Lung carcinogenicity by whole body inhalation exposure to Anatase-type Nano-titanium Dioxide in rats

Abstract

To investigate the carcinogenicity of anatase-type nano-titanium dioxide (aNTiO₂), F344/DuCrlCrlj rats were exposed to aNTiO₂ aerosol at concentrations of 0, 0.5, 2, and 8 mg/m³. The rats were divided into 2 groups: carcinogenicity study groups were exposed for two years, and satellite study groups were exposed for one year followed by recovery for 1 day, 26 weeks, and 52 weeks after the end of exposure. In the carcinogenicity groups, bronchiolo-alveolar carcinomas were observed in two 8 mg/m³-exposed males, showing an increasing trend by Peto's test. However, this incidence was at the upper limit of JBRC’s historical control data. Bronchiolo-alveolar adenomas were observed in 1, 2, 3, and 4 rats of the 0, 0.5, 2, and 8 mg/m³-exposed females and were not statistically significant. However, the incidence in the 8 mg/m³-exposed females exceeded JBRC’s historical control data. Therefore, we conclude there is equivocal evidence for the carcinogenicity of aNTiO₂ in rats. No lung tumors were observed in the satellite groups. Particle-induced non-neoplastic lesions (alveolar epithelial hyperplasia and focal fibrosis) were observed in exposed males and females in both the carcinogenicity and satellite groups. Increased lung weight and neutrophils of bronchoalveolar lavage fluid were observed in the 8 mg/m³-exposed carcinogenicity groups. The aNTiO₂ deposited in the lungs of the satellite group rats was decreased at 26 weeks after the end of exposure compared to 1 day after the end of exposure. At 52 weeks after the end of exposure, the decreased level was the same at 26 weeks after the end of exposure.

INTRODUCTION

Titanium dioxide (TiO₂) has been industrially produced as a white pigment for approximately 100 years and is used in a wide range of products such as paints, inks, plastics, paper, textiles, sunscreens, cosmetics, food additives (white colorants), pharmaceuticals, and toothpaste. TiO₂ is broadly classified into two groups, rutile and anatase, according to its crystal structure. In recent years TiO₂ nanoparticles have been produced with various functions added by nanoparticle nanotechnology (IARC, 2010; Nakata and Fujishima, 2012; McIntyre, 2012; Ağçeli et al., 2020; Ziental et al., 2020). Rutile-type titanium dioxide is widely used because of its high refractive index and its effectiveness in scattering and reflecting light to prevent damage caused by ultraviolet rays. Due to its excellent photocatalytic, antibacterial, and UV shielding properties, it is used for industrial catalyst carriers (solar cells), food packaging materials, and interior materials (tiles, wallpaper) (Xue et al., 2010; ACGIH, 1992; IARC, 2010).

In humans, titanium dioxide is often detected in the lungs of workers exposed to TiO₂ even years after exposure has ceased, and varying degrees of fibrotic changes in the lungs have been reported after short-term or long-term exposure to high concentrations of TiO₂ (Määttä et al., 1971; Rode et al., 1981; Redline et al., 1986; Yamadori et al., 1986; Ohno et al., 1996; Moran et al., 1991; Iijima et al., 2020). The carcinogenicity of TiO₂ in rodents has been demonstrated by three whole-body inhalation exposure studies and two intratracheal administration studies (Lee et al., 1985; Muhle et al., 1991; Heinrich et al., 1995; Borm et al., 2000; Pott and Roller, 2005). The results of these studies are considered sufficient evidence for the carcinogenic potential of TiO₂ in animals, and the International Agency for Research on Cancer (IARC) working group in 2010 raised the carcinogenic classification of TiO₂ pigment-grade particles and ultrafine particles (nanoparticles) from Group 3 (unclassifiable carcinogenicity to humans) to Group 2B (possibly carcinogenic to humans) (Volume 93, IARC, 2010). In the above carcinogenicity studies, one inhalation study showed an increased incidence of benign lung tumors in a 250 mg/m³ - dose group of female rats exposed to pigment-grade TiO₂ (rutile; 99%), and another study showed an increased incidence of benign and malignant lung tumors in female rats exposed to 10 mg/m³ - P25 nanoscale TiO₂ which was a mixture of rutile and anatase-type TiO₂ (Lee et al., 1985; Heinrich et al., 1995). Thus, the studies referred to by the IARC did not actually include studies that examined the carcinogenicity of anatase-type TiO₂ nanoparticles. Importantly, it is generally agreed that the toxicities of particles generally become stronger with decreasing size (Oberdörster et al., 1992; Brown et al., 2001; Renwick et al., 2004; Sager et al., 2008; Noël1 et al., 2013), and because a vast amount of anatase TiO₂ nanoparticles (aNTiO₂) are now being produced, it is essential that the long-term toxicity of aNTiO₂ is determined for risk assessment of workers exposed to aNTiO₂ for extended periods of time.

We have developed a dry aerosol generation and exposure system (cyclone sieve method) for whole-body inhalation exposure to multi-walled carbon nanotubes (Kasai et al., 2014). For aNTiO₂ exposure studies, the instrument was modified and optimized for aNTiO₂ aerosol generation. Using this system, we initially conducted a thirteen-week inhalation study of aNTiO₂ in rats (Yamano et al., 2022). This study revealed a concentration-dependent increase in aNTiO₂ deposition, and aNTiO₂-containing alveolar macrophages in the lung. In particular, alveolar epithelial type 2 cell hyperplasia was observed. Therefore, we conducted a 2-year carcinogenicity study by whole body inhalation with satellite groups added to investigate associations between lung burden and lung lesions. In this study, F344 male and female rats, 6-weeks-old at the commencement of the study, were exposed to aNTiO₂ aerosol for 6 hr/day, 5 days/week at concentrations of 0, 0.5, 2, and 8 mg/m³ for 104 weeks. In addition, satellite groups with recovery periods of 1 day, 26 weeks, and 52 weeks after 1 year of exposure were added to the carcinogenicity study.

MATERIALS AND METHODS

The present study was conducted with reference and compliance to the Organisation for the Economic Co-operation and Development (OECD) Guideline for Testing of Chemicals 451 (OECD, 2009), OECD Principles of Good Laboratory Practices (GLP) (OECD, 1998) and “Standards to be Observed by Testing Institutions” of the Ministry of Labour, Japan. The rats were cared for in accordance with the Standards relating to the Care and Keeping and Reducing Pain of Laboratory Animals (Ministry of the Environment, 2013) and was reviewed and approved by the Institutional Animal Care and Use Committee of the Japan Bioassay Research Center (JBRC) (approved No.0158).

Test Substance

aNTiO₂ was purchased from Tayca Co., Ltd. (AMT-600, Osaka, Japan), and was used in the present study without further purification or sieving. According to the inspection sheet of Tayca Co., Ltd., the aNTiO₂ had a purity of > 97.9% (Lot No. 6545), an average primary particle size of 30 nm, and a surface area of 63 m²/g. When measured in our laboratory, using Transmission Electron Microscope (JEM-1400; JEOL Ltd., Tokyo), the primary particle size of the aNTiO₂ was about 30 nm (Yamano et al., 2022 see additional file 11); however, many agglomerates approximately 1 μm or larger in diameter were also observed. The weight and number of aNTiO₂ samples used for exposure were measured using an electronic balance and scanning electron microscope (SEM: SU-8000; Hitachi High Tech Corporation, Tokyo, Japan), and the number of particles was 4.2 × 10⁶ in 1 μg aNTiO₂.

Animals

Two-hundred-forty 4-week-old F344/DuCrlCrlj (SPF) rats of each sex were purchased from Charles River Japan Inc. (Kanagawa, Japan). The animals were quarantined and acclimated for 2 weeks before the commencement of the experiment (inhalation exposure was started at 6 weeks of age). The animals were divided by stratified randomization into 8 body weight-matched groups. Four groups of 50 rats of each sex were used for inhalation exposure to 0, 0.5, 2 or 8 mg/m³ in the 2-year carcinogenicity study groups and 4 groups of 10 rats of each sex for inhalation exposure to 0, 0.5, 2 or 8 mg/m³ in the satellite study groups. The rats were housed individually in stainless-steel wire hanging cages placed in pyramid-shaped stainless steel inhalation exposure chambers (10 m³ in volume). The chambers were maintained at a temperature of 23 ± 2°C and a relative humidity of 50 ± 20%, with 10 air changes per hour during the experimental period. There were 7 to 9 air changes per hour in the barrier system animal room. Fluorescent lighting in the animal room was automatically controlled to give a 12-hr light/dark cycle. All rats had free access to a γ-irradiation-sterilized commercial pellet diet (CRF-LPF, Oriental Yeast Co., Ltd., Tokyo, Japan) and sterilized water.

Experimental design

In the carcinogenicity study groups, 50 rats of each sex were exposed to aNTiO₂ aerosol for 104 weeks (6 hr/day, 5 days/week exposure cycles) at target concentrations of 0, 0.5, 2, and 8 mg/m³. In the satellite groups, 10 rats of each sex were exposed to aNTiO₂ aerosol for 52 weeks (6 hr/day, 5 days/week exposure cycles) at target concentrations of 0, 0.5, 2, and 8 mg/m³. The carcinogenicity study groups, and satellite study groups were exposed to aNTiO₂ in the same inhalation chambers. The control groups (both the carcinogenicity control group and satellite control group) were kept in chambers installed in a clean room separate from the aNTiO₂-exposed groups and exposed to clean air during each exposure cycle. The aNTiO₂ concentrations used in this inhalation exposure study were determined on the basis of the results of our previous 13-week inhalation study using aNTiO₂ concentrations of 0, 6.3, 12.5, 25, and 50 mg/m³ (Yamano et al., 2022).

At the end of the 52-week exposure period, the exposed satellite groups were transferred from the exposure chambers to other chambers. The satellite control groups, on the other hand, remained in the same chamber as the 104-week control group until their respective sacrifice periods. The satellite group rats were sacrificed after 3 different recovery periods with the following plan: Recovery period 1, one-day after the 52 weeks exposure period (3 males and 3 females from each group); recovery period 2, 26 weeks after the 52 weeks exposure period (3 males and 3 females from each group); recovery period 3, 52 weeks after the 52 weeks exposure period (4 males and 4 females from each group). However, due to health of the male rats after the 52-week exposure period, the male satellite groups were changed to 4 rats in the 8 mg/m³ recovery period 1 group, 4 rats in the control and 2 mg/m³ recovery period 2 groups, and 3 rats in the control, 2 mg/m³, and 8 mg/m³ recovery period 3 group (see Table 3-3). In the carcinogenicity study groups, from week 105, rats were sacrificed on 5 separate days.

Inhalation exposure to aNTiO₂

Aerosol generation and inhalation exposure to aNTiO₂

The aerosol generator (cyclone sieve method) developed for exposure to multi-walled carbon nanotubes was improved and optimized for aNTiO₂ exposure and used to generate aNTiO₂ aerosols (Fig. 1). In this study, a DF-3 dust feeder (Sibata Scientific Technology, Ltd., Tokyo, Japan) was used for aNTiO₂ dust feeding instead of the FO-type dust feeder (Funken Powtex Corporation, Tokyo, Japan). In the DF-3 dust feeder, the aNTiO₂ is first aerosolized and introduced into the central portion of a cylindrical vessel for particle size separation. An upward spiral airflow facilitates dry aerosolization of aNTiO_2, and particles are sorted by gravity and centrifugal forces. One of the improvements of the system is that the pore size of the sieve in the sieving unit of the aerosol generator was changed from 53 μm to 25 μm based on the results of preliminary tests (un-published).

Fig. 1

Layout of the aerosol generator (cyclone sieve method) and inhalation exposure system for aNTiO₂.The dust feeder at the bottom of the figure first aerosolizes the aNTiO₂ (bulk aNTiO₂) and feeds it into the sieving unit in the aerosol generator. Here, the aNTiO₂ is further aerodynamically processed, and the aerosol is transported to the inlet of the inhalation chamber to the right of the aerosol generator. The black lines and arrows leading from the dust feeder indicate the flow of the aNTiO₂ aerosol. The black lines and arrows leading from the dust feeder indicate the flow of the aNTiO₂ aerosol. The device has a feedback system (dashed arrow) to control the supply of aNTiO₂ fed into the sieving unit by the dust feeder, maintaining a constant aerosol concentration in the inhalation chamber.

Monitoring and regulation of aNTIO₂ aerosol in the inhalation chamber

Chamber atmosphere samples were taken from the animals’ breathing zone. Monitoring and regulation of aNTiO₂ aerosols in the inhalation exposure chambers is described in detail by Kasai et al. (2014). Particle concentrations of aNTiO₂ in the chambers were monitored in real-time with an optical particle controller (OPC, OPC-AP-600, Sibata Scientific Technology, Ltd., Tokyo, Japan). The OPC data were collected every ten seconds. The mass concentrations of aNTiO₂ were determined gravimetrically by collecting aerosols on Teflon-binder filters for each target exposure concentration after 1, 3, and 5 hr of exposure every 2 weeks during the exposure period. The mass per particle (the K value) was calculated based on the particle concentration data (particles/m³) collected by the OPC and the gravimetric results (mass/m³). Using this K-value, particle concentrations were converted to mass concentrations every 30 min for each group during the exposure cycles. The chamber aNTiO₂ concentrations were held constant by controlling the dust feeder using a feedback loop from the OPC: when the chamber concentration rose above the upper limit of a designated concentration range, the dust feeder stopped supplying aNTiO₂ aerosol to the sieving unit and when the chamber concentration fell below the lower limit, the dust feeder resumed supplying aNTiO₂ aerosol to the sieving unit. The size distribution and morphology of the aNTiO₂ particles in each inhalation chamber were determined every 13 weeks (8 times in total) throughout the experimental period. The size distribution of the particles was ascertained using a micro-orifice uniform deposit cascade impactor (MOUDI) (Model 122R NanoMoudi-II, MSP, Shoreview, MN, USA), and the morphology of the particles was examined by SEM (SU8000, Hitachi Ltd., Tokyo, Japan).

Clinical observations and urinary, hematological, and blood biochemical analyses

The animals were observed daily for clinical signs and mortality. Moribund animals were promptly euthanized. Body weight and food consumption were measured once a week for the first 14 weeks and every 4 weeks thereafter. At the terminal necropsy of the carcinogenicity groups, blood was collected for hematology and blood biochemistry from the abdominal aorta under isoflurane anesthesia after overnight fasting. The blood sample was analyzed with an automatic cell analyzer (Automated hematology analyzers for Animal, XN-2000V, Sysmex Corp., Kobe, Japan) and an automatic chemistry analyzer (Hitachi 7080, Hitachi, Ltd., Ibaraki, Japan) for blood biochemistry.

Organ weights, macroscopical and microscopical examinations

To avoid contamination with aNTiO₂, animals from each administration group were anatomized separately. Organs, including the adrenal, testis, ovary, heart, lung, kidney, spleen, liver and brain, were collected at the terminal necropsy and weighed and examined for macroscopic lesions. For animals that underwent bronchoalveolar lavage fluid (BALF) testing, only the right lung was weighed. After macroscopic examination, the nasal cavity, nasopharynx, larynx, trachea, lung, bronchus, bone marrow, lymph node (including lung-associated lymph nodes), thymus, spleen, heart, tongue, salivary glands, esophagus, stomach, small intestine (including the duodenum), large intestine, liver, pancreas, kidney, urinary bladder, pituitary gland, thyroid, parathyroid, adrenal gland, testis, epididymis, seminal vesicle, prostate, ovary, uterus, vagina, mammary gland, brain, spinal cord, peripheral nerve, eye, Harderian gland, and muscles were examined histopathologically. The organs and tissues were fixed in 10% neutral buffered formalin. The nasal cavity was decalcified by immersion in formic acid-formalin solution prior to trimming and was transversely trimmed at three levels according to the procedure described in our previous paper (Nagano et al., 1997): at the level of the posterior edge of the upper incisor teeth (Level 1), at the incisive papilla (Level 2), and at the level of the anterior edge of the upper molar teeth (Level 3). The tissues were embedded in paraffin and 3 μm -thick sections were prepared and stained with Hematoxylin and Eosin (H&E). Nasal lesions were diagnosed with reference to the criteria set by the International Classification of Rodent Tumours (IARC, 2010). Histopathological diagnosis was performed by pathologists certified by the Japanese Society of Toxicologic Pathology and peer reviewed by outside pathologists.

Observation of cells in the bronchoalveolar lavage fluid (BALF) and cytological and biochemical analyses of the BALF

BALF was collected from the left lung of the carcinogenicity and satellite groups according to the method previously described by Kasai et al. (2016). In the carcinogenicity groups, 25 animals were used to determine lung weight and up to eight of the remaining animals in each group were used for BALF analysis. The right bronchus was tied with a thread in order to lavage only the left lung. The lung was lavaged 2 times with 5 mL (males) and 4 mL (females) isotonic sodium chloride solution (OTSUKA NORMAL SALINE, ph 4.5-8.0, Otsuka Pharmaceutical Factory, Inc., Tokushima, Japan). The wash-out was collected as the BALF and used for cytological and biochemical analyses. The total number of cells and differential white blood cells were counted with an automatic cell analyzer (Automated hematology analyzers for Animal, XN-2000V, Sysmex Corp., Kobe, Japan). For biochemical analysis, the BALF was centrifuged at 1,960 rpm (800 g) at 4°C for 10 min, and aliquots of the acellular supernatant were examined with an automatic chemical analyzer (Hitachi 7080, Hitachi, Ltd., Ibaraki, Japan). γ-Glutamyl transpeptidase (γ-GTP), alkaline phosphatase (ALP), and phospholipid (PL) levels were measured by conventional biochemical methods. γ-GTP is a marker for damage to Clara and type 2 epithelial cells (Ma-Hock et al., 2009), ALP is a marker of type 2 epithelial cell toxicity (Henderson, 1984), and PL is a marker for fibrosis or the fibrotic process (Onodera et al., 1983; Carter et al., 2017). In the above analyses, only animals which did not have leukemia were selected. Consequently, the number of BALF specimens in the male 2 mg/m³ in the carcinogenicity group had only two animals, although the other groups had 4 or more animals.

In the satellite group, to confirm the presence of macrophages phagocytosing aNTiO₂, after lung weight was measured BALF was collected from the left lung of one male exposed to 0, 0.5, 2, and 8 mg/m³ aNTiO₂ at recovery period 2 (26 weeks after 52 weeks of exposure). BALF was collected using the same isotonic sodium chloride solution and volume as described for the carcinogenicity group; however, the washing method was performed by the forced method using a syringe, not the water pressure method.

aNTiO₂ lung burden analysis

To measure the lung burden in the carcinogenicity and satellite group rats exposed to aNTiO₂, the accessory lobe of the right lung and the entire lobe were weighed, and the accessory lobe was used for analysis. The analyzed values for the accessory lobe were corrected for the whole lung values and presented as deposition per lung. In the carcinogenicity groups, lung burden was measured at the terminal necropsy for five animals in each of the 0.5, 2, and 8 mg/m³ aNTiO₂ particle-exposed groups. In the satellite groups, all animals were examined at the three necropsy periods. The lung tissue was put into a glass vessel, treated with 3 mL of distilled water, 3 mL of sulfuric acid, and 1 mL of nitric acid at 270°C for 1 hr. Samples were then diluted to 30–50 mL with 3% sulfuric acid. The samples were further diluted 2 to 50-fold to keep the concentration within the calibration curve, and the TiO₂ concentration in the samples was determined by Zeeman atomic absorption spectrometry (Z-5010; Hitachi High Tech Corporation, Tokyo, Japan) with a Hitachi High Tech lamp for Ti (part#207–2012 Serial 0,490,158,100). Absorbance of the digested samples was detected at 364.3 nm. Quantification was performed using a seven-point calibration curve prepared by diluting appropriate volumes of a 1000 mg/L stock solution (Kanto Chemical Co., Inc., Tokyo, Japan) to 0.025, 0.05, 0.1, 0.15, 0.2, 0.3, and 0.4 µg/mL. TiO₂ concentrations were calculated from the corresponding molecular weight ratio of TiO₂ to Ti. The values obtained were calculated as the amount of Ti per gram. Finally, the lung weight was multiplied by this value to calculate the amount of deposited aNTiO₂ per lung.

Calculation of aNTiO₂ deposition fraction in the lungs in the satellite and carcinogenic groups

Lung deposition fraction was calculated based on lung burden data in Table 4 for 104 weeks in the carcinogenicity group and for 52 weeks in the satellite group (recovery period 1; 1 day observation after 52 weeks exposure) exposure to 0.5, 2, and 8 mg/m³ aNTiO₂ particles. The parameters used to calculate deposition fraction are shown in Table 5. The lung deposition fraction was calculated by dividing the experimental lung burden (Table 4) by virtual total particle inhalation and expressing the result as a percentage. The virtual total particle inhalation was calculated as the product of the exposure concentration (0.5, 2, and 8 mg/m³) with the total respiratory volume (the deposition fraction of aNTiO₂ taken up through respiration is assumed to be 100%). To calculate total respiratory volume, the body weight in each study was used in Bide's formula Vm= 0.499BW^0.809 (BW: body weight (kg)) (Bide et al., 2000) to obtain minute respiratory volume (L/min), which was then multiplied by the total exposure time during the exposure period (6 hr/day × 5 day/week × 52 or 104 weeks). Figure 8 shows a graph of the deposition (%) on the vertical axis and the exposure concentration on the horizontal axis to determine the correlation coefficient.

Statistical analysis

Survival curves were plotted according to the Kaplan-Meier method, (Kaplan and Meier, 1958). Two-tailed tests were used for all statistical analyses except for Peto’s test and Fisher’s exact test. P values less than 0.05 were considered statistically significant. Body weight, organ weights, food consumption, hematological and blood biochemical parameters, and biochemical and cytological parameters in the BALF (N=4 or more) were analyzed by Bartlett's test to test whether the variance was homogeneous. When the variance was homogeneous, one-way ANOVA was used to test for statistical differences between groups, and when the variance was not homogeneous, the Kruskal-Wallis rank sum test was used. Statistical difference from the control group was analyzed by Dunnett's multiple comparison test when the variance was homogeneous and Dunnett's multiple comparison test by rank when the variance was not homogeneous. The incidences of neoplastic lesions (except for the satellite group rats) were statistically analyzed by Fisher’s exact test. The incidences of non-neoplastic lesions (except for the satellite group rats) were statistically analyzed by Chi-square Test. Positive trends for neoplastic incidence were analyzed using Peto’s test (Peto et al., 1980). Tumor incidence was also evaluated by comparison with incidences of the same tumor type from the Japan Bioassay Research Center historical control data (JBRC’s historical control data). The use of historical control data for evaluation of tumor induction is described by Haseman et al. (1984, 1985, and 1995).

RESULTS

Environmental conditions in the inhalation chambers

All inhalation chambers of the carcinogenicity group and the satellite group were maintained with an air exchange rate of 10 times/hr (1,667-1,673 L/min), a temperature of 22.7-22.9°C, and a humidity of 56.5–57.8% during each exposure cycle (Table 1).

Table 1. Characterization of the chamber the environment (temperature, humidity, ventilation, concentration, particle size).

aNTiO₂ concentrations and particle size distributions

The aNTiO₂ aerosol was freshly generated at each concentration during each exposure cycle of 6 hr using our aerosol generator (Fig. 1), and the aNTiO₂ concentrations in the inhalation chambers were kept at the target doses throughout the experimental period (Table 1 and Fig. 2); Figure 2 shows the mean concentrations of the aNTiO₂ aerosol in the inhalation chambers over the course of the 2-year exposure period. The concentrations of the aNTiO₂ aerosols in the inhalation chambers throughout the initial 52-week experimental period were 0.51 ± 0.02 (mean ± SD) mg/m³ for the 0.5 mg/m³ group, 2.04 ± 0.08 mg/m³ for the 2 mg/m³ group, and 7.99 ± 0.24 mg/m³ for the 8 mg/m³ group (Table 1); the carcinogenicity and satellite groups were exposed to the aNTiO₂ aerosol in the same inhalation chambers. The mean concentrations over the 104 weeks exposure period were: 0.51 ± 0.02 (mean ± SD) mg/m³ in the 0.5 mg/m³ group, 2.03 ± 0.06 mg/m³ in the 2 mg/m³ group, and 7.99 ± 0.21 mg/m³ in the 8 mg/m³ group (Table1); only the carcinogenicity groups were exposed to the aNTiO₂ aerosol for 104 weeks. The mass median aerodynamic diameters (MMAD) and the geometric standard deviations (GSD) of the aNTiO₂ aerosol were in the range of 0.9-1.1 μm and 2.2-2.4, respectively, in all aNTiO₂-exposed groups (Table 1). SEM observations of aNTiO₂ particles was performed eight times every 13 weeks throughout the experimental period. Figure 2B shows SEM photos of aNTiO₂ particles collected at week 1 and week 92. The morphology of the aNTiO₂ particles was the same in the different concentration groups, and no changes were observed in the morphology of the aNTiO₂ particles throughout the experimental period.

Fig. 2

Control of the aNTiO₂ concentration and observation of aNTiO₂ particles in the inhalation chambers. Panel A shows the measured concentration of the aNTiO₂ aerosol in the chamber at each target concentration. The concentrations of aNTiO₂ particles in the inhalation chamber were monitored by the optical particle controller (OPC in Fig. 1) throughout each 6-hr exposure period during the 104 weeks of whole body exposure. The mass concentration of aNTiO₂ in the chamber (mg/m³) is calculated by multiplying the mass per particle (K-value) concentration data (particles/m³) by the particle concentration measured by the OPC. The target concentrations of aNTiO₂ were 0.5, 2, and 8 mg/m³. Panel B shows SEM images of aNTiO₂ in the inhalation chamber. The particles shown in the figure were collected on a polycarbonate filter at 1 week and 92 weeks. The aNTiO₂ aerosols were in the form of nano- sized to μm-sized aggregates, common to all groups. The bar is 20 µm.

Clinical observations and, urinary, hematological and blood biochemical analyses

Neither aNTiO₂ related deaths nor abnormal clinical signs were observed in any aNTiO₂-exposed male or female animals in either the carcinogenicity or satellite groups.

Survival rates and body weight curves for male and female rats in the carcinogenicity groups are shown in Table 2 and Fig. 3. Although the survival rates of the 2 mg/m³ group were slightly lower in both males and females, there were no significant differences in survival rates of exposed male and female rats and their control groups. The number of surviving animals in the 0, 0.5, 2, and 8 mg/m³ groups at the end of experiment were 30, 32, 27, and 37 for males and 42, 43, 31, and 44 for females; the survival rates exceeded 54% for males and 62% for female at the end of 104-week experimental period (Table 2).

Fig. 3

Survival curves and body weight changes of rats exposed to aNTiO₂ or clean air for 104 weeks. The survival rates of the rats are illustrated in panels A (male) and B (female). The survival rate exceeded 60% in males and 62% in females at the end of the 104 week experimental period. The growth curves of the rats are illustrated in panels C (male) and D (female). Body weights were measured once a week for the first 14 weeks and every 4 weeks thereafter. There was no growth retardation in any of the groups.

A slight suppression of body weight gain in the 0.5 mg/m³-exposed male group and the 0.5 mg/m³- and 2 mg/m³-exposed female groups was observed during the experiment period. However, this was not considered to be caused by exposure to aNTiO₂ because there was little or no change in the body weight gain in either the male or female 8 mg/m³-exposed groups compared to the clean air controls. The relative body weights of the 0.5, 2, and 8 mg/m³-exposed animals at the end of experiment were 93, 95, and 99% for males and 93, 91, and 97% for females compared to their respective controls (Table 2). Hematological analysis, blood biochemical analyses, and urinalysis of the aNTiO₂-exposed animals did not show any indications of toxicity compared to the clean air controls (data not shown).

Table 2. Animal survival, body weights, and lung weights (absolute and relative weights) in rats exposed to aNTiO₂ for 104 weeks.

Macroscopic findings, organ weights, and microscopic findings

At the terminal necropsy, in both the carcinogenicity and satellite groups, multiple white-spots and patches were found on the visceral pleural costal side of the lung in nearly all male and female rats exposed to 2 and 8 mg/m³ aNTiO₂. No aNTiO₂-specific macroscopic findings were observed in other organs, including the parietal pleura and peritoneum, in either the carcinogenicity or satellite groups (Fig. 4-2 and 4-7).

In the carcinogenicity groups, absolute and relative lung weights were significantly increased in male and female rats exposed to 8 mg/m³ aNTiO₂ (Table 2). No aNTiO₂ exposure-related increases or decreases were found in the weights of the other organs. In the satellite groups, no changes in the absolute or relative organ weights, including the lung, were found at the terminal necropsies after the three recovery periods (data not shown).

The major neoplastic lesions in the carcinogenicity group are shown in Table 3-1, and selected non-neoplastic lesions in the lung of the animals in the carcinogenicity groups are shown in Table 3-2. In males, bronchiolo-alveolar carcinomas were observed 0, 0, 0, and 2 animals in the 0 (clean air control), 0.5, 2, and 8 mg/m³ groups, respectively (Fig. 4-4). This incidence showed a positive trend by the Peto's test (Table 3-1). Bronchiolo-alveolar adenomas were observed 4, 5, 7 and 2 animals in the 0, 0.5, 2, and 8 mg/m³ groups, respectively, and combined incidences of bronchiolo-alveolar carcinomas and bronchiolo-alveolar adenomas were 4, 5, 7, and 4 animals in the 0, 0.5, 2, and 8 mg/m³ groups, respectively.

In females, bronchiolo-alveolar adenomas were observed 1, 2, 3, and 4 animals in the 0 (clean air control), 0.5, 2, and 8 mg/m³ groups, respectively (Table 3-1 and Fig. 4-3). Although the incidence in the 8 mg/m³ group (4/50) was not statistically significant, it did exceed the incidence (15/600: incidence 2.5%, range 0–6%) of bronchiolo-alveolar adenomas in female rats in the JBRC historical control data. Bronchiolo-alveolar carcinomas were observed 0, 0, 1, and 0 animals in the 0, 0.5, 2, and 8 mg/m³ groups, respectively, and adenosquamous carcinomas were observed 0, 0, 1, and 0 animals in the 0, 0.5, 2, and 8 mg/m³ groups, respectively. The combined incidences of bronchiolo-alveolar adenomas, bronchiolo-alveolar carcinomas and adenosquamous carcinomas were 1, 2, 5, and 4 animals in the 0, 0.5, 2, and 8 mg/m³ groups, respectively.

Non-neoplastic lesions observed in the carcinogenicity and satellite groups were predominantly alveolar epithelial hyperplasia (alveolar epithelial type 2 cell hyperplasia) which were observed with particle deposition in all exposure groups, and the particles were found in the alveolar space in a phagocytosed by alveolar macrophages. Alveolar epithelial hyperplasia found on the visceral pleural side, and alveoli close to the bronchi tended to be intact (Fig. 4-5 and 4-6). The alveolar epithelial hyperplasia, which is characteristic of particle exposure, seen in this experiment showed a growth pattern that was distinct from the growth pattern of bronchiolo-alveolar hyperplasia. Therefore, these two hyperplasias were distinguishable (Tables 3-2 and 3-3). The incidence of alveolar epithelial hyperplasia in the lung were 31/50 males and 42/50 females in the 0.5 mg/m³ group, 44/50 males and 47/50 females in the 2 mg/m³ groups, and 49/50 males and 50/50 females in the 8 mg/m³ groups (Table 3-2). In the recovery period 1 (one day after the 52-weeks exposure) satellite groups, these lesions were observed in 1/3 males and 0/3 females in the 0.5 mg/m³ groups, 3/3 males and 2/3 females in the 2 mg/m³ groups, and 4/4 males and 3/3 females in the 8 mg/m³ groups; in the recovery period 2 (26 weeks recovery after the 52 weeks exposure) animals, these lesions were observed in 2/3 males and 2/3 females in the 0.5 mg/m³ groups, 3/4 males and 3/3 females in the 2 mg/m³ groups, and 3/3 males and 3/3 females in the 8 mg/m³ groups; and in the recovery period 3 (52 weeks recovery after the 52 weeks exposure) animals, these lesions were observed 1/4 males and 2/4 females in the 0.5 mg/m³ groups, 1/3 males and 3/4 females in the 2 mg/m³ groups, and 3/3 males and 4/4 females in the 8 mg/m³ groups (Table 3-3). The histopathological findings in the satellite groups showed that the severity of the lesions decreased over time after the end of exposure. In the carcinogenicity and satellite control groups, no alveolar epithelial hyperplasia was observed in either sex.

Focal fibrosis at alveolar wall of the lung was observed in all exposed groups; the incidences were 2/50 males and 1/50 females in the 0.5 mg/m³ group, 2/50 males and 4/50 females in the 2 mg/m³ groups, and 28/50 males and 48/50 females in the 8 mg/m³ groups (Table 3-2). In the satellite groups, focal fibrosis was only observed in the recovery period 3 groups (52 weeks recovery after the 52 weeks exposure); 2/3 males in the 8 mg/m³-exposed group and 1/4 females in the 0.5 mg/m³, 1/4 females in the 2 mg/m³, and 3/4 females in the 8 mg/m³ (Table 3-3). No fibrosis in the lung was observed in either sex in the control groups.

In the carcinogenicity and satellite groups, macroscopic examination revealed white-spots and patches, and microscopic examination revealed alveolar epithelial hyperplasia and focal fibrosis (Fig. 4-5, 4-6 and 4-8) that contained aNTiO₂, and these lesions increased in size and number with exposure concentration. In addition, aNTiO₂ particles were observed in the alveoli (phagocytosed alveolar macrophages) and in the bronchus-associated lymphoid tissue (BALT) of the lungs and lymph nodes of mediastinal in all exposed carcinogenicity groups and satellite groups at terminal necropsy of male and female rats.

Fig. 4

Representative Macroscopic and microscopic pathological images. Fig. 4-1. Macroscopic image of the normal lung, satellite group (52w+52w), control, female. Fig. 4-2. Macroscopic image of the lung with white patches, carcinogenicity study group, exposed to 8 mg/m³ aNTiO₂, female. Fig. 4-3. Microscopic image of Bronchiolo-alveolar adenoma of the lung, carcinogenicity study group, exposed to 8 mg/m³ aNTiO₂, female. Fig. 4-4. Microscopic image of Bronchiolo-alveolar carcinoma of the lung, carcinogenicity study group, exposed to 8 mg/m³ aNTiO₂, female. Fig. 4-5. Microscopic image of alveolar epithelial hyperplasia (particle induced alveolar epithelial type 2 cell hyperplasia) of the lung, carcinogenicity study group, exposed to 8 mg/m³ aNTiO₂, female. Fig. 4-6. Microscopic image of Focal fibrosis of the lung, carcinogenicity study group, exposed to 2 mg/m³ aNTiO₂, female. Fig. 4-7. Macroscopic image of the lung with white patches, recovery period 3 satellite group (52w+52w), exposed to 8 mg/m³ aNTiO₂, female. Fig. 4-8. Microscopic image of Focal fibrosis of the lung, recovery period 3 satellite group (52w+52w), exposed to 2 mg/m³ aNTiO₂, female.

Table 3.

Cytological and biochemical analyses of bronchoalveolar lavage fluid (BALF)

Macrophages with phagocytosed aNTiO₂ were present in males and females from all exposed groups in the carcinogenicity group. Macrophages filled with aNTiO₂ in the cytoplasm were observed in the BALF from males and females in the 2 and 8 mg/m³-exposed groups, and this was particularly evident in the 8 mg/m³-exposed groups, where macrophage enlargement and disintegration were observed (Fig. 5-A). However, macrophage numbers were not significantly increased in the exposed groups. Neutrophils were increased in both sexes in the 8 mg/m³-exposed groups. Lymphocytes were also increased in females in the 8 mg/m³-exposed group (Fig. 5-B, left panel).

Fig. 5

Cytological and biochemical analyses of bronchoalveolar lavage fluid (BALF) in carcinogenicity study. A: Alveolar macrophages in the BALF of male rats exposed to aNTiO₂ at concentrations of 0, 0.5, 2, and 8 mg/m³ for 2 years. In the 2 and 8 mg/m³-exposed groups, macrophages were observed with their cytoplasm filled with aNTiO₂, especially prominent in the 8 mg/m³-exposed groups. Enlarged and disintegrated macrophages were also observed. Cells stained with May–Grünwald–Giemsa stain. Bars indicate 50 µm in the control group image and 20 µm in the exposure group images. B: The number of inflammatory cells/mL in the BALF is shown in the left panel. The numbers of neutrophils are increased in male and female groups exposed to 8 mg/m³ aNTiO₂. Lymphocytes are increased in the 8 mg/m³ female group. **: p < 0.01 by Dunnett’s multiple comparison test. The right panel shows the Biochemical analyses: Alkaline phosphatase (ALP), γ-Glutamyl transpeptidase (γ-GTP), and Phospholipid (PL) levels in the BALF were determined for male and female rats. Error bars indicate the SD for 2 rats in the male 2 mg/m² group and the SD for 4 or more rats in the other groups. *: p < 0.05 and **: p < 0.01 by Dunnett’s multiple comparison test.

γ-GTP levels in the BALF of males in the 8 mg/m³ aNTiO₂-exposed group and females in the 2 mg/m³ aNTiO₂ and 8 mg/m³-exposed groups (Fig. 5-B, right panel) were significantly elevated. ALP and PL levels in the BALF were not elevated in the exposed groups.

To determine the presence or absence of macrophages with phagocytosed aTiO₂ after the end of exposure, BALF was collected from the satellite group 26 weeks after the end of exposure (satellite recovery period 2). Macrophages with phagocytosed aNTiO₂ were present in males from all exposed groups (Fig. 6). The number of macrophages phagocytosed aNTiO₂ increased with exposure concentration. After measuring the lung weight, lung lavage was performed, and the lavage method was the forced lavage method using a syringe, not the water pressure method. Therefore, unlike the standard method, a large number of red blood cells were observed due to blood contamination.

Fig. 6

BALF photographs collected from recovery period 2 satellite group males (26 weeks after the end of 52 weeks of exposure) exposed to 0, 0.5, 2, and 8 mg/m³ aNTiO₂. Since the lungs were lavaged after the lung weight was measured, the lavage contained a large number of red blood cells. Macrophages phagocytosing aNTiO₂ are observed even 26 weeks after the end of exposure in all exposed groups. Bars indicate 20 µm.

Lung burden analysis of aNTiO₂

The average weight of the accessory lobe of the right lung was 0.15 g for males and 0.11 g for females, both the carcinogenicity and satellite group animals. Table 4 shows the aNTiO₂ lung burden in rats in the carcinogenicity groups and the satellite groups, and Fig. 7 shows graphs of the aNTiO₂ lung burden in rats in the satellite groups after the three recovery periods. Comparing the aNTiO₂ deposition levels in the lungs of the satellite recovery period 1 groups (52 weeks exposure) and the carcinogenicity groups (104 weeks exposure), the lung burdens in the 0.5 and 2 mg/m³ groups was related to the duration of exposure for both sexes. Lung burdens were about twice as high in the carcinogenicity group as in the satellite recovery period 1 groups. However, the 8 mg/m³ carcinogenicity groups (11.9 and 10.10 mg per lung for males and females) showed higher aNTiO₂ retention values than expected when compared to the satellite recovery period 1 group (4.34 and 3.50 mg/ lung for males and females).

Compared to recovery period 1 rats, the deposition percentage of aNTiO₂ after recovery period 2 was 40% for males and 36% for females in the 0.5 mg/m³ groups, 45% for males and 58% for females in the 2 mg/m³ groups, and 67% for males and 67% for females in the 8 mg/m³ groups. Thus, during the 26-week recovery period, 60% of the aNTiO₂ was removed from lung in males and 64% in females in the 0.5 mg/m³-exposed groups, 55% in males and 42% in females in the 2 mg/m³ groups, and 37% in males and 33% in females in the 8 mg/m³ groups. Notably, comparing the lung burden after recovery period 2 (26 weeks) and recovery period 3 (52 weeks) the amount of aNTiO₂ in the lungs did not decrease from week 26 to week 52 in any of the exposed male or female groups (Table 4 and Fig. 7).

Fig. 7

Lung burden analysis of the satellite groups from the three recovery periods. Intrapulmonary aNTiO₂ was measured in rats at 1 day, 26 weeks, and 52 weeks after the end of the 52-week exposure. The top graphs show the results for males, and the bottom graphs show the results for females.

Table 4. Lung burden analysis of Anatase-type Nano-titanium Dioxide (aNTiO₂).

Calculation of aNTiO₂ deposition fraction in the lungs in satellite and carcinogenic groups

Table 5 shows the parameters used to calculate the deposition fraction of aNTiO₂ in the lungs of the carcinogenicity group rats at 104 weeks exposure and the satellite groups at 52 weeks exposure: mean body weight (g) (1 week to 104 weeks or 1 week to 52 weeks), total respiratory volume (m³), virtual total particle inhalation (mg), and aNTiO₂ deposition (%). The deposition (%) was 1.33, 1.52, and 3.45 for males and 1.47, 2.11, and 4.58 for females in the 0.5, 2, and 8 mg/m³ groups. In the 52 weeks exposure groups, the deposition (%) was 1.54, 2.17, and 2.75 for males and 1.80, 1.90, and 3.75 for females in the 0.5, 2, and 8 mg/m³ groups. These data were plotted with the deposition (%) on the vertical axis and the exposure concentration on the horizontal axis to determine the correlation coefficients (Fig. 8). The correlation coefficient was 0.99 for males and 1 for females in the 104-week exposure group, indicating a high correlation between deposition (%) and exposure concentration. In the 52-week exposure group, the correlation coefficient was 0.88 for males and 0.980 for females, which also shows a high correlation between deposition (%) and exposure concentration. When the aNTiO₂ deposition (%) is compared between the 52 week and 104 week groups, deposition (%) was higher in the 104 week 8 mg/m³-exposed group for both males and females. In the 0.5 and 2 mg/m³-exposed groups, the deposition (%) at 52 weeks was almost the same or even higher than the deposition (%) at 104 weeks.

Fig. 8

aNTiO₂ deposition fraction for males and females exposed to 0.5, 2, and 8 mg/m³ aNTiO₂ for 52 and 104 weeks. The virtual total particle inhalation was determined from total respiratory volume obtained from Bide's formula (Bide et al., 2000) and each exposure concentration, and the deposition fraction was calculated by dividing the measured particle deposition value by the virtual total particle inhalation. The vertical axis is the deposition (%), and the horizontal axis is the exposure concentration. The correlation coefficient is shown in the graph.

DISCUSSION

To investigate the potential hazard of airborne aNTiO₂, rats were exposed to aNTiO₂ aerosol for 6 hr/day, 5 days/week for 104 weeks. There were no abnormal clinical signs, such as suppression of body weight gain, or decreased survival rate in the aNTiO₂-exposed groups. Therefore, the concentration used in this study fulfills the maximum tolerated dose (MTD) criteria (Sontag et al., 1976; Bannasch et al., 1986) for 104-week bioassay studies, including rodent carcinogenicity studies.

After 104 weeks exposure, bronchiolo-alveolar carcinoma was observed in two males (2/50, 4%) exposed to 8 mg/m³ aNTiO₂, which the Peto test (prevalence method) identified as an increasing trend. However, this incidence is at the upper end of the range of JBRC’s historical control data (5/599: range 0–4%). In females, bronchiolo-alveolar adenomas were observed 1, 2, 3, and 4 animals in the 0 (control), 0.5, 2, and 8 mg/m³-exposed groups, respectively. Although there was no statistical difference in the incidence of bronchiolo-alveolar adenomas in the exposed groups compared to the control group, the incidence in the 8 mg/m³ group (4/50, 8% in the 8 mg/m³) exceeded that of JBRC's historical control data (15/600: range 0–6%). Based on these data, we concluded that at 8 mg/m³ aNTiO₂ showed equivocal evidence of lung carcinogenicity in male and female F344 rats.

We also found evidence of chronic toxicity. Lung lesions, alveolar epithelial hyperplasia (alveolar epithelial type 2 cell hyperplasia) and focal fibrosis, were found close to the visceral pleural side of the lungs and tended to be intact in the alveoli close to the bronchioles. These findings suggest that these lesions were associated with migration of macrophages that had phagocytosed the particles or migration of the particles by lymphatic flow. In the carcinogenicity groups exposed to 2 and 8 mg/m³ for 104 weeks, macrophages containing phagocytosed aNTiO₂ were observed within epithelial hyperplasia and fibrotic lesions. However, a statistically significant induction of fibrotic lesions was observed only in the 8 mg/m³ group. In the recovery period 2 group (26 weeks recovery after the end of 52 weeks exposure), aNTiO₂ phagocytosing macrophages were observed in the BALF of all exposed groups (Fig. 6), and in the recovery period 3 group (52 weeks observation after end of 52 weeks exposure), delayed fibrotic lesions were observed in the 8 mg/m³ group. Finally, an increase in neutrophils was observed in the 8 mg/m³-exposed carcinogenicity groups, indicating prolonged neutrophil inflammation (Fig. 5). Furthermore, biochemical analysis of the BALF showed that γ-GTP was elevated in males exposed to 8 mg/m³ and in females exposed to 2 mg/m³ or higher aNTiO₂. γ-GTP is a marker of Clara and type 2 epithelial cell damage, and ALP is a marker of type 2 epithelial cell toxicity. Our exposure experiments showed an increase only γ-GTP, but no notable change in ALP. Accordingly, aNTiO₂ cytotoxicity to Clara cells was suggested.

Lung particle overload is associated with inflammation, epithelial hyperplasia, and, in extreme cases, lung cancer. “Lung particle overload” refers to the impaired lung particle clearance and increased particle retention occurring with inhalation of high concentrations of poorly soluble low toxicity particles (PSLTs) (ILSI Risk Science Institute Workshop Participants, 2000; Driscoll and Borm, 2020). Rats are sensitive to PSLT-induced lung particle overload, and PSLTs such as TiO₂ have been demonstrated to cause lung cancer in rats (Lee et al., 1985; Heinrich et al., 1995). A recent workshop on the inhalation toxicity of PSLTs concluded that rats can be used as a sensitive species for PSTL inhalation toxicology, but that study designs for chronic inhalation toxicity should include both overload (impaired clearance) and non-overload doses (Driscoll and Borm, 2020). Importantly, the workshop also concluded that overload-associated lung tumors in rats does not imply a human hazard unless supported by human data or data from other species. On the other hand, regulatory agencies have considered overload-associated lung tumors in rats as evidence of possible human hazard (IARC, 2010; ECHA, 2017). However, these regulations are open to change (CURIA, 2022; Driscoll, 2022). The relationship between lung tumor development and lung overload must be carefully evaluated when assessing the lung hazard of PSTLs.

Lung overload is considered in terms of volume and mass. Experimental evidence suggests that the volume of particles phagocytized by alveolar macrophages is critical for causing impaired clearance, and that the condition of lung overload is reached once the retained lung particle burden reaches a level equivalent to a volume of approximately 1 μL/g of lung tissue (Oberdörster, 1995). From the MMAD of 1.0 μm in the present 2-year carcinogenicity study, the mean volume of the aNTiO₂ particles (using the formula for the volume of a perfect circle of radius 0.5 μm, V = 4/3 πr³) is calculated as 0.524 μm³. The total particle volume (μm³) in the lung was calculated by multiplying 0.524 μm³ by the lung burden (mg/lung) in Table 4 and the number of particles per 1 μg of aNTiO₂. As described in Test substance of MATERIALS AND METHODS, the number of particles per 1 μg of aNTiO₂ was 4.2×10⁶. Using this total volume and the lung weights in Table 2, the particle volume in μL (1 μL = 10⁹ μm³) in 1 g of lung tissue in each group was calculated. Exposure to 0.5 mg/m³ in males and females resulted in 0.45 µL/g of lung for males and 0.47 µL/g of lung for females; exposure to 2 mg/m³ resulted in 2.05 µL/g of lung for males and 2.51 µL/g of lung for females; exposure to 8 mg/m³ resulted in 17.20 µL/g of lung for males and 20.08 µL/g of lung for females. Retained particle mass is also associated with lung overload. Muhle et al. (1990) report that alveolar clearance retardation was detectable above a retained pulmonary burden of 0.5 mg per rat lung, and a substantial decrease in the clearance rate (about a factor of 6) was observed following heavy dust loading, exceeding 10 mg dust per rat lung. Using these criteria, overload might have occurred at exposure concentrations of 2 and 8 mg/m³ in our study: 1.30 mg/lung and 11.90 mg/lung in the 2 and 8 mg/m³-exposed male groups, and 1.13 mg/lung and 10.10 mg/lung in the 2 and 8 mg/m³-exposed female groups. Analysis of the volume and mass of the retained aNTiO₂ in our study suggest that lung overload occurred at 2 mg/m³ aNTiO₂ exposure, with clear lung overload in the 8 mg/m³ groups.

Analysis of lung burden data in the satellite and carcinogenicity groups in Table 4 and Fig. 6 indicated that: (1) The clearance half-time was calculated to be within 26 weeks for male and female rats in the 0.5 mg/m³ group and male rats in the 2 mg/m³ group. However, for female rats exposed to 2 mg/m³ and male and female rats exposed to 8 mg/m³, the half-time was not determined because lung burden was not reduced by more than half after either the 26-week or 52-week recovery periods. (2) There was little or no clearance of aNTiO₂ in any of the satellite exposure groups from week 26 to week 52 after the end of exposure.

Two different clearance mechanisms are known to be involved in the clearance of particles deposited in the alveolar region. One is associated with phagocytic clearance by macrophages, which generally has a half-time of 60 to 80 days (Driscoll and Borm, 2020). The other is associated with lymphatic clearance, a slower mechanism with a half-time of months or years. The percentage of pulmonary dust cleared via the lymphatic vessels increases with the total lung dust burden (Lippmann et al., 1980; Muhle et al., 1990). The pathological observations in the carcinogenicity study groups (104 weeks exposure) showed exposure to increasing levels of aNTiO₂ resulted in dose-dependently increased deposition of aNTiO₂ in the lymph nodes (Table 3-2). aNTiO₂ deposition in the lymph nodes was also observed in 1/4 of males and 1/3 of females in the 8 mg/m³-exposed recovery period 1 satellite groups (52 weeks exposure), but no aNTiO₂ deposition in the lymph nodes in either males or females in the 0.5 mg/m³ or 2 mg/m³ recovery period 1 groups (Table 3-3). Although a direct comparison of the carcinogenicity groups with the satellite groups is not possible due to the small number of animals in the satellite groups, the higher percentage of carcinogenicity group animals with aNTiO₂ deposited in the lymph nodes compared to the satellite group animals (compare Tables 3-2 and 3-3) suggests that aNTiO₂ deposition in the lymph nodes was significantly higher after 104 weeks exposure compared to 52 weeks exposure. These data suggest that elimination of aNTiO₂ by the lymphatics occurred, but that continued inhalation of aNTiO₂ into the lung resulted in inhibiting/blocking clarence via the lymphatics. Notably, a significant number of animals in the 0.5 mg/m³ and 2 mg/m³ carcinogenicity groups had particle deposition in the lymph nodes (Table 3-2), suggesting that lymphatic flow obstruction occurred even at the lowest exposure concentration of 0.5 mg/m³. Notably, macrophages in the BALF of recovery period 2 animals contained phagocytosed particles (Fig. 6); however, the lung burden in the recovery period 3 (52 weeks post-exposure) animals was similar to the lung burden in the recovery period 2 (26 weeks post-exposure) animals (Table 4 and Fig. 7). These data indicate that clearance by macrophages and lymphatic clearance were both blocked. The exact mechanism by which aNTiO₂ particles interact with cells/tissue in the lung that results in blocking particle clearance from the lung is currently unknown. Importantly, this blockage is not caused by lung overload as it occurred in rats exposed to 0.5 mg/m³ aNTiO₂, which as discussed above did not result in lung overload. The blockage is also not specific to rats as titanium dioxide is often detected in the lungs of workers exposed to TiO₂ years after exposure has ceased (Määttä et al., 1971; Rode et al., 1981; Redline et al., 1986; Yamadori et al., 1986; Ohno et al., 1996; Moran et al., 1991; Iijima et al., 2020).

The question of whether aNTiO₂ causes cytotoxicity in alveolar macrophages is an also important factor when considering impaired clearance. One of the mechanisms underlying impaired clearance is direct damage to alveolar macrophages by chemical-specific toxicity. Our carcinogenicity study confirmed that neutrophils and lymphocytes were increased in the BALF in the 8 mg/m³ group, suggesting that inflammation was continuously induced in the lung from early in the exposure period. However, examination of the BALF showed no change in the morphology of most macrophages even in the highest concentration, 8 mg/m³, exposure groups; however, some macrophages appeared to have expanded due to excessive particle phagocytosis and some appeared to have ruptured and disintegrated into pieces in severe cases. These suggests that particle toxicity to macrophages is not very strong.

We calculated the lung deposition fraction at the end of the 52-week exposure period in the recovery period 1 satellite groups and at the end of the 104-week exposure period in the carcinogenicity groups. The deposition fraction was calculated as follows: The Respiratory minute Volume (Vm: liters/minute) was calculated using Bide's formula: Vm = 0.499BW^0.809 with Body Weight (BW) in Kg (Bide et al., 2000). The body weights used in the above calculations are the average weights of males and females during the exposure period, 52 or 104 weeks, shown in Table 5. The total respiratory volume was calculated by multiplying Vm (L/min) by the exposure time (in minutes). The virtual total particle inhalation was calculated by multiplying the total respiratory volume by the exposure concentration. The deposition percentage was calculated by dividing the measured particle deposition values (Table 4) by the virtual total particle inhalation. Figure 8 shows a graph with the deposition (%) on the vertical axis and the exposure concentration on the horizontal axis to determine the correlation coefficient. The deposition percentages for the 0.5, 2, 8 mg/m³ groups in the 2-year exposure group were 1.33, 1.52, 3.45 for males and 1.47, 2.11, 4.58 for females, showing a good correlation with the exposure concentration. The aNTiO₂ lung deposition percentages in the 2-year exposure groups correlated with the exposure concentration (0.99 for males and 1 for females) better than the lung deposition percentages in the 52-week exposure groups (0.88 for males and 0.98 for females); although, while the 52 week exposure correlation coefficient is less than the 104 week correlation coefficient, the correlation of exposure concentration and lung deposition for males in the 52-week exposure group is excellent. The difference between the 52-week exposure and 104-week exposure correlation coefficients is likely due to the small number of animals in the 52-week exposure satellite groups. Analysis of the deposition percentages in the carcinogenicity and satellite groups did not appear to show an unnatural increase in particle retention or deposition percentages, even at exposure concentrations causing high lung loads that the volume and mass calculations discussed above suggest may have caused lung particle overload.

In this study, rats exposed to 2 mg/m³ and 8 mg/m³ aNTiO₂ had lung burdens greater than 0.5 mg per lung and volumes of aNTiO₂ exceeding 1 μL/g lung tissue, suggesting the possibility of lung particle overload in these exposure groups. However, exposure to 2 mg/m³ aNTiO₂ did not cause increased neutrophil infiltration into the lung (inflammation) or lung cancer. In addition, development of carcinogenic and non-carcinogenic lesions in the 0.5 and 2 mg/m³ aNTiO₂ exposure groups was similar, and the slightly higher non-carcinogenic lesions observed in the 2 mg/m³ aNTiO₂ exposure groups compared to the 0.5 aNTiO₂ exposure groups are consistent with non-particle overload exposure (see Table 3-1 and 3-2). Therefore, if exposure to 2 mg/m³ aNTiO₂ did cause lung overload, the overload did not result in increased toxicity. The correlation between exposure and lung deposition also suggest that the increased lung burden in the 8 mg/m³ aNTiO₂ groups was not caused by lung overload. Importantly, impaired particle clearance was observed in all exposure groups, indicating that the impaired clearance was not caused by possible lung overload in the 8 mg/m³ aNTiO₂ groups. Therefore, the lesions that developed in the 8 mg/m³ aNTiO₂ groups were very likely to not be caused by possible lung overload.

Table 5. Parameters to determine aNTiO₂ deposition fraction after 52 and 104 weeks of exposure to aNTiO₂ at 0.5, 2, and 8 mg/m^3.

It is important to assess the occupational exposure limits (OEL) of TiO₂ nanoparticles in workplaces for protection of workers from occupational cancer. To calculate the OEL for carcinogens, it is necessary to know whether the substance itself is genotoxic or non-genotoxic. There have been various reports on the genotoxicity of nano-TiO₂, with some reports claiming genotoxicity and others that report that nano-TiO₂ is not genotoxic. IARC has classified nano-TiO₂ as a non-genotoxicity substance (IARC, 2010); however, there are many papers suggesting that nano-TiO₂ is genotoxic (Bhattacharya et al., 2005; Jaeger et al., 2012; Lu et al., 1998; Gurr et al., 2005; Wang et al., 2007; Huang et al., 2009; Trouiller et al., 2009; Petković et al., 2011; Wilhelmi et al., 2012). These papers note that the genotoxicity induced by TiO₂ nanoparticles in vivo or in vitro includes DNA double strand breaks, DNA fragmentation, disturbance of cell cycle progression, induction of a duplicated genome, and mis-segregation of chromosomes, which leads to chromosomal instability and cell transformation. In addition, aNTiO₂ strongly induces intracellular reactive oxygen species (ROS) (Gurr et al., 2005; Wang et al., 2007; Shukla et al., 2011). According to a recent review paper (Kirkland et al., 2022), the ability of TiO₂ NPs to inflict DNA damage has been observed on numerous occasions and is thought to be driven by secondary, non-genotoxic mechanisms, such as particle mediated ROS production. Existing evidence does not therefore support a direct DNA damaging mechanism for TiO₂ (nano and other forms), and we strongly speculate that the genotoxicity of TiO₂ nanoparticles is likely due to secondary mechanisms induced by inflammation and/or oxidative stress. This type of carcinogen will have a practical threshold (Bolt and Hermann, 2008; Bolt, 2016; Fukushima et al., 2010, 2016; Safe Work Australia, 2018). Consequently, risk assessment can be performed based on the carcinogenicity threshold and a No Observed Adverse Effect Level (NOAEL) is applicable for calculation. In the present study, since the incidence of tumors in the 8 mg/m³ group was not statistically significant but was not negligible when compared with the controls, 8 mg/m³ aNTiO₂ is considered to have carcinogenic risk, and 2 mg/m³ is the No Observed Adverse Effect Concentration (NOAEC) for carcinogenicity. This suggests that the NOAEC for carcinogenicity of aNTiO₂ in rats lies between 2 mg/m³ and 8 mg/m³. Using the NOAEL-based approach, the NOAEC for human workers is estimated to be 0.027 to 0.107 mg/m³. This was value is obtained by dividing 2 mg/m³ and 8 mg/m³ by 75, where 75 is calculated using a species difference uncertainty factor of 10 (the default value of species difference uncertainty factors) and a carcinogenic severity of 10 and an exposure time of 6 hours per day, 5 days per week for rats, and 8 hours/day, 5 day/week for humans (10 × 10 × 6/8).

The American Conference of Governmental Industrial Hygienists (ACGIH) has set a threshold (TLV) limit of TiO₂ particle (nanoscale particles) concentration of 0.2 mg/m³ for a time-weighted average (TWA: calculated as 8 hr/day and 40 hr/week) (ACGIH, 2021). In November 2005, the National Institute for Occupational Safety and Health (NIOSH) proposed a recommended exposure limit (REL) for TiO₂ NP of 0.3 mg/m³ (NIOSH, 2011). The Japan Society for Occupational Health (JSOH) recommended an REL of 0.3 mg/m³ for TiO₂ NP (Sangyo Eiseigaku Zasshi, 2013). The values we calculated for the NOAEC for lung cancer development, 0.027 to 0.107 mg/m³, are much lower than those recommended by Japanese and U.S. professional organizations.

In conclusion, the carcinogenicity and chronic toxicity of anatase-type nano-titanium dioxide (aNTiO₂) was investigated using an aerosol generator based on dry aerosolization using the aerodynamic cyclone principle. Exposures were carried out at concentrations of 0, 0.5, 2, and 8 mg/m³. In addition to the usual 2-year carcinogenicity study, satellite groups were set up that included three recovery periods after 1 year of exposure: recovery periods were 1 day, 26 weeks, and 52 weeks after the end of aNTiO₂ exposure. The study was carried out in accordance with GLP principles and OECD test guidelines.

Bronchiolo-alveolar carcinoma was observed in two males exposed to 8 mg/m³ aNTiO₂ for 104 weeks. The incidence showed an increasing trend by the Peto's test. However, this incidence was at the upper limit of JBRC’s historical control data. In females exposed to aNTiO₂ for 104 weeks, bronchiolo-alveolar adenomas were observed in 1, 2, 3, and 4 rats in the 0 (control), 0.5, 2, 8 mg/m³ groups, respectively. Although the incidence was not statistically significant, the incidence in 8 mg/m³-exposed females exceeds the JBRC historical control data.

Chronic toxicity was evidenced by multiple white-spots and patches, particle-induced alveolar epithelial hyperplasia, and focal fibrosis in the lungs of rats exposed to 0.5, 2, and 8 mg/m³ aNTiO₂, in both the carcinogenicity and satellite groups. In the satellite groups, focal fibrosis was observed only in the recovery period 3 groups (52 weeks observation after end of exposure). Increased lung weight and neutrophilic inflammation of the BALF were observed in the 8 mg/m³ carcinogenicity groups. Lung burden analysis of the satellite groups indicated that the aNTiO₂ in the lungs decreased after the end of exposure from recovery period 1 (1-day recovery after the end of exposure) to recovery period 2 (26 weeks recovery after end of exposure). However, the aNTiO₂ that remained deposited in the lungs after the 26 weeks recovery period did not decrease from week 26 to week 52 after the end of exposure.

Based on these data, we conclude that aNTiO₂ showed equivocal evidence of lung carcinogenicity in male and female F344 rats under possible lung particle overload conditions, but conditions that fulfill the MTD criteria. Increased lung weight and neutrophilic inflammation were observed in the 8 mg/m³-exposed groups. Therefore, aNTiO₂ is toxic to the lungs of exposed rats and may represent a toxic hazard to humans.

ACKNOWLEDGMENT

The present study was contracted and supported by the Ministry of Health, Labour and Welfare, Japan. We would like to thank Dr. Shoji Fukushima, former director of our institute and project leader of the aNTiO₂ inhalation research project at the time, for his insightful advice. We wish to thank Dr. David B. Alexander (Nanotoxicology Project, Nagoya City University) for his valuable comments and advice. We also thank all members of our laboratory, especially Kyohei Koizumi, Hirofumi Fujita, and Masahiro Yamamoto for their technical support.

Conflict of interest

The authors declare that there is no conflict of interest.

REFERENCES

• ACGIH. (1992): Threshold limit values and biological exposure indices for 1992–1993: American Conference of Governmental industrial hygienists, Cincinnati,Ohio.
• ACGIH. (2021): Titanium dioxide. In Documentation of the threshold limit values for chemical substances. American Conference of Governmental Industrial Hygienists, Cincinnati, Ohio.
•  Ağçeli,  G.,  Hammachi,  H.,  Kodal,  S.P., et al. (2020): Novel Approach to Synthesize TiO₂ Nanoparticles: Biosynthesis by Using Streptomyces sp. HC1. J. Inorg. Organomet. Polym. Mater., 3221-3229.
• Bannasch, P., Griesemer, R.A., Anders, F., et al. (1986): Long-term assays for carcinogenicity in animals. In: Long-term and short-term assays for carcinogens (Montesano, R., Bartsch, H., Vainio H. et al., ed.), pp 13–83, IARC Sci Pub, Lyon.
•  Bhattacharya,  K.,  Davoren,  M.,  Boertz,  J.,  Schins,  R.P.,  Hoffmann,  E. and  Dopp,  E. (2009): Titanium dioxide nanoparticles induce oxidative stress and DNA-adduct formation but not DNA-breakage in human lung cells. Part. Fibre Toxicol., 6, 17.
•  Bide,  R.W.,  Armour,  S.J. and  Yee,  E. (2000): Allometric respiration/body mass data for animals to be used for estimates of inhalation toxicity to young adult humans. J. Appl. Toxicol., 20, 273-290.
•  Bolt,  H.M. and  Hermann,  M. (2008): The concept of “practical thresholds” in the derivation of occupational exposure limits for carcinogens by the scientific committee on occupational exposure limits (SCOEL) of the European Union. Genes Environ., 30, 114-119.
• Bolt, H.M. (2016): Practical thresholds in the derivation of occupational exposure limits (OELs) for carcinogens. In: Thresholds of genotoxic carcinogens of DNA-reactive compounds. (Nohmi, T. and Fukushima, S., ed.) pp. 117-126. Elesevier Publisher, Amsterdam.
•  Borm,  P.J.,  Höhr,  D.,  Steinfartz,  Y.,  Zeitträger,  I. and  Albrecht,  C. (2000): Chronic inflammation and tumor formation in rats after intratracheal instillation of high doses of coal dusts, titanium dioxides, and quartz. Inhal. Toxicol., 12 (Suppl 3), 225-231.
•  Brown,  D.M.,  Wilson,  M.R.,  MacNee,  W.,  Stone,  V. and  Donaldson,  K. (2001): Size-dependent proinflammatory effects of ultrafine polystyrene particles: a role for surface area and oxidative stress in the enhanced activity of ultrafines. Toxicol. Appl. Pharmacol., 175, 191-199.
•  Carter,  C.L.,  Jones,  J.W.,  Farese,  A.M.,  MacVittie,  T.J. and  Kane,  M.A. (2017): Lipidomic dysregulation within the lung parenchyma following whole-thorax lung irradiation: markers of injury, inflammation and fibrosis detected by MALDI-MSI. Sci. Rep., 7, 10343.
• Court of Justice of the European Union (CURIA). (2022): Press release No 190/22. Luxembourg, 23, November. https://curia.europa.eu/jcms/upload/docs/application/pdf/2022-11/cp220190en.pdf.
•  Driscoll,  K.E. and  Borm,  P.J. (2020): Expert workshop on the hazards and risks of poorly soluble low toxicity particles. Inhal. Toxicol., 32, 53-62.
•  Driscoll,  K.E. (2022): Review of lung particle overload, rat lung cancer, and the conclusions of the edinburgh expert panel—it’s time to revisit cancer hazard classifications for titanium dioxide and carbon black. Front. Public Health, 10, 907318.
• ECHA. (2017): RAC committee for risk assessment. Opinion proposing harmonised classification and labelling at EU level of titanium dioxide. EC Number: 236-675-5 CAS Number: 13463-67-7, CLH-O-0000001412-86-163/F. https://www.mhlw.go.jp/content/11201000/000358565.pdf.
• Fukushima, S., Wei, M., Kakehashi, A., et al. (2010): Threshold for genotoxic carcinogen: Evidence from mechanism-based carcinogenicity studies. Cancer risk assessment. https://doi:10.1002/9780470622728.ch8.
• Fukushima, S., Gi, M., Kakehashi, A., et al. (2016): Qualitative and quantitative assessment on loe-dose carcinogenicity of genotoxic hepatocarcinogens: Dose-response for key events in rat hepatocarcinogenesis. In: Thresholuds of genotoxic carcinogens of DNA-reactive compounds (Nohmi, T. and Fukushima, S. ed.), pp. 1-18, Elesevier Pub, Amsterdam.
•  Gurr,  J.R.,  Wang,  A.S.,  Chen,  C.H. and  Jan,  K.Y. (2005): Ultrafine titanium dioxide particles in the absence of photoactivation can induce oxidative damage to human bronchial epithelial cells. Toxicology, 213, 66-73.
•  Haseman,  J.K.,  Huff,  J. and  Boorman,  G.A. (1984): Use of historical control data in carcinogenicity studies in rodents. Toxicol. Pathol., 12, 126-135.
•  Haseman,  J.K. (1985): Issues in carcinogenicity testing: dose selection. Fundam. Appl. Toxicol., 5, 66-78.
•  Haseman,  J.K. (1995): Data analysis: statistical analysis and use of historical control data. Regul. Toxicol. Pharmacol., 21, 52-59, discussion, 81-86.
•  Heinrich,  U.,  Fuhst,  R.,  Rittinghausen,  S., et al. (1995): Chronic inhalation exposure of wistar rats and two different strains of mice to diesel engine exhaust, carbon black, and titanium dioxide. Inhal. Toxicol., 7, 533-556.
•  Henderson,  R.F. (1984): Use of bronchoalveolar lavage to detect lung damage. Environ. Health Perspect., 56, 115-129.
•  Huang,  S.,  Chueh,  P.J.,  Lin,  Y.W.,  Shih,  T.S. and  Chuang,  S.M. (2009): Disturbed mitotic progression and genome segregation are involved in cell transformation mediated by nano-TiO₂ long-term exposure. Toxicol. Appl. Pharmacol., 241, 182-194.
• IARC. (2010): Carbon black, titanium dioxide, and talc. IARC Monographs on the evaluation of carcinogenic risks to humans. IARC Scientific Publications., 93, LYON, FRANCE.
•  Iijima,  Y.,  Tateishi,  Y.,  Tsuchiya,  K., et al. (2020): Pneumoconiosis caused by inhalation of metallic titanium grindings. Intern. Med., 59, 425-428.
• ILSI Risk Science Institute Workshop Participants (2000): The relevance of the rat lung response to particle overload for human risk assessment: a workshop consensus report. Inhal. Toxicol., 12, 1-17.
•  Jaeger,  A.,  Weiss,  D.G.,  Jonas,  L. and  Kriehuber,  R. (2012): Oxidative stress-induced cytotoxic and genotoxic effects of nano-sized titanium dioxide particles in human HaCaT keratinocytes. Toxicology, 296, 27-36.
• Japan Society for Occupational Health. (2013): Sangyo Eisegaku Zasshi (In Japanese)., 55. https://www.nite.go.jp/chem/chrip/chrip_search/dt/pdf/CI_04_002/OEL_13463677.pdf
•  Kaplan,  E.L. and  Meier,  P. (1958): Nonparametric estimation from incomplete observations. J. Am. Stat. Assoc., 53, 457-481.
•  Kasai,  T.,  Gotoh,  K.,  Nishizawa,  T.,  Sasaki,  T.,  Katagiri,  T.,  Umeda,  Y.,  Toya,  T. and  Fukushima,  S. (2014): Development of a new multi-walled carbon nanotube (MWCNT) aerosol generation and exposure system and confirmation of suitability for conducting a single-exposure inhalation study of MWCNT in rats. Nanotoxicology, 8, 169-178.
•  Kasai,  T.,  Umeda,  Y.,  Ohnishi,  M.,  Mine,  T.,  Kondo,  H.,  Takeuchi,  T.,  Matsumoto,  M. and  Fukushima,  S. (2016): Lung carcinogenicity of inhaled multi-walled carbon nanotube in rats. Part. Fibre Toxicol., 13, 53.
•  Kirkland,  D.,  Aardema,  M.J.,  Battersby,  R.V.,  Beevers,  C.,  Burnett,  K.,  Burzlaff,  A.,  Czich,  A.,  Donner,  E.M.,  Fowler,  P.,  Johnston,  H.J.,  Krug,  H.F.,  Pfuhler,  S. and  Stankowski,  L.F. Jr. (2022): A weight of evidence review of the genotoxicity of titanium dioxide (TiO₂). Regul. Toxicol. Pharmacol., 136, 105263.
•  Lee,  K.P.,  Trochimowicz,  H.J. and  Reinhardt,  C.F. (1985): Pulmonary response of rats exposed to titanium dioxide (TiO₂) by inhalation for two years. Toxicol. Appl. Pharmacol., 79, 179-192.
•  Lippmann,  M.,  Yeates,  D.B. and  Albert,  R.E. (1980): Deposition, retention, and clearance of inhaled particles. Br. J. Ind. Med., 37, 337-362.
•  Lu,  P.J.,  Ho,  I.C. and  Lee,  T.C. (1998): Induction of sister chromatid exchanges and micronuclei by titanium dioxide in Chinese hamster ovary-K1 cells. Mutat. Res., 414, 15-20.
•  Määttä,  K.,  Elo,  R.,  Arstila,  A. and  Uksila,  E. (1971): Pulmonary changes induced by titanium dioxide. Duodecim, 87, 1435-1444.
•  Ma-Hock,  L.,  Burkhardt,  S.,  Strauss,  V.,  Gamer,  A.O.,  Wiench,  K.,  van Ravenzwaay,  B. and  Landsiedel,  R. (2009): Development of a short-term inhalation test in the rat using nano-titanium dioxide as a model substance. Inhal. Toxicol., 21, 102-118.
•  McIntyre,  R.A. (2012): Common nano-materials and their use in real world applications. Sci. Prog., 95, 1-22.
• Ministry of Labour, Japan (2000): Standards to be observed by testing institutions of notification No.76 of the Ministry of Labour, Japan, September 1, 1988 (amendment: Notification No.13 of the Ministry of Labour, Japan, March 29, 2000) (in Japanese).
• Ministry of the Environment. (2013): Standards relating to the care and keeping and reducing pain of laboratory animals (Notice of the Ministry of the Environment No. 88 of 2006). Latest revision: Notice of the Ministry of the Environment No. 84 of 2013. https://www.env.go.jp/nature/dobutsu/aigo/2_data/laws/nt_h25_84_en.pdf (accessed. 2023.11.7)
•  Moran,  C.A.,  Mullick,  F.G.,  Ishak,  K.G.,  Johnson,  F.B. and  Hummer,  W.B. (1991): Identification of titanium in human tissues: probable role in pathologic processes. Hum. Pathol., 22, 450-454.
• Muhle, H., Creutzenberg, O. and Bellmann, B. (1990): Pulmonary dust overload: an investigation of various materials, species differences, and irreversibility of effects. J. Aerosol Med., 3, Supplement 1.
•  Muhle,  H.,  Bellmann,  B.,  Creutzenberg,  O.,  Dasenbrock,  C.,  Ernst,  H., et al. (1991): Pulmonary response to toner upon chronic inhalation exposure in rats. Fundam. Appl. Toxicol., 17, 280-299.
•  Nagano,  K.,  Katagiri,  T.,  Aiso,  S.,  Senoh,  H.,  Sakura,  Y. and  Takeuchi,  T. (1997): Spontaneous lesions of nasal cavity in aging F344 rats and BDF1 mice. Exp. Toxicol. Pathol., 49, 97-104.
•  Nakata,  K. and  Fujishima,  A. (2012): TiO₂ photocatalysis: design and applications. J. Photochem. Photobiol., C, 169-189.
• NIOSH. (2011): Occupational exposure to titanium dioxide. NIOSH current intelligence bulletin 63. https://www.cdc.gov/niosh/docs/2011-160/pdfs/2011-160.pdf.
• Noël1, A., Charbonneau, M., Cloutier, Y., et al. (2013): Rat pulmonary responses to inhaled nano-TiO2: effect of primary particle size and agglomeration state. Part. Fibre. Toxicol., 10, 48.
•  Oberdörster,  G.,  Ferin,  J.,  Gelein,  R.,  Soderholm,  S.C. and  Finkelstein,  J. (1992): Role of the alveolar macrophage in lung injury: studies with ultrafine particles. Environ. Health Perspect., 97, 193-199.
•  Oberdörster,  G. (1995): Lung particle overload: implications for occupational exposures to particles. Regul. Toxicol. Pharmacol., 21, 123-135.
•  Ohno,  S.,  Hagiwara,  S.,  Kobayashi,  J.,  Sugiyama,  Y.,  Kitamura,  S.,  Kanai,  N. and  Saitoh,  K. (1996): [Tonoko pneumoconiosis with deposition of titanium]. Nihon Kyobu Shikkan Gakkai Zasshi, 34, 1234-1238.
• Onodera, T., Nakamura, M., Sato, T., et al. (1983): Biochemical characterization of pulmonary washings of patients with alveolar proteinosis, interstitial pneumonitis and alveolar cell carcinoma. PMID: 6304935 DOI: 10.1620/tjem.139.245.
• Organisation for Economic Co-operation and Development (OECD). (1998): OECD principles of good laboratory practice.as revised 1997. OECD., Paris.
• Organisation for Economic Co-operation and Development (OECD). (2009): OECD Guideline for testing of chemicals 451 for “Carcinogenicity studies” Adopted Sep. 2009. OECD., Paris.
• Peto, R., Pike, M.C., Day, N.E., et al. (1980): Guidelines for simple, sensitive significance tests for carcinogenic effects in long-term animal experiments. In: Long-term and short-term screening assays for carcinogens: A critical appraisal, IARC Monographs, Suppl 2, pp. 311–426, IARC, Lyon.
•  Pott,  F. and  Roller,  M. (2005): Carcinogenicity study with nineteen granular dusts in rats. Eur. J. Oncol., 10, 249-281.
•  Petković,  J.,  Zegura,  B.,  Stevanović,  M.,  Drnovšek,  N.,  Uskoković,  D.,  Novak,  S. and  Filipič,  M. (2011): DNA damage and alterations in expression of DNA damage responsive genes induced by TiO₂ nanoparticles in human hepatoma HepG2 cells. Nanotoxicology, 5, 341-353.
•  Redline,  S.,  Barna,  B.P.,  Tomashefski,  J.F. Jr. and  Abraham,  J.L. (1986): Granulomatous disease associated with pulmonary deposition of titanium. Br. J. Ind. Med., 43, 652-656.
•  Renwick,  L.C.,  Brown,  D.,  Clouter,  A. and  Donaldson,  K. (2004): Increased inflammation and altered macrophage chemotactic responses caused by two ultrafine particle types. Occup. Environ. Med., 61, 442-447.
•  Rode,  L.E.,  Ophus,  E.M. and  Gylseth,  B. (1981): Massive pulmonary deposition of rutile after titanium dioxide exposure: light-microscopical and physico-analytical methods in pigment identification. Acta Pathol. Microbiol. Scand. [A], 89, 455-461.
• Safe Work Australia. (2018): Non-threshold based genotoxic carcinogens Accessory document to Recommending health-based workplace exposure standards and notations, Australian workplace exposure standards and advisory notations., Australia. https://www.safeworkaustralia.gov.au/system/files/documents/1806/non-threshold_based_genotoxic_carcinogens_-_accessory_document_to_recommending_health-based_workplace_exposure_standards_and_notations.pdf
•  Sager,  T.M.,  Kommineni,  C. and  Castranova,  V. (2008): Pulmonary response to intratracheal instillation of ultrafine versus fine titanium dioxide: role of particle surface area. Part. Fibre Toxicol., 5, 17.
• Science Council of Japan. (2006): Guideline for proper conduct of Animal Experiments. http://www.scj.go.jp/info/kohyo/pdf/kohyo-20-k16-2 (in Japanese).
•  Shukla,  R.K.,  Sharma,  V.,  Pandey,  A.K.,  Singh,  S.,  Sultana,  S. and  Dhawan,  A. (2011): ROS-mediated genotoxicity induced by titanium dioxide nanoparticles in human epidermal cells. Toxicol. In Vitro, 25, 231-241.
• Sontag, M., Page, N.P. and Saffiotti, U. (1976): Guidelines for carcinogen bioassay in small rodents. NCI-CG-TR-1. DHEW, Washington DC. https://ntp.niehs.nih.gov/sites/default/files/ntp/htdocs/lt_rpts/tr001.pdf
•  Sun,  Q.,  Tan,  D.,  Zhou,  Q.,  Liu,  X.,  Cheng,  Z.,  Liu,  G.,  Zhu,  M.,  Sang,  X.,  Gui,  S.,  Cheng,  J.,  Hu,  R.,  Tang,  M. and  Hong,  F. (2012): Oxidative damage of lung and its protective mechanism in mice caused by long-term exposure to titanium dioxide nanoparticles. J. Biomed. Mater. Res. A, 100, 2554-2562.
•  Trouiller,  B.,  Reliene,  R.,  Westbrook,  A.,  Solaimani,  P. and  Schiestl,  R.H. (2009): Titanium dioxide nanoparticles induce DNA damage and genetic instability in vivo in mice. Cancer Res., 69, 8784-8789.
•  Wang,  J.J.,  Sanderson,  B.J. and  Wang,  H. (2007): Cyto- and genotoxicity of ultrafine TiO₂ particles in cultured human lymphoblastoid cells. Mutat. Res., 628, 99-106.
•  Wilhelmi,  V.,  Fischer,  U.,  van Berlo,  D.,  Schulze-Osthoff,  K.,  Schins,  R.P. and  Albrecht,  C. (2012): Evaluation of apoptosis induced by nanoparticles and fine particles in RAW 264.7 macrophages: facts and artefacts. Toxicol. In Vitro, 26, 323-334.
•  Winter,  M.,  Beer,  H.D.,  Hornung,  V.,  Krämer,  U.,  Schins,  R.P. and  Förster,  I. (2011): Activation of the inflammasome by amorphous silica and TiO₂ nanoparticles in murine dendritic cells. Nanotoxicology, 5, 326-340.
•  Xue,  C.,  Wu,  J.,  Lan,  F.,  Liu,  W.,  Yang,  X.,  Zeng,  F. and  Xu,  H. (2010): Nano titanium dioxide induces the generation of ROS and potential damage in HaCaT cells under UVA irradiation. J. Nanosci. Nanotechnol., 10, 8500-8507.
•  Yamadori,  I.,  Ohsumi,  S. and  Taguchi,  K. (1986): Titanium dioxide deposition and adenocarcinoma of the lung. Acta Pathol. Jpn., 36, 783-790.
•  Yamano,  S.,  Goto,  Y.,  Takeda,  T.,  Hirai,  S.,  Furukawa,  Y.,  Kikuchi,  Y.,  Kasai,  T.,  Misumi,  K.,  Suzuki,  M.,  Takanobu,  K.,  Senoh,  H.,  Saito,  M.,  Kondo,  H. and  Umeda,  Y. (2022): Pulmonary dust foci as rat pneumoconiosis lesion induced by titanium dioxide nanoparticles in 13-week inhalation study. Part. Fibre Toxicol., 19, 58.
•  Yin,  J.J.,  Liu,  J.,  Ehrenshaft,  M.,  Roberts,  J.E.,  Fu,  P.P.,  Mason,  R.P. and  Zhao,  B. (2012): Phototoxicity of nano titanium dioxides in HaCaT keratinocytes--generation of reactive oxygen species and cell damage. Toxicol. Appl. Pharmacol., 263, 81-88.
•  Yoo,  K.C.,  Yoon,  C.H.,  Kwon,  D.,  Hyun,  K.H.,  Woo,  S.J.,  Kim,  R.K.,  Lim,  E.J.,  Suh,  Y.,  Kim,  M.J.,  Yoon,  T.H. and  Lee,  S.J. (2012): Titanium dioxide induces apoptotic cell death through reactive oxygen species-mediated Fas upregulation and Bax activation. Int. J. Nanomedicine, 7, 1203-1214.
•  Ziental,  D.,  Czarczynska-Goslinska,  B.,  Mlynarczyk,  D.T.,  Glowacka-Sobotta,  A.,  Stanisz,  B.,  Goslinski,  T. and  Sobotta,  L. (2020): Titanium dioxide nanoparticles: prospects and applications in medicine. Nanomaterials (Basel), 10, 387.