Changes in the Testicular Damage Caused by Indium Arsenide and Indium Phosphide in Hamsters during Two Years after Intratracheal Instillations

Transcription

1 J Occup Health 2000; 42: Journal of Occupational Health Changes in the Testicular Damage Caused by Indium Arsenide and Indium Phosphide in Hamsters during Two Years after Intratracheal Instillations Minoru OMURA 1, Koji YAMAZAKI 1, 2, Akiyo TANAKA 1, Miyuki HIRATA 1, Yuji MAKITA 1 and Naohide INOUE 1 1 Department of Hygiene, 2 Department of General Surgery, Graduate School of Medical Sciences, Kyushu University Abstract: Changes in the Testicular Damage Caused by Indium Arsenide and Indium Phosphide in Hamsters during Two Years after Intratracheal Instillations: Minoru OMURA, et al. Department of Hygiene, Graduate School of Medical Sciences, Kyushu University Change in the testicular damage caused by indium arsenide (InAs) and indium phosphide (InP) was examined during two yr after repetitive intratracheal instillations in hamsters. In this study, 4.0 mg/kg body weight/day of InAs or 3.0 mg/kg body weight/day of InP was instilled intratracheally twice weekly for eight wk. A single instillation dose of indium was 2.4 mg/kg body weight in both groups. Testicular damage was evaluated 0, 8, 16, 40, 64 and 88 wk after the last instillation. Both InAs and InP were proved to be definite testicular toxicants. Both materials decreased reproductive organ weight and caudal sperm count, and caused severe histopathologic changes in the testes. InAs-induced testicular damage was always more serious than InP-induced testicular damage. The serum indium concentration in the InAs group was always higher than that in the InP group, and indium was probably a toxic element in both materials. In the histopathologic examination, vacuolization of seminiferous epithelium was frequently observed as an early histopathologic change and spermatogonia remained in general even in the seminiferous tubules with severe histopathologic changes in both groups. It is therefore estimated that Sertoli cells, not stem cell spermatogonia, were the target cells of these indiumcontaining compound semiconductor materials. The threat of InAs and InP to male reproduction was proved in this study. We concluded that male reproductive disorders should not be overlooked when severe exposure to indium-containing compound Received March 22, 2000; Accepted April 24, 2000 Correspondence to: M. Omura, Department of Hygiene, Graduate School of Medical Sciences, Kyushu University, Fukuoka , Japan semiconductor materials is apparent in human subjects. (J Occup Health 2000; 42: ) Key words: Indium arsenide, Indium phosphide, Indium, Intratracheal instillation, Testicular toxicity, Long term observation, Sertoli cell, Spermatogonia III-V compound semiconductor materials, such as gallium arsenide (GaAs), are now extensively used for manufacturing light-emitting diodes, semiconductor lasers and microcircuits 1). This means that the health risk due to these materials needs to be properly assessed. Indium is an element in the IIIa column of the periodic table and is incorporated into III V compound semiconductor materials, such as indium arsenide (InAs) and indium phosphide (InP). Only a few data are available to assess the health risk of indium-containing compound semiconductor materials 2 7), especially to assess the risk to the reproductive system 6, 7). We previously examined the testicular toxicity of GaAs and InAs immediately after eight-week repetitive intratracheal instillation (7.7 mg/kg body weight/day, twice a week) in rats 6) and hamsters 7). GaAs showed a definite testicular toxicity in both species. InAs showed only a weak testicular toxicity in rats and did not show any sign of testicular toxicity in hamsters. But our study on hamsters was inadequate to evaluate the testicular toxicity of InAs. In the study, eight-wk intratracheal instillation could not be completed because of emaciation of the animals 7). We also thought that our study on rats might have underestimated the testicular toxicity of InAs because high concentrations of indium and arsenic were still detected in the serum of hamsters a long time after the intratracheal instillations of InAs 7). Chemical-induced adverse effects on the male reproductive system are drawing attention in the field of industrial hygiene in Japan because of 2-bromopropane-induced reproductive

2 Minoru OMURA, et al.: Testicular Damage Caused by InAs and InP in Hamsters 197 hazards in a Korean electronic company 9 11) and widespread anxiety about endocrine-disrupting chemicals. We therefore decided to re-examined the testicular toxicity of InAs in hamsters. Testicular toxicities of InP were also examined and were compared with those of InAs in order to clarify the contribution of indium to the testicular toxicity of indium-containing compound semiconductor materials. Materials and Methods Test materials InAs and InP were provided by Mitsuwa Chemicals (Osaka, Japan) and had a purity of more than 99.99%. These materials were finely pulverized in an agate mortar. The mean count diameter for InAs and InP was 1.58 µm [σg (geometric standard deviation): 2.15] and 1.06 µm [σg: 1.80], respectively. Both powders were analyzed in an energy dispersive X-ray fluorescence element analyzer (MESA-500, Horiba, Ltd., Kyoto, Japan). InAs powder contained 0.01% (wt%) of zirconium and 0.01% of yttrium, and InP powder contained 0.01% of zirconium and a trace amount of yttrium. Test materials were suspended in particle form in pathogen-free phosphate buffer solution (0.025 M, ph 6.86) just before instillation. The concentrations of InAs and InP in the suspension were 2.0 mg/ml and 1.5 mg/ml, respectively. Animal Male Syrian golden hamsters were purchased at six weeks of age from Japan SLC, Inc., (Shizuoka, Japan). The animals were housed four per stainless cage and were maintained in a specific pathogen-free laboratory room of the Laboratory of Animal Experiments, Faculty of Medicine, Kyushu University. The light cycle was 12 h: 12 h (light/dark), the temperature was C, and the air humidity was 50 60%. The animals were provided with CE-2 feed (Clea Japan Inc., Tokyo, Japan) and tap water ad libitum. This experiment was reviewed by the Committee of Ethics on Animal Experiments in the Faculty of Medicine, Kyushu University and was carried out under the Guidelines for Animal Experiment of the Faculty of Medicine, Kyushu University and the Law (No. 105) and Notification (No. 6) of the Government of Japan. Treatment After a two-week acclimation period, 144 hamsters, eight weeks of age, were randomized into three groups by weight: a control group, an InAs group and an InP group. The mean body weights of hamsters in the control group, InAs group and InP group were g, g, and g, respectively. Hamsters in each group were intratracheally instilled with a 2.0 ml suspension/kg body weight under ether anesthesia. A single instillation dose of test material was 4.0 mg/kg body weight in the InAs group and 3.0 mg/kg body weight in the InP group. The single instillation dose of indium was 2.4 mg/kg body weight in both groups. Hamsters in the control group were given phosphate buffer solution only. Hamsters were thus treated twice weekly for eight weeks. No animals died during the administration period. Examination of the effects on the male reproductive system Eight hamsters died of emaciation (five in the InAs group and three in the InP group), seven were cannibalized (four in the control group and three in the InAs group), and four were killed accidentally before examination in the control group during the observation period. These animals were excluded from all evaluations. Four to eight hamsters in each group were weighed and killed with an overdose of ether 0, 8, 16, 40, 64 and 88 wk after the last instillation. The testes and epididymes were removed and weighed. The right cauda epididymidis was homogenized in saline containing 0.05% (v/v) Triton X- 100 in a blender and homogenization-resistant sperm were counted in a hemocytometer. The right testis was fixed in Bouin s solution, embedded in paraffin wax, thinly sectioned and stained with periodic acid Schiff reagent (PAS) and hematoxylin. Histopathologic changes in the testis were examined under an optical microscope. All cross-sections of seminiferous tubules in one transverse section of the testis (about cross-sections of the tubule) were examined and the seminiferous tubules with histopathologic changes were counted. Degeneration of one or two germ cells was not regarded as a histopathologic change. The results of this study concerning the pulmonary toxicity and serum concentrations of indium and arsenic have already been reported by Yamazaki et al. 8). Statistical analysis Data were analyzed for the mean and standard deviation and F test was performed to evaluate equality of variance. If a significant difference was found in a variance, t test with Welch s correction was used for the statistical analysis, and otherwise a t test without a correction was used. The differences were interpreted as significant at p<0.05. There were only two hamsters adequate for evaluation in the control group 8 wk after the last instillation, (four were killed accidentally before examination and two had marked unilateral atrophy of the testis and the epididymis). Therefore, statistical analysis was not performed at this point. Results Body weight, reproductive organs weights, and sperm count in the cauda epididymidis Body weight in the InAs group had already decreased significantly immediately (0 wk) after the last instillation

3 198 J Occup Health, Vol. 42, 2000 and was always approximately 70% of the control value thereafter (Fig. 1). Body weight in the InP group was compatible with that in the control group immediately after the last instillation. Body weight in this group decreased significantly from 16 wk to 64 wk after the last instillation and was 80 90% of the control value in this period. In the InP groups, body weight became compatible with the control value again 88 wk after the last instillation (Fig. 1). Weights of the two reproductive organs in the InAs group had already decreased significantly immediately after the last instillation. Reproductive organ weights in this group decreased further afterwards and were 30 50% of the control values from 16 wk to 88 wk after the last instillation (Figs. 2, 3). Reproductive organ weights in the InP group were compatible with those in the control group from 0 wk to 8 wk after the last instillation. Reproductive organ weights in this group decreased significantly from 16 wk to 64 wk after the last instillation and were 60 70% of the control values in this period. In the InP group, reproductive organ weights became compatible with the control values again 88 wk after the last instillation (Figs. 2, 3). The sperm count in the cauda epididymidis in the InAs group had already decreased significantly immediately after the last instillation. The caudal sperm count in this group decreased further afterwards and was 10 30% of the control value from 16 wk to 88 wk after the last instillation (Fig. 4). Unlike body weight and reproductive organ weights, the caudal sperm count in the InP group Fig. 1. Body weight change in two yr after the repetitive intratracheal instillations of indium arsenide (4.0 mg/ kg body weight/d, ) and indium phosphide (3.0 mg/ kg body weight/d, ) twice weekly for eight wk ((A) absolute value ( ; control group), (B) % mean control value). Error bars indicate standard deviation. The number under a mean bar is the number of animals. Statistical significance is analyzed with the t-test; *p<0.05, **p<0.01. Statistical analysis is not performed 8 wk after the last instillation because of the small number of animals in the control group. Fig. 2. Testes weight change in two yr after the repetitive intratracheal instillations of indium arsenide (4.0 mg/ kg body weight/d, ) and indium phosphide (3.0 mg/ kg body weight/d, ) twice weekly for eight wk ((A) absolute value ( ; control group), (B) % mean control value). Error bars indicate standard deviation. The number under a mean bar is the number of animals. Statistical significance is analyzed with the t-test; *p<0.05, **p<0.01. Statistical analysis is not performed 8 wk after the last instillation because of the small number of animals in the control group.

4 Minoru OMURA, et al.: Testicular Damage Caused by InAs and InP in Hamsters 199 Fig. 3. Epididymes weight change in two yr after the repetitive intratracheal instillations of indium arsenide (4.0 mg/kg body weight/d, ) and indium phosphide (3.0 mg/kg body weight/d, ) twice weekly for eight wk ((A) absolute value ( ; control group), (B) % mean control value). Error bars indicate standard deviation. The number under a mean bar is the number of animals. Statistical significance is analyzed with the t-test; *p<0.05, **p<0.01. Statistical analysis is not performed 8 wk after the last instillation because of the small number of animals in the control group. Fig. 4. Change in sperm count in cauda epididymidis in two yr after the repetitive intratracheal instillations of indium arsenide (4.0 mg/kg body weight/d, ) and indium phosphide (3.0 mg/kg body weight/d, ) twice weekly for eight wk ((A) absolute value ( ; control group), (B) % mean control value). Error bars indicate standard deviation. The number under a mean bar is the number of animals. Statistical significance is analyzed with the t-test; *p<0.05, **p<0.01. Statistical analysis is not performed 8 wk after the last instillation because of the small number of animals in the control group. had already decreased significantly immediately after the last instillation. The caudal sperm count in this group decreased further thereafter and was 40 50% of the control value from 16 wk to 64 wk after the last instillation. In the InP group, like body weight and reproductive organ weights, the caudal sperm count recovered to the control level 88 wk after the last instillation (Fig. 4). Histopathologic changes in the testis In the InAs group, the frequency of seminiferous tubules with histopathologic changes had already increased immediately after the last instillation (Fig. 5). The frequency of seminiferous tubules with histopathologic changes in the InP group were compatible with the control value until 8 wk after the last instillation and, first increased significantly 16 wk after the last instillation (Fig. 5). Both in the InAs and InP groups, vacuolization of seminiferous epithelium was frequently observed as an early histopathologic change (Figs. 6b, 6c). During the observation period, severe histopathologic changes were always observed in the InAs group. In this group, 70 90% of seminiferous tubules had histopathologic changes from 16 wk to 88 wk after the last instillation (Fig. 5). These histopathologic changes included degeneration and loss of germ cells, exfoliation and disarrangement of seminiferous epithelium, and atrophy of seminiferous tubules, but

5 200 J Occup Health, Vol. 42, 2000 Fig. 5. Change in the frequency of seminiferous tubules with histopathologic lesions in two yr after the repetitive intratracheal instillations of indium arsenide (4.0 mg/kg body weight/d, ) and indium phosphide (3.0 mg/kg body weight/d, ) twice weekly for eight wk ( ; control group). Error bars indicate standard deviation. The number under a mean bar is the number of animals. Statistical significance is analyzed with the t-test; *p<0.05, **p<0.01. Statistical analysis is not performed 8 wk after the last instillation because of the small number of animals in the control group. spermatogonia remained in general even in the seminiferous tubules with severe histopathologic changes (Fig. 6d). Severe histopathologic changes were also observed in the InP group from 16 wk to 64 wk after the last instillation, although the frequency was always less than that in the InAs group (30 50%, Fig. 5). Spermatogonia remained in general in the seminiferous tubules in the InP group, also (Fig. 6e). In the control group, the frequency of seminiferous tubules with histopathologic changes gradually increased with age and 13.6 ± 8.9% of seminiferous tubules had histopathologic changes 88 wk after the last instillation (104 wk of age, Fig. 5). At this point, the frequency of seminiferous tubules with histopathologic changes was 31.2 ± 38.7% in the InP group, but the frequency was 14.2 ± 8.0% in this group at this point when one animal with extremely high frequency (99.4%) was excluded. In the InP group, the histopathologic findings 88 wk after the last instillation were apparently different from those until 64 wk after the last instillation. At this point, most of seminiferous tubules were normal and a few seminiferous tubules with severe atrophy were mingled in some places in the InP group (Fig. 6g). Discussion In the previous study, we could not prove the testicular toxicity of InAs in hamsters 7), but that study was inadequate to evaluate the testicular toxicity of InAs because planned repetitive intratracheal instillations could not be completed because of emaciation of the animals. We therefore re-examined the testicular toxicity of InAs in hamsters in this study. We used approximately half the daily dose of InAs in this study to prevent animal emaciation. No animals died during the administration period and more than 80% survived until the planned day of euthanasia in the InAs and InP groups, so that the testicular toxicities of InAs and InP were thought to be adequately evaluated in this study. The Table 1 summarizes the results of this study. Both InAs and InP were proved to be definite testicular toxicants. Testicular toxicity of InAs was first proved in hamsters and the toxicity was far more greater than that seen in rats in our previous study 6). In our two previous studies, testicular toxicity of InAs was once evaluated immediately after the last of the repetitive instillations 6, 7). In this study, the testicular toxicity of InAs was repeatedly evaluated until 88 wk after the last of the repetitive instillations and, as shown in Figs. 2 5, InAs-induced testicular damage became more severe until 16 wk after the last instillation. Testicular damage was still in progress immediately after the last instillation of InAs in this study, and therefore the testicular toxicity of InAs had been underestimated in our two previous studies. We proved the testicular toxicity of InP for the first time in this study. As shown in Figs. 2 5, InP-induced testicular damage was always less than InAs-induced damage. There are two possible ways to explain this difference. One is the existence of internal exposure to arsenic in the InAs group and the absence of this exposure in the InP group. In this study, 0.4 µm arsenic was constantly detected in the serum of InAs-treated hamsters and the serum concentration of arsenic in the InP group (0.05 µm) was similar to the control value (0.04 µm) during the observation period (Fig. 7) 8). In our previous study 7), the mean serum concentration of arsenic in As 2 O 3 -treated hamsters was 1.31 µm immediately after the last instillation. This concentration was three times higher than that in the InAs group in this study, although no sign of testicular damage was found in As 2 O 3 -treated hamsters in our previous study 7). These results indicate that the internal exposure to arsenic did not fully explain the more potent testicular toxicity of InAs. The other cause is higher internal exposure to indium in the InAs group than in the InP group. In this study, a single instillation dose of indium was the same in both groups. But, as shown in Fig. 7, the serum indium concentration in InAs-treated hamsters was constantly more than twice that in InP-treated hamsters 8). This is probably due to a greater influx of indium into serum in the InAs group than in the InP group 8). The hypothesis that the difference in the internal exposure to indium was the cause of the difference in the testicular toxicity of the two materials

6 Minoru OMURA, et al.: Testicular Damage Caused by InAs and InP in Hamsters 201 Fig. 6. Histopathologic findings in the testis in the control group (a), indium arsenide (InAs) group (b, d, f) and indium phosphide (InP) group (c, e, g). Vacuolizations of seminiferous epithelium (arrow heads) are observed as an early histopathological change (b; InAs group at 0 wk, c; InP group at 16 wk). Spermatogonia (arrow heads) remain in the seminiferous tubule with severe histopathological changes (d; InAs group at 16 wk, e; InP group at 16 wk). At 88 wk after the last instillation, almost all seminiferous tubules are severely affected in the InAs group (f), but most of the seminiferous tubules are normal and a few seminiferous tubules with severe atrophy (arrow heads) are mingled in some places in the InP group (g). Bar indicates 100 µm.

7 202 J Occup Health, Vol. 42, 2000 Table 1. Summary of the results of this study Weeks after the last instillation InAs Body weight reduction a) Testes weight reduction a) Epididymes weight reduction a) Caudal sperm count reduction a) Increase in histopathlogic changes in the testis b) InP Body weight reduction a) Testes weight reduction a) Epididymes weight reduction a) Caudal sperm count reduction a) Increase in histopathlogic changes in the testis b) c) a) Severity of the effect is classified as follows: : The reduction was not significant or less than 20% of the control value. +: The reduction was significant or more than 20% of the control value. ++: The reduction was more than 30% of the control value. +++: The reduction was more than 50% of the control value. b) Severity of the effect is classified as follows: : Less than 20% of the seminiferous tubules were damaged. +: More than 20% of the seminiferous tubules were damaged. ++: More than 30% of the seminiferous tubules were damaged. +++: More than 50% of the seminiferous tubules were damaged. c) One animal with an extremely high frequency of damaged tubules is excluded. Fig. 7. Changes in the serum indium concentration ( ; indium arsenide group, ; indium phosphide group) and serum arsenic concentration ( ; indium arsenide group) in two yr after the repetitive intratracheal instillations of indium arsenide (4.0 mg/kg body weight/d) and indium phosphide (3.0 mg/ kg body weight/d) twice weekly for eight wk. Error bars indicate standard deviation (n=4 8). (Figure is reproduced from Yamazaki et al. 7) and used with permission of the Japan Society for Occupational Health).

8 Minoru OMURA, et al.: Testicular Damage Caused by InAs and InP in Hamsters 203 was supported by other results in this study. In the InP group, testicular damage was found from 0 wk to 64 wk after the last instillation and recovered 88 wk after the last instillation (Table 1). In line with the recovery of testicular damage, the serum indium concentration decreased by approximately half in the InP group (Fig. 7) 8). These results indicate that indium was primarily responsible for the toxic effect of InP, and that the difference in the internal exposure to indium was the main cause of the difference in the testicular toxicities of InAs and InP. Body weight of the hamsters decreased to a maximum of 70% of the control value in the InAs group and decreased up to 80 90% of the control value in the InP group in this study (Fig. 1). Feed restriction and body weight reduction impair the reproduction of animals 12 16). Feed restriction which decreased body weight to 70% of the control value caused little testicular damage in rats 12). Feed restriction to the same degree caused a 10 20% decrease in reproductive organ weights and caudal sperm count in mice 14). Feed restriction and body weight reduction severely affect the reproduction of hamsters. Eskes reported that feed restriction which decreased body weight to 70% of the control value decreased the width of the testis to 60 80% of the control value 15). Pieper and colleagues reported that feed restriction which decreased body weight to 70% of the control value halved testis size (=the length times the width of the testis) 16). We must therefore consider the possibility that a decrease in food intake and a body weight reduction were responsible for testicular damage in the two groups. In the InP group, body weight began to decrease 16 wk after the last instillation but caudal sperm reduction had already been apparent immediately after the last instillation. Testicular damage preceded body weight reduction in the InP group and then it was improbable that a decrease in food intake and a body weight reduction were responsible for testicular damage in this group. Testicular damage in the InP group was primarily due to the toxic effect of indium. High internal exposure to indium was found in the InAs group, also, and, as described below, the same histopathologic changes were seen in the testes in the InAs and InP groups, so that it is reasonable to estimate that testicular damage in the InAs group was primarily due to the toxic effect of indium as in the InP group. But both testicular damage and a body weight reduction had already been apparent immediately after the last instillation in the InAs group. The possibility that a decrease in food intake and a body weight reduction were responsible for testicular damage in the InAs group could not be excluded in this study. Further study with smaller instillation doses of InAs is needed to verify this possibility. Testicular damage lasted for even more than one yr after the cessation of the instillations in both groups. Does this mean that testicular damage caused by InAs and InP was irreversible? The answer is no. In the histopathologic examination, severe tubular damage was always observed in both groups, but seminiferous tubules containing only Sertoli cells were rarely observed and spermatogonia remained in general even in the seminiferous tubules with severe histopathologic changes in both groups. This means that seminiferous tubules did not lose their ability to regenerate germ cells and that spermatogenesis always occurred in both groups during the observation period. In fact, testicular damage in the InP group recovered 88 wk after the last instillation, in line with the decrease in the serum indium concentration. Probably the internal exposure to indium was sufficiently high to prevent spermatogenesis for even more than one yr after the cessation of the instillations in both groups, and then testicular damage lasted. The serum indium concentration decreased with age in the InAs group as in the InP group, although testicular damage in the InAs group did not recover during the observation period. The serum indium concentration in the InAs group 88 wk after the last instillation was higher than that in the InP group 40 wk after the last instillation. This means that the serum indium concentration in the InAs group was always high enough to prevent spermatogenesis throughout the observation period, and testicular damage did not recover in this group. Both in the InAs and InP groups, vacuolization of seminiferous epithelium was frequently observed as an early histopathologic change. Vacuolization of seminiferous epithelium is a sign of Sertoli cell damage 17, 18). Considering histopathologic changes in the testes, we estimate that Sertoli cells were first damaged but that stem cell spermatogonia were rarely damaged in the InAs and InP groups. We reported that gallium arsenide, other III V compound semiconductor material, causes spermatid retention in seminiferous tubules and that gallium plays the main role in the testicular toxicity of gallium arsenide in hamsters 6, 7). Because spermatid retention in seminiferous tubules is also a sign of Sertoli cell damage 17, 18), elements in the III column of the periodic table might have toxic effects on Sertoli cells. Indium-containing compound semiconductor materials, InAs and InP, were proved to be definite testicular toxicants. It is estimated that indium was the main toxic element in the two materials and that Sertoli cells were damaged by them. Actual inhalation doses of InAs and InP in workplaces would be much lower than the doses used in this study and further studies are needed to assess the risk of low-dose indium-containing compound semiconductor materials on the human male reproductive function. Nevertheless, the threat of these materials to male reproductive function was proved in this study. Therefore, we thought that male reproductive disorders should not be overlooked when severe exposure to

9 204 J Occup Health, Vol. 42, 2000 indium-containing compound semiconductor materials is apparent in human subject. Acknowledgments and grant information: We thank Mrs. Yoshiko Hirose for technical assistance. This study was supported in part by Grant No from the Ministry of Education, Science and Culture, Japan. References 1) Ishiguro S. Industries using arsenic and arsenic compounds. Appl Organomet Chem 1992; 6: ) Tanaka A, Hisanaga A, Hirata M, Omura M, Inoue N, Ishinishi N. Pulmonary toxicity of indium arsenide and arsenic selenide following repeated intratracheal instillations to the lungs of hamsters. Appl Organomet Chem 1994; 8: ) Tanaka A, Hirata M, Omura M, et al. Comparative toxic study of the effect of gallium arsenide, indium arsenide and arsenic trioxide following intratracheal instillations to the lung of Syrian golden hamsters. Fukuoka Acta Medica 2000; 91: ) Uemura T, Oda K, Omae K, et al. Effects of intratracheally administered indium phosphide on male Fisher 344 rats. J Occup Health 1997; 39: ) Oda K. Toxicity of a low level of indium phosphide (InP) in rats after intratracheal instillation. Industrial Health 1997; 35: ) Omura M, Tanaka A, Hirata M, et al. Testicular toxicity of gallium arsenide, indium arsenide, and arsenic oxide in rats by repetitive intratracheal instillation. Fundam Appl Toxicol 1996; 32: ) Omura M, Hirata M, Tanaka A, et al. Testicular toxicity evaluation of arsenic-containing binary compound semiconductors, gallium arsenide and indium arsenide, in hamsters. Toxicol Lett 1996; 89: ) Yamazaki K, Tanaka A, Hirata M, et al. Long term pulmonary toxicity of indium arsenide and indium phosphide instilled intratracheally in hamsters. J Occup Health 2000; 42: ) Kim Y, Jung K, Hwang T, et al. Hematopoietic and reproductive hazards of Korean electronic workers exposed to solvents containing 2-bromopropane. Scand J Work Environ Health 1996; 22: ) Park J, Kim Y, Park D, et al. An outbreak of hematopoietic and reproductive disorders due to solvents containing 2-bromopropane in an electronic factory, South Korea: epidemiological survey. J Occup Health 1997; 39: ) Ichihara G, Asaeda N, Kumazawa T, et al. Testicular and hematopoietic toxicity of 2-bromopropane, a substitute for ozone layer-depleting chlorofluorocarbons. J Occup Health 1997; 39: ) Chapin RE, Gulati DK, Barnes LH, Teague JL. The effects of feed restriction on reproductive function in Sprague-Dawley rats. Fundam Appl Toxicol 1993; 20: ) Seki M, Yamaguchi K, Marumo H, Imai K. Effects of food restriction on reproductive and toxicological parameters in rats In search of suitable feeding regimen in long term tests. J Toxicol Sci 1997; 22: ) Chapin RE, Gulati DK, Fail PA, et al. The effects of feed restriction on reproductive function in Swiss CD- 1 mice. Fundam Appl Toxicol 1993; 20: ) Eskes GA. Gonadal responses to food restriction in intact and pinealectomized male golden hamsters. J Reprod Fert 1983; 68: ) Pieper DR, Lobocki CA, Karo KH. Olfactory bulbectomy counteracts inhibitory effect of food restriction on reproductive function. Am J Physiol 1994; 266: R1891 R ) Nolte T, Harleman JH, Jahn W. Histopathology of chemically induced testicular atrophy in rats. Exp Toxic Pathol 1995; 47: ) Creasy DM. Evaluation of testicular toxicity in safety evaluation studies: The appropriate use of spermatogenic staging. Toxicol Pathol 1997; 25: