Source :
SCCS (2012)

Résumé & Détails :
Media Consulta

A propos...

Zinc oxide (nano form)

Previous Question

Level 3 Questions

Next Question

4. How have ZnO nanoparticles been tested for safety?

Acute toxicity

Acute oral toxicity

The acute oral toxicity of one ZnO nanoparticle product (FINEX-50) was investigated with a dose of 2,000 mg/kg.

Study Design
Guideline/method:	OECD Guideline 423 (2001).
Species/strain:	Rat/Sprague-Dawley (Crl:CD(SD)
Group size:	Three males
Test substance:	FINEX-50 ZnO (supplied by Sakai Chemical Industry Co., primary particle size: 20 nm, non-coated).
Batch:	OZ52
Purity:	≥96% as indicated in the submission
Dose levels:	2,000 mg/kg bw as 20% (W/V) aqueous suspension.
Dose volume:	10 mL/kg bw
Vehicle:	Water for injection.
Route:	Oral (gavage)
Exposure:	Single application
Observation period:	14 days
GLP:	No
Published:	No
Reference:	99

Study period Eight days
Date of report: March 2006

The maximum dose (2,000 mg/kg body weight) as specified in OECD Guideline 423 for acute oral testing was used. The dose was administered as a 25% aqueous suspension by oral gavage. The animals were observed prior to dosing and for treatment-related effects after 2 hours and on day 1, and then at least once a day for a total of one week. Body weights were determined on days 1, 2, 3, 4, 6 and 8 after application. Gross pathology was performed in all rats at two weeks after oral administration.

Results

No animals died during the course of the study. On day 3 of observation, decreases in body weight and fecal excretion were observed in one animal, but these findings disappeared four days after dosing. There were no abnormal findings at necropsy at two weeks after oral administration.

Conclusion

The approximate acute oral toxicity (LD₅₀) was >2,000 mg/kg bw for male Sprague-Dawley rats.

(Reference: 99)

Comment

For the ZnO used, information on surface area and number of particles per mass was not provided. Also, information on the dose expressed as surface area and number of particles was not provided.

The purity of the FINEX-50 used was indicated in the submission to be ≥96%. However, data on the purity is not provided in the physico-chemical characterization of the FINEX-50 (section 3.1.4), and was also not provided in the study reference 99.

The SCCS agrees that the acute oral toxicity of the ZnO materials tested in rat is >2,000 mg/kg body weight.

Additional study

An exploratory acute oral toxicological investigation comparing the effects of nano-scale (20 nm) or pigmentary (submicron-scale 120 nm) ZnO (Produced by Jiangsu Haitai Nanomaterials Co. Ltd. Jiangsu Province, China) was performed in mice as reported by Wang et al. (Reference: 118). The primary mean size as determined by TEM was 20 ± 5 nm and 120 ± 20 nm, respectively. Surface area was 4.4×10⁵ cm²/g and 9.1×10⁴ cm²/g, respectively. The particle number concentration was 2.2×10¹⁶/g and 2.0×10¹⁴/g, respectively. When administered in a 1% sodium carboxymethylcellulose solution the size range of these materials was between 30 nm and 70 nm, and between 186 nm and 190 nm, respectively. Male (n=5) and female (n=5) CD-ICR mice received a single oral application of 20 nm ZnO or 120 nm ZnO powder at dose levels of 1,000, 2,000, 3,000, 4,000 and 5,000 mg/kg bw suspended in 1% sodium carboxymethylcellulose. Doses expressed as surface area were: 4.4-8.8-13.2-17.6-22x10⁵ cm²/kg bw for the 20 nm ZnO; and 9.1-18.2-27.3-36.4-45.5x10⁴ cm²/kg bw for the 120 nm ZnO. Doses expressed as number of particles were: 2.2-4.4-6.6-8.8-11x10¹⁶ particles/kg bw for the 20 nm ZnO; and 2-4-6-8-10x10¹⁴ particles/kg bw for the 120 nm ZnO. The control animals received the vehicle only. The animals were observed for 14 days and at termination blood was sampled for selected hematological and clinical pathology examinations. Several organs were collected for histopathological investigations and ZnO content analysis. For the 120 nm sized ZnO the highest dose (5,000 mg/kg) induced the highest mortality; three out of five females and two out of five males died, whereas one out five females died in the 4,000 mg/kg group, and one out of five females died in the 2,000 mg/kg group. For the 20 nm sized ZnO, one out of five males died in the 5,000 mg/kg group, and one out of five females died in the 2,000 mg/kg group. In addition, effects were reported showing alterations in individual hematology, clinical pathology parameters, and pathological findings (stomach, liver, heart, spleen). Indications of Zn accumulation in selected tissues (bone, pancreas and kidney) were observed for both 20 nm and 120 nm ZnO.

(Reference: 118)

Comment

In view of the size ranges reported in Reference 118, the larger ZnO cannot be considered as pigment grade. The published paper describes the particles as submicron-sized with a mean size of 120 nm. Although not all effects showed a clear dose response relationship, many effects were noted at the highest dose administered that should be considered dose related even when they only occur at the highest dose. The pathological findings indicate a dose response relationship with the highest dose inducing the most severe alterations. The lowest dose investigated (1,000 mg/kg) also induced pathological effects in the animals. Most effects showed some minor differences between the 20 and 120 nm sized ZnO. Target organs for toxicity were liver, heart, spleen, pancreas and bone. A no observed adverse effect level (NOAEL) was not dentified because alterations were also observed at the lowest dose investigated (1,000 mg/kg).

No deaths were observed in a similar study using a single oral dose of 5 mg/kg in mice (Reference: AR16). ZnO nanoparticles were approximately 50 nm in size (TEM evaluation), which were compared to ZnO microparticles showing at least one diameter >100 nm (TEM evaluation). DLS evaluation showed an average hydrodynamic diameter of 1,226 ± 120 nm for the ZnO microparticles, and an average hydrodynamic diameter of 93 ± 14 nm for the ZnO nanoparticles. After oral administration for both ZnO nanoparticles and microparticles, Zn could be observed in serum indicating uptake from the GI–tract, either as particulate material, dissolved Zn ions, or in both forms. In liver, spleen and lung, ZnO nanoparticles showed higher Zn distribution compared to ZnO microparticles. Serum liver enzyme analysis (aspartate aminotransferase (AST), alanine aminotransferase (ALT) and lactate dehydrogenase (LDH)) indicated liver toxicity due to both ZnO nanoparticles and ZnO microparticles which was confirmed by histopathology (Reference AR16). Histopathological lesions were only observed for ZnO nanoparticles in the liver (Reference 118).

Additional studies in the open literature

An exploratory study of acute oral toxicity with a commercial ZnO nanomaterial is summarized below.

Study Design

Guideline/method: OECD Guideline 423 (2001)

Species/strain: Rats, Sprague-Dawley, in house animal facility.

Group size: Five males and five females per group.

Test substance: ZnO nanoparticles (20 nm, 63 nm in SEM) from Nanostructured and Amorphous Materials Inc., USA), surface area 50 m²/g. Micro- sized ZnO from Sigma- – Aldrich (product no. ZO385).

Batch: Stock no. 5180HT

Purity: Not stated

Dose levels: 5, 50, 300, 1,000, 2,000 mg/kg bw as 20% (W/V) aqueous suspension, dose as surface area 0.25, 2.5, 15, 50, and 100 m²/kg bw.

Dose volume: 10 mL/kg bw

Vehicle: Distilled water

Route: Oral (gavage)

Exposure: Single application

Observation period: 14 days

GLP: No

Published: Yes

Study period: 14 days

Date of report: 2011

In this study, the acute oral toxicity of ZnO nanomaterial was compared to a ZnO preparation from Sigma-Aldrich. The study was performed according to the OECD guideline for acute oral toxicity OECD Guideline 423 (2001). Toxicity and animals were evaluated up to 14 days after a single administration by oral gavage.

Results

There was an inverse relationship with serum levels of liver enzymes (aspartate aminotransferase (AST) and alanine aminotransferase (ALT), the lower doses inducing the highest increase. On a weight basis the nano-sized material induced serum alterations in AST and ALT that were not observed with the commercial non-nano ZnO (Sigma-Aldrich). The incidences of microscopic lesions in the liver, pancreas, heart and stomach were higher at lower doses of nano-size zinc oxide compared to higher doses. It was concluded that nano-sized zinc oxide exhibited toxicity at lower doses probably due to lower aggregation of the material at low concentrations. No measurements of Zn ions (or particles) in blood were reported. Thus on a worst-case basis, it can be assumed that the translocating ZnO was in nanoparticulate form (SCCS Guidance on Nanomaterials in Cosmetics SCCS/1484/12). The nano-ZnO induced histopathological lesions in the liver, pancreas, stomach and heart at all the doses investigated (lowest dose 5 mg/kg), whereas the commercial ZnO did so only at the highest dose investigated (2,000 mg/kg).

Conclusion

The nano-sized ZnO induced toxicity at the lowest dose investigated (5 mg/kg body weight).

(Reference AR20)

Comment

In this study, the size of the commercial micro-ZnO is not reported. Also for the ZnO information on number of particles per mass was not provided. Information on dose expressed as number of particles was not provided. The acute toxicity evaluation was performed based on blood parameters and histopathology. Body weights were not reported.

Exploratory study

Study Design

Guideline/method: No

Species/strain: ICR mice

Group size: Five males and five females.

Test substance: ZnO nanoparticles (50 nm, TEM evaluation) and ZnO microparticles (1,226i nm, DLS measurement) from Top Nanotechnology Co. Ltd., Taiwan.

Batch: Not stated

Purity: Not stated

Dose levels: 1.25, 2.5, 5 g/kg bw.

Dose volume: 10 mL/kg bw

Vehicle: Sterile water plus 1% hydroxypropyl methyl cellulose.

Route: Oral (gavage) and intraperitoneal administration.

Exposure: Single application

Observation period: 14 days

GLP: No

Published: Yes

Study period: 14 days

Date of report: 2011

In this study, ZnO nanoparticles used were approximately 50 nm in size (TEM evaluation), which were compared to ZnO microparticles showing at least one diameter >100 nm (TEM evaluation). DLS evaluation showed an average hydrodynamic diameter of 1,226 ± 120 nm for the ZnO microparticles and an average hydrodynamic diameter of 93 ± 14 nm for the ZnO nanoparticles. Toxicity was evaluated by analysing serum biochemistry parameters and histopathology. Genotoxicity was evaluated using the Ames test and the micronucleus test. Zinc absorption and tissue distribution was evaluated by ICP-MS determination of elemental Zn. In addition an in vitro cytotoxicity assay was performed.

Results

All animals survived a single oral or intraperitoneal dose of 5 g/kg body weight of both nano-sized ZnO (50 nm) and micro-sized ZnO (1,226 nm), although the large ZnO particles induced a body weight reduction in both male and female animals. After oral and intraperitoneal administration of a single dose of 2.5 g/kg bw for both ZnO nanoparticles and microparticles, Zn could be observed in serum indicating uptake from the GI-tract. For ZnO nanoparticles, the systemic availability was significantly higher (approximately 24% to 50%) at 2, 4 and 6 hours after administration compared to that of ZnO microparticles as indicated by Zn measurements by ICP-MS. In the liver, spleen and lung, treatment with ZnO nanoparticles showed a higher distribution of Zn compared to treatment with ZnO microparticles. Serum liver enzyme analysis (aspartate aminotransferase (AST), alanine aminotransferase (ALT) and lactate dehydrogenase (LDH)) indicated liver toxicity at 24, 48 and 72 hours after treatment, both for ZnO nanoparticles and ZnO microparticles. This was confirmed by histopathology. Histopathological lesions were only observed for ZnO nanoparticles in the liver.

The in vitro studies revealed a dose related induction of cytotoxicity for both nano-ZnO and micro-ZnO up to a concentration of 100 μg/mL. The Ames test (in vitro) and the in vivo micronucleus test were negative for both nano-ZnO and micro-ZnO.

(Reference: AR16)

Comment

For the ZnO nanoparticles and microparticles, information on surface area and number of particles per mass was not provided. Information on dose expressed as surface area and number of particles was not provided.

The results indicate a similar uptake, tissue distribution, and toxicity of ZnO nanoparticles and ZnO microparticles. As Zn was measured by ICP-MS, it cannot be concluded whether it was present in particulate form or as dissolved ionic Zn. Although serum levels of aspartate aminotransferase, alanine aminotransferase and lactyate dehydrogenase indicated similar liver toxicity for both ZnO nanoparticles and ZnO microparticles, histopathological lesions were only observed for ZnO nanoparticles in the liver.

Acute dermal toxicity

No information provided.

Acute inhalation toxicity

No information provided.

Irritation and corrosivity

Skin Irritation

The skin irritation potential was determined for one ZnO nanomaterial (FINEX-50) according to Japanese guidelines.

Study Design
Guideline/method:	Guide to marketing and manufacturing of cosmetics and quasi drugs in Japan.
Species/strain:	Guinea pig/Hartley
Group size:	Three males per concentration.
Test substance:	FINEX-50 ZnO (supplied by Sakai Chemical Industry Co., primary particle size: 20 nm, non-coated).
Batch:	OZ52
Purity:	≥96% as indicated in the submission
Vehicle:	Ethanol
Dose level:	25% or 40% dispersion in ethanol.
Dose volume:	0.05 mL, once a day for three consecutive days.
Exposure:	Intact skin (open)
GLP:	No
Study period:	Up to three days

Date of report: February 2009

The cumulative acute dermal irritation potential was investigated in individually housed Hartley Guinea pigs. The hair was clipped on the dorsal area of the trunk one day prior to application and prior to examination on day 3. An amount of 0.05 mL of the test substance as a 25% or 40% ethanolic dispersion was applied to the test site (2 × 2 cm) on the right back of each of the six animals and left exposed without dressing. Percutaneous administration was used for the substance. The administration was conducted once daily for three consecutive days. The animals were examined for erythema, incrustation and edema on each day of administration.

Results

There were no signs of systemic toxicity and no mortality. Slight erythema was observed in one of three animals in the 40% test substance group on day 3 of the administration. No other abnormal dermal changes were observed at any time during the administration and observation periods. No abnormal dermal changes were observed at any time during the administration and observation periods in the 25% test substance group.

Conclusion

The three day consecutive skin irritation study in male Guinea pigs led to skin irritation scores of 0.1 and 0.0 for the 40% and 25% test substance dispersions indicating a weak irritation according to the Japanese guideline used.

(Reference: 101)

Comment

For the ZnO information on surface area and number of particles per mass was not provided. Information on dose expressed as surface area and number of particles was not provided.

This test does not correspond to OECD Guideline 404 which indicates that rabbits should be used for the in vivo irritation test, and gives a dermal dose of 0.5 mL per application site using a gauze patch.

Mucous membrane irritation

The mucosal irritation potential was determined for one ZnO nanomaterial (FINEX-50) according to Japanese guidelines.

Study Design
Guideline/method:	Guide to marketing and manufacturing of cosmetics and quasi drugs in Japan and according to Draize (1959) and Gullot et al. (1982) in total comparable to OECD Guideline 405.
Species/strain:	Rabbit/Japanese White
Group size:	Three males per concentration.

Test substance: FINEX-50 ZnO (supplied by Sakai Chemical Industry Co., primary particle size: 20 nm, non-coated).

Batch: OZ52

Purity: ≥96% as indicated in the submission

Vehicle: Olive oil

Dose level: 0.1 g of neat (unchanged) substance or 0.1 mL of a 25% solution in olive oil.

Dose volume: 0.1 mL

GLP: No

Study period: Up to seven days

Date of report: April 2006

The potential irritant effect on the mucous membrane was investigated by instillation of either 0.1 g of the neat test substance or 0.1 mL of a 25% oily solution into the right conjunctival sac of the eye of each of the three rabbits. The 25% oily solution was chosen as the proposed consumer use concentration of the product. The eyes were not washed. The left eyes remained untreated and served as controls. Both eyes of the animals were examined within 1 and 4 hours after application and twice daily on days 1, 2, 3, 6 and 7 according to the method and scoring system of Draize.

Results

Neat test substance group

At 1 hour after administration, redness of the conjunctiva was observed in all animals, and slight or apparent edema in two of three animals and edema accompanying one-half occlusion of the eyelid in one of three animals. Slight discharge was observed in two of three animals. At 4 hours after administration, apparent or diffuse beef-like redness of the conjunctiva was observed in all three animals, and slight or apparent edema in two of three animals and edema accompanying more than one-half occlusion of the eyelid in one of three animals. Slight or apparent discharge was observed in two of three animals. On day 1, apparent redness and slight edema were observed in two of three animals, and slight discharge in one of three animals. On day 2, apparent redness was observed in two of three animals, and slight edema and slight discharge in one of three animals. On day 3, slight or apparent redness was observed in two of three animals. There were no adverse effects recorded in observations made on day 6 post-administration and thereafter.

25% test substance group

Apparent redness of the conjunctiva, slight edema, and slight discharge were observed in all three animals at 1 hour after administration. At 4 hours after administration, there was slight or apparent redness of the conjunctiva in two of three animals and diffuse beef-like redness in one of three animals. In addition, slight or apparent edema was observed in two of three animals. Slight discharge was observed in one of three animals. On day 1 post-administration, slight redness was observed in one of three animals. There were no adverse effects recorded in observations made on day 2 post-administration and thereafter.

The calculated eye irritation indeces based on the findings at 1 hour after administration of the test substance and 25% test substance solution were 11.3 and 8.0, respectively.

Conclusion

The neat and the 25% solution of the test substance were concentration dependent, slightly and transiently irritating to the eyes of three Japanese White rabbits. Therefore, both were assessed as "mild irritant”.

(Reference: 96)

Comment

For the ZnO used in this study, information on surface area and number of particles per mass was not provided. Information on dose expressed as surface area and number of particles was not provided.

In Reference 96, the number of animals investigated was six (three per group) and the numbering is from 1 to 6. However, the results are presented for animal numbers 7 to 12.

Skin sensitization

The skin sensitization potential of one ZnO nanomaterial (FINEX-50) was evaluated using a modification of the guinea pig maximization test as a short-term adjuvant and patch test.

Study Design

Date of report: February 2006

Guideline/method: Short-term adjuvant and patch test method (s-APT according to Reference: AR25).

Species/strain: Guinea pig/Hartley

Group size: Ten females in total (five per group).

Test substance: FINEX-50 ZnO (supplied by Sakai Chemical Industry Co., primary particle size: 20 nm, non-coated).

Batch: OZ52

Purity: ≥96% as indicated in the submission

Vehicle: Distilled water

Concentration:
Induction:
Group A) distilled water
Group B) 40% aqueous test substance suspension

Challenge:
Group A) 5%, 12.5%, 25%, 40% aqueous test substance suspension.
Group B) 1.25%, 2.5%, 5%, 12.5%, 25%, 40% aqueous test substance suspension.

Route/Exposure:
Induction:
1^st induction: intradermal injection with Freund’s complete adjuvant followed by occlusive topical application of the vehicle or 40% aqueous test concentration for 72 hours.
2^nd induction: six days after 1^st induction occlusive topical application of the vehicle or 40% aqueous test concentration for 48 hours.

Challenge: 13 days after 1^st induction, open application of test concentrations.

Skin evaluations: 24, 48, 72 hours after challenge.

Study period: 16 days

GLP: No

Published: No

Date of report: February 2006

The sensitizing potential of the test substance was tested in female Hartley Guinea pigs according to the short-term adjuvant and patch test method (s-APT according to Yanagi et al. (Reference: AR25). In total ten animals were tested, divided in two groups of five animals each. Prior to the 1^st induction the hair was removed and the animals were shaved, followed by an intradermal injection of Freund’s complete adjuvant at four sites. (The Guinea pig maximization test according to OECD Guideline 406 uses three injections as induction; one being a mixture of Freund’s complete adjuvant and water, one a mixture of FCA and the test substance, and one of the test substance only). Thereafter, one group received the vehicle (group A) or a 40% aqueous test substance suspension (group B) for 72 hours under an occlusive dressing. Six days after the 1^st induction, the animals were induced a 2^nd time by occlusive topical application of either the vehicle or the 40% aqueous test substance concentration.

The challenge was performed by open topical application of 5%, 12.5%, 25%, 40% aqueous test substance suspension (group A) or 1.25%, 2.5%, 5%, 12.5%, 25%, 40% aqueous test substance suspension (group B), and the skin was evaluated at 24, 48 and 72 hours after the challenge.

Results

Neither the group induced with the 40% aqueous test substance suspension or with the vehicle showed any skin reaction to the challenge concentrations at the evaluation time points of 24, 48 or 72 hours.

Conclusion

It was concluded that the test substance exhibited no potential to induce dermal sensitization in Guinea pigs under the conditions of the study.

(Reference: 95)

Comment

For the ZnO used in this study, information on surface area and number of particles per mass was not provided. Information on the doses expressed as surface area and number of particles was not provided.

The sensitizing potential of the test substance was tested in female Hartley Guinea pigs according to the short-term adjuvant and patch test method (s-APT), which is not a recognized guideline assay. In the classical Guinea pig maximization test (OECD Guideline 406) the induction is performed with a mixture of the test material and FCA, and ten animals are used in the test group. Mixing of the test substance with FCA provides optimal conditions for inducing an immune response, i.e. induction of sensitization if allergic substances are present in the test material.

A concurrent positive control with a well known weak sensitizer as indicated in OECD Guideline 406 (hexylcinnamic aldehyde (CAS No. 101-86-0), mercaptobenzothiazole (CAS No. 149-30-4) and benzocaine (CAS No. 94-09-7)) was not included in this assay, so there is no certainty as to whether the test system used was able to identify weak sensitizers. However, in the paper of Yanagi et al. (Reference: AR25) eight contact sensitizers were found to show similar reactions in the GPMT performed according to the OECD Guideline 406, and in this shortened test.

The SCCS considers that the validity of this (or any) test for demonstrating sensitization potency of nanomaterials has not yet been demonstrated. The inclusion of a positive particle control might overcome this problem. However, no positive particle control has been identified thus far.

Human repeat insult patch test

Study Design

Date of report: October 2007

Guideline/method: According to internal laboratory methodology.

Species: Human

Group size: 50 volunteers (six males and 44 females)

Test substance:
A) ZnO, coated (Zano 10 Plus, Umicore) (former Zano® Plus, ZnO coated with octyl triethoxysilane).
B) ZnO uncoated (Zano 10, Umicore).

Batch:
A) 22-5D
B) No data

Route: Dermal occlusive application by patch.

Procedure:
Induction:
Nine consecutive 24 hour exposures on Monday, Wednesday and Friday for three consecutive weeks.

Rest period: 10–14 days

Challenge: After rest period one 24 hour exposure as during the induction period.

Reaction scoring: 24 and 48 hours after challenge application.

Concentration: 25% in corn oil (0.2 mL dispensed on occlusive, hypoallergenic patch).,

Vehicle: Corn oil

GLP: Yes

Published: No

The test solutions were freshly prepared prior to applications. Of 52 volunteers enrolled for the study, 50 volunteers (six males and 44 females aged 21–66 years; Caucasian and Hispanic) completed the study.

The oily test material dispersions were dispensed onto an occlusive, hypoallergic patch. After 24 hours the patch was removed and the procedure was repeated until a series of nine consecutive 24 hour exposures were made on Monday, Wednesday and Friday for three consecutive weeks. The volunteers were given a rest period of 10–14 days after which a challenge was applied once to a previously unexposed test site. The challenge was equivalent to any of the original procedures. The skin reactions were scored 24 and 48 hours after application on a scale graded 0–4.

Results

Under the selected conditions of this study none of the 50 investigated volunteers showed any skin reaction at any time. All volunteers were consequently scored with grade 0.

Conclusion

The applicant concluded that coated and uncoated ZnO tested as 25% oily dispersion did not produce any skin irritation or skin sensitization under the conditions of this repeat human insult patch test.

(References: 105, 106)

Comment

For the ZnO used in this study, information on surface area and number of particles per mass was not provided. Information on dose expressed as surface area and number of particles was not provided.

Two volunteers discontinued the treatment at the third application after two applications without any clinical signs. Any follow-up and reasons for discontinuation were not presented in the report.

Repeated dose toxicity

Repeated dose (5 days) inhalation toxicity

Exploratory 5-day lung toxicity study

Study Design

Date of publication: March 2010

Guideline/method: Exploratory study with inhalation exposures according to OECD Guideline 412.

Species/strain: Rat/Wistar

Group size: 17 males per group.

Test substance:
a) Z-COTE® HP1 coated with triethoxycaprylylsilane.
b) Pigmentary zinc oxide powder.

Batch:
a) CNFC0701 (97.3 g Zn/100 g ZnO)
b) S35583-206 (>99.9%, <1 μ)

Route: Inhalation

Concentrations:
a) 0, 0.5, 2.5, 12.5 mg/m³, corresponding to 22,126, 87,044, 233,360 particles/cm³ (SMPS measurement during exposure).)

Information on dose expressed as surface area was not provided.

b) 0, 12.5 mg/m³, corresponding to 219,031 particles/cm³ (SMPS measurement during exposure).

Exposure period: Five days

Frequency of exposure: 6 hours/day

Type of exposure: Head-nose

Exposure conditions: Head-nose exposure systems: aerodynamic exposure systems (INA 60, volume V ≈ 90 L, BASF SE).

Generator systems: Solid particle generators and glass cyclonic separators.

Generation procedure: The test substance was used unchanged. For each concentration, a solid particle feeder was used for generating the dust. The control group was exposed to conditioned air.

Observations: The examinations were restricted to clinical observation, body weight data, broncho-alveolar lavage with clinico-chemical and cytological evaluation of lavage fluid, gross necropsy and histopathological examination of the respiratory tract.

Recovery period: About 14 days

GLP: Yes

Published: No

The purpose of this study was to determine the pulmonary toxicity of Z-COTE® HP1 coated with triethoxycaprylylsilane in rats using a short-term bioassay including broncho-alveolar lavage with clinico-chemical and cytological evaluation of lavage fluid and serum, as well as pathological examination of the lung and pulmonary cell proliferation measurements. The concentration response relationship should be established as well as a potential No Observed Adverse Effect Concentration (NOAEC). To elucidate the influence of particle size, pigmentary ZnO power was tested at one concentration, which is comparable to the high concentration of Z-COTE HP1. Seventeen male Wistar rats per test group and time point were head-nose exposed to respirable dusts for 6 hours per day, on five consecutive days. The target concentrations for Z-COTE® HP1 were: 0.5, 2.5 and 12.5 mg/m³ and for pigmentary zinc oxide, powder < 1μ was: 12.5 mg/m³. A concurrent control group was exposed to conditioned air. Animals were sacrificed on study days 4 and 25. On each sacrificing day, nine animals per group were designated for histopathological examination. Moreover, organ burdens were determined in three animals per group. On study days 7 and 28, the lungs of the five animals per group were lavaged, and the broncho-alveolar lavage fluid (BALF) was analyzed for markers indicative of injury of the broncho-alveolar region. On exposure days clinical examination was performed before, during and after exposure. During the post exposure period clinical findings were recorded once on each working day. The body weight of the animals was determined. The concentrations were produced with brush particle generators. The dust concentration was determined by gravimetrical measurements. Particle size was determined by gravimetrical cascade impactor measurements, optical particle counter and a scanning mobility particle sizer.

Results

The inhalation of Z-COTE® HP1 coated with triethoxycaprylylsilane for five days resulted in local inflammation in the lungs of the rats, indicated by changes in several parameters in the BALF and histological examinations.

Secondary to the effects in the lung, activation of the draining lymph nodes was noted. Moreover, minimal to moderate necrosis of the olfactory epithelium was observed. The effects occurred in a concentration-related manner and were reversible within the recovery period. Only a multifocal increase in alveolar macrophages was still present at the end of the recovery period. Similar effects were also observed in the animals exposed to pigmentary ZnO powder. At the lowest concentration of 0.55 mg/m³, increased levels of a few mediators in the BALF and in serum were determined. Moreover, minimal (grade 1) multifocal necrosis of the olfactory epithelium was noted in the nasal cavity in one of the six animals treated with the lowest dose. Therefore, a No Observed Adverse Effect Concentration (NOAEC) could not be established in this study. The lowest concentration of 0.55 mg/m³ is considered to be the Low Observed Adverse Effect Concentration (LOAEC).

Conclusion

ZnO induced a concentration-related inflammation reaction in the lung which was associated with dose-dependent increases in BALF markers. In addition to the inflammation reaction, necrosis was detected in the lung and the nose. There was no biologically relevant difference between the nano-sized and pigmentary ZnO. As ZnO is soluble in lung fluid and zinc ions are cytotoxic at higher concentrations necrosis can be attributed to the zinc ions dissolved from the ZnO particles. Likewise, elevated zinc levels were detected in various organs, most likely due to zinc ions dissolved from the ZnO particles. There was, however, no indication of systemic effects.

(References: 30, 71)

Comment

For the ZnO used in this study information on surface area and number of particles per mass was not provided. Information on dose expressed as surface area was not provided.

At the lowest dose investigated there were minimal alterations as indicated by the presence of an increased number of alveolar macrophages after the recovery period at day 25.

Repeated dose (28 days) dermal toxicity

Additional study in the open literature

Exploratory in vivo study on repeated dose dermal toxicity

Study Design

Guideline/method: Exploratory study on repeated dose dermal toxicity of zinc oxide

nanoparticles based on OECD Guideline 410 with modifications for dose levels, biochemical parameters and measurement of collagen content.

Species: Sprague-Dawley rats

Group size: Ten animals (five males and five females).

Test substance ZnO nanoparticles were obtained from Nanostructured and Amorphous Materials Inc., USA, stock no. 5810HT, surface area 50 m²50m²/g. ZnO was obtained from Sigma-Aldrich, product no. ZO385.

Particle sizes: ZnO nanoparticles; 20 nm according to manufacturer; 63 nm by SEM evaluation; 224.7 nm by DLS in aqueous solution.

Dose applied: ZnO nanoparticle doses, 75, 180, and 360 mg/kg bw, surface area 3.75, 9, and 18 m²/kg bw., ZnO 2,000 mg/kg bw.

Skin area: 10% of the total body surface

Route: Topical application

Exposure time: 6 hours per day for five days per week for a 28 day period.

Evaluation: Blood samples on day 29 and for satellite groups at day 42. Skin and tail for collagen content. Gross pathology and organ wet weight. Histopathology of organs.

GLP: No

Date of report: 2012

Published: Yes

In this exploratory study, rats were treated for five days per week at 10% of the total body skin area. Animals were treated for a total of 28 days and evaluated for toxicity at day 29. After a recovery period satellite groups were evaluated at day 42. The dosages administered were calculated to be similar to the daily uptake of nanomaterial per kg body weight per day. The average uptake (administration) for a woman was determined to be 12 mg/kg bw using a 2% concentration of ZnO in a sunscreen product, and a recommended use of 36,000 mg per day. Using conversion factors for the rat of 1, 2.5, and 5 and a body weight of approximately 180 g, the calculated doses were: 72 mg/kg bw (rounded to 75 mg/kg bw) for the low dose; 180 mg/kg bw for the intermediate dose; and 360 mg/kg bw for the high dose. Toxicity was evaluated at days 28 and 42. Hydroxyproline content of the skin and tail was estimated, and the collagen content was calculated.

Results

No significant changes were observed in the clinical chemistry parameters in both micro and nano zinc oxide treated rats. There were no statistically significant changes in the hematologic parameters compared with the control. A statistically significant increase in clotting time was observed in all treatment groups of nano zinc oxide compared to the micro-sized zinc oxide.

No gross pathology or histopathological lesions were observed in any of the organs investigated.

There was a significant decrease in the collagen content of the skin and the tail in all the nano ZnO treated groups of rats compared to the control, as well as with the micro-sized zinc oxide treated groups. The loss was higher in the skin than in the tail. There was an inverse dose relationship with the higher doses inducing a lower decrease. The decrease induced by the micro-sized ZnO (2,000 mg/kg bw) was similar to the decrease of the highest dose of nano-sized ZnO (360 mg/kg bw). A maximum decrease of more than 50% was observed in three treatment groups (skin 75 mg/kg bw males, 180 mg/kg bw males, and tail 75 mg/kg bw males).

Conclusion

The decrease in collagen content of the low dose group of nano ZnO was significant compared to the high dose group and control group in both the tail and the skin. It was suggested that this effect was due to potential skin penetration of ZnO nanoparticles due to partial dissolution, followed by induction of reactive oxygen species.

(Reference: AR23)

Comment

For the ZnO used in this study information on number of particles per mass was not provided. Information on dose expressed as number of particles was not provided.

This is the only study available on repeated dose toxicity in the skin. In general, no systemic effects were observed. The data indicate a decrease in skin collagen content. Although statistics were not presented in the paper, the data suggest statistically significant differences. The inverse dose response relationship for the decrease in the collagen content cannot be explained and needs further confirmation.

Repeated dose (30 days) oral toxicity

No data available.

Sub-chronic (90 days) toxicity (oral, dermal)

No data available.

Chronic (>12 months) toxicity

No data available.

Mutagenicity/genotoxicity

Mutagenicity/genotoxicity in vitro

Bacterial gene mutation assays

Guideline/method: OECD Guideline 471 (1997)

Species/strain: Salmonella typhimurium strains TA 98, TA 100, TA 102, TA 1535, TA 1537.

Replicates: Triplicate plates, two independent experiments (SPT, PIT).

Test substance: ZnO (Z-COTE® Max) (ZnO coated with dimethoxydiphenylsilanetriethoxy-caprylylsilane cross-polymer).

Batch: FCGB0701

Purity: 100.4 g/100 g).

Solvent: DMSO (standard plate test) or fetal calf serum (pre-incorporation test)

Concentrations: Standard plate test (SPT): 0, 20, 100, 500, 2,500, 5,000 μg/plate both without and with metabolic activation (S9-mix)).

Preincubation test (PIT): 0, 20, 100, 500, 2,500, 5,000 μg/plate (with/without metabolic activation (S9-mix)).

Treatment: direct plate incorporation with 48 - 72 h incubation without and with S9-mix pre-incubation method was used with 20 minutes pre-incubation and at least 48 – 72 h incubation time both without and with S9-mix

GLP: in compliance

Date of report: March 2009

The test substance was tested for mutagenicity in the reverse mutation assay on bacteria with and without metabolic activation (S9-mix prepared from phenobarbital/β-naphthoflavone induced male Wistar rat liver) according to the standard plate test (SPT) and the plate incorporation test (PIT). The Salmonella typhimurium strains were exposed to the test substance dissolved in DMSO (SPT) or in fetal calf serum (PIT, added to avoid aggregate formation, and to mimic the protein containing body fluids (e.g. blood)) at concentrations ranging from 20–5,000 μg/plate. For control purposes, a sterility and solvent (DMSO) and positive controls (NOPD, MNNG, AAC, MIT.C, 2-AA) were also investigated. In the standard plate test the revertant colonies were counted after incubation at 37°C for 48

– 72 hours in the dark; in the preincubation test, the duration of preincubation was about 20 minutes whereas the revertant colonies were again counted after incubation at 37°C for 48 – 72 hours in the dark. Negative and positive controls were in accordance with the OECD guidelines.

Results

No sign of bacteriotoxicity seen as reduced his- background growth, a decrease in the number of his+ revertants, or reduction in the titer, was noted when tested up to the highest required concentration in the absence and presence of metabolic activation.

Precipitation of the test substance was recorded at and above 2,500 μg/plate with or without metabolic activation.

The test substance did not induce a biologically relevant increase in revertant colony numbers in the bacterial strains at any concentration tested in the presence or absence of metabolic activation compared to the background control. The sensitivity and validity of the test system used was demonstrated by the expected induction of a significantly increased number of revertants with the positive controls.

Conclusion

The test substance did not induce gene mutations in the bacterial strains used either in the presence or absence of S9-mix up to precipitation concentrations. Thus, it was shown to be non-mutagenic in this bacterial gene mutation test.

(Reference: 27)

Comment

For the ZnO used in this study information on surface area and number of particles per mass was not provided. For the exposures information on dose expressed as surface area and number of particles was not provided.

The positive control used represents a generally accepted chemical positive control for this test and demonstrated its performance. However it does not provide proof that a negative response of a nanomaterial/nanoparticle is really negative. There is presently no accepted nanoparticle positive control that demonstrates whether the assay is generally suitable for the mutagenicity testing of insoluble/poorly soluble nanoparticles.

The behaviour of the ZnO in the test was evaluated as indicated by the observation that there was precipitation at the two highest concentrations of ZnO investigated. Contact of bacterial DNA (i.e. nanoparticle uptake by bacteria) with the ZnO nanomaterials was not demonstrated. Negative findings in the Ames tests have also been shown for coated ZnO (i.e. tetramethylammoniumhydroxide-capped ZnO) (Reference: AR26). Bacterial mutagenicity assays are considered to be less appropriate for the testing of nanoparticles compared to mammalian cell systems (Reference: 72) due to the lack of endocytosis (Reference: AR4).

It is uncertain whether Zn ions that might be available from ZnO nanoparticles may cause mutagenicity in bacterial assays. Reachable intra-bacterial zinc concentrations will depend on the solubility and dissolution kinetics of ZnO nanoparticles (see section 3.1.6) and the available zinc transporter systems. Ionic zinc (e.g. zinc acetate) has tested negative in bacterial mutagenicity tests, while it tested positive in other mutagenicity tests (Reference: AR24).

The SCCS considers that the results of the used bacterial gene mutation assay can be considered negative. However, as there is no certainty on the exposure of the bacterial DNA to the added ZnO nanoparticles this negative result is of a limited value.

Comet assay in human epidermal cells

Guideline/method: According to published protocol

Species/strain: Human epidermal cell line (A431).

Replicates: Duplicate slides per tissue culture well

Test substance: ZnO nanopowder

Batch: No data

Purity: >99%), Sigma–Aldrich, St. Louis, MO, USA.

Solvent: DMEM

Concentrations: 0.001, 0.008, 0.08, 0.8, 5 μg/mL.

Exposure: cells were exposed for 6 hours

Sampling: immediately after the end of treatment

GLP: Not in compliance

Date of publication: January 2009

The DNA damaging potential of ZnO nanoparticles was investigated in a Comet assay in a human epidermal cell line (A431) according to a previously published protocol. At 24 hours after seeding, cells were exposed to different concentrations of ZnO nanoparticles (0.001, 0.008, 0.08, 0.8, 5 μg/mL) for 6 hours. Cell viability was determined with the MTT assay, LDH relase, NR uptake and trypan blue exclusion. After exposure two slides were prepared from each well (one well/concentration) for analysis. These 2 slides were exposed to alkali (pH>10), followed by electrophoresis for 25 minutes at 0.7 V/cm and 300 mA, stained with ethidium bromide and scored for comets. The Comet parameters used to measure DNA damage in the cells were % tail DNA (fraction of DNA in the tail) and Olive tail moment (OTM; arbitrary units, the product of the distance of DNA migration from the body of nuclear core and the total fraction of DNA in the tail). Images from 50 random cells (25 from each replicate slide) were analyzed for each experiment.

Results

The mean hydrodynamic diameter and zeta potential of the nanoparticle suspension in deionized water, as determined by dynamic light scattering (DLS) measurement, was 165 nm and −26 mV, respectively. The average size analyzed by transmission electron microscopy (TEM) was 30 nm. A decrease in cell viability was noted from a concentration of 0.8 μg/mL and higher at an exposure time of 24 h and 48 h. At an exposure time of 6 h, a decrease in viability occurred at 8 μg/mL and higher. The cell viability in the Comet assay exceeded 90% for all experimental groups before and after the treatment as assessed by Trypan blue dye exclusion assay.

A statistically significant and concentration dependent increase in DNA damage was observed compared to the control; in cells exposed to ZnO for 6 hours at 5 and 0.8 μg/mL. ZnO nanoparticles were also found to induce oxidative stress in cells indicated by depletion of glutathione (59% and 51%); catalase (64% and 55%) and superoxide dismutase (72% and 75%) at 0.8 and 0.08_g/ml respectively.

Conclusion

Under the experimental conditions used ZnO nanopowder induced DNA damage in A431 cells and, consequently, ZnO nanopowder was genotoxic (clastogenic and/or mutagenic) in these cells. However, the authors emphasized that the observed genotoxic potential in human epidermal cells may be mediated through lipid peroxidation and oxidative stress.

(Reference: 94)

Comment

For the ZnO used in this study information on surface area and number of particles per mass was not provided. Information on dose expressed as surface area and number of particles was not provided.

S9-mix fraction was not mentioned and presumed not used. However, the usefulness of S9-mix fraction in assays investigating ZnO nanoparticles may be doubtful as ZnO nanoparticles will not be metabolised.

The use of two slides from a tissue culture well cannot be considered a replicate of two. It consists of one single incubation for which two read out slides were used.

The applicant stated that depletion of the anti-oxidative defence systems and cellular protection mechanisms of mitochondria in situ particularly under stress, is a major intracellular source of oxidative stress which is obviously no longer controlled by cellular mechanisms and therefore leads directly to DNA damage. This is therefore not related to nanoparticles per se, but is probably due to the imbalance of Zn ions under artificial in vitro conditions. The applicant thus concluded that the results are a set of cellular effects that are not linked to ZnO nanoparticles, but to Zn solubility under in vitro test conditions.

The authors suggest that the DNA damaging effects may be attributed to lipid peroxidation and oxidative stress as one of the probable causes. This is a plausible explanation for the induction of the observed DNA damage. However, the Zn ions were delivered by the ZnO nanoparticles in this test situation so the damage induced is of relevance for ZnO nanoformulations, irrespective of the mechanism.

In view of this, the SCCS concludes that the used ZnO nanoparticles were genotoxic in the comet assay with A431 human epidermal cells.

Additional studies in the open literature

Comet assay in human nasal mucosal cells

Guideline/method:	According to published protocol
Species/strain:	Human nasal mucosal cells obtained from ten surgery patients (three female and seven male).
Group size:
Test substance:	ZnO nanoparticles (<100 nm, surface area 15–25 m²/g) and ZnO powder (<5 μm) were obtained from Sigma–Aldrich (Steinheim, Germany). Sizes: mean longitudinal diameter of 86 ± 41 nm (mean ± SE) and a mean lateral diameter of 42 ± 21 nm (mean ± SE).
Batch:	Not stated
Purity:	Not stated
Dose levels:	Diluted nanoparticle or powder suspension at end concentrations of 0.01, 0.1, 5, 10 and 50 μg/mL in the well, corresponding to 0.15–-0.25, 1.5–-2.5, 75–-125, 150–-250, 750–1,250-1250 mm²/mL in the
	well.
Route:	In vitro
Vehicle:	Distilled water
Exposure:	Single exposure of 24 hour.

Sampling: immediately after the end of treatment

GLP: Not in compliance

Date of report: 2011

The DNA damaging potential of nano-sized and micro-sized ZnO was investigated, in human mucosal cells. The cells were exposed to ZnO nanoparticles or ZnO powder dilutions for 24 h at 37^oC in a humidified incubator with 5% CO₂. The MTT assay as well as the trypan blue exclusion test were performed in order to measure cell viability of human nasal mucosa cells after exposure to ZnO nanoparticles and ZnO powder. The cells were detached from the wells by trypsinization with trypsin–EDTA (0.05% ) and centrifuged at 500g for 5 min. After removal of the supernatant, the cell pellet was resuspended at 10⁵ cells/ml.

Slides were prepared and placed in the ice-cooled Plexiglas gel electrophoresis chamber (distance between electrodes: 30 cm) for 20 min. Electrophoresis was conducted for another 20 min at 25 V and 300 mA.. The following parameters were analyzed to quantify the induced DNA fragmentation: tail DNA (TD), tail length (TL), and Olive tail moment (OTM) as a product of the median migration distance and the percentage of DNA in the tail. The OTM was applied for statistical analysis.

Cells were prepared for TEM for ZnO particle uptake by the cells. Airway Epithelial Cell Growth-Medium (BEGM) served as negative control, and directly alkylating methyl methane sulphonate (MMS) at 100 μM was used as a reliable positive control of genotoxicity without cytotoxic effects.

Results

The frequency of cells with intracytoplasmatic ZnO nanoparticles was 10%, while particle transfer into the cell nucleus could be observed in 1.5% of the cells. While no cytotoxicity or genotoxicity was observed for the ZnO powder (<5 μm) in the tetrazolium (MTT) assay, in the trypan blue exclusion test, and in the Comet assay, cytotoxic effects were shown at a ZnO nanoparticles concentration of 50 μg/mL (P-value <0.01).

In comparison to the control, a ZnO nanoparticles concentration dependent increase in the Olive tail moment (OTM) as an indicator for genotoxic effects could be seen. The enhanced DNA migration was statistically significant at 10 μg/mL (P-value <0.05) and 50 μg/mL (P-value <0.01).

Also, the proinflammatory cytokine interleukin 8 secretion into the basolateral culture medium was increased at ZnO nanoparticles concentrations of 5 μg/mL. In contrast, the pigmentary grade ZnO did not induce cytotoxicity (MTT assay) or DNA damage (comet assay).

Conclusion

Under the experimental conditions the used ZnO nanoparticles induced DNA damage in human nasal mucosal cells and, consequently, ZnO nanoparticles was genotoxic (clastogenic and/or mutagenic) in these cells.

Under the same experimental conditions, the used ZnO powder did not induce DNA damage and, thus ZnO powder was not genotoxic (clastogenic and/or mutagenic) in these cells.

(Reference: AR11)

Comment

For the ZnO used in this study, information on number of particles per mass was not provided. Information on dose expressed as number of particles was not provided.

S9-mix fraction was not mentioned and presumably not used. However, the usefulness of S9-mix fraction in assays investigating ZnO nanoparticles may be doubtful as most ZnO nanoparticles will not be metabolised.

The SCCS agrees with the conclusion that the used ZnO nanoparticles were genotoxic in the comet assay with primary human nasal mucosal cells.

Comet assay in human nasal mucosal cells

Guideline/method:	According to published protocol
Species/strain:	Human nasal mucosal cells.
Group size:	Ten patients (seven female and three male).
Test substance:	ZnO nanoparticles (<100 nm, surface area 15–25 m²/g) were obtained from Sigma–Aldrich (Steinheim, Germany).
Batch:	Not reported
Purity:	Not reported
Dose levels:	500 μL of diluted nanoparticle or powder suspension at end concentrations of 0.1, and 5 μg/mL in the well, corrersponding to 1.5–-2.5, and 75–-125 mm²/mL in the well.
Route:	In vitro
Vehicle:	Distilled water
Exposure:	Repeated 3 consecutive exposures of 1 h with a washing step in
Observation period:	between
Observation period:
GLP:	Not in compliance
Published:	Yes
Study period:	27 hours
Date of report:	2011

To perform repetitive exposures, seven day old mini organ cultures (MOCs) (three-dimensional cultures of human nasal mucosa) were incubated with 0.1 and 5 μg/mL of ZnO NP suspension, without S9-mix, bronchial epithelium growth medium (BEGM), or methyl methanesulfonate (MMS) for 1 hour followed by a washing step with fresh BEGM. After the initial exposure step, the first cyto- and genotoxicity assays were performed in four MOCs for each concentration (control, 0.1 μg/mL, 5 μg/mL, MMS). The other 48 MOCs were again exposed to ZnO nanoparticles, BEGM, or MMS for another hour. This procedure was repeated one more time. Following the third exposure, the remaining 16 MOCs were further incubated in fresh BEGM for a 24 hour regeneration period. The experiment was performed on cells after 1 h exposure, after two consecutive 1 h exposures, after three consecutive 1 h exposures, and after three consecutive 1 h exposures followed by a regeneration period of 24 h.

Cells were prepared for TEM for ZnO particle uptake by the cells. Cell viability was determined by the trypan blue exclusion test.

After a 20-min DNA "unwinding’’period, the electrophoresis was performed under standard conditions (25 V, 300 mA, distance between electrodes 30 cm) for 20 minutes. The following parameters were analyzed to quantify the induced DNA fragmentation: tail DNA (TD), tail length (TL), and Olive tail moment (OTM) as a product of the median migration distance and the percentage of DNA in the tail. The OTM was applied for statistical analysis.

Results

ZnO nanoparticles had an oval shape with a mean longitudinal diameter of 76 ± 41 nm (mean ± SE) and a mean lateral diameter of 53 ± 22 nm (mean ± SE). After the exposure period it was observed that the ZnO nanoparticles were distributed to the cytoplasm and the nucleus of the cells in some of the cultures, but this was more pronounced for the 5 μg/mL exposure culture. There was no significant enhancement of DNA migration as determined by the Comet assay in nasal mucosa cells from MOCs exposed to 0.1 μg/mL of ZnO nanoparticles. At a ZnO nanoparticle concentration of 5 μg/mL, a significant increase in OTM was observed after each of one, two, and three consecutive 1 hour exposure periods compared to the control when measured directly after exposure. At both concentrations of ZnO nanoparticles, DNA fragmentation increased when measured at a 24 hours regeneration period after 3 repeated 1 hour exposures. A dose dependent increase in DNA fragmentation was present.

Conclusion

At both highest concentrations (5 μg/mL and 10 μg/mL) of ZnO-nanoparticles, DNA fragmentation increased after 24 hr of regeneration. In contrast, DNA damage which was induced by the positive control, methyl methanesulfonate, was significantly reduced after 24-hr regeneration.The authors concluded that the results suggest that repetitive exposure to low concentrations of ZnO nanoparticles results in persistent or ongoing DNA damage.

(Reference: AR12)

Comment

For the ZnO used, information on number of particles per mass was not provided. Information on dose expressed as number of particles was not provided.

Although it was suggested that a repeated exposure was used, the exposure was actually either 1, 2 or 3 hours with only a minimal non-exposed period with a washing step between the exposures.

The SCCS agrees with the conclusion that in this in vitro study, the ZnO exposure resulted in DNA damage. There fore, the ZnO was genotoxic in the used comet assay with human mucosal cells.

3.3.6.2 Mutagenicity/genotoxicity in vivo

Mouse Micronucleus Assay

Guideline/method:	OECD Guideline 474 (1997), EC Regulation 440/2008
Species/strain:	Mouse/NMRI
Group size:	Five male mice per dose group/sacrifice time.
Test substance:	ZnO (Z-COTE® HP1) (ZnO coated with triethoxycaprylylsilane).
Batch:	CNFC0701
Dose levels:	0, 15, 30, 60 mg/kg bw

Application volume: 10 mL/kg body weight.

Exposure: Single application

Route: Intraperitoneal (ip) injection

Vehicle: Fetal calf serum (FCS).

Sacrifice Times: Micronucleus test: 24 hours post-dose (vehicle, positive controls, low, mid and high) and 48 hours post-dose (vehicle and high dose only).

Positive controls: Cyclophosphamide (CPP): 20 mg/kg bw, vincristine sulfate (VCR):

0.15 mg/kg bw.

GLP: Yes

Date of report: August 2009

The potential of the test substance to cause chromosomal damage or spindle poison effects in vivo was investigated in the mouse bone marrow micronucleus test. Based on the results of a dose range-finding study, where mortality was observed at and above 80 mg/kg bw and where 60 mg/kg bw led to distinct clinical signs with no differences between sexes, each of the five male NMRI mice per dose group received a single intraperitoneal injection of the test substance suspension at 15, 30 or 60 mg/kg bw. Five male mice were used as controls receiving the vehicle (fetal calf serum), while each of the five positive control mice received 20 mg/kg bw cyclophospamide/kg bw or 0.15 mg/kg bw vincristine sulfate. In each case, the application volume was 10 mL/kg bw. All animals were observed for clinical signs of intoxication at regular intervals throughout the study period.

Bone marrow cells were harvested for evaluation of micronuclei at 24 hours post-dose (vehicle, positive controls, low, mid and high) and 48 hours post-dose (vehicle and high dose only).

Bone marrow for micronuclei examination was prepared, spread on a glass slide, fixed and stained with eosin and methylene blue, rinsed followed by staining with Giemsa solution and fixed in Corbit Balsam. Using a light microscope, the slides were evaluated. Two-thousand polychromatic erythrocytes (PCE) per animal were scored for the presence of micronuclei (10,000 per treatment group). The number of normochromatic erythrocytes (NCE) was also scored. The results obtained for the test groups, the negative and positive controls were compared to historical control data for validation purpose.

Results

The formulation analysis showed the correctness of the injected dose level and stability in the vehicle. The single intraperitoneal injection of the test substance preparations led to a dose-dependent increase in distinct clinical signs, while the animals of the negative and positive control groups revealed no findings. There were no statistically significances or biologically relevant differences in the number of erythrocytes containing micronuclei either between the vehicle control groups and the three dose groups, or between the two sacrifice intervals (24 and 48 hours). The number of normochromatic or polychromatic erythrocytes containing small micronuclei or large micronuclei did not deviate from the vehicle control values at any of the sacrifice intervals and was within the historical vehicle control data range. Both of the positive control substances, cyclophosphamide and vincristine sulfate, induced a statistically significant increase in the number of PCEs containing small and/or large micronuclei within the range of (or above) the historical positive control data indicating the suitability and sensitivity of the test system.

Conclusion

The test substance did not induce an increase in the number of polychromatic erythrocytes with micronuclei in the bone marrow of intraperitoneally treated mice and did not lead to any impairment of chromosome distribution in the course of mitosis. Thus, there was no indication for a clastogenic potential or any aneugenic activity in vivo under the conditions of the study.

(Reference: 26)

Comment

For the ZnO used in this study, information on surface area and number of particles per mass was not provided. Information on dose expressed as surface area and number of particles was not provided.

Intraperitoneal administration is proposed in OECD Guideline 474 but the text also states,

"If there is evidence that the test substance, or a reactive metabolite, will not reach the target tissue, it is not appropriate to use this test”. For intraperitoneal ZnO, exposure of bone marrow to the nanoparticles has not been demonstrated, although it can be assumed that Zn ions released from the ZnO nanoparticles may have resulted in exposure of the bone marrow. The value of this test is thus limited.

The SCCS disagrees with the conclusion of the authors. This assay does not prove that ZnO nanoparticles lack clastogenic or aneugenic activity in vivo as exposure of the target organs was not demonstrated.

Carcinogenicity

No data available.

Reproductive toxicity

No data specific for ZnO nanoparticles have been submitted or identified from the open literature.

Toxicokinetics

Additional studies in the open literature

No specific data for ZnO nanoparticles have been submitted. However, an exploratory study was recently published indicating systemic availability of Zn after ZnO nanoparticles were administered orally and intraperitoneally.

ZnO nanoparticles were approximately 50 nm in size (TEM evaluation), which were compared to ZnO microparticles showing at least one diameter >100 nm (TEM evaluation). DLS evaluation for the ZnO microparticles showedi an average hydrodynamic diameter of 1,226 ± 120 nm, and for the ZnO nanoparticles an average hydrodynamic diameter of 93 ± 14 nm was reported. Information on surface area and number of particles per mass was not provided.

After oral and intraperitoneal administration of a single dose of 2.5 g/kg bw for both ZnO nanoparticles and microparticles, Zn could be observed in serum indicating uptake from the GI–tract. For ZnO nanoparticles the systemic availability was somewhat higher compared to that of ZnO microparticles as indicated by Zn measurements by ICP-MS. A higher distribution of Zn in the liver, spleen and lung was shown after treatment with ZnO NP compared to treatment with ZnO microparticles. Serum liver enzyme analysis (aspartate aminotransferase (AST), alanine aminotransferase (ALT), lactate dehydrogenase (LDH)) indicated liver toxicity due to both ZnO NP and ZnO MP treatments which was confirmed by

histopathology. Histopathological lesions were only observed for ZnO NP treatments in the liver.

(Reference: AR16)

Comment

The results indicate a similar uptake, tissue distribution, and toxicity for ZnO nanoparticles and ZnO microparticles. As Zn was measured by ICP-MS, it cannot be concluded whether it was present as particulate material or as dissolved Zn ions. Although serum levels of aspartate aminotransferase, alanine aminotransferase, and lactyate dehydrogenase indicated similar liver toxicity for both ZnO nanoparticles and ZnO microparticles, histopathological lesions were only observed for ZnO nanoparticles in the liver.

Photo-induced toxicity

Phototoxicity/photoirritation and photosensitisation

Since submission I (2001) for ZnO, which served as a basis for the former opinion (SCCNFP opinion dated 24-25 June 2003, SCCNFP/0649/03, Reference: 87) and submission II (2005), no further studies have been performed or provided by the sponsors.

The studies presented in opinion SCCNFP/0649/03 (Reference: 87) did not indicate a photoirritant or photosensitizing potential of micronized ZnO (not further specified).

Phototoxicity/photomutagenicity/photoclastogenicity

Photomutagenicity in a Salmonella typhimurium Reverse Mutation Assay

Study Design
Date of report:	August 2006
Guideline/method:	Considerations on photochemical genotoxicity: Report of the International Workshop on Genotoxicity Test Procedures Working Group, 2000; Japanese Ministry of Health and Welfare: Guidelines for genotoxicity studies of drugs, 1999.
Species/strain:	Salmonella typhimurium TA1537, TA98, TA100, TA102.
Replicates:	Triplicate plates
Test substance:	FINEX-50 ZnO (supplied by Sakai Chemical Industry Co., primary particle size: 20 nm, non-coated).
Batch:	OZ52 (purity: ≥96% as indicated in the submission)
Concentrations:	Dose-ranging test: 5, 15, 50, 150, 500, 1,500, 5,000 μg/plate (with/without light irradiation).
Main test:	313, 625, 1,250, 2,500, 5,000 μg/plate (with/without light irradiation).
Vehicle:	DMSO
Irradiation:
Source of light:	Sunlight simulator (SOL 500) equipped with a metal halide lamp with emission of a continuous spectrum of simulated sunlight (50% at 335 nm wavelength).
Intensity of irradiation:	1.6–1.7 mW/cm²

Bacterial Bacterial Strains

TA 1537 (8 minutes 20 seconds) TA 98 (8 minutes 20 seconds) TA 100 (8 minutes 20 seconds)

UVA dose (mJ/cm²) 0.80–0.85 0.80–0.85 0.80–0.85

	TA 102 (50 minutes)	4.8–5.1
Positive controls:	Without irradiation	TA 100: AF-2, 0.01 μg/plate
		TA 1537: 9AA, 80 μg/plate
		TA 98: AF-2, 0.1 μg/plate
		TA 102 MMC, 0.05 μg/plate
	With irradiation:	TA 98, TA 100, TA 1537: CPZ, 1.0 μg/plate
		TA 102: 8-MOP, 0.05 μg/plate
GLP:	No
Published:	No

The test substance was investigated for its potential to induce gene mutations under irradiation with simulated sunlight according to considerations on photochemical genotoxicity using the Salmonella typhimurium strains TA 1537, TA 98, TA 100, and TA 102. The tests were performed in the absence of S9 mix. Each concentration, including the controls, was tested in triplicate and at concentrations of 5, 15, 50, 150, 500, 1,500, 5,000 μg/plate in the dose range-finding part of the study and concentrations of 313, 625, 1,250, 2,500, 5,000 μ g/plate in the main study in the absence of metabolic activation (S9 mix), under the light irradiation or without light irradiation. The test substance was dissolved in DMSO. Petri dishes containing the test substance suspension, test bacterial suspension and phosphate buffered saline were either irradiated with a sunlight simulator (SOL 500) or remained not irradiated. The irradiation time was the treatment time. After the treatments, aliquots of the mixtures were dispensed into test tubes, mixed with top agar, and spread on the minimum glucose agar plates. The plates were incubated at 37°C for 48 hours and revertant colonies on the plates were counted.

Results

The dose-range finding study was conducted at 7 dose levels, i.e., 5, 15, 50, 150, 500, 1500 and 5000μg ZnO/plate. Growth inhibition was not observed in each tester strain at any doses in both the light irradiated group and the light-unirradiated group. The plates incubated with the test item showed normal background growth up to 5,000 μ g ZnO/plate in all strains used. In the main test no substantial increase in revertant colony numbers of any of the four tester strains was observed following treatment with the test substance under irradiation with simulated sunlight at any dose level. Also in the absence of irradiation no effects were found. In both the dose–range finding study and the main test, the precipitation of FINEX-50 zinc oxide was seen on the plate at 5000 and 2500 μg/plate in both the light-irradiated group and the light-unirradiated group

The sensitivity and validity of the test system used was demonstrated by the expected induction of a significantly increased number of revertants with the appropriate positive controls.

Conclusion

The test substance did not induce gene mutations both in the presence and absence of irradiation. It was therefore considered to be non-photomutagenic in this Salmonella typhimurium photomutagenicity test.

(Reference: 98)

Comment

For the ZnO used in this study, information on surface area and number of particles per mass was not provided. Information on dose expressed as surface area and number of particles was not provided.

S9-mix fraction was not used. However, the usefulness of S9-mix fraction in assays investigating ZnO nanoparticles may be doubtful as ZnO nanoparticles will not be metabolised.

Contact of bacterial DNA (i.e. nanoparticle uptake by bacteria) with the ZnO nanomaterials was not demonstrated. This test has been considered inappropriate for nanoparticles because bacteria do not take up nanoparticles due to the lack of endocytosis (Reference: AR4). However, Released Zn ions might be available as suggested for in vitro exploratory Comet assay.

The SCCS considers that the results can be considered negative in the used photomutagenicity bacterial gene mutation assay. However, as there is no certainty on the exposure of the bacterial DNA to the added ZnO nanoparticles, this negative result is of a limited value.

Photo-chromosomal aberration test with cultured mammalian cells

Study Design

Date of report: August 2006

Guideline/method:
Considerations on photochemical genotoxicity:
Report of the International Workshop on Genotoxicity Test Procedures Working Group, 2000; Japanese Ministry of Health and Welfare: Guidelines for genotoxicity studies of drugs, 1999

Species/strain: Chinese Hamster lung fibroblasts (CHL cells).

Replicates: Duplicate experiments.

Test substance: FINEX-50 ZnO (supplied by Sakai Chemical Industry Co., primary particle size: 20 nm, non-coated).

Batch: OZ52 (purity: ≥96% as indicated in the submission)

Concentrations: Growth inhibition test: 9.77, 19.53, 39.06, 78.13, 156.25, 312.5, 625, 1,250, 2,500, 5,000 μg/mL (with/without light irradiation).

Chromosomal aberration test: 2.44, 4.88, 9.77, 19.53 μg/mL (with/without light irradiation).

Vehicle: 0.5% aqueous carboxymethyl-cellulose solution.

Irradiation:

Source of light:	Sunlight simulator (SOL 500) equipped with a metal halide lamp with emission of a continuous spectrum of simulated sunlight (50% at 335 nm wavelength).
Intensity of irradiation:	1.6–1.7 mW/cm²
UV doses:	4.8–5.1 J/cm²
Preincubation time:	60 minutes
Irradiation time:	50 minutes
Recovery:	After irradiation reaction solutions were removed and the cells were further cultured for another 22 hours.
Negative control:	0.5% aqueous carboxymethyl-cellulose solution.
Positive controls:	With irradiation: 8-Methoxypsoralene (8-MOP), 0.01 μg/mL.
	Without irradiation: MNNG, 2 μg/mL.
GLP:	No
Published:	No

The cytotoxicity of the ZnO test substance was evaluated by measuring cell density. The cell density was determined as a percentage to the untreated control value (100%) by measuring the degree of staining on each dish with a single-layer culture cell densimeter. On the basis of the measured values, an approximate 50% inhibitory concentration (IC50) of the test substance was estimated. This assay was conducted in duplicate.

The test substance was investigated for its potential to induce structural chromosomal aberrations in Chinese Hamster lung fibroblasts (CHL cells) in the absence and the presence of simulated sunlight in two independent experiments. A sunlight simulator (SOL 500) equipped with a metal halide lamp with emission of a continuous spectrum of simulated sunlight (50% at 335 nm wavelength) was used as a light source. The cultures were pre-incubated with the test item for 60 minutes. After pre-incubation, the cultures were exposed to the solar simulator at an irradiation intensity of 1.6–1.7 mW/cm² resulting in a UVA dose of 4.8–5.1 J/cm². Thereafter, the cultures were washed. Corresponding cultures with the test item were kept in the dark for the 50 minute exposure period. The chromosomes were prepared 22 hours after washing. Two parallel cultures were investigated and at least 100 metaphase per plate were scored for structural chromosome aberrations in each culture.

Results

Cell growth inhibition test

The cell growth ratio was affected by precipitation of the test substance at 312.5 μg/mL and above in the cultures without irradiation, while no precipitation was observed at 156.25 μg/mL and below. The cell growth ratios by test substance were reduced to 31% at 19.5 μg/mL and 77% at 9.8 μ g/mL. In the irradiated cultures, the cell growth ratio was affected by precipitation at 312.5 μg/mL and above, while no precipitation was observed at 156.25 μg/mL and below. Thus, concentrations of 19.5 μg/mL (as the maximum concentration), followed by serial 2-fold dilution to give 2.4, 4.9, 9.8 and 19.5 μg/mL were selected for the aberration test in the presence or absence of irradiation.

Chromosomal aberration test

The non-irradiated cultures revealed structural aberrations at a concentration of 19.5 μg/mL and a ratio of 13% between the non-irradiated and the irradiated groups was calculated, which was statistically significant positive. The ratios of numerical aberrations were not

statistically significant and increased at any dose level. With irradiation, the ratios of structural aberrations at 19.5 and 9.8 μg/mL were 17 and 14%, respectively, and attained statistical significance. The ratios of numerical aberrations were not affected at any dose.

In order to clarify whether the test substance was photo-clastogenic or not, the difference in the ratios of aberrations between the irradiated cultures and those which were not irradiated was analyzed, showing a significant increase at 9.8 μg/mL. The sensitivity of the system was demonstrated since the positive controls induced statistically significant increases in cells showing structural chromosome aberrations.

Conclusion

Under the conditions of the study, the test substance induced structural chromosome aberrations in the absence or presence of simulated sunlight as determined by the photo-chromosomal aberration test in Chinese Hamster lung fibroblasts. According to the authors, the test substance was also shown to be photo-clastogenic when tested up to cytotoxic concentrations. However, the study with non-, pre- or simultaneously irradiated mammalian cells (Reference: 37 below) showed that treatment with UV light resulted in an increased susceptibility to ZnO clastogenicity and thus, is not photoclastogenic per se, according to the applicant.

(Reference: 97)

Comment

For the ZnO used in this study, information on surface area and number of particles per mass was not provided. Information on dose expressed as surface area and number of particles was not provided.

The study showed that test substance induced structural chromosome aberrations in the absence or presence of simulated sunlight. Without irradiation, the highest dose investigated (19.5 μg/mL) was positive in the test, while after irradiation a dose dependent positive response was seen for the two highest doses investigated (9.8 μ g/mL and 19.5 μg/mL). Even if treatment with UV light increased the sensitivity of the cells, it is the presence of the ZnO nanoparticles that induced the chromosomal aberrations, and this cannot be ignored.

The SCCS considers that ZnO induced chromosomal aberrations in Chinese hamster lung fibroblasts both in the absence and presence of light irradiation.

Photo-chromosomal aberration test with non-, pre- or simultaneous irradiated mammalian cells

Study Design

Date of report: January 2007

Guideline/method: Requirements of the EMEA Committee for Proprietary Medicinal Products, Note for Guidance on Photosafety Testing, 27 June 2002, Considerations on Photochemical Genotoxicity: Report of the

	International Workshop on Genotoxicity Test Procedures Working Group. (Gocke et al. 2000).
Species/strain:	Chinese Hamster ovary (CHO) cells.
Replicates:	Duplicate cultures, two flasks (test substance, positive controls), four flasks (negative control).
Test substance:	ZnO, uncoated (Z-COTE®), size less than 200 nm according to materials and methods section (100 nm is mentioned in the abstract).
Batch:	EHDE0402 (purity: >99% (i.e. 100.1 g/100 g)).
Concentrations:	Phototoxicity range finder: 8.201, 13.67, 22.78, 37.97, 63.28, 105.5, 175.8, 293.0, 488.3, 813.8 μg/mL (with simultaneous and pre-imposed UV irradiation).
Chromosomal aberration test:	27.49, 34.36, 42.95, 53.69, 67.141, 83.89, 104.9, 131.1, 163.8, 204.8, 256.0, 320.0, 400, 500 μg/mL (non-irradiated, with simultaneous and pre-imposed UV irradiation).
Vehicle:	McCoys 5A medium
Irradiation:
Source of light:	Atlas Suntest CPS+ solar simulator (Heraeus Equipment Limited, Brentwood, UK). The intensity of UVA and UVB was measured using a Dr Gröbel RM 21 UV meter. The ratio of UVB:UVA was in the range of approximately 1:30.
UV doses:	Phototoxicity range finder: UVA: 400 and 800 mJ/cm².
Chromosomal aberration test:	UVA: 350 and 700 mJ/cm².
Groups:
Non-irradiated:	Without irradiation
Pre-irradiated (PI):	Irradiation 2–3 hours (350 mJ/cm²) or 1–2 hours (700 mJ/cm²) prior to treatment with ZnO concentration.
Simultaneous irradiated (SI):	350 mJ/cm² and 700 mJ/cm² cultures received simultaneous treatment with ZnO concentration.
Incubation time:	Cells were exposed to ZnO for 3 hours, after washing, cells were harvested after a further incubation for 17 hours in tissue culture medium.
Negative control:	McCoys 5A medium.
Positive controls:	8-Methoxypsoralene (8-MOP): 0.5 and 1.0 μg/mL (with/without irradiation).
	4-Nitroquinoline-1-oxide (NQO): 0.25 and 0.3 μg/mL (without irradiation).
GLP:	Yes
Published:	Yes

ZnO was tested in an in vitro cytogenetic assay using duplicate cultures of Chinese Hamster Ovary (CHO) cells in the presence and absence of UV light to clarify whether the slightly pronounced clastogenicity in vitro in the dark is a genuine photo-genotoxic effect.. The effect of ZnO exposure was investigated in the cells in the dark (D), under pre-imposed irradiation (PI, i.e. UV irradiation of cells followed by treatment with ZnO 1-3 hours later) and under simultaneous irradiation conditions (SI, i.e. ZnO treatment concurrent with UV irradiation).

Concentrations used in the main experiment were selected on the basis of cytotoxicity (expressed as a decrease in population doublings relative to controls). The number of cells/mL were measured in trypsinised samples of cell suspension using a Coulter counter. The highest concentration of ZnO used in the range-finder was 813.8 μg/mL (equivalent to 10 mM) for zinc. After three hours of exposure to ZnO or control chemicals, cell monolayers from all cultures were washed with sterile saline, and re-fed with fresh McCoy’s 5A medium containing foetal calf serum and gentamycin, and incubated for 17 hours before cell harvesting. Approximately 1.5 h prior to harvest, colchicine was added to give a final concentration of approximately 1_g/mL to arrest dividing cells in metaphase.

At the defined sampling times, monolayers of these cultures were removed using trypsin/EDTA and a measurement of cell counts/mL was performed on an aliquot of cell suspension using a Coulter counter. The remaining cell suspensions from each flask were harvested and slides prepared for chromosome aberrations (CA) analysis using standard operating procedures.

Negative (solvent) control cultures were included in the test system under each treatment condition. 4-Nitroquinoline-1- oxide (NQO) and 8-methoxypsoralen (8-MOP) were employed as positive control chemicals in the absence and presence of UV light. The slides for CA analysis were prepared 17 hours after washing. Two parallel cultures were investigated and at least 100 metaphase plates were scored for structural chromosome aberrations in each culture.

Results

In the absence of UV light, the cytotoxicity of ZnO was concentration dependent, with 40-60% cytotoxicity observed in the concentration range 256–320 μg/mL. In SI cultures, the cytotoxicity was more pronounced than in the dark and related to the UV dose: 40-60% cytotoxicity was observed in the concentration range 131.1–256 μg/mL (350 mJ/cm²) or 83.89–13 1.1 μ g/mL (700 mJ/cm²). Overall, cytotoxicity in SI cultures increased with ZnO concentrations, although a large variability in ZnO cytotoxicity was observed at the high UV dose (700 mJ/cm²) and at high ZnO concentrations (≥204.8 μg/mL).

In PI cultures, cytotoxicity of ZnO was concentration dependent and comparable at the low UV dose (350 mJ/cm²) to that observed in SI cultures (40-60% cytotoxicity was observed in the concentration range 104.9–256.0 μg/mL). At the high UV dose (700 mJ/cm²) and at low ZnO concentrations (<204.8 μg/mL), cytotoxic effects of ZnO in PI cultures were intermediate between those observed in the dark and SI cultures, with 40-60% cytotoxicity observed in the concentration range 131–256 μg/mL. At higher concentrations (≥204.8 μg/mL), ZnO cytotoxicity was similar in the PI and SI cultures. For each individual irradiation dose and condition, the concentrations analyzed for CA (chromosome aberrations) covered a range of cytotoxicity from little or none, to maximum effects.

Treatment of cultures with ZnO in the absence of UV light resulted in statistically significant increases in the frequencies of cells with structural aberrations at 104.9 μg/mL (giving 13% cytotoxicity, as measured by population doubling) and above, in a concentration dependent manner. The frequencies of cells with structural aberrations (excluding gaps) exceeded the historical negative control (normal) range in both cultures analyzed at 163.8 μg/mL and above, and also in single cultures analyzed at several lower concentrations. These observations were considered biologically relevant and ZnO was considered clastogenic in the dark. Under SI conditions, treatment with ZnO induced biologically relevant increases in structural aberrations at 104.9 μg/mL (23% cytotoxicity) and 53.69 μg/mL (19% cytotoxicity) following 350 and 700 mJ/cm² UV radiation.

Under PI conditions, treatment with ZnO resulted in biologically relevant increases in the frequencies of cells with structural aberrations at 104.9 μg/mL and above (36% cytotoxicity) and at 53.69 μg/mL and above (0% cytotoxicity) following 350 and 700 mJ/cm² UV radiation, respectively. When compared at similar cytotoxic concentrations, the incidence of chromosome aberrations following PI or SI was generally similar at 700 mJ/cm².

The proportion of cells with structural aberrations treated with the vehicle (negative control) fell within historical solvent control ranges. Both treatments with the positive controls induced increases in the proportion of cells with structural aberrations. When added to cultures treated in the absence of UV light, 8-MOP induced frequencies of cells with structural aberrations that were similar to those seen in concurrent solvent control cultures (non irradiated). Thus, the validity and sensitivity of the test system was demonstrated.

ZnO induced clastogenic activity in CHO cells, both in the absence and in the presence of UV light. The cytotoxicity of ZnO to CHO cells under the different irradiation conditions was as follows: SI > PI > D. The effects were more pronounced at 700 mJ/cm². The incidence of chromosome aberrations in SI or PI cells was generally higher than in the dark. At similar ZnO concentrations, SI conditions generally produced the greatest incidence of chromosome aberrations. However, at similar cytotoxic concentrations, the incidence of chromosome aberrations following PI or SI was generally similar for 700 mJ/cm².

Conclusion

The authors concluded that ZnO was clastogenic to CHO cells in the dark and after irradiation of the cellular test system, either prior to (PI conditions) or concurrently with (SI conditions). The treatment with UV light resulted in an increased susceptibility of CHO cells to ZnO clastogenicity, as indicated by clastogenic responses of higher magnitude and/or observed at lower concentrations in both SI and PI conditions when compared to those obtained in the dark.

Finally, the results provided evidence that, under conditions of in vitro photo-clastogenicity tests, UV irradiation of the cellular test system per se may produce a slight increase in the genotoxic potency of compounds that are clastogenic in the dark. Therefore, minor increases in clastogenic potency under conditions of photo-genotoxicity testing do not necessarily represent a photogenotoxic effect, but may occur due to an increased sensitivity of the test system subsequent to UV irradiation.

(References: 37, 67)

Comment

For the ZnO used in this study, information on surface area and number of particles per mass was not provided. Information on dose expressed as surface area and number of particles was not provided. The uptake of the ZnO by the exposed cells was not demonstrated.

An increase in the frequency of cells with structural chromosomal aberrations was noted after UV irradiation both in simultaneous irradiation and in pre-irradiation culture conditions. In addition, clastogenic activity was also demonstrated without UV treatment. The observed clastogenic effect during irradiation may not be a photo-genotoxic effect of the ZnO as also in the pre-irradiated cultures (in which ZnO was not irradiated) a clastogenic effect was seen.

The SCCS consideres that based on the results presented ZnO has clastogenic activity in Chinese hamster ovary cells both in the presence and absence of UV.

Photo-micronucleus test in mice

Study Design
Date of report:	August 2008
Guideline/method:	Photo-micronucleus test in mice according to published methodology (Reference: AR13).
Species/strain:	Hairless mouse/HR-1
Group size:	Three male mice/group.
Test substance:	FINEX-50 ZnO (supplied by Sakai Chemical Industry Co., primary particle size: 20 nm, non-coated)
Batch:	OZ52 (purity: ≥96% as indicated in the submission)
Dose levels:	Preliminary toxicity study: 2.5%, 5%, 10%, 20% (w/v) concentration of the test sample in Aceton Olive Oil (AOO) (with irradiation).
Micronucleus test:	5%, 10%, 20% (w/v) concentration (with/without irradiation).
Vehicle:	Acetone/olive oil (4:1)
Exposure:	Once daily for two consecutive days.
Route:	Topical on the skin (open without occlusion of the test samples)
Application volume:	0.2 ml
Application area:	3 × 4 cm
Irradiation:

Source of light:	Sunlight simulator (light emission equipment for animals (LDS-20, ABLE) with six xenon lamps) with emission of a continuous spectrum of simulated sunlight between 300–800 nm, visible light (48,800–67,600 lux).
Intensity of irradiation:	UVA: 2.27–2.59 mW/cm²
	UVB: 88.6–17.2 μW/cm²
Irradiation time:	2 hours, once a day for two days at an interval of 24 hours.
Negative control:	Acetone/olive oil (4:1, with/without irradiation).
Positive controls:	With irradiation: 8-Methoxypsoralene (8-MOP), 0.0015%.
	Without irradiation: Mitomycin C (MMC), 0.05%.
GLP:	No

The ability of the test substance to cause photo-chromosomal damage in vivo was investigated in the photo-micronucleus test with mouse epidermal cells after topical application on the skin of male hairless HR-1 mice.

After a preliminary toxicity test with light irradiation to observe skin effects and to determine the maximum tolerated dose of the test article, the micronucleus test (with/without light irradiation) was carried out. In the preliminary toxicity test, concentrations of 2.5, 5, 10 and 20% (w/v) of the test article suspended in acetone/olive oil (4:1) were used.

The test article suspensions were applied on the dorsal skin of mice (3–4 cm) once a day for two days at an interval of 24 hours followed by light irradiation using a sunlight simulator (Light Emission Equipment for Animals, LDS-20: with six xenon lamps, ABLE), within 30 minutes after each application. As neither abnormal signs nor dead animals were observed in the dosing groups, concentrations of 5%, 10% and 20% were selected for the micronucleus test.

The micronucleus test was carried out with ten experimental groups of mice, five with and five without light irradiation: the negative control groups [vehicle (acetone/olive oil, 4:1)], the test article groups [5, 10 and 20 % (w/v)] and the positive control groups (with light irradiation: 8-methoxypsoralen; without light irradiation: Mitomycin C).

The dosing formulation (0.2 mL) was applied on the dorsal skin of mice once a day for two days at an interval of 24 hours. For the light-irradiated group, the mice were exposed to light from the sunlight simulator within 30 minutes after each application for 2 hours (UVA intensity: 2.27–2.59 mW/cm², UVB intensity: 88.6–117.2 μW/cm²). For the non-light irradiated group, the mice were returned to their home cages immediately after application and kept in the animal room under white fluorescent light.

All animals were euthanized about 48 hours after the last treatment. The skin was processed and epidermal cell suspensions were prepared. The cell suspensions were placed on glass slides, dried and stained with acridine orange solution. Each of the 2,000 epidermal cells were examined with a fluorescence microscope and the number of micronucleated epidermal cells was recorded.

Results

In the preliminary toxicity test, after the topical application on the skin, abnormal behavioral signs or dead animals as indication for toxicity were not observed, so the test sample did not induce overt toxicity in the used application.

The fluorescence microscopic observation revealed no test substance (ZnO) related increase in the number of micronucleated epidermal cells outside the historical control data, neither with nor without UVA and UVB light irradiation.

In the vehicle group the incidence of micronucleated cells did not exceed the historical control group ranges, while the respective positive controls induced a significant increase in the number of micronucleated epidermal cells with and without simulated light irradiation indicating the suitability and sensitivity of the test system.

Conclusion

The test substance did not increase the incidence of micronucleated epidermal cells in exposed male mice with or without simulated light irradiation. From these results, it was concluded that FINEX-50 ZnO Batch OZ52 (purity: ≥96%) is not clastogenic or photo-clastogenic in epidermal cells of mice under these experimental conditions.

(Reference: 100)

Comment

For the ZnO used in this study, information on surface area and number of particles per mass was not provided. Information on dose expressed as surface area and number of particles was not provided.

The purity of the FINEX-50 used was indicated in the submission to be ≥96%. However, data on the purity is not provided in the physico-chemical characterization of the FINEX-50 (section 3.1.4). The submitted study report in reference 100 indicates a content of 100%.

A different product/batch is mentioned in the reference to that included in the dossier. Product/batch 05M099 is mentioned in the results and conclusions, but not in materials and methods. Product/batch 05M99 (lot no. 2Z51, content 100%) is mentioned in Reference: 100.

The ZnO was applied topically on the skin. Uptake by living cells in the skin was not investigated. As there is no information whether the substance reaches the target cells (or the nucleus), the value of this test is very limited.

The results show that the ZnO tested did not increase the number of micronuclei in the skin after topical application indicating that the ZnO is not a positive mutagenic substance. However, from the negative results it cannot be concluded that ZnO is non-mutagenic.

The SCCS considers the value of this test limited.

Human data

Comparative human inhalation study

Study Design

Date of publication: February 2005

Guideline/method: According to internal laboratory methodology.

Species: Human

Group size: 12 volunteers (six males and six females, one male/female in each exposure sequence).

Test substance:
A) Fine ZnO particles (A) ZnO, coated (Zano 10 Plus, Umicore) (former Zano® Plus, ZnO coated with octyl triethoxysilane).
B) Ultrafine ZnO particles.

ZnO particles were generated by an electric arc discharge system between two consumable zinc electrodes.

Particles sizes: For the ultrafine particle exposures, the count median diameter was 40.4 ± 2.7 nm geometric standard deviation (GSD) 1.7, whereas for the fine particle exposures, the count median diameter was 291.2 ± 20.2 nm GSD 1.7.

Batch: Not applicable

Route: Inhalation (by mouth piece)

Frequency: 2 hour inhalation at rest on three exposure days.

Concentration: 500 μg/m³ fine and ultrafine ZnO particles, or 4.6×10¹⁰6x10¹⁰/m³ and 1.9×10⁸9x10⁸/m³ particles, respectively, as measured during exposure.

Examinations: Clinical symptoms, heart rate, blood pressure, mouth temperature, oxygen saturation at rest, differential blood cell counts, expression of activation markers and adhesion molecules, coagulation factors, inflammation markers, electrocardiogram (ECG) parameters, sputum induction.

GLP: No

Published: Yes

A comparative exploratory inhalation study was performed in 12 human volunteers (six male and six female, age range 23–52 years) with ultrafine and fine ZnO particles generated by an electric arc discharge system between two consumable zinc electrodes. The exposure response relationships for respiratory, hematologic, and cardiovascular endpoints between ultrafine and accumulation mode zinc oxide particles were compared. Twelve healthy adults inhaled 500 μ g/m³ of ultrafine zinc oxide, the same mass of fine zinc oxide, and filtered air while at rest for 2 hours. Pre-exposure and follow-up studies of symptoms, leukocyte surface markers, homeostasis, and cardiac electrophysiology were conducted up to 24 hours post-exposure. Induced sputum was sampled 24 hours after exposure.

Conclusion

Freshly generated zinc oxide in the fine or ultrafine fractions inhaled by healthy subjects at rest at a concentration of 500 μg/m³ for 2 hours is below the threshold for acute systemic effects as detected by these endpoints.

(Reference: 32)

Comment

For the ZnO used in this study, information on surface area and number of particles per mass was not provided. Information on dose expressed as surface area was not provided.

The ZnO materials given as test substances in the dossier (A and B in the table) were not used in the study.

Reference 32 only describes the effect of ZnO generated at the research facility itself. Ultrafine and fine ZnO were generated by an electric arc discharge system (Palas generator; Palas, Karlsruhe, Germany) between two consumable zinc electrodes. The ZnO materials used in the reference do not relate to any of the ZnO materials submitted and included in the dossier by the applicant.

In view of the different type of ZnO used (spark generated) in the study, the SCCS considers this study not relevant for the evaluation of the use of ZnO as cosmetic ingredient.

Special investigations

Experimental study on toxicity after IV administration

Study Design
Date of report:	December 2009
Guideline/method:	Study according to internal laboratory methodology for investigative purposes considering OECD Guideline 407, EC Commission Regulation No. 440/2008, EC Commission Directive 87/302/EEC, OECD Guideline 417
Species/strain:	Rat/Wistar (Crl:WI (Han))
Group size:	Groups of five male and five female rats.
Test substances:	a) ZnO nanoscale, coated with triethoxyoctylsilane (Z-COTE®HP1)
	b) ZnO, nanoscale uncoated (Z-COTE®)
	c) ZnO pigment
	d) Zinc sulfate
Batches:	a) CNBG1602 (ZnO >97.1%, triethoxyoctylsilane 3.4%)
	b) EHFA0907
	c) S35583-206 (>99%)
	d) 02616TC-126 (100%)
Dose levels:	a) 1 or 5 mg/kg bw (subsets 0, 1, 2), single administration day 28
	b) 1 or 5 mg/kg bw (subsets 0, 1, 2)
	c) 5 mg/kg bw (subsets 0, 1, 2)
	d) 17.6 mg/kg bw (subset 2) or 4 × 4.4 mg/kg bw on days 0, 7, 14, 21 (subsets 0, 2) as an equimolar zinc ion concentration related to the dose level of 5 mg/kg bw of ZnO.
Exposure:	I) Single application for test item groups a, b and c (subsets 0, 1, 2) and test item group d (subset 1).
	II) Four applications on study days 0, 7, 14 and 21 for test item group d (subsets 0, 2).
Route:	Intravenous
Application volume:	1 mL/kg bw
Vehicle:	a–c) Inactivated calf serum
	d) Physiological (0.9%) NaCl solution
	Application and sampling time-points: subset 0 (five males/five females per group): application on study day 0, urine sampling on study days 24/25, blood sampling and necropsy on study day 29.
	Subset 1 (five males per group): application on study day 28, blood sampling and perfusion fixation on study day 29.
	Subset 2 (five males per group): application on study day 0, perfusion fixation on study day 29.

GLP: Yes

The study was divided into separate subsets according to the dosing and sampling schedule.

Subset 0: ZnO coated and uncoated, and pigment ZnO dosing at day 0 and sampling at days 24/25 (urine) and day 29 (blood). As control Zn sulfate was used that was administered at days 0–7–14–21 and sampled on the same days.
Subset 1: ZnO coated and uncoated, pigment ZnO and Zn sulfate on day 28 and sampling at day 29 (day 1 after exposure).
Subset 2: Same as subset 0 treatment at day 0, but animals were sacrificed for histopathology after perfusion fixation at day 29.

The total dose of Zn sulphate (17.6 mg/kg bw) was administered in four dosages of 4.4 mg/kg bw at days 0, 7, 14, and 28.

Results

The analysis confirmed the homogenous distribution of the test items in the vehicles and the correctness of the concentrations.

Significantly increased body weights were determined on study day 28 for female animals in test group 2 (zinc oxide [nanoscale, coated], 5 mg/kg bw) and test group 3 (zinc oxide [nanoscale, uncoated], 1 mg/kg bw), as well as on study days 21 and 28 for female animals in test group 5 (zinc oxide [pigment], 5 mg/kg bw) compared to the control. For male animals in test group 26 (zinc sulfate, 4 × 4.4 mg/kg bw), a significantly lower mean body weight was detected on study day 21.

Significantly lower mean body weight changes were determined on study days 7, 14 and 21 for male animals in test group 14 (zinc oxide [nanoscale, uncoated], 5 mg/kg bw).

Significantly higher mean body weight change was determined on study day 28 for female animals in test group 2 (zinc oxide [nanoscale, coated], 5 mg/kg bw).

Significant effects on body weight were noted in various groups after ZnO treatment compared to the vehicle (calf serum inactivated, 1 ml/kg bw) control animals.

Test group 2, females, ZnO nanoparticles, coated, 5 mg/kg bw, single dose day 0, increased bw day 28.

Test group 3, females, ZnO nanoparticles, uncoated, 1 mg/kg bw, single dose day 0, increased bw day 28.

Test group 5, females, ZnO-pigment, 5 mg/kg bw), single dose day 0, increased bw days 21 and 28.

Test group 26, males, Zn-sulfate, dose 4 × 4.4 mg/kg bw, decreased bw day 21.

Test group 14, males, ZnO nanoparticles, uncoated, 5 mg/kg bw, decreased bw days 7, 14, and 21.

The values reflect the normal range of biological variation inherent in the strain of rats used for this study and, therefore, were assessed as incidental.

There was no mortality, no clinical findings, no effect on food consumption or body weight gain, no impaired organ weight and no gross pathological findings in the organs of any of the investigated in either group in the respective subsets, irrespective of whether the animals were killed after one day or after four weeks following the single (ZnO) or four (zinc sulfate) intravenous test item injections. In addition, the male and female animals of subset 0 (ZnO coated and uncoated, and pigment ZnO) showed no treatment related findings in the urinalysis and the comprehensive clinical pathology examinations including additional serum parameters (haptoglobulin, alpha2-macroglobulin, troponin I) at the end of the observation period of four weeks.

However, the animals of subset 1, which were killed and investigated by clinical pathology at day 1 after injection of the test items, revealed some minor deviations in single parameters of differential blood count, enzymes, clinical chemistry and the additional serum parameter haptoglobin. See table below (5.12.1.).

Conclusion

The comparative toxicity study of ZnO (coated or uncoated nanoparticles), ZnO (inndin pigment form) and zinc and sulfate, showed no persistent effects at the end of the observation period of four weeks when injected as a single dose intravenously at dose levels of 1 or 5 mg/kg bw.

Very mild, and in any case not persistent, deviations in single hematology parameters (monocyte, large unstained cells) and few clinical chemistry parameters, probably indicative of a minimal impairment of liver function (ZnO preparations) or of kidney function in addition (zinc sulfate) were observed one day after intravenous injection of 5 mg/kg bw pigmented or triethoxyoctylsilane coated nanoscale ZnO or an equimolar zinc sulfate dose. At 1 mg/kg bw no biologically relevant effects were observed. Among the known and very

sensitive markers of acute phase inflammatory reactions towards nanoparticles, only haptoglobulin was increased. Alpha2-macroglobulin and troponin were not affected.

Finally, the comparative screening showed only minor and in any case transient effects without toxicological relevance. There was no biologically relevant difference in the injected form of ZnO, and in particular, no enhancement of the observed effects due to the injection of nanoscale material. The equimolar injection of the zinc ion as zinc sulfate led qualitatively and quantitatively to a slight enhancement of the observed variation in homoeostasis. In any case, four weeks after treatment there was virtually no difference in comparison to the respective control group, irrespective of the injected dose level, the usage of ZnO as coated or uncoated nanoparticles, as ZnO pigment or as zinc sulfate.

(Reference: 28)

Comment

For the ZnO used in this study, information on surface area and number of particles per mass was not provided. Information on dose expressed as surface area and number of particles was not provided.

In general similar minimal effects were observed for both ZnO coated and uncoated nanoparticles. Some effects on body weight were observed, but they were not consistent. Minimal alterations were observed in blood parameters (clinical pathology) at one day after treatment. For these effects a dose response relationship could not be established as most alterations were only observed at the highest dose (5 mg/kg bw) administered.

For both coated and uncoated nano-ZnO indications for acute liver damage (alterations in blood parameters ALP, AST and bilirubin) were observed at the highest dose of 5 mg/kg bw administered.

Indications for kidney damage (increase in blood urea and creatinine) were only observed for zinc sulfate after the single high dose (17.6 mg/kg bw).

Four weeks after the administration, no alterations were observed in blood parameters evaluated for clinical pathology and in histopathology in the major organs evaluated (brain, lung, liver, spleen and kidney). Thus, there was no indication for persistent toxic effects.

The intravenous administration provided a 100% bioavailability of the nano-ZnO. A limitation of the study is that only a single administration was investigated, while for zinc sulfate the equimolar dose was administered both as a single dose or divided over four administrations in four weeks. The comparison of coated and uncoated ZnO is limited to acute reactions as indicated by several alterations in blood parameters at day 1 after treatment with 1 mg/kg bw or 5 mg/kg bw. A real comparison between coated and uncoated ZnO cannot be performed as toxic reactions were not observed.

In view of the indications for liver damage, a repeated dose toxicity study would have provided better information on potential toxicity.

In conclusion, after intravenous administration obtaining an internal dose of 5 mg/kg bw alterations in clinical pathology were observed that were indicative of liver damage. The liver damage appeared to be transient as similar alterations were not observed at four weeks after the single administration.

Based on this study, the SCCS considers a NOAEL of 1 mg/kg bw for acute toxicity of ZnO in rats after intravenous administration.

Additional studies submitted

Additional information was included in the dossier on ZnO commercially obtained from Sigma-Aldrich and not related to the ZnO presented in the dossier.

(References: 85, 94, 119)

In vitro cytotoxicity studies

Exploratory Comet assay in human epidermal cells

These studies demonstrated dose and time dependent cytotoxicity of ZnO nanoparticles. LDH release started at 5, 8 and 20 μg/mL exposure. Information on doses expressed as surface area and number of particles was not provided. NRU assay showed cytotoxicity at 8–20 μg/mL exposure. Oxidative stress markers were increased at a dose of 0.008 μg/mL exposure.

According to the applicant the cytotoxic responses in terms of disturbance of cellular homeostasis were observed at an overdosing of ZnO nanoparticles. The effect was attributed not to the nanoparticles per se, but to the released Zn ions as the soluble part of the ZnO nanoparticles.

(Reference: 94)

Comment

The SCCS does not consider an effect induced at doses as low as 0.08 μg/mL or even 5 μg/mL as a result of overdosing. Therefore, the SCCS considers ZnO nanoparticles at these doses cytotoxic. Even if the effects were due to the release of toxic ions, these were introduced by the added ZnO nanoparticles so the toxicity is indirectly due to the ZnO nanoparticles.

Comparison of in vivo and in vitro studies

For the development of a predictive in vitro assay to assess the lung hazard potential of nanomaterials, a comparative screening study was performed. The main objectives were to compare lung toxicity impacts of nanoscale (NZnO) versus fine zinc oxide (FZnO) particles, assess predictability of in vitro cell culture systems, and compare the effects of instillation versus inhalation exposures in rats. The fine and nano-sized ZnO materials were obtained from Sigma-Aldrich, St. Louis, USA (purity: >99%). The surface area of the fine and nano-ZnO was 9.6 m²/g and 12.1 m²/g, respectively. Information on the number of particles per mass was not provided. The authors concluded that the comparisons of in vivo and in vitro toxicity measurements following nano- or fine-ZnO exposures demonstrated little convergence and few differences in potency. In addition, the implementation of the current cell culture methodologies tested in this study did not accurately predict the pulmonary hazards associated with in vivo exposures to ZnO.

(References: 85, 119)

Method for screening of photoactivity of nanoparticles

Undoped and doped zinc oxide was prepared by proprietary methods and characterized chemically and physically. Besides size, no other information is presented on the ZnO used. Information on surface area and number of particles per mass was not provided. In contrast to a comparative study cited in the dossier, the screening method revealed photoactivity behaviour increasing (one to twofold) with decreasing crystallite size for crystallites <100 nm. A further series of nanophase ZnO powders doped with metal ions showed that, depending on the type and/or level of dopant, the photoactivity behaviour of nanoparticles could be reduced by more than an order of magnitude. According to the dossier, this enzymatic method should be considered as an initial exploratory screening study on photoreactivity in vitro with questionable relevance for in vivo effects on the skin.

(Reference: 33)

Comment

The paper was published in 2004 (Reference 33, submission III) and no follow up studies were cited In view of the rather limited characterization of the ZnO preparations used, the study does not contribute to the evaluation of the use of ZnO as a UV-filter in sunscreen formulations. However, scientifically the results indicate that the photoactivity of the ZnO nanoparticles can be manipulated by doping of the particles.

The SCCS considers this study not relevant for the evaluation of the use of ZnO as a cosmetic ingredient.

Discussion

Several submissions on ZnO have been received during recent years. Submission I dealt with ZnO in general, the main focus was on zinc ions as being responsible for Zn toxicity. The ZnO used was only marginally characterized. Submission II relied mainly on the report on ZnO of the regular risk assessment prepared in the context of Council Regulation (EEC) No. 793/93 on the evaluation and control of existing substances. This risk assessment was peer reviewed by the EU Commission’s Scientific Committee on Toxicity, Ecotoxicity and the Environment (CSTEE). Submission III, evaluated in this opinion, deals with ZnO nanoparticles from four different sources, either coated or uncoated, with a mean size range of 30 nm to 55 nm as indicated by their number size distribution.

Regarding the dosimetry of nanoparticles it is not yet clear which metric is the best dose descriptor for nanomaterials. However, it is assumed that for evaluation of toxicity endpoints the dose metric of mass, i.e. mg/kg body weight or μg/ml in in vitro systems, might not be the best parameter for expression of the dose response relationship. The total surface area administered and or the number of particles have been suggested as parameters to better describe dose response relationships for nanoparticles as it is the unit of one particle that interacts with the biological system and not the number of molecules of that particle, in contrast to soluble chemicals in which the total number of molecules present can initiate such interactions. However, in the present Opinion, the dose metric used for the risk assessment has only been described in mass units.

Physicochemical properties

It should be noted that it is technically possible to produce nano-dispersed ZnO with different physicochemical properties. Therefore, this opinion relates to those coated and uncoated ZnO nanomaterials for which data have been submitted in this dossier. It is required that manufacturers carefully analyze the raw materials which they obtain, particularly if they come in a finally stabilized formulation in liquid dispersions.

It should be noted that this assessment applies to those ZnO nanoparticles that are included in this dossier, and similar materials that have the following characteristics:

ZnO nanoparticles of purity ≥99%, with wurtzite crystalline structure, and physical appearance as described in the dossier, i.e. clusters that are rod-like, star-like and/or isometric shapes.
ZnO nanoparticles with a median diameter (D50: 50% of the particles below this diameter) of the particle number size distribution between 30 nm and 55 nm, and D1 (1% of the particles below this diameter) above 20 nm.
ZnO nanoparticles that are either uncoated, or coated with triethoxycaprylylsilane, dimethicone, dimethoxydiphenylsilanetriethoxycaprylylsilane cross-polymer, or octyl triethoxy silane.
ZnO nanoparticles that have a comparable solubility to that reported in the dossier, i.e. below 50 mg/L (approximately the maximum solubility of the ZnO nanomaterials for which data are provided in the dossier).

In the submitted dossier on ZnO nanomaterials the purity of one of the ZnO nanomaterials used in some safety assays was indicated to be ≥96%. However, data on the purity is not provided in the physico-chemical characterization section of the dossier (section 3.1.4), and was also not provided in the various study reports cited although in one report a content of 100% was indicated.

As discussed in section 3.1.6, the parameter of equilibrium solubility describes the static dissolved fraction in a saturated solution. However, since ZnO nanoparticles and/or their agglomerates/aggregates in the body will not be in a static equilibrium of dissolution, Zn ions will be continuously generated and taken into the endogenous Zn pool, and the ZnO particles will continue to dissolve until completeness. Hence the parameter of dissolution rate is more appropriate than the static solubility parameter when considering the fate of ZnO nanoparticles in the body. As the dermal penetration, if any, seems to be associated with the Zn ions released, the solubility of ZnO nanomaterials used as cosmetic ingredients should be below 50 mg/L because for this solubility level the data presented in this Opinion show an acceptable Margin of Safety.

Irritation, sensitization

A skin irritation study (not according to a guideline) was performed in which male Guinea pigs were exposed to 25% and 40% of 20 nm ZnO dispersed in ethanol. No effects were observed at any time during the administration and observation periods in the 25% test substance group. Slight erythema was observed in one of three animals in the 40% test substance group on day 3 of administration.

The ZnO nanomaterial, both as a neat dispersion and 25% solution as used in sunscreen, was slightly and transiently irritating to the eyes when tested in rabbits.

The sensitization assay was not performed according to a recognized OECD guideline. Furthermore, a concurrent positive control with a well known weak sensitizer was not included in this assay, so there is no certainty as to whether the test system used was able to identify weak sensitizers. However, for eight contact sensitizers, similar responses were found in the GPMT performed according to the OECD Guideline 406, and in a shortened test.

The validity of this (or any) test for demonstrating sensitization potency of nanomaterials has not yet been demonstrated. The inclusion of a positive particle control might overcome this problem. However, no positive particle control has been identified thus far.

General toxicity

In an acute oral toxicity study a dose of 2 g/kg body weight in rats induced some body weight decrease in one animal, although no deaths were observed. In one study all animals (mice) survived a single oral dose of 5 g/kg body weight of both nano-sized ZnO (50 nm) and micro-sized ZnO (1,226 nm), although the large ZnO particles induced a body weight reduction in both male and female animals (Reference: AR16). After oral administration of 2.5 g/kg body weight, systemic uptake was demonstrated and indications for acute liver toxicity (at 24-72 hours after administration) were noted in serum enzyme values and histopathology (Reference: AR16). However, in another study a similar oral dose of 5 g/kg body weight of ZnO nanoparticles in mice resulted in the death of one out of five animals when ZnO of 20 nm in size was administered, whereas five out of ten animals died after administration of 120 nm diameter ZnO (Reference: 118). In addition, in several organs e.g. liver and spleen, toxicity was observed for both the 20 nm and 120 nm ZnO particles. Similar results were reported after oral administration in a recent rat study indicative of liver toxicity following oral administration of 5 mg/kg body weight (Reference: AR20). Overall, it can be concluded that after oral administration Zn can become systemically available resulting in liver toxicity.

A comparative (exploratory) intravenous toxicity study of ZnO as coated or uncoated nanoparticles, as a pigment form or as zinc sulfate was performed. No persistent effects were observed at the end of the four week observation period when the various ZnO preparations were injected intravenously as a single dose at levels of 1 or 5 mg/kg bw. Deviations in some hematology parameters (monocyte, large unstained cells) and clinical chemistry parameters (alterations in blood parameters ALP, AST and bilirubin) indicative of an effect on liver function (coated and uncoated ZnO-nano preparations) and kidney function (zinc sulfate) were observed 1 day after intravenous injection of 5 mg/kg bw pigmented or coated nanoscale ZnO or an equimolar zinc sulfate dose. As at 1 mg/kg bw, no effect was observed and a dose response relationship could not be established. There was no indication for persistent toxic effects and, as at four weeks after the administration, no alterations were observed in blood parameters evaluated for clinical pathology and for histopathology in the major organs evaluated (brain, lung, liver, spleen and kidney). In general there was no difference in the reactions between the various injected forms of ZnO. There were no indications that nanoscale ZnO induced different effects compared to pigmentary ZnO or zinc sulfate.

The intravenous administration provided 100% bioavailability of the nano-ZnO. A limitation of the study is that only a single administration was investigated, while for the Zn sulfate the equimolar dose was administered both as a single dose or divided over four administrations in four weeks. The comparison of coated and uncoated ZnO is limited to acute reactions as indicated by several alterations in blood parameters at day 1 after treatment with 1 mg/kg bw or 5 mg/kg bw. A real comparison between coated and uncoated ZnO cannot be performed as toxic reactions were not observed. In view of indications for liver damage, a repeated dose toxicity study would have provided better information on potential toxicity.

In general it can be concluded that based on the observations on serum liver enzyme levels and histopathology, the systemic availability of either ZnO nanoparticles or Zn ions has the potential to induce liver toxicity.

No oral repeated dose toxicity study with ZnO nanoparticles is available.

Inhalation

In a 5 day exposure study in rats, ZnO induced a concentration-related inflammation reaction in the lung which was associated with dose-dependent increases in BALF. In addition to the inflammation reaction, necrosis was detected in the lung and the nose. There was no difference between the nano-sized and pigmentary ZnO. Local effects in the lung were still present 25 days after the exposure. There was no indication of systemic effects after the 5 day inhalation exposure.

After inhalation exposure of human volunteers, no acute toxic effects were observed with a dose of 500 μg/m³ air with 2 hours of exposure.

In view of the observed effects in the respiratory tract a caution is warranted against the use of ZnO nanoparticles in spray application despite the potentially low inhalation exposure.

Mutagenicity

Mutagenicity has been addressed in the dossier by both in vitro and in vivo tests.

The bacterial reverse mutation assay (according to OECD Guideline 471) was performed with dimethoxydiphenylsilanetriethoxycaprylylsilane coated ZnO nanoparticles in a panel of Salmonella typhimurium strains, both with and without S9-mix and was negative..The material tested negative for base pair changes and frame shift mutations, in contrast to the specific positive controls (MNNG, MIT.C, NOPD, AAC and 2-AA). Bacterial mutagenicity assays are judged less appropriate for testing of nanoparticles in comparison to mammalian cell systems in view of different internalisation pathways with eukaryotes and specific bactericidal effects of specific types of nanoparticles such as silver. According to the study provided in the dossier, no signs of bacteriology were observed. However, uptake of the nanoparticles by the bacteria and the role of Zn ions were not investigated. Negative findings in Ames tests (frame shifts, base pair substitutions) have also been reported in the literature for tetramethylammoniumhydroxide-capped ZnO nanoparticles (Reference: AR26) as well as ionic zinc (Reference: AR24).

A DNA damage potential of ZnO nanoparticles was revealed in A431 human epidermal cells by an alkakine Comet assay. Lipid peroxidation and oxidative stress induction were postulated as underlying mechanisms in the dossier. It should be emphasized however, that this does not exclude a genotoxic potential of the treatment agent, i.e. the applied ZnO nanoparticles. Several other studies in the open literature have confirmed DNA damaging effects with ZnO nanoparticles with the alkaline Comet assay (References: AR6, AR11, AR12, AR14, AR22).

The mouse micronucleus assay (according to OECD Guideline 474) using intraperitoneal administration showed an absence of clastogenic and aneugenic effects of triethoxycaprylylsilane coated ZnO nanoparticles, whereas the controls cyclophosphamide and vincristine sulfate tested positive. No evidence was provided that the ZnO nanoparticles (or dissolved ions) reached the bone marrow after the intraperitoneal applications, hence the results of this study are of limited value.

The SCCS concludes that there is evidence for in vitro genotoxicity of ZnO nanoparticles as demonstrated in the positive comet assay, but does not consider the negative results of the in vivo assay valid as exposure of the target organs was not demonstrated.

Both in vitro and in vivo data on photo-genotoxicity/mutagenicity were provided in the dossier. In Salmonella typhimurium mutation assays, ZnO revealed no gene mutations by base pair changes or frame shifts in the presence or absence of irradiation. In other investigations ZnO showed some clastogenic and photo-clastogenic potential. In CHO fibroblasts, increased structural chromosome aberrations were found both in the absence or presence of simulated sunlight. In CHO cells, clastogenic effects were observed after treatment with ZnO in the presence or absence of simultaneous UV irradiation. Increased chromosome aberrations were also observed in cells in which UV irradiation was performed prior to ZnO treatment, suggesting that radiation increases the susceptibility of the CHO cells to the clastogenic activity of ZnO. In vivo, ZnO was investigated in a photo-micronucleus test in hairless mice after topical application and did not increase the incidence of micronucleated epidermal cells in exposed animals with or without simulated light irradiation.

The SCCS is of the view that ZnO nanoparticles are (photo)clastogenic in vitro. The negative results in the in vivo micronucleus assay for epidermal cells can not be accepted as actual exposure of the living epidermal cells was not demonstrated.

Although the conclusion that UV-irridiation increases the sensitivity of the photomutagenicity/clastogenicity assay may be valid, the induction of chromosomal aberration in the absence of UV exposure should not be ignored. This should be considered a positive genotoxic activity of ZnO nanoparticles. Further studies available in the open literature have shown that ZnO nanoparticles do not cause mutations in FE1-Muta™Mouse lung epithelial cells (Reference: AR1), and induce micronuclei in A549 human lung epithelial cells only at a concentration where marked toxicity was observed (Reference: AR2).

It has been discussed in the open literature that part of the toxicity of nano ZnO is due to generation of reactive oxygen species, which has been associated with DNA damage (AR28). Scavenging of these oxygen radicals can be done by adding antioxidants or other components to the formulations. The impact of this in relation to potential harmful effects of nano ZnO in dermally applied cosmetic products is sofar not known.

Based on the available database and additional in depth evaluation of the studies, and in view of the uncertainties over whether or not nanoparticles reached the target cells/DNA in the tests, there is no conclusive evidence to conclude whether or not micro-or nano-sized ZnO particles pose a mutagenic/genotoxic, photo-toxic or photo-mutagenic/genotoxic risk to humans. However, where ZnO nanoparticles are applied on the skin in a sunscreen formulation, there is sufficient evidence to conclude that due to the very low if any systemic exposure, the risk to the consumer is negligible. The evidence from in vitro and in vivo studies presented in this dossier (section 3.3.4), and other studies on different metal/metal oxide nanoparticles (e.g. titanium dioxide – Nanoderm Project¹) shows that penetration of nano or larger particles is generally limited to the upper few layers of the stratum corneum and there is no significant dermal penetration of the particles to systemic circulation. Whilst this leaves the possibility for nanoparticle mediated local effects, it diminishes the possibility for any harmful effects at the systemic level.

Toxicokinetics

After oral exposure there is some uptake of Zn in the systemic circulation. ZnO nanoparticles of approximately 50 nm in size (TEM evaluation) were compared to ZnO microparticles showing at least one diameter >100 nm (TEM evaluation). After oral and intraperitoneal administration for both ZnO nanoparticles and microparticles, Zn could be observed in serum indicating uptake from the GI–tract, either as particulate materials or as dissolved Zn ions. For ZnO nanoparticles the systemic availability was somewhat higher compared to that of ZnO microparticles as indicated by Zn measurements by ICP-MS. Zn showed a higher distribution in the liver, spleen and lung after treatment with ZnO nanoparticles compared to treatment with ZnO microparticles.

After inhalation exposure elevated zinc levels were detected in various organs, most likely due to zinc ions dissolved from the ZnO particles.

There are no indications for significant if any penetration of nanoparticles through the skin, most likely only a minimal amount of Zn ions released from the nanoparticles may be available for systemic exposure.

Carcinogenicity

Specific data on carcinogenicity studies of ZnO nanomaterials are not available. In view of the occurring dissolution of the ZnO nanoparticles it can be assumed that the carcinogenic risk is similar to the conventionally manufactured ZnO preparations. According to the EU¹ Risk assessment report (Reference 44, submission III), there is no clear experimental or epidemiological evidence for a direct carcinogenic action of zinc or its compounds.

¹ Nanoderm Project: http://ec.europa.eu/research/environment/pdf/env_health_projects/nanotechnology/n-nanoderm.pdf

La structure à trois niveaux utilisée dans la communication de l’avis du comité scientifique pour la sécurité des consommateurs (CSSC) est protégée par les droits d’auteur de Cogeneris SPRL.