Comparative Pulmonary Absorption, Distribution, and Toxicity of Copper Gallium Diselenide, Copper Indium Diselenide, and Cadmium Telluride in Sprague–Dawley Rats

TOXICOLOGY AND APPLIED PHARMACOLOGY ARTICLE NO.

147, 399 – 410 (1997)

TO978267

Comparative Pulmonary Absorption, Distribution, and Toxicity of Copper Gallium Diselenide, Copper Indium Diselenide, and Cadmium Telluride in Sprague–Dawley Rats Daniel L. Morgan,* Cassandra J. Shines,* Shawn P. Jeter,* Mark E. Blazka,* Michael R. Elwell,* Ralph E. Wilson,* Sandra M. Ward,* Herman C. Price,† and Paul D. Moskowitz‡ *National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709; †ManTech Environmental Technology, Inc., Research Triangle Park, North Carolina; and ‡Brookhaven National Laboratory, Upton, New York Received March 7, 1997; accepted July 18, 1997

Comparative Pulmonary Absorption, Distribution and Toxicity of Copper Gallium Diselenide, Copper Indium Diselenide, and Cadmium Telluride in Sprague-Dawley Rats. Morgan, D. L., Shines, C. J., Jeter, S. P., Blazka, M. E., Elwell, M. R., Wilson, R. E., Ward, S. M., Price, H. C., and Moskowitz, P. D. (1997). Toxicol. Appl. Pharmacol. 147, 399 – 410. Copper gallium diselenide (CGS), copper indium diselenide (CIS), and cadmium telluride (CdTe) are novel compounds used in the photovoltaic and semiconductor industries. This study was conducted to characterize the relative toxicities of these compounds and to evaluate the pulmonary absorption and distribution after intratracheal instillation. Female Sprague–Dawley rats were administered a single equimolar dose (70 mM) of CGS (21 mg/kg), CIS (24 mg/kg), CdTe (17 mg/kg), or saline by intratracheal instillation. Bronchoalveolar lavage fluid (BALF) protein, fibronectin, inflammatory cells, lung hydroxyproline, and tissue distribution were measured 1, 3, 7, 14, and 28 days after instillation. Relative lung weights were significantly increased in CISand CdTe-treated rats at most time points. Inflammatory lesions in the lungs consisting of an influx of macrophages, lymphocytes, and PMNs were most severe in CdTe-treated rats, intermediate in CIS-treated rats, and minimal in rats receiving CGS. Hyperplasia of alveolar type 2 cells was present in CIS- and CdTe-treated rats and was greatest in CdTe-treated rats. Pulmonary interstitial fibrosis was observed in CdTe-treated rats at all time points. All three compounds caused marked increases in total BALF cell numbers, with the greatest increase observed in CIS-treated rats. BALF protein, fibronectin, and lung hydroxyproline were significantly increased in all treated animals and were highest in CdTetreated animals. There was no apparent pulmonary absorption or tissue distribution of CGS. Indium levels increased in extrapulmonary tissues of CIS-treated rats, although Cu and Se levels remained unchanged. CdTe was absorbed from the lung to a greater extent than CGS and CIS. Cd and Te levels decreased in the lung and increased in extrapulmonary tissues. Of these compounds CdTe presents the greatest potential health risk because it causes severe pulmonary inflammation and fibrosis and because it is readily absorbed from the lung may potentially cause extrapulmonary toxicity.

In the photovoltaic and semiconductor industries a number of unique materials are used for which even the most basic toxicological information is lacking. Consequently, the health hazard to individuals exposed to these materials is not known. Several novel compounds of Group IIIA and IVA elements, including copper gallium diselenide (CGS), copper indium diselenide (CIS), and cadmium telluride (CdTe), have demonstrated significant advances in semiconductor efficiency and will likely see increased use in the future (Service, 1996). The primary health risk from CGS, CIS, and CdTe is to workers in the photovoltaic and semiconductor industries. The risk to the public is relatively small; however, accidental releases from production facilities could occur and, as with workers, the primary route of exposure in these instances would be by inhalation. The pulmonary toxicity of CGS, CIS, and CdTe was previously shown to be related to compound solubility (Morgan et al., 1995). Although CGS, CIS, and CdTe are considered insoluble in water, in vitro incubation in aqueous media was shown to release the free elements (Morgan et al., 1995). The solubility of these compounds may be facilitated after deposition in the lung. Dissolution of relatively insoluble metal compounds can occur in lung lining fluids (Hadley et al., 1980) or after phagocytosis by alveolar macrophages (AMs) (Lundborg et al., 1984). The free elements may be absorbed from the lung and distributed to extrapulmonary tissues where accumulation of these elements could result in toxicity. Cadmium (Driscoll et al., 1992), tellurium (Geary et al., 1978), and indium (Blazka et al., 1994) all cause pulmonary toxicity when administered by intratracheal instillation or by inhalation. Systemic administration of cadmium results in accumulation and toxicity in the kidneys (Lauwerys et al., 1984); tellurium causes neurotoxicity after systemic administration (Duckett, 1982). Acute selenium toxicity results in neurotoxicity (Patty, 1963), and ingestion of high levels of copper compounds can cause hepatotoxicity (Chuttani et al., 1965). In a previous study (Morgan et al., 1995) the acute pulmonary toxicity of CGS, CIS, and CdTe was evaluated 4 days

399

0041-008X/97

400

MORGAN ET AL.

after intratracheal instillation in rats. All three compounds caused a marked inflammatory response in the lung characterized by a massive influx of PMNs. The objective of the current study was to compare the pulmonary absorption, distribution, and potential systemic toxicity of lower doses of these compounds and to evaluate the progression of pulmonary lesions over a longer time period. The pulmonary absorption and distribution of these compounds were measured to determine extrapulmonary tissue doses. Potential extrapulmonary toxicity was evaluated by histopathology and clinical chemistry indices. Because of the previously observed acute pulmonary inflammation caused by these compounds, pulmonary inflammation as well as biochemical and histological indices of fibrosis were evaluated at various times after treatment. METHODS Chemicals. Copper gallium diselenide (CuGaSe2), copper indium diselenide (CuISe2), and cadmium telluride were provided by Cerac, Inc. (Milwaukee, WI). All compounds were at least 99.9% pure, and mean particle sizes were CGS, 2.47 mm; CIS, 2.55 mm; and CdTe, 1.70 mm. All three compounds were used in particulate form. Equimolar concentrations (70 mM) of each chemical were prepared just prior to dosing by suspending the chemical in sterile phosphate-buffered saline, pH 7.2, at concentrations of 21 mg/ml CGS, 24 mg/ml CIS, and 17.1 mg/ml CdTe. In a previous dose–response study (Morgan et al., 1995) these doses were shown to cause no histopathological effects and only a slight influx of PMNs into the lungs. The chemical suspensions were thoroughly mixed by vortexing immediately prior to each instillation. Animals. Specific pathogen-free female Sprague–Dawley rats (Charles River Breeding Laboratories, Raleigh, NC) weighing 180 –200 g (42– 48 days old) were acclimated for 10 to 14 days after arrival in the animal facility with a 12-hr light– dark light cycle. During acclimation, rats were randomized by weight into groups, five rats/group. Animals were provided NIH-07 diet and tap water ad libitum throughout the study. Intratracheal instillation. Intratracheal instillations were performed on lightly anesthetized (Metofane) animals using a 1-ml syringe and a 20-gauge needle connected to 2–3 cm Teflon tubing placed about 1 cm into the trachea. Each rat received a single instillation at a dosing volume of 0.1 ml/100 g body wt and equimolar concentrations of CGS, CIS, or CdTe (corresponds to 21, 24, and 17.1 mg/kg body wt, respectively). Control rats received 0.1 ml sterile PBS/100 gm body wt. Rats were weighed just prior to treatment and weekly thereafter. All animals were observed for overt signs of toxicity twice daily throughout the study. Clinical pathology. At 7, 14, and 28 days after dosing, five treated and five control rats were weighed and anesthetized with a 70:30 mixture of CO2:O2. Blood (about 1 ml) was collected via cardiac puncture into serum separator vials, and serum samples were stored at 270°C until analyzed at the end of the study. Serum was analyzed for albumin, total protein, creatinine, urea nitrogen, alanine aminotransferase, sorbitol dehydrogenase, 59-nucleotidase, total bile acids, and creatine kinase using an automated analyzer (Monarch System 2000, Instrumentation Laboratory, Lexington, MA) and commercially available reagents. One week after dosing, blood (about 1 ml) was collected from an additional five treated and five control rats for hematology. Complete blood counts and cell differentials were performed using a Technicon H1 (Bayer Corp., Tarrytown, NY), and reticulocyte counts were performed manually. Histopathology. After collecting blood for clinical pathology, rats were euthanized by CO2 inhalation and exsanguination. Lung, liver, right kidney, and spleen weights were recorded and then tissues were trimmed and fixed in 10% buffered neutral formalin. Lungs were infused with formalin to the

normal inspiration volume. Paraffin-embedded sections were stained with hematoxylin and eosin for light microscopic examinations. Additional lung sections were stained with Masson’s Trichrome for demonstration of collagen formation (fibrosis). Bronchoalveolar lavage. At 1, 3, 7, 14, and 28 days after treatment, five treated and five control rats were euthanized by Metofane anesthesia and exsanguination. The additional early time points (days 1 and 3) were included in order to collect data for the early inflammatory response observed previously (Morgan et al., 1995). The lungs were lavaged in situ three times with 6 ml cold, Ca21/Mg21-free Hanks’ balanced salt solution (HBSS) after cannulating the trachea with a 16-gauge needle. Bronchoalveolar lavage fluid (BALF) cellularity and differential. The BAL fluids were centrifuged (10 min, 2000 rpm, 4°C), and the cell pellets were combined for total and differential cell counts. The cell pellets were suspended in 5 ml of sterile HBSS and the total number of cells was determined electronically (Coulter ZB, Coulter Electronics Inc., Marietta, GA). Aliquots of the cell suspensions were used to prepare slides for differential counts using a cytocentrifuge. The cytospin preparations were fixed in methanol and stained with Diff-Quik (Baxter, Miami, FL), and 300 cells were counted in each preparation. BALF protein and fibronectin. The cell-free supernatant of the first lavage fraction was analyzed for total protein and fibronectin. Total BALF protein is a sensitive indicator of damage to the alveolar epithelial barrier (Henderson, 1989). Total protein was measured using an automated analyzer (Monarch System 2000, Instrumentation Laboratory) and Bio-Rad reagent using the manufacturer’s instructions. To better characterize the fibrotic response fibronectin in the BALF was quantified by an indirect ELISA using the method of Driscoll et al. (1990a,b). Although 4-hydroxyproline is a good indicator of collagen content, it does not distinguish between extracellular collagen deposition resulting from increased synthesis and intracellular collagen due to increased numbers of fibroblasts (Witschi et al., 1985). Previous reports have demonstrated that increased fibronectin levels are indicative of pulmonary fibrosis (Driscoll et al., 1990a,b). 4-Hydroxyproline analysis. Collagen deposition was estimated by determining the total 4-hydroxyproline content of the lung. After bronchoalveolar lavage, the lungs were excised, weighed, homogenized, and hydrolyzed in 6 N HCl overnight at 110°C. Hydroxyproline content was assessed colorimetrically at 560 nm with p-dimethylaminobenzaldehyde as described by Woessner (1985). Data are expressed as grams of 4-hydroxyproline per lung. Chemical absorption and distribution. Parallel studies were conducted to quantify the pulmonary absorption and extrapulmonary tissue distribution of the elemental components of CGS, CIS, and CdTe. Groups of five rats/ compound/time point were euthanized by CO2 asphyxiation 1, 3, 7, 14, and 28 days after chemical instillation. From each animal, blood samples were collected by cardiac puncture, and then lung, liver, kidney, spleen, and femur were collected and stored at 270°C until metal analyses. All tissues were thawed, weighed, homogenized, and microwave digested overnight in closed vessels containing concentrated nitric acid. If needed, 30% hydrogen peroxide was added to complete the digestion. After tissue digestion, lung, liver, kidney, spleen, and blood samples were analyzed for copper by inductively coupled plasma optical emission spectrometry and for indium or cadmium by inductively coupled plasma and mass spectrometry. Femur samples contained very low levels of selenium and were analyzed by hydride generation atomic fluorescence spectrometry. Graphite furnace atomic absorption spectrometry was used to measure tissue levels of gallium and tellurium and selenium in tissues other than femur. Statistics. Analysis of variance procedures were used to assess the significance of differences among days and treatment effects (Snedecor and Cochran, 1980). Pairwise comparisons were made by Dunnett’s test or by Fisher’s least significant difference test (Miller, 1966). In some instances the variancestabilizing logarithmic transformation was used prior to data analysis.

401

COMPARATIVE PULMONARY TOXICITY

TABLE 1 Body and Selected Organ Weights in CdTe-Treated Rats Days after CdTe instillation

Body weight gain (g) Spleen weight (mg/g) Kidney weight (mg/g)

Dose (mg/kg)

7

0 17 0 17 0 17

23.2 6 3.7 (5) 23.2 6 7.5 (4)* 2.58 6 0.13 (5) 2.93 6 0.22 (4) 4.25 6 0.13 (4) 4.63 6 0.11 (2)

14 39.3 6 1.7 (5) 24.8 6 2.3 (5)* 2.32 6 0.13 (5) 4.08 6 0.28 (5)* 4.36 6 0.16 (5) 4.42 6 0.06 (5)

28 86.5 6 8.5 (5) 57.2 6 6.5 (5)* 2.44 6 0.13 (5) 4.48 6 0.37 (5)* 3.88 6 0.08 (5) 4.68 6 0.14 (5)*

Note. Spleen and kidney weights are expressed as mg organ weight/g body weight. Values represent mean weights 6 SE (n). * Significantly different from respective control (p , 0.05).

RESULTS

measured 7, 14, or 28 days after treatment with CGS, CIS, or CdTe (data not shown).

Body and Organ Weights Treatment with CGS or CIS had no significant effect on body weights (data not shown); however, CdTe caused significant reductions in body weight gain when measured at 7, 14, and 28 days after instillation (Table 1). Relative liver, spleen, and kidney weights were not significantly different from controls in CGS- and CIS-treated rats (data not shown). CdTe caused significant increases in relative spleen weights when examined 14 and 28 days after instillation (Table 1). Relative kidney weights were significantly increased only at 28 days after treatment. However, because the kidney weights of controls were lower than previous values and the percentage of change was very small, the biological significance of this slight effect is questionable. Relative lung weights were significantly increased in CISand CdTe-treated rats when examined at 7, 14, and 28 days after treatment (Table 2). In contrast, CGS had no effect on relative lung weights at any of the time points evaluated. Clinical Pathology No statistically significant treatment-related effects were observed in serum chemistry or hematology parameters when

Histopathology Light microscopic evaluation of liver, kidney, and spleen from CGS-, CIS-, and CdTe-treated and control rats revealed no significant lesions in liver or kidney. In spleen, there was a mild hyperplasia of the periarteriolar lymphoid sheath in the spleen of CdTe-treated rats at days 7 (four of five), 14 (five of five), and 28 (five of five treated rats). The following changes were observed in the lung following CGS, CIS, and CdTe administration. CGS. By 7 days after instillation, minimal to mild foci of inflammation were present in lungs of CGS-treated rats. No change in pulmonary inflammation was observed between days 7 and 28. By day 28 there was a minimal infiltrate of macrophages in the interstitium and alveolar spaces around the terminal bronchioles (Fig. 1A). Fibrosis was not a component of the CGS-induced pulmonary inflammation. Black CGS particles or aggregates of particles were present in the interstitium, in the cytoplasm of alveolar macrophages, and in the bronchial lymph nodes. CIS. By day 7 an inflammatory response was observed in the lungs and consisted of an infiltration of macrophages,

TABLE 2 Relative Lung Weights in Rats Treated with CGS, CIS, or CdTe Days after CGS, CIS, or CdTe instillation Chemical CGS CIS CdTe

Dose (mg/kg)

7

0 21 0 24 0 17

6.79 6 0.22 (5) 6.75 6 0.38 (5) 6.69 6 0.50 (5) 11.56 6 1.11 (5)* 9.96 6 1.48 (5) 13.87 6 2.21 (4)

14 6.72 6 0.47 (5) 7.71 6 0.67 (5) 6.97 6 0.30 (5) 12.76 6 0.90 (5)* 5.82 6 0.41 (5) 15.40 6 1.15 (5)*

Note. Relative lung weights are expressed as mg lung weight/g body weight. Values represent means 6 SE (n). * Significantly different from respective control (p , 0.05).

28 6.99 6 0.67 (5) 7.69 6 0.53 (5) 6.77 6 0.58 (5) 14.59 6 0.53 (5)* 6.78 6 0.49 (5) 11.18 6 0.95 (5)*

402

MORGAN ET AL.

FIG. 1. Lung sections from rats administered CGS, CIS, or CdTe by intratracheal instillation 28 days prior to necropsy. Adjacent to the pulmonary acinus (*) of a rat administered CGS (A) there are focal aggregates of black CGS particles (arrows) in the interstitium as well as in some macrophages. (B) Centered around the pulmonary acinus (*) of a rat administered CIS is an accumulation of a granular protein exudate, macrophages, and black CIS particles in the alveolar spaces. Many alveolar are lined by a cuboidal type 2 epithelium (arrows). With CdTe (C), there is moderate fibrosis (arrows) surrounding the pulmonary acinus (*) and an accumulation of macrophages, lymphocytes, and plasma cells in the interstitium of the alveolar septae. (D) CdTe also caused dense fibrotic inflammatory nodules (arrows) with a central area containing black CdTe particles and necrotic cell debris. Short bands of mature fibrous tissue (arrowheads) are adjacent to the fibrotic inflammatory nodule. Bar, 50 mm.

lymphocytes, and a few neutrophils in the interstitium and alveolar spaces around the terminal bronchioles. CIS particles and aggregates of particles were present in alveolar spaces and the interstitium. An eosinophilic granular exudate that was periodic acid schiff positive and morphologically consistent with a lipoproteinosis was present in the alveolar spaces. Hyperplasia of type 2 alveolar epithelium occurred in the areas with the inflammatory exudate. By 14 and 28 days after treatment, the eosinophilic exudate and interstitial inflammation were increased in severity and were most prominent at day 28 (Fig. 1B). CIS particles were present in the bronchial lymph nodes of all treated rats at these time points. CdTe. On day 7 large, confluent foci of inflammation centered around terminal bronchioles were observed throughout the lungs. Some of the larger foci extended to the pleural surface. Inflammation consisted of an infiltration of macro-

phages, lymphocytes, and neutrophils in the alveolar and bronchiolar septae and lumen. An eosinophilic granular exudate and fibrin and necrotic cell debris were present in the lumen of some distal airways. Hyperplasia of alveolar type 2 cells was extensive in and adjacent to the foci of inflammation. CdTe particles or aggregates of particles up to 100 mm in greatest dimension were scattered throughout the foci of inflammation. Trichrome staining revealed minimal interstitial fibrosis in the alveolar septae. By day 14 there was some resolution of inflammatory cell infiltrate and cell debris in airways. CdTe particles were sometimes present within nodules composed of fibrous connective tissue and macrophages. Alveolar type 2 cell hyperplasia was still prominent, and fibrosis (trichrome staining) within the alveolar septae was increased compared to day 7. By day 28 there was further resolution of the inflammatory

COMPARATIVE PULMONARY TOXICITY

403

FIG. 1—Continued

exudate with distinct foci of inflammation and fibrosis separated by areas of relatively normal appearing alveolar tissue (Fig. 1C). Type 2 cell hyperplasia persisted in the residual areas of inflammation and distinct inflammatory nodules consisting of focal fibrosis with a central area of granulomatous inflammation, necrosis, and CdTe particles were often present near terminal bronchioles (Fig. 1D). Focal aggregates of lymphocytes with plasma cell differentiation were a prominent component of the pulmonary inflammation. The lumen of terminal bronchioles or alveolar ducts were often compressed or distorted by the presence of focal inflammation and fibrosis. Lymphoid hyperplasia and CdTe particles were present in the bronchial/mediastinal lymph nodes in treated rats at days 7, 14, and 28. Cellularity and Differential of BALF CGS. Relative to controls, the numbers of PMNs and AMs in the BALF of CGS-treated rats were significantly increased at all time points (Fig. 2). At days 1, 3, and 7, the total cell numbers were about 5-fold higher than controls. The increase was attributed primarily to a significant infiltration of PMNs.

By day 28 numbers of PMNs and AMs decreased to 2.8- and 1.5-fold higher than controls, respectively. CIS. The numbers of PMNs and AMs were significantly increased in the BALF of CIS-treated rats at all time points (Fig. 3). The increased cell numbers were due primarily to a massive influx of PMNs, although numbers of AMs were also significantly increased at all time points except day 1. The influx of PMNs was considerably greater than that observed in CGS- or CdTe-treated rats. Numbers of PMNs were about 30-fold greater than controls on day 7 and were still more than 10-fold greater than controls on day 28. CdTe. In CdTe-treated rats, the increased number of BALF cells was slightly less than that observed in CGS-treated rats (Fig. 4). A maximum increase in PMNs of about five-fold was observed on day 7. The numbers of AMs were significantly increased only on days 3 and 7 and were only about two-fold greater than controls. BALF Protein The amount of protein in BALF was increased in rats treated with CGS, CIS, and CdTe (Fig. 5). CGS and CIS

404

MORGAN ET AL.

FIG. 2. Cellularity and differential of BAL fluid following a single intratracheal instillation of CGS. Values represent means 6 SE of five animals per group. *Significantly greater than controls (p , 0.05).

caused about a 4- to 5-fold increase in BALF protein when examined on days 1, 3, 7, and 14. In both CGS- and CIS-treated rats BALF protein levels continued to increase

and were 7- to 8-fold greater than controls by day 28. Initial increases in BALF protein were considerably greater in CdTe-treated rats. Protein levels were about 24-fold greater

FIG. 3. Cellularity and differential of BAL fluid following a single intratracheal instillation of CIS. Values represent means 6 SE of five animals per group. *Significantly greater than controls (p , 0.05).

COMPARATIVE PULMONARY TOXICITY

405

FIG. 4. Cellularity and differential of BAL fluid following a single intratracheal instillation of CdTe. Values represent means 6 SE of five animals per group. *Significantly greater than controls (p , 0.05).

than controls when evaluated on days 1 and 3. However, protein levels rapidly decreased and were about 4-fold greater than controls on day 7 and 2- to 3-fold greater than controls on days 14 and 28.

BALF Fibronectin CGS, CIS, and CdTe treatments resulted in increased the levels of fibronectin in the BALF of treated rats (Fig. 6). BALF

FIG. 5. Total protein in BAL fluid following a single intratracheal instillation of CGS, CIS, or CdTe. Values represent means 6 SE of five animals per group. *Significantly greater than controls (p , 0.05).

406

MORGAN ET AL.

FIG. 6. Fibronectin levels in BAL fluid following a single intratracheal instillation of CGS, CIS, or CdTe. Values represent means of five animals per group.

fibronectin levels in CGS-treated rats were increased 10- to 14-fold relative to controls throughout the study. In CIS-treated rats BALF fibronectin levels were increased by about 7-fold on day 1, 40-fold on days 3 and 7, and about 70-fold at 14 and 28-days after treatment. CdTe caused an initial increase in fibronectin of about 400-fold greater than controls on day 3, followed by a rapid decrease in levels to about 10-fold greater than controls on day 28.

4). Increased indium levels were most apparent in kidney and at 14 and 28 days after dosing in all tissues. Cadmium and tellurium levels decreased significantly in the lungs after dosing, and concomitant increases in cadmium and tellurium levels were detected in all extrapulmonary tissues evaluated (Table 5).

DISCUSSION

4-Hydroxyproline Lung hydroxyproline levels following CGS, CIS, and CdTe administration are shown in Fig. 7. Statistically significant increases in 4-hydroxyproline were observed in lungs of CGStreated rats 14 and 28 days after treatment. CIS caused a significant increase in lung 4-hydroxyproline at all time points, with a maximum of about a two-fold increase on day 28. CdTe caused significant increases in lung 4-hydroxyproline levels on days 7, 14, and 28. Of the three compounds, CdTe caused the greatest increase in 4-hydroxyproline with a maximum of about a five-fold increase detected on day 28. Chemical Absorption and Distribution There was no appreciable pulmonary clearance or extrapulmonary accumulation of copper, gallium, or selenium in CGStreated rats in the 28 days after dosing (Table 3). In CIS-treated animals, changes in lung levels of copper, indium, or selenium could not be detected; however, the amounts of indium in extrapulmonary tissues increased with time after dosing (Table

Short-term studies of CGS, CIS, and CdTe were conducted because of a lack of basic toxicity data for these relatively new compounds. The pulmonary absorption, distribution, systemic toxicity, and progression of pulmonary lesions of these three compounds were evaluated and compared after intratracheal instillation of equimolar doses. The doses of CGS, CIS, and CdTe used in this study are considerably higher than the inhalation doses to which workers could be exposed and would be more representative of an accidental exposure situation. These doses caused a minimal inflammatory response, with minimal or no effect on lung histopathology in an initial dose–response study (Morgan et al., 1995). For these initial chemical characterization studies, CGS, CIS, and CdTe were administered by intratracheal instillation. A major drawback to intratracheal instillation is that it is not possible to obtain an even distribution of particles in the lungs as occurs during inhalation of an aerosol. Uneven administration of a large dose of particulate suspension can result in a ‘‘bolus’’ effect in the larger airways, resulting in localized

407

COMPARATIVE PULMONARY TOXICITY

FIG. 7. Lung hydroxyproline levels following a single intratracheal instillation of CGS, CIS, or CdTe. Values represent means 6 SE of five animals per group. *Significantly greater than controls (p , 0.05).

lesions due to high concentrations of particles. There was some histological evidence that a uniform distribution throughout the lung was not obtained. Portions of some lobes were not affected or the size and number of lesions in lobes was somewhat variable. However, there were no aggregates of particles or

inflammatory infiltrates within the larger airways. Inflammation and particles were observed primarily in the alveoli around the terminal bronchioles. Some aggregates of CGS, CIS, and CdTe particles were also observed in these areas at the same apparent incidence for all three compounds.

TABLE 3 Tissue Distribution of Cu, Ga, and Se after Intratracheal Instillation of Copper Gallium Diselenide Days after intratracheal instillation Tissue Lung

Spleen

Kidney

Femur

Liver

Blood

Element

Control

1

3

7

14

28

Cu Ga Se Cu Ga Se Cu Ga Se Cu Ga Se Cu Ga Se Cu Ga Se

2.20 6 0.12 0.09 6 0.01 0.28 6 0.02 1.23 (1) 5.38 (1) ,MDL 17.20 (1) ,MDL 0.329 (1) 1.08 6 0.10 0.01 (1) ,MDL 4.80 6 0.18 ,MDL 1.18 6 0.03 0.73 (1) ,MDL 0.22 (1)

266 6 126 303 6 143 624 6 285 1.30 6 0.10 0.12 6 0.06 ,MDL 12.0 6 1.27 ,MDL 0.20 6 0.01 1.09 6 0.06 0.19 6 0.08 ,MDL 4.74 6 0.29 ,MDL 1.33 6 0.03 1.21 6 0.40 ,MDL 0.33 (1)

372 6 36 444 6 40 777 6 32 1.70 6 0.19 0.19 6 0.08 ,MDL 9.55 6 2.28 ,MDL 0.21 6 0.04 1.09 6 0.04 0.28 6 0.02 ,MDL 4.54 6 0.18 ,MDL 1.11 6 0.11 0.97 6 0.04 ,MDL 0.24 6 0.03

344 6 28 389 6 25 904 6 348 1.69 6 0.19 0.57 6 0.25 ,MDL 12.37 6 1.85 ,MDL 0.25 6 0.03 1.02 6 0.05 0.22 6 0.02 ,MDL 4.12 6 0.17 ,MDL 1.19 6 0.10 0.74 6 0.09 ,MDL 0.25 6 0.03

280 6 21 338 6 25 989 6 242 1.23 6 0.08 0.14 6 0.02 ,MDL 13.29 6 1.31 ,MDL 0.21 6 0.06 0.97 6 0.05 0.24 6 0.03 ,MDL 4.34 6 0.15 ,MDL 1.33 6 0.16 0.79 6 0.03 ,MDL 0.29 6 0.05

261 6 36 335 6 47 559 6 88 1.62 6 0.32 0.18 6 0.05 ,MDL 15.52 6 1.63 ,MDL 0.33 6 0.03 1.38 6 0.47 0.39 6 0.15 ,MDL 4.20 6 0.23 ,MDL 1.01 6 0.16 1.24 6 0.40 ,MDL 0.31 6 0.07

Note. Values represent means 6 SE of five animals/group, mg/g of tissue. ,MDL, less than minimum detection limit for that element in a specific tissue.

408

MORGAN ET AL.

TABLE 4 Tissue Distribution of Cu, In, and Se after Intratracheal Instillation of Copper Indium Diselenide Days after intratracheal instillation Tissue Lung

Spleen

Kidney

Femur

Liver

Blood

Element

Control 1.69 6 0.27 (5) 0.005 6 0.004 (5) 0.46 6 0.015 (5) 1.65 6 0.43 (4) 0.098 6 0.05 (3) 0.40 6 0.07 (4) 12.65 6 2.18 (5) 0.04 6 0.01 (5) 1.03 6 0.06 (5) 1.35 6 0.17 (4) 0.002 6 0.001 (5) 0.18 6 0.02 (5) 3.82 6 0.11 (5) 0.004 6 0.001 (5) 1.19 6 0.03 (5) 0.83 6 0.22 (5) 0.0005 6 0.0001 (5) 0.60 6 0.03 (5)

Cu In Se Cu In Se Cu In Se Cu In Se Cu In Se Cu In Se

1 64 6 22 (3) 132 6 47 (3) 169 6 56 (3) 1.28 6 0.20 (3) 0.24 6 0.02 (3) 0.40 6 0.01 (3) 10.49 6 1.60 (3) 0.28 6 0.26 (3) 1.14 6 0.06 (3) 0.88 6 0.24 (3) 0.01 6 0.01 (3) 0.23 6 0.03 (3) 3.53 6 0.08 (3) 0.009 6 0.004 (3) 1.30 6 0.10 (3) 0.58 6 0.05 (3) 0.001 6 0.0003 (3) 0.56 6 0.01 (3)

3

7

314 (1) 298 (1) 779 (1) 1.15 (1) 0.017 (1) 0.41 (1) 8.67 (1) 0.06 (1) 0.89 (1) 1.11 (1) 0.01 (1) 0.22 (1) 4.53 (1) 0.02 (1) 0.67 (1) nd 0.006 (1) 0.43 (1)

14

244 6 26 (4) 384 6 59 (4) 641 6 62 (4) 1.09 6 0.07 (4) 0.19 6 0.03 (4) 0.42 6 0.08 (4) 11.69 6 1.16 (4) 0.43 6 0.05 (4) 0.89 6 0.04 (4) 1.99 6 0.36 (4) 0.06 6 0.007 (4) 0.19 6 0.008 (4) 6.80 6 2.42 (4) 0.15 6 0.01 (4) 0.75 6 0.04 (4) 2.08 6 0.26 (3) 0.05 6 0.02 (4) 0.62 6 0.08 (4)

28

122 6 40 (3) 274 6 91 (3) 364 6 143 (3) 1.59 6 0.29 (3) 0.39 6 0.21 (3) 0.67 6 0.37 (3) 17.51 6 5.50 (3) 0.31 6 0.05 (3) 0.83 6 0.04 (3) 3.38 6 2.20 (3) 0.06 6 0.008 (3) 0.21 6 0.01 (3) 3.99 6 0.21 (3) 0.15 6 0.06 (3) 0.74 6 0.07 (3) 1.51 6 0.23 (3) 0.01 6 0.002 (3) 0.73 6 0.01 (3)

138 6 66 (2) 310 6 149 (2) 351 6 162 (2) 1.68 6 0.03 (2) 0.80 6 0.80 (2) 0.99 6 0.35 (2) 16.70 6 2.56 (2) 0.38 6 0.14 (2) 1.12 6 0.16 (2) 1.08 6 0.02 (2) 0.20 6 0.10 (2) 0.17 6 0.01 (2) 3.80 6 0.12 (2) 0.43 6 0.23 (2) 0.94 6 0.14 (2) 1.00 6 0.01 (2) 0.07 6 0.03 (2) 0.28 6 0.04 (2)

Note. Values represent means 6 SE (n), mg/g of tissue. nd, not determined (sample lost).

Although administration of a particle suspension by intratracheal instillation is not a good physiological substitute for inhalation of an aerosol, this method offered several desirable advantages. In addition to being a technically simple method, intratracheal instillation allowed accurate delivery of measured amounts of each compound thereby facilitating comparison of pulmonary absorption and the potential toxicity of these compounds. Administration of equivalent doses of CIS, CGS, and CdTe in an inhalation study would be difficult because a number of variables such as breathing patterns and the aero-

dynamic diameters of the different particles all affect pulmonary deposition. Of the three compounds, CdTe was clearly the most toxic for the lung. Inflammation and hyperplasia of alveolar type 2 cells were considerably more extensive in the lungs of CdTe-treated rats. In addition, BALF protein levels, a sensitive indicator of damage to the alveolar epithelial barrier (Henderson, 1989), was highest in CdTe-treated rats and peaked on day 3 corresponding to the peak in PMN influx. Although lung tissue damage was extensive in CdTe-treated rats, the numbers of

TABLE 5 Tissue Distribution of Cd and Te After Intratracheal Instillation of Cadmium Telluride Days after intratracheal instillation of CdTe Tissue Lung Spleen Kidney Femur Liver Blood

Element

Control

1

3

7

14

28

Cd Te Cd Te Cd Te Cd Te Cd Te Cd Te

,MDL ,MDL 0.21 6 0.12 ,MDL ,MDL ,MDL 0.26 6 0.03 ,MDL 0.06 6 0.06 ,MDL ,MDL ,MDL

404 6 30 477 6 38 0.08 6 0.08 0.85 6 0.08 0.41 6 0.03 3.26 6 0.34 0.34 6 0.03 0.28 6 0.02 2.11 6 0.13 0.53 6 0.03 ,MDL 1.33 6 0.08

328 6 21 358 6 24 0.86 6 0.20 1.46 6 0.27 4.11 6 0.31 3.83 6 0.60 0.36 6 0.03 0.56 6 0.05 4.63 6 0.30 0.96 6 0.06 ,MDL 2.66 6 0.14

298 6 22 329 6 26 28.4 6 5.4 35.1 6 7.1 15.1 6 0.14 4.80 6 0.58 0.91 6 0.08 1.48 6 0.14 14.56 6 0.87 8.82 6 0.60 0.04 6 0.04 4.25 6 0.22

223 6 8 253 6 10 62.0 6 6.0 82.8 6 10.2 62.8 6 6.4 8.10 6 1.3 1.65 6 0.24 3.48 6 0.51 23.34 6 0.51 7.08 6 0.73 0.08 6 0.05 5.33 6 0.19

141 6 22 159 6 31 51.9 6 6.0 64.6 6 6.9 107.1 6 14.6 4.82 6 0.83 1.47 6 0.29 2.85 6 0.69 30.36 6 4.25 5.96 6 0.85 0.12 6 0.05 4.93 6 0.47

Note. Values represent means 6 SE of five animals per group, mg/g of tissue. ,MDL, less than minimum detection limit.

409

COMPARATIVE PULMONARY TOXICITY

inflammatory cells recovered in BALF were lower than those in CGS- and CIS-treated rats. A decreased yield of BALF cells was also reported in the previous study (Morgan et al., 1995) and was attributed to the presence of fibrin plugs in the small airways of CdTe-treated rats which prevented collection of inflammatory cells from the small airways during lavage. Increased levels of BALF fibronectin have been observed in many fibrotic lung disorders (Rennard and Crystal, 1981; O’Connor et al., 1988; Maasilita et al., 1991) and may contribute to the fibrotic process by recruiting and stimulating the proliferation of fibroblasts (Rennard et al., 1981). Peak levels of BALF fibronectin in CdTe-treated rats were 10- to 20-fold higher than in rats treated with CGS and CIS, respectively. Levels of fibronectin in BALF peaked in the first few days after CdTe treatment and remained at elevated levels for the duration of the study. The persistent release of fibronectin likely provided a continuous stimulation of fibrosis in the lung. CdTe was a potent initiator of fibrogenesis in the lung and caused considerably greater deposition of collagen than either CIS or CGS. Levels of lung 4-hydroxyproline were still increasing 28 days after treatment with CdTe, indicating that the fibrosis was not resolving. The pulmonary fibrosis caused by CdTe is similar to that reported for a number of cadmium compounds (Dervan and Hayes, 1979; Niewoehner and Hoidal, 1982; Kutzman et al., 1986; Damiano et al., 1990, Driscoll et al., 1992). Cadmium compounds differ in their abilities to cause fibrosis, and these differences may be related to the ease by which these compounds dissolve and release free cadmium. The fibrogenicity of tellurium has not been established, and the contribution of tellurium to this lesion could not be determined from this study. Of the three compounds CdTe was previously shown to be the most soluble (Morgan et al., 1995), and in this study cadmium and tellurium were absorbed from the lung and distributed to extrapulmonary tissues in considerably greater quantities than the elements of either CGS or CIS. Lung levels of cadmium and tellurium decreased throughout the 28-day study, while increasing amounts of these elements were detected in extrapulmonary tissues. However, histological evaluations and clinical chemistry indices indicated no adverse effects on extrapulmonary organ systems, suggesting that the accumulation of cadmium and/or tellurium in these tissues had not attained toxic levels. One exception was the decreased weight and histological effects observed in the spleens of CdTe-treated rats. Although these effects were not necessarily indicative of toxicity, they represent an extrapulmonary response to CdTe that could potentially progress to an adverse effect. CIS caused a significantly greater influx of PMNs into the lung than either CGS or CdTe. An influx of PMNs into the lung is normally associated with tissue injury; however, only minimal structural damage to the lung was associated with CIS treatment. CIS also caused a significantly greater influx of PMNs than CdTe or CGS in an earlier study (Morgan et al.,

1995). A similar persistent inflammatory response was observed in rats after intratracheal administration of indium trichloride (Blazka et al., 1994). Although supporting evidence could not be found in the literature, it is possible that indium may be more inflammatory (i.e., elicit a greater chemotactic response) than the other elements present in these compounds. CIS was only slightly fibrogenic in the lung, and this correlates with its limited solubility (Morgan et al., 1995). The fibrosis caused by CIS may have been caused by small amounts of free indium. The water-soluble compound indium trichloride was highly fibrogenic when instilled into the lungs rats (Blazka et al., 1994), whereas chronic inhalation of the insoluble indium sesquioxide (In2O3) caused a sustained inflammatory response with no fibrosis (Leach et al., 1961). CGS was the least toxic of the compounds causing only a mild, transient inflammatory response in the lung. CGS was less fibrogenic than CIS although the chemical structures and solubilities of these two compounds are similar. Differences in the toxicity of these two compounds are likely due to differences in the toxicities of gallium and indium. The differences in the fibrotic responses to CGS and CIS may be attributed to the greater fibrogenic effect of indium. Indeed, highly soluble indium compounds are fibrogenic (Blazka et al., 1994), whereas free gallium does not cause fibrosis when instilled into the lungs (Webb et al., 1986). These intratracheal instillation studies provide evidence that exposure to high concentrations of CGS, CIS, and CdTe may cause significant pulmonary inflammation and fibrosis with limited pulmonary absorption and systemic effects. Although aerosol inhalation studies of CGS, CIS, and CdTe are needed to evaluate the health hazard of occupational exposure conditions, these acute toxicity data suggest that appropriate safety precautions should be followed to avoid inhalation exposure during the preparation, use, and disposal of materials containing these compounds. ACKNOWLEDGMENTS The authors acknowledge the technical support of N. Bernholc, B. Collins, D. Crawford, P. Dixon, R. Gay, J. Haseman, M. Moorman, R. O’Connor, S. Philpot, P. Rydell, W. Stephens, T. Walser, and T. Ward. The authors thank Dr. J. Bonner and Dr. R. Chapin for critical review of the manuscript. Chemical analyses were conducted under contract to Research Triangle Institute, and Radian, Inc., Research Triangle Park, North Carolina.

REFERENCES Blazka, M. E., Dixon, D., Haskins, E., and Rosenthal, G. J. (1994). Pulmonary toxicity to intratracheally administered indium trichloride in Fischer 344 rats. Fundam. Appl. Toxicol. 22, 231–239. Chuttani, H. K., Gupti, P. S., and Gultati, S. (1965). Acute copper sulfate poisoning. Am. J. Med. 39, 849 – 854. Damiano, V. V., Cherian, P. V., Frankel, F. R., Steeger, J. R., Sohn, M., Oppenheim, D., and Weinbaum, G. (1990). Intraluminal fibrosis induced unilaterally by lobar instillation of CdCl2 into the rat lung. Am. J. Pathol. 137, 883– 894.

410

MORGAN ET AL.

Dervan, P. A., and Hayes, J. A. (1979). Peribronchiolar fibrosis following acute experimental lung damage by cadmium aerosol. J. Pathol. 128, 143–150. Driscoll, K. E., Maurer, J. K., Lindenschmidt, R. C., Romberger, D., Renard, S. I., and Crosby, L. (1990a). Respiratory tract responses to dust: Relationships between dust burden, lung injury, alveolar macrophage fibronectin release, and the development of pulmonary fibrosis. Toxicol. Appl. Pharmacol. 106, 88 –101. Driscoll, K. E., Lindenschmidt, R. C., Maurer, J. K., Higgins, J. M., and Ridder, G. (1990b). Pulmonary response to silica or titanium dioxide: Inflammatory cells, alveolar macrophage-derived cytokines, and histopathology. Am. J. Respir. Cell. Mol. Biol. 2, 381–390. Driscoll, K. E., Maurer, J. K., Poynter, J., Higgins, J., Asquith, T., and Miller, N. S. (1992). Stimulation of rat alveolar macrophage fibronectin release in a cadmium chloride model of lung injury and fibrosis. Toxicol. Appl. Pharmacol. 116, 30 –37. Duckettt, S. (1982). The distribution and localization of 127mtellurium in normal and pathological nervous tissues of young and adult rats. Neurotoxicology 3, 63–74. Geary, D. L., Myers, R. C., Nachreiner, D. J., and Carpenter, C. P. (1978). Tellurium and tellurium dioxide: Single endotracheal injection to rats. Am. Ind. Hyg. Assoc. J. 39, 100 –109. Henderson, R. (1989). Bronchoalveolar lavage: A tool for assessing the health status of the lung. In Concepts in Inhalation Toxicology (R. O. McClellan and R. F. Henderson, Eds.), pp. 415– 444. Hemisphere, New York. Kutzman, R. S., Drew, R. T., and Shiotsuka, R. N. (1986). Pulmonary changes resulting from subchronic exposure to cadmium chloride aerosol. J. Toxicol. Environ. Health 17, 175–189. Lauwerys, R. R., Bernard, A., Roels, H. A., Buchet, J. P., and Viau, C. (1984). Characterization of cadmium proteinuria in man and rat. Environ. Health Perspect. 54, 147–152. Leach, L. J., Scott, J. K., Armstrong, R. D., Steadman, L. T., and Maynard, E. A. (1961). The Inhalation Toxicity of Indium Sequioxide in the Rat, The University of Rochester Atomic Energy Project Report No. UR-590. University of Rochester, Rochester, NY. Lundborg, M., Birger, L., and Cammer, P. (1984). Ability of rabbit alveolar macrophages to dissolve metals. Exp. Lung Res. 7, 11–22.

Maasilita, P. E., Saslonen, M., Vaheri, A., Kivisaari, L., Holsti, L. R., and Mattson, K. (1991). Pro-collagen III in serum, plasminogen activation and fibronectin in bronchoalveolar lavage fluid during and following irradiation of human lung. Int. J. Radiat. Oncol. Biol. Phys. 20, 973–980. Miller, R. G., Jr. (1966). Simultaneous Statistical Inference. McGraw-Hill, New York, NY. Morgan, D. L., Shines, C. J., Jeter, S. P., Wilson, R. E., Elwell, M. P., Price, H. C., and Moskowitz, P. D. (1995). Acute pulmonary toxicity of copper gallium diselenide, copper indium diselenide, and cadmium telluride intratracheally instilled into rats. Environ. Res. 71, 16 –24. Niewoehner, D. E., and Hoidal, J. R. (1982). Lung fibrosis and emphysema: divergent responses to a common injury? Science 217, 359 –360. O’Connor, C., Odlum, C., Van Breda, A., Power, C., and Fitzgerald, MX. (1988). Collagenase and fibronectin in bronchoalveolar lavage fluid in patients with sarcoidosis. Thorax 43, 393– 400. Patty, F. A. (1963). Arsenic, phosphorus, selenium, sulfur, and tellurium. In Industrial Hygiene and Toxicology (D. W. Fassett and D. D. Irish, Eds.), 2nd ed., pp. 871–910. Interscience, New York. Rennard, S. I., and Crystal, R. G. (1981). Fibronectin in human bronchopulmonary lavage fluid: Elevation in patients with interstitial lung disease. J. Clin. Invest. 69, 113–122. Rennard, S. I., Hunninghake, G. W., Bitterman, P. B., and Crystal, R. G. (1981). Production of fibronectin by the human alveolar macrophage: Mechanism for the recruitment of fibroblasts to sites of tissue injury in interstitial lung diseases. Proc. Natl. Acad. Sci. USA 78, 7147–7151. Service, R. F. (1996). New solar cells seem to have power at the right price. Science 272, 1744 –1745. Snedecor, G. W., and Cochran, W. G. (1980). Statistical Methods. Iowa State Univ. Press, Ames, IA. Webb, D. R., Wilson, S. E., and Carter, D. E. (1986). Comparative pulmonary toxicity of gallium arsenide, gallium (III) oxide, or arsenic (III) oxide intratracheally instilled into rats. Toxicol. Appl. Pharmacol. 82, 405– 416. Witschi, H. P., Tryka, A. F., and Lindenschmidt, R. C. (1985). The many faces of an increase in lung collagen. Fundam. Appl. Toxicol. 5, 240 –250. Woessner, J. F. (1985). The determination of hydroxyproline in tissue and protein samples containing small proportions of this amino acid. Arch. Biochem. Biophys. 93, 440 – 447.