fun damental differences in crystal structure and crystal chemistry. Collaborative efforts between minerals scien tists and life scientists would be extremely effective at achieving this. Unfortunately, very few minerals scien tists are involved directly in such research, primarily be cause of scientific language barriers and a lack of famil iarity with current issues in health-related mineral-dust research. In order to address the latter problem, this paper will review the health risks associated with a variety of min erals. Ross (1981, 1984) reviewed the human health haz ards associated with the asbestos minerals in a form ac cessible to geoscientists, and Mossman et al. (1990) presented a review of the current issues in this research field. These papers are useful starting points for anyone interested in the extensive literature available on asbes tos. Likewise, the research on the the biological effects of silica is extensive, and Heppleston (1984) provides an introduction to this literature. However, no such com pilation exists for the literature on numerous other fine grained minerals. The intentions of this paper are (1) to compile the available data on the biological effects of clays (exclusive of chrysotile), zeolites, and several other fine grained minerals; (2) to facilitate a better understanding of the data by introducing the research techniques used in the studies from which the data were drawn; and (3) to provide a foundation that can be used by minerals scientists interested in pursuing this area of research. The intentions of this paper are to avoid any discussion of mineral regulation or the regulatory implications of the material reviewed. Though many o f the minerals covered in this review are toxic, carcinogenic, or both in some tests, the risk to humans exposed under normal condi tions may be minimal. As will be noted, many of the experiments used to assess the pathogenicity of a mineral investigate the mechanisms of pathogenicity or evaluate the pathogenicity of a specific mineral relative to other minerals. These experiments do not necessarily emulate typical conditions of human exposure; hence the results may not reflect the exact response that would be expected in humans. The assessment o f risk from exposure to a specific mineral is an extremely involved task. The inter ested reader is directed to the article by Mossman et al. (1990) for an introduction to the problem and a list of pertinent references. Current federal regulations, includ ing those for minerals, can be found in Codes o f Federal Regulations (CFR 29, part 1910.1000). CFR 29 is revised annually on July 1 and is found in most libraries. Introduction to biological reports Determination of a substance's biological activity Exposure to mineral dusts has been linked with a va riety of lung diseases. Exposure to a fibrogenic mineral can result in fibrosis (production of scar tissue) in the lung, which can impair the function of the lung. Exposure to a tumorigenic (or carcinogenic) mineral can result in cancer, such as lung cancer or mesothelioma. Lung cancer is associated with exposure to a variety of substances (e g , cigarette smoke) in addition to minerals, but mesotheli oma, or cancer of the mesothelium (lining of the abdom inal wall), is commonly associated exclusively with ex posure to fibrous minerals, predominantly the asbestos minerals. The potential for a specific mineral to induce these diseases can be evaluated by numerous techniques, each o f which provides different information and has dif ferent factors that complicate interpretation. As pointed out by Rail (1988), there are four basic groups of methods used to determine the carcinogenic potential of a sub stance: epidemiological studies, in vitro studies, in vivo studies, and prediction of biological activity by compar ison with a similar mineral (structure-activity relation ship). Epidemiological studies. In an epidemiological study, a substance's health risks are evaluated by determining the relationships between human exposure to a substance and the potential health effects. One approach to epide miological studies uses cohorts, or groups of individuals whose lifestyles are similar, to monitor the incidence of disease in response to exposure to a substance. Ideally, two groups are chosen such that their difference is only in the exposure to a specific substance; thus one group serves as a control against the health effects from other agents. In many studies, however, no explicit control group is used, but the study group is compared with national averages. Epidemiological data are commonly reported using a standardized mortality ratio (SMR) that com pares the observed death rate from a disease in the study group with the rate expected based on the control group. Another approach to epidemiological studies uses case studies to determine lifestyle patterns of individuals af flicted with a disease. In both approaches it can be diffi cult to assess the effect of other harmful substances to which individuals were exposed, such as other mineral dusts or tobacco. In other words, an epidemiological study looks for patterns in the incidence of a disease and, thus, permits only an indirect determination of the cause of that disease. An obvious advantage of an epidemiological study, however, is that it attempts to determine the actual effect of mineral-dust exposure on humans exposed under typical conditions. An important aspect of epidemiological studies is the characterization of exposure, which can be achieved by a number of techniques. A dust's mineral content, particle size and shape, and areal distribution can be measured directly in an environment by collecting air samples or soil or dust samples. Such measurements are useful for providing detailed information concerning current ex posure conditions. However, onset of disease can occur 20-30 yr after exposure to a mineral dust, and current exposure conditions may differ significantly from previ ous exposure conditions. Furthermore, dust characters- GUTHRIE: EFFECTS OF INHALED MINERALS 227 tics may fluctuate in some environments, causing an in correct estimation of exposure. A direct characterization of the dusts to which an in dividual was exposed can be made by analyzing lung tis sue or expectorated sputum. A variety of processes occur in the lung following exposure to dust (Lehnert, 1990; Schlesinger and Driscoll, 1989), and many of these result in a clearance of dust particles via the mucociliary esca lator within the lung and trachea. Thus, sputum provides a means of directly sampling some of the dust particles being cleared from the respiratory system. This process can continue over a prolonged period after exposure. Fre quently, these particles become coated with ferruginous material believed to be derived from proteins; such coat ed particles are referred to as ferruginous bodies. Particle concentration in the sputum has been used as an indi cation of current particle content in the lung (Sbastien et al., 1984). Analysis of sputum samples is compara tively simple and inexpensive. However, high lung bur dens--i.e., -10000 ferruginous bodies per gram of dry lung (Sbastien et al., 1984)--are required before particles are detected, and exposure estimates are biased to the extent that the sample includes only those particles being cleared. An accurate assessment of the mineral content of lungs can be obtained even for lower lung burdens by analyzing lung tissue obtained during surgery or a post mortem examination (e.g., Churg et al., 1984). However, biopsy of lung tissue is complicated by the heterogeneous vanation of particle deposition and retention at different sites within the lung, so lung samples are needed from several locations to estimate an average lung burden. Incidence of disease is typically estimated by several methods. Clinical examination of at-risk individuals combined with chest X-rays can often detect early indi cations of disease, but to assess the subjective factors as sociated with the grading of a chest X-ray, the same set of X-rays must be read by several individuals. Death cer tificates provide another means of estimating the inci dence of disease. However, this method can result in in correct estimates, since the accurate classification of a specific disease often requires more extensive analysis than is commonly performed during a postmortem examina tion. Finally, in some cases, biopsies can be performed on lung tissue from diseased individuals, and lungs and other organs can be removed and examined after an in dividual from an exposed group has died, as is done with asbestos workers in South Africa, for example. Even so, there is some concern that neoplasias (malignant tumors) can be classified incorrectly, resulting in either an over estimation or underestimation of the incidence of disease. In vivo studies. Animal models are used extensively to study the effects of exposure to mineral dusts. Ideally, an animal species is chosen such that its response to a spe cific substance closely resembles the response observed in humans under similar conditions. With such an animal model, the complete biological effect of a substance can be studied under various exposure conditions. In prac tice, however, responses observed in animal models are not identical to responses observed in humans, so pre diction of a human response using results from an in vivo experiment is not always straightforward. In nuneral-dust research, rats and mice are the most commonly used an imal models; however, guinea pigs, sheep, dogs, ham sters, monkeys, and rabbits have also been used. Pott (1980) reviewed some of the in vivo experiments con cerning the biological effects of mineral fibers and dis cussed the differences in response among species. Because disease must be induced more rapidly in ani mals than it is induced in humans, in vivo experiments commonly use exposure methods that differ significantly from exposure conditions expenenced by humans. Dis ease in humans often occurs up to 20-30 yr after initial exposure to a dust; however, most lab animals live less than 20-30 yr. Hence, disease is induced more rapidly in an m vivo experiment than would be expected under nat ural exposure conditions. Typical in vivo experiments employ one of three exposure methods: (1) intratracheal injection of a dust-saline solution into the target organ; (2) direct application of the dust to the target organ (e.g , intrapleural and intrapentoneal instillations), or (3) in halation in a dust-rich environment (e.g., 1-50 mg/m3or ~ 500-2500 fibers/mL). (The current regulatory standard for occupational asbestos exposure in the U.S. is <0,2 fibers/mL; however, 50% of all asbestos exposure levels in U.S. schools lie within lO ^ -lO '3fibers/mL, according to the Environmental Protection Agency, 198ti.) The route of entry for the dusts in such experiments clearly differs substantially from the typical route of entry in humans (inhalation in a comparatively dust-poor environment) However, even for inhalation experiments conducted with reasonable dust levels, exposure would differ from that expected in humans, since the differences between the respiratory systems of laboratory animals and humans (e.g., airway size and shape, breathing patterns, clearance mechanisms) introduce a sampling bias on the particles reaching the lungs (e.g., Oberdorster and Lehnert, 1990) Short-term (2-3 yr) animal experiments also inadequate ly model human exposure in that short-term expenments do not consider the long-term (10-30 yr) alteration of mineral particles (such as dissolution or surface modifi cation) by the biological medium. In vivo studies can provide important information, such as (1) the effect of mineral dusts on a living organism including types, incubation periods, and severity of dis eases, translocation (migration) and clearance rates of particles from the site of initial exposure, and cellular responses to exposure; (2) the effect of various exposure conditions; (3) an evaluation of the risk to humans, (4) the elucidation of pathogenic (disease causing) mecha nisms; and (5) the identification of potential treatment methods. However, such experiments are time consum ing (up to several years duration), expensive, and difficult to interpret (unless a very strong or very weak effect is found) (Rail, 1988). Furthermore, the use of data from in vivo experiments to predict human response can be c o m plicated by a variety of experimental factors, including 228 GUTHRIE: EFFECTS OF INHALED MINERALS differences between human and animal response to ex posure, the degree and method of exposure, and the use of animal strains particularly susceptible or resistant to disease. These two last factors additionally make com parison of results from different studies difficult unless identical experimental procedures were used. When pro vided in the original report, animal strains will be in cluded here in the review of in vivo data to allow com parisons to be made between studies. In vitro studies. In vitro experiments use specific cells to determine a mineral's biological activity and are com monly used because they are rapid and relatively inex pensive. One of the more commonly used in vitro meth ods is the Ames test (Ames et al., 1975), which uses mutation rates in bacteria as a measure of carcinogenic potential. However, the asbestos minerals are one of only two suspected carcinogens (the other being conjugated estrogen) that do not appear mutagenic in a bacterial as say (Chamberlain and Tarmy, 1977; Shelby, 1988). The implications of this remain incompletely understood. Eukaryotic mammalian cells are also used for in vitro experiments to test a mineral's biological activity, with red blood cells (RBCs), macrophages, and epithelial cells being the most commonly used cell types. These cell types are also found in the lung where they can potentially in teract with inhaled dusts. Hemolysis experiments test the ability of a substance to destroy or lyse RBCs by incu bating RBCs in contact with the substance and then mea suring cell viability (the release o f hemoglobin is an index of cell destruction). In vitro experiments with other cell lines also test for cytotoxicity (the ability of a substance to kill a cell) normally either by (1) the cellular exclusion of a vital dye (where dead cells allow penetration of the dye but living cells do not); (2) the reduction in the ability of the cells to develop into colonies; or (3) the release of enzymes indicative of an increase in cell-membrane per meability or cell death, where membranes of healthy cells are impermeable to the enzymes. Heppleston (1984) pre sented an excellent discussion of information that can be derived from cytotoxicity experiments. In addition to their use in determining cytotoxicity, in vitro experiments can be used to determine the genotoxicity (ability to affect genetic material) or mutagenicity (ability to cause mutations) of a substance. As mentioned above, the Ames test is one such technique. Other meth ods involve a quantification of mutation by measuring processes involving chromosomes or DNA directly, such as sister chromatid exchange (SCE) or unscheduled DNA synthesis. O radicals, such as the superoxide anion, are believed by some investigators to participate in both fi brosis and carcinogenesis (Mossman et al., 1989; Mossman and Marsh, 1989), and the measurement of the gen eration of such anions by a catalytic reaction involving a mineral surface (Pezerat et al., 1989) is another potential indicator of a particle's mutagenic activity. Structure-activity relationship. The simplest and least expensive method to determine the potential health haz ards of a mineral dust may eventually prove to be the prediction of a mineral's biological activity by compari son with known structure-activity relationships. How ever, an accurate knowledge of the mechanisms by which a mineral is toxic is essential for this method to be effec tive. Several mechanisms are currently proposed to ex plain the biological activities of minerals (Table 1 and Fig. 1). However, these mechanisms remain too poorly understood to allow an accurate prediction based on a mineral's structure. Nevertheless, such predictions are made, primarily based upon the observed correlation be tween biological activity and particle shape and size (e.g , Stanton et al., 1981). It is interesting to note that many of the proposed mechanisms for mineral-induced disease involve chem ical reactions (e.g., oxidation/reduction) that are similar to reactions that occur in a geological environment. A geochemist's approach to the study of such reactions dif fers greatly from the approach taken by most biological scientists. Consequently, geochemists can contribute enormously to the study of mineral-induced diseases. One role for the geochemist is the identification and charac terization of active sites on mineral surfaces (the results of which would be of interest to both geologists and bi ologists). By characterizing the appropriate physical and chemical aspects of each sample, biological activity can be related to a measurable parameter (e.g., surface area, Lewis-acid/base sites per surface area, Bronsted-acid sites per surface area) and a model for biological activity can be developed. Furthermore, a sound mineralogical ap proach would facilitate the design of biological experi ments that control most potentially active mineralogical characteristics while allowing the active site of interest to be studied. For example, numerous experiments have been conducted to test the biological activity of fibrous am phiboles. As shown in Figure 1, however, amphiboles have numerous potentially active sites. Hence, the results of a study on the toxicity of amphibole cannot be asso ciated with one particular active site. Conversely, if ex periments were conducted using two amphiboles that dif fered only in the composition of the octahedral site, then the activity of polyvalent cations in amphiboles (active site 3 on Fig. 1) could be determined. The diversity of mineral species offers a unique potential for characteriz ing the activity of numerous mineralogical characteris tics. In addition to natural samples, synthetic minerals could be used effectively, e.g., by growing zeolites with identical framework topologies but with various amounts of tetrahedral Al. Mineralogical aspects Mineral names. In principle, usage of nomenclature to describe a mineral should follow strict guidelines For species names, these guidelines are widely accepted and have been developed so that a mineral name provides important information concerning both structure and com position. F or exam ple, the nom enclature for am phi boles is extremely complex and relies on a knowledge of the composition and structure of a given amphibole spec- GUTHRIE: EFFECTS OF INHALED MINERALS 229 T able 1. Mechanisms of mineral-induced disease Proposed mineralogical mechanisms Proposed biological mechanisms Bransted-actd sites proton-donor sites associated with underbonded O atoms resulting from either cation substitutions or broken bonds at a mineral surface Lewis-acid/base sites electron-acceptor/donor sites associated with polyvalent cations Alkali-cation sites easily exchangeable cations, e.g , amphibole A site or ze olite cage site Specific surface periodicities periodicities that promote a specific interaction with a particular molecule, e.g., DNA-particle interactions re ported by Appel et al (1988) Oxidative stress resulting from the catalytic pro duction of oxygen radicals at a mineral surface Genetic alteration by a number of mechanisms, in cluding transfection of Inter cellular DNA and DNA-particle interaction dunng mitosis Incomplete phagocytosis resulting In the release of cyto toxic enzymes by the cell ;Note. There is no implied correlation between opposing mineralogical and biological mechanisms. imen (Leake, 1978). Hence, use of the species name riebeckite defines a mineral as a clinoamphibole (i.e., a dou ble-chain silicate with a specific stacking order) with a composition of Na2Fe5Si80 22(0H )2. Varietal names, how ever, are not so strictly defined. So, although the varietal name "crocidolite" is commonly used to describe asbestiform riebeckite, the identification can be based on the blue color and asbestiform habit but not on composition al or structural information. Numerous examples illustrate the problems that can result from inaccurate usage of mineral-species nomen clature. "Amosite" is a term used to describe brown as bestos. The term originates from an acronym for the As bestos Miners of South Africa. Although the term is frequently used to describe asbestiform cummingtomtegrunerite, it has also been used to describe ores containing mixtures of asbestiform amphiboles. This ambiguity has led to conflicting definitions of "amosite," i.e., "amosite" has been defined as a varietal name for asbestiform ferrogedrite by some (e.g., Roberts et al., 1990). Hence, the term "amosite" provides limited information pertaining to the mineral content of the sample. In reports on the effects of zeolites, erionite and mordenite are often referred to as fibrous and nonfibrous equivalents, despite the fact that both are normally fi brous and have different structures and composition. Furthermore, the identification of one or the other is typ ically made using qualitative analytical TEM (ATEM) based on the presence or absence of specific exchangeable cations. Similar examples of the incorrect usage of min eral names occur throughout the literature on the health effects of dusts. Hence, the reported mineral content is potentially suspect, unless adequate data were used for identification. The lack of interest in the importance of a mineral's structure and composition is further demonstrated by data Fig. 1. Schematic diagram o f the amphibole structure viewed along the c axis Numbers indicate possible active sites (1) A-site cations, (2) protons associated with Al substitution m the tetrahedral sites, (3) polyvalent cations in the octahedral sites, (4) protons associated with underbonded O atoms sheets for the asbestos minerals. Both the National Insti tute for Occupational Safety and Health (NIOSH) and at least one major supplier of asbestos for biological re search provide data sheets listing incorrect mineral for mulae for the asbestos minerals. Such incorrect infor mation may form the basis of mineral identification or interpretation of results used in some studies on the health effects of asbestos. Because of this casual usage of mineral-species nomen clature, it is difficult to interpret the results of a study m terms of mineralogical properties, such as structure and composition. The improper use of mineral-species names is comparable to an incorrect usage of nomenclature for cell type or animal species. If a report stated that "rodents were used in the experiments" or, worse, "rats were used in the experiments" when in fact the experiments were conducted on mice, the data would be extremely difficult to interpret. Most reports on the in vivo effects of min erals are extremely specific in describing animal species and cell types, yet they often fail to identify mineral spe cies correctly. Unfortunately, the advantages of a strict usage of mineral-species nomenclature are not fully ap preciated by many scientists involved in health-related mineral research, including both minerals scientists and nonminerals scientists. Sample purity. Another source of uncertainty in these studies concerns purity of the mineral dust. Detailed de scription of the mineral content is rarely given, and iden tification of the minerals generally relies upon that made by the supplier. Even when obtained from reliable sources, however, the exact mineral content of a dust is suspect Most suppliers provide mineral samples that contain mixtures of m in e rals, e v e n th o u g h one m in e ral m ay be 230 GUTHRIE: EFFECTS OF INHALED MINERALS the dominant constituent. Mineral impurities, even in small quantities, may have a significant effect on a bio logical response. For example, recent studies (e.g., Churg et al., 1984) show that the lungs of workers exposed to chrysotile contain abundant amounts of fibrous tremolite, a minor contaminant of chrysotile ores. Churg et al. (1989) found that the rate of mesothelioma is strongly correlated with tremolite but much less so with chryso tile. Continuing with the analogy above, the use of im pure mineral samples is equivalent to running an in vitro assay with lung cells, which would consist of several types of cells (e.g., macrophages and epithelial cells). This would never be done in an in vitro experiment, as each cell type responds differently Mineral identification. Most of the studies reviewed below generally characterized samples by light micros copy or transmission electron microscopy (TEM) and less commonly by ATEM. Electron diffraction analysis is rarely used to determine mineral content, despite the fact that most of the particles cannot be identified uniquely based on morphological and compositional data alone. Particle identification by TEM can be time consuming when large numbers of particles are included, especially when ATEM and electron diffraction are used. Hence, results are potentially affected by poor counting statistics and incorrect identification of the particles (when electron diffraction is not used). Minerals present in dust can also be determined by quantitative X-ray diffraction (XRD). Davis (1990) measured reference intensity ratios (RIRs) for chrysotile and amphibole asbestoses and demonstrat ed that the minerals in asbestos mixtures can be deter mined with a lower detection limit of ~ 0 .5-2.0 wt%. Puledda and Marconi (1990) also reported a low detec tion limit (2 fig) for chrysotile in various matrices. Chipera and Bish (1989) and Bish and Chipera (1991) used RIRs to determine erionite concentrations in dusts and reported detection limits as low as 100 ppm. Quantitative analysis of mineral content using the Rietveld method with X-ray diffraction data (Snyder and Bish, 1989) may prove to be even more successful than the use of RIRs, since the Rietveld method effectively addresses problems associated with peak overlap and compensates somewhat for the high degree of preferred orientation exhibited by fibrous and platy minerals. However, no biological stud ies using this approach have been reported to date. Assessing human risk As indicated above, many of the results of experiments on mineral-induced pathogenesis do not relate directly to risks to humans. For example, though a mineral may be highly active in an in vitro assay, it may pose little risk to humans under normal conditions (e.g., kaolinite). Hu man response relates to a number of factors, including the pathogenicity of the sample, residence time in vivo, dose, and variations in individual response (e.g., physical condition of the person, smoking history, propensity to ward a specific disease). Hence, a mineral such as kaolin ite may be highly active at the cellular level, but it does not reside long enough in the lungs to induce disease. Alternatively, a mineral may show a positive response in a given assay, but the mechanism tested by the assay is not involved in pathogenesis in humans. Review of data Numerous data are reported that describe the biologi cal effects of mineral dusts. However, most such studies provide only limited mineralogical detail Hence, the re sults from these studies may not reflect results one would anticipate from a mineralogically pure sample. Studies that reported only limited mineralogical data may have used samples that were (1) mineralogically homogeneous, containing the mineral described; (2) mineralogically het erogeneous, containing numerous minerals; or (3) mmeralogically homogeneous, containing a mineral different from the one described. In general, it is not possible based on the published data to determine to which category a sample belongs. In the absence of data to the contrary, I believe it should be assumed that the mineral content of a sample is as described by the original investgalo: s Nevertheless, this adds uncertainty to anv interpretation based on these data, and this uncertainty should be rec ognized. In order to clanfy these uncertainties, some in vestigators have supplied me with their samples, and these samples are being characterized using quantitative X-ray diffraction analysis (Guthrie and Bish, unpublished data). Where preliminary results are available, these will be giv en. Those investigators that have supplied samples are commended for their desire to augment their studies with detailed mineralogical information. For those studies that did present mineralogical infor mation (ranging from a generalized locality to complete descriptions, including compositional analysis), the rameralogical detail will be included with the summary of results. When no mineralogical information was given, only a mineral or material name will be listed. Related to the poor mineralogical detail is potentially inaccurate usage of mineral-species names. The use of unaccepted mineral names, rock names, and compound names is rampant in the biological literature. Some as sumptions, therefore, were made in order to categorize these studies with respect to mineralogy. For instance, for the purpose of categorizing it was assumed that " Fe20 , " and "iron-ore dust" refer to hematite, "bentonite" refers to montmorillonite, and "attapulgite" refers to palygorskite; the quoted terms are retained in the review to allow the reader to recognize assumptions by the author. Fur thermore, when mineral group terms (such as carbonate or mica) were used, these are also retained in the review. This is not meant to imply that either the quoted terms or the mineral group names are mineral-species names, but rather it serves to categorize the study as appropri ately as possible based on the probable major mineral. Oxides and hydroxides Most studies on oxide- and hydroxide-bearing dusts suggest that some samples can produce fibrosis in vivo GUTHRIE: EFFECTS OF INHALED MINERALS 231 but are generally relatively inactive minerals. Exceptions may include oxides containing Cr (not necessarily min eral samples) and fibrous brucite. Fibrous brucite may, in fact, be extremely active, with exposure resulting in fibrosis, carcinogenesis, or both. Hematite. Epidemiological studies suggest that expo sure to hematite-bearing dust alone under modem min ing conditions--i.e., low dust exposure, ventilation, wet drilling--does not increase the risk of lung cancer. How ever, individuals exposed under unfavorable mining con ditions--i.e., high exposures to dusts (especially when contaminated with silica), radon, and tobacco sm okedo show a higher than expected incidence of respiratory disease. Lawler et al. (1985) found no overall excess of mortal ity due to lung cancer (SMR = 0.97 for all cancers and SMR = 0.94 for lung cancer; SMRs based on U.S. white males) among 10403 underground miners in Minnesota who were exposed to iron-ore dust (hematite + "limomte," silica, phosphates, and other oxides, as given by Lawler et al., 1985). In fact, they found a lower than expected mortality for respiratory disease overall (SMR = 0.79). However, they reported no information about exposure conditions, i.e., quantitative mineral content of the dusts, particle sizes, or mass concentrations. Lawler et al. agreed with suggestions by previous workers (e.g., Boyd et al., 1970; Radford and Renard, 1984) that ob served excesses in mortality due to lung cancer in ironore miners reflect exposures to other factors, such as ra don and radon daughter products (RRDs), silica, tobacco smoke, and diesel fuel, since exposure to these factors was minimal in their cohort compared with cohorts of ironore miners studied previously. Chen et al. (1990) also suggested that exposure to RRDs and silica could explain excesses of mortality due to lung cancer in hematite miners. They studied 6444 male work ers associated with hematite mining in China, 5406 of whom were involved in underground operations, and re ported SMRs (based on age-specific death rates for Chi nese males) for lung cancer ranging from ~ 1.0 at the 95% confidence level for zero-to-medium dust exposure to >2 7 for heavy dust exposures. However, exposure to RRDs is highly correlated with dust exposure; hence, the individual effects of the two could not be separated. In cidences of other respiratory diseases, such as silicosis and tuberculosis, are also correlated with the incidence of lung cancer, but no data on silica exposure were given. Cigarette smoking shows a positive relationship with lung cancer as well, as indicated by an SMR for lung cancer in smokers of 2.7-6.3 (95% confidence interval or Cl). In the cohort of Chen et al. the use of modem mining con ditions (wet drilling and ventilation) has lowered airborne dust exposure significantly (from > 100 mg/m3to < 5 mg/ m3), and workers exposed only under the improved con ditions may show a lower SMR for lung cancer (1.1-4.6, 95% Cl) than those first exposed prior to the use of wet drilling and ventilation (2.9-7.4, 95% Cl). In vivo experiments indicate that hematite-bearing samples are biologically inactive. Pott and coworkers (Pott and Friedrichs, 1972; Pott et al., 1974) found no fibrosis or tumors in 80 Wistar rats 530 d after intrapentoneal injections of hematite (source not given), whereas in the same study, Wistar rats showed fibrosis and up to a 40% incidence of tumors when injected with chrysotile and a 55% incidence of tumors following the injection of silica glass. Vorwald and Karr (1938) found that hematite dust induces no tumors in guinea pigs or rats following inha lation of the dust; no information was given concerning the source of dusts or exposure levels. However, asbestos (type not given) also failed to induce tumors in guinea pigs in the same experiment. Finally, Mossman and Craighead (1982) found that hematite (IIT Research In stitute, Chicago) does not induce tumors in golden Synan hamsters following subcutaneal implantation of in vitroexposed tracheas; however, hematite is nearly as effective a cocarcinogen as riebeckite asbestos [Union Internatio nale Contre le Cancer (UICC) standard) when pretreated with a polycyclic aromatic hydrocarbon. In vitro experiments suggest that hematite-bearing samples are noncytotoxic and nongenotoxic Dubes and Mack (1988) used an in vitro technique to test the ability of a variety of materials to mediate the transfection of mammalian cell cultures; transfection of cells is the pro cess of introducing foreign genetic material into a cell and is one proposed mechanism for the carcinogenicity of mineral dusts (e.g., Appel et al., 1988) Their data show that Fe20 3(reagent grade from Matheson, Coleman, and Bell Company) is only slightly effective as a mediator For comparison, asbestos (from a vanety of sources) and Cr20 , (from Matheson, Coleman, and Bell) were ~3-7 times more effective mediators. Witmer and Cooper (1983) also reported that Fe20 3is nonmutagemc whereas Cr20 3 is mutagenic, as determined in a modified Ames test. Boehmite, goethite, and lepidocrocite. In vivo experi ments suggest that samples containing iron and alumi num hydroxides may be slightly active in the lung. In halation experiments by Gardner et al. (1944) showed no effect of boehmite laths [measuring ~7 5 x 30 0 nm; mineral identification confirmed by King et al. (1955) using TEM and XRD] on the lungs of guinea pigs. How ever, using the same material, King et al. (1955) observed that severe and rapid pulmonary fibrosis de\ elops in rats after a direct injection of boehmite-saline solution into the lungs. Stacy et al. (1959) extended the study by King et al. (1955) to include an additional sample of boehmite (better crystallized and having a mean size of - 3 am) and samples of goethite and "lepidocracite" (presumably lepidocrocite) measuring <0.5 am and ~0.5 x 2 am, respectively. They demonstrated that the boehmite sam ple used in the experiments by Gardner et al (1944) pro duces a dose-dependent fibrogenic response, whereas the larger-grained, better-crystallized boehmite is much less fibrogenic. Goethite- and lepidocrocite-beanng samples, however, are only slightly fibrogenic, even at higher doses than boehmite. Inhalation experiments by Campbell 232 GUTHRIE- EFFECTS OF INHALED MINERALS (1940) using "the precipitated brown oxide of iron (Fe20 3, H20), British Drug Houses" produced a 22.7% incidence of lung tumors in mice, compared with 6.8% in the con trol group and 17.6% in a group exposed to "precipitated silica" dust; tumors did not develop until after 300 d, with most developing 600-900 d after first exposure. An imals were exposed to 0.5 g of dust per hour, 6 h per day, 5 d per week for 1 yr. Brucite. In vivo experiments suggest that samples con taining fibrous brucite (described as "nemalite" in most studies) are both fibrogenic and carcinogenic. Pott and coworkers (Pott and Friedrichs, 1972; Pott et al., 1974) found that fibrous brucite, when intrapleurally injected in Wistar rats, produces fibrosis comparable to that pro duced by silica or chrysotile injections and exhibits a tu mor rate (62.5%) exceeding those of silica (55%) and chrysotile (40%); palygorskite, however, exhibits a 65% tumor rate in the same test. Wagner et al. (1973) found that 20 mg of a sample containing fibrous brucite admin istered to Fischer 344 rats by intrapleural injection in duces mesotheliomas at a rate of 56%. Their sample was obtained from a Canadian mine and contained chrysotile, though no estimate for the amount of chrysotile contam ination was given. It should also be noted that chrysotile from this same mine induces mesotheliomas under the same conditions at a slightly higher rate (61%). In vitro experiments further indicate that samples con taining fibrous brucite are cytotoxic to a variety of cell lines. Chamberlain and Brown (1978) showed that fi brous brucite (same material as used by Wagner et al. 1973) reduces the colony-forming efficiency of Chinesehamster lung cells at doses roughly equivalent to those for a comparable cytotoxic response in experiments with either amphibole or serpentine asbestos. The dose re quired to reduce cloning efficiency to 50% is 12 Mg/mL for fibrous brucite compared with 9 ug/mL for riebeckite asbestos and 17-26 g/mL for chrysotile; talc is nontoxic at 50 ig/mL, the highest dose used. Jaurand et al. (1980) found that fibrous brucite is cytotoxic to human RBCs and rabbit alveolar macrophages. They demonstrated that both the hemolytic and cytotoxic activities of fibrous brucite are intermediate to those of chrysotile or nebecklte asbestos. Finally, Pezerat et al. (1989) found that fi brous brucite is comparable to chrysotile and much more effective than amphibole asbestos in its ability to catalyze the production of O radicals, a step of potential impor tance in both fibrogenesis and carcinogenesis. The 1:1 layer silicates and chlorite Most studies of samples containing 1:1 layer silicates or chlorite suggest that some samples can produce fibrosis or tumors in vivo and can be highly active in vitro. How ever, epidemiological data on exposure to kaolinite-bearmg dusts suggest that fibrosis is induced only in extraor dinary conditions, i.e., high exposures or in the presence of other pulmonary complications. Although these min erals may be cleared rapidly from the lung (and hence are not pathogenic in humans), their in vivo and in vitro activities may provide clues to the mechanisms of min eral-induced pathogenesis. Kaolinite and halloysite. Epidemiological studies sug gest that kaolinite-bearing dust is fibrogenic only under extraordinary conditions, i.e., high dust conditions or ex posure combined with another respiratory disease, such as tuberculosis. Hale et al. (1956) reported case studies of seven kaolinite workers (primarily baggers exposed to extremely high dust conditions) who showed indications of respiratory disease, as determined by clinical exami nations including chest X-rays. Autopsies of two of the men revealed fibrosis associated with large amounts of kaolinite, mica, and amorphous silica. One of the autop sies, however, also noted tuberculosis; dead tuberculosis bacilli enhance the fibrogenic effect of kaolin dust in an imals (Kettle, 1934; Attygalle et al., 1954). Similar ob servations were reported m a case study by Lynch and Mclver (1954). In a cohort study, Sheers (1964) found fibrosis in up to 13% o f kaolinite workers exposed to high dust levels. Tuberculosis was uncommon in his cohort However, fibrosis was highly correlated with high dust exposures and length of employment. More recent studies confirm that exposure to high dust levels (particularly silica-bearing dusts) during kaolin mining can be corre lated with abnormalities m chest X-rays (e.g., Kennedy et al., 1983; Oldham, 1983; Sepulveda et al., 1983, Ogle et al., 1989). However, some of these abnormalities may not reflect the onset of fibrosis (Oldham, 1983). Lapenas et al. (1984) confirmed the presence of kaolinite in pul monary tissue from five kaolin workers with pneumo coniosis; silica was not present in the lung samples. In vivo experiments reported thus far on the fibrogenic potential of kaolinite-bearing dusts are inconclusive. Ket tle (1934) observed no fibrosis in guinea pigs following intratracheal injection of kaolinite (British Drug Houses; the sample contained quartz and "very numerous" sencite fibers), though, as indicated above, he did find that exposure to kaolinite and dead tuberculosis bacilli does result in fibrosis. King and Harrison (1948) used direct injection into the lung to study the effects of two kaolinite samples on rats. Unfortunately, one of the experiments used kaolinite samples containing 35.68 wt% carbonate minerals (species not given), whereas in the other exper iment, which used a comparatively pure sample of ka olinite, only two rats survived more than 10 d after ex posure. Neither of these rats developed fibrosis. Mossman and Craighead (1982) found that kaolinite (3-5 Min in diameter, Georgia Kaolin Company) does not induce tu mors in golden Syrian hamsters following subcutaneal implantation of in vitro-exposed tracheas and is a slight ly less effective cocarcinogen than UICC crocidolite when pretreated with a polycyclic aromatic hydrocarbon. In halation experiments by Wagner (1990) produced no lung tumors in 20 rats (probably from the Wistar strain) ex posed over a period of 3-24 months, but a slight fibro genic re sp o n se was o b se rv e d . His sa m p le s c o n ta in e d 8 5 95% kaolinite, with the remainder consisting of mica, feldspar, and quartz. For comparison, a "nonfibrous ze GUTHRIE: EFFECTS OF INHALED MINERALS 233 olite" and a "long attapulgite" produced more severe fibrogenic responses. Wastiaux and Daniel (1990) also used inhalation-exposure methods to assess the fibrogenicity of kaolin. They reported that their kaolin sample (Cor nish kaolin dust) induces a moderate fibrogenic response in Wistar rats. Long-term experiments currently in prog ress by Maltoni and coworkers (Maltom et al., 1982; Maltoni and Minardi, 1989) may provide additional infor mation on the in vivo activity of kaolinite. In vivo studies using halloysite-bearing samples, how ever, suggest that this kaolin-group mineral may be car cinogenic. Stanton et al. (1981) found that two samples of halloysite (obtained from the water supply of Hong Kong) induce a tumor rate of ~20% in Osborne-Mendel rats exposed by direct application of the dust to the pleu ral surface. For comparison, in the same expenments, amphibole asbestoses induce tumors at rates ranging from 0 to 100%. Wagner (1982) also reported in vivo data on halloysite. He observed no mesotheliomas in 40 Fischer 344 rats treated by intrapleural inoculation, whereas chrysotile (UICC standard B; derived from Canadian de posits) induces 22.5% mesotheliomas by the same tech nique. Whether the observed difference in the pathoge nicities of kaolinite and halloysite is related to particle morphology or other mineralogical properties (e.g., sur face characteristics) is not known. In vitro experiments show that kaolinite-bearing sam ples are cytotoxic to most cell types studied, though some materials are noncytotoxic. Low et al. (1980) found that kaolinite is cytotoxic to rabbit alveolar macrophages; their sample was a 99% pure kaolinite (as determined by XRD and energy-dispersive spectrometry) obtained from the Georgia Kaolin Company. Davies (1983) showed that ka olinite is also cytotoxic to mouse pentoneal macrophages but that treatment of the dust with poly(2-vinylpyndine N oxide) (PVPNO), a class of polymers that inhibit the cytotoxicity of quartz (Holt et al., 1970), almost com pletely eliminates kaolinite's cytotoxicity. Davies noted that his sample contained 2% mica, as determined by XRD. Dubes and Mack (1988) found that kaolinite (J. T. Baker Chemical Company) is --4-5 times more effective than asbestos in mediating transfection of mammalian cell cultures. Gormley and Addison (1983) also found that the kaolinite standards of the CMS Clay Mineral Repository, KGa-1 and KGa-2, are cytotoxic to a macro phagelike mouse cell line only at high doses. In contrast to the above studies, however, Marks and Nagelschmidt (1959) found that kaolinite is much less cytotoxic to guin ea pig peritoneal macrophages than silica minerals. Woodworth et al. (1982) used the release of 5lCr to mon itor changes in cell-membrane permeability and cell death in Syrian hamster tracheal epithelial cells. They found that kaolinite (Georgia Kaolin Company) will cause the release of s'Cr. Kaolinite is less effective than chrysotile and montmorillonite but more effective than silica. There is some indication that kaolinite's cytotoxicity is in part related to broken Si-0 bonds at the crystallite edges. As noted above, Davies (1983) found that the treatment of kaolinite with PVPNO reduces kaolinite's cytotoxicity at amounts less than the total amount of polymer that can be adsorbed, implying that only some of the polymer-binding sites may be related to kaohnite's cytotoxic activity. Furthermore, PVPNO is effective at inhibiting the cytotoxic activity of quart? (Holt et al., 1970), suggesting that the mechanisms by which quartz and kaolinite exert their cytotoxic effects are related. Steel and Anderson (1972) found that the addition of a bac terium, Staphylococcus aureus, to a kaolimte-NaCl solu tion at low NaCl concentrations (14 mM) inhibits floc culation, possibly because of an interaction between the bacterium and the kaolinite crystal edges. Others (Ken nedy et al., 1989; Ghio et al., 1990) have shown that kaolinite (noncalcined Georgian sample) and other pneumoconiosis-causing minerals function as Fenton catalysts (electron transfer by Fe2+ - Fe3+ + e~), possibly as a result of Fe3+adsorbed on its surface. Their studies illus trate that biochemical mechanisms can be probed effec tively if mineral samples are selected carefully Serpentine, berthierine, and chlorite. Chrysotile is the serpentine mineral that has been studied in greatest detail as a potential health hazard. As indicated above, because of the enormous body of literature on the health effects of chrysotile, this mineral is not addressed directly in this paper. The interested reader, however, is directed to the recent article by Mossman et al. (1990) and papers by Ross (1981, 1984) for reviews of the research on chrys otile. With respect to other serpentine minerals, Woodworth et al. (1983) found that antigorite (Ward's Scientific; sam ple from Arizona) does not induce metaplasia (prolifer ation of cells) in tracheal mucosa of the golden Syrian hamster in vitro, whereas nebeckite asbestos does, the tracheal mucosa, or lining of the trachea, is a part of the respiratory tract with which inhaled dusts interact. Using the same material, Mossman and Sesko (1990) found that antigorite does not cause the release of slCr from hamster tracheal epithelial cells, whereas nebeckite asbestos and chrysotile do. Hence, if antigorite is cytotoxic, it is much less active than chrysotile. A berthierine-rich iron ore (40% berthienne; sample from Lorraine, France) and two Fe-nch chlontes (Pyr nes and Anjou) were studied by Costa et al (1990), using a chemical assay to measure the production of activated O species. Production of activated O species is a mech anism by which a material can induce a toxic response They found that the samples were highly active, and they associated this activity with the high Fe content. Fe con tents (reported as FeO) were in the range 12.5-30.0% for the three samples, but they did not show that the Fe was directly responsible for the observed activity. For com parison, they found that kaolinite (St Austelle) and quartz (DQ 12) are inactive in the assay. Despite the poor qual ity of the specimens used, their study is a good example of the type of mineralogical-based research that can ben efit the field of mineral-induced pathogenesis. The bio chemical mechanisms they investigated with their assay 234 GUTHRIE: EFFECTS OF INHALED MINERALS involved an electron-transfer process at the mineral sur face, a process similar to many geochemical processes. The 2:1 layer silicates Most studies of samples containing 2:1 layer silicates suggest that some samples can produce fibrosis in vivo and can be highly active in vitro. However, epidemiolog ical data suggest that fibrosis may not be a problem in modem mining conditions. As with the 1:1 layer silicates, though 2:1 layer silicates may be cleared rapidly from the lung (and, hence, are not pathogenic in humans), their activity may provide clues to the mechanisms of mineralinduced pathogenesis. Talc. In general, epidemiological studies suggest that exposure to talc-bearing dusts elicits a dose-dependent (albeit minor) response. Kleinfeld et al. (1967, 1974) studied the mortality in a group of 220 1alc miners em ployed for at least 15 yr between 1940 and 1965. Dust exposures in this group were very high before 1945 (102105 particles/mL) but dropped substantially after 1945 (--101--103 particles/mL); furthermore, miners were ex posed to a variety of dusts, including talc, serpentine, tremolite, carbonates, and silica. During 1945-1959, the mortality rate due to lung cancer was 3.4 times the rate expected based on U.S. white males in 1957, but the rate dropped to near normal during 1960-1969, possibly be cause of lower exposure levels. However, because of the exposure to dusts other than talc (particularly tremolite), it is not possible to assign this elTect to talc exposure alone. Other studies (Selevan et al., 1979; Brown and Wagoner, 1980; Leophonte and Didier, 1990) of talc miners and millers in New England reported similar ob servations, but one study (Stille and Tabershaw, 1982) found no increases m mortality from lung cancers among workers at one New York mine who had no prior work exposure. Coexposure to amphiboles is likely in many studies of talc-exposed workers. Cullinan and McDonald (1990) separated seven studies of talc workers on the basis of suspected amphibole exposure. Among the three studies with no suspected amphibole exposure, no mesothelio mas were reported out of a total of 2540 workers, though a slight increase in other respiratory malignancies was observed in two studies (Cullinan and McDonald, 1990). In vivo experiments on talc-bearing dusts suggest that talc is nonfibrogenic and noncarcinogenic. Pott and Friedrichs (1972) observed no fibrosis or abdominal tu mors following intraperitoneal injection of talc in Wistar rats. In a later experiment using the same technique, how ever, Pott et al. (1974) observed a slight incidence of tu mors (2.5%) with a latency period (587 d) twice that ob served for chrysotile or fibrous brucite. Wehner (1980) observed no significant changes in golden Syrian ham sters exposed to talc baby powder (presumably obtained from the funding agency, Johnson and Johnson); ham sters exposed to asbestos cement (mineral content not described) exhibited a response similar to the talc re sponse at comparable exposures. Wehner used asbestos dust (mineral content not detailed) as a positive control, but comparison is hindered because the control experi ments used exposures 8 times those used in the talc ex periments. Stanton et al (1981) observed a statistically insignificant tumor rate m Osbome-Mendel rats exposed by direct application of talc to the pleural surface. Wagner et al. (1979) and Wagner (1990) found no lung tumors among 96 Wistar rats exposed to talc by inhalation for 3-12 months. Their sample contained 8% impurities in cluding silica, chlorite, and carbonate. Endo-Capron et al. (1990) induced no pleural tumors after intrapleural injection of 20-mg talc (Luzenac, France). In vitro experiments are inconclusive regarding the cy totoxic activity of talc-bearing dusts. Chamberlain and Brown (1978) found that Italian talc (commercial cos metic grade; source not given) is noncytotoxic to Chinesehamster lung cells at concentrations up to 50 mg/mL However, using the same assay, Pigott and Pinto (1983) reported that talc (source not given) is slightly cytotoxic at the same concentration, for comparison, Pigott and Pinto found that nebeckite asbestos is highly cytotoxic and calcium carbonate is noncytotoxic. Talc is much less hemolytic than kaolinite or montmorillomte (Woodworth et al., 1982; Brown et al., 1980). Despite its weak cytotoxicity, talc is an effective mediator in transfection. Dubes and Mack (1988) showed that talc (talcum pow der, Mallinckrodt Chemical Works) is --2 times more effective than asbestos but ~2 times less effective than kaolinite in mediating transfection of mammalian cell cultures. Woodworth et al. (1982) found that talc (Cyprus Industnal Minerals Company, Los Angeles) will affect cell membrane (in Syrian-hamster tracheal epithelial cells) as monitored by the release of 51Cr. Talc is approximately as active as kaolinite in this assay. However, Endo-Cap ron et al. (1990) found that talc (Luzenac, France) pro duces no SCEs in rat pleural mesothehal cells Phlogopite, muscovite, illite, smectite, and vermiculite. Only a few epidemiological studies of respiratory disease resulting from exposure to dusts containing micas or micalike clays have been published, and some of these suggest that such samples can elicit a mild, dose-dependent fibrogenic response at high exposure levels (e.g., Vestal et al., 1943). Exposure to mica-like minerals is generally accompanied by an exposure to other minerals (e.g , silica and amphiboles), and the response to these minerals complicates the interpretation of the data (e.g., Heimann et al., 1953; McDonald et al., 1988). For example, some cases of vermiculite-related mesothelioma may be cor related with amphibole contamination (see Cullman and McDonald, 1990, for a review of the studies). In vivo experiments suggest that samples containing micas or mica-like clays are slightly fibrogenic. King et al. (1947) found injection of 50 mg of illite dust (sepa rated from shales in southern Wales) into the lungs of rats produces no fibrosis unless the clay is pretreated in an HC1 solution Policard (1934) used exposure by in halation to study the short-term effects (3-30 d) of ground white mica (from Madagascar; light microscopy showed GUTHRIE: EFFECTS OF INHALED MINERALS 235 the dust to contain both fibrous and polyhedral particles) on the lungs of rats, ground white mica induces a cellular response similar to that observed with quartz. Pott et al. (1974) found that biotite is inactive following intraperitoneal injection in Wistar rats. Sykes et al. (1982) used intratracheal instillation to study the short-term (<7 d) and medium-term (< 100 d) effects of "bentonite" on Alderley-Park-denved rats (strain 1; specific pathogen free). Though these results show that "bentonite" induces a greater pulmonary response than quartz in the short term, medium-term effects indicate that "bentonite" induces a response similar to a saline control. Rosmanith et al. (1990) studied the relationship between surface area and activity using intratracheal installation of a well-charac terized muscovite sample in SPF-Wistar rats (Rosmanith et al., 1990; Schyma, 1990). They found that the finest material elicits the greatest fibrogenic response. Brambilla et al. (1979) reported mild pulmonary lesions in zoo animals exposed to mica dusts. Mineralogical analysis of lung contents indicated the presence of muscovite and lllite In vitro experiments suggest that samples containing micas and mica-like clays may be slightly cytotoxic, though some studies suggest that phlogopite and montmorillonite may be highly cytotoxic. Pigott and Pinto (1983) studied the cytotoxicity of phlogopite, "hydrophlogopite" (?), and biotite (distinction not explained) and muscovite using Chinese-hamster lung cells. All four mi cas are slightly cytotoxic, with muscovite showing the greatest effect and being comparable to talc in activity. As mentioned above, riebeckite asbestos is highly cyto toxic in the same study. Gormley and Addison (1983) found that samples SAz-1 and STx-1 (calcium montmorillonites), SWy-1 (sodium montmonllonite), and SHCa-1 (hectonte) from the CMS Clay Repository exhibit a range in toxicities, with SHCa-1 being slightly cytotoxic (roughly comparable to kaolinite samples K.Ga-1 and K.GA-2) and STx-1 being highly cy totoxic (more cytotoxic than their positive control, quarz). It should be noted, however, that they reported 10% cristobalite in STx-1, as determined by XRD, and cristoballte is even more cytotoxic than quartz (Marks and Nagelschmidt, 1959). Adamis and Timr (1978) used peritoneal macro phages from Sprague-Dawley rats to show that both quartz and "bentonite" (Istenmezeje, Hungary; obtained from Z. Juhsz) are cytotoxic, but their modes of action are different. Although quartz alters the permeability of cell membranes to the enzyme lactate dehydrogenase (LDH), "bentonite" does not. However, "bentonite" does signif icantly lower the intracellular activity of LDH. Costa et al. (1990) used a chemical assay to determine the role of Fe2+ in the production of activated O species. As found for Fe-rich chlorite and berthierine, Fe-rich bi otite (Razs) is an effective catalyst in this assay, whereas an Fe-poor montmorillonite (Maroc) is an ineffective cat alyst. In light of the purity of other samples used in the study (i,e., iron ore to test berthierine and granite to test biotite and muscovite), the mineralogical punty of these specimens may be of some concern Woodworth et al. (1982) found that montmonllonite (American Colloid Company, Skokie, Illinois) will affect cell membrane (in Synan-hamster tracheal epithelial cells) as monitored by the release of 51Cr. Montmonllonite is roughly as active as chrysotile in this assaj. However, Dubes and Mack (1988) found that "bentonite" (ob tained from Fisher Scientific Company) is approximately one-tenth as effective as asbestos at mediating transfec tion of mammalian cell cultures. In contrast, Holopainen et al. (1990) found that phlog opite (phl87 ann,,) is almost twice as hemolytic as quartz and as cytotoxic as quartz to rat alveolar macrophages (as determined by the release of LDH). After treatment with nitric and sulfuric acids, the phlogopite is more he molytic but less cytotoxic. In their assay, the hemolytic and cytotoxic activities of muscovite are comparable to those of rutile (a negative control). Modulated 2:1 layer silicates Most studies on samples containing modulated 2:1 lay er silicates suggest that some samples can produce fibrosis or tumors in vivo and can be highly active in vitro. How ever, epidemiological data suggest these minerals are at most mildly active in humans. Sepiolite. One epidemiological study suggests that ex posure to sepiolite-bearing dust does not increase the risk of pulmonary disease. Baris et al. (1980) studied 63 se piolite workers in Turkey involved in trimming, cleaning, and polishing sepiolitic stones. Ten of the 63 showed signs of pulmonary fibrosis, but no relationship was established between exposure to sepiolite and fibrosis Sputum was analyzed from one of the ten, but no ferruginous bodies were observed. In vivo experiments by Wagner (1982) using Fischer 344 rats exposed for 1 yr through inhalation showed that sepiolite (termed by Wagner as "European sepiolite," possessing a fibrous morphology) is as fibrogenic as rie beckite asbestos. However, sepiolite induces no mesothe liomas in Fischer 344 rats exposed by intrapleural inoc ulation, whereas chrysotile (UICC standard B) induces mesotheliomas at a rate of 22.5%. Pott et al. (1990) found that the response elicited by sepiolite is highly sample dependent. The two sepiolite samples studied by Pott et al. (1990) showed tumor rates of 6% (Finland) and 67% (Uicaluaro) following intrapleural injection in female Wistar rats. Preliminary results of a powder X-ray dif fraction study indicate that these samples contain signif icant amounts of other minerals (Guthrie and Bish, un published data); quantitative mineral content data are not available yet, so it is not possible to correlate purity with biological activity. In vitro experiments by Hansen, Mossman, and co workers (Hansen and Mossman, 1987; Mossman et al., 1989) indicate that sepiolite (Minerals Research) is ca pable of inducing the release of the superoxide anion from both hamster and rat alveolar macrophages in a dose- 236 GUTHRIE: EFFECTS OF INHALED MINERALS dependent manner. In hamster alveolar macrophages, the release induced by sepiolite is comparable to the release induced by erionite and riebeckite asbestos; however, m rat alveolar macrophages, sepiolite is less active than er ionite or riebeckite asbestos in eliciting a response. Chamberlain et al. (1982) found that long-fiber sepiolite (source not given) is as cytotoxic as riebeckite asbestos (UICC standard) to mouse peritoneal macrophages (as determined by release of LDH) and human type II alve olar cells (as determined by the formation of giant cells), but less cytotoxic than nebeckite asbestos to Chinesehamster lung cells (as determined by reduction in cloning efficiency); short-fiber sepiolite (source not given), how ever, was determined to be noncytotoxic in the same ex periments. Palygorskite ("attapulgite"). Epidemiological data sug gest that exposure to palygorskite-beanng dusts may in crease the risk of lung cancer among whites (Waxweiler et al., 1988). Waxweiler et al. studied a cohort of 2302 miners and millers from an "attapulgite" company in the United States. They reported SMRs of <1.0 (based on U.S. males) for nonmalignant respiratoiy disease in all races (0.23-0.76, 90% Cl) and race-specific SMRs for lung cancer (1.21-2.93 in whites, 90% Cl; 0.21-1.12 in non whites, 90% Cl). Respirable dust exposures were <5 mg/ m 3, but no information concerning mineral content was given except to note that the only fibrous mineral ob served is "attapulgite clay." No mineral content is re ported for the dust to which their cohort was exposed. Instead, reference was made to the "typical" mineral con tent of "attapulgite" clay mined in the United States as reported by Haden and Schwint (1967); a "typical" dust would thus consist of "70-80% attapulgite; 1-15% montmorillonite, sepiolite, and other clays; 4-8% quartz; and 1-5% calcite or dolomite" (Waxweiler et al., 1988). Sors et al. (1979) reported a case study of a mining engineer who exhibited signs of respiratory disease following a 2-yr exposure to "attapulgite." Lung lavage fluids suggested heavy particle burdens; XRD gave a pattern "similar to those of mineral attapulgite." In vivo experiments have suggested that palygorskitebeanng dusts are mildly active in the lung, though some samples can be very active. Stanton et al. (1981) showed that "attapulgite" is slightly tumorigenic in OsbomeMendel rats following direct application of the dust to the lungs. Experiments using two different samples resulted in tumor rates of 8 5.3% and 11 7.5%; samples were from Attapulgus, Georgia, and >90% pure, the remain der consisting of quartz. Jaurand et al. (1987) found that "attapulgite" (French; obtained from a deposit in Mormoiron) is nontumorigenic following intrapleural injection in specific pathogen-free Sprague-Dawley rats, whereas in the same experiments, chrysotile induces tumors at a rate of 19-52%, depending on particle size. However, Wagner (1982) observed mesothelioma rates of 12.5-25% (de pending on specim en preparation m ethod) for " attapul gite" (Spanish) following intrapleural inoculation of Fi scher 344 rats; chrysotile (UICC standard B) exhibits a comparable mesothelioma rate (22.5%). In his inhalation experiments, Wagner (1982) showed that "attapulgite" (Spanish) is as fibrogenic as nebeckite asbestos, but he reported negative results in the two experiments for an other "attapulgite" (American). Bgin et al. (1987, 1990) used a bronchoalveolar lavage technique to monitor the cellular response in lungs of sheep exposed to "attapul gite" (from northern Florida). Exposure results in increas es in cell numbers and enzyme levels, comparable to those observed after similar experiments using the UICC as bestos standards. No fibrosis was observed at the end of the study, but the elevated levels of enzymes indicate that the "attapulgite" is cytotoxic in vivo. Coffin et al. (1989a) reported a 1.4% incidence of mesotheliomas in rats in jected intrapleurally with "attapulgite" from Georgia and Florida compared with a 1.3% incidence in the control group. Pott et al. (1974, 1990) found that "attapulgite" is carcinogenic at rates from 3.5-40% following intra pleural injection in Wistar rats, i.e., the response is sam ple dependent. Preliminary results of a powder X-ray dif fraction study indicate that these samples contain significant amounts of other minerals (Guthne and Bish, unpublished data). In vitro experiments have indicated that palygorskite is as hemolytic as chrysotile, but in other nonerythrocyte cell types palygorskite is at most slightly cytotoxic and is nongenotoxic. Bignon et al. (1980) showed that "attapul gite" (Spanish) was ~8 times more hemolytic to human red blood cells than chrysotile (UICC standard A; derived from Rhodesian deposits). Perderiset et al. (1989) found that "attapulgite" (Senegalese; obtained from Rhone Poulenc) is hemolytic but that pretreatment of the dust with lipids or proteins (material similar to cell mem branes or extracellular lung fluid) reduces the hemolytic activity. Nadeau et al. (1983) reported that "attapulgite" is as hemolytic as chrysotile (UICC standaid B) and more hemolytic than sepiolite or erionite. In contrast, in vitro experiments using cells other than RBCs have suggested that palygorskite-beanng dusts are generally inactive, although the activity vanes greatly as a function of the surface charactenstics of the sample. Woodworth et al. (1983) found that palygorskite (CMS Clay Repository sample from Nevada) might have a slight effect on cultured hamster trachea, but the effect is not statistically different from the control group, nebeckite asbestos and fiber glass, however, test positive statisti cally in the same assay. Jaurand et al. (1987) found that "attapulgite" (French) may be cytotoxic to rat pleural mesothelial cells only at high doses, whereas chrysotile is generally much more cytotoxic. Reiss et al. (1980) dem onstrated that palygorskite (Attapulgus, Georgia) is much less cytotoxic than "amosite" asbestos to human-embry onic, intestine-derived epithelial cells. Pezerat et al. (1989) found that "attapulgite" (Senegalese) is inactive in cata lyzing the production of O radicals, and Achard et al. (1987) found that the sam e m a te n a l does not induce SCEs in rat pleural mesothelial cells. Renier et al (1990) found no unscheduled DNA repair synthesis in rat pleural me- GUTHRIE: EFFECTS OF INHALED MINERALS 237 sothehal cells following treatment with "attapulgite" (Mormoiron region, France). Chamberlain et al. (1982) found that long-fiber "attapulgite" (source not given) is more cytotoxic than riebeckite asbestos (UICC standard) to mouse peritoneal macrophages (see description of as says in the section on sepiolite) but less cytotoxic than riebeckite asbestos to Chinese-hamster lung cells; shortfiber "attapulgite" (source not given), however, is slightly cytotoxic to mouse peritoneal macrophages and noncytotoxic to Chinese-hamster lung cells. Nolan et al. (1991) further demonstrated that the in vitro activity of palygorskite varies between samples by showing that among nine palygorskites that possess different surface charac teristics there is a corresponding range in hemolytic ac tivity. Zeolites The biological activity of erionite has been studied ex tensively, and all data indicate that it is extremely active in humans, in vivo, and in vitro. Data on other zeolites are less conclusive, particularly in light of the poor qual ity of the samples studied. Erionite. Epidemiological data suggest that exposure to eriomte-beanng dusts increases the risk of mesothelioma in humans, even at much lower exposure levels than re quired for amphibole asbestos-induced mesothelioma. Earlier epidemiological studies in the Cappadocian re gion of Turkey revealed outbreaks of asbestos-related re spiratory diseases, including mesothelioma (Baris et al., 1979). Initially, it was assumed that asbestos present in the stucco used in that region was the cause of these dis eases; however, Bans et al. (1979) found outbreaks in villages in which the stucco does not contain asbestos. Later studies focused on exposure to erionite as the cause. Baris et al. (1987) found zeolite fibers (as determined by ATEM) in air samples from affected villages, although air samples from some of these villages additionally contain fibrous tremolite. Lung contents as determined from spu tum samples (Sebastien et al., 1981, 1984; identification by ATEM and electron diffraction) and biopsies (Baris et al., 1987; identification by ATEM) also indicate that in dividuals from these regions have been exposed to both erionite and asbestos (chrysotile, tremolite asbestos, and riebeckite asbestos), though the amount of zeolite ex ceeded the combined amounts of asbestos (on a per-fiber basis) in the two samples of human lung contents. Mumpton (1979) investigated the mineral content of dusts in the Cappadocian region. He found that erionite was present in samples from two villages in which meso thelioma rates are high, although erionite is abundant only in one of those villages. He also found that erionite is abundant in a third village, Sarihidir, which at that time had no reported cases of mesothelioma. Mumpton concluded, therefore, that the geographic distribution of erionite is inconsistent with the distribution of mesothe lioma. Subsequently, however, Baris et al. (1987) sur veyed Sarihidir and reported three cases of mesothelio ma. Bans et al. also reported fiber characteristics from the affected villages, indicating that zeolite fibers were present ubiquitously but other types of fibers varied be tween villages. The most disturbing implication of the observations in Turkey is that if enomte is indeed the cause of the high rates of mesothelioma, then erionite is capable of induc ing mesothelioma in humans at low exposures. Bans et al. (1987) reported total fiber levels in the villages from 0.004 to 0.175 libers/mL, and these measurements in cluded other dusts in addition to zeolite. Simonato et al. (1989) reported newer estimates of fiber characteristics in Karain and Sanhidir (two of the affected villages) and found levels to be 0.002-0.010 (~80% zeolite) and 0.0010.029 (~60% zeolite), respectively. In vivo expenments have further demonsnaied the high fibrogenic and carcinogenic potential of enomte-beanng dusts. Suzuki and Kohyama (Suzuki, 1982; Suzuki and Kohyama, 1984, 1988) studied the effects on mice of in trapleural injections of two erionite samples, one from Needle Peak, Nevada (Minerals Research; listed by Su zuki and Kohyama as "Needle Park"), and one from an unknown locality (Resource International Company). In mice injected intrapleurally with 2 mg of dust, Needle Peak erionite induces tumors at a rate of 54 5%, com pared with 0-25% for chrysotile and 40 5% for "amosite" asbestos, fibrosis also develops after injection of any of the dusts. They also performed experiments using higher doses, but a low percentage of mice survived long enough to develop tumors (>7 months). Maltom et al. (1982) are investigating the effect of method of exposure to enomte on the induction of mesotheliomas in rats. For their ini tial results, they reported that erionite (sedimentary, ob tained from G. Gottardi, University of Modena, Italy) induces mesotheliomas by intrapleural injection at a rate of 90% (nine of ten rats), whereas riebeckite asbestos in duces mesothelioma by intraperitoneal injection at a rate of 100% (12 of 12). Wagner and coworkers have also studied the in vivo effects of exposure to enomte using specific pathogen-free Sprague-Dawley rats (Wagner, 1982) and Fischer 344 rats (Wagner et al., 1985). They used intrapleural injection (20 mg/rat) and inhalation to study the carcinogenic po tential of four enonite samples; Oregonian enomte (from F. Mumpton, Minerals Research), nonfibrous synthetic zeolite (chemically identical to enomte, from R. Taylor, Laporte Industnes), a New Zealand enomte (similar size distribution to the Oregonian sample, though fibers are slightly thicker), and a Turkish (Karain) rock determined to consist of "poorly consolidated rock . . [made of] in completely formed enonite . . . in an amorphous matrix which has the same composition as enonite" (Wagner ct al., 1985). In the rats intrapleurally injected with Orego nian enonite and Turkish rock, 40 of 40 and 38 of 40 developed mesotheliomas, respectively, with mean sur vival times of 390 and 435 d; the New Zealand enonite was --Vi as potent as the Oregonian and Turkish eno- nites. For comparison, in the same experiment, chrysotile induced 19 mesotheliomas in 40 rats with a mean sur 238 GUTHRIE: EFFECTS OF INHALED MINERALS vival time of 678 d. The same effect was observed in the inhalation experiments: 27 of 28 rats exposed to Orego nian erionite developed mesotheliomas with a mean sur vival time of 580 d compared with one of 28 rats exposed to riebeckite asbestos (UICC standard) with a mean sur vival time of 917 d. The synthetic zeolite was also tested in the inhalation experiments and induced two tumors in 28 rats with a mean survival time of 784 d. With respect to the erionite sample, they stated "No other dusts we have investigated have produced this high incidence of tumours particularly following inhalation" (Wagner et al., 1985). Coffin et al. (1989a) confirmed 1he observation that erionite-treated rats develop mesotheliomas at a higher rate and in a shorter time than chrysotile asbestosor riebeckite asbestos-treated rats; the erionite used was from Rome, Oregon (Minerals Research), and was pre pared by either H20 sedimentation or air elutriation. In two later studies, the Wagner group confirmed their original finding. Johnson and Wagner (1989) exposed Fi scher 344 rats to erionite from Rome, Oregon (obtained from Minerals Research), by inhalation in a dust-rich en vironment (10 mg/m3) and found that erionite exposure produces both fibrosis and mesothelioma; three of the three rats exposed to dust for 12 weeks and allowed to recover for 12 months developed mesothelioma. Hill et al. (1990) used intrapleural injection of Oregonian enomte in Porton rats to determine the dose-response rela tionship for induction of mesotheliomas. They found a sharp rise in mesothelioma rate from 0% mesotheliomas at 0.01 mg/rat to ~90% at 1.0 mg/rat; they stated "erio nite is over 200 times more tumourogenic than crocidohte." In vitro studies have demonstrated that erionite-bearing dusts are both cytotoxic and genotoxic. Poole et al. (1983b) studied the genotoxic effects of Oregonian erio nite (Minerals Research) by monitoring morphological transformations and unscheduled DNA repair in mouseembryo fibroblasts (cells that reside in the connective tis sue and that are responsible for collagen production, i.e., that are involved in fibrosis). Erionite was found to be active in both of these tests, whereas amphibole asbestos does not cause morphological transformations (Poole et al., 1983a). Numerous studies have also demonstrated that erionite is cytotoxic (Palekar et al., 1988; Brown et al., 1989) and genotoxic (Palekar et al., 1987; 1989a, 1989b) to Chinese-hamster lung cells. Brown et al. (1989) found that the cytotoxic effects are related to long, thin fibers, since milling of the sample to reduce the fiber lengths also reduces its activity. The cytotoxic and geno toxic activities of erionite are slightly less than those of asbestos when compared on a mass basis, but comparison on a per-fiber basis shows that the activities of erionite are much greater than those of asbestos (Palekar et al., 1988; Brown et al., 1989). In other tests for genotoxicity, Hansen, Mossman, and coworkers (Hansen and Mossman, 1987; Mossman et al., 1989; Mossman and Sesko, 1990) have shown that er ionite (Rome, Oregon; from R. Davies) is as effective as riebeckite asbestos in catalyzing the production of the superoxide anion from hamster and rat alveolar macro phages. In contrast, however, Pezerat et al. (1989) found that though Oregonian erionite is nearly as effective as most types of asbestos at catalyzing the production of O radicals from an aqueous medium, it is inactive when compared with chrysotile (UICC standard B) or fibrous brucite. Kelsey et al. (1986) found that erionite (Rome, Oregon; obtained from V. Timbrell) induces SCEs slight ly in Chinese-hamster ovary cells, whereas riebeckite as bestos (UICC standard) does not; ultraviolet light, how ever, is much more effective at inducing SCEs. Both Oregonian erionite and nebeckite asbestos (UICC stan dard) induce low levels of chromosomal aberrations in the Chinese-hamster ovary cells (Kelsey et al., 1986). Cof fin et al. are attempting to relate the biological activity of erionite to its mineralogical characteristics (Coffin et al., 1989b). Mordenite and other zeolites. In vivo experiments by Suzuki and Kohyama (Suzuki, 1982; Suzuki and Kohyama, 1984, 1988) suggest that a mordenite-bearing dust and zeolite 4A are fibrogemc but noncarcinogemc. Suzuki and Kohyama used intraperitoneal injection in mice to test the biological response to various zeolites. Included in their experiments was a mordenite-beanng sample (Resource International Company, Denver) that con tained both granular and fibrous morphologies and a syn thetic zeolite, 4A (Union Carbide Corporation); quanti tative X-ray diffraction of the mordenite-bearing sample has shown that it contains --63.5% impurities, including clinoptilolite, feldspar, opal-CT, and gypsum (Guthrie and Bish, unpublished data). In fact, mordenite was used m these experiments to test the relationship between parti cle shape and activity among zeolites. The mordenite sample was described as a mixture of fibrous and nonfibrous forms, despite the fact that mordenite is uniquely fibrous. Hence, it should be recognized that their results apply to an impure mordenite sample. Their experiments show that 10-mg doses of either the mordenite-beanng sample or zeolite 4A induce no tumors m mice for expenments up to 23 months in duration; fibrosis, however, does result from the exposure. On the other hand, the same experiments showed that 0.5- and 2 0-mg doses of enomte (Needle Peak, Nevada) induce tumors at 33 3 and 54.5%, respectively; also, fibrosis in enomte-exposed mice is more pronounced. In their group exposed to a 10mg dose of enonite, a 37.5% rate o f tumor induction was observed, but only eight rats survived to > 7 months, 1 e , exposure to erionite may elicit a dose-dependent response in the lungs of mice, but this relationship cannot be tested because of the poor statistical significance of the results from the group exposed to high concentrations of eno nite. Maltom and Minardi (1988, 1989) studied the biolog ical activity of synthetic zeolites used in detergents (MS 4A and MS 5A, Na and Ca rich, respectively; source not given) by intraperitoneal, intrapleural, and subcutaneous injection in rats. For both zeolites and all routes of ex GUTHRIE: EFFECTS OF INHALED MINERALS 239 posure, tumors appeared but not at rates significantly dif ferent from those observed in the control groups that were injected with H20 (--25--50%). In the same assay, how ever, riebeckite asbestos induces tumors following injec tion into the abdominal cavity at a rate of 97.5%. In vitro experiments suggest that mordenite-bearing dust is much less active than erionite. Hansen, Mossman, and coworkers (Hansen and Mossman, 1987; Mossman et al., 1989; Mossman and Sesko, 1990) found that a mordenite-bearing sample (source not given) is less effec tive than sepiolite and much less effective than erionite at stimulating the release of the superoxide anion from rat alveolar macrophages. Palekar et al. (1988) found that this same mordenite is noncytotoxic to Chinese-hamster lung cells. Quantitative X-ray diffraction of this morden ite-bearing sample has shown that it contains ~ 50.5% impurities, including clinoptilolite, feldspar, and opal-CT (Guthrie and Bish, unpublished data). As was the case for the in vivo experiments of Suzuki and Kohyama, mor denite was used as a nonfibrous-zeolite control in the in vitro assays, despite that mordenite is uniquely fibrous. Rom et al. (1983) discussed the implications of fibrous zeolite health hazards with respect to the western United States and stressed the need for epidemiological studies in this region. T able 2. Summary of data on the biological activities of clays and zeolites Mineral Epidemio logical In vivo in vitro Hematite Goethite Lepidocrocite Boehmite Fibrous brucite Kaolimte Halloysite Antigorlte Berthierine Chlorite (Fe-rich) Talc Mica/mica-like clays Sepiolite Palygorskite Erionite Mordenite Zeolites 4A and 5A + 3 - n.d n.d n d n.d + (t) n.d n.d n.d n.d. - to 4- - to + 4-4-4n d n d __ + to + + (t) + (t) + to + + (<) - to + + (9 4- 4- 4- (t) - tO 4- 4 (t) - to + (t) n.d n.d n.d - - to + (f) 4- 4- 4- (t) - (t) - to + + + (t) - to + 4 + (h o 4-4-4- (t) + + + + (f) m - (t) 4- - (t) - n d n d n d 4- 4+ to + + + n.d 4-4-4++ + - tO + + + - to 4* - to + + + - to 4- 4- + 4-4-4- - tO 4- n d NoteSymbols: - indicates inactive; + indicates active, n d indicates no data available, f and t indicate fibrogemc and tumorigemc, respectively D is c u s s io n The wide range of minerals that have been studied by various techniques offers the potential for revealing the causes of a mineral's biological activity. Indeed, it is clear from the dusts studied already that minerals exhibit dif ferent activities and elicit different biological responses (Table 2). In fact, differences in biological response can be found both between mineral species and between dif ferent samples of the same mineral species. The varia tions in response likely reflect variations in the interac tions between the mineral surfaces and biological components (i.e., cell surfaces, enzymes, proteins, DNA, etc.). Ideally, if this observed variation can be related to differences in the physical and chemical properties of the minerals, then the mechanisms of mineral toxicity may be elucidated. Unfortunately, several mineralogical prob lems are present in the studies reviewed above that make such inferences difficult if not impossible. Generally, the primary mineralogical aspects that are controlled in most biological experiments are the particle shape and size dis tribution and mass concentration or dose employed, since these parameters appear to relate to the material's bio logical activity as determined by in vivo methods (e.g., Stanton et al., 1981). The exact mineral content of the dusts, however, is rarely characterized. In other words, little attention is generally given to the identification and amount of contaminants in the dust sample. Instead, it is assumed that the mineral content of the sample match es the information provided by the supplier. However, samples obtained from most suppliers potentially contain a mixture of minerals and often are simply a different mineral from the one listed on the label. The mordenite samples used in both the in vitro experiments (Hansen and Mossman, 1987; Palekar et a l , 1988; Mossman et al., 1989; Mossman and Sesko, 1990) and the in vivo experiments (Suzuki, 1982, Suzuki and Kohyama, 1984, 1988) illustrate this well. Each sample actually contains a mixture of mordenite, clinoptilolite, feldspar, opal-CT, and gypsum (Guthrie and Bish, unpublished data). The published mordenite-toxicity data actually apply to a mixture of minerals that is 50% mordenite Another mineralogical problem in biological studies is that the surface properties of the samples are generally not adequately characterized. Recent work (e.g., Pezerat, 1990) has suggested that although the fibrous shape of a material may be important in maintaining the particle in ihe target organ or in enhancing surface area, the mech anisms by which minerals are toxic relate to their surface properties, such as active oxidation/reduction sites. In deed, the activity of a mineral varies with surface state (Nolan et al., 1991) and surface area (Gormley and Ad dison, 1983), which in turn can vary substantially be tween samples. Thus, it is not only important to control the surface aspects of a mineral during an experiment, but it is important to characterize these properties so that they can be related to biological activity. These mineralogical deficiencies in biomedical re search can be rectified through collaborative efforts be tween minerals scientists and health scientists. Such col laboration should involve both characterization of the mineralogical aspects of the experiment and design of ex periments that will allow mechanistic questions to be ad dressed. For example, an amphibole has numerous prop 240 GUTHRIE: EFFECTS OF INHALED MINERALS erties that might contribute to its activity (e.g., broken Si-0 bonds, "exchangeable" cations in the A site, poly v a le n t catio n s in th e octahedral sites, underbonded O re sulting from A1 substitution in the tetrahedral sites, spe cific surface periodicities). Hence, the results of a study on amphibole-induced pathogenesis may record effects from several mineralogical properties. On the other hand, it is possible to isolate the effects of a specific mineral ogical characteristic by an appropriate choice of mineral pairs. For instance, the role of polyvalent cations in the octahedral site can be determined by comparing the ac tivities between two minerals that differ only in the com position of the octahedral site (e.g., tremolite and ferroactinolite; annite and phlogopite). This type of approach could be extremely effective for determining mineralogi cal mechanisms of mineral-induced pathogenesis. Such information will lead to both a better understanding of diseases such as cancer and more effective regulation of minerals, since regulations can be based on mineralogical properties additional to particle size and shape. Furthermore, it should be recognized that though a mineral is active in a particular assay, it may pose limited risk to humans. A complete understanding o f the numer ous factors that contribute to mineral-induced pathogen esis is essential before the results of any one assay can be used to predict nsk to humans. Acknowledgments I would like to thank D.L. Bish, G.D. Guthrie, Sr., and C.S. NicholsonGuthne for extensive discussions and encouragement dunng the prepa ration of this manuscript. I also benefited from discussions with F. Mumpton and B. Lehnert. Thoughtful reviews of the manuscnpt were provided by D.L Bish, J. Hughes, B. Lehnert, W Moll, B T Mossman, M. Ross, and D. Vaniman. My time was supported by a postdoctoral fellowship from the Director's office at the Los Alamos National Labo ratory References cited Achard, S., Perdenset, M., and Jaurand, M.-C. (1987) Sister chromatid exchanges in rat pleural mesothelial cells treated with crocidohte, attapulgite, or benzo 3-4 pyrene British Journal of Industrial Medicine, 44, 281-283 Adamis, Z., and Timr, M. (1978) Studies on the effect of quartz, ben tonite and coal dust mixtures on macrophages in utro British Journal of Experimental Pathology, 59, 411-415. Ames, B N , McCann, J , and Yamasaki, E. (1975) Methods for detecting carcinogens and mutagens with the Salmonella/mammahan-microsome mutagenicity test. Mutation Research, 31, 347-364 Appel, J D , Fasy, T M , Kohtz, D S , Kohtz, J D , and Johnson, E.M (1988) Asbestos fibers mediate transformation of monkey cells by ex ogenous plasmid DNA Proceedings of the National Academy of Sci ences U.S A , 85, 7670-7674 Attygalle, D., Hamson, C V., Kang, E.J , and Mohanty, G P. (1954) In fective pneumoconiosis British Journal of Industrial Medicine, 11, 245259. Baris, 1., Simonato, L., Artvinli, M., Pooley, F., Saracci, R., Skidmore, J., and Wagner, C (1987) Epidemiological and environmental evidence of the health effects of exposure to enomte fibres A four-year study in the Cappadocian region o f Turkey International Journal of Cancer, 39, 10-17. Bans, Y.I., Anvinli, M., and Sahm, A.A. (1979) Environmental meso thelioma in Turkey. Annals of the New York Academy of Science, 330, 423-432. Bans, Y I , Sahin, A A , and Erkan, M L. (1980) Clinical and radiological study in sepiohte workers. Archives of Environmental Health, 35, 343-- 346 Bgin, R., Mass, S , Rola-Pleszczynski, M , Geoffroy, M , Martel, M , Desmarais, Y , and Sbastien, P (1987) The lung biological activity of American attapulgite Environmental Research, 42, 328-339 Bgin, R , Mass, S , and Sebastien, P (1990) Assexsment of the fibrogemcity of attapulgite In J Bignon, Ed , Health related effects of phyllosihcates, NATO ASI Senes, vol G 21,p 395-403 Springer-Verlag, Berlin Bignon, J , Sbastien, P , Gaudichet, A , and Jaurand, M C (1980) Bio logical effects of attapulgite. In J C Wagner, Ed , Biological effects of mineral fibres, vol. 1, IARC Scientific Publication 30, p 163-181 World Health Organization, Lyon, France Bish, D L , and Chipera, S.J. (1991) Detection of trace enomte using X-ray powder diffraction Enomte m tuffs of Yucca Mountain, Nevada, and central Turkey Clays and Clay Minerals 39. 437-445 Boyd, J.T , Doll, R , Faulds, J.S., and Leiper, J (1970) Cancer of the lung in iron ore (haematite) miners. Bntish Journal of Industrial Medicine, 27, 97-105 Brambilla, C , Abraham, J., Brambilla, E., Benirschke, K., and Bloor, C (1979) Comparative pathology of silicate pneumoconiosis Amencan Journal of Pathology, 96, 149-163 Brown, D.P., and Wagoner, J.K. (1980) Occupational exposure to talc containing asbestos: III. Retrospective cohort study of mortality U S Department of HEW (NIOSH) Publication 80-115, 31-33. Brown, R C , Chamberlain, M , Davies, R , Morgan, D M L , Pooley, F D ., and Richards, R J (1980) A comparison of 4 in vilro systems applied to 21 dusts In E.G Beck and J Bignon, I d s , In vitro effects of mineral dusts, NATO ASI Senes, p. 47-52 Spnnger-Verlag, Berlin Brown, R.C , Davies, R., and Rood, A.P. (1989) Modification of fibrous Oregon enomte and its effects on in vitro activity In J Bignon, J Peto and R. Saracci, E d s, Non-occupational exposure to mineral fibres, p 74-80. International Agency for Research on Cancer, Lyon, France Campbell, J.A. (1940) Effects of precipitated silica and of iron oxide on the incidence of pnmary lung tumours in mice Bntish Medical Jour nal, 2, 275-280 Chamberlain, M., and Brown, R.C. (1978) The cytotoxic effects of asbes tos and other mineral dust in tissue culture cell lines Bntish Journal of Expenmental Pathology, 59, 183-189 Chamberlain, M., and Tarmy, E.M. (1977) Asbestos and glass fibers in bacterial mutation tests. Mutation Research, 43, 159-164 Chamberlain, M., Davies, R , Brown, R C , and Gnffiths, D M (1982) In vitro tests for the pathogenicity of mineral dusts \nnals of Occupa tional Hygiene, 26, 583-592. Chen, S.Y , Hayes, R.B., Liang, S.R., Li, Q G., Stewart, P A., and Blair, A. (1990) Mortality experience of haematite mine workers in China. Bntish Journal of Industrial Medicine, 47, 175-181. Chipera, S.J , and Bish, D L (1989) The occurrence and distnbution of enomte at Yucca Mountain, Nevada Los Alamos National Laboratory Report LA-11663-MS, 20 p. Churg, A , Wiggs, B., DepaoU, L., Kampe, B , and Stevens, B (1984) Lung asbestos content in chrysotile workers with mesothelioma Amer ican Reviews of Respiratory Disease, 130, 1042-1045 Churg, A , Wnght, J L , Depaoli, L , and Wiggs, B (1989) Minralogie correlates of fibrosis in chrysotile miners and millers Amencan Re views of Respiratory Disease, 139, 891-896 Coffin, D L , Palekar, L.D , Cook, P M , and Creason, J P (1989a) Companson of mesothelioma induction in rats by asbestos and nonasbestos mineral fibers1Possible correlation with human exposure data In A P Wehner, Ed., Biological interaction of inhaled mineral fibers and cig arette smoke, p. 347-354. Battelle Press, Seattle Coffin, D L , Peters, S E , Palekar, L D , and Stahel, E P (1989b) A study of the biological activity of enomte in relation to its chemical and structural characteristics. In A P Wehner, Ed , Biological interaction of inhaled mineral fibers and cigarette smoke, p 355-372 Battelle Press, Seattle. Costa, D., Guignard, J , and Pezerat, H (1990) Oxidi/ing surface prop erties of divalent iron-nch phyllosilicates in relation to their toxicity by oxidative stress mechanism In J. Bignon, Ed., Health related effects GUTHRIE: EFFECTS OF INHALED MINERALS 241 of phyllosilicates, NATO ASI Senes, vol. G21, p. 129-134. SpnngerVerlag, Berlin Cullman, P., and McDonald, J C. (1990) Respiratory disease from occu pational exposure to non-fibrous phyllosilicates. In J. Bignon, Ed., Health related effects of phyllosilicates, NATO ASI Series, vol G21, p 161-- 178. Sponger*Verlag, Berlin. Davies, R (1983) Factors involved m the cytotoxicity of kaohmte to wards macrophages in vitro Environmental Health Perspectives, 51, 249-252. Davis, B L (1990) Quantitative analysis of asbestos minerals by the ref erence intensity X-ray diffraction procedure. Journal of the American Industnal Hygiene Association, 51, 297-303. Dubes, G R , and Mack, L.R. (1988) Asbestos-mediated transfection of mammalian cell cultures. In Vitro Cellular and Developmental Biolo gy, 24, 175-182 Endo-Capron, S., Fleury-Feith, J., Nebut, M., De Neef, R., and Jaurand, M.C (1990) Some in vivo and m vitro studies earned out with talc samples. In J. Bignon, Ed , Health related effects of phyllosilicates, NATO ASI Senes, vol. G21, p. 369-375. Spnnger-Verlag, Berlin. Environmental Protection Agency (1986) Airborne asbestos health as sessment update, EPA/600,8-84/003F. Gardner, L V , Dworski, M., and Deiahanl, A.B (1944) Aluminum ther apy in silicosis. Journal of Industnal Hygiene and Toxicology, 26, 211223 Ghio, A , Kennedy, T , Baser, M , Rawlings, W R , Rao, N V , and Hoidal, J.R. (1990) Kaolin generates *OH and causes hemolysis by acting as a Fenton reagent In J Bignon, Ed., Health related effects of phyllosihcates, NATO ASI Senes, vol. G21, p 361-368 Spnnger-Verlag, Berlin Gormley, I.P , and Addison, J (1983) The in vitro cytotoxicity of some standard clay mineral dusts of respirable size. Clay Minerals, 18, 153163 Haden, W L., Jr., and Schwint, I.A. (1967) Attapulgite: Its properties and applications. Industnal and Engmeenng Chemistry Research, 59, 5869 Hale, LW ., Gough, J., King, E.J., and Nagelschmidt, G. (1956) Pneu moconiosis of kaolin workers Bntish Journal of Industnal Medicine, 13, 251-259 Hansen, K , and Mossman, B.T. (1987) Generation of superoxide (0 2~) from alveolar macrophages exposed to asbestiform and nonfibrous parucles. Cancer Research, 47, 1681-1686. Heimann, H., Moskowitz, S , Hanhara, C.R , Gupta, M.N., and Mankiker, N S (1953) Silicosis m mica mining in Bihar, India. Archives of Industnal Hygiene and Occupational Medicine, 8, 420-435. Heppleston, A G (1984) Pulmonary toxicology of silica, coal and asbes tos. Environmental Health Perspectives, 55, 111-127. Hill, R.J., Edwards, R E., and Carthew, P (1990) Early changes m the pleura! mesothehum following intrapleural inoculation of the mineral fibre enomte and the subsequent development of mesotheliomas Jour nal of Expenmental Pathology, 71, 105-118. Holopainen, M , Vallyathan, V., Hedenborg, M., and Klockars, M. (1990) Toxicity of phlogopite and muscovite in vitro In J. Bignon, Ed , Health related effects of phyllosilicates, NATO ASI Senes, vol. G21, p 349360 Spnnger-Verlag, Berlin. Holt, P.F , Lindsay, H., and Beck, E G (1970) Some derivatives of polyvinylpyndine 1-oxides and their effect on the cytotoxicity of quartz in macrophage cultures. Bntish Journal of Pharmacology, 38, 192-201. Jaurand, M.C., Magne, L,, and Bignon, J. (1980) Effects of well-defined fibres on red blood cells and alveolar macrophages. In J.C. Wagner, Ed., Biological effects of mineral fibres, vol. 1, IARC Scientific Publi cation 30, p. 441-450 World Health Organi?ation, Lyon, France Jaurand, M .-C, Fleury, J., Monchaux, G , Michehne, N , and Bignon, J. (1987) Pleural carcinogenic potency of mineral fibers (asbestos, attapulgite) and their cytotoxicity on cultured cells. Journal o f the National Cancer Institute, 79, 797-804. Johnson, N.F., and Wagner, J.C. (1989) Effect of enomte inhalation on the lungs of rats In A.P. Wehner, Ed., Biological interaction of inhaled mineral fibers and cigarette smoke, p. 355-372. Battelle Press, Seattle. Kelsey, K.T., Yano, E., Liber, H.L., and Little, J B. (1986) The in vitro genetic effects of fibrous enomte and crocidohte asbestos Bntish Jour nal of Cancer, 54, 107-114. Kennedy, T P., Rawhngs, W., Jr., Baser, M , and Tockrnan, M (1983) Pneumoconiosis in Georgia kaolin workers Amencan Reviews of Re spiratory Disease, 127, 215-220 Kennedy, T.P., Dodson, R., Rao, N.V., Ky, H , Hopkms, C , Baser, M . Tolley, E , and Hoidal, J.R. (1989) Dusts causing pneumoconiosis gen erate *OH and produce hemolysis by acting as Fento i catalysts Ar chives of Biochemistry and Biophysics, 269, 359-364 Kettle, EH . (1934) Expenmental pneumoconiosis Infective silicatosis Journal of Pathology and Bactenology, 38, 201-208 King, EJ., and Hamson, C V (1948) The effects of kaolin on the lungs of rats Journal of Pathology and Bactenology, 60, 435-440 King, E.J., Gilchnst, M., and Rae, M.Y (1947) Tissue reaction to sencite and shale dusts treated with hydrochlonc acid An expenmental in vestigation on the lungs of rats. Journal of Pathology and Bactenology, 59, 325-327 King, E J , Hamson, C V., and Mohanty, G.P (1955) The effect of vanous forms of alumina on the lungs of rats. Journal of Pathology and Bacteriology, 69, 81-93 Kleinfeld, M , Messite, J , Kooyman, O , and Zak>, M H (1967) Mortaln> among talc miners and millers in New York State Archives of Envi ronmental Health, 14, 663-667 KJemfeld, M , Mcssite, J , and Zaki, M H (J974) Mortality experiences among talc workers: A follow-up study Journal of Occupational Med icine, 16,345-349 Langer, A.M., and Nolan, R.P (1989) Fibre type and burden in paren chymal tissues of workers occupationally exposed to asbestos in the United States. In J. Bignon, J. Peto, and R. Saracci, Eds., Non-occupational exposure to mineral fibres, p 330-335 International Agency for Research on Cancer, Lyon, France Lapenas, D , Gale, P , Kennedy, T , Rawlings, W , J r , and Dietnch, P (1984) Kaolin pneumoconiosis Amencan Reviews of Respiratory Dis ease, 130, 282-288 Lawler, A.B., Mandel, J.S., Schuman, L M , and Lubin, J H (1985) A retrospective cohort mortality study of iron ore (hematite) miners in Minnesota Journal of Occupational Medicine, 27, 507-517 Leake, B E. (1978) Nomenclature ofamphiboles Canadian Mineralogist, 16, 501-520. Lehnert, B.E (1990) Lung defense mechanisms against deposited dusts Problems in Respiratory Care, 3, 130-162. Leophonte, P,, and Didier, A. (1990) French talc pneumoconiosis In J Bignon, Ed., Health related effects of phyllosilicates, NATO ASI Senes, vol G21, p. 203-209 Spnnger-Verlag, Berlin Low, R B , Leffingwell, C M , and Bulman, C A (1980) Effects of kaolinite on amino acid transport and incorporation into protein by rabbit pul monary alveolar macrophages Archives of Environmental Health, 35, 217-223 Lynch, K.M., and Mclver, F.A (1954) Pneumoconiosis from exposure to kaolin dust* Kaolinosis Amencan Journal of Pathology, 30, 11171122. Maltom, C , and Minardi, F (1988) First available results o f long-term carcinogenicity bioassay on detergency zeolites (MS 4A and MS 5A) Annals of the New York Academy of Science, 534, 978-985 -------- (1989) Recent results of carcinogenicity bioassays of fibres and other particulate matenals. In J Bignon, J Peto, and R Saracci, Eds , Non-occupational exposure to mineral fibres, p 46-5 1 International Agency for Research on Cancer, Lyon, France Maltom, C., Minardi, F , and Monsi, L (1982) Pleural mesothelioma in Sprague-Dawley rats by enomte First expenmental evidence Envi ronmental Reseaich, 29, 238-244. Marks, J , and Nagelschmidt, G. (1959) Study of the toxicity of dust with use of the in vitro dehydrogenase technique Amencan Medical Asso ciation Archives of Industrial Health, 20, 37/383-43/389 McDonald, J.C., McDonald, A.D., Sebastien, P., and May, K ( 1988) Health of vermicuhte miners exposed to trace amounts o f fibrous tremolite. Bntish Journal oflndustnal Medicine, 45, 630-634 Mossman, B.T., and Craighead, J.E. (1982) Comparative cocarcinogemc effects of crocidohte asbestos, hematite, kaolin and carbon in implanted tracheal organ cultures Annals of Occupational Hygiene, 26, 553-567 Mossman, B.T., and Marsh, J.P. (1989) Evidence supporting a role for active oxygen species in asbestos-induced toxicity ard lung disease Environmental Health Perspectives, 81, 91--94 242 GUTHRIE: EFFECTS OF INHALED MINERALS Mossman, B T , and Sesko, A.M. (1990) In vitro assays to predict the pathogenicity of mineral fibers. Toxicology, 6 0 ,5 3 -6 1 Mossman, B.T., Hansen, K.., Marsh, J P., Brew, M.E., and Petruska, J. (1989) Mechanisms of fibre-induced superoxide release from alveolar macrophages and induction of superoxide dismutase in the lungs of rats inhaling crocidolite. In J Bignon, J Peto, and R. Saracci, Eds., Non-occupational exposure to mineral fibres, p 81-92 International Agency for Research on Cancer, Lyon, France Mossman, B T , Bignon, J , Com, M , Seaton, A., and Gee, J B.L (1990) Asbestos1Scientific developments and implications for public policy. Science, 247, 294-301 Mumpton, F.A (1979) A reconnaissance study of the association of ze olites with mesothelioma occurrences in central Turkey. U.S Geolog ical Survey, Open-file Report 79-954, 55 p Nadeau, D , Paradis, D , Gaudreau, A , Pele, J P , and Calvert, R (1983) Biological evaluation of various natural and man-made mineral fibers. Cytotoxicity, hemolytic activity and chemilunescence study Environ mental Health Perspectives, 51, 374. Nolan, R P , Langer, A.M., and Herson, G.B. (1991) Characterisation of palygorskite specimens from different geological locales for health haz ard evaluation British Journal oflndustnal Medicine, 48, 463-475 Oberdorster, G., and Lehnert, B.E. (1990) Particle properties, tumors and fibrosis (abs). Mechanisms in fibre carcinogenesis, p. 57. NATO Ad vanced Research Workshop, Albuquerque, New Mexico Ogle, C J , Rundle, E M., and Sugar, E.T. (1989) China clay workers in the south west of England: Analysis of chest radiograph readings, ven tilating capacity and respiratory symptoms in relation to type and du ration of occupation. British Journal of Industrial Medicine, 46, 261-- 270 Oldham, P D (1983) Pneumoconiosis m Cornish china clay workers. British Journal oflndustnal Medicine, 40, 131-137 Palekar, L D , Eyre, J F , Most, B M , and Coffin, D L (1987) Metaphase and anaphase analysis of V79 cells exposed to enonite, UICC chrysotile and UICC crocidolite Carcinogenesis, 8, 553-560. Palekar, L D., Most, B.M , and Coffin, D L (1988) Significance of mass and number ol fibers in the correlation of V79 cytotoxicity with tumongemc potential of mineral fibers. Environmental Research, 46, 142152. Palekar, L D , Eyre, J F., and Coffin, D. (1989a) Transplantation and chromosomal analysis of cell lines derived from mesotheliomas in duced in rats with enonite and UICC chrysotile asbestos. In J. Bignon, J Peto, and R Saracci, Eds., Non-occupational exposure to mineral fibres, p. 161-172. International Agency for Research on Cancer, Lyon, France -------- (1989b) Chromosomal changes associated with tumongemc min eral fibers. In A P. Wehner, Ed , Biological interaction of inhaled min eral fibers and cigarette smoke, p 355-372. Battelle Press, Seattle. Perdenset, M., Saint Etienne, L., Bignon, J., and Jaurand, M C. (1989) Interactions of attapulgite (fibrous clay) with human red blood cells. Toxicology Letters, 47, 303-309. Pezerat, H. (1990) Surface activity of mineral dusts and processes of ox idative stress (abs.). Mechanisms in fibre carcinogenesis, p. 60. NATO Advanced Research Workshop, Albuquerque, New Mexico Pezerat, H., Zalma, R., and Guignard, J (1989) Pioduction of oxygen radicals by the reduction of oxygen arising from the surface activity of mineral fibres In J Bignon, J. Peto, and R Saracci, Eds., Non-occupational exposure to mineral fibres, p. 100-111. International Agency for Research on Cancer, Lyon, France. Pigott, G H., and Pinto, P.J. (1983) Effects of nonfibrous minerals in the V79-4 cytotoxicity test Environmental Health Perspectives, 51, 173-- 179 Pohcard, A (1934) The action of mica dust on pulmonary tissue Journal o f Industrial Hygiene, 16, 160-164. Poole, A., Brown, R .C , and Fleming, G.T.A (1983a) Study of the celltransforming ability of amosite and crocidolite asbestos and the ability to induce changes in the metabolism and macromolecular binding of benzo(a)pyrene in C 3H 10T`/2cells Environmental Health Perspectives, 51,319-324. Poole A Brown. R C . Turver, C J . Skidmore, J W , and Griffiths, D M (1983b) In vitro genotoxic activities of fibrous enonite. Bntish Journal of Cancer, 47, 697-705- Pott, F. (1980) Animal expenments on biological effect of mineral fibres In J C Wagner, Ed , Biological effects of mineral fibres, vol l. IARC Scientific Publication 30, p. 261-272. World Health Organization, Lyon, France Pott, F., and Friedrichs, K H (1972) Tumoren der ratte nach 1 p -mjektion faserformiger Stube. Die Naturwissenschaften, 59, 318. Pott, F , Huth, F , and Fnednchs, K H (1974) Tumongemc effect of fibrous dusts in experimental animals Environmental Health Perspec tives, 9, 313-315 Pott, F , Bellmann, B , Muhle, H., Rddelsperger, K.., Rippe, R M , Roller, M., and Rosenbruch, M. (1990) Intrapentoneal injection studies for the evaluation of the carcinogenicity of fibrous phyllosilicates In J Bignon, Ed., Health related effects of phyllosilicates, NATO AS1 Senes, vol G21, p. 319-329 Springer-Verlag, Berlin Puledda, S , and Marconi, A (1990) Quantitative X-ray diffraction anal ysis of asbestos by the silver membrane filter method Application to chrysotile. Journal o f the Amencan Industrial Hygiene Association, 51, 107-114 Radford, E.P., and Renard, K.G.St.C. (1984) Lung cancer m Swedish iron miners exposed to low doses of radon daughters. New England Journal of Medicine, 310, 1485-1494 Rail, D.P. (1988) Laboratory animal toxicity and carcinogenesis testing Underlying concepts, advantages, and constraints Annals of the New York Academy of Science, 534, 78-83 Reiss, B., Millette, J R., and Williams, G.M. (1980) The activity of en vironmental samples in a cell culture test for asbestos toxicity Envi ronmental Research, 22, 315-321. Renier, A., Levy, F., Pilliere, F., and Jaurand, M.C (1990) Unscheduled DNA synthesis in rat pleural mesotheha! cells treated with mineral fibres Mutation Research, 241, 361-367 Roberts, W L., Campbell, T J , and Rapp, G R , Jr (1990) Encyclopedia of minerals (2nd edition), 979 p Van Nostrand Reinhold, New York Rom, W.N., Casey, K.R., Parry, W T , Mjaatvedt, C H , and Moatamed, F (1983) Health implications of natural fibrous zeontes for the mtermountain west. Environmental Research, 30, 1-8 Rosmamth, J., Hilscher, W , and Schyma, S B (1990) The effect of the surface quality on the fibrogemcity of the phyllosilicates muscovite and kaolimte. In J. Bignon, Ed., Health related effects of phyllosilicates, NATO ASI Senes, vol G21, p. 123-128. Spnnger-Verlag, Berlin Ross, M. (1981) The geologic occurrences and health hazards of amphi bole and serpentine asbestos. In Mineralogical Society of Amenca Re views in Mineralogy, 9A, 279-323. -------- (1984) A survey of asbestos-related disease in trades and mining occupations and in factory and mining communities as a means of predicting health nsks of nonoccupational exposure to fibrous minerals In B Levadle, E d , Definitions for asbestos and other health-related silicates, ASTM STP 834, p. 51-104. Amencan Society for Testing and Materials, Philadephia Schlesinger, R.B., and Driscoll, K.E. (1989) Respiratory tract defense mechanisms and their interaction with air pollutants In M A Hollinger, Ed , Focus on pulmonary pharmacology and toxicology, p 37-72 CRC Press, Boca Raton, Flonda. Schyma, S B. (1990) The physical characterization of muscovite and kaolinite dusts. In J. Bignon, Ed., Health related effects of phyllosilicates, NATO ASI Senes, vol G 2 l,p 97-105 Spnnger-Verlag, Berlin Sbastien, P , Gaudichet, A , Bignon, J , and Bans, Y I (1981) Zeolite bodies in human lungs from Turkey. Laboratory Investigations, 44, 420-425 Sbastien, P., Bignon, J , Bams [sic], Y I , Awad, L , and Petit, G (1984) Ferruginous bodies in sputum as an indication of exposure to airborne mineral fibers in the mesothelioma villages of Cappadocia Archives of Environmental Health, 39, 18-23 Selevan, S G., Dement, J M., Wagoner, J K , and Froines, J R (1979) Mortality patterns among miners and millers of non-asbestiform talc Preliminary report. Journal of Environmental Pathology and Toxicol ogy, 2, 273-284 Selikoff, I.J , Churg, J , and Hammond, E.C (1964) Asbestos exposure and neoplasia. Journal of the Amencan Medical Association, 188, 2226 -------- (1965) Relation between exposure to asbestos and mesothelioma New England Journal of Medicine, 272, 560-565. GUTHRIE. EFFECTS OF INHALED MINERALS 243 Sepulveda, M.H., Vallyathan, V., Attfield, M D., Piacitelh, L., Tucker, J.H (1983) Pneumoconiosis and lung function in a group of kaolin workers. American Reviews of Respiratory Disease, 127, 231-235. Sheers, G (1964) Prevalence of pneumoconiosis in Cornish kaolin work ers. British Journal of Industrial Medicine, 21, 218-225. Shelby, M D. (1988) The genetic toxicity of human carcinogens and its implications Mutation Research, 204, 3-15. Shishido, S., Iwai, K , and Tudagoshi, K. (1989) Incidence of ferruginous bodies in the lungs dunng a 45-year pcnod and mineralogical analysis of the core fibres and uncoated fibres. In J. Bignon, J Peto, and R Saracci, E ds, Non-occupational exposure to mineral fibres, p 229-- 238 International Agency for Research on Cancer, Lyon, France Simonato, L , Bans, R., Saracci, R., Skidmore, J., and Winkelmann, R. (1989) Relation of environmental exposure to erionite fibres to risk of respiratory cancer. In J. Bignon, J. Peto, and R Saracci, Eds., Nonoccupational exposure to mineral fibres, p 398-405. International Agency for Research on Cancer, Lyon, France Snyder, R L , and Bish, D.L (1989) Quantitative analysis In Mineral ogical Society of Amenca Reviews in Mineralogy, 20, 101-144 Sors, H., Gaudichet, A., Sbastien, P., Bignon, J., and Even, P (1979) Lung fibrosis after inhalation of fibrous attapulgite Thorax, 34, 695. Stacy, B D , King, E J , and Harrison, C.V. (1959) Tissue changes in rats' lungs caused by hydroxides, oxides, and phosphates of aluminium and iron Journal of Pathology and Bacteriology, 77, 417-426 Stanton, M .F, Layard, M., Tegens, A., Miller, E., May, M , Morgan, E., and Smith, A (1981) Relation of particle dimension to carcinogenicity in amphibole asbestos and other fibrous minerals Journal of the Na tional Cancer Institute, 67, 965-975. Steel, R F , and Anderson, W (1972) The interaction between kaohnitc and Staphylococcus aureus. Journal of Pharmacy and Pharmacology, 24, 129 Stille, W T , and Tabershaw, I.R (1982) The mortality experience of up state New York talc workers. Journal of Occupational Medicine, 24, 480-484. Suzuki, Y (1982) Carcinogenic and fibrogemc effects of zeolites Prelim inary observations. Environmental Research, 27, 433-445. Suzuki, Y , and Kohyama, N (1984) Malignant mesothelioma induced by asbestos and zeolite in the mouse pentoneal cavity Environmental Research, 35, 277-292. -------- (1988) Carcinogenic and fibrogemc effects of erionite, mordemte, and synthetic zeolite 4A In D Kali and H S Sherry, E ds, Occur rence, properties and utilization of natural zeolites, p. 829-840. Akadmiai Kiad, Budapest. Sykes, S.E., Morgan, A., Evans, J .C , Evans, N., Holmes, A , and Moores, S.R (1982) Use of an in vivo test system to investigate the acute and sub-acute responses of the rat lung to mineral dusts Annals of Occu pational Hygiene, 26, 593-605. Vestal, T F , Winstead, J A , and Joliet, P V (1943) Pneumoconiosis among mica and pegmatite workers Industrial Medicine, 12, '1 -1 4 Vorwald, A.J., and Karr, J.W (1938) Pneumoconiosis and pulmonary carcinoma American Journal of Pathology, 14,49-57 Wagner, J.C (1982) Health hazards of substitutes. World Symposium on Asbestos, Proceedings, May 1982, Montreal, p 244-266 Canadian Asbestos Information Centre, Montreal -------- (1990) Review on pulmonary effects of phyllosilicates after inha lation In J Bignon, Ed , Health related effects of phyllosihcates, NATO ASI Senes, vol. G21, p. 309-318 Springer-Verlag, Berlin Wagner, J.C., Sleggs, C.A., and Marchand, P. (1960) Diffuse pleural meso thelioma and asbestos exposure in the north western (2ape Province British Journal of Industrial Medicine, 17, 260-271. Wagner, J.C., Berry, G , and Timbrell, V (1973) Mesothehomata in rats after inoculation with asbestos and other materials. British Journal of Cancer, 28, 173-185. Wagner, J.C., Berry, G , Hill, R J., Skidmore, J W , and Pooley, F D (1979) An animal model for inhalation exposure to tak In R Lemen and J M Dement, Eds , Dusts and disease, p 389-392 Pathotox Pub lishers, Park Forest South, Illinois. Wagner, J.C., Skidmore, J.W., Hill, R J., and Griffiths, D M (1985) Erlomte exposure and mesotheliomas in rats. Bntish Journal of Cancer, 51, 727-730. Wastiaux, A., and Daniel, H. (1990) Pulmonary toxicity of kaolin m rats exposed by inhalation In J. Bignon, Ed , Health related effects of phyl losihcates, NATO ASI Senes, vol. G21, p 405-414 Spnnger-Verlag, Berlin. Waxweiler, R.t Zumwalde, R D , Ness, G O , and Brown, D P (1988) A retrospective cohort mortality study of males mining and milling at tapulgite clay Amencan Journal of Industrial Medicine, 13,305-315 Wehner, A.P. (1980) Effects of inhaled asbestos, asbesto-v plus cigarette smoke, asbestos-cement and talc baby powder m hamsters In J C Wagner, Ed., Biological effects of mineral fibres, vol 1 ,1ARC Scientific Publication 30, p 373-375 World Health Organization Lyon, France Witmer, C , and Cooper, K (1983) Lack of mutagemcry of baghouse samples m a modified Ames plate assay. Environmental Health Per spectives, 51, 373 Woodworth, C.D., Mossman, B.T , and Craighead, J.E. ( 982) Compar ative effects of fibrous and nonfibrous minerals on cells and liposomes Environmental Research, 27, 190-205 -------- (1983) Induction of squamous metaplasia in organ cultures of hamster trachea by naturally occurring and synthetic fibers. Cancer Research, 43, 4906-4912 Manuscript received A uoust 14, 1991 Manuscript accepted N ovember 4, 1991