Nano Titanium Dioxide Particles Promote Allergic Sensitization and Lung Inflammation in Mice

The purpose of this study was to investigate whether photocatalytic TiO2 nanoparticles have adjuvant effect, when administered in combination with ovalbumin (OVA) in mice. Mice were immunized via intraperitoneal injections of OVA, OVA + TiO2 or OVA + Al(OH)3 and challenged with aerosols of OVA. At the end of the study, serum was analysed for content of OVA-specific IgE, IgG1 and IgG2a antibodies, and the bronchoalveolar lavage fluid (BALF) was analysed for content of inflammatory cells and levels of interleukin (IL)-4, IL-5, IL-10 and interferon-γ. The TiO2 particles promoted a Th2 dominant immune response with high levels of OVA-specific IgE and IgG1 in serum and influx of eosinophils, neutrophils and lymphocytes in BALF. The TiO2 particles induced a significantly higher level of OVA-specific IgE than the standard adjuvant Al(OH)3. However, the two substances were comparable regarding the level of eosinophilic inflammation and interleukins present in BALF.

In contrast to e.g. diesel exhaust particles and carbon nanotubes, titanium dioxide (TiO 2 ) particles are generally accepted to be low-toxic to humans as well as animals [15][16][17]. The presumed low toxicity of TiO 2 is reflected in the wide use of this compound, for example as a pigment in paints, food, medicine and sunscreens [18][19][20]. Micrometersized TiO 2 particles showed no adjuvant effect when administered to mice in combination with ovalbumin (OVA) [21].
For many substances, including TiO 2 , a reduction of particle size is associated with increased pro-inflammatory properties [22] and oxidative damage in human bronchial epithelial cell line [23]. With the development of nanotechnology and manufacturing of products containing nanoparticles, knowledge of the role of particle size on the toxicity is of increasing importance. Only one previous study [21] has investigated the effect of nano-sized TiO 2 on the immune response to simultaneously administered allergen, but very fine particles of widely different composition possessed adjuvant effect in a mouse model using OVA as model allergen [24]. Also, adjuvant effect of particles on allergic sensitization has been shown to increase with decreasing particle size [21,25].
Some of the newly marketed 'nano' household spray products with self-cleaning properties contain nano-sized photocatalytic TiO 2 , why human airway exposures to such particles may be increasing. Similarly, paints are deliberately added nanoparticulate TiO 2 to obtain improved properties and self-cleaning effects. Numerous other products may also contain TiO 2 nanoparticles. The present study aimed to investigate whether nano-sized photocatalytic TiO 2 possess adjuvant effect and promote allergic sensitization when administered in combination with OVA in mice.

Materials and Methods
Animals. Inbred female BALB ⁄ cJ mice aged 5-6 weeks were purchased from Taconic M&B, Ry, Denmark, and were housed as described previously [26]. Treatment of the animals followed procedures approved by The Animal Experiment Inspectorate, Denmark.
The TiO 2 was VP Disp. W 2730 X from Degussa AG (Frankfurt am Main, Germany), now Evonik Industries, which is a 30 wt% dispersion of fumed TiO 2 . The hydrodynamic particle number size distribution of the TiO 2 particles were determined by photon correlation spectroscopy using a Dynamic Laser Scatter (DLS) Zetasizer nano ZS (Malvern Inc., Malvern, UK) equipped with a 4 mW, 633 nm He-Ne Class I Laser. The results were calculated using the Dispersion Technology Software (DTS) versus 5.0 (Malvern Instruments Ltd.). TiO 2 was suspended in 0.45 lm MilliQ filtered water, treated with ultrasound and measured in 1 ml disposable polystyrene cuvettes. For calculations, we used the refractive (R i ) and absorption indices (R s ) for rutile TiO 2 (R i = 2.903; R s = 0.100) and standard properties for H 2 O. The density of TiO 2 was set to 4.25 g ⁄ cm 3 .
Immunization and airway challenge procedures. Mice (n = 7-8 per group) were primarily immunized on day 0 with 1 lg OVA alone or with 1 lg OVA in combination with either 2, 10, 50 or 250 lg TiO 2 intraperitoneal (i.p.). A group receiving 1 lg OVA in combination with 270 lg Al(OH) 3 served as positive adjuvant control.
To avoid further particle agglomeration, the hydrous TiO 2 suspension was diluted in sterile, pyrogen-free water as vehicle. Pyrogen-free saline was used for booster injections where OVA was administered without particles.
On days 27 and 34, all mice were challenged 20 min. with an aerosol of 1% OVA solution in saline.
The experiment was terminated on day 35. Blood samples were collected by heart puncture followed by bronchoalveolar lavage (BAL) as previously described [26].
Sera were assayed for content of OVA-specific IgE, IgG1 and IgG2a. The measurement of IgE was performed using a commercial IgE ELISA kit (OVA-IgE96 MD Biosciences, St Paul, MN, USA) according to manufacturer's description. IgG1 and IgG2a were assayed as described previously [27]. Levels of interleukin (IL)-4, IL-5, IL-10 and interferon (IFN)-c in BAL fluids were assayed by ELISA kits (eBioscience, San Diego, CA, USA) according to manufacturer's description.
Statistics. Antibody levels, numbers of inflammatory cells and levels of cytokines in BAL fluid in the four TiO 2 groups were compared to the OVA control group by the Kruskal-Wallis test. If a statistically significant difference was apparent, the individual TiO 2 groups were further compared to the OVA control group by the Mann-Whitney's U-test. A p-value of less than 0.05 was considered statistically significant. Calculations were performed using the Minitab Statistical Software, Release 14 Xtra (Minitab Inc., State College, PA, USA).

Characterization of particles.
Dynamic Laser Scatter analysis showed that the size spectrum of the ultrasound-treated TiO 2 particles suspended in MilliQ-filtered water was highly dominated by nano-sized particles ( fig. 2). In the number size distribution, showed a single modal peak occurred around 28 nm. Plotted by volume, a polymodal size distribution with coarser submicron particles was resolved giving an average zeta-size of 140.04 € 1.36 nm (Polydispersivety Index: 0.22). However, even by volume, more than 50% of the particles were smaller than 100 nm. This is in agreement with the manufacturer's specification claiming that the mean aggregate size (d-50) was below 100 nm.
The surface area of the particles was calculated to be 34.1 m 2 ⁄ g.

OVA-specific antibodies in serum.
The IgE and IgG1 levels in the three lower doses of TiO 2 were not statistically different from the OVA control group. The highest dose of TiO 2 , 250 lg, gave rise to significantly increased levels of IgE and IgG1 compared to the OVA control group (fig. 3). The Al(OH) 3 adjuvant group gave rise to a significantly higher level of IgG1, whereas the increase in IgE was not statistically significant compared to the OVA control group. Furthermore, the level of IgE, but not IgG1 was significantly higher in the 250 lg TiO 2 when compared to the Al(OH) 3 group (p = 0.009).
Levels of IgG2a were below the limit of detection in all samples (data not shown).

Inflammatory cells in BAL.
A significantly increased level of eosinophils, neutrophils and lymphocytes were seen in both the 250 lg TiO 2 group and the adjuvant control group when compared to the OVA  control group (confer fig. 4). Comparing the responses in the adjuvant control group and the 250 lg TiO 2 groups revealed that the number of neutrophils were significantly higher in the adjuvant control group, whereas no difference was seen on the numbers of eosinophils, lymphocytes and macrophages.
Cytokines in BAL supernatant. The Kruskal-Wallis analyses revealed a significant difference among groups in the levels of both the IL-4 (p = 0.032) and IL-5 (p = 0.002) (confer fig. 5); the highest levels of IL-4 and IL-5 were seen in the 250 lg TiO 2 group and Al(OH) 3 group. However, the pairwise comparison of exposure groups with the OVA control groups revealed no significant differences, i.e. none of the individual exposure groups differed from the OVA control group. IL-10 and IFN-c were below the limits of detection in all BAL samples (data not shown).

Discussion
Our study showed that the photocatalytic TiO 2 nanoparticles delivered as suspended in water possess adjuvant effect when administered in combination with OVA in mice. The adjuvant effect was suggested from the IgE and IgG1 levels, i.e. TiO 2 primes a Th2-dominant immune response. Photocatalytic TiO 2 nanoparticles were significantly more potent than Al(OH) 3 to stimulate the IgE production, whereas the two compounds gave rise to an approximately equal increase in production of IgG1. Furthermore, TiO 2 increased the influx of inflammatory cells in the lung. The increased level of eosinophils further supports that TiO 2 stimulates an allergyrelated immune response. Administration of 250 lg TiO 2 gave rise to a 5-fold increase in the level of eosinophils, a response comparable to that induced by 270 lg Al(OH) 3 . As IgE is not significantly increased in the Al(OH) 3 group, it could be speculated that the eosinophilic inflammation observed in this group is, at least in part, driven by the IgG1 antibody, which can be anaphylactic in the mouse after exposure to high allergen levels [28,29].
The mechanisms behind the adjuvant effect of particles have not been fully elucidated, but studies have shown that the size of a particle is important for its adjuvant activity. In general, adjuvanticity increases with decreasing particle size [21,25]. Due to the higher specific surface area of nanoparticles as compared to micron-sized particles, more antigen can be adsorbed to the same amount of particles [25]. Thus, nano (below 30 nm in diameter) TiO 2 particles were found to bind more OVA per mass unit than fine (over 200 nm in diameter) TiO 2 particles [21]. This may lead to the formation of an antigen depot with a prolonged release of antigen; a process that may increase antigenicity. Adsorption of antigen to a particle may also per se increase the antigenicity, as the most important professional antigen presenting cell type, the dendritic cells, are most effectively stimulated by antigen in particulate form [30]. Another proposal for the adjuvant mechanism is that ultra fine TiO 2 may induce oxidative damage and inflammation; processes that increase the level of activity of immune competent cells. This is supported by an unpublished study conducted at our institute (Dr. A.T. Saber) showing that a single intratracheal instillation of 54 lg photocatalytic TiO 2 nanoparticles induced acute neutrophilic inflammation in mice.
In conclusion, although micrometer-sized TiO 2 particles in general are considered to be biologically relatively inert, this study has shown that TiO 2 nanoparticles, which is a component in e.g. self-cleaning products, has adjuvant effect in mice. Thus, for TiO 2 , as shown for other substances, the toxicological risk assessments based on fine particles may not simply be extended to nanoparticles.