Titanium Dioxide Particle Type and Concentration Influence the Inflammatory Response in Caco-2 Cells

Titanium dioxide (TiO2) nanoparticles are widely used in cosmetics, sunscreens, biomedicine, and food products. When used as a food additive, TiO2 nanoparticles are used in significant amounts as white food-coloring agents. However, the effects of TiO2 nanoparticles on the gastrointestinal tract remain unclear. The present study was designed to determine the effects of five TiO2 particles of different crystal structures and sizes in human epithelial colorectal adenocarcinoma (Caco-2) cells and THP-1 monocyte-derived macrophages. Twenty-four-hour exposure to anatase (primary particle size: 50 and 100 nm) and rutile (50 nm) TiO2 particles reduced cellular viability in a dose-dependent manner in THP-1 macrophages, but in not Caco-2 cells. However, 72-h exposure of Caco-2 cells to anatase (50 nm) TiO2 particles reduced cellular viability in a dose-dependent manner. The highest dose (50 µg/mL) of anatase (100 nm), rutile (50 nm), and P25 TiO2 particles also reduced cellular viability in Caco-2 cells. The production of reactive oxygen species tended to increase in both types of cells, irrespective of the type of TiO2 particle. Exposure of THP-1 macrophages to 50 µg/mL of anatase (50 nm) TiO2 particles increased interleukin (IL)-1β expression level, and exposure of Caco-2 cells to 50 µg/mL of anatase (50 nm) TiO2 particles also increased IL-8 expression. The results indicated that anatase TiO2 nanoparticles induced inflammatory responses compared with other TiO2 particles. Further studies are required to determine the in vivo relevance of these findings to avoid the hazards of ingested particles.


Introduction
Engineered nanoparticles (NPs), defined as particles with diameters of less than 100 nm, exhibit new physicochemical features at the nanoscale, such as large surface area, altered electronic properties,

Characterization of Suspensions of Titanium Dioxide (TiO 2 ) Particles
We selected 50 nm of anatase and rutile TiO 2 particles as the nano-sized TiO 2 particles and 100 nm of anatase and 250 nm of rutile TiO 2 particles as the large-sized TiO 2 particles. We also used TiO 2 NPs (Degussa, P25; 21 nm), which are the standard materials in the field of photocatalytic reactions, contain anatase and rutile phases in a ratio of about 4:1 w/w. Nano-sized TiO 2 particles were dispersed in the each culture medium of THP-1 and Caco-2 cells. The intensity-weighted hydrodynamic average diameter (z-average) of dispersed NPs was measured by the dynamic light scattering (DLS) technology, as described previously [14,15]. Table 1 shows the mean hydrodynamic diameters, polydispersity index (PdI), and ζ potential of dispersed TiO 2 particles in each medium. The DLS data of TiO 2 NPs indicated that the mean hydrodynamic diameter was >150 nm and confirmed the presence of nano-sized particles in the medium (Figure 1). The ζ potential of all particles in both cell media ranged from´11 to´14 mV (Table 1). No association between size or crystal structures and electrophoretic mobility of the particles was found.

Effects of Exposure on Cell Viability
THP-1 macrophages and Caco-2 cells were exposed to TiO2 particles at a concentration ranging from 1 to 50 µg/mL for 24 or 72 h.

Effects of Exposure on Cell Viability
THP-1 macrophages and Caco-2 cells were exposed to TiO 2 particles at a concentration ranging from 1 to 50 µg/mL for 24 or 72 h.

Effects of Exposure on Accumulation of Reactive Oxygen Species (ROS)
ROS production occurs as the initial cellular response to foreign materials and the maximum ROS levels were found at 2-6-h exposure to TiO 2 NPs by a previous in vitro study [16]. Therefore, we examined the effects of 3-h exposure of THP-1 macrophages and Caco-2 cells to TiO 2 particles (25 and 50 µg/mL) on ROS production. Exposure to each type of TiO 2 particles significantly increased ROS levels in THP-1 macrophages ( Figure 5A), especially A50-exposed cells, and Caco-2 cells ( Figure 5B).

Effects of Exposure on Accumulation of Reactive Oxygen Species (ROS)
ROS production occurs as the initial cellular response to foreign materials and the maximum ROS levels were found at 2-6-h exposure to TiO2 NPs by a previous in vitro study [16]. Therefore, we examined the effects of 3-h exposure of THP-1 macrophages and Caco-2 cells to TiO2 particles (25 and 50 µg/mL) on ROS production. Exposure to each type of TiO2 particles significantly increased ROS levels in THP-1 macrophages ( Figure 5A), especially A50-exposed cells, and Caco-2 cells ( Figure 5B).

Effects of Exposure on Interleukin (IL)-1β Levels in THP-1 Macrophages
Increased production of inflammatory cytokine, IL-1β, was noted in THP-1 macrophages after 24-h exposure to A50, and such increase was dose-dependent, and the increase was significant at 50 µg/mL of A50 ( Figure 6). However, exposure to other TiO2 particles had no significant effect on IL-1β level.

Effects of Exposure on Expression of IL-8 in Colorectal Adenocarcinoma (Caco-2) Cells
We measured IL-8 expression in Caco-2 cells after 3-and 6-h exposure to particles because previous reports showed that inflammatory cytokine expression induced by NPs was detected after 1-6-h exposure [17,18]. There were no significant changes in IL-8 mRNA expression level in Caco-2 cells following 3-h exposure, irrespective of the type of TiO2 particle (25 or 50 µg/mL) (data not shown). However, exposure to 50 µg/mL of A50 for 6 h significantly increased IL-8 mRNA expression in Caco-2 cells (Figure 7).

Effects of Exposure on Interleukin (IL)-1β Levels in THP-1 Macrophages
Increased production of inflammatory cytokine, IL-1β, was noted in THP-1 macrophages after 24-h exposure to A50, and such increase was dose-dependent, and the increase was significant at 50 µg/mL of A50 ( Figure 6). However, exposure to other TiO 2 particles had no significant effect on IL-1β level.

Effects of Exposure on Accumulation of Reactive Oxygen Species (ROS)
ROS production occurs as the initial cellular response to foreign materials and the maximum ROS levels were found at 2-6-h exposure to TiO2 NPs by a previous in vitro study [16]. Therefore, we examined the effects of 3-h exposure of THP-1 macrophages and Caco-2 cells to TiO2 particles (25 and 50 µg/mL) on ROS production. Exposure to each type of TiO2 particles significantly increased ROS levels in THP-1 macrophages ( Figure 5A), especially A50-exposed cells, and Caco-2 cells ( Figure 5B).

Effects of Exposure on Interleukin (IL)-1β Levels in THP-1 Macrophages
Increased production of inflammatory cytokine, IL-1β, was noted in THP-1 macrophages after 24-h exposure to A50, and such increase was dose-dependent, and the increase was significant at 50 µg/mL of A50 ( Figure 6). However, exposure to other TiO2 particles had no significant effect on IL-1β level.

Effects of Exposure on Expression of IL-8 in Colorectal Adenocarcinoma (Caco-2) Cells
We measured IL-8 expression in Caco-2 cells after 3-and 6-h exposure to particles because previous reports showed that inflammatory cytokine expression induced by NPs was detected after 1-6-h exposure [17,18]. There were no significant changes in IL-8 mRNA expression level in Caco-2 cells following 3-h exposure, irrespective of the type of TiO2 particle (25 or 50 µg/mL) (data not shown). However, exposure to 50 µg/mL of A50 for 6 h significantly increased IL-8 mRNA expression in Caco-2 cells (Figure 7).

Effects of Exposure on Expression of IL-8 in Colorectal Adenocarcinoma (Caco-2) Cells
We measured IL-8 expression in Caco-2 cells after 3-and 6-h exposure to particles because previous reports showed that inflammatory cytokine expression induced by NPs was detected after 1-6-h exposure [17,18]. There were no significant changes in IL-8 mRNA expression level in Caco-2 cells following 3-h exposure, irrespective of the type of TiO 2 particle (25 or 50 µg/mL) (data not shown). However, exposure to 50 µg/mL of A50 for 6 h significantly increased IL-8 mRNA expression in Caco-2 cells (Figure 7).

Discussion
In the present study, we examined the effects of exposure to different crystal structures and sizes of TiO2 particles in Caco-2 cells and THP-1 monocyte-derived macrophages. Our results indicated that anatase TiO2 NPs induced inflammatory responses compared with other TiO2 particles.
Although the primary particle size of TiO2 particles is around 200-300 nm, smaller TiO2 NPs measuring 1-50 nm are currently used for the purpose of ultraviolet (UV) protection and photocatalytic activity. Previous studies examined the effects of inhalation exposure on inflammation and showed that ultrafine particles induced stronger inflammatory responses than fine particles [19]. On the other hand, pulmonary instillation studied showed that nanoscale particle types of TiO2 were not more cytotoxic or inflammogenic to the lung compared with larger sized particles of similar composition [20]. Other studies showed that inhalation of rutile ultrafine-TiO2 particles was less likely to be associated with adverse pulmonary health effects compared with anatase ultrafine-TiO2 particles [21]. Sayer et al. [22] showed that anatase TiO2 particles were 100 times more toxic than an equivalent sample of rutile TiO2 particles in human dermal fibroblasts and human lung epithelial cells. However, these studies focused mainly on the pulmonary toxicity of TiO2 NPs following intratracheal instillation and inhalation. Since TiO2 NPs have also been used recently as a white pigment and as a food additive for food coloring, determination of the effects of TiO2 NPs on the intestine is urgently needed for safety assessment of these particles.
The present study investigated the effects of exposure to different sizes of anatase and rutile TiO2 particles. The MTS assay showed that incubation of THP-1 macrophages in the presence of anatase TiO2 particles significantly reduced cell viability compared with rutile TiO2 particles. Moreover, incubation of Caco-2 cells in the presence of anatase TiO2 particles, especially anatase TiO2 NPs of primary particle size of 50 nm, significantly reduced cell viability after exposure for 72 h. We have recently reported the cytotoxicity of zinc oxide (ZnO) NPs, but not P25 TiO2 NPs, on endothelial cells [23]. The typical crystalline composition of P25 TiO2 NPs was around 80% anatase and 20% rutile [24]. The present study also showed that 24-h-exposure to P25 TiO2 NPs was not cytotoxic for both THP-1 macrophages and Caco-2 cells. The results indicate that anatase TiO2 NPs is more toxic than rutile TiO2 particles, suggesting that TiO2 particle toxicity in human intestinal cells depends on the particle size and crystalline structure.
Previous acute oral toxicity studies showed that TiO2 NPs had very low toxicity in animals [25,26]. Moreover, oral administration of TiO2 NPs showed low absorption and narrow range of organ distribution [27], but slow tissue elimination [28]. Although data using cultured cells are not a substitute for whole animal studies, the use of simple cell culture models with endpoints that can identify the mechanism of cellular responses or toxicity can be the basis for further assessment of the potential risk of material exposure. Previous cell culture studies showed that TiO2 NPs induced oxidative stress and increased IL-1β levels in murine dendritic cells [29]. Yazdi et al. [30] showed that TiO2 NPs activated the NLR pyrin domain containing 3 (Nlrp3) inflammasome, leading to IL-1β

Discussion
In the present study, we examined the effects of exposure to different crystal structures and sizes of TiO 2 particles in Caco-2 cells and THP-1 monocyte-derived macrophages. Our results indicated that anatase TiO 2 NPs induced inflammatory responses compared with other TiO 2 particles.
Although the primary particle size of TiO 2 particles is around 200-300 nm, smaller TiO 2 NPs measuring 1-50 nm are currently used for the purpose of ultraviolet (UV) protection and photocatalytic activity. Previous studies examined the effects of inhalation exposure on inflammation and showed that ultrafine particles induced stronger inflammatory responses than fine particles [19]. On the other hand, pulmonary instillation studied showed that nanoscale particle types of TiO 2 were not more cytotoxic or inflammogenic to the lung compared with larger sized particles of similar composition [20]. Other studies showed that inhalation of rutile ultrafine-TiO 2 particles was less likely to be associated with adverse pulmonary health effects compared with anatase ultrafine-TiO 2 particles [21]. Sayer et al. [22] showed that anatase TiO 2 particles were 100 times more toxic than an equivalent sample of rutile TiO 2 particles in human dermal fibroblasts and human lung epithelial cells. However, these studies focused mainly on the pulmonary toxicity of TiO 2 NPs following intratracheal instillation and inhalation. Since TiO 2 NPs have also been used recently as a white pigment and as a food additive for food coloring, determination of the effects of TiO 2 NPs on the intestine is urgently needed for safety assessment of these particles.
The present study investigated the effects of exposure to different sizes of anatase and rutile TiO 2 particles. The MTS assay showed that incubation of THP-1 macrophages in the presence of anatase TiO 2 particles significantly reduced cell viability compared with rutile TiO 2 particles. Moreover, incubation of Caco-2 cells in the presence of anatase TiO 2 particles, especially anatase TiO 2 NPs of primary particle size of 50 nm, significantly reduced cell viability after exposure for 72 h. We have recently reported the cytotoxicity of zinc oxide (ZnO) NPs, but not P25 TiO 2 NPs, on endothelial cells [23]. The typical crystalline composition of P25 TiO 2 NPs was around 80% anatase and 20% rutile [24]. The present study also showed that 24-h-exposure to P25 TiO 2 NPs was not cytotoxic for both THP-1 macrophages and Caco-2 cells. The results indicate that anatase TiO 2 NPs is more toxic than rutile TiO 2 particles, suggesting that TiO 2 particle toxicity in human intestinal cells depends on the particle size and crystalline structure.
Previous acute oral toxicity studies showed that TiO 2 NPs had very low toxicity in animals [25,26]. Moreover, oral administration of TiO 2 NPs showed low absorption and narrow range of organ distribution [27], but slow tissue elimination [28]. Although data using cultured cells are not a substitute for whole animal studies, the use of simple cell culture models with endpoints that can identify the mechanism of cellular responses or toxicity can be the basis for further assessment of the potential risk of material exposure. Previous cell culture studies showed that TiO 2 NPs induced 7 of 12 oxidative stress and increased IL-1β levels in murine dendritic cells [29]. Yazdi et al. [30] showed that TiO 2 NPs activated the NLR pyrin domain containing 3 (Nlrp3) inflammasome, leading to IL-1β release in murine and human macrophages and human keratinocytes. Moreover, comparison of IL-1β production in response to exposure to various engineered NPs showed that high concentration (500 µg/mL) of smaller anatase and larger rutile TiO 2 particles induced high production of IL-1β [31]. Another study showed that TiO 2 nanobelts, but not P25 or anatase TiO 2 , induced IL-1β in THP-1 cells [32]. Yazdi et al. [30] also demonstrated that chemical ROS scavenger diminished IL-1β secretion triggered by TiO 2 NPs in THP-1 cells, suggesting that ROS production induced inflammatory cytokine production after exposure to TiO 2 NPs. On the other hands, other previous reports showed that ROS was not essential for IL-1β production via the Nlrp3 inflammasome [33,34]. In the present study, the production of IL-1β was significantly increased in THP-1 macrophages after exposure to 50 µg/mL of anatase TiO 2 NPs. Since the level of ROS was also most elevated in THP-1 macrophages exposed to 50 µg/mL of anatase TiO 2 NPs compared with other particles, it seems that anatase TiO 2 nanoparticles induce inflammatory responses through accumulation of ROS in THP-1 macrophages. However, ROS might be not necessarily the main contributing factor of particles-induced IL-1β production in THP-1 macrophages because ROS level was increased after exposure to all TiO 2 particles.
Orally ingested NPs are uptaken by epithelial cells and M cells in Peyer's patches through the process of endocytosis, invasion by over-adsorption in cell gaps, and/or intrusion by passing through the tight junctions between cells [35]. Fine and ultrafine particles are potent adjuvants in antigen-mediated immune responses and are increasingly associated with inflammatory bowel diseases, such as Crohn's disease [36]. ZnO NPs have been shown to induce cytotoxicity associated with overproduction of ROS in Caco-2 cells [37,38]. ZnO NPs has also been reported to induce inflammatory responses and increase the release of IL-8 in Caco-2 cells [37,39]. The present study found IL-8 over-expression in Caco-2 cells exposed to anatase TiO 2 NPs for 6 h. Interestingly, a previous similar study showed that exposure to 10 µg/mL of P25 TiO 2 NPs for 24 h led to increased IL-8 production in Caco-2 cells [40]. On the other hand, De Angelis et al. [39] demonstrated the induction of IL-8 production after 6-h exposure to ZnO NPs, but not anatase TiO 2 NPs. The smaller mean hydrodynamic diameter of anatase TiO 2 NPs estimated in the present study, relative to that of the above study [40] could perhaps explain the differences between the two studies. In the present study, there was no correlation between the induction of IL-8 production and ROS production in Caco-2 cells after TiO 2 particles exposure. As shown in a previous report [41], oxidative stress induced by various NPs is an early event as the initial cellular response and ROS might not play a role in the impairment of inflammation-related pathway. TiO 2 NPs can induce nuclear factor (NF)-κB activity by subsequent degradation of inhibitor (I)κ-B in airway epithelial and endothelial cells [42,43]. The presence of a binding site for NF-κB in the promoter region of IL-8 and enhanced IL-8 transcription following NF-κB binding, suggests that IL-8 expression could be up-regulated through NF-κB activation following exposure to anatase TiO 2 NPs.
The present study indicated that anatase TiO 2 NPs induced inflammatory responses compared with other TiO 2 particles. However, Zijno et al. [44] recently compared the genotoxicity of TiO 2 and ZnO NPs and demonstrated that only ZnO NPs were genotoxic, including destruction of micronuclei and DNA damage, although both NPs produced ROS in Caco-2 cells. Moreover, native TiO 2 NPs and pretreated TiO 2 NPs with the digestion simulation fluid or bovine serum albumin did not show significant toxicity in both Caco-2 cells and Caco-2 monolayers [45]. Janer et al. [46] suggested that the Caco-2 monolayer system is likely to underestimate the effects of oral absorption of NPs due to the fact that NPs were observed in Peyer Patch cells in the oral absorption study. The development of safe and effective NPs is important for advancement of technology and for healthy lives. There is no doubt a need to elucidate the effects and mechanisms of TiO 2 NPs in the intestine using co-culture models, such as microfold (M) cells or intestinal epithelial cells ingested particles in vivo.

TiO 2 Particles Preparation and Characterization
The TiO 2 particles used in the present study were A50 (anatase, primary diameter: 50 nm) (mkNano, Mississauga, ON, Canada), A100 (anatase, primary diameter: 100 nm) (mkNano), R50 (rutile, primary diameter: 50 nm) (mkNano), R250 (rutile, primary diameter: 250 nm) (mkNano), and P25 (80% anatases/20% rutile, primary diameter: 21 nm) (Degussa, Germany). We characterized previously P25 TiO 2 NPs from the same lot by DLS as well as by transmission electron microscope (TEM, JEM-1011; JEOL, Tokyo, Japan), and then established a suitable protocol for the preparation of a suspension of TiO 2 NPs [14]. NPs were suspended in serum-containing culture media and dispersed using a sonicator (model 450, Branson Sonifier, Danbury, CT, USA) set at 80% pulsed mode, 100 W, and 15 min. The hydrodynamic size of the particles in the medium was measured four times after 1 h on standing using DLS technology with zetasizer Nano-S (Malvern Instruments, Worcestershire, UK). The dispersion status was described by the intensity-weighted hydrodynamic average diameter (z-average) and PdI, which reflect the broadness of the size distribution (scale range from 0 to 1, with 0 being monodispersion and 1 being polydispersion). To investigate the electrophoretic mobility of the particles, ζ potential of the particles in each medium was measured three times with Photal LEZA-600 (Otsuka Electronics, Tokyo, Japan).

Cell Viability Assay
THP-1 monocytes were seeded at 1.5ˆ10 4 cells/well on 96-well plates and differentiated to macrophages with PMA before the experiment as described above. Caco-2 cells were seeded overnight at 1.5ˆ10 4 cells per well on 96-well plates before the experiment. Particles were dispersed in each serum-containing cell culture medium at a final concentration ranging from 1 to 50 µg/mL. The previous studies demonstrated that TiO 2 NPs induced a pronounced inflammatory response at the concentration of 10-200 µg/mL in in vitro models of gut epithelium [47]. Also, the concentration range corresponded to the dose used in our previous study [23]. Cell viability was determined after incubation with the dispersed TiO 2 particles for 24 or 72 h, by MTS assay based on the CellTiter 96 AQueous One Solution (Promega, Madison, WI, USA), which measures mitochondrial function; the latter correlates with cell viability. The serum-containing cell culture medium was used during incubation with the particles. After the incubation, the cells were incubated with fresh medium (phenol red-free) containing MTS reagent for 1 h before measurements at an absorbance of 490 nm. The effect of particles on cell proliferation was expressed as percentage of inhibition of cell growth relative to the control.

Measurement of ROS Production
Cellular ROS production triggered by TiO 2 particles was assayed by staining with 5-(and-6)-chloromethyl-2 1 ,7 1 -dichlorodihydro fluorescein diacetate, acetyl ester (CM-H 2 DCFDA) (Life Technologies) followed by flow cytometry (FACS CantoII, BD Bioscience, Franklin Lakes, NJ, USA). Before the experiment, THP-1 monocytes were seeded at 3ˆ10 5 cells/well onto 24-well plates and allowed to differentiate into macrophages using PMA as described above. Caco-2 cells were seeded overnight at 3ˆ10 5 cells/well onto 24-well plates before the experiment. After exposure to TiO 2 particles for 3 h, THP-1 macrophages and Caco-2 cells (3ˆ10 5 cells) were loaded with 5 µM CM-H 2 DCFDA for 30 min at 37˝C and analyzed by flow cytometry. Ten thousand cells per sample were acquired in histograms using FlowJo software (Flowjo, Ashland, OR, USA). Dead cells and debris were excluded by electronic gating using forward and side scatter measurements.

Measurement of IL-1β Production
Before the experiment, THP-1 monocytes were seeded at 1.5ˆ10 4 cells/well onto 96-well plates and differentiated to macrophages using PMA as described above. THP-1 macrophages were exposed to 25 or 50 µg/mL of the suspended particles for 24 h. The cell culture medium was collected and centrifuged at 10,000ˆg to remove cell debris and suspended TiO 2 particles. The final supernatant was stored at´20˝C until cytokine analysis. The amount of IL-1β in the cell medium was measured using ELISA (Biolegend, San Diego, CA, USA) according to the protocol supplied by the manufacturer. Changes in color intensity were quantified by a plate reader (Bio-Rad Laboratories, Hercules, CA, USA).

Analysis of IL-8 Expression
Caco-2 cells (2ˆ10 5 cells) were seeded onto 12-well plates and exposed to 25 or 50 µg/mL of the suspended particles for 3 or 6 h. Total RNA from the cells was isolated by using ReliaPrep RNA cell miniprep system (Promega) using the protocol provided by the manufacturer. The concentration of total RNA was quantified by spectrophotometry (ND-1000; NanoDrop Technologies, Wilmington, DE, USA). RNA was reverse transcribed to single-strand cDNA using SuperScript III First-Strand Synthesis System for RT-PCR (Life Technologies). cDNA (n = 4 in each group) was subjected to quantitative PCR analysis with FastStart Universal Probe Master Mix (Roche, Basel, Switzerland) and primers for IL-8 using an ABI 7000 Real-Time PCR system (Life Technologies), as described previously [48]. The gene expression level was normalized to that of β-actin in the same cDNA.

Statistical Analysis
All parameters were expressed as mean˘standard deviation (SD). Differences between groups were analyzed by one-way analysis of variance (ANOVA) followed by Dunnett's post hoc test. A p value less than 0.05 was considered statistically significant.

Conclusions
Since TiO 2 nanoparticles are widely used in various fields, including the food industry, understanding their behavior and effects on the intestine is essential for risk assessment. In this study, we examined the effects of TiO 2 particles of different crystal structures and sizes in Caco-2 cells and THP-1 monocyte-derived macrophages. Exposure to 50 µg/mL of anatase TiO 2 nanoparticles increased the production of IL-1β in THP-1 macrophages and increased IL-8 expression in Caco-2 cells. These results indicate that anatase TiO 2 nanoparticles, but not other TiO 2 particles, seem to induce inflammatory response.

Conflicts of Interest:
The authors declare no conflict of interest.

Abbreviations
NPs nanoparticles TiO 2 titanium dioxide DLS dynamic light scattering PdI polydispersity index ROS reactive oxygen species