Registration Dossier

Toxicological information

Genetic toxicity: in vivo

Currently viewing:

Administrative data

Endpoint:
in vivo mammalian cell study: DNA damage and/or repair
Remarks:
Type of genotoxicity: DNA damage and/or repair
Type of information:
experimental study
Adequacy of study:
weight of evidence
Study period:
not specified
Reliability:
3 (not reliable)
Rationale for reliability incl. deficiencies:
significant methodological deficiencies
Remarks:
The study shows deficiencies in design and reporting. E171 supplier is not specified. Number of analysed cells is too low (100; 150 are recommended by the OECD*). The authors tested only one particle suspension dose (at least three are recommended by the OECD*). Positive controls and historical data are not specified. Exact time point of sampling and comet analysis is not specified. Cytotoxicity measurements are described insufficiently. Hedgehog occurrence is not specified. Information on animal husbandry are largely missing. No details on environmental conditions, housing, diet supply, water supply, acclimation period, and test group randomization. Age of test animals is only specified as adult. Animal body weights are only specified for the study initiation. Methods are described insufficiently. Details on cell lysis, electrophoresis, and tissue and slide preparation are missing. Concentration of gavaged solution is not specified. Evaluation criteria description is insufficient. Rationale for species, dose, and vehicle selection is not justified or only insufficiently. General clinical observations are not specified. Individual rat data are not presented. *OECD TG489: In Vivo Mammalian Alkaline Comet Assay, 2016
Cross-reference
Reason / purpose for cross-reference:
reference to same study
Reference
Endpoint:
sub-chronic toxicity: oral
Remarks:
100 days and 7 days
Type of information:
experimental study
Adequacy of study:
disregarded due to major methodological deficiencies
Study period:
not specified
Reliability:
3 (not reliable)
Rationale for reliability incl. deficiencies:
significant methodological deficiencies
Remarks:
Due to the following major study restictions the study is not relevant for human hazard assessment and therefore disregarded. The study design was not in accordance with any accepted guideline. The study did not include three dose groups, as recommended by OECD 408, only two dose groups were used in the 100 days toxicity study and one dose group in the 7 days toxicity study. The two concentration that were used are not in accordance with the dose setting recommended by the OECD guideline 408 (two or four fold intervals). Additionally, only male mice were used and the significance of this study is clearly reduced due to the low number of animals per group. No ophthalmological examination and haematological determination were conducted. No clinical chemistry examination was performed and no full gross necropsy and histopathological examination was conducted. E171 characterization was performed but no concentration determination or stability data of E171 dispersion prior intragastric administration or oral exposure via drinking water was stated. Further, no data about animal housing conditions, animal diet and body weight or food/water consumption were given. Beside of these study design restriction additional general deficiencies need to be mentioned. 1. First of all the source of the employed E171 is not stated. Despite the author´s efforts in characterising the test item, it is not possible to reconcile their results with those of the producer. 2. In the main study E171 was administered via drinking water. Considering the low solubility of TiO2 in water, it may be anticipated that the test item settled in the water bottle, leading to a discontinuous exposure of E171. Additional to that, no data on water consumption throughout the 100 days study are available and therefore no dose administration determination is possible. 3. Both modes of exposure (via gavage and drinking water) are strikingly different from the typical mode of human exposure in which E 171 is delivered to the GI tract as a constituent of food preparations. The surface structure of TiO2 molecules is rather malleable to alteration by surrounding macromolecules, resulting in widely different coronas depending on the preparation of TiO2, which could impact the way TiO2 particles interact with the host gut environment. 4. Delivery of TiO2 in water can greatly change the nature of the particles themselves, as TiO2 tends to aggregate and agglomerate in water suspension, raising the possibility that delivery of E 171 in an aqueous solution could lead to clumps of TiO2 passing through to the gut environment. 5. E 171 administered in drinking water, requires sonication, a process with unknown effects on the form of TiO2. Agglomeration and particle settling could also occur over time. This would represent a key difference in how TiO2 is seen by the host when consumed in food, where more particles would be part of the food matrix resulting in a more uniform profile of single particles as compared to clumps. 6. According to the author adult male Wistar rats were used for both studies. However, the given body weight range of 175-200 g is in fact characteristic for a pre-adolescent rats at the age of 5.5-6 weeks and therefore clearly too young for this study. 7. Cytokine concentrations in different tissues were evaluated with cytokine-kits which are not suitable/validated for tissues homogenates. Due to that, results obtained with these kits are questionable. 8. T helper cell detection in flow cytometry was conducted by using the cell surface marker CD4 and CD25. These surface marker are not only characteristic to T helper cells and therefore cannot be used for specific detection of T helper cells. 9. Findings for the divergent relative immune cell distribution of the Peyer´s patches in control animals after 7 and 100 days exposure are not explained. It remains unclear why the %DC increased by more than 7-fold between the control animal groups. 10. The food-grade TiO2 used in this study is not representative for E171. Electron microscopy and stability testing of food-grade TiO2 (E171) suggests that approximately 36% of the particles are less than 100 nm. In contrast to that, the E171 sample used in this study had a nano particle amount of 44.7 % and clearly exceeded the average nano particle amount of 36%. 11. NP-105 was used as reference material. This compound is characterized by 100 % NP´s, a mean size of 23 nm, a specific surface area of 50 m²/g, a mixture of anatase and rutile grains (85/15) and an isoelectric point at pH 6.5. Therefore NM-105 samples clearly distinguish from E171 samples by all parameters taken into account and it is therefore concluded that NM-105 does not appear to be the most suitable reference material for toxicity studies by ingestion. 12. While DMH is a known, potent, genotoxic intestinal carcinogen in rodents, variability in numbers and types of lesions in the intestine are usually seen in bioassays of DMH and other carcinogens, particularly when a relatively low dose is administered as in this study. There is also considerable variability in the number and size of ACF that are induced by DMH. Further, in the rats without DMH pre-treatment there are typically also a few ACF (background). However, no groups administered only E171 were available in this study. Bettini et al. only evaluated groups pre-treated with DMH. 13. It is also noteworthy, that in this study animals were exposed to E171 with and without pre-treatment of DMH, but the results with number of ACF or number of aberrant crypts per colon were only reported for animals pre-treated with DMH and not for those not pre-treated. For those not pre-treated only the number of animals with and without ACF was reported. This raises the question why the results were reported in a different way. And despite of that, as mentioned before, the number of ACF per animal is not a representative parameter, since ACFs are also noted in colons of untreated animals (background level). 14. It was not mentioned whether the highly acidic DMH injection solution was neutralized before i.p. injection during pre-treatment. However, they almost certainly did as the acid without neutralization is highly irritating and highly toxic to the rats when injected i.p. Also, they stated that they injected 180mg/kg of DMH. However, they used must have used DMH·2HCI, and at that level of DMH (which would be 398mg/kg DMH·2HCI), there is severe toxicity. Thus, it is likely that they used 180mg DMH·2HCI, which is 81mg/kg DMH. 15. Comparing the results of Bettini et al. with other similar research for example the work by Nogueira et al. or Carpenter et al. the divergent findings in immune cell response as well as in the cytokine levels are disturbing and raise questions as to whether the reported findings by Bettini et al. are in fact test-substance related or more likely chance findings of uncertain physiological relevance.
Justification for type of information:
The study published by Blevins et al. (see cross-reference) has been conducted to recapitulate the findings of Bettini et al. using a model of TiO2 delivery in food containing E171.
Reason / purpose for cross-reference:
reference to other study
Reason / purpose for cross-reference:
reference to same study
Qualifier:
no guideline followed
Principles of method if other than guideline:
Bettini et al evaluated the impact of food-grade titanium dioxide (E171) and NM-105 on rats in a 7 and 100 day oral toxicity study. Adult male Wistar rats were exposed to E171 or the TiO2 nano particle reference substance NM-105 by intragastric gavage in a concentration of 10 mg/kg bw/day for seven days. After seven days animals were sacrificed and used for tissue imaging, flow cytometry and cytokine assays and to carry out tissue inflammation and gut permeability measurements.
In the main experiment rats were treated or not with 1,2-Dimethylhydrazine (DMH) to induce colon carcinogenesis. Seven days later these rats were exposed to E171 at 200 µg/kg bw/day or 10 mg/kg bw/day though drinking water for 100 days. Control animals received water only. After 100 days rats were sacrificed and used for flow cytometry and cytokine assays as well as for gut inflammation and ACF assessments.
In an additional part, untreated rats were used for ex vivo cytotoxicity and prolifertive assays on isolated immune cells.
GLP compliance:
not specified
Limit test:
no
Specific details on test material used for the study:
not applicable
Species:
rat
Strain:
Wistar
Sex:
male
Details on test animals or test system and environmental conditions:
TEST ANIMALS
- Source: Janvier Labs (France)
- Weight at study initiation: 175 - 200 g
- Acclimation period: 5 days (7-day study), similar period (100-day study)
Route of administration:
other: Main study: 100 day exposure to TiO2 via drinking water Additional study: 7 days exposure to TiO2 via gavage
Vehicle:
water
Details on oral exposure:
PREPARATION OF DOSING SOLUTIONS: according to the generic Nanogenotox dispersion protocol (Final protocol for producing suitable manufactured nanomaterial exposure media. The generic NANOGENOTOX dispersion protocol. July, 2011)

Pre-treatment of 100-day study animals:
Rats were given a single i.p. injection of 1,2-dimethylhydrazine (DMH; 180 mg/kg i.p.; Sigma Aldrich) in NaCl (9 g/L) to induce colon carcinogenesis. Seven days later, they were randomly allocated to one of the hree groups and were exposed to 0, 0.2 or 10 mg/kg bw/day of E171 in drinking water for 100 days. As control further animals were not pre-exposed to DMH but received only 0 or 10 mg/kg bw/day of E171 in drinking water for 100 days.
Analytical verification of doses or concentrations:
not specified
Duration of treatment / exposure:
main study: 100 days
additional study: 7 days
Frequency of treatment:
daily
Dose / conc.:
0.2 mg/kg bw/day (nominal)
Remarks:
100 days oral toxicity study - via drinking water
Dose / conc.:
10 mg/kg bw/day (nominal)
Remarks:
100 days oral toxicity study - via drinking water
Dose / conc.:
10 mg/kg bw/day (nominal)
Remarks:
7 days oral toxicity study - via gavage (200µl)
No. of animals per sex per dose:
Main study 100 days: 11 to 12 animals/group
Additional study 7 days: 10 animals/group
Control animals:
yes, concurrent vehicle
Positive control:
not specified
Observations and examinations performed and frequency:
CAGE SIDE OBSERVATIONS: Not specified

DETAILED CLINICAL OBSERVATIONS: Not specified

BODY WEIGHT: Not specified

FOOD CONSUMPTION AND COMPOUND INTAKE (if feeding study):
- Food consumption for each animal determined and mean daily diet consumption calculated as g food/kg body weight/day: Not specified
- Compound intake calculated as time-weighted averages from the consumption and body weight gain data: Not specified

FOOD EFFICIENCY:
- Body weight gain in kg/food consumption in kg per unit time X 100 calculated as time-weighted averages from the consumption and body weight gain data: Not specified

WATER CONSUMPTION AND COMPOUND INTAKE (if drinking water study): Not specified

OPHTHALMOSCOPIC EXAMINATION: Not specified

HAEMATOLOGY: Not specified

CLINICAL CHEMISTRY: Not specified

URINALYSIS: Not specified

NEUROBEHAVIOURAL EXAMINATION: Not specified

IMMUNOLOGY: Yes
- Time schedule for examinations: study termination
- How many animals: 10-12 animals/group
- Dose groups that were examined: all dose groups after 7 and 100 days
- Parameters checked in table were examined:
- 100 and 7 days: quantification of dendritic cells, Treg and Th cells. Jejunal and ileal Peyer´s patches were washed 5 times with PBS 3 mM EDTA, once with PBS, and digested with collagenase for 40 min. Cell suspensions were filtered and purified by centrifugation on a 40–80% Percoll gradient. Spleen cells were filtered in PBS containing 1% FCS. Cell phenotypes were analysed with antibodies directed against CD103, MHC-II and CD11b/c to determine DC phenotypes; CD4 and CD25 were used to identify Th cells, and association with intracellular staining for FoxP3 was used to detect Tregs.
Isolated cells from PP and spleen were seeded on 24-well plates coated with anti-CD3/CD28 antibodies at 1 * 106 cells/well. After 4 days of antibody stimulation, the culture supernatants were collected and cytokine assays were performed.
The levels of IL-6, IFN-γ , TNF-α , IL-1β , IL-10, IL-8 (CINC-1), and IL-17 in cell culture supernatants and/or tissue extracts were determined using commercial ELISA kits.
- 7 days: MPO, TNFa, IL-10, IL-1ß, IFN-y, IL-17 (in jejunal and colonic tissue)
Sacrifice and pathology:
When the experiments were completed, the animals were killed in random order.
Tissue samples were fixed in 4 % formaldehyde, equilibrated in 30 % sucrose in phosphate buffer, frozen in liquid nitrogen and stored at -80 °C. Cryo-sections (15 µm thickness) were fixed in acetone at -20 °C (5 min), rehydrated in PBS (10 min), and then mounted in Pro-long gold antifade medium.
The colons were coded, stained with methylene blue (0.1 %) for 6 min, and scored for preneoplastic lesions (ACF).
Other examinations:
Confocal microscopy and micro X-ray fluorescence imaging
Tissue samples were fixed in 4% formaldehyde, equilibrated in 30% sucrose in phosphate buffer, frozen in liquid nitrogen and stored at − 80 °C. Cryo-sections (15 μ m thickness) were fixed in acetone at − 20 °C (5 min), rehydrated in PBS (10 min), and then mounted in Pro-long gold antifade medium (Life Technologies). Tissue sections were examined under a confocal microscope (Leica SP8) at 488/BP 488–494 nm to detect light scattering TiO2 particles and at 514/BP 560–660 nm to monitor auto fluorescence in the tissue. Similar measurements of E171-TiO2 light scattering at 488/BP 488– 494 nm were performed for particle size determination in the E171 water suspensions given to the rats and in the recovered intestinal luminal contents (jejunum and colon) after their spreading, drying and mounting in Pro-long medium. The particles were detected using a 63X objective and a magnification factor of 1 pixel to 50 nm. The μ XRF analyses were performed at the SOLEIL Synchrotron (DiffAbs beamline) using a 4-element silicon drift detector (Vortex-ME4, Hitachi). The Ti distribution in the tissue sections was determined using μ XRF in micro-beam mode by adding secondary focusing optics. At the sample position, the beam size was 10.4 × 7.0 μ m2 (horizontal and vertical, respectively) with a flux close to 1010 phs−1. The energy of the monochromatic X-ray beam was set at 7.5 keV, and a 4-element silicon drift detector allowed for fluorescence collection.


Ti distribution determination via NanoSIMS imaging
The SIMS image analyses were performed on 200-300 nm ultra-thin sections prepared according to standard transmission electronic microscopy protocols and using NanoSIMS50 equipped with a caesium source. With an impact energy of 16 keV and a primary current of 1pA, the Cs+ beam was rastered over a 20 × 20 μ m² area on the sample surface. The images were recorded in 256 × 256 pixels with a counting time of 20 ms/pixel. The beam size was in the range of 80–100 nm. The instrument was tuned for a mass resolution M/Δ M > 5000. The four masses recorded were the negative clusters 12C14N− (m/z = 26.003 u), 31P16O− (m/z = 46.969 u), 46Ti16O− (m/z = 61.948 u) and 48Ti16O− (m/z = 63.943 u). Due to his low affinity with one electron, the titanium (electron affinity Ea = 0.079 eV) was recorded as TiO clusters (Ea = 1.30 eV). Mass calibrations were achieved using standard references (Ti-sheet, Goodfellow). Both clusters of Ti isotopes (46Ti and 48Ti) were recorded simultaneously to verify the isotopic ratio (46Ti/48Ti = 0.112) and the separation of the isobaric masses (e.g., 32S16O2, and 48Ti16O).


Intestinal permeability measurement
In vivo intestinal permeability was assessed using 51Cr-EDTA (Perkin Elmer Life Sciences), a marker of paracellular permeation. Briefly, rats were placed in metabolic cages for 3 days to become accustomed to the environment, and 0.7 μ Ci of 51Cr-EDTA diluted in 0.5 mL of saline was administered by gavage, and then urine samples were collected for 24 h. The total intestinal permeability to 51Cr-EDTA is expressed as a percentage of the administered radioactivity recovered from the 24 h urine samples as measured by a gamma counter (Cobra II, Packard).


Cell isolation and flow cytometry analysis
Jejunal and ileal PP were washed 5 times with PBS 3 mM EDTA, once with PBS, and digested with collagenase for 40 min. Cell suspensions were filtered and purified by centrifugation on a 40–80% Percoll gradient. Spleen cells were filtered (40-μ m nylon mesh filter) in PBS containing 1% FCS. Cell phenotypes were analysed with antibodies directed against CD103 (OX-62, Biolegend), MHC-II (OX-6, Biolegend) and CD11b/c (OX-42, BD Pharmingen) to determine DC phenotypes; CD4 (W3/25, BD Pharmingen) and CD25 (OX-39, BioLegend) were used to identify Th cells, and association with intracellular staining for FoxP3 (FJK-16s, eBioscience) was used to detect Tregs. The data were collected using a MACSQuant analyser (Miltenyi Biotec) and were analysed via VenturiOne software (AppliedCytometry).

Ex Vivo Cytotoxicity and Proliferation Assay
Isolated cells from PP were labeled with CellTrace® Violet (Life Technologies) to assess proliferation. Cells were seeded on 24-well plates at 6*105 cells/well in Cerrotini medium, and incubated with 2 μg/mL concanavalin A for 4 days with E171 or NM-105 (0, 37, 75 or 150 μg/mL). After washing, T cells were stained with anti-TCRαβ (R73) and with To-pro-3 (Life Technologies) to follow cell viability by flow cytometry.

Cytokine assays
The levels of IL-6, IFN-γ , TNF-α , IL-1β , IL-10, IL-8 (CINC-1), and IL-17 in tissue extracts were determined using commercial ELISA kits (Duoset R&D Systems) according to the manufacturer’s instructions. For tissue samples, segments of jejunum and colon were prepared in RIPA buffer (0.5% deoxycholate, 0.1% SDS, and 1% Igepal in TBS) containing complete protease inhibitor cocktail (Roche), and protein concentrations were measured using a BCA Optima kit (Interchim). The data are expressed as pg per milligram of tissue protein.


Aberrant crypt foci
To examine the impact of E171 on initiation of ACF´s rats were exposed to E171 for 100 days. When the experiments were completed, the animals were killed in random order. The colon were coded, stained with methylene blue (0.1 %) for 6 min, and scored for preneoplatic lesions following Bird´s procedure. The number of ACF per colon were counted under a light microscope at 40X magnification in duplicate by two readers who were blinded to sample origins.

Western blot analysis for caspase-1.
Colons of rats treated with vehicle or E171 (n=4 per group) for 100 days were collected in dry ice and stored at -80°C until processing. Proteins were extracted with RIPA buffer containing protease inhibitor cocktail (Roche Diagnostics), and total protein concentrations were assessed using a BCA Protein Assay kit (Thermo Scientific, Rockford, USA). Equal amounts of protein extracts (30 μg) were separated in 12% FastCast TGX Bio-Rad gels and transferred onto a blotting membrane. The membrane was blocked for 1 h at RT with Odyssey blocking buffer (Rockland, Tebu-bio, France) and then incubated overnight at 4°C with primary mouse anti-caspase-1 antibody (R&D Systems; 1:1000 in blocking buffer). After washing, the membranes were incubated for 1 h at RT with secondary fluorescent IRDye *680 anti-mouse antibody (LI-COR, Lincoln, USA; 1:10000 in blocking buffer). An Odyssey Infrared Imaging Scanner (LiCor ScienceTec, Les Ulis, France) was used to detect the intensity of full length (45 kD) and cleaved (20 kD) caspase-1 protein bands.

Myeloperoxidase Activity Assay
For assessment of neutrophilia in intestinal mucosa, the activity of the enzyme myeloperoxidase (MPO), a marker of polymorphonuclear neutrophils granules, was determined in tissue segments of the jejunum and colon (1 cm).
Statistics:
The results are expressed as the mean ± s.e.m. Statistical significance was assessed using Prism 4 software (GraphPad). Student´s t-test, One-way-ANOVA or the Kruskal-Wallis test, followed by post hoc tests were used, or contingency analysis followed by Fisher´s exact test was used, when appropriate. A P value of <0.05 was considered significant.
Clinical signs:
not specified
Mortality:
not specified
Body weight and weight changes:
not specified
Food consumption and compound intake (if feeding study):
not specified
Food efficiency:
not specified
Water consumption and compound intake (if drinking water study):
not specified
Ophthalmological findings:
not specified
Haematological findings:
not specified
Clinical biochemistry findings:
not specified
Urinalysis findings:
not specified
Behaviour (functional findings):
not specified
Immunological findings:
effects observed, treatment-related
Description (incidence and severity):
Frequency of dendritic cells, T-regs and T-helper cells- 7- and 100-days:
After 7 days of oral exposure, both NM-105 and E171 induced a significant increase in dendritic cell (DC) frequency in Peyer´s patches (PP). After chronic E171 treatment, early effects on DC in the PP were found to be transient, as they were not detected in rats exposed for 100 days through drinking water. Regarding Tregs, the NM-105 nanomaterial had no effect on PP after 7 days of oral exposure while the same duration of treatment with E171 additive led to significant decrease in this cell subset that was still observed in PP after 100 days of exposure. The decrease of Tregs appeared concomitantly with a decrease in T helper cells. To determine whether TiO2 particles directly mediated T cell depletion, cells isolated from PP of untreated rats were exposed ex vivo to E171 particles or NM-105 TiO2-NPs, and cell viability and proliferation were compared. A dose-dependent cytotoxic and anti-proliferative effect on the T cells was observed, and this effect was found to be more pronounced with E171 compared to the NM-105 TiO2-NP model.

Mucosal inflammation and immune cell responses- 7-days:
No change in myeloperoxidase (MPO) activity or the content of basal cytokines TNF-a, IL-10, IL-1ß, IFN-y and IL-17 in mucosa of the small and large intestine relative to the control rats could be observed.

Ex-vivo immune cell response (cells isolated from PP and spleen)- 7-days:
Total immune cells were isolated from PP and the spleen and then cultured with anti-CD3/CD28 antibodies to induce cytokine secretion into the culture media. In the PP, all TiO2 materials attenuated inflammatory IFN-γ secretion relative to the controls while the IL-17 response remained unchanged. In the spleen, both NM-105 and E171 elicited a potent Th1/Th17 immune response through increased production of IFN-γ and IL-17.

Cytokine assay in colonic mucosa- 100-days:
Cytokine assays showed moderate but significant increase in TNFa (+26 %), IL-8 (+ 45 %), and IL-10 (+ 26 %) in the colonic mucosa of non DMH pre-treated E171-treated rats (100 days) relative to controls.
Organ weight findings including organ / body weight ratios:
not specified
Gross pathological findings:
not specified
Neuropathological findings:
not specified
Histopathological findings: non-neoplastic:
not specified
Histopathological findings: neoplastic:
not specified
Other effects:
effects observed, treatment-related
Description (incidence and severity):
Tissue distribution of E171 in the rat intestine and liver after 7 days
Light-diffracting TiO2 particles were found in the Peyer´s patches along the small intestine as well as in the colonic mucosa and liver of rats orally given E171 for 7 days but not in the controls. µXRF measurement for Ti element revealed Ti in the gut lumen and PP as well as in colon mucosa. In addition Ti was found in the liver, with the highest density found close to the portal vein sinus.

NanoSIMS anaylses of subcellular Ti distribution in PP after 7 days
NanoSIMS imaging revealed Ti in the PP of all TiO2-treated rats, with similar distribution patterns between the NM-105 and E171 TiO2 sources. The highest Ti density was found in the cytoplasm, Ti-rich regions were identified in the nuclei of PP cells and were closely associated with the phosphorus-positive chromatin.

Aberrant crypt foci after 100 days
After induction with DMH E171 treatment at 10 mg/kg of bw/day significantly increased the total number of aberrant crypts per colon as well as the number of large ACF per colon (i.e., more than three aberrant crypts per ACF) relative to the control and 200 μ g/kg of bw/day groups. Despite an increasing trend at the highest dose, no significant difference in the number of ACF per colon was observed between the groups of rats.
The E171-treatment (with DMH pre-treatment) at 10 mg/kg bw/day led to an increase in aberrant crypt foci in the colon of treated rats. Four of the 11 animals spontaneously developed one to three ACF per colon. Three of the 4 rats developed lesions of 1 to 3 aberrant crypt(s) per ACF, and 1 rat developed a severe lesion of 12 aberrant crypts.
Details on results:
Gut permeability after 7 days
No significant change in the epithelial paracellular permeability to 51Cr-EDTA that was orally given to rats was observed in the E171 group in comparison to the controls (1.41 ± 0.08 vs. 1.63 ± 0.09 % of total radioactivity recovered in 24 h urine samples, respectively).

Western blot after 100 days
Western blotting for caspase-1 did not show cleaved caspase-1 in the colons of E171-treated rats relative to control animals.
Remarks on result:
not determinable because of methodological limitations
Critical effects observed:
not specified
Conclusions:
Bettini, S. et al. (2017): The study conducted by Bettini et al. evaluated the toxicity of orally administered food-grade TiO2 and nano TiO2 (P25) after 7 days via gavage as well as of E171 after 100 days via drinking water. According to the author TiO2 was detected in the immune cells of Peyer´s patches (PP) in rats orally exposed for one week to E171. Dendritic cell frequency increased in PP regardless of the TiO2 treatment, while regulatory T cells decreased with E171 only, an effect still observed after 100 days of treatment. In all TiO2-treated rats, stimulation of immune cells isolated from PP showed a decrease in T-helper (Th)-1 IFN-γ secretion, while splenic Th1/Th17 inflammatory responses sharply increased. E171 or NM-105 for one week did not initiate intestinal inflammation, while a 100- day E171 led to an increase in TNF-a, IL-10 and IL-8 in the colonic mucosa and initiated preneoplastic lesions while also fostering the growth of aberrant crypt foci in a chemically induced carcinogenesis model.

In contrast to that, no changes in gut permeability could be observed and E171 administration for 100 days did not initiate genotoxicity (Comet assay) nor an increase in MPO activity. Western blotting for caspase-1 did not show cleaved caspase-1 in the colons of E171-treated rats relative to control animals.

Due to the following major study restrictions, the study is not considered relevant for human health hazard assessment and therefore disregarded: The study design was not in accordance with any accepted guideline. The study did not include three dose groups, as recommended by OECD 408, only two dose groups were used in the 100 days toxicity study and one dose group in the 7 days toxicity study. The two concentration that were used are not in accordance with the dose setting recommended by the OECD guideline 408 (two or four fold intervals). Additionally, only male mice were used and the significance of this study is clearly reduced due to the low number of animals/tests per group/experiment (in vivo n=10-12, ex vivo n=5). No ophthalmological examination and haematological determination were conducted. No clinical chemistry examination, gross necropsy and histopathology were performed. E171 characterization was performed but no concentration determination or stability data of E171 dispersion prior intragastric administration or oral exposure via drinking water was stated. Further, no data about animal housing conditions, animal diet and body weight or food/water consumption were given.


Beside of these study design restriction additional general deficiencies need to be mentioned.
1. First of all the source of the employed E171 is not stated. Despite the author´s efforts in characterising the test item, it is not possible to reconcile their results with those of the producer.
2. The food-grade TiO2 used in this study is not representative for E171. Electron microscopy and stability testing of representative food-grade TiO2 (E171) samples suggests that approximately 36% of the particles are less than 100 nm. In contrast to that, the E171 sample used in this study had a nano particle amount of 44.7 % and clearly exceeded the average nano particle amount of 36%.
3. NP-105 was used as reference material. This compound is a mixture of anatase and rutile grains (85/15), shows a mean primary particle size of 23 nm, a specific surface area of 50 m²/g and an isoelectric point at pH 6.5. Therefore NM-105 samples clearly distinguish from E171 samples by all parameters taken into account and it is therefore concluded that NM-105 does not appear to be the most suitable reference material for toxicity studies by ingestion.
4. In the main study E171 was administered via drinking water. Considering the low solubility of TiO2 in water, it may be anticipated that the test item settled in the water bottle, leading to a discontinuous exposure of E171. Additional to that, no data on water consumption throughout the 100 days study are available and therefore no dose administration determination is possible.
5. Delivery of TiO2 in water can greatly change the nature of the particles themselves, as TiO2 tends to aggregate and agglomerate in water suspension, raising the possibility that delivery of E 171 in an aqueous solution could lead to clumps of TiO2 passing through to the gut environment.
6. E 171 administered in drinking water, requires sonication, a process with unknown effects on the form of TiO2. Agglomeration and particle settling could also occur over time. This would represent a key difference in how TiO2 is seen by the host when consumed in food, where more particles would be part of the food matrix resulting in a more uniform profile of single particles as compared to clumps.
7. Both modes of exposure (via gavage and drinking water) are strikingly different from the typical mode of human exposure in which E 171 is delivered to the GI tract as a constituent of food preparations.
8. According to the author adult male Wistar rats were used for both studies. However, the given body weight range of 175-200 g is in fact characteristic for a pre-adolescent rats at the age of 5.5-6 weeks and therefore clearly too young for this study.
9. Cytokine concentrations in different tissues were evaluated with cytokine-kits which are not suitable/validated for tissues homogenates. Due to that, results obtained with these kits are questionable.
10. T-helper cell detection in flow cytometry was conducted by using the cell surface marker CD4 and CD25. Both surface marker are not only characteristic to T helper cells and therefore cannot be used for specific detection of T helper cells. Indeed, age-related changes in %DC are much greater than the changes suggested in this paper to result from exposure to TiO2.
11. Findings for the divergent relative immune cell distribution of the Peyer´s patches in control animals after 7 and 100 days exposure are not explained. It remains unclear why the %DC increased by more than 7-fold between both control groups.
12. It is interesting to compare the effects on the cytokine profile reported by Bettini et al. (2017) with other similar research for example the work by Nogueira et al. (2012). They treated male Bl 57/6 mice orally with 100 mg/kg bw/day of either (i) uncoated Anatase TiO2 (MPTiO2, 260nm), (ii) KRONOS 171 titanium dioxide E 171) or (iii) uncoated self-synthesised nanoparticles (NPTiO2, 66nm) for 10 consecutive days. Immediately after the last treatment the duodenum, jejunum and ileum were extracted for assessment of cytokines and inflammatory cells. Cytokines were evaluated via ELISA and CD4+ (Th and Tregs) and CD8+ (cytotoxic T cells) T cells, natural killer cells, and dendritic cells were evaluated in the fixed samples by immunohistochemistry.
- Nogueira et al. (2012) measured a statistically significant increase in T CD4+ (Th and Tregs) cells in the duodenum, jejunum and ileum of mice treated with MPTiO2 and NPTiO2, compared to the control group (Fig. 2). No significant difference was observed between the MPTiO2 and NPTiO2 groups. Mice treated with MPTiO2 or NPTiO2 showed no increase in T CD8+ (cytotoxic T cells), natural killers, or dendritic cells in the small intestine.
- The Bettini et al. (2017) results showed that both NM-105 and E171 induced a significant increase in DC frequency in PP (Fig. 3a) after 7 days of oral exposure. Regarding Tregs, the NM-105 had no effect on Peyer’s Patches (PP) after 7 days of oral exposure (Fig. 3b) while E171 led to a significant decrease in Tregs and Th (Fig. 3b,d and Fig. 3c, e).
- Nogueira et al. (2012) measured a statistically significant increase in inflammatory cytokines in MPTiO2 and NPTiO2 treated animals, compared to the control group. Specifically, enhanced concentrations of IFN-γ were shown in the ileum for groups receiving MPTiO2 and NPTiO2 (Figure 1c and d), increasing by approx. 1500% and 675% respectively.
- The Bettini et al. (2017) study on the other hand showed that in the PP, NM-105 and E171 both attenuated inflammatory IFN-γ secretion relative to the controls, decreasing by approx. 58% and 48% respectively.
Given the similarity in treatment duration and the titanium dioxide E171 samples used for the treatment by Bettini and Nogueira, the obviously divergent findings in immune cell response as well as in the cytokine levels are disturbing and raise questions as to whether the reported findings are in fact test-substance related or more likely chance findings of uncertain physiological relevance. Considering that the dose administered by Nogueira et al. was even an order of magnitude higher than that of Bettini et al., the findings by Bettini et al. cannot be interpreted as positive evidence of any adverse inflammatory response in the small intestine with relevance for humans.
13. The suggested preneoplastic lesions and subtle alterations of inflammation markers observed in ex-vivo, in-vitro experiments that are discussed in the paper are an interesting model, but as yet there is no agreed link to cancer in humans. The reported slight alterations in inflammatory cytokines also do not appear to follow any consistent pattern. The authors report on levels of 8 different cytokines in colonic mucosa after 100 days of administration of 10 mg/kg bw/d of TiO2, whereby no statistically significant differences were found for IL-6, IL-17, IL-18, IFN-γ and IL-1β; only slight increases for IL-8, IL-10 and TNF-α. In contrast, under conditions of chronic inflammation such levels are usually elevated by one or several orders of magnitude. The low group size does not allow a statistically robust statement regarding any potential carcinogenic effect.
Without attempting to provide an exhaustive overview of the relationship between colorectal cancer (CRC) and interleukin levels, two examples may be discussed at some length here:
(i) IL-6 signalling is involved in sporadic colorectal cancer: serum IL-6 levels were significantly elevated in patients with sporadic colorectal adenoma compared to normal controls (Uchiyama et al., 2012). IL-6 amount of the serum and tumoural tissue in the patients with colorectal cancer correlate significantly with the staging of the tumour and with each other, with IL-6 level rising to as much as 4-fold during progression of colorectal cancer growth (Esfandi et al., 2006). Several experimental and clinical studies have linked interleukin-6 to the pathogenesis of sporadic and inflammation-associated colorectal cancer (CRC), and an increased IL-6 expression has been related to advanced stage of disease and decreased survival in CRC patients (Waldner et al., 2012). In a large meta-analysis of Fourteen studies comprising 1,245 patients, serum IL-6 expression was highly correlated with overall survival rate, tumour invasion, distant metastasis and tumour stage (Wang et al. 2015).
(ii) It is widely accepted that chronic inflammation plays an active role in cancer, and interleukin-17 as a proinflammatory cytokine can promote cancer-elicited inflammation and is generally considered to be a promoter in CRC progression (Wu et al., 2013).
Despite the established link between inflammation, colorectal cancer and elevated IL-6 and IL-17, there are no statistically significant increases of these two interleukins in the work by Bettini et al.
14. The significance of aberrant crypt foci (ACF) has been investigated over the last several decades with literally hundreds of publications with our knowledge not leading to a consistent opinion, leaving the relevance to humans to still be in question. Extensive epidemiological studies on more than 24,000 titanium dioxide workers over several decades who have been inadvertently exposed orally shown no indication whatsoever of any forms of cancer arising from exposure to titanium dioxide.
Aberrant crypt foci (ACF) were first discovered in mice in 1987 (Bird, 1987) and in colonic mucosa of humans suffering from colorectal cancer in 1991 (Pretlow et al., 1991). Since then, they have been widely investigated as a putative precursor for colorectal cancer. ACF are found predominantly in the distal colon and can for example be recognised in mice and rats as early as two to four weeks after dosing with a colon carcinogen. However, based on knowledge obtained during the last decade it can be stated that an unequivocal correlation between occurrence of ACF (neither qualitatively nor quantitatively) and later development of colorectal tumours has not been demonstrated. Moreover, neither the total number nor the crypt multiplicity of the ACF is correlated with the tumour outcome. Even though a lot of knowledge about ACF has accumulated during the last decade, the outcome of the studies is still inconclusive concerning the neoplastic potential of the ACF (Thorup, 1996).
In several recent reviews, the relevance of ACF as a surrogate endpoint for colorectal cancer is explicitly questioned (Lance & Hamilton, 2017; Khare et al., 2009).
Despite the above, ACF in rats are nevertheless proposed by Bettini et al. as pre-neoplastic markers for colon cancer and said to increase with TiO2 exposure in their study. The dose dependency for this effect is at best weak with only modest increments in response being observed after 50-fold increases in dose. Resistance of preneoplastic cells to toxicity from TiO2 relative to normal cells is proposed to modulate this effect, but the cell culture study data presented exhibit inverse relationships between dose and cytotoxicity with greater toxicity being associated with lower TiO2 concentrations, thus lacking plausibility.
15. It is also noteworthy, that in this study animals were exposed to E171 with and without pre-treatment of DMH, but the results with number of ACF or number of aberrant crypts per colon were only reported for animals pre-treated with DMH and not for those without pre-treatment. For those not pre-treated only the number of animals with and without ACF was reported. This raises the question why the results were reported in a different way. And despite of that, as mentioned before, the number of ACF per animal is not a representative parameter, since ACFs are also noted in colons of untreated animals (background level).
16. While DMH is a known, potent, genotoxic intestinal carcinogen in rodents, variability in numbers and types of lesions in the intestine are usually seen in bioassays of DMH and other carcinogens, particularly when a relatively low dose is administered as in this study. There is also considerable variability in the number and size of ACF that are induced by DMH. Further, in the rats without DMH pre-treatment there are typically also a few ACF (background). However, no groups administered only E171 were available in this study. Bettini et al. only evaluated groups pre-treated with DMH.
17. It was not mentioned whether the highly acidic DMH injection solution was neutralized before i.p. injection during pre-treatment. However, they almost certainly did as the acid without neutralization is highly irritating and highly toxic to the rats when injected i.p. Also, they stated that they injected 180mg/kg of DMH. However, they used must have used DMH·2HCI, and at that level of DMH (which would be 398mg/kg DMH·2HCI), there is severe toxicity. Thus, it is likely that they used 180mg DMH·2HCI, which is 81mg/kg DMH.
18. Scientific considerations: generally, the observation that TiO2 particles can be found in Peyer’s Patches of the small intestine has already been reported several times previously elsewhere. This is also neither alarming nor unexpected, since the cellular composition of these patches is specifically designed to allow for a wide range of particulates to be internalised in order to facilitate an antigen recognition and a protective immunological response, as part of a natural innate immune response. Under these conditions, it is also not particularly unexpected that some (not all) cytokines are slightly elevated, with some cells internalising “foreign matter” and processing it. We fail to see why this should be considered an adverse effect, since it constitutes an expected result of a normal innate immune response.
19. Oral uptake: to date, several reputable scientific organisations (EFSA, RIVM) have considered oral uptake negligible based on toxicokinetic investigations reporting a lack of analytically detectable titanium in blood/tissues even after repeated oral administration of high doses of TiO2 (EFSA, 2016; Geraets et al., 2014), which is in strict contrast to Bettini et al. microscopic findings of TiO2 particulates in liver tissue, for example. No true quantitative estimate of uptake is provided with even microscopic detection of particles in the liver demonstrating sparse distribution of presumed TiO2 particles in what is described as a “Ti-rich” area of the liver. The finding of such sparse distribution of particles does not provide evidence that systemic uptake exceeds the extremely limited uptake findings of other studies.
20. Extrapolation from laboratory animals (rats) to humans: the authors do not adequately acknowledge the inherent difficulties in extrapolating any observations of gastrointestinal effects in rats to humans. When particles such as TiO2 are ingested, they “rapidly adsorb proteins from biological fluids forming a protein ‘corona’. This protein corona alters the size, aggregation state and interfacial composition of a nanomaterial, giving it a biological identity that is distinct from its synthetic nature. The biological identity determines the physiological response, including signalling, kinetics, transport, accumulation and toxicity” (EFSA, 2016). Within the environment of the gastrointestinal tract, the composition of this corona likely includes both proteins and bile salts (McCracken et al., 2013). There are highly significant differences in the kinetics of bile salt production, concentration and composition between rats and humans (Tanaka et al., 2012) that would be expected to yield a species-specific modulation of any gastrointestinal impacts observed. Failure to acknowledge such basic impediments to inter-species extrapolation does not provide the reader with an objective contextual framework from which to evaluate the present work. It also questions the relevance of ex vivo, in vitro experiments exposing extracted cells to unmodified TiO2 outside of a biologically relevant environment.
21. Comparing the results of Bettini et al. with other similar research for example the work by Blevins et al, Nogueira et al. or Carpenter et al. the divergent findings in immune cell response as well as in the cytokine levels are disturbing and raise questions as to whether the reported findings by Bettini et al. are in fact test-substance related or more likely chance findings of uncertain physiological relevance.

References:
Bernard, B. K. et al. (1990): Toxicology and Carcinogenesis Studies of Dietary Titanium Dioxide-Coated Mica in Male and Female Fischer 344 Rats. J. Toxicol. Environ. Health 29, 417-429
Bird, R.P. (1987): Observation and quantification of aberrant crypts in the murine colon treated with a colon carcinogen: preliminary findings, Cancer Lett. 37, 147
Carpenter, C. (2011): H-29865: Subchronic toxicity 90-day gavage study in rats. Testing laboratory: E. I. du Pont de Nemours and Company, DuPont Haskell Global Centers for Health & Environmental Sciences, P. O. Box 50, Newark, Delaware 19714, U. S. A. Report no.: DuPont-19267-1026. Report date: 2011-08-26
Charles River, Wistar Rat Crl:WI Growth Chart (accessed January 2017). Available from: http://www.criver.com/products-services/basic-research/find-a-model/wistar-rat
EFSA (2016): Re-evaluation of titanium dioxide (E 171) as a food additive, EFSA Panel on Food Additives and Nutrient Sources added to Food (ANS), EFSA Journal 2016,14(9): 4545
Esfandi, F et al. (2006): Interleukin-6 level in patients with colorectal cancer, Cancer Letters 244 (2006) 76–78
Geraets et al. (2014): Tissue distribution and elimination after oral and intravenous administration of different titanium dioxide nanoparticles in rats, Particle and Fibre Toxicology, 11:30
Khare, S. et al. (2009): Aberrant crypt foci in colon cancer epidemiology, Methods Mol Biol, cancer Epidemiol 472, 373-386
Lance, P.; Hamilton, S.R. (2008): Sporadic aberrant crypt foci are not a surrogate for colorectal adenoma prevention, Cancer Prev Res June 1 2008 1(1) 4-8; DOI:10.1158/1940-6207
McCraken, C. et al (2013): Minimal intestinal epithelial cell toxicity response to short- and long-term food-relevant inorganic nanoparticle exposure. Chem. Res. Toxicol. 26, 1514 – 1525.
National Cancer Institute (1979): Bioassay of Titanium Dioxide for possible Carcinogenicity. Testing laboratory: National Cancer Institute, National Institute of Health. Report no.: (NIH) 79-1347
Nogueira, C.M. et al. 2012. Titanium dioxide induced inflammation in the small intestine. World J. Gastroenterol. 18, 4729-4735
Pretlow, T.P. et al. (1991): Aberrant crypts: putative preneoplastic foci in human colonic mucosa, Cancer Res. 51, 1564-1567
Sengupta, P. (2013). The Laboratory Rat: Relating Its Age With Human’s. International Journal of Preventive Medicine, 4(6), 624–630
Tanaka, Y., et al (2012): Regional differences in the components of the luminal water from rat gastrointestinal tract and comparison with other species, J. Pharm. Phamaceut Sci. 15(4), 510 – 518
Thorup, I. (1996): Aberrant crypt foci in the colo-rectal mucosa as reliable markers of tumor development - with special reference to histomorphological and immunohistochemical characterization of aberrant crypt foci in rats, Institute of Toxicology Institute of Life Sciences and Chemistry,
National Food Agency of Denmark Roskilde University Center, Denmark, 1996
Uchiyama, T. et al. (2012): IL-6 Plays Crucial Roles in Sporadic Colorectal Cancer through the Cytokine Networks including CXCL7, Journal of Cancer Therapy (3), 874-879
Waldner, M. et al. (29012): Interleukin-6 - A Key Regulator of Colorectal Cancer Development, Int J Biol Sci 8(9), 1248–1253.
Wang, Z. et al. (2015): Prognostic and clinicopathological significance of serum interleukin-6 expression in colorectal cancer: a systematic review and meta-analysis, OncoTargets and Therapy 2015:8 3793–3801
Warheit (2015): Acute and subchronic oral toxicity studies in rats with nanoscale and pigment grade titanium dioxide particles D. B. Warheit, S. C. Brown, E. M., Donner Food Chem Toxicol. 2015 Oct; 84: 208–224
Wu D et al. (2013): Interleukin-17 - a promoter in colorectal cancer progression, Clin Dev Immunol. 2013, 436307

Data source

Reference
Reference Type:
publication
Title:
Food-grade TiO2 impairs intestinal and systemic immune homeostasis, initiates preneoplastic lesions and promotes aberrant crypt development in the rat colon
Author:
Bettini, S. et al.
Year:
2017
Bibliographic source:
Nature, Scientific reports, 7: 40373

Materials and methods

Test guideline
Qualifier:
no guideline followed
Principles of method if other than guideline:
The study investigated potential genotoxic effects of two different TiO2 particles (P25 Aeroxide (nanoscale) and E171 (pigment-scale, containing a nanoscale fraction). Male adult rats got a daily intragastric gavage of TiO2 particle suspensions (10 mg/kg of BW/day) for seven days, a comet assay was performed subsequently with Peyer’s patches (PP).
GLP compliance:
not specified
Remarks:
not specified in the publication
Type of assay:
mammalian comet assay

Test material

Constituent 1
Chemical structure
Reference substance name:
Titanium dioxide
EC Number:
236-675-5
EC Name:
Titanium dioxide
Cas Number:
13463-67-7
Molecular formula:
O2Ti
IUPAC Name:
dioxotitanium
Test material form:
other: solid: nanomaterial/bulk
Details on test material:
- Name of test materials (as cited in the study report) (i) P25 Aeroxide (NM-105), (ii) E171
- Primary particle diameters: (i) 10-45 nm, (ii) 20-340 nm (44.7% were below 100 nm)
- Crystallite phase: (i) 85% anatase, 15% rutile (ii) mainly anatase
Specific details on test material used for the study:
not specified

Test animals

Species:
rat
Strain:
Wistar
Details on species / strain selection:
not specified
Sex:
male
Details on test animals or test system and environmental conditions:
TEST ANIMALS
- Source: Janvier Labs (France)
- Weight at study initiation: 175-200 g
- Age at study initiation: adult (not further specified)

All animal experiments were performed in accordance with the guidelines of European legislation (Council Directive 2010/63/UE) and French Decree 2013-118 on the protection of animals used for scientific purposes and were approved by the Local Animal Care and Use Committee (TOXCOM-0036-EH-EH) of Toulouse Midi-Pyrénées (agreement CEEA-86). The animal facilities used are licensed by the relevant local authorities for rodents (agreement C31 555 13).All animal experiments were performed in accordance with the guidelines of European legislation (Council Directive 2010/63/UE) and French Decree 2013-118 on the protection of animals used for scientific purposes and were approved by the Local Animal Care and Use Committee (TOXCOM-0036-EH-EH) of Toulouse Midi-Pyrénées (agreement CEEA-86). The animal facilities used are licensed by the relevant local authorities for rodents (agreement C31 555 13).

Administration / exposure

Route of administration:
oral: gavage
Vehicle:
- Vehicle used: water
- Amount of vehicle: 200 µl
Details on exposure:
PREPARATION OF DOSING SOLUTIONS:
The TiO2 products were prepared following the generic Nanogenotox dispersion protocol [*][**].

*Final protocol for producing suitable manufactured nanomaterial exposure media. The generic NANOGENOTOX dispersion
protocol. July, 2011. http://www.nanogenotox.eu/files/PDF/Deliverables/nanogenotox%20deliverable%203_wp4_%20dispersion%
20protocol.pdf.

**Jensen, K. A. et al. Final protocol for producing suitable manufactured nanomaterial exposure media (Nanogenotox, 2011). Available
at: http://www.nanogenotox.eu/files/PDF/Deliverables/nanogenotox%20deliverable%203_wp4_%20dispersion%20protocol.pdf
(Accessed: june 19th 2012).
Duration of treatment / exposure:
7 days
Frequency of treatment:
daily
Post exposure period:
not specified
Doses / concentrations
Dose / conc.:
10 mg/kg bw/day
Remarks:
Total dose: 70 mg/kg bw/day
No. of animals per sex per dose:
5 male rats
Control animals:
yes, concurrent vehicle
Positive control(s):
not specified

Examinations

Tissues and cell types examined:
DNA damage was assessed on PP. Fifty cells per slide and two slides per sample were analyzed.
Details of tissue and slide preparation:
CRITERIA FOR DOSE SELECTION:
- E171: human relevant levels
- not further specified

TREATMENT AND SAMPLING TIMES:
A parallel digestion with formamido pyrimidine glycosylase (FPG, gift from Serge Boiteux), allowing the detection of oxidative damages, was performed as already described*.

DETAILS OF SLIDE PREPARATION:
PP and scraped colonic mucosa were collected in cold NaClethylenediamine tetra-acetic acid (EDTA) (0.075 M/0.024 M, respectively; pH=7.5). Cells
were isolated using a Dounce homogenizer, embedded in 0.7% low melting point agarose (Sigma-Aldrich) and laid on CometAssay® HT Slide (Trevigen).

METHOD OF ANALYSIS:
Fifty cells per slide and 2 slides per sample were analyzed. The extent of DNA damage was evaluated for each cell through the measurement of intensity of all tail pixels divided by the total intensity of all pixels in head and tail of comet. The median from these 100 values was calculated, and named % tail DNA.

*Azqueta, A., Arbillaga, L., Lopez de Cerain, A. & Collins, A. Enhancing the sensitivity of the comet assay as a genotoxicity test, by combining it with bacterial repair enzyme FPG. Mutagenesis 28, 271-277 (2013).
Evaluation criteria:
see 'Statistics'
Statistics:
The results are expressed as the mean ± s.e.m. Statistical significance was assessed using Prism 4 software (GraphPad). Kruskal-Wallis test, followed by post hoc tests were used, or contingency analysis followed by Fisher’s exact test was used, when appropriate. A P value of < 0.05 was considered significant.

Results and discussion

Test results
Sex:
male
Genotoxicity:
negative
Toxicity:
not specified
Remarks:
A dose-dependent cytotoxic and anti-proliferative effect on T-cells was observed. This effect was found to be more pronounced with E171 compared to the NM-105 TiO2-NP model. No further statements on cytotoxicity. Clinical observations are not specified.
Vehicle controls validity:
valid
Negative controls validity:
not specified
Positive controls validity:
not specified
Additional information on results:
COMET ASSAY
No increase in DNA damage was detected in PP cells of the E171- and NM-105-treated rats

Applicant's summary and conclusion

Conclusions:
Male rats got daily intragastric gavages of two different TiO2 particle types (suspensions of: NM-105 (P25) or E171) for seven days at a concentration of 10 mg/kg bw/day. Afterwards, comet assay was performed and 100 Peyer's Patch (PP) cells were analyzed regarding their % tail DNA. NM-105 and E171 cytotoxicity was measured ex vivo in cells isolated from PP of untreated cells.

According to the authors, treatment with the two different TiO2 suspensions (without as well as with Fpg treatment) did not result in increased DNA damage in PP cells when compared to controls. A dose-dependent cytotoxic and anti-proliferative effect on the T cells was observed, and this effect was found to be more pronounced with E171 compared to the NM-105 TiO2-NP model.