Contribution of toxicological pathology to occupational health: lung carcinogenicity of fibrous and particulate substances in rats

Abstract

In this review, we focus on the rat pulmonary carcinogenicity of two solid substances, fibrous multi-walled carbon nanotube (MWCNT) and particulate indium tin oxide (ITO). Inhalation exposure to MWNT-7, a type of MWCNTs, and ITO induced lung carcinogenicity in both male and female rats. Toxicity to the alveolar epithelium is induced by macrophages undergoing frustrated phagocytosis or frustrated degradation of engulfed particles (referred to as frustrated macrophages). Melted macrophage contents contribute significantly to development of hyperplasia of the alveolar epithelium, which eventually results in the induction of lung carcinoma. MWNT-7 and ITO induce secondary genotoxicity; consequently, a no-observed-adverse-effect level can be applied to these materials rather than benchmark doses that are used for non-threshold carcinogens. Thus, establishing occupational exposure limit values for MWNT-7 and ITO based on the existence of a carcinogenic threshold is reasonable.

Introduction

Chemical carcinogens in humans are identified by epidemiological data: for example, cancer that develops in factory workers (occupational cancer). However, epidemiological data is not always available. Therefore, in general, the carcinogenic hazards of chemical substances to humans are evaluated based on animal carcinogenicity studies. Moreover, it is particularly important to determine whether genotoxicity contributes to the harmful effects elicited by the tested chemicals, such as their carcinogenicity and reproductive toxicity. In animal carcinogenicity studies, rodents exposed to solid substances, such as fibrous and particulate substances, develop respiratory diseases, particularly tumors, such as lung carcinoma and mesothelioma. Thus, the inhalation route is associated with tumors located in the respiratory tract. Japan Bioassay Research Center (JBRC) has investigated the inhalation carcinogenicity of many chemical substances in rats and mice. In this review, we focused on the rat pulmonary carcinogenicity of two solid substances, fibrous multi-walled carbon nanotube (MWCNT)¹ and particulate indium tin oxide (ITO)². Then, we discuss about the allowable occupational exposure limit (OEL) using the no-observed-adverse-effect level (NOAEL) or benchmark dose (BMD) as a point of departure to derive occupational health guidance values.

Carcinogenicity of Fibrous Substances

Background

Asbestos is a naturally occurring fibrous silicate mineral that was used as a soundproofing and heat insulating material. Asbestos does not represent a human hazard owing to its chemical properties, but the fibers themselves represent one when dispersed at the worksite and inhaled by workers, causing mesothelioma and lung cancer. Asbestos fibers are not only an occupational health issue, but also an environmental issue.

In Great Britain, the incidence of deaths from mesothelioma increased linearly until around 2017, then leveled off and is expected to decline progressively after 2020³. The total ban on the use of asbestos in Great Britain was implemented in 1999. In Japan, asbestos-induced pleural mesothelioma and lung carcinoma have been reported⁴. While the number of cases is lower than that in Great Britain, the incidence has not decreased during the past five years. Notably, a total ban on the use of asbestos was implemented in Japan in 2004.

The fact that asbestos, a fibrous material, causes mesothelioma and lung carcinoma in humans strongly suggests that long-term accumulation of inhaled thin and long fibers in the body could cause the same diseases. Moreover, the causal factor of fiber-induced disease is not the chemical properties of the fibrous substances; disease induction is rather based on the ratio of the length to the width of the fiber (Stanton-Pott theory)^{5, 6}.

Nanomaterial science has brought considerable benefits to the society. However, many studies have suggested that nanomaterials adversely affect human health. For example, in rodents, intraperitoneal or intraabdominal administration of MWNT-7, one of the representative MWCNTs that are fibrous nanomaterials, causes peritoneal or abdominal mesothelioma^{7, 8, 9}. Based on these results, the question was whether inhalation exposure to MWCNT would cause lung carcinoma or pleural mesothelioma in humans, such as asbestos. To resolve this issue, the JBRC conducted an inhalation carcinogenicity study of MWNT-7 using rats¹.

Carcinogenicity of MWCNT in rats by inhalation exposure

Development of an inhalation exposure device for MWCNT

Two methods exist for evaluating inhalation exposure to dust in rodents: whole-body and nose-only inhalation exposure system. The whole-body system allows free movement of rodents, resulting in minimal stress, and is tolerable for long continuous periods of exposure of the experimental animal to the test materials. Therefore, whole-body inhalation exposure was selected for our 2-year carcinogenicity/chronic inhalation exposure to MWCNT in rats.

MWCNT are very fine and light, and human exposure most likely occurs owing to inhalation of their non-agglomerated forms. Therefore, it is essential to conduct animal inhalation studies by using dry air. However, difficulties arise due to technical limitations in generating MWCNT aerosols suitable for animal toxicological studies. The major problem is that MWCNT do not retain their original single form at high concentrations, owing to agglomeration. Therefore, an aerosol generator that can separate single fibers from agglomerates is required to deliver single-fiber MWCNT to the inhalation chamber at various exposure concentrations. In addition, the generator must allow aerosols to be consistently introduced into the inhalation chamber throughout the two-year period.

Three of the basic conditions that must be met when performing solid dust inhalation tests are concern the control of the concentration of the exposure dust and dust particle size. The OECD guidelines¹⁰ state that, first, the exposure concentration should be kept constant, with less than 20% fluctuation in the chamber concentration; the JBRC aims to keep the coefficient of variation (CV) within 10%. Second, the exposure particles must be respirable. The OECD guidelines state that the mass median diameter (MMAD) should be 2 µm or less; the JBRC attempts to maintain the MMAD of the test substance at 1.5 μm or less. Third, in chambers with different particle concentrations, the size distributions of the particles must be the same.

The JBRC has developed a novel system for MWCNT fiber generation¹¹. The layout of the aerosol generation and exposure device for MWCNT dust inhalation is shown in Fig. 1. This is a dry powder system that uses the cyclone sieve method to generate MWCNT aerosols. Briefly, a dust feeder supplied bulk MWCNT, MWNT-7 (Hodogaya Chemical Co. Ltd, Tokyo, Japan) to the aerosol generator. The generator dry-aerosolized MWNT-7 with upward spiraling air from the slits at the bottom of the generator. Near the top of the generator was a partitioning sieve (mesh size of 53 μm), and only the MWNT-7 fibers that passed through the sieve reached the inhalation chamber. The fiber concentration in the chamber was constantly monitored using an optical particle controller (OPC). In the final stage of development, we tested this system to ensure that exposure concentrations were constant for 6 h and confirmed that each of 0.2, 1, and 5 mg/m³ concentrations was constant (Fig. 2). The fluctuations in mass concentration were within 5% of the set concentration with CV. Moreover, using two different measuring instruments, a micro-orifice uniform deposit cascade impactor and a scanning mobility particle sizer, we confirmed that the size distribution of the MWNT-7 fibers was the same at all three concentrations¹¹.

Fig. 1.

Aerosol generator and exposure system. Left: aerosol generator; cyclone sieve method. Right: layout of aerosol generator and exposure system.

Fig. 2.

Control of MWNT-7 concentration in the inhalation chamber for 6 hr measured by optical particle controller (OPC). Concentration values are shown as concentration per minutes (CPM).

MWNT-7 2-year carcinogenicity study in rats¹

The JBRC undertook an OECD and GLP guideline-compliant 2-year carcinogenicity study of MWNT-7 in rats, based on the results of a 2-week and a 13-week preliminary studies^{12, 13}. In the 2-year study, 6-week-old male and female F344 rats were divided into four groups and exposed to MWNT-7 (average fiber width of 83.8–90.7 nm and length of 5.2–5.7 μm) by whole body inhalation at exposure concentrations of 0 (control), 0.02, 0.2, and 2 mg/m³ for 6 hours a day, 5 days a week for 2 years. MWNT-7 concentrations and sizes were constant throughout the 2-year duration of the study. For morphological examination, MWNT-7 was collected from the inhalation chamber, and its morphology was confirmed by scanning electron microscopy (SEM). No aggregated MWNT-7s were observed in any of the inhalation chambers from the first week to the 104th week of the study. The SEM images of MWNT-7 demonstrated the presence of single fibers or two or three entwined fibers, and each group showed similar images (Fig. 3).

Fig. 3.

Scanning electron microscope (SEM) images of MWNT-7 in the inhalation chamber at weeks 1 and 102. Individual MWNT-7 fibers are seen at each concentration. (Quotation from Particle and Fibre Toxicology 13: 53. 2016)

There were no differences in the general condition or survival of the rats in the male and female groups compared to the controls. No growth retardation was observed in either the male or female groups.

The induction of tumors in the lungs of rats exposed to MWNT-7 is shown in Table 1. Exposure to MWNT-7 induced bronchioloalveolar carcinoma, total of carcinoma, and adenoma and/or carcinoma in a concentration-dependent manner in both male and female rats (Fig. 4A and 4B). The incidence of these tumors was significantly increased in the middle- and high-concentration male groups, and in the high-concentration female group compared to each control group. Thus, lung tumor induction owing to inhaled MWNT-7 differed with gender.

Table 1. Neoplastic Lesions of the Lung in Rats Exposed to MWNT-7 by Inhalation for Two Years.

Fig. 4.

Histopathological changes in the lung of rats exposed to MWNT-7. A–D: (A) Bronchioloalveolar carcinoma in the lung of the male rat exposed to 2 mg/m³ MWNT-7. Carcinoma cells invaded the wall of the vein. Bar, 50 µm; H&E staining. (B) Bronchioloalveolar carcinoma in the lung of a female rat exposed to 2 mg/m³ MWNT-7. Carcinoma cells are accompanied by severe fibrous connective tissue. Bar, 50 µm; H&E staining. (Fig. 4A and 4B: Quotation from Particle and Fibre Toxicology 13: 53. 2016). (C) Bronchioloalveolar hyperplasia in the lung of the male rat exposed to 2 mg/m³ MWNT-7. Bar, 50 µm; H&E staining. (D) Granulomatous change in the lung of a male rat exposed to 2 mg/m³ MWNT-7. Bar, 100 µm; H&E staining.

MWNT-7 fibers were also found in many organs and tissues, including the pleural and abdominal cavities. However, no MWNT-7-related tumors were observed in any organ other than the lung, including the pleural and peritoneal mesothelium, in any of the male or female groups.

The frequencies of hyperplastic epithelial and non-epithelial lesions in the lungs of rats exposed to MWNT-7 are shown in Table 2. Males exposed to concentrations of MWNT-7 at 0.2 mg/m³ and higher presented a significant induction of bronchioloalveolar hyperplasia (Fig. 4C) and alveolar hyperplasia, while the 2 mg/m³ group presented significantly increased bronchiolar hyperplasia and atypical hyperplasia. In females, significant induction of these hyperplastic epithelial lesions were also observed in the 2 mg/m³ group but not in the 0.2 mg/m³ group. Importantly, bronchioloalveolar hyperplasia and alveolar hyperplasia are presumed to be preneoplastic lesions. In addition, the induction of epithelial hyperplasia and tumor in the lung is related to the amount of MWNT-7 deposited in the lung, as described below.

Table 2. Hyperplastic Epithelial and Non-epithelial Lesions of the Lung in Rats Exposed to MWNT-7 by Inhalation for Two Years.

The accumulation of alveolar macrophages, the presence of granulomatous changes, and fibrosis are shown in Table 2. Accumulation of alveolar macrophages was observed in the 2 mg/m³ male and female groups. Granulomatous changes (Fig. 4D), which are associated with exposure to inhaled fibrous substances, and fibrosis of the alveolar wall were observed at concentration at 0.2 mg/m³ and higher in both male and female rats. These lesions were observed particularly in the alveolar ducts and alveoli near bronchioles. Fibrous thickening of the visceral pleura was also observed, but it was mild and localized. No fibrous thickening was observed in the parietal pleura.

Fate of inhaled MWNT-7

In dust inhalation studies, the interaction of the test particle with the pulmonary system is an important aspect of the investigation of particle-associated pathology and its mode of action on lung toxicity.

In the carcinogenicity study of MWNT-7 inhalation exposure, important points in the interaction of the fibers with the pulmonary system are (i) MWNT-7 deposition in the lung (lung burden), (ii) phagocytosis by macrophages in the lung, (iii) translocation from the lung to the pleural cavity and other tissues and organs, (iv) lung overload, and (v) the physico-chemical properties of the fibers, including shape and size (length and diameter), impurities, and surface characteristics¹⁴. In this review, deposition, phagocytosis, and translocation are discussed below.

(i) MWNT-7 deposition in the lung (lung burden)

Inhaled MWNT-7 fibers are excluded from the lung by clearance via the mucociliary escalator of the tracheobronchial epithelium of MWNT-7 itself, or MWNT-7 phagocytosed by macrophages adhering to the mucosa of the terminal bronchus. The fibers are also translocated from the lung to the bronchus-associated lymphoid tissue and mediastinal lymph nodes through intrapulmonary lymphatic channels, and to other organs, including the liver and kidney, hematogenously via the blood vessels of the alveolar wall. Deposition in the lung, also known as lung burden, is the result of the fibers remaining in the lung despite the removal of inhaled MWNT-7 by these mechanisms.

Lung burden is related to MWNT-7 exposure concentration and overall exposure duration. In the 2-week and 13-week studies conducted as preliminary experiments of the 2-year carcinogenicity study^{12, 13}, deposition of MWNT-7 in the lung increased with exposure concentration. In the 13-week and 2-year studies, we measured the amounts of MWNT-7 in the rat lung after 13 weeks and 104 weeks of exposure, respectively, using the measurement method described by Ohnishi et al.¹⁵ which uses benzo [ghi]perylene as a marker. In both studies, the amount of MWNT-7 deposited in the lung increased with increasing exposure concentration in both males and females, and the amount increased as the administration period became longer^{1, 13}. In the 2-year study, the size of the MWNT-7 fibers collected from the lungs was measured by SEM: MWNT-7 in the lungs had an average length of 5.8–5.9 μm. This result supports the fiber paradigm theory that fiber length is an important factor in fiber-induced carcinogenesis¹⁶.

Female rats required a higher amount of MWNT-7 for tumor development than male rats (Table 3). This is supported by a lower incidence of neoplastic and proliferative non-neoplastic lesions in females compared to males (Tables 1–3). The fact that the incidence of spontaneous lung tumors is higher in male F344 rats than in females¹⁷ suggests that the gender difference is most likely due to the basic susceptibility of male rats to lung carcinogenesis.

Table 3. Relation Between Incidences of Lung Tumors and MWNT-7 Lung Burden at Week 104.

(ii) Phagocytosis of MWNT-7 by macrophages in the lungs

A SEM image of an alveolar macrophage from a rat in the 2-year carcinogenicity study that was incompletely digested by enzymatic treatment is shown in Fig. 5A. MWNT-7 fibers were clearly attached to the cell membrane. A macrophage phagocytosed with MWNT-7 after removal of the cell membrane by enzymatic treatment is presented in Fig. 5B. Many MWNT-7 fibers were entwined in a spherical, coconut-like morphology. In this SEM image, we present a macrophage that is unable to degrade the material it has phagocytosed, resulting in oxidative stress.

Fig. 5.

Images of scanning electron microscope (SEM) and smear of alveolar macrophages phagocytizing MWNT-7. (A) Incomplete digestion of a macrophage. Some MWNT-7 fibers are seen on the surface. (B) Complete digestion image of a macrophage. Large cocoon-like mass of MWNT-7 fibers are observed. (Quotation from Particle and Fibre Toxicology 13: 53. 2016) (C) May-Giemsa staining finding of a smear specimen of bronchoalveolar lavage. An aggregated macrophage phagocytizing MWNT-7 fibers. (D) Polarizing microscope finding of a smear specimen: The fibers appear to shine brilliantly.

We also confirmed that alveolar macrophages phagocytosed MWNT-7 fibers by analyzing smear specimens of the bronchoalveolar lavage fluid of rats in the 2-year carcinogenicity study (Fig. 5C and 5D).

(iii) Translocation from the lung to the pleural and abdominal cavities

MWCNT reaches the subpleural tissue of mice after inhalation exposure¹⁸. Furthermore, inhaled MWCNT are transported to the parietal pleura¹⁹. These findings indicate that MWCNT reaches the pleural cavity, although they do not completely explain the route they undertake. In addition, the MWCNT fibers that reach the pleura are cleared through the pleural exits and lymphatic stomata in the parietal pleura. Long fibers that are unable to be cleared accumulate in the stomata, leading to inflammation²⁰. The most important concern is whether MWNT-7 which translocate in the pleural cavity, induces pleural mesothelioma in humans similar to asbestos fibers.

To investigate the transfer of inhaled MWNT-7 into the pleural and abdominal cavities in the 2-year study, we counted the number of fibers present in the pleural and abdominal lavage fluids using SEM. MWNT-7 fibers accumulated in the pleural and abdominal cavities in a concentration-dependent manner in both males and females. In the pleural cavity, the average width and length of the MWNT-7 fibers was 89.8 nm and 6.7 μm, respectively. The MWNT-7 fibers that accumulated in the abdominal cavity were longer than those that accumulated in the pleural cavity, with an average width of 90.0 nm and an average length 9.5 µm. Mild fibrous thickening of the pleura was observed in males but not in females, with a more pronounced thickening of the parietal pleura. The abdominal cavity showed a normal histology.

Pathogenesis of histopathological changes in the lung

The pathogenesis of lung tumor development in the 2-year MWNT-7 inhalation study and the key histopathological changes are summarized in Fig. 6. Inhaled MWNT-7 is phagocytosed by alveolar macrophages. Some macrophages are cleared from the alveolar duct through the terminal bronchi by the mucociliary escalator. Most macrophages that are not cleared accumulate in pulmonary alveoli. We also found that phagocytosed MWNT-7-enclosing macrophages migrated from the pulmonary alveoli into the parenchyma of the lungs, which contains a network of lymphatic vessels. They could migrate via lymphatic flow through the intrapulmonary lymphatic channels to the lymph nodes belonging to the lungs and mediastinal lymph nodes. Macrophages which phagocytosed MWNT-7 also spread from the lungs to other organs via the circulatory system and hematogenous translocation. Furthermore, macrophages which phagocytosed MWNT-7 remaining in the alveoli led to the development of alveolar granulomatous changes, fibrosis, and epithelial hyperplasia. Alveolar hyperplasia and bronchial hyperplasia are proliferative lesions that have the potential to develop into tumors, and bronchioloalveolar hyperplasia and atypical hyperplasia are presumed to be preneoplastic lesions.

Fig. 6.

The pathogenesis of lung tumor development in the 2-year MWNT-7 inhalation study in rats.

Genotoxicity of MWCNT

Genotoxicity studies provide important information on MWCNT toxicity, especially regarding the mechanisms underlying its carcinogenicity. MWCNT was negative in bacterial reverse mutation assays using Salmonella typhimurium TA98, TA100, TA1535, TA 1537, and Escherichia coli WP2uvrA in the presence and the absence of a metabolic activation system^{21, 22}. However, the results of the in vitro genotoxicity tests of fibers such as MWCNT should not be considered conclusive, as the in vitro effects of a fiber on bacteria could very well differ from its effects on tissues in vivo. In contrast to the results of the in vitro reverse mutation assays, gpt mutation frequencies were significantly increased in the lungs of MWCNT-treated gpt-delta transgenic mice, compared to non-treatment mice²³. Transversions were predominant, and G:C to C:G transversions were substantially increased by the MWCNT.

In human mesothelial (MeT-5A) cells, MWCNT-induced DNA damage was detected using the comet assay. However, it had no effect on bronchial epithelial (BEAS 2B) cells²⁴. The formation of 8-hydrooxy-2’-deoxyguanosine (8-OHdG), a marker of oxidative stress, was observed in the lungs of gpt-delta transgenic mice exposed to MWNT-7, and DNA damage in the lung was increased²⁵. These data indicate that reactive oxygen species (ROS) induced by MWNT-7 may play an important role in the carcinogenicity. This was confirmed in a co-culture system of murine lung resident cells (GDL1) and immune cells (RAW264.7) treated with MWNT-7²⁵.

Thus, MWNT-7 genotoxicity, particularly its mutagenicity, was positive in the lungs of gpt-delta transgenic mice exposed to MWNT-7²³. However, since the mutations induced by MWNT-7 are inflicted by generation of ROS, known genotoxic agents, it can be considered MWNT-7 possess a secondary genotoxic mechanism.

We compared the genotoxicity of MWNT-7 to that of chrysotile asbestos²⁶. MWNT-7 and chrysotile induced no structural chromosomal abnormalities in the chromosomal aberration assay using Chinese hamster lung (CHL/IU) cells. In addition, both types of fibers induced polyploid formation, characterized by numerical chromosome aberration; neither MWNT-7 nor chrysotile directly damaged chromosomes in vitro, but they did inhibit cell division. This phenomenon was also confirmed by SEM. As shown in Fig. 7, a large number of MWNT-7 were phagocytosed by CHL/IU cells. Two CHL/IU cells were firmly connected via a bridge of a bundle of MWNT-7 fibers, showing inhibition of cell division. In addition, the micronucleus test using CHL/IU cells was positive for MWNT-7 (unpublished data). Moreover, exposure of A549 monocultures to MWNT-7 induced a highly significant increase in nucleoplasmic bridges, indicating that cytokinesis was blocked, resulting in aneugenic/clastogenic effects²⁷. In another study studying on the influence of the shape and size of the MWNT-7 fibers for the physical interference of cytokinesis using CHL/IU cells, we examined chromosomal aberrations caused by seven types of MWCNTs²⁸. Four straight fibrous MWCNTs, including MWNT-7, induced polyploidy, whereas only one of the three tangled types of MWCNTs did so.

Fig. 7.

Scanning electron microscope (SEM) image of CHL/IU cells which phagocytosed MWNT-7. Two CHL/IU cells are firmly connected via a bridge of a bundle of MWNT-7 fibers.

The findings discussed above indicate that the genotoxicity of MWNT-7 is not caused by the interaction of the fibers with the DNA of the target cells. In the lungs of gpt-delta transgenic mice exposed to MWNT-7, the mutagenicity of MWNT-7 was caused by the generation of ROS, most likely by macrophages, as they attempt to destroy the MWNT-7 fibers, and not by the direct action of MWNT-7 on the DNA of the target cell. The aneugenic/clastogenic effects of MWNT-7 occur owing to the interaction of the fiber with cellular cytokinesis and not by interaction with the DNA of the target cell. In conclusion, MWNT-7 is genotoxic owing to secondary mechanisms.

Key characteristics of MWCNT inhalation exposure: adverse outcome pathway (AOP) development for assessment of fiber-induced carcinogenicity

The mode of action of non-genotoxic carcinogens needs to be clarified, and AOP should be developed. An AOP for fiber carcinogenicity links molecular events to the final adverse outcome. Furthermore, AOPs should be also developed for genotoxic carcinogens, as they could induce an adverse outcome via a non-genotoxic mechanism.

The hierarchy of biological responses to inhaled MWCNT and the resultant AOP is described below. This AOP illustrates the mode of action by which various inhaled fibrous substances induce toxicological effects in the lung and enables the identification of potentially carcinogenic fibers.

(i) Molecular level

As mentioned above, MWNT-7 was genotoxic both in vitro and in vivo; however, its in vivo genotoxic effects involve ROS generation. In addition, inflammatory cytokines such as interleukin (IL)‐1β and tumor necrosis factor (TNF‐α) were produced by RAW264.7 cells treated with MWNT-7²⁵. IL-1β was also produced by the human THP-1 macrophage cell line exposed to other high-aspect-ratio MWCNTs²⁹. These findings indicate that, at molecular level, MWCNT can induce the generation of genotoxic agents and alveolar inflammation.

(ii) Cellular level

Inefficient removal of inhaled MWCNT from the lung and the alveolar accumulation of macrophages enclosing phagocytosed MWNT-7 are reported. These findings indicate that MWCNTs that are not cleared from the lungs are phagocytosed by alveolar macrophages and induce the molecular events noted above.

(iii) Organ level

Macrophages that phagocytose MWNT-7 undergo either frustrated phagocytosis or frustrated degradation of engulfed fibers. This leads to the production of genotoxic substances and inflammatory cytokines. Subsequently, alveolar epithelial hyperplasia, inflammation with granulomatous changes, and alveolar fibrosis occur. These biological responses in rat lung epithelial hyperplasia, especially occurrence of bronchioloalveolar epithelial hyperplasia (Fig. 6), are considered to be closely linked to MWNT-7-induced lung carcinogenesis.

Thus, inefficient removal of inhaled fibers from the lung and interaction of the retained fibers with macrophages are key events in the AOP of fiber-induced carcinogenicity. Accordingly, inhaled fibers that are inefficiently removed from the lung, causing accumulation of frustrated macrophages in the alveoli, generation of ROS and inflammatory cytokines, and epithelial hyperplasia, are highly related to lung carcinogenesis in rats.

Carcinogenicity of Non-fibrous Particles

Background

Non-fibrous particles can also cause lung disease in humans. One such particle is indium tin oxide (ITO), a mixture of indium oxide (In₂O₃) and tin oxide (SnO₃). ITO particles have 0.1-1 μm size, are yellow-green, and are used in conductive films such as LCD, plasma display, smartphone, and solar panels.

A case of interstitial pneumonia in a 27-year-old, male worker engaged in the wet surface polishing process of ITO for approximately four years was reported in 2003³⁰. More recently, development of interstitial pneumonia was reported in other workers exposed to ITO³¹. The interstitial pneumonia was accompanied by pulmonary alveolar proteinosis. Nakano et al.³² examined the relationship between workers exposed to ITO and the health effects on the lungs and reported dose-dependent harmful effects of ITO exposure in the human lung.

However, no data exist on cancer occurrence in humans exposed to ITO. Therefore, to prevent occupational cancer in workers exposed to ITO, an experimental carcinogenicity study was conducted by the JBRC.

Carcinogenicity of ITO by inhalation exposure in rats

ITO 2-year carcinogenicity study in rats²

An OECD and GLP guideline-compliant 2-year carcinogenicity study of ITO (JX Nippon Mining & Metals Co., Ltd. Tokyo, Japan) in rats was undertaken by the JBRC, based on a 2-week and 13-week preliminary studies³³. The ITO powder was produced by grinding sintered ITO plates. The mean powder diameter was 3.5 μm with a 90% cumulative diameter of 8.9 μm and purity of 99.8% (90.06% IO and 9.74% SnO₂). Whole-body inhalation exposure was the method selected to perform the carcinogenicity study in rats. In this study, 6-week-old male and female F344 rats were used. The exposure concentration of ITO was 0 (control), 0.01, and 0.03 mg/m³ for 6 hr/day, 5 days/wk for 104 wks, and 0.1 mg/m³ for 6 hr/day, 5 days/wk for 26 wks. The highest exposure group, 0.1 mg/m³, received clean air for the remaining 78 wk. The exposure concentrations were controlled within 10% of the variation coefficient and were accurate within a 10% deviation from the target concentration. Amounts of indium in the lungs of rats exposed to the 0.1 mg/m³ concentration, were measured at week 26 and were similar in males and females (20.2 μg/g/lung in male and 21.1 μg/lung in female).

There were no significant differences in the survival rates between ITO-exposed groups and the respective control of males and females. Compared to the control, a slight but significant decrease in body weight was observed in the ITO-exposed male groups during the first half of the 104-wk exposure period, but not in the females. Irregular respiration was noted in several rats in the 0.03 mg/m³ male and female groups.

Exposure to ITO induced bronchioloalveolar carcinoma, total of carcinoma, and adenoma and/or carcinoma in a concentration-dependent manner in both male and female rats (Table 4, Fig. 8A). Bronchioloalveolar adenoma incidence was significantly increased in the 0.03 mg/m³ and 0.1 mg/m³ male groups and in the 0.1 mg/m³ female group compared to the controls. The incidences of bronchioloalveolar carcinomas in the 0.03 and 0.1 mg/m³-exposed male and female were significantly increased compared to the controls. Total of carcinomas were significantly increased even in the male 0.01 mg/m³group compared to the control group. Incidences of adenoma and/or carcinoma were significantly increased in all exposed groups of male and female compared to the controls. There were no clear differences in lung tumor induction between males and females. No ITO-related tumors developed in any other tissue, including the pleural mesothelium, in either the male or female rats.

Table 4. Neoplastic Lesions of the Lung in Rats Exposed to Indium Tin Oxide (ITO) by Inhalation for Two Years.

Fig. 8.

Histopathological changes in the lung of rats exposed to ITO: (A) Bronchioloalveolar carcinoma in the lung of the male rat exposed to ITO at 0.03 mg/m³. Carcinoma cells are accompanied by severe fibrous connective tissue. Bar, 200 µm; H&E staining. (B) Bronchioloalveolar hyperplasia in the lung of a male rat exposed to 0.1 mg/m³ ITO. Bar, 200 µm. H&E staining. (C) Particle-phagocytizing alveolar macrophage in the lung of a male rat exposed to ITO at 100 mg/m³ for 2 weeks. Bar, 30 µm; H&E staining. (D) Alveolar proteinases in the lung of a male rat exposed to ITO at 0.1 mg/m³ for 2 weeks. Bar, 30 µm; H&E staining (Fig. 8A and 8B: Quotation from J Occup Health 53: 175–187. 2011).

The non-neoplastic lesions observed in the lungs are listed in Table 5. Epithelial hyperplasia was observed in the lungs of both male and female rats exposed to ITO (Fig. 8B). Compared to control, bronchioloalveolar hyperplasia, which is a preneoplastic lesion, was significantly increased in the male groups exposed to concentrations ≥0.03 mg/m³ and in the female groups exposed to concentrations ≥0.01 mg/m³. Most cases of bronchioloalveolar hyperplasia were adjacent to alveolar wall fibrosis. A few cases of atypical hyperplasia were also observed in male and female rats exposed to ITO. A high incidence of alveolar hyperplasia was observed in all the ITO-exposed groups of male and female compared to each control. In addition, high incidences of alveolar macrophage accumulation (Fig. 8C), alveolar proteinosis (Fig. 8D), and fibrosis of the alveolar wall were observed in all the exposure groups of male and female compared to each control.

Table 5. Hyperplastic Epithelial and Non-epithelial Lesions of the Lung in Rats Exposed to Indium Tin Oxide (ITO) by Inhalation for Two Years.

Pathogenesis of histopathological changes in the lung

The histopathological changes associated with the pathogenesis of lung tumor development induced by inhalation exposure of rats to ITO are very similar to those induced by MWNT-7. Both types of particles cause accumulation of alveolar macrophages, fibrosis of the alveolar wall, and epithelial hyperplasia. The primary difference is that while granulomatous changes were observed in the lungs of MWNT-7-exposed rats, alveolar proteinosis developed in the lungs of ITO-exposed rats.

Genotoxicity of ITO

The results of the Ames test using S. typhimurium strains TA98 and TA100, were negative for ITO, both in the absence and presence of the metabolic activation S9 mixture³⁴. However, as mentioned above, for particles such as ITO the results of the in vitro Ames test should not be considered conclusive, as the in vitro effects of an ITO particle on bacteria could be different from its effects on cells and tissues.

No data exist on the in vivo genotoxicity of ITO in transgenic rats or mice. However, ROS production was assessed using electron paramagnetic resonance spectrometry in an acellular system. Carbon-centered radicals (COO-) and Fenton-like activity were detected in the presence of ITO particles, indicating ROS production³⁵ and genotoxic potential. Tabei et al.³⁶ reported ITO nanoparticles induced oxidative stress and DNA damage in human lung adenocarcinoma A549 cells, by increasing the intracellular levels of ROS, confirming its possible genotoxicity. In another study, Tabei et al.³⁷ reported that in A549 cells exposed to ITO nanoparticles, ITO was taken up by the cells and solubilized by acidic lysosomes, releasing indium and tin ions. The intracellular accumulation of indium ions, but not that of tin ions, induces extensive oxidative stress and DNA damage. Exposure to ITO also significantly increased the mitotic index and total chromosomal aberrations in the root cells of Allium cepa in the Allium test, and a significant increase in DNA damage was also observed using the Comet assay³⁸. These results indicate that ITO exhibits genotoxic activity, suggesting secondary mechanism.

In addition to the results discussed above, other findings suggest that ITO is a secondary genotoxic substance. In a micronucleus test using human peripheral blood lymphocytes isolated from the blood of healthy volunteers, only the highest concentration of ITO (750 µg/mL) caused a significant increase in the formation of micronuclei (MN)³⁴. Furthermore, MN formation using this concentration of ITO was much lower than MN formation caused by 0.2 µg/mL mitomycin C, a known primary genotoxin, suggesting that there was relatively little uptake and/or solubilization of ITO particles by normal lymphocytes. In addition, ITO particles induced an increased frequency of micronuclei in type II pneumocytes in vivo (rats), but not in vitro, in lung epithelial cells originating from type II lung epithelial cells³⁵. Thus, while ITO uptake by the lung cancer cell line A549 results in solubilization of ITO, production of ROS, and generation of DNA damage, ITO showed little activity in isolated normal cells. Taken together with the ability of ITO to induce micronuclei formation in vivo, these results suggest the existence of a secondary mechanism for ITO genotoxicity.

Key characteristics of ITO inhalation exposure: generation of AOP for assessment of particle-induced carcinogenicity

The generation of an AOP for ITO-induced carcinogenicity, taken together with the AOP for MWCNT-induced carcinogenicity, is a good basis for assessing the mode of action of secondary genotoxic carcinogens. The hierarchy of the biological responses to inhaled ITO and the resultant responses are described below. This AOP illustrates the mode of action by which various inhaled particles induce toxicological effects in the lungs and enables the identification of potentially carcinogenic particles.

(i) Molecular levels

As mentioned above, ITO is genotoxic both in vitro and in vivo, and this genotoxicity involves ROS production. In ITO-exposed workers, increases in 8-OHdG levels in urine and peripheral white blood cells were observed^{39, 40}. In addition, increased DNA damage was observed in the whole blood of ITO-exposed workers, as measured by the comet assay³⁹.

RAW 264.7 mouse macrophage and BEAS-2B human bronchial epithelial cells exposed to sintered ITO produced cytokines, such as IL-1β, IL-6, TNFα, and IL-8, via nuclear factor kappa B pathway, suggesting inhaled ITO particles initiate and propagate inflammation⁴¹. In addition, IL-1β, which is released from macrophages stimulated with ITO, is able to induce epithelial-mesenchymal transition progression in human alveolar epithelial cells, A549 cells, thereby potentially triggering the genesis and development of pulmonary fibrosis⁴².

(ii) Cellular level

The accumulation of alveolar macrophages is a fundamental phenomenon in the inhalation toxicity of both fibrous and non-fibrous particles. ITO is readily phagocytosed by alveolar macrophages. Gwinn et al.^{43, 44} reported the uptake of ITO by macrophages (RAW 264.7 cells) and lung-derived epithelial cells in vitro. In macrophages, phagolysosomal acidification results in the solubilization of ITO and the release of cytotoxic indium ions into the extra-cellular medium by dying cells. Indium ions were not released into the extracellular medium by epithelial cells, supporting the hypothesis of a secondary mechanism underlying ITO genotoxicity. Thus, macrophage uptake of ITO and their solubilization by phagolysosomal acidification are required for the occurrence of cytotoxic effects.

(iii) Organ level

Pulmonary alveolar proteinosis, bronchioloalveolar hyperplasia, alveolar hyperplasia, and fibrosis of the alveolar wall were observed in the lungs of ITO-exposed rats. Proteinosis may be an important trigger of lung carcinogenesis. Epithelial hyperplasia is also closely linked to ITO-induced carcinogenesis in rat lungs.

Thus, macrophage uptake of ITO and their solubilization by phagolysosomal acidification, followed by macrophage death and release of indium ions into the extracellular space become the key events in the AOP of ITO carcinogenicity. Accordingly, inhaled particles cause accumulation and death of macrophages in the alveoli, pulmonary alveolar proteinosis (likely related to impaired macrophage function), generation of inflammatory cytokines, and epithelial hyperplasia. These events are highly associated with ITO-induced lung carcinogenesis in rats.

Concept of a Threshold Dose for Risk Assessment of Carcinogens: Low-dose Carcinogenicity of Genotoxic (Mutagenic) Carcinogens

Most carcinogens are evaluated in 2-year carcinogenicity studies using high-dose levels based on the maximally tolerated dose in rats and mice. Carcinogenic curves obtained from carcinogenicity studies are used to assess carcinogenic risk in humans. Importantly, the curves were based on high to low dose extrapolation to obtain acceptable data for human exposure levels.

Carcinogens are generally classified into genotoxic and non-genotoxic types. For quantitative risk assessment to humans, the theory that genotoxic carcinogens do not have a threshold dose is applied (non-threshold concept), because the effect of a small number of mutations on cancer induction is difficult to evaluate: since cells acquire mutations throughout the life, it is theoretically possible for a single additional mutation to induce tumorigenesis; by using large enough populations, this effect can be quantified. However, we investigated the low-dose carcinogenicity of 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx), which is a known genotoxic, and determined that there is a carcinogenic threshold, at least a practical threshold^{45, 46}. Moreover, the International Workshop on Genotoxicity Testing (IWGT)⁴⁷ analyzed the dose-genotoxicity/carcinogenicity relationships of MeIQx using BMDL₁₀ (benchmark dose lower bound; confidence limit value on the safe side of 95% lower confidence limit) in BMD (benchmark dose). This analysis showed a quantitative dose-response relationship, the increase of the dose leading progressively to the induction of DNA adducts << mutations < preneoplastic foci, glutathione S-transferase positive (GST-P) foci << cancer. Therefore, these data indicate that there is a practical carcinogenic threshold for MeIQx, and that such a threshold might very likely exist for all genotoxic carcinogens. This study proposes an important recommendation regarding actual risk assessment and management from the perspective of occupational health: the determination of acceptable human exposure levels to carcinogens and exposure-related risk of genetic damage.

Occupational Exposure Limits for MWCNTs and ITO

For numerous substances, regulatory agencies have proposed a non-threshold theory for risk assessment. However, this is not realistic from the perspective of occupational health. Bolt HM⁴⁸ reported that in the development of occupational exposure limits (OELs), the Scientific Committee on Occupational Exposure Limits (SCOEL) of the European Union recommended practical thresholds for genotoxic carcinogens. SCOEL stresses that even if a substance is genotoxic, if its genotoxicity is due to a secondary mechanism, there is a carcinogenic threshold. The threshold levels for occupational exposure to carcinogens are very important. This is a realistic response, and a no-observed-adverse-effect level (NOAEL) can be obtained for these carcinogens. As mentioned above, the results⁴⁶ of research on setting practical thresholds for the carcinogenicity of genotoxic carcinogens strongly support the development of OELs. This allows NOAELs to be applied rather than BMDs, which are currently used for non-threshold carcinogens.

The International Agency for Research on Cancer (IARC) evaluated MWNT-7 as possibly carcinogenic to humans (Group 2B)⁴⁹. In our carcinogenicity study¹, the NOAEL of MWNT-7 was determined to be 0.02 mg/m³, based on lung carcinoma induction in male rats. Using the NOAEL-based approach, the occupational exposure limit of MWNT-7 is calculated to be 0.0006 mg/m³ (formula: 0.02 mg/m³ × 1/25 (uncertainty factor) × 6/8 hr/day (correction of working hours).

IARC also evaluated ITO as being possibly carcinogenic to humans (Group 2B)⁵⁰. For the occupational exposure limit of ITO, the Ministry of Health, Labor and Welfare of Japan adopted a lowest-observed-adverse-effect level (LOAEL) of 0.01 mg/m³, which is the LOAEL based on the carcinogenic threshold in our 2-year ITO carcinogenicity study in rats. The allowable exposure concentration of indium was consequently set to 0.0003 mg/m^{3 51}. The Japan Society for Occupational Health set an allowable biological tolerance of serum ionic indium of 3 μg/L⁵².

Liquid Inhalation Carcinogenicity Studies in Rats

In liquid inhalation carcinogenicity studies, airflow containing the liquid substance at target concentrations was generated according to a vaporization technique. Thirty-one vaporized liquid substances were assessed by JBRC in inhalation carcinogenicity studies using rats. A total of 19 substances were positive for carcinogenicity, but only one substance targeted the lungs. In contrast, 7 substances targeted the nasal cavity.

Conclusion

In the 2-year inhalation carcinogenicity studies of MWNT-7 and ITO in rats, alveolar macrophage phagocytosis of fibrous or particle substances triggered lung carcinogenesis. Phagocytosis of MWNT-7 resulted in frustrated degradation of the fibers and production of ROS. Phagocytosis of ITO resulted in the solubilization of ITO and the release of cytotoxic indium ions. Toxicity to the alveolar epithelium by frustrated macrophages or melted macrophage content contributed significantly to hyperplasia of the alveolar epithelium, resulting in the induction of lung carcinomas.

Regarding the genotoxicity of MWNT-7 and ITO, both substances are secondary genotoxins, and NOAELs can be applied to them rather than the BMDs used for non-threshold carcinogens. Therefore, the occupational exposure limit values based on the existence of a carcinogenic threshold are reasonable to apply.

Carcinogenicity studies are collaborative efforts; in particular, inhalation studies require close collaboration between physical researchers (inhalation toxicologists) and biological researchers (toxicology pathologists). The collaborative engagement of these researchers has led to the production of accurate carcinogenic data.

Disclosure of Potential Conflicts of Interest

This review paper was written mainly based on the results of contracted carcinogenicity studies (refs. ¹ & ²) conducted by the JBRC. The carcinogenicity study of MWCNT was supported by the Ministry of Health, Labour and Welfare of Japan, and that of ITO was supported in collaboration with JX Nippon Mining & Metal, Corp. and the other nine companies. The authors declare no potential conflicts of interest.