Document nkLYmn3oGe4x9wEyZ4EB92EnR

[CANCER RESEARCH 43, 4906-4912, October 1983] Induction of Squamous Metaplasia in Organ Cultures of Hamster Trachea by Naturally Occurring and Synthetic Fibers1 Craig D. Woodworth, Brooke T. Mossman,2 and John E. Craighead Department of Pathology, University of Vermont College of Medicine, Burlington, Vermont 05405 ABSTRACT Asbestos exhibits many properties of classical tumor pro moters. These characteristics include the ability to stimulate proliferation and inhibit normal differentiation of cells. In organ cultures of trachea, crocidolite and amosite asbestos stimulate squamous metaplasia, a pathological process in which a rapidly proliferating squamous epithelium replaces the normal epithe lium. We hypothesized that the induction of metaplasia depends upon the fibrous nature of asbestos. Accordingly, several natu rally occurring and synthetic fibrous materials and their nonfibrous analogues were assessed for their ability to induce meta plastic changes in tracheal mucosa of the Syrian hamster. Ex posure to both crocidolite asbestos and fiberglass resulted in significant increases (p < 0.05) in squamous metaplasia over a range of dosages (1.0, 4.0, 16.0 mg/mi). Attapulgite (palygorskite) and both "long-" and "short-" fiber preparations of chrysotile asbestos had similar but less marked effects. Nonfibrous analogues of each material (riebeckite, antigorite, and glass particles) failed to produce metaplasia. Asbestos, and fibrous materials in general, appear to stimulate squamous metaplasia because of their fibrous geometry. INTRODUCTION Asbestos has enormous commercial importance (3, 21) be cause it is durable and resistant to both heat and fire. Asbestos acts as a cocarcinogen in the respiratory tract (6, 9, 24), and exhibits properties of classical tumor promotors (for review, see Refs 6, 17, and 28). These characteristics include the ability to stimulate proliferation of cells and alter their normal differentia tion. For example, crocidolite and amosite asbestos induce squa mous metaplasia in cultured tracheal epithelium (18). Although the chemical composition of asbestos minerals might influence pathogenicity (11), most experimental evidence indicates that physical parameters such as the length and diameter of fibers are important (10,13, 26, 33). Fiber geometry is critical in the experimental induction of mesotheliomas in rats (26). Intrapleural inoculation of materials comprised of fibers which are long and thin results in tumors, whereas short, thick fibers and particles are less tumorigenic. We hypothesized that fibrous morphology also might be impor tant in the induction of proliferation and squamous metaplasia by asbestos. To test this, several different fibrous materials and their nonfibrous analogues were examined for their ability to induce these alterations in organ culture^ of hamster trachea. We used chrysotile and crocidolite, the 2 types of asbestos of ' Supported by National Institute for Occupational Safety and Health Grant PHSROI00888, Grant PHSROI33501 from the National Cancer Institute and Grant BC-415 from the American Cancer Society. 2 To whom requests for reprints should be addressed. Received February 2,1983; accepted July 12, 1983. greatest commercial importance. Attapulgite, a naturally occur, ring fibrous clay mineral, and 2 forms of fiberglass also wen examined. MATERIALS AND METHODS Preparation of Minerals. The sources of materials tested in these studies are listed in Table 1. Fibrous materials were prepared as follow*. /3-Fiberglass yam was cut to varying lengths using a Sorvall TC-2 tissut sectioner. Code-100 fiberglass wool was ground gently in a tisau*, homogenizer since the random orientation of fibers made cutting ineffto* >five. Attapulgite, crocidolite, and "long" and "short" chrysotile (Tabtoi)' 2 and 3) were used without further modification. Samples of the miners* antigorite and riebeckite were prepared from rocks selected because of' their mineralogical purity, and glass particles were produced by fusta code-100 fiberglass at 750. These latter materials were ground in a bitfv mill for 15 min, yielding powders with heterogeneous size distribution!,'' The large particles (>5 pm in diameter) then were separated by aqueout i sedimentation (27). The suspensions of minerals were dried, sterizad;-, (125 for 16 hr), and stored in powdered form. ;l; Materials were dispersed in HBSS3 (Grand Island Biological Co., GrmQ'j' Island, N. Y.) by a 2-min bath sonication (Model B-22-4 Branson Ultfj$ sonic Cleaner; Branson Cleaning Equipment Co., Shelton, Conn.) " addition to organ cultures. Chrysotile was not sonicated because w< procedure alters the size of the fibers (25). .}! Characterization of Minerals'. The mineralogical purity, surfaot'/i charge, and size distributions of the particles were characterized. W* used X-ray diffraction to assess mineralogical purity (Table 1). Th# electrophoretic mobility, an indication of net surface charge (22), wm measured after suspending the particles in MEM (Grand Island Biologlcii. Co.) and the zeta potential (22) was assessed after dispersal of select*! fibers in distilled HaO (Table 1). Zeta potential and electrophoretic mobity are similar; however, the latter is applicable when particles are suspended in solutions containing organic compounds. When suspended in MEM, particles had no detectable net surface charge. Size distributions were assessed using SEM (Tables 2 and 3). Sus pensions of each material (10 pg/ml) were collected on Nucleopore filters (Nucleopore Filter, Pleasanton, Calif.) by pressure filtration. Regions from each filter, which appeared to represent the normal distribution of fibers, were photographed at magnifications ranging from xlOO to xl 0,000. Photographs of adjacent regions were assembled as a montage to alow measurement of long fibers. A planimeter was used to calculate the length of long, curly fibers of chrysotile, and diameters of nonfibrous particles were determined by the technique of Cadle (5). Eight hundred fibers were counted, and particle size distributions were comirted aft* each experiment to standardize results. Tracheal Organ Culture. The procedures used to prepare and main tain organ cultures of hamster trachea have been described previously (16). Briefly, random bred male Syrian hamsters between 6 and 10 weeks of age were sacrificed by an i.p. injection of 0.1 mi euthanasia solution (Taylor Pharmacal Co., Decatur, III.). Each trachea was isofatad aseptically, and the adherent connective tissue was removed by dissec tion. The trachea was divided longitudinally at the cartilagenous discon* 3 The abbreviations used are; HBSS, Hanks' balanced salt solution. MEM* Eagle's minimal essential medium; SEM, scanning electron microscopy 4906 CANCER RESEARCH VOL. *3 Mineral Fiber-induced Squamous Metaplasia and then dissected into 14 to 16 explants measuring approxir 1.5 x 2.5 mm. crimental Design. The study was carried out as 14 separate .intents because of procedural considerations. In each, 2 materials ithe appropriate controls were evaluated. Explants from 16 hamsters -divided randomly into groups and then transferred to 60- x 15-mm dishes (approximately 36 to 40/dish). Suspensions of each material I (1.0,4.0, and 16.0 mg/ml; 2 ml total volume) were added to the 'dish and allowed to deposit onto the mucosal surface of organ for 1 hr. The amounts of minerals used in our studies are ntially higher than those causing biological effects in monolayers :. However, the mucociliary clearance of tracheal organ cultures is Table 1 Source and net surface charge of materials us (A) and nonfi; (6) analogues Source Net surface0 charge dolite asbesutes ,, skite otile asbestos ' fibers " fibers iriterf glass ;-100 fiber > particles ulgite Union Internationale Contre Can cer Wards Natural Science Est., , Rochester, N. Y. (sample of fluor-riebeckite from El Paso county, Co.) Manville Corp., Denver, Co. (sam ples from the Jeffrey Mine in Quebec) Wards Scientific Est. (sample from Arizona) Owens Coming Fiberglas, Toledo, Ohio Manville Corp. Prepared from samples of code- 100 fiberglass Clay Mineral Society, Columbia, Mo. (sample from Nevada) 0 (-42.0 3.2) 0 (+49.4 2.1) NAe 0 (-48.6 4.1) 0 ophoretic mobility (22) of particles which were suspended in MEM. ibers in parentheses, zeta potential (22) determinations of fibers in distilled S.E. iple of riebeckite contained <1% fibers (aspect ratio >3). orite contained <3% fibers and small amounts of the minerals picrolite lite. No contamination was detected in samples of the other minerals, not applicable, fl Fiberglass rapidly settled from suspension due to the iiameter of fibers. This precluded accurate determinations of zeta potential. very efficient, and only a small fraction of the original amount remains on the surfaces of explants after 1 hr. We chose the dosages above on the basis of previous experiments using crocidolite which indicated that a 4.0-mg/ml dose was most effective in inducing metaplasia (18). For each type of dust, 12 explants were assessed at each concentration. All experiments were repeated twice. After experimental treatments, 4 explants were transferred to 30- x 15-mm plastic Petri dishes (Costar, Cambridge, Mass.) the surface of which had been scored to facilitate attachment. We added 0.5 ml of MEM containing 25 mM A/-2-hydroxyethylpiperazine-A/'-2-ethanesulfonic acid buffer (Sigma Chemical Co., St. Louis, Mo.), gentamicin (100 pg/ ml), and nystatin (25 units/ml). This volume of medium was sufficient to wet but not submerge the mucosal surface of the explant. Cultures were maintained at 35-36 in an humidified atmosphere containing 5% C02, and the medium was changed every 3 days. Explants maintain normal mucociliary differentiation for at least 4 weeks under these conditions (16). Representative explants were examined at intervals of 2, 4, and 6 weeks after exposure to particles. Assessment of Metaplasia. The extent of squamous metaplasia was evaluated by SEM. Explants were rinsed twice in HBSS to remove adherent mucin and were placed in modified Kamovsky's fixative for 12 hr. The specimens were dehydrated in ethanol, critical-point dried, sput ter-coated with gold-palladium, and examined in a JEOL JSM 350 SEM. The method used to grade the extent of the epithelium on the explants exhibiting squamous metaplasia has been described previously (32). An image of the entire mucosal surface was centered on the viewing screen of the SEM at a 0 tilt. A rectangular area (0.138sq mm, which comprised approximately 5% of the mucosal surface) in the center of each specimen then was defined after increasing the magnification 10-fold. The central region of the explant was analyzed since artifactual metaplastic changes usually develop at the cut margin (presumably due to traumatization of the tissue during preparation). The percentage of mucosal surface within this area which showed either squamous differentiation or cytotoxic changes (cell necrosis and sloughing) was measured by placing a sheet of clear plastic over the viewing screen and outlining the distribution of lesions. The areas de marcated in this way were then converted to numerical values using a graphics tablet (Apple Computer, Inc., Cupertino, Calif.) and image analysis software (Optomax, Inc., Hollis, N. H.). Squamous cells were distinguished by their polygonal configuration and large diameter (>15 Table 2 Cumulative frequency distribution of fiber length % of particles at following fiber length est material <1 Mm <5 iim <10 /im <20 Mm <30 fim <40 urn 94 100 " chrysotile 59 100 "e 21 69 87 95 98 99 (00 fiberglass 2 31 54 78 87 92 .Chrysotile 0 0 31 43 50 56 ass 0 0 1 5 10 16 es indicate cumulative percentage of particles equal to or less than a given size. <50 /jm <100 Mm 100 94 100 60 84 20 43 <200 Mm 98 77 <300 Mm 100 90 <500 Mm 100 Table 3 Cumulative frequency distribution of fiber/particle diameter % of particles at following fiber/particle diameter0 ! Test material <0.2 Mm <0.4 Mm <0.6 Mm <0.8 Mm <1.0 Mm <2.0 Mm <3.0 Mm <4.0 Mm <5.0 Mm t<` "Short" chrysotile Attapulgite 89 1 99 100 4- - - r 25 60 100 "Long" chrysotile 65 82 90 93 96 98 99 100 Antigorite 47 71 85 93 99 99 100 Crocidolite 64 83 89 94 98 100 Code-100 50 73 83 89 94 98 100 Glass powder 7 20 33 51 91 98 99 100 Riebeckite 14 27 47 57 88 95 97 100 4-fiberglass 00 0 000 3 97 100 0 Values indicate the cumulative percentage of particles equal to or less than a given size. Mm). Stratified squamous metaplasia was conspicuous because the superficial cells "heaped up" to form mounds. However, we did not attempt to distinguish simple squamous metaplasia from stratified squa mous metaplasia by SEM. Autoradiography. Explants were labeled for 5 hr with [3H]thymidine (20 MCi/ml; specific activity, 49 Ci/mmol; Amersham-Searle Corp., Arling ton Heights, III.) and then rinsed twice in HBSS at 37 before fixation. Using this protocol, approximately 0.5 to 2.0% of the tracheal epithelial cells are labeled in MEM after 2 weeks in culture (16). After examination by SEM, selected specimens were processed for light microscopy (2). Tissues were rehydrated and embedded in Paraplast (American Scientific Products, Bedford, Mass.), and sections of 5-Mm thickness were cut. Sections mounted on glass slides were dipped in Kodak NTB immulsion (Eastman Kodak Co., Rochester, N. Y.) and then exposed at 4 for 1 week. Autoradiographs were developed in Kodak D-19 and stained with Harris hematoxylin. One hundred epithelial cells (basal and suprabasal) on each side of he approximate center of the explant in 4 sections (a total of 800 epithelial cells per specimen) were counted to determine the labeling index (18). RESULTS Untreated explants exhibit normal mucociliary differentiation for 4 weeks in vitro although small foci of metaplasia occasionally develop over this period. In contrast, explants exposed to as bestos and other fibrous materials underwent both proliferative and metaplastic alterations. Analysis of variance and Fisher's 10 5 0 # c aas> E 10 least significant difference procedures were used to experimental groups (15). Crocidolite and fiberglass i significant increases (p < 0.05) in the proportion c -'-e involved by squamous metaplasia after both 2 ar(. : ~w culture (Chart 1). Metaplastic changes were most proafter 4 weeks whereas degeneration of the epithelium terized by cell sloughing and/or a spindly epithelium of thickness, often was evident after 6 weeks. Both the "lono* "short" fibers of chrysotile asbestos induced a signify crease in metaplasia at low dosages (1.0 and 4.0 mg/mh tively). Exposure to attapulgite resulted in similar effect a| the increase was not statistically significant. High concent (16.0 mg/ml) of "long" chrysotile were markedly cytotoxic determined by desquamation and necrotic alterations in < ' cells. The labeling index of epithelial cells was increased signH in cultures exposed to both crocidolite asbestos and fib for 2 weeks (Chart 2). Attapulgite and "long" chrysotile caused increases in the labeling index, although the were not statistically significant. No change was observed' exposure for 2 weeks to "short" chrysotile. After 4 wc culture, labeling indices were low, and experimental grou| )' not differ from controls. The decrease in labeling of epi ' cells after extended periods in vitro has been observed prev in this system (16). Nonfibrous analogues (riebeckite, antigorite, and glass cles) failed to induce significant increases in both the la ' index and extent of squamous metaplasia over a dosages and durations of exposure. The pattern of deposition of materials on the mucosal su was studied by SEM to determine the association of fibers metaplastic lesions. Most fibers aggregated at the ma the explant, although small numbers of individual fibers also *, distributed randomly on the mucosal surface. These fibers i rested on nonciliated cells (Fig. 1) or protruded into the mu surface where they often were encompassed by accumula" of epithelial cells (Figs. 2 to 5). Metaplastic foci were us m CO 5 ara 0) 2 0 c3o O E 15 c3o cr CO 10 5 0 0 1 4 16 0 1 4 16 0 1 4 16 Exposure Dose (mg/ml) Chart 1. Development of squamous metaplasia 4 weeks after exposure of tracheal organ cultures to fibrous and nonfibroas materials. The percentage of squamous metaplasia represents the area of the mucosa showing squamous change divided by the total area examined. *, responses which are statistically different (p < 0.05) than the corresponding nonexposed controls (); bars, S.E.; the number of observations vaned between 19 and 26. numbers within columns. percentage of mucosal surface with cytotoxic alterations (f.e., loss of epithelium). Type of Material Chart 2. Percentage of epithelial cells with labeled nuclei on explants of hamsw trachea 2 weeks after exposure to fibrous and nonfibrous matenals. The expiana were exposed to (A) nebeckite, (S) crocidolite, (C) antigorite, (D) "short" chrysotf*. (E) "long" chrysotile, (F) glass particles, (G) /S fiberglass, (H) code-100 fibergfat*. and (/) attapulgite at 4.0 mg/ml medium. *, responses which were significant^ different than the corresponding nonexposed controls (); bars, S E.; number of observations varied between 19 and 26. 4908 CANCER RESEARCH VOL. 43 Mineral Fiber-induced Squamous Metaplasia Jj (<50 Mnn in diameter), but occasionally larger expanses of Jjnucosa were affected (approximate diameter > 300 /urn; Fig. fjjythough many metaplastic lesions developed in association fibers, foci also were located at sites where deposits of ; could not be found. ; noted above, "long" fibers of chrysotile (in contrast to the rminerals) induced cell distortion and sloughing. These fibers |p formed aggregates which were associated with clusters of otic cells (Fig. 6). Many fibers became coated with material a smooth, bead-like appearance, suggesting deposition jjjiicin or proteinaceous material (Fig. 7). Cytotoxicity of nat- occurring asbestos has been described by others in a Bty of cell types (11,19). ! surface of explants treated with nonfibrous materials was lively free of particles. When present, this material accumu- I into aggregations which were scattered sparingly over the osal surface. JSSION ibestos and cigarette smoke act synergistically in the develsnt of bronchogenic carcinoma (9, 24). The stimulation of ious metaplasia by asbestos might predispose epithelial to transformation by carcinogens in cigarette smoke. The lies recorded here were designed to investigate the role of le geometry in the induction of squamous metaplasia by is materials. Fibers of differing physicochemical composijjtinduced hyperplasia and metaplasia in cultured tracheal its. Their nonfibrous analogues do not. rplastic and metaplastic lesions are found in the airways its after the inhalation of asbestos (4, 31). Therefore, our ations using cultured tissues are consistent with findings fiimals. ie importance of fiber morphology in the induction of bronlenic carcinoma is unclear, but fibrous shape is critical in the imental production of both mesotheliomas (26) and pulfibrosis (33). Intrapleural innoculation of materials comof long, thin fibers results in mesotheliomas, whereas blocky fibers are less tumorigenic (26). In the studies of ton et at. (26), maximum carcinogenic activity was observed ll fibers <0.25 /tm in diameter and >8 txm in length; however, i; documented the metaplastic capability of /3 fiberglass, a ' ation containing fibers >3 ^m in diameter. Thus, fiber in comparison to diameter seems more important in ig squamous metaplasia. Phagocytosis and subsequent ation of short fibers might explain biological inactivity (13, |$3), but there has been no satisfactory explanation for the ogenicity of long fibers in the pleural and peritoneal cavity, er or not long fibers of types other than asbestos are ogenic in tracheobronchial epithelium is unproven in anibecause of the difficulty in generating sized preparations of for inhalation studies. r morphological observations suggest a hypothetical model [(Which long fibers stimulate squamous metaplasia (Chart 3). (have found that fibers occur between cells of the mucociliary ilium where they appear to interact with the underlying cells (19, 32). This process is accompanied by necrosis of cial epithelial cells (19). Although no definitive evidence to explain how fibers stimulate squamous metaplasia, the ss probably involves more than compensatory changes Chart 3. Hypothetical mechanism by which long fibers stimulate squamous metaplasia. Fibers occurred between cells of the mucociliary epithelium and ap peared to interact with the underlying basal cells. This process causes injury and compensatory hyperplasia of adjacent basal cells (19). Alternatively, the long fibers which interact with the mucosal surface might serve as an artificial substrate which stimulates the outward migration of epithelial cells. Subsequently, these cells divide and differentiate into squamous cells. resulting from cell death. For example, "long" chrysotile was the most cytotoxic material, but both crocidolite asbestos and fiber glass induced more extensive squamous metaplasia. The nature of the substrate upon which cells rest influences their type of differentiation (8). Long fibers of glass provide a unique substrate for attachment of cultured fibroblasts and also act as a stimulus to promote cell division (13). Epithelial cells which interact with fibers might respond similarly. The failure of nonfibrous materials to stimulate metaplasia could reflect their rapid elimination from the mucosa by either phagocytosis or mucociliary clearance. Squamous metaplasia in the human respiratory tract is topo graphically associated with squamous cell carcinoma (29). It might predispose the respiratory epithelial cells to malignant transformation by carcinogens in cigarette smoke. The rapid proliferation of cells which accompanies metaplasia (12) can enhance DNA damage by chemical carcinogens (14). Alterna tively, chronic hyperplasia might promote neoplastic progression by previously initiated cells. Squamous differentiation also results in loss of protective mucociliary function, possibly leading to both the accumulation and prolonged interaction of carcinogens with epithelial cells (32). Cigarette smoking also causes squamous metaplasia (1) and modifies the deposition and clearance of inhaled particles (23). Hence, both substances could interact to increase retention of chemical carcinogens in the respiratory tract (6). Because occupational exposure to asbestos is a recognized health hazard, a variety of nonasbestiform fibers recently have been promoted as substitutes (3, 21). These include both natu rally occurring fibrous minerals (/.e., attapulgite) and a number of man-made mineral fibers (i.e., fiberglass). Although these mate rials are used widely, their biological effects are incompletely understood (7, 10, 20, 26, 30). Our results suggest that the ability of inorganic particulates to induce squamous metaplasia depends upon their fibrous geometry and not on any particular chemical or structural property unique to asbestos. Thus, other fibrous materials have the potential to induce hyperplasia and OBER 1983 4909 o. u. wooawonn et ai. metaplastic alterations in respiratory epithelium. Our results should be useful in predicting the potential of specific inorganic particulates to induce proliferative and metaplastic changes in epithelial cells. ACKNOWLEDGMENTS The authors would like to thank J. Carrassi, B. Clements, and P. Gale for their excellent technical assistance. G. Badger helped with the statistical analyses. REFERENCES 1. Auerbach, 0., Stout, A. P., Hammond, E. C,, and Garfinkel, L. Changes in bronchial epithelium in relation to cigarette smoking and in relation to lung cancer. N. Engl. J. Med., 265: 253-267,1961. 2. Ayres, A., Allen, J. M., and Williams, A. E. A method of obtaining conventional histological sections from specimens after examination by scanning electron microscopy. J. Microsc. (Oxf.), 93: 247-250,1971. 3. Background information on substitutes for asbestos, Environmental Protection Agency, Office of Toxic Substances. Washington, D. C.: Government Printing Office, 1980. 4. Botham, S. K,, and Holt, P. F. The effects of inhaled crocidolites from Transvaal and northwest cape mines on the lungs of rats and guinea pigs. Br. J. Exp. Pathol., 53:612-620, 1972. 5. Cadle, R. D. Particle Size: Theory and Industrial Applications, pp. 1-38. New York: Reinhold, 1965. 6. Craighead, J. E., and Mossman, B. T. The pathogenesis of asbestos-associ ated diseases. N. Engl. J. Med., 306:1446-1455,1982. 7. Davis, J. M. G. The biological effects of mineral fibers. Ann. Occup. Hyg., 24: 227-234,1981. 8. Gospodarowicz, D. The extracellular matrix and the control of cell proliferation. In: J. B. Martin, S. Reichlin, and K. L. Bick (ed.), Neurosecretion and Brain Peptides, pp. 243-261. New York: Raven Press, 1981. 9. Hammond, E. C., Selikoff, I. J., and Seidman, H. Asbestos exposure, cigarette smoking, and death rates. Ann. N. Y. Acad. Sci., 330: 473-490,1979. 10. Harington, J. S. Fiber carcinogenesis: epidemiologic observations and the Stanton hypothesis. J. Natl. Cancer Inst., 67: 977-988,1981. 11. Harington, J. S., Allison, A. C., and Badami, B. V. Mineral fibers: chemical, physicochemical, and biological properties. Adv. Pharmacol. Chemother., 72: 291-402,1975. 12. Harris. C. C,, Silverman, T., Smith, J. M., Jackson, F., and Boren, H. G. Proliferation of tracheal epithelial cells in normal and vitamin A deficient Syrian golden hamsters. J. Natl. Cancer Inst., 51:1059-1062,1973. 13. Maroudas, N. G., O'Neill, C. H., and Stanton, M. F. Fibroblast anchorage in carcinogenesis by fibers. Lancet, 1: 807-809,1973. 14. Marquardt, H. Cell cycle dependence of chemically induced malignant trans formation in vitro. Cancer Res., 34:1612-1615,1974. 15. Montgomery, D. Design and Analysis of Experiments. New York: John Wiley & Sons, 1976. 16. Mossman, B. T., and Craighead, J. E. Long-term maintenance of d :> sntiatart respiratory epithelium in organ culture. I. Medium composition. Proc oc Biol. Med., 149: 227-233,1975. "' 17. Mossman, B. T., and Craighead, J. E. Mechanisms of asbestos carcinoaenesi. Environ. Res., 25:269-280,1981. "** 18. Mossman, B. T,, Craighead, J. E., and MacPherson, B. V. Asbestos-induced epithelial changes in organ cultures of hamster trachea: inhibition of retta methyl ether. Science (Wash. D. C.), 207:311 -313,1980. v 19. Mossman, B. T,, Kessler, J. B,, Ley, B. W., and Craighead, J. E. Interaction of crocidolite asbestos with hamster respiratory mucosa in organ culture b* Invest., 36:131-139,1977. '^ 20. Occupational exposure to fibrous glass-proceedings of a symposium. UHm States Department of Health, Education, and Welfare, Washington, D. c- Government Printing Office, 1976. ''" 21. Pye, A. M. Alternatives to asbestos in industrial applications. In: L. Michaah and S. S. Chissick (eds.), Asbestos: Properties, Applications, and Haza*. Vol. 1, pp. 339-373. New York: John Wiley & Sons, 1979. ^ 22. Riddick, T. M. Zeta Meter Manual, pp. 43-73. New York: Zeta Meter Ire. 1961. '^ 23. Sanchis, J., Dolovich, M., Chalmers, R,, and Newhouse, M. T. Regiortf deposition and lung clearance mechanisms in smokers and nonsmokers. In W. H. Walton (ed.), Inhaled Particles III, Vol. 1, pp. 183-191. Surrey. England; Unwin Brothers Limited, 1971. ' 24. Saracci, R. Asbestos and lung cancer: an analysis of the epidemidoglat' evidence on the asbestos-smoking interaction. Int. J. Cancer, 20: 323-331 1977. V. 25. Spumy, K. R., Opiela, H., and Weiss, G. On the milling and ultrasonic treatmrtU of fibers for biological and analytical applications. In: J. C. Wagner (ed.), WHO)4;1? Scientific Publication No. 30, Vol. 2, pp. 931-933. Lyon, France: Intematkwl ` ' Agency for Research on Cancer, 1980. . 26. Stanton, M. F., Layard, M., Tegeris, A., Miller, E., May, M., Morgan, E,, d Smith, A. Relation of particle dimension to carcinogenicity in amphibote asbtt*'' loses and other fibrous minerals. J. Natl. Cancer Inst., 67:965-975,1981. > 27. Tanner, C. B., and Jackson, M. L. Nomographs of sedimentation times for *o|/Vv particles under gravity or centrifugal acceleration. Soil Sci. Soc. Am. Proc., It 60-65,1947. 28. Topping, D. C., and Nettesheim, P. PromotionJike enhancement of tracfl*#i(i'l{ carcinogenesis in rats by 12-O-tetradecanoylphorbol-l 3-acetate. Cancer 40:4352-4355,1980. 29. Valentine, E. H. Squamous metaplasia of the bronchus. A study of metapltlf$||>f changes occurring in the epithelium of the major bronchi in cancerous st! noncancerous cases. Cancer (Phila.), 10:272-279,1957. . 30. Wagner, J. C,, Berry, G,, and Pooley, F. D. Carcinogenesis and mineral Br. Med. Bull., 36: 53-56,1980. ;',f 31. Wagner, J. C,, Berry, G., Skidmore, J. W., and Timbrell, V. The effects oft<i| inhalation of asbestos in rats. Br. J. Cancer, 29:252-269,1974 '(fy 32. Woodworth, C. D., Mossman, B. T., and Craighead, J. E. Squamous metapMiVc1 of the respiratory tract: possible pathogenic role in asbestos-associated twin* chogenic carcinoma. Lab. Invest., 48:578-584,1983. 33. Wright, G. W,, and Kuschner, M. The influences of varying lengths of ghM :. ; and asbestos fibers on tissue response in guinea pigs. In: W. H. Walton (*dL`:ii|' Inhaled Particles IV, Vol. 2, pp. 455-474. New York: Pergamon Press, 1977,wy a. 4910 CANCER RESEARCH VOL- Cancer Research VOLUME 43* NO. 10 CNREA 8 PP 4525-5048 October 1983