A 104-week pulmonary toxicity assessment of long and short single-wall carbon nanotubes after a single intratracheal instillation in rats

Abstract

We compared long-term pulmonary toxicities after a single intratracheal instillation of two types of dispersed single-wall carbon nanotubes (SWCNTs), namely, those with relatively long or short linear shapes with average lengths of 8.6 and 0.55 µm, respectively. Both types of SWCNTs were instilled intratracheally in male F344 rats at 0.2 or 1.0 mg/kg (long SWCNTs) or 1.0 mg/kg (short SWCNTs). Pulmonary responses were characterized at 26, 52 and 104 weeks after a single instillation. Inflammatory changes, test substance deposition, test substance engulfment by macrophages, and alveolar wall fibrosis were observed in the lungs of almost all test rats at 52 and 104 weeks after short nanotube instillation. The incidences of these changes were much lower in the long nanotube-treated groups. In almost all rats of the long nanotube-treated groups, fibrosis and epithelium loss in the terminal bronchiole with test substance deposition were observed. These bronchiolar changes were not observed after administering short nanotubes. Both bronchiolo-alveolar adenoma and carcinoma were found in the negative-control group, the high-dose long-nanotube group, and the short-nanotube group at 104 weeks post-instillation, although the incidences were not statistically different. The genotoxicity of the SWCNTs was also evaluated by performing in vivo comet assays with lung cells obtained 26 weeks post-instillation. No significant changes in the percent tail deoxyribonucleic acid were found in any group. These findings suggested that most long SWCNTs were deposited at the terminal bronchioles and that a considerable amount of short SWCNTs reached the alveolus, resulting in chronic inflammatory responses, but no genotoxicity in the lungs.

Keywords:

• Intratracheal instillation
• length effect
• long-term study
• lung toxicity/inflammation
• nanomaterial
• single-walled carbon nanotube

Introduction

The rapidly developing field of nanotechnology, which involves the creation of materials with size-dependent properties, has resulted in a recent increase in nanomaterial (NM) exposure. Because of their unique physico-chemical properties, such as increased strength, reactivity, or conductivity, engineered NMs (ENMs) can be used in a wide range of products to improve their performance and provide benefits to the consumer (Thomas, 2014). Human exposure to ENMs can occur at any stage during ENM synthesis/manufacture, use and disposal (Jiménez et al., 2014). Occupational-exposure levels are potentially greater than consumer- or environmental-exposure levels because of the potentially larger quantity of NMs handled in their pure form in an industrial setting. Introducing these novel materials into work environments and consumer products necessitates appropriate safety evaluations and a clearer understanding of their potential impact on human health.

More than a quarter of a century has passed since the discovery of carbon nanotubes (CNTs) (Iijima, 1991), and various types of CNTs have been developed as industrial materials. Workers may be at risk for exposure to nano-sized particles through their respiratory organs during the manufacture, handling, and cleanup of ENMs (Nakanishi et al., 2015; NIOSH, 2013). Airway exposure is most relevant for ENMs, as inhalation represents the main route of exposure in an occupational setting. Thus, investigating the respiratory toxicity of CNTs is of high priority.

Many studies have been conducted to elucidate the pulmonary toxicity of multi- and single-walled CNTs (MWCNTs and SWCNTs, respectively) using in vivo rodent models (Fujita et al., 2015a, 2015b, 2016; Lam et al., 2004; Ma-Hock et al., 2009; Morimoto et al., 2012a, 2012b; Muller et al., 2009; Pauluhn, 2010; Shvedova et al., 2014; Warheit et al., 2004). Some reports have presented evidence that 1 type of MWCNT was carcinogenic to rodents (Sakamoto et al., 2009; Sargent et al., 2014; Takagi et al., 2008, 2012), resulting in its classification as a Group 2B agent by the World Health Organization (WHO)/the International Agency for Research on Cancer (IARC) (Grosse et al., 2014). However, the pulmonary toxicity of SWCNTs during a long-term observation period remains unclear. Here, we performed a 104-week toxicity study for SWCNTs, using a single intratracheal instillation in rats.

The physical and chemical properties of NMs, including their size, shape, charge, agglomeration status and metal impurity, affect their toxicity (Donaldson et al., 2006; Johnston et al., 2010). Length is a critical factor underlying the potential toxicity of fibrous NMs (Hamilton et al., 2009; Ye et al., 1999), and fiber length is accepted as the major contributing factor to fiber pathogenicity (Schinwald et al., 2012). Therefore, we prepared a SWCNT solution containing with long fibers of a length thought to pose a potential asbestos-like hazard, based on the fiber paradigm (Donaldson et al., 2010; Murphy et al., 2012; Poland et al., 2008; Schinwald et al., 2012). The primary objective of this study was to investigate the long-term pulmonary toxicity of SWCNTs to the lungs and/or pleural tissues in rats. This is the first report to describe long-term pulmonary toxicities after a single intratracheal instillation using such long, linear-shaped SWCNTs.

Materials and methods

Test materials

SWCNTs, synthesized using a catalytic chemical vapor-deposition method, were obtained from Nikkiso Co., Ltd. (Tokyo, Japan). A detailed description of the characterization of bulk SWCNTs was provided in our previous study (Morimoto et al., 2012b). The geometric mean diameter of the tubes was 1.8 nm, and the specific surface area of the bulk SWCNTs was 878 m²/g, as determined using a Brunauer–Emmett–Teller method (Autosorb-1-C; Quantachrome Instruments; Boynton Beach, FL).

Preparation of SWCNT suspensions

Bulk SWCNTs were dispersed into 10-fold diluted phosphate-buffered saline (PBS) containing 1% salmon-serum deoxyribonucleic acid (DNA; Wako Pure Chemical Industries Ltd., Osaka, Japan), as a dispersant. The DNA–dispersant solution was prepared by ultrasonic homogenization (60% output Sonifier 250D, Branson) for 9 h to reduce its viscosity, followed by centrifugal separation (134,000×g, 4 min). A food processor (IFM-800, Iwatani, Tokyo, Japan) was used for the dispersion process.

A “long” SWCNT solution was obtained by purification with centrifugation (10,000×g, 10 min), followed by adsorption column chromatography on a silica-gel column (Wakosil C-200, Wako Pure Chemical Industries Ltd., Osaka, Japan), using PBS as the mobile phase. The retained fraction was eluted with distilled water.

We also prepared a test solution containing relatively short SWCNT fibers with an average length of ∼1 μm. Most in vivo studies of CNTs have involved test materials of ∼1 μm or with shorter lengths because of the difficulty in obtaining well-dispersed CNT samples with a length larger than a few microns. The “short” tube solution was obtained from a raw CNT solution by ultrasonic homogenization for 5 h, followed by purification via centrifugation to eliminate small metal fragments derived from the homogenizer tip.

The SWCNT concentration in the suspension was determined by measuring the optical absorbance at 800 nm, and then three test solutions were prepared for animal testing: 0.2 and 1 mg/ml long CNTs (SWCNT-1 and SWCNT-2, respectively) and 1 mg/ml short CNTs (SWCNT-3).

The physical dimensions of the long and short SWCNTs were evaluated by atomic force microscopy (AFM, S-image, SII, Chiba, Japan). The dispersed liquid samples were dripped onto a hydrophilic treated silicon substrate. CNTs were deposited on a silicon surface, lightly rinsed in purified water, blow-dried, and used for measurements. CNT lengths were evaluated from the obtained AFM images by using publically available software (UTHSCSA ImageTool, UT Health Science Center San Antonio).

Determination of metal impurities in SWCNT suspensions

The metal contents in the dispersed SWCNT solutions were determined by inductively coupled plasma-mass spectrometry (ICP-MS) and inductively coupled plasma-emission spectrometry (ICP-ES). A 0.5 mg portion of the sample solution (weighed in a quart-sized melting pot) was heated in a mixture of sulfuric acid and nitric acid to decompose the CNTs. The residue was re-dissolved in 10 ml of dilute nitric acid and analyzed for metal impurities. The Co, Ni, Ti, V, Cr, Mn and Al concentrations were estimated by ICP-MS, whereas the Fe concentration was so high that ICP-ES was necessary for the analysis.

Animals

All animal tests were performed at the Public Interest Incorporated Foundation BioSafety Research Center (BSRC, Iwata, Japan). The rats were housed in an animal facility with positive-pressure, air-conditioned units (22.2–23.2 °C, 43.6–68.8% relative humidity), and a 12-h/12-h light/dark cycle. A standard rodent pellet diet (CRF-1, Oriental Yeast Co., Ltd., Tokyo, Japan) and drinking water were provided ad libitum. The pellet diet was used for the first 37 weeks and was thereafter changed to the powder diet because of throat occlusion and death from suffocation. After changing the diet, no throat occlusion occurred.

After a 1-week acclimation, 245 9-week-old male F344/DuCrlCrlj male rats (Charles River Laboratories Japan, Yokohama, Japan) were divided into five groups (Table 1). Four groups consisted of 60 animals each, including a negative-control group (vehicle-administered group) and three SWCNT-administered groups. A positive-control group, consisting of five animals, was used for the comet assay. The average body weight (bw) of the rats before instillation treatment was approximately 195 ± 9 g (minimum 178 g, maximum 212 g). The numbers of animals in each group and the sacrifice time points are summarized in Table 1.

Table 1. Number of rats autopsied at each time point.

Sample	Week 26	Week 52	Weeks 104–105	Total number of rats
Negative control (PBS containing 1% salmon DNA, 1 ml/kg bw)	5	5	50	60
SWCNT-1 (long tubes, 0.2 mg/[ml·kg bw])	5	5	50	60
SWCNT-2 (long tubes, 1 mg/[ml·kg bw])	5	5	50	60
SWCNT-3 (short tubes, 1 mg/[ml·kg bw])	5	5	50	60
Positive control for the comet assay (Ethyl methanesulfonate, 5 mg/[ml·kg bw])	5			5

All procedures and animal care were approved by the Institutional Animal Care and Use Committee of AIST and BSRC, in compliance with the Guideline for Animal Experimentation (1987), the Law Concerning the Protection and Control of Animals (1973) and the Standards Relating to the Care and Management of Experimental Animals (1980).

Experimental design

The SWCNT suspensions were lightly sonicated periodically during the administrations to minimize aggregation. For the negative-control and SWCNT groups, the test solutions were intratracheally instilled once in rats in 0.1 ml vehicle per 100 g bw. The dosages administered were 0.2 and 1.0 mg/kg bw for the SWCNT-1 and SWCNT-2 groups, respectively, and 1.0 mg/kg bw for SWCNT-2 group. The maximum dose of 1.0 mg/kg bw was selected for our previous study (Ema et al., 2017), which induced lung inflammation for up to 26 weeks after single intratracheal instillation. The negative-control group was administered 1% DNA in 10-fold diluted PBS. All rats used were anesthetized by isoflurane inhalation. A metal tube was used for the instillation. Ethyl methanesulfonate (EMS) at a dosage of 500 mg/kg bw was used as a positive control for the comet assay. A polytetrafluoroethylene tube was used, and 0.1 ml of 50 mg/ml EMS aqueous solution was orally administrated per 100 g bw.

After instillation, the viability and general conditions of the rats were observed once daily until they were sacrificed. The bw of each rat was measured before instillation, once a week until 27 weeks post-exposure, and once during odd-numbered weeks beginning at 29 weeks post-exposure. Moribund or dead rats were immediately evaluated in terms of their general conditions and bw after either condition was observed. The general conditions and bws were also measured in rats on the day that they were scheduled to be sacrificed.

Animals in the negative-control and SWCNT groups were sacrificed by exsanguination from the abdominal aorta under deep anesthesia induced by isoflurane at 26, 52 or 104–105 weeks post-exposure. EMS administration was conducted 3 h prior to sacrifice at 26 weeks post-instillation. For humane reasons, moribund rats were euthanized by exsanguination under deep anesthesia with isoflurane, after which they were autopsied.

Histopathological examinations

The external surfaces of the rats and all organs were grossly examined. The lungs were weighed. Organs (including the brain, lungs, liver, spleen and kidneys) were fixed in 10% neutral buffered solution at each time point, embedded in paraffin, sectioned into 3 µm pieces, and stained with hematoxylin and eosin (HE) or Masson’s trichrome stain for histopathological examination.

Comet assay

The comet assay was conducted using the right lobes of all rats at 26 weeks after instillation. The comet assay was conducted in accordance with the standard protocol “International Validation of the In Vivo Rodent Alkaline Comet Assay for the Detection of Genotoxic Carcinogens” issued by the Japanese Center for the Validation of Alternative Methods (JaCVAM). The procedure was described in detail in our previous report (Ema et al., 2013a).

Transmission electron microscope (TEM) observations of the lungs

Lung sections and mediastinal lymph nodes in rats intratracheally instilled with the SWCNTs, were fixed using 2.5% (v/v) glutaraldehyde for 2 h at 4 °C and 1% osmium oxide solution for 2 h at 4 °C, dehydrated in ethanol, and embedded in a commercially available epoxy resin (TAAB Laboratories Equipment Ltd., Reading, England). Samples were transferred to fresh resin in capsules and polymerized at 60 °C for 48 h. A TEM system (H-7000; Hitachi, Tokyo, Japan) was used at 75 kV to observe the tissue distribution of SWCNTs.

Statistical analysis

Data from the SWCNT-exposed groups were compared to the corresponding negative-control data. Bw, bw gain and absolute and relative organ weights were analyzed by the Tukey–Kramer test with the Shaffer–Royen procedure. Incidences of non-neoplastic lesions were analyzed by Fisher’s exact test. The log-rank test was used to detect statistically significant differences in survival rates at the end of observation period. Incidences of neoplastic lesions were analyzed by Peto’s log-rank test. The data obtained from the positive-control and negative-control groups were analyzed by the Aspen–Welch t-test. Two-tailed tests were used for all statistical analyses except for Peto’s log-rank test. p values less than .05 were considered statistically significant.

Results

Characterization of SWCNT suspensions

Table 2 summarizes the physical dimensions of SWCNTs in test solutions and the dosages administered to the animals. The average lengths and standard deviations were determined by measuring 222 short SWCNTs from 1 AFM image and 116 long SWCNTs from 4 AFM images. The heights (diameters) of SWCNTs were also evaluated using AFM section-image analysis. The average heights and deviations were obtained by analyzing 101 short SWCNTs from 1 AFM image and 105 long SWCNTs from 2 AFM images. The length distributions of the short and long SWCNTs obtained from the AFM analyzes are depicted in Figure 1. Well-dispersed suspensions were obtained using salmon-serum DNA as a dispersant and were purified by column chromatography.

Figure 1. Length distribution of short and long SWCNTs in dispersed solutions.

Table 2. Physical dimensions of SWCNTs in working solutions and the dosages administered to animals.

Sample	Average length (μm)	Average width (nm)	Dose (mg/kg)
SWCNT-1 (long tubes)	8.6 ± 4.3	2.4 ± 2.0	0.2
SWCNT-2 (long tubes)	8.6 ± 4.3	2.4 ± 2.0	1.0
SWCNT-3 (short tubes)	0.55 ± 0.36	1.4 ± 0.7	1.0

Both short and long SWCNTs were well dispersed in the test solutions, as indicated in the TEM images (Figure 2). As shown in Figure 1, 89% of the short fibers were ≤1 μm in length and their length distribution was similar to that described in our previous report (Ema et al., 2017). Fifty-two percent of the long fibers were ≥10 μm in length. The value of 52% was substantially larger than the 16% value for the test sample used in our previous study.

Figure 2. TEM images of CNTs in a working solution. Long SWCNT-2 (a) and short SWCNT-3 (b).

The length distributions of the test solutions appeared to be maintained during the injection procedure. The average length and standard deviation of the short SWCNT solution ejected from the syringe used for the instillation were 0.54 and 0.58 μm, respectively. The average length and standard deviation of the long SWCNT solution ejected from the syringe were 10.4 and 4.8 μm, respectively. These average lengths were not significantly different from the corresponding values for the test solutions after the preparation.

The metal contents in the CNT suspensions and medium solution were determined, as summarized in Table 3.

Table 3. Concentration of metal impurities in the working solutions.

Elements	Short SWCNT solution (µg/g)	Long SWCNT solution (µg/g)	DNA/PBS (µg/g)
Fe	27	20	0.03
Co	0.0014	0.010	<0.0005*
Ni	0.060	0.54	<0.002*
Ti	1.0	0.73	0.23
V	0.044	0.032	0.014
Cr	0.12	1.7	–
Mn	0.018	0.12	–
Al	0.37	0.50	–

–: Not measured. *Under the limit of quantification. The values shown were obtained by averaging three measurements.

General conditions, survival rates, bws and lung weights

Neither SWCNT-related death, nor clinical signs were observed in any SWCNT-instilled groups throughout the experimental period. The survival rates were 82, 86, 82 and 74% in the control, SWCNT-1, SWCNT-2 and SWCNT-3 groups, respectively. No significant differences were found between the rates in the control and exposure groups.

Figure 3 shows changes in bw occurring in each group over time. Compared to the negative-control group, the bws in the SWCNT-2-administered group was significantly lower at 2 weeks after the administration. No significant differences in bws were observed between the control group and each SWCNT-administered group during the remainder of the experimental period.

Figure 3. Bws of rats after a single intratracheal instillation of SWCNTs. Sixty animals were included in all groups at the beginning of the experiment. The number of rats remaining at 104 weeks was 41, 43, 41 and 37 in the DNA/PBS, SWCNT-1, SWCNT-2 and SWCNT-3 groups, respectively. The error bars indicate standard deviations. *Denotes significant differences (p ≤ .05, Fisher’s exact test) relative to the DNA/PBS group.

The rat lung weights after a single intratracheal instillation of SWCNTs are summarized in Figure 4. Lung weights in the SWCNT-2 and SWCNT-3 groups at 104–105 weeks post-administration showed significant increases compared to those of the negative-control group, suggesting that lung inflammation occurred in the CNT-administered groups.

Figure 4. Normalized lung weights of rats to the body weights after a single intratracheal instillation of SWCNTs. The error bars indicate standard deviations. * and ** denote significant differences p ≤ .05 and p ≤ .01 (Fisher’s exact test) from the DNA/PBS group, respectively.

Macroscopic findings

Black, brown or gray patches were observed in all lungs of the SWCNT-treated groups, suggesting that the test material reached most parts of the lungs. These patches were still present at 104–105 weeks after the instillation (Figure 5). Only rats administered short SWCNTs showed black- or gray-colored lymph nodes in the lung, mediastinum and bilateral thymus, while the other rats did not show such colored lymph nodes. The colored areas of the lungs tended to be larger in the SWCNT-3 group than in the SWCNT-2 group.

Figure 5. Necropsy findings for the lungs in the SWCNT-treated groups. DNA/PBS (a), SWCNT-1 (c), SWCNT-2 (e) and SWCNT-3 (g) at 26 weeks after the instillation. DNA/PBS (b), SWCNT-1 (d), SWCNT-2 (f) and SWCNT-3 (h) at 104–105 weeks after the instillation.

Microscopic findings

Inflammatory changes, test substance deposition, test substance engulfment by macrophages, alveolar epithelialization, cell debris formation and alveolar wall fibrosis were observed in the lungs of almost all animals at 52 and 104 weeks after short SWCNT instillation (Table 4). Figure 6 displays rat lung tissues that were stained with HE 104 weeks after a single intratracheal instillation of SWCNTs. These findings were less pronounced in the long tube-treated groups, although macrophages engulfing the test substance were also found. Almost all animals in the long tube-treated groups showed fibrosis in the terminal bronchioles with test substance deposition and lost epithelium in the terminal bronchioles. These bronchiolar changes were not observed after treatment with the short SWCNTs.

Figure 6. Histopathology of the rat lungs (HE staining) 104–105 weeks after a single intratracheal instillation of SWCNTs. Arrows indicate fibrosis at terminal bronchiole with deposit of the test substance in SWCNT-2 group (a), mineralization at terminal bronchiole in SWCNT-2 group (b), macrophage engulfing the test substance in SWCNT-3 group (c) and fibrosis at alveolar wall in SWCNT-3 group (d).

Table 4. Test substance-related non-neoplastic findings.

Test substance	DNA/PBS	SWCNT-1	SWCNT-2	SWCNT-3
Weeks	26 52 104–105	26 52 104–105	26 52 104–105	26 52 104–105
Lungs
(No. animals examined)	(5) (5) (50)	(5) (5) (50)	(5) (5) (50)	(5) (5) (50)
Inflammatory change, focal	0 0 7	0 0 3	0 0 0*	0 0 0*
Inflammatory change with deposited test substance	0 0 0	0 0 3	0 0 4	5* 5* 50*
Macrophage engulfing test substance	0 0 0	5*4*19*	0 0 6*	5*5*50*
Fibrosis, terminal bronchioles with deposited test substance	0 0 0	5*5*49*	5* 5*  49*	0 0 0
Loss epithelium, terminal bronchioles	0 0 0	5*5*47*	4* 5* 45*	0 0 0
Mineralization	0 0 0	0 0 0	0 1 27*	0 0 0
Alveolar epithelialization	0 0 0	0 0 0	0 0 0	5* 5* 45*
Cell debris	0 0 0	0 0 0	0 0 0	5* 5* 50*
Fibrosis, alveolar wall	0 0 0	0 0 0	0 0 0	5* 5* 47*
Lymph nodes
(No. animals examined)	(0) (0) (15)	(0) (0) (17)	(0) (0) (21)	(2 3 32)
Aggregation, macrophage with deposited test substance	0	0	0	1 3 28*

Significantly different from the DNA/PBS group *p ≤ .05 (Fisher’s exact test).

Neoplastic findings

Bronchiolo-alveolar adenoma and carcinoma were found in 2, 1 and 9 cases in the negative-control, the high-dose long SWCNT, and the short SWCNT groups at 104 weeks after instillation, respectively, although these numbers were statistically nonsignificant (Tables 5 and 6 and Figure 7). One example of mesothelioma was observed in the peritoneum of the SWCNT-1- and the SWCNT-2-administered groups, but these numbers showed no significant difference compared to the negative-control group.

Figure 7. A bronchiolo-alveolar carcinoma 104 weeks after instillation with short SWCNT-3. Necropsy photograph for the lung (a) and histopathological specimen (b).

Table 5. Neoplastic findings (all animals).

Test substance	DNA/PBS	SWCNT-1	SWCNT-2	SWCNT-3
No. of animals examined	50	50	50	50
Lungs	(50)	(50)	(50)	(50)
Adenoma, bronchiolo-alveolar	1	0	1	6
Carcinoma, bronchiolo-alveolar	1	0	0	3
Thyroid glands	(8)	(16)	(17)	(12)
Adenoma, C-cell	3	11*	13*	8
Pancreas (endocrine)	(9)	(11)	(5)	(13)
Carcinoma, islet cell	1	4	4$	3

Significantly different from the DNA/PBS group *p ≤ .05 (Fisher’s exact test).

Significantly different from the DNA/PBS group $p ≤ .05 (Peto’s mortality-prevalence test, fixed-sequence procedure).

Table 6. Detailed findings of bronchiolo-alveolar carcinoma.

Test substance	Findings
DNA/PBS	Nodule, white, single, left, 12 × 10 × 10 mm
SWCNT-3	Nodule, white, single, left, 7 × 7 × 6 mm
SWCNT-3	Nodule, white, single, left, 12 × 10 × 10 mm
SWCNT-3	Nodule, white, single, right, diameter, 5 mm

TEM images of lungs

TEM images of lung parenchyma in the SWCNT-treated groups at 104–105 weeks after instillation are shown in Figure 8, indicating that both the short and long tubes reached the alveolus. Short fibers were also found in the mediastinum lymph nodes (Figure 9), although the long fibers were not found when observing five samples.

Figure 8. Long (a,b) and short (c,d) SWCNTs (arrows) found in the lung parenchyma. Images (a) and (c) are shown enlarged in (b) and (d), respectively.

Figure 9. Short SWCNTs (arrows) found in the mediastinal lymph node. The image in (a) is enlarged in (b).

Comet assay

SWCNT genotoxicity was also evaluated in vivo with comet assays, using rat lung cells. Previously, we showed that short SWCNTs induced chronic inflammation in rat lungs 26 weeks after a single intratracheal instillation (Ema et al., 2017). Thus, the comet assay was performed at 26 weeks to determine whether sustained stimulation by SWCNTs induced DNA damages accompanied by the chronic inflammation. The average tail-DNA values in the lung cells were 4.1 and 22.8% in the negative- and positive-control groups, respectively. Those values were 3.6, 3.6 and 4.4% in the SWCNT-1, SWCNT-2 and SWCNT-3 groups, respectively. No significant difference in percent tail DNA was noted between the negative-control group and the SWCNT-treated groups.

Discussion

Previously, we performed a 26-week observational study after a single intratracheal instillation, which demonstrated that SWCNTs with an average length of 2.8 μm induced weaker inflammation in rat lungs than did SWCNTs with an average length of 0.4 μm (Ema et al., 2017). We have also reported that a single intratracheal administration of SWCNTs with an average length of 0.69 μm did not induce lung tumors at 108 weeks after instillation (Fujita et al., 2015a). However, a substantial number of SWCNTs accompanied with chronic inflammation and fibrotic changes were observed in the lungs at 1-year post exposure to SWCNTs (Fujita et al., 2015a; Shvedova et al., 2014). We observed a continuous elevation in the levels of lactose dehydrogenase, total protein levels, and some cytokines in bronchoalveolar lavage fluid when SWCNTs were administrated by intratracheal instillation (Ema et al., 2017; Fujita et al., 2015a, 2015b, 2016). Those elevations suggest that persistent lung injury may have occurred. Furthermore, chronic inflammation and fibrotic changes in the lungs are considered significant risk factors for pulmonary carcinogenesis (Hubbard et al., 2000; Shvedova et al., 2008). Therefore, a 2-year observational study was performed here to evaluate the lung carcinogenicity of SWCNTs after a single intratracheal instillation. This is the first report showing long-term pulmonary effects using long, linear-shaped SWCNT fibers.

We improved the preparation procedure for the long-SWCNT suspension, yielding a dispersion solution with 52% of the fibers having a length of ≥10 μm. This value was much higher than the value of 16%, which were reported previously (Ema et al., 2017). A food processor was used for the dispersion process to prevent CNT shortening, which often occurred when using a conventional homogenizer in our previous studies. Purifying the dispersion solvent by column chromatography also increased the long fiber content.

Measuring inhalation is the gold standard for determining the potential toxicity of an inhalant and represents the main route of NM exposure in an occupational setting. In this study, we adopted an intratracheal-instillation method to evaluate the long-term toxicity of SWCNTs. Intratracheal instillation is recognized as an administration procedure that usually reproduces the effects of inhalation well and can provide relevant information regarding the basic toxicity of an inhalable material (Costa et al., 2006; Driscoll et al., 2000). This method has been regarded as useful for screening in vivo to explore the respiratory toxicity of NMs (METI, 2016; Suzui et al., 2016).

In this study, the short fibers induced stronger persistent pulmonary inflammation than did the long fibers. This result was similar to those observed in our previous experiments involving 13-week (Fujita et al., 2015b) and 26-week (Ema et al., 2017) observation studies, in which short SWCNTs induced higher pulmonary inflammatory responses that did longer SWCNTs. It was estimated that approximately 40-fold more fibers were present in the short-SWCNT solution than in the long-SWCNT solution at the same concentration (1 mg/ml), assuming that the SWCNTs adopted cylindrical form with the average lengths and diameters presented in Table 2. The larger number of fibers in the short-SWCNT solution would affect the activity of macrophages engulfing the nanotubes, which was presumably related to the observed stronger pulmonary inflammatory responses. The short fibers adopted well-dispersed, fiber-like forms in the lung tissues, as shown by a TEM observation (Figure 8). It has been reported that administering well-dispersed CNTs induced lung fibrosis, but not lung granulomas (Mercer et al., 2008). These findings are consistent with our results showing that fibrosis was frequently observed in the short fiber-administered group.

The short fibers were found also in the mediastinal lymph nodes, as indicated in Figure 9. They were probably translated with the lung tissues, the interstitium, and the lymph system. Translated fibers can potentially reach the peritoneum. Data from some intraperitoneal-injection experiments revealed that MWCNTs injected into the peritoneum-induced mesothelioma (Nagai et al., 2011, 2013; Takagi et al., 2008, 2012). Thus, mesothelioma at the peritoneum could be a sensitive indicator of the carcinogenicity of the CNTs. In this study, one example of mesothelioma was observed in each of the SWCNT-2 and SWCNT-3 groups, but mesothelioma was not significantly associated with SWCNT instillation. Further pharmacokinetics studies are needed on the translating CNTs, as well as their whole-body effects. No reliable method for determining SWCNTs in tissues was available when this study was started, although recent data show promise that such quantitative methods will be developed soon (Doudrick et al., 2013; Ohnishi et al., 2016).

The long tubes induced inflammatory responses at the terminal bronchioles, suggesting that most particles were deposited there, although the TEM images indicated that some reached the alveoli. The longer tubes are expected to be more likely to deposit at the trachea and/or alveoli during breathing. Even if workers were exposed to the same concentrations of short and long CNTs in a work environment, a minor fraction of long tubes reaching the interstitium or deep parts of the lungs would result in a reduced risk for adverse reactions. However, high deposition of CNTs should be avoided to prevent the workers from exposure to an unnecessary health hazard.

Data regarding the carcinogenicity of airway exposure to MWCNTs have just begun to be published. Exposure to Mitsui MWCNTs (MWNT-7) at 5 mg/m³ for a total of 15 d (5 h/d, 5 d/week) was reported to promote methylcholanthrene-initiated lung carcinogenicity in mice (Sargent et al., 2014). Intratracheal instillation of Nikkiso MWCNTs at 1 mg/rat (8 instillations of 125 μg/rat) over a 2-week period increased the incidences of malignant mesothelioma and lung tumors (bronchiolo-alveolar adenoma and carcinomas) in male rats (Suzui et al., 2016). Rats were exposed to MWNT-7 in a whole-body inhalation chamber for 104 weeks (6 h/d, 5 d/week) at 0.02, 0.2 or 2 mg/m³ (Kasai et al., 2016). The incidences of lung carcinoma (mainly bronchiolo-alveolar carcinoma) and combined carcinomas and adenomas increased in males at 0.2 and 2 mg/m³, and in females at 2 mg/m³. These findings indicated that airway exposure to MWCNTs can induce lung carcinomas and pleural mesotheliomas. Kasai et al. (2016) suggested that MWNT-7 was carcinogenic, based on their findings form a 2-year inhalation study.

However, no definitive carcinogenicity studies have been conducted with SWCNTs. Only a few studies on the long-term effects of airway exposure to SWCNTs have been published. Data from a 1-year observational study in mice on purified HiPco SWCNTs (1–3 μm in length) revealed no increase in the incidence of lung tumors after inhalation at 5 mg/m³ (5 h/d for 4d) or a single pharyngeal aspiration at 40 μg/mouse (Shvedova et al., 2014). In rats, no pulmonary tumors were detected 754 d after a single intratracheal instillation of SWCNTs at 0.2 or 0.4 mg/rat (0.69 μm of average length) (Fujita et al., 2015a). In this study, we prepared long (average length, 8.6 μm) and short (average length, 0.55 μm) SWCNTs and performed a 2-year observational study after a single intratracheal instillation in rats. We found that airway exposure to either long or short SWCNTs did not cause a statistically significant increase in the incidence of lung adenomas or carcinomas. These findings suggested that SWCNTs induce chronic lung inflammation, but not tumor formation after long-term accumulation.

We previously reported that SWCNTs were not genotoxic under the conditions tested with an in vitro bacterial reverse-mutation assay, a chromosomal-aberration assay, or an in vivo bone marrow-micronucleus test (Ema et al., 2013b; Naya et al., 2011). Furthermore, intratracheal instillation of SWCNTs did not induce DNA damage in a comet assay using lung tissues of rats singly instilled at 0.3 or 1.0 mg/kg or repeatedly (once a week for 5 weeks) at 0.04 or 0.2mg/(kg·d) (Naya et al., 2012). These negative results were observed even at doses that elicited pulmonary inflammation. In this study, we also found that no DNA damage was induced in lung tissues by intratracheal instillation of SWCNTs during a long-term observation period. These findings suggested that SWCNTs are potentially not genotoxic or directly DNA-reactive.

CNTs with a high aspect ratio and long, thin, rigid and biopersistent properties may induce pulmonary responses similar to asbestos (Braakhuis et al., 2014; Poland et al., 2008). CNTs can exist as compact tangles of nanotubes that are essentially particles, or as longer, straighter fibers, and the toxicities of these two different forms are anticipated to be clearly different (Donaldson et al., 2010). Long fibers (>20 μm) are more slowly cleared because they cannot be easily enclosed by macrophages, leading to frustrated phagocytosis compared to SWCNTs, which are flexible, bendable, often entangled and more easily cleared by phagocytic cells (Donaldson et al., 2006, 2010). SWCNTs are essentially more similar to particles and, thus, SWCNT tangles would not obey the fiber-toxicity paradigm because of their non-fibrous geometry (Donaldson et al., 2006; Kolosnjaj-Tabi et al., 2010). Curled and tangled CNTs, rather than straight fibers, probably do not induce asbestos-like pulmonary responses (Braakhuis et al., 2014). Indeed, inflammation and fibrosis were caused in the peritoneal cavity after intraperitoneal injection of MWCNTs containing long fibers and long asbestos particles, but not short-tangled MWCNTs or short asbestos particles (Poland et al., 2008). Mesothelioma was induced in rats by the intraperitoneal injection of highly crystalline MWCNTs (∼50 nm in diameter), but not thick (∼150 nm in diameter) or tangled (∼2–20 nm in diameter) MWCNTs (Nagai et al., 2011, 2013). These findings suggest that SWCNTs lack potential for inducing mesothelioma.

However, we should emphasize that proper risk management is important for suppressing adverse effects on the lungs. It is known that lung tumors similar to that observed with MWNT-7 can occur when a large amount of poorly soluble substances such as titanium dioxide and carbon blacks remain in the lung (Heinrich et al., 1995; Kasai et al., 2016). Work places where CNTs are produced should be managed to avoid unprotected large exposures to the materials.

The dose of 1 mg/kg used in the SWCNT-2 and -3 groups corresponded to approximately 0.2 mg per rat with an average bw of 195 ± 9 g at the time of instillation. The CNT suspensions were too viscous to prepare a suspension with a higher CNT concentration. Pleural malignant mesothelioma and lung tumors were induced by MWCNTs intratracheally instilled into rat lungs with a total dose of 1 mg/rat (Suzui et al., 2016). This dosage was achieved by performing eight instillations over a 2-week period. We used 20% of the reported dose in our experiments. However, we think this dose was appropriate for exploring long-term pulmonary toxicity. As shown in the TEM images, a considerable amount of SWCNTs remained in the rat lungs, suggesting that an overload of the materials might have occurred to some extent. It has been said that one should be careful to avoid overly high dosages, which might give rise to artificial biological effects (Morimoto et al., 2016).

As shown in Table 3, the iron and nickel concentrations were higher in the CNT test solutions than in the negative-control solution. These metals may have originated from a catalyst used during CNT synthesis. The concentrations of cobalt, nickel, and chromium were higher in the long-tube solution than in the short-tube solution. Probably these metals were derived from the dispersion and/or column chromatography procedure. Some reports have shown that iron impurities possibly affect CNT toxicities (Liu et al., 2013). The iron concentrations determined in this study were much lower than the value of 13,700 µg/g detected in previous in vivo studies involving SWCNTs, in which significant malignant tumors were not observed (Fujita et al., 2015a; Morimoto et al., 2012b). The aluminum content was also 10-fold lower than measured in those SWCNT studies, while the nickel, chromium and manganese contents were 100-fold lower. It is hypothesized that even iron-free MWCNTs could adsorb iron-containing proteins such as hemoglobin and transferrin onto the surface and that protein-CNT complexes could transport iron into rat peritoneal mesothelial cells during endocytosis (Wang et al., 2016). We did not detect such protein-SWCNT complexes, but the comet assay used in this study did not indicate the occurrence of genotoxicity.

Conclusions

Most long SWCNT tubes deposited at the terminal bronchioles, resulting in fibrosis and epithelium loss, although some of the test sample was found in the rat lungs in the TEM images. A considerable amount of the short SWCNTs appeared to reach the alveolus, resulting in chronic inflammatory responses. Neither the long nor the short SWCNTs induced a significant occurrence of malignant mesotheliomas, lung tumors or genotoxicity. WHO/IARC have classified MWNT-7 and the other CNTs into Group 2B (possibly carcinogenic to humans) and Group 3 (not classifiable in terms of carcinogenicity to humans) carcinogens, respectively. In a recent report, it was shown that malignant mesothelioma and lung tumors were induced by MWCNTs other than MWNT-7, suggesting that the IARC classification of MWNT-7 as a group 2B carcinogen could be expanded to other species of MWCNTs (Suzui et al., 2016). However, no carcinogenicity to mammals has yet been ascribed to SWCNTs. Thus, the IARC classification should be separately evaluated with MWCNTs and SWCNTs. In addition, proper risk management should be performed at work places where CNTs are handled to avoid unprotected exposures to large quantities of SWCNTs because the short tubes can translate to the lymph system, and their biological effects are still unclear.

Acknowledgments

The authors would like to thank Dr. Junko Nakanishi for advice on the study design, Dr. Shuji Abe and Dr. Kazuhiro Yamamoto for helpful discussions, Ms. Emiko Kobayashi for help in TEM analysis, and Ms. Sawae Obara for revising the manuscript. The pathological findings were reviewed by Dr. Isao Narama, D.V.M., Diploma of the Japanese Society of Toxicologic Pathology, Diploma of the Japanese College of Veterinary Pathologist. A part of this study was based on the results obtained from a project entitled: “Innovative Carbon Nanotubes Composite Materials Project toward Achieving a Low-Carbon Society (P10024),” which was commissioned by the New Energy and Industrial Technology Development Organization (NEDO) in Japan.

Disclosure statement

No potential conflict of interest was reported by the authors.