Analyst View Article Online
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
PAPER
Cite this: Analyst, 2014, 139, 964
View Journal | View Issue
Using citrate-functionalized TiO2 nanoparticles to study the effect of particle size on zebrafish embryo toxicity M.-S. Kim,a K. M. Louis,b J. A. Pedersen,bcd R. J. Hamers,b R. E. Petersonad and W. Heideman*ad TiO2 nanoparticles (NPs) are photoactive, potentially producing toxicity in vivo in the presence of sunlight. We have previously demonstrated photodependent toxicity in zebrafish embryos exposed to TiO2 NPs. Here we investigate the effect of particle size on developing zebrafish exposed to 6, 12 and 15 nm citrate-functionalized anatase TiO2 NPs under either simulated sunlight illumination or in the dark. All three sizes of TiO2 NPs caused photo-dependent toxicity. Under simulated sunlight illumination, the acute toxicity of the 6 nm citrate-TiO2 NPs (120 h LC50 of 23.4 mg L1) exceeded that of the 12 and 15 nm citrate-TiO2 NPs. Exposure to 6 nm particles under illumination also caused a higher incidence of developmental defects than the larger particles. These abnormalities included pericardial edema, yolksac edema, craniofacial malformation, and opaque yolk. To gain insight into the mechanisms of toxicity, we measured hydroxyl radicals (cOH) generated by NPs in vitro and reactive oxygen species (ROS) produced in vivo. We found that on a mass basis, smaller particles generated higher levels of ROS both in vitro and in vivo, and the 6 nm citrate-TiO2 NPs induced more oxidative stress than larger particles in the zebrafish embryo. We examined oxidative DNA damage by measuring 8-hydroxydeoxyguanosine in zebrafish exposed to different-sized citrate-TiO2 NPs and found that 6 nm particles caused more DNA damage than did larger particles (12 and 15 nm) under illumination. Our results indicate a photo-
Received 19th October 2013 Accepted 4th December 2013
dependent toxicity of citrate-TiO2 NPs to zebrafish embryos, with an inverse relationship between particle size and toxicity. Production of more ROS, resulting in more oxidative stress and more DNA damage, represents one possible mechanism of the higher toxicity of smaller citrate-TiO2 NPs. These
DOI: 10.1039/c3an01966g www.rsc.org/analyst
results highlight the relationship between citrate-TiO2 NP size and toxicity/oxidative stress in developing zebrafish embryos.
Introduction Manufactured titanium dioxide (TiO2) nanoparticles (NPs) are used in the pharmaceutical, biomedical, electrical, and environmental elds.1–4 These particles are employed as ultraviolet (UV) protective screens or as photocatalysts.5 They have high photocatalytic activity, high UV radiation absorption, wide band gaps and high refractive index, and they are chemically stable. TiO2 as a material is generally considered to be environmentally benign.6 The rapidly increasing use of TiO2 in the nanoparticle form has prompted questions about the potential risks of adverse
a
Pharmaceutical Science Division, School of Pharmacy, University of Wisconsin, Madison, USA. E-mail: [email protected]; Fax: +1-608-265-3316; Tel: +1-608262-1795
b c
Department of Chemistry, University of Wisconsin, Madison, USA
Department of Soil Sciences, University of Wisconsin, Madison, USA
d
Molecular and Environmental Toxicology Center, University of Wisconsin, Madison, USA
964 | Analyst, 2014, 139, 964–972
effects on human health and the environment. Several recent studies have reported that TiO2 NPs exhibit photo-enhanced toxicity to several aquatic species including marine phytoplankton (Isochrysis galbana, Thalassiosira pseudomana, Dunaliella lertiolecta), microcrustaceans (Daphnia magna), amphibians (Xenopus laevis), and sh (Oryzias latipes, Danio rerio).7–12 When TiO2 absorbs photons with energy exceeding the band gap energy (3.2 eV for bulk anatase and 3.02 eV for bulk rutile), electrons are promoted from the valence to the conduction band, leaving a hole (h+) in the valence band.13 Electron–hole pairs may directly recombine, but at the particle surface in aqueous environments they can react with water and oxygen to form reactive oxygen species (ROS), such as superoxide (O2c), hydrogen peroxide (H2O2), singlet oxygen (1O2) and hydroxyl radical (cOH): the surface h+ can oxidize H2O, while the e can reduce O2 to generate ROS.14,15 As TiO2 nanoparticle size decreases, the surface area of the particle increases relative to mass. Therefore, smaller particles have an increased contact area with the aqueous environment per unit mass, and might be expected to produce more ROS
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
Paper
upon photoactivation on a mass basis. Jiang et al.16 found that the ROS activity of illuminated anatase TiO2 NPs in the size ranges of 4–10 nm and 30–200 nm changed with the surface area; over some ranges, smaller TiO2 NPs produced more ROS at the same mass concentration. Since ROS production is strongly related to TiO2 NP toxicity, this leads to the expectation that, on a mass basis, smaller TiO2 NPs may be more toxic than larger TiO2 NPs. Several studies have shown that nano-scale TiO2 nanoparticles are more toxic than ne-sized bulk TiO2 particles.17,18 However, at present, little information is available about the effect of TiO2 NP size on photo-dependent toxicity in a whole organism. Citrate is commonly used as a stabilizing or capping agent for nanoparticles, and has been frequently used as a stabilizer for commercially available silver and gold nanoparticles. Citrate is a non-toxic organic acid, inexpensive, and easily removed by acidication. Zebrash (Danio rerio) embryos are becoming a very useful vertebrate model in toxicology because of their small size, optical clarity, rapid development, fundamental similarity to other vertebrates including humans, and low experimental costs. Results obtained with zebrash are relevant to human health and wild sh populations.19–21 The objective of the present study was to investigate the inuence of particle size on photo-dependent toxicity of citrate-TiO2 nanoparticles using zebrash embryos. Our results indicate that particle size does indeed affect ROS production in vitro, ROS damage in vivo, and toxicity.
Experimental Synthesis of citrate-functionalized titanium dioxide nanoparticles (TiO2 NPs) Citrate-TiO2 NPs were synthesized according to the method reported by Kotsokechagia et al.22 One milliliter of TiCl4 (9.12 103 mol) was slowly added into a glass vial containing 5 mL of anhydrous ethanol. Twenty mL of anhydrous benzyl alcohol (0.19 mol) was added with stirring, and the ask was le open at 80 C and kept stirring for 6–15 h until the average particle diameter of the nanoparticles had reached the target value (6–15 nm). The particles were precipitated out of solution using a 1 : 2 (v/v) ratio of diethyl ether to nanoparticle suspension. The mixture was centrifuged until the supernatant was clear and the nanoparticles created a white solid pellet at the bottom. The liquid supernatant was decanted and the nanoparticles were re-suspended in water. Determination of ROS generated in citrate-TiO2 NPs suspensions Hydroxyl radicals (cOH) were measured quantitatively using a uorescent probe, 30 -(p-aminophenyl) uorescein (APF) purchased from Life Technologies Corporation. Aer samples were exposed to light or dark conditions, 10 mM APF and 100 mM phosphate buffer pH 7.2 were added to each sample. The samples were then transferred to amber vials to prevent additional exposure to light. Steady-state
This journal is © The Royal Society of Chemistry 2014
Analyst
uorescence measurements were taken using an ISS K2 TimeResolved Spectrouorometer using slit sizes of 2.0 mm and 0.5 mm for the excitation and emission respectively. The samples were excited at 490 nm and emission peak occurred between 510 and 525 nm. To prevent oversaturation of the photon-counting detector, each of the samples was diluted by a factor of 10. A linear calibration curve was used to determine concentrations of cOH present in each sample. To create the linear calibration curve, 5 mL solutions of 0 to 1000 nM H2O2 concentrations in 100 nM increments were made using ultrapure (18 MU cm resistivity) H2O that was purged with Ar gas and le out of the light for at least one month. Each solution contained 50 mM of pH 7.2 phosphate buffer, 2.95 units per mL of horseradish peroxidase, and 10 mM of APF. To correlate to similar quantities of APF as in the TiO2 samples, each of the standards was diluted by a factor of 10. Fluorescence measurements were performed using the same conditions as the TiO2 samples, and the intensity of uorescence emission at 520 nm was used to create a standard curve.
Characterization of citrate-TiO2 NPs The primary particle diameters of citrate-functionalized TiO2 NPs were determined by transmission electron microscopy (TEM) (High-Resolution TEM Philips CM200, Eindhoven, The Netherlands). Dispersed citrate-TiO2 NPs were deposited on carbon-coated copper grids and observed under TEM with a Gatan CCD camera (Gatan, Pleasanton, CA) at an accelerating voltage of 200 kV. Images sizes were determined using ImageJ soware http://rsb.info.nih.gov/ij/. Particle averages and standard deviations were 5.7 1.2 nm, n ¼ 109; 12.4 2.6 nm, n ¼ 51; 15.0 2.2 nm, n ¼ 47. The hydrodynamic diameters and zeta potentials of the TiO2 NPs were determined by dynamic light scattering (DLS) and electrophoretic light scattering (ELS) using a Malvern Zetasizer Nano (Malvern Instruments Ltd, Worchestershire, UK) using a refractive index of nR ¼ 2.49. Measurements were made using nanoparticles suspended in exposure water (vide infra for composition) as they were used in toxicity experiments. Suspensions were prepared for dosing with a 20 min sonication to disaggregate particles. The dosing suspension was sonicated in a water bath sonicator (Laboratory Supplies Inc., Hicksville, NY; Model no. G112SPIT, 600 volts, 80 kHz, and 0.5 Amps). Samples were collected immediately aer initial suspension in exposure medium and also aer 24 hours with and without illumination with a metal halide lamp, exactly as the particles were used during the toxicity studies. For each time point, three samples were collected, and three DLS measurements were made on each sample. To determine the zeta potential of citrate-TiO2 NPs suspensions in exposure medium, samples were prepared as described above and introduced into disposable capillary cells and measured using a Malvern Zetasizer Nano. All measurements were conducted at 25 C, using a concentration of 500 mg mL1. Analyst, 2014, 139, 964–972 | 965
View Article Online
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
Analyst
Paper
Zebrash husbandry and dosing
Oxidative stress detection using CM-H2DCFDA
Adult AB strain wild type zebrash (Danio rerio) were maintained at 30 C in glass recirculation aquaria containing reverse osmosis puried water to which 58 mg of Instant Ocean™ salt (Aquarium Systems; Mentor, OH) and 47.6 mg NaHCO3 (Fisher Scientic) per L. This lightly buffered water was used in all suspensions of nanoparticles as well, and is a standard in zebrash husbandry around the world. It yields an approximate nal ionic composition of 4.4 mM Cl, 4.6 mM Na+, 0.78 mM HCO3, 0.22 mM Mg2+, 87 mM K+, 81 mM Ca2+, 27 mM CO32, 5.7 mM Br, 4.2 mM B(OH)3 + B(OH)4, 0.81 mM Sr2+, 0.43 mM F, 0.35 mM Li+, and 15 nM I (ionic strength ¼ 0.07 mM, pH ¼ 7–7.4). Fish were fed twice daily with “Sprirulina” ake food (O.S.L. Marine Lab., Inc., Burlingame, USA) and housed with a 14 : 10 h light–dark cycle. Embryos were obtained using group spawns in 10 gallon aquaria, at a ratio of 1 male to 2 females. Embryos were collected at 4 hours post fertilization (hpf), and distributed (10–12 embryos per well) in 24-well at-bottom cell culture plates (BD Biosciences, Franklin Lakes, NJ, USA) using a sterile plastic pipette. The plates were housed at 30 C. During the course of the experiments, the embryos/larvae grew on yolk nutrients, and were not fed. The protocol for zebrash use and maintenance was approved by the Research Animal Resources Center of the University of Wisconsin-Madison (protocol # M00489). Nanoparticle suspensions were prepared new each day in the water used to house the sh. Embryos were collected from spawns as described and at 4 hours post fertilization (hpf) placed in 24 well plates containing different concentrations (0–1000 mg mL1) of the TiO2 NPs (6 wells per concentration; total 60–70 embryos per concentration). Dosing solutions (1 mL per well) were prepared and renewed each day.
Oxidative stress induced by citrate-TiO2 NPs nanoparticles was assessed by the uorescence produced by the oxidized form of the cell-permeant ROS indicator acetyl ester of 5-(and 6-) chloromethyl-20 ,70 -dichlorodihydrouorescein diacetate (CM-H2DCFDA). This molecule reacts with many different ROS, and it is a general indicator of oxidative stress (Invitrogen). Embryos at 4 hpf were exposed to 500 mg mL1 of citrate-TiO2 NPs for 120 h under illumination or in the dark and collected at 120 hpf. Embryos exposed to water with no TiO2 NPs during same periods served as the control group. Dosing solutions were renewed every 24 h. Collected larvae were washed with ultra pure H2O three times, and excess water was removed. Larvae were incubated in 1 mg mL1 CM-H2DCFDA for 2 h and washed again with ultra pure H2O three times, then immobilized in 3% methylcellulose. Images of live samples were captured using a confocal microscope (excitation/emission ¼ 485 nm/535 nm) on an Olympus FV 1000.
Illumination “Illuminated” embryos/larvae were placed 45 cm below a bluespectrum metal halide lamp (XM 250 W, 10 000 K; electronic ballast; hellolights.com) that simulates the slightly blue sunlight in (z1 m) water. Fish were illuminated for 14 h per day. Non illuminated sh were placed beside the illuminated sh, but the plates were covered with aluminium foil. Temperatures were maintained at 30 C for both illuminated and control groups. Mortality and sublethal toxicity assays Embryos/larvae were screened and scored daily for survival. To measure sublethal toxic defects, surviving embryos/larvae were collected at 120 hpf, and immobilized in 3% methylcellulose and photographed live using a MicroFire camera (Optronics, Goleta, CA, USA) mounted onto a Leica MZ16 stereomicroscope (Meyer Instruments, Houston, TX, USA). Toxic defects included pericardial edema, yolk sac edema, craniofacial malformation, opaque yolk and altered axial curvature. All experiments were conducted in triplicate. LC50 values and their 95% condence intervals for exposure to different-sized citrate-TiO2 NPs were assessed by Probit analysis using POLO-PC soware (LeOra Soware, Berkeley, CA).
966 | Analyst, 2014, 139, 964–972
8-OHdG ELISA assay Embryos were exposed to citrate-TiO2 NPs (500 mg mL1) from 4 to 120 hpf with either illumination or in the dark as described above. Dosing solutions were changed every 24 h. Embryos (n ¼ 50–60 per treatment) were collected at 120 hpf, rinsed with ultrapure H2O, and ash frozen. DNA was extracted using DNA extraction buffer (50 mM Tris, 1 mM EDTA, 20 mM NaCl, 0.25% SDS) containing Proteinase-K (0.2 mg mL1; AMRESCO, Solon, OH, USA) at 55 C for 3 h, heated at 95 C for 10 min to inactivate Proteinase-K, and then precipitated with glycogen (0.2 mg mL1). Extracted DNA was digested using nuclease P1, and then used as ELISA samples using a 96-well competitive 8-OHdG EIA Kit according to the manufacturer's instructions (StressMarq Biosciences, Victoria, BC, Canada). Plates were read at 420 nm using a SpectraMax M2 microplate reader (Molecular Devices, Sunnyvale, CA, USA).
Results Nanoparticle properties We synthesized citrate-TiO2 NPs of three sizes: 6, 12 and 15 nm. Citrate is a commonly used ligand in nanoparticle syntheses to control nanoparticle size and shape, and is easily removed if desired by acidication.23 To conrm the primary particle sizes, we used transmission electron microscopy (TEM). Fig. 1 shows particle size distribution histograms derived from TEM micrographs of the citrate-TiO2 NPs used for this study. Mean primary particle diameters for the three anatase TiO2 NP preparations were 5.7 1.2 nm (n ¼ 109), 12.4 2.6 nm (n ¼ 51), and 15.0 2.2 nm (n ¼ 47), that we nominally refer to as 6, 12, and 15 nm respectively. The hydrodynamic diameters and zeta potentials of the TiO2 NPs were determined by DLS and ELS. We made these measurements under conditions that correspond to the conditions used in zebrash exposure experiments. This meant measurement immediately following suspension in water, and again aer 24 h, corresponding to the time when the TiO2
This journal is © The Royal Society of Chemistry 2014
View Article Online
Paper
Analyst
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
size of the 12 nm and 15 nm particles increased substantially, and became multimodal. Aer 24 h with illumination, the 12 nm citrate-TiO2 NPs had an average aggregate size of 380 34 nm: the average hydrodynamic diameter without illumination at 24 h was 400 120 nm. However, these averages were made from several distinct peaks. Aer 24 h with illumination the 15 nm particles had an average aggregate size of 700 60 nm: without illumination this was only 220 40 nm. For reasons we do not understand, the average hydrodynamic diameter of 6 nm particles changed little over time. At 24 h it was 23.9 0.7 with illumination and 18.8 0.8 nm without illumination. In most cases we observed shoulders or multiple peaks, indicating particle aggregates of discrete sizes, suggesting that over time the particles formed aggregates of specic sizes.
Fig. 1 Histograms of primary particle size. Particles were prepared for TEM as described and counted for size using morphometric tools provided by the microscope software. Each bar represents the fraction of the total particles found in the indicated size interval.
nanoparticles suspension would be discarded and renewed in the zebrash exposures. Initially, at t ¼ 0 h, the average hydrodynamic diameters of the 6, 12, and 15 nm particles in the exposure water were approximately 2–3 times that of the measured primary particle size at 18.1 0.7 nm, 19.8 0.9 nm and 28.9 2.2 nm, respectively (Fig. 2). Aer 24 h in exposure water, the aggregate size of the 6 nm particles did not change dramatically, but the
This journal is © The Royal Society of Chemistry 2014
Hydrodynmaic diameter distributions. Particles were prepared as they would be for dosing zebrafish embryos. DLS measurements are shown for each particle size at times representing exposure conditions, 0 h when suspensions were made, and 24 h when suspensions were replenished.
Fig. 2
Analyst, 2014, 139, 964–972 | 967
View Article Online
Analyst Table 1
Paper Zeta potential of citrate-TiO2 NPs of different sizes
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
Zeta potentiala Size
0h
24 h + illumination
6 nm 12 nm 15 nm
35.62 (0.61) 30.17 (1.02) 25.95 (0.81)
42.75 (0.64) 39.00 (0.63) 31.17 (0.96)
a Data is presented as means (S.D.) and values at the same time point are signicantly different from each other (p < 0.05) in ANOVA and Tukey's test.
The zeta potentials of citrate-TiO2 NPs at 0 h, were 36 1.8 mV, 30 2.5 mV and 26 2.0 mV for the 6 nm, 12 nm, 15 nm particles, respectively. As with the DLS experiments, the zeta potentials changed aer 24 h with illumination, becoming more negative (Table 1). At both 0 and 24 h, the smaller citrateTiO2 NPs showed lower zeta potentials values than those of larger particles. Size-dependent toxicity We exposed zebrash embryos to graded concentrations of the different sized nanoparticles over a period of 5 days, monitoring survival over that period. Because TiO2 nanoparticles have been previously shown to produce photo-dependent toxicity, we exposed sh to TiO2 NPs with and without illumination from a metal halide lamp simulating sunlight. In triplicate experiments, the 6 and 12 nm sizes of citrate-TiO2 NPs showed photoand dose-dependent toxicity, with the 6 nm particles more potent than the other sizes of particles (Fig. 3). On a mass basis, the LC50 value for 6 nm particles with illumination was 23 mg mL1 (95% CI: 1.1–190), while the value for the 12 nm particles
was 610 mg mL1 (95% CI: 58–8700), and was not measurable for the 15 nm citrate-TiO2 NPs. The large condence interval for the LC50 value for the 12 nm NPs made it impossible to state with statistical condence that there was a difference between the values for 6 and 12 nm LC50s. However the trend was apparent, and the two had signicantly different degrees of toxicity at the highest doses (P < 0.05). For the 6 and 12 nm particle sizes, illumination increased toxicity. While we could identify clear differences in toxicity between the 6 and 15 nm TiO2 nanoparticles with illumination, we observed no statistically signicant effect of size on toxicity without illumination, even at the highest doses. The toxic effects of the TiO2 NPs and illumination were manifested as malformations during development. These included pericardial and yolk sac edema that are caused by uid accumulation, and are characteristic of cardiovascular failure. We observed craniofacial malformation, and altered axial body curvature, suggesting possible effects on cartilage deposition. Also noted was opaque and unabsorbed yolk, a somewhat unique response to TiO2 nanoparticles under illumination.10–12 This failure to absorb the yolk could be due to circulation failure and could be a cause of the stunted growth observed. All of these effects can be seen in a representative 120 hpf zebrash exposed since fertilization to 6 nm particles with illumination (Fig. 4A). We recorded the incidences of these toxic effects produced by each size of TiO2 nanoparticle at 500 mg mL1 in surviving sh at 120 hpf (Fig. 4B). Approximately 85% of the control sh had no observable defects, and were scored as normal. It is not uncommon to nd a few individuals within a large group of developing sh with some defect, even in the controls. However in contrast to the controls in which the large majority were scored without any defect, less than 15% of the sh exposed to the 6 nm particles with illumination appeared normal. Among these sh, 77% (3.9) had pericardial edema, 60% (15) had yolk sac edema, 61% (9.1) had craniofacial malformation, 64% (8.3) had opaque yolk, and 27% (1.0) had altered axial curvature. Embryos exposed to 12 nm particles with illumination were scored as 50% (10) pericardial edema, 31% (9.8) yolk sac edema, 22% (5.2)% craniofacial malformation, 27% (2.8) opaque yolk, and 14% (4.8) altered axial curvature. Consistent with the lower mortality, the 15 nm also produced fewer malformations, with 36% (6.4) pericardial edema, 13% (1.4) yolk sac edema, 11% (2.3) craniofacial malformation, 15% (3.2) % opaque yolk, and 8% (2.4) altered axial curvature. The nominal 6 nm particles produced more toxicity in all categories than the larger particles.
ROS and oxidative stress
Fig. 3 Toxicity of different sized citrate-TiO2 NPs at 120 hpf. Zebrafish embryos were incubated with the indicated concentration of TiO2 NP from 4 to 120 hpf with and without illumination and mortality was recorded. Asterisks indicate significant difference from particle of the same size and concentration without illumination (p < 0.05).
968 | Analyst, 2014, 139, 964–972
To gain deeper insight into the mechanisms of size-dependent toxicity of citrate-TiO2 NPs, we measured the ability of citrateTiO2 NPs to generate ROS upon illumination. We used APF, a uorogenic ROS indicator, to measure ROS production in vitro, matching the conditions of the toxicity assays. Fig. 5 shows the levels of hydroxyl radical (cOH) generated by citrate-TiO2 NPs with illumination in the exposure suspension. Consistent with
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
Paper
Analyst
Fig. 4 Forms of developmental toxicity caused by citrate-TiO2 NPs. Images at left show a normal zebrafish at 120 hpf and one that had been exposed to 6 nm TiO2 NPs (500 mg mL1) with illumination. The bar graph shows incidence of malformations for each size particle with and without illumination. PE, pericardial edema; YSE, yolk sac edema; CFM, craniofacial malformation; OY, opaque yolk; AC, axial curvature. Letters indicate values that are not significantly different from each other (p > 0.05; ANOVA, Tukey's test). Mean SEM (n ¼ 3, 40–50 larvae for each n).
signs of oxidative stress is the formation of 8-OHdG DNA adducts. We used an ELISA to measure the formation of 8OHdG in sh exposed to the citrate-TiO2 NPs. As shown in Fig. 7, the 8-OHdG levels in the embryos exposed to citrate-TiO2 NPs with illumination were signicantly higher than the illuminated controls. As with the other assays, the 6 nm particles induced signicantly higher level of 8-OHdG than the larger
Production of hydroxyl radicals by TiO2 NPs. Waterborne TiO2 NPs (1000 mg mL1) were placed in 24-well plates and incubated under simulated sunlight illumination for 24 h and hydroxyl radical (cOH) concentrations were determined using APF (10 mM) and a luminometer. Fig. 5
its higher potency in causing toxicity, on a mass/volume basis the 6 nm citrate-TiO2 NPs, produced signicantly higher levels of hydroxyl radicals than the 12 or 15 nm particles. We also used a uorescent ROS indicator, CM-H2DCFDA, to determine whether the nanoparticles produce ROS in vivo. This indicator produces a uorescent green product when oxidized by ROS. As with the in vitro assays, we observed higher level of uorescence intensity in embryos exposed to 6 nm particles compared to those exposed to the 12 and 15 nm particles, and the ROS signal was illumination-dependent (Fig. 6). In all sh, including the controls with no added TiO2 NPs, we observed some signal in control embryos within the gut and intestine. This background signal was observed with and without illumination, so this signal apparently reects natural oxidative processes. The long-term damaging effect of ROS exposure in vivo is the formation of oxidized macromolecules. One of the hallmark
This journal is © The Royal Society of Chemistry 2014
Fig. 6 In vivo production of ROS by TiO2. CMH2DCFDA staining of embryos exposed to citrate-TiO2 NPs of the indicated size (500 mg mL1). Shown are overlays of fluorescence and bright field images taken at 120 hpf. White filled arrows indicate non-specific fluorescence observed in all fish tested. Black arrows indicate TiO2 NPspecific signal. Segments were processed in the same manner, and neither brightness, nor contrast was manipulated differently between images.
Analyst, 2014, 139, 964–972 | 969
View Article Online
Analyst
particles. Taken together, we nd a consistent pattern of photodependent ROS production and DNA damage, with the smaller particles producing more ROS and oxidative DNA damage.
Discussion
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
Size-dependent toxicity of TiO2 NPs In the present study, we showed that, under simulated sunlight illumination, 6 nm citrate-TiO2 NPs were more potent on a mass basis than the 12 and 15 nm particles in causing pericardial edema, yolk sac edema, craniofacial malformation and opaque yolk in developing zebrash embryos. In previous work we have shown that low dose-exposure to TiO2 nanoparticles produces cumulative toxicity that results in mortality aer several weeks. Toxic effects are more readily observed, and occur earlier at higher doses. These exposure levels might not be observed in nature, but allowed us to make comparisons between the individual size groups. Smaller particles have larger surface area per unit mass for interaction with the aqueous environment aer photoactivation. It is therefore possible that the smaller particles are more efficient in producing ROS. This is borne out by our in vitro and in vivo assays: the smaller (6 nm) citrate-TiO2 NPs induced more ROS in vitro and more oxidative stress measured in vivo than the larger (12 and 15 nm) particles. Oxidative stress mediated by nanoparticles has been proposed as one of the mechanisms of nanoparticle toxicity.16,24,25 TiO2 nanoparticles are both photoactivatable and small. This may alter their ability to produce ROS and to interact with biological structures. Effects of oxidative stress during developmental process have
Paper
been well studied, and ROS are known to play important roles in both normal and altered embryonic development.26,27 Uptake measurements using ICP-OES showed no signicant trends separating the uptake of one size particle from another (not shown). We therefore cannot attribute the differences in potency to differences in particle uptake, suggesting that the particles themselves differ in toxicity within the sh tissues. Depending on pH, ionic strength, electrolytes, concentration, surface functionalization, and natural organic matter TiO2 NPs can aggregate or agglomerate in aquatic environments.12,28 Aggregation or agglomeration of TiO2 NPs can be difficult to control, and may affect toxicity, since it can change size, effective surface area, and other physicochemical properties. Surface coatings can be used to reduce or control agglomeration. For reasons that we cannot explain, we found that the smaller particles were less likely to aggregate during our exposures. This may have had an inuence on particle uptake, distribution, and toxicity. Despite the size of aggregates, we have previously found TiO2 NPs distributed throughout developing zebrash tissues, despite formation of 1 micron sized aggregates.11
Potency We note that this work was aimed at determining whether one size of particle is more toxic than another. How differences in potency arise is a logical, but separate question. As mentioned above, size could inuence toxicity in complex ways, through effects on uptake, distribution, and elimination, to say nothing of the cellular interactions that produce toxicity. Our work used lethality, specic malformations, in vivo ROS generation, and oxidative DNA damage as readouts to determine whether, with simple aqueous exposure, one size particle produced a larger nal effect than another size, regardless of body burden or mechanism.
Mass, particle number, and surface area
Fig. 7 DNA damage induced by different sizes of citrate-TiO2 NPs exposure with or without illumination. Shown are ELISA results using DNA extracted from embryos exposed to the indicated TiO2 NPs (500 mg mL1) from 4–120 hpf with and without illumination. DNA was extracted at 120 hpf for the 8-OHdG ELISA assay (n ¼ 3; 30–50 embryos pooled for each n). Value ¼ Mean SEM. * indicates different from control and all other values; ** indicates significantly different from control (p > 0.05; ANOVA, Tukey's test).
970 | Analyst, 2014, 139, 964–972
The results of this study have been presented using conventional mass dose metrics. For each particle size, the LC50 values were presented as 6 nm < 12 nm < 15 nm based on mass concentration. Fig. 8A shows the original dose response curve from Fig. 3 leading to that conclusion using particle mass as the X-axis. However, one can also plot mortality against surface area concentration (e.g., cm2 mL1) (Fig. 8B). In this case, the rankorder potency seems to disappear between the 6 and 12 nm particles; the 15 nm particles were not sufficiently toxic to produce a curve. We also plotted toxicity against particle number (Fig.8C). In this case all of the nanoparticles appeared to fall roughly along the same curve, with the 12 nm particle points lying only slightly to the le of the others. These results show that the dose metrics chosen to characterize the toxicity of TiO2 NPs can signicantly inuence the interpretation of the ndings. From a practical point of view, measurement by an easily established metric such as mass is desirable. We measure exposure and discharge in mass units, so measuring toxicity using these units seems useful, and will probably continue.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Paper
Analyst
It appears that examining potency from different perspectives provides useful information, but this brings up an intrinsic experimental problem: matching doses in terms of mass, area and particle number in a single experiment is quite difficult. At the highest dose, we had far more of the smaller particles than the larger ones. Had we chosen our doses based on particle number the larger sizes would have been used at very high mass concentrations compared to the smaller ones.
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
Conclusions Our work shows that from a mass standpoint, 6 nm TiO2 NPs were more potent in producing ROS in vivo and in vitro and in producing malformation and mortality in developing sh. As shown previously for TiO2 NPs, this was photodependent. However, the mechanism of toxicity may depend more on the surface area or particle number than the total TiO2 in suspension.
Acknowledgements We thank Dorothy Nesbit for maintaining lines and for help with zebrash husbandry, and Ofek Bar-Ilan for helpful discussions. This study was supported by Nanoscale Science and Engineering Center (NSEC) funded by the National Science Foundation (NSF) (DMR-0832760).
Notes and references
Fig. 8 Dose response curves showing mortality induced by TiO2 NPs. Shown are the same results as in Fig. 3, with different concentration scales for the X-axes.
However, using metrics other than mass may shed far more light on mechanism. For example, the fact that our curves appear to coalesce when plotted on the basis of particle number may indicate that each unit particle causes an equivalent amount of toxicity, and the particle concentration is what determines the lethal effects. Similar argument can be made for examining surface area, where again the differences in potency seem to diminish between sizes. If the effect is actually due to total surface area or particle number, plotting by mass can be misleading.
This journal is © The Royal Society of Chemistry 2014
1 R. J. Aitken, M. Q. Chaudhry, A. B. A. Boxall and M. Hull, Manufacture and use of nanomaterials: current status in the UK and global trends, Occup. Med., 2006, 56, 300–306. 2 R. Anselmann, Nanoparticles and nanolayers in commercial applications, J. Nanopart. Res., 2001, 3, 329–336. 3 J. S. Chen, M. C. Liu, J. D. Zhang, X. Y. Ying and L. T. Jin, Photocatalytic degradation of organic wastes by electrochemically assisted TiO2 photocatalytic system, J. Environ. Manage., 2004, 70, 43–47. 4 T. Maier and H. C. Korting, Sunscreens – which and what for?, Skin Pharmacol. Physiol., 2005, 18, 253–262. 5 T. Kaida, K. Kobayashi, M. Adachi and F. Suzuki, Optical characteristics of titanium oxide interference lm and the lm laminated with oxides and their applications for cosmetics, J. Cosmet. Sci., 2004, 55, 219–220. 6 N. S. Allen, M. Edge, J. Verran, J. Stratton, J. Maltby and C. Bygott, Photocatalytic titania based surfaces: environmental benets, Polym. Degrad. Stab., 2008, 93, 1632–1646. 7 R. J. Miller, S. Bennett, A. A. Keller, S. Pease and H. S. Lenihan, TiO2 nanoparticles are phototoxic to marine phytoplankton, PLoS One, 2012, 7, e30321. 8 C. Ma, J. L. Marlowe and A. Puga, The aryl hydrocarbon receptor at the crossroads of multiple signaling pathways, EXS, 2009, 99, 231–257. 9 J. L. Zhang, M. Wages, S. B. Cox, J. D. Maul, Y. J. Li, M. Barnes, et al., Effect of titanium dioxide nanomaterials
Analyst, 2014, 139, 964–972 | 971
View Article Online
Analyst
10
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
11
12
13
14
15
16
17
18
and ultraviolet light coexposure on African clawed frogs (Xenopus laevis), Environ. Toxicol. Chem., 2012, 31, 176–183. O. Bar-Ilan, C. C. Chuang, D. J. Schwahn, S. Yang, S. Joshi, J. A. Pedersen, et al., TiO2 nanoparticle exposure and illumination during zebrash development: mortality at parts per billion concentrations, Environ. Sci. Technol., 2013, 47, 4726–4733. O. Bar-Ilan, K. M. Louis, S. P. Yang, J. A. Pedersen, R. J. Hamers, R. E. Peterson, et al., Titanium dioxide nanoparticles produce phototoxicity in the developing zebrash, Nanotoxicology, 2012, 6, 670–679. S. P. Yang, O. Bar-Ilan, R. E. Peterson, W. Heideman, R. J. Hamers and J. A. Pedersen, Inuence of humic acid on titanium dioxide nanoparticle toxicity to developing zebrash, Environ. Sci. Technol., 2013, 47, 4718–4725. H. Lin, C. P. Huang, W. Li, C. Ni, S. I. Shah and Y. H. Tseng, Size dependency of nanocrystalline TiO2 on its optical property and photocatalytic reactivity exemplied by 2chlorophenol, Appl. Catal., B, 2006, 68, 1–11. O. Carp, C. L. Huisman and A. Reller, Photoinduced reactivity of titanium dioxide, Prog. Solid State Chem., 2004, 32, 33–177. S. G. Kumar and L. G. Devi, Review on modied TiO2 photocatalysis under UV/visible light: selected results and related mechanisms on interfacial charge carrier transfer dynamics, J. Phys. Chem. A, 2011, 115, 13211–13241. J. Jiang, G. Oberdorster, A. Elder, R. Gelein, P. Mercer and P. Biswas, Does nanoparticle activity depend upon size and crystal phase?, Nanotoxicology, 2008, 2, 33–42. D. W. Xiong, T. Fang, L. P. Yu, X. F. Sima and W. T. Zhu, Effects of nano-scale TiO2, ZnO and their bulk counterparts on zebrash: acute toxicity, oxidative stress and oxidative damage, Sci. Total Environ., 2011, 409, 1444–1452. J. Wang, G. Zhou, C. Chen, H. Yu, T. Wang, Y. Ma, et al., Acute toxicity and biodistribution of different sized titanium dioxide particles in mice aer oral administration, Toxicol. Lett., 2007, 168, 176–185.
972 | Analyst, 2014, 139, 964–972
Paper
19 A. J. Hill, H. Teraoka, W. Heideman and R. E. Peterson, Zebrash as a model vertebrate for investigating chemical toxicity, Toxicol. Sci., 2005, 86, 6–19. 20 A. Nasevicius and S. C. Ekker, The zebrash as a novel system for functional genomics and therapeutic development applications, Curr. Opin. Mol. Ther., 2001, 3, 224–228. 21 H. Teraoka, W. Dong and T. Hiraga, Zebrash as a novel experimental model for developmental toxicology, Congenital Anomalies, 2003, 43, 123–132. 22 T. Kotsokechagia, F. Cellesi, A. Thomas, M. Niederberger and N. Tirelli, Preparation of ligand-free TiO2 (anatase) nanoparticles through a nonaqueous process and their surface functionalization, Langmuir, 2008, 24, 6988–6997. 23 I. A. Mudunkotuwa and V. H. Grassian, Citric acid adsorption on TiO2 nanoparticles in aqueous suspensions at acidic and circumneutral pH: surface coverage, surface speciation, and its impact on nanoparticle– nanoparticle interactions, J. Am. Chem. Soc., 2010, 132, 14986–14994. 24 H. J. Johnston, G. R. Hutchison, F. M. Christensen, S. Peters, S. Hankin and V. Stone, Identication of the mechanisms that drive the toxicity of TiO2 particulates: the contribution of physicochemical characteristics, Part. Fibre Toxicol., 2009, 6, 33. 25 G. Oberdorster, E. Oberdorster and J. Oberdorster, Nanotoxicology: an emerging discipline evolving from studies of ultrane particles, Environ. Health Perspect., 2005, 113, 823–839. 26 P. A. Dennery, Effects of oxidative stress on embryonic development, Birth Defects Res., Part C, 2007, 81, 155–162. 27 J. M. Hansen, Oxidative stress as a mechanism of teratogenesis, Birth Defects Res., Part C, 2006, 78, 293–307. 28 V. K. Sharma, Aggregation and toxicity of titanium dioxide nanoparticles in aquatic environment – a review, J. Environ. Sci. Health, Part A: Toxic/Hazard. Subst. Environ. Eng., 2009, 44, 1485–1495.
This journal is © The Royal Society of Chemistry 2014
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
PAPER
Cite this: Analyst, 2014, 139, 964
View Journal | View Issue
Using citrate-functionalized TiO2 nanoparticles to study the effect of particle size on zebrafish embryo toxicity M.-S. Kim,a K. M. Louis,b J. A. Pedersen,bcd R. J. Hamers,b R. E. Petersonad and W. Heideman*ad TiO2 nanoparticles (NPs) are photoactive, potentially producing toxicity in vivo in the presence of sunlight. We have previously demonstrated photodependent toxicity in zebrafish embryos exposed to TiO2 NPs. Here we investigate the effect of particle size on developing zebrafish exposed to 6, 12 and 15 nm citrate-functionalized anatase TiO2 NPs under either simulated sunlight illumination or in the dark. All three sizes of TiO2 NPs caused photo-dependent toxicity. Under simulated sunlight illumination, the acute toxicity of the 6 nm citrate-TiO2 NPs (120 h LC50 of 23.4 mg L1) exceeded that of the 12 and 15 nm citrate-TiO2 NPs. Exposure to 6 nm particles under illumination also caused a higher incidence of developmental defects than the larger particles. These abnormalities included pericardial edema, yolksac edema, craniofacial malformation, and opaque yolk. To gain insight into the mechanisms of toxicity, we measured hydroxyl radicals (cOH) generated by NPs in vitro and reactive oxygen species (ROS) produced in vivo. We found that on a mass basis, smaller particles generated higher levels of ROS both in vitro and in vivo, and the 6 nm citrate-TiO2 NPs induced more oxidative stress than larger particles in the zebrafish embryo. We examined oxidative DNA damage by measuring 8-hydroxydeoxyguanosine in zebrafish exposed to different-sized citrate-TiO2 NPs and found that 6 nm particles caused more DNA damage than did larger particles (12 and 15 nm) under illumination. Our results indicate a photo-
Received 19th October 2013 Accepted 4th December 2013
dependent toxicity of citrate-TiO2 NPs to zebrafish embryos, with an inverse relationship between particle size and toxicity. Production of more ROS, resulting in more oxidative stress and more DNA damage, represents one possible mechanism of the higher toxicity of smaller citrate-TiO2 NPs. These
DOI: 10.1039/c3an01966g www.rsc.org/analyst
results highlight the relationship between citrate-TiO2 NP size and toxicity/oxidative stress in developing zebrafish embryos.
Introduction Manufactured titanium dioxide (TiO2) nanoparticles (NPs) are used in the pharmaceutical, biomedical, electrical, and environmental elds.1–4 These particles are employed as ultraviolet (UV) protective screens or as photocatalysts.5 They have high photocatalytic activity, high UV radiation absorption, wide band gaps and high refractive index, and they are chemically stable. TiO2 as a material is generally considered to be environmentally benign.6 The rapidly increasing use of TiO2 in the nanoparticle form has prompted questions about the potential risks of adverse
a
Pharmaceutical Science Division, School of Pharmacy, University of Wisconsin, Madison, USA. E-mail: [email protected]; Fax: +1-608-265-3316; Tel: +1-608262-1795
b c
Department of Chemistry, University of Wisconsin, Madison, USA
Department of Soil Sciences, University of Wisconsin, Madison, USA
d
Molecular and Environmental Toxicology Center, University of Wisconsin, Madison, USA
964 | Analyst, 2014, 139, 964–972
effects on human health and the environment. Several recent studies have reported that TiO2 NPs exhibit photo-enhanced toxicity to several aquatic species including marine phytoplankton (Isochrysis galbana, Thalassiosira pseudomana, Dunaliella lertiolecta), microcrustaceans (Daphnia magna), amphibians (Xenopus laevis), and sh (Oryzias latipes, Danio rerio).7–12 When TiO2 absorbs photons with energy exceeding the band gap energy (3.2 eV for bulk anatase and 3.02 eV for bulk rutile), electrons are promoted from the valence to the conduction band, leaving a hole (h+) in the valence band.13 Electron–hole pairs may directly recombine, but at the particle surface in aqueous environments they can react with water and oxygen to form reactive oxygen species (ROS), such as superoxide (O2c), hydrogen peroxide (H2O2), singlet oxygen (1O2) and hydroxyl radical (cOH): the surface h+ can oxidize H2O, while the e can reduce O2 to generate ROS.14,15 As TiO2 nanoparticle size decreases, the surface area of the particle increases relative to mass. Therefore, smaller particles have an increased contact area with the aqueous environment per unit mass, and might be expected to produce more ROS
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
Paper
upon photoactivation on a mass basis. Jiang et al.16 found that the ROS activity of illuminated anatase TiO2 NPs in the size ranges of 4–10 nm and 30–200 nm changed with the surface area; over some ranges, smaller TiO2 NPs produced more ROS at the same mass concentration. Since ROS production is strongly related to TiO2 NP toxicity, this leads to the expectation that, on a mass basis, smaller TiO2 NPs may be more toxic than larger TiO2 NPs. Several studies have shown that nano-scale TiO2 nanoparticles are more toxic than ne-sized bulk TiO2 particles.17,18 However, at present, little information is available about the effect of TiO2 NP size on photo-dependent toxicity in a whole organism. Citrate is commonly used as a stabilizing or capping agent for nanoparticles, and has been frequently used as a stabilizer for commercially available silver and gold nanoparticles. Citrate is a non-toxic organic acid, inexpensive, and easily removed by acidication. Zebrash (Danio rerio) embryos are becoming a very useful vertebrate model in toxicology because of their small size, optical clarity, rapid development, fundamental similarity to other vertebrates including humans, and low experimental costs. Results obtained with zebrash are relevant to human health and wild sh populations.19–21 The objective of the present study was to investigate the inuence of particle size on photo-dependent toxicity of citrate-TiO2 nanoparticles using zebrash embryos. Our results indicate that particle size does indeed affect ROS production in vitro, ROS damage in vivo, and toxicity.
Experimental Synthesis of citrate-functionalized titanium dioxide nanoparticles (TiO2 NPs) Citrate-TiO2 NPs were synthesized according to the method reported by Kotsokechagia et al.22 One milliliter of TiCl4 (9.12 103 mol) was slowly added into a glass vial containing 5 mL of anhydrous ethanol. Twenty mL of anhydrous benzyl alcohol (0.19 mol) was added with stirring, and the ask was le open at 80 C and kept stirring for 6–15 h until the average particle diameter of the nanoparticles had reached the target value (6–15 nm). The particles were precipitated out of solution using a 1 : 2 (v/v) ratio of diethyl ether to nanoparticle suspension. The mixture was centrifuged until the supernatant was clear and the nanoparticles created a white solid pellet at the bottom. The liquid supernatant was decanted and the nanoparticles were re-suspended in water. Determination of ROS generated in citrate-TiO2 NPs suspensions Hydroxyl radicals (cOH) were measured quantitatively using a uorescent probe, 30 -(p-aminophenyl) uorescein (APF) purchased from Life Technologies Corporation. Aer samples were exposed to light or dark conditions, 10 mM APF and 100 mM phosphate buffer pH 7.2 were added to each sample. The samples were then transferred to amber vials to prevent additional exposure to light. Steady-state
This journal is © The Royal Society of Chemistry 2014
Analyst
uorescence measurements were taken using an ISS K2 TimeResolved Spectrouorometer using slit sizes of 2.0 mm and 0.5 mm for the excitation and emission respectively. The samples were excited at 490 nm and emission peak occurred between 510 and 525 nm. To prevent oversaturation of the photon-counting detector, each of the samples was diluted by a factor of 10. A linear calibration curve was used to determine concentrations of cOH present in each sample. To create the linear calibration curve, 5 mL solutions of 0 to 1000 nM H2O2 concentrations in 100 nM increments were made using ultrapure (18 MU cm resistivity) H2O that was purged with Ar gas and le out of the light for at least one month. Each solution contained 50 mM of pH 7.2 phosphate buffer, 2.95 units per mL of horseradish peroxidase, and 10 mM of APF. To correlate to similar quantities of APF as in the TiO2 samples, each of the standards was diluted by a factor of 10. Fluorescence measurements were performed using the same conditions as the TiO2 samples, and the intensity of uorescence emission at 520 nm was used to create a standard curve.
Characterization of citrate-TiO2 NPs The primary particle diameters of citrate-functionalized TiO2 NPs were determined by transmission electron microscopy (TEM) (High-Resolution TEM Philips CM200, Eindhoven, The Netherlands). Dispersed citrate-TiO2 NPs were deposited on carbon-coated copper grids and observed under TEM with a Gatan CCD camera (Gatan, Pleasanton, CA) at an accelerating voltage of 200 kV. Images sizes were determined using ImageJ soware http://rsb.info.nih.gov/ij/. Particle averages and standard deviations were 5.7 1.2 nm, n ¼ 109; 12.4 2.6 nm, n ¼ 51; 15.0 2.2 nm, n ¼ 47. The hydrodynamic diameters and zeta potentials of the TiO2 NPs were determined by dynamic light scattering (DLS) and electrophoretic light scattering (ELS) using a Malvern Zetasizer Nano (Malvern Instruments Ltd, Worchestershire, UK) using a refractive index of nR ¼ 2.49. Measurements were made using nanoparticles suspended in exposure water (vide infra for composition) as they were used in toxicity experiments. Suspensions were prepared for dosing with a 20 min sonication to disaggregate particles. The dosing suspension was sonicated in a water bath sonicator (Laboratory Supplies Inc., Hicksville, NY; Model no. G112SPIT, 600 volts, 80 kHz, and 0.5 Amps). Samples were collected immediately aer initial suspension in exposure medium and also aer 24 hours with and without illumination with a metal halide lamp, exactly as the particles were used during the toxicity studies. For each time point, three samples were collected, and three DLS measurements were made on each sample. To determine the zeta potential of citrate-TiO2 NPs suspensions in exposure medium, samples were prepared as described above and introduced into disposable capillary cells and measured using a Malvern Zetasizer Nano. All measurements were conducted at 25 C, using a concentration of 500 mg mL1. Analyst, 2014, 139, 964–972 | 965
View Article Online
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
Analyst
Paper
Zebrash husbandry and dosing
Oxidative stress detection using CM-H2DCFDA
Adult AB strain wild type zebrash (Danio rerio) were maintained at 30 C in glass recirculation aquaria containing reverse osmosis puried water to which 58 mg of Instant Ocean™ salt (Aquarium Systems; Mentor, OH) and 47.6 mg NaHCO3 (Fisher Scientic) per L. This lightly buffered water was used in all suspensions of nanoparticles as well, and is a standard in zebrash husbandry around the world. It yields an approximate nal ionic composition of 4.4 mM Cl, 4.6 mM Na+, 0.78 mM HCO3, 0.22 mM Mg2+, 87 mM K+, 81 mM Ca2+, 27 mM CO32, 5.7 mM Br, 4.2 mM B(OH)3 + B(OH)4, 0.81 mM Sr2+, 0.43 mM F, 0.35 mM Li+, and 15 nM I (ionic strength ¼ 0.07 mM, pH ¼ 7–7.4). Fish were fed twice daily with “Sprirulina” ake food (O.S.L. Marine Lab., Inc., Burlingame, USA) and housed with a 14 : 10 h light–dark cycle. Embryos were obtained using group spawns in 10 gallon aquaria, at a ratio of 1 male to 2 females. Embryos were collected at 4 hours post fertilization (hpf), and distributed (10–12 embryos per well) in 24-well at-bottom cell culture plates (BD Biosciences, Franklin Lakes, NJ, USA) using a sterile plastic pipette. The plates were housed at 30 C. During the course of the experiments, the embryos/larvae grew on yolk nutrients, and were not fed. The protocol for zebrash use and maintenance was approved by the Research Animal Resources Center of the University of Wisconsin-Madison (protocol # M00489). Nanoparticle suspensions were prepared new each day in the water used to house the sh. Embryos were collected from spawns as described and at 4 hours post fertilization (hpf) placed in 24 well plates containing different concentrations (0–1000 mg mL1) of the TiO2 NPs (6 wells per concentration; total 60–70 embryos per concentration). Dosing solutions (1 mL per well) were prepared and renewed each day.
Oxidative stress induced by citrate-TiO2 NPs nanoparticles was assessed by the uorescence produced by the oxidized form of the cell-permeant ROS indicator acetyl ester of 5-(and 6-) chloromethyl-20 ,70 -dichlorodihydrouorescein diacetate (CM-H2DCFDA). This molecule reacts with many different ROS, and it is a general indicator of oxidative stress (Invitrogen). Embryos at 4 hpf were exposed to 500 mg mL1 of citrate-TiO2 NPs for 120 h under illumination or in the dark and collected at 120 hpf. Embryos exposed to water with no TiO2 NPs during same periods served as the control group. Dosing solutions were renewed every 24 h. Collected larvae were washed with ultra pure H2O three times, and excess water was removed. Larvae were incubated in 1 mg mL1 CM-H2DCFDA for 2 h and washed again with ultra pure H2O three times, then immobilized in 3% methylcellulose. Images of live samples were captured using a confocal microscope (excitation/emission ¼ 485 nm/535 nm) on an Olympus FV 1000.
Illumination “Illuminated” embryos/larvae were placed 45 cm below a bluespectrum metal halide lamp (XM 250 W, 10 000 K; electronic ballast; hellolights.com) that simulates the slightly blue sunlight in (z1 m) water. Fish were illuminated for 14 h per day. Non illuminated sh were placed beside the illuminated sh, but the plates were covered with aluminium foil. Temperatures were maintained at 30 C for both illuminated and control groups. Mortality and sublethal toxicity assays Embryos/larvae were screened and scored daily for survival. To measure sublethal toxic defects, surviving embryos/larvae were collected at 120 hpf, and immobilized in 3% methylcellulose and photographed live using a MicroFire camera (Optronics, Goleta, CA, USA) mounted onto a Leica MZ16 stereomicroscope (Meyer Instruments, Houston, TX, USA). Toxic defects included pericardial edema, yolk sac edema, craniofacial malformation, opaque yolk and altered axial curvature. All experiments were conducted in triplicate. LC50 values and their 95% condence intervals for exposure to different-sized citrate-TiO2 NPs were assessed by Probit analysis using POLO-PC soware (LeOra Soware, Berkeley, CA).
966 | Analyst, 2014, 139, 964–972
8-OHdG ELISA assay Embryos were exposed to citrate-TiO2 NPs (500 mg mL1) from 4 to 120 hpf with either illumination or in the dark as described above. Dosing solutions were changed every 24 h. Embryos (n ¼ 50–60 per treatment) were collected at 120 hpf, rinsed with ultrapure H2O, and ash frozen. DNA was extracted using DNA extraction buffer (50 mM Tris, 1 mM EDTA, 20 mM NaCl, 0.25% SDS) containing Proteinase-K (0.2 mg mL1; AMRESCO, Solon, OH, USA) at 55 C for 3 h, heated at 95 C for 10 min to inactivate Proteinase-K, and then precipitated with glycogen (0.2 mg mL1). Extracted DNA was digested using nuclease P1, and then used as ELISA samples using a 96-well competitive 8-OHdG EIA Kit according to the manufacturer's instructions (StressMarq Biosciences, Victoria, BC, Canada). Plates were read at 420 nm using a SpectraMax M2 microplate reader (Molecular Devices, Sunnyvale, CA, USA).
Results Nanoparticle properties We synthesized citrate-TiO2 NPs of three sizes: 6, 12 and 15 nm. Citrate is a commonly used ligand in nanoparticle syntheses to control nanoparticle size and shape, and is easily removed if desired by acidication.23 To conrm the primary particle sizes, we used transmission electron microscopy (TEM). Fig. 1 shows particle size distribution histograms derived from TEM micrographs of the citrate-TiO2 NPs used for this study. Mean primary particle diameters for the three anatase TiO2 NP preparations were 5.7 1.2 nm (n ¼ 109), 12.4 2.6 nm (n ¼ 51), and 15.0 2.2 nm (n ¼ 47), that we nominally refer to as 6, 12, and 15 nm respectively. The hydrodynamic diameters and zeta potentials of the TiO2 NPs were determined by DLS and ELS. We made these measurements under conditions that correspond to the conditions used in zebrash exposure experiments. This meant measurement immediately following suspension in water, and again aer 24 h, corresponding to the time when the TiO2
This journal is © The Royal Society of Chemistry 2014
View Article Online
Paper
Analyst
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
size of the 12 nm and 15 nm particles increased substantially, and became multimodal. Aer 24 h with illumination, the 12 nm citrate-TiO2 NPs had an average aggregate size of 380 34 nm: the average hydrodynamic diameter without illumination at 24 h was 400 120 nm. However, these averages were made from several distinct peaks. Aer 24 h with illumination the 15 nm particles had an average aggregate size of 700 60 nm: without illumination this was only 220 40 nm. For reasons we do not understand, the average hydrodynamic diameter of 6 nm particles changed little over time. At 24 h it was 23.9 0.7 with illumination and 18.8 0.8 nm without illumination. In most cases we observed shoulders or multiple peaks, indicating particle aggregates of discrete sizes, suggesting that over time the particles formed aggregates of specic sizes.
Fig. 1 Histograms of primary particle size. Particles were prepared for TEM as described and counted for size using morphometric tools provided by the microscope software. Each bar represents the fraction of the total particles found in the indicated size interval.
nanoparticles suspension would be discarded and renewed in the zebrash exposures. Initially, at t ¼ 0 h, the average hydrodynamic diameters of the 6, 12, and 15 nm particles in the exposure water were approximately 2–3 times that of the measured primary particle size at 18.1 0.7 nm, 19.8 0.9 nm and 28.9 2.2 nm, respectively (Fig. 2). Aer 24 h in exposure water, the aggregate size of the 6 nm particles did not change dramatically, but the
This journal is © The Royal Society of Chemistry 2014
Hydrodynmaic diameter distributions. Particles were prepared as they would be for dosing zebrafish embryos. DLS measurements are shown for each particle size at times representing exposure conditions, 0 h when suspensions were made, and 24 h when suspensions were replenished.
Fig. 2
Analyst, 2014, 139, 964–972 | 967
View Article Online
Analyst Table 1
Paper Zeta potential of citrate-TiO2 NPs of different sizes
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
Zeta potentiala Size
0h
24 h + illumination
6 nm 12 nm 15 nm
35.62 (0.61) 30.17 (1.02) 25.95 (0.81)
42.75 (0.64) 39.00 (0.63) 31.17 (0.96)
a Data is presented as means (S.D.) and values at the same time point are signicantly different from each other (p < 0.05) in ANOVA and Tukey's test.
The zeta potentials of citrate-TiO2 NPs at 0 h, were 36 1.8 mV, 30 2.5 mV and 26 2.0 mV for the 6 nm, 12 nm, 15 nm particles, respectively. As with the DLS experiments, the zeta potentials changed aer 24 h with illumination, becoming more negative (Table 1). At both 0 and 24 h, the smaller citrateTiO2 NPs showed lower zeta potentials values than those of larger particles. Size-dependent toxicity We exposed zebrash embryos to graded concentrations of the different sized nanoparticles over a period of 5 days, monitoring survival over that period. Because TiO2 nanoparticles have been previously shown to produce photo-dependent toxicity, we exposed sh to TiO2 NPs with and without illumination from a metal halide lamp simulating sunlight. In triplicate experiments, the 6 and 12 nm sizes of citrate-TiO2 NPs showed photoand dose-dependent toxicity, with the 6 nm particles more potent than the other sizes of particles (Fig. 3). On a mass basis, the LC50 value for 6 nm particles with illumination was 23 mg mL1 (95% CI: 1.1–190), while the value for the 12 nm particles
was 610 mg mL1 (95% CI: 58–8700), and was not measurable for the 15 nm citrate-TiO2 NPs. The large condence interval for the LC50 value for the 12 nm NPs made it impossible to state with statistical condence that there was a difference between the values for 6 and 12 nm LC50s. However the trend was apparent, and the two had signicantly different degrees of toxicity at the highest doses (P < 0.05). For the 6 and 12 nm particle sizes, illumination increased toxicity. While we could identify clear differences in toxicity between the 6 and 15 nm TiO2 nanoparticles with illumination, we observed no statistically signicant effect of size on toxicity without illumination, even at the highest doses. The toxic effects of the TiO2 NPs and illumination were manifested as malformations during development. These included pericardial and yolk sac edema that are caused by uid accumulation, and are characteristic of cardiovascular failure. We observed craniofacial malformation, and altered axial body curvature, suggesting possible effects on cartilage deposition. Also noted was opaque and unabsorbed yolk, a somewhat unique response to TiO2 nanoparticles under illumination.10–12 This failure to absorb the yolk could be due to circulation failure and could be a cause of the stunted growth observed. All of these effects can be seen in a representative 120 hpf zebrash exposed since fertilization to 6 nm particles with illumination (Fig. 4A). We recorded the incidences of these toxic effects produced by each size of TiO2 nanoparticle at 500 mg mL1 in surviving sh at 120 hpf (Fig. 4B). Approximately 85% of the control sh had no observable defects, and were scored as normal. It is not uncommon to nd a few individuals within a large group of developing sh with some defect, even in the controls. However in contrast to the controls in which the large majority were scored without any defect, less than 15% of the sh exposed to the 6 nm particles with illumination appeared normal. Among these sh, 77% (3.9) had pericardial edema, 60% (15) had yolk sac edema, 61% (9.1) had craniofacial malformation, 64% (8.3) had opaque yolk, and 27% (1.0) had altered axial curvature. Embryos exposed to 12 nm particles with illumination were scored as 50% (10) pericardial edema, 31% (9.8) yolk sac edema, 22% (5.2)% craniofacial malformation, 27% (2.8) opaque yolk, and 14% (4.8) altered axial curvature. Consistent with the lower mortality, the 15 nm also produced fewer malformations, with 36% (6.4) pericardial edema, 13% (1.4) yolk sac edema, 11% (2.3) craniofacial malformation, 15% (3.2) % opaque yolk, and 8% (2.4) altered axial curvature. The nominal 6 nm particles produced more toxicity in all categories than the larger particles.
ROS and oxidative stress
Fig. 3 Toxicity of different sized citrate-TiO2 NPs at 120 hpf. Zebrafish embryos were incubated with the indicated concentration of TiO2 NP from 4 to 120 hpf with and without illumination and mortality was recorded. Asterisks indicate significant difference from particle of the same size and concentration without illumination (p < 0.05).
968 | Analyst, 2014, 139, 964–972
To gain deeper insight into the mechanisms of size-dependent toxicity of citrate-TiO2 NPs, we measured the ability of citrateTiO2 NPs to generate ROS upon illumination. We used APF, a uorogenic ROS indicator, to measure ROS production in vitro, matching the conditions of the toxicity assays. Fig. 5 shows the levels of hydroxyl radical (cOH) generated by citrate-TiO2 NPs with illumination in the exposure suspension. Consistent with
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
Paper
Analyst
Fig. 4 Forms of developmental toxicity caused by citrate-TiO2 NPs. Images at left show a normal zebrafish at 120 hpf and one that had been exposed to 6 nm TiO2 NPs (500 mg mL1) with illumination. The bar graph shows incidence of malformations for each size particle with and without illumination. PE, pericardial edema; YSE, yolk sac edema; CFM, craniofacial malformation; OY, opaque yolk; AC, axial curvature. Letters indicate values that are not significantly different from each other (p > 0.05; ANOVA, Tukey's test). Mean SEM (n ¼ 3, 40–50 larvae for each n).
signs of oxidative stress is the formation of 8-OHdG DNA adducts. We used an ELISA to measure the formation of 8OHdG in sh exposed to the citrate-TiO2 NPs. As shown in Fig. 7, the 8-OHdG levels in the embryos exposed to citrate-TiO2 NPs with illumination were signicantly higher than the illuminated controls. As with the other assays, the 6 nm particles induced signicantly higher level of 8-OHdG than the larger
Production of hydroxyl radicals by TiO2 NPs. Waterborne TiO2 NPs (1000 mg mL1) were placed in 24-well plates and incubated under simulated sunlight illumination for 24 h and hydroxyl radical (cOH) concentrations were determined using APF (10 mM) and a luminometer. Fig. 5
its higher potency in causing toxicity, on a mass/volume basis the 6 nm citrate-TiO2 NPs, produced signicantly higher levels of hydroxyl radicals than the 12 or 15 nm particles. We also used a uorescent ROS indicator, CM-H2DCFDA, to determine whether the nanoparticles produce ROS in vivo. This indicator produces a uorescent green product when oxidized by ROS. As with the in vitro assays, we observed higher level of uorescence intensity in embryos exposed to 6 nm particles compared to those exposed to the 12 and 15 nm particles, and the ROS signal was illumination-dependent (Fig. 6). In all sh, including the controls with no added TiO2 NPs, we observed some signal in control embryos within the gut and intestine. This background signal was observed with and without illumination, so this signal apparently reects natural oxidative processes. The long-term damaging effect of ROS exposure in vivo is the formation of oxidized macromolecules. One of the hallmark
This journal is © The Royal Society of Chemistry 2014
Fig. 6 In vivo production of ROS by TiO2. CMH2DCFDA staining of embryos exposed to citrate-TiO2 NPs of the indicated size (500 mg mL1). Shown are overlays of fluorescence and bright field images taken at 120 hpf. White filled arrows indicate non-specific fluorescence observed in all fish tested. Black arrows indicate TiO2 NPspecific signal. Segments were processed in the same manner, and neither brightness, nor contrast was manipulated differently between images.
Analyst, 2014, 139, 964–972 | 969
View Article Online
Analyst
particles. Taken together, we nd a consistent pattern of photodependent ROS production and DNA damage, with the smaller particles producing more ROS and oxidative DNA damage.
Discussion
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
Size-dependent toxicity of TiO2 NPs In the present study, we showed that, under simulated sunlight illumination, 6 nm citrate-TiO2 NPs were more potent on a mass basis than the 12 and 15 nm particles in causing pericardial edema, yolk sac edema, craniofacial malformation and opaque yolk in developing zebrash embryos. In previous work we have shown that low dose-exposure to TiO2 nanoparticles produces cumulative toxicity that results in mortality aer several weeks. Toxic effects are more readily observed, and occur earlier at higher doses. These exposure levels might not be observed in nature, but allowed us to make comparisons between the individual size groups. Smaller particles have larger surface area per unit mass for interaction with the aqueous environment aer photoactivation. It is therefore possible that the smaller particles are more efficient in producing ROS. This is borne out by our in vitro and in vivo assays: the smaller (6 nm) citrate-TiO2 NPs induced more ROS in vitro and more oxidative stress measured in vivo than the larger (12 and 15 nm) particles. Oxidative stress mediated by nanoparticles has been proposed as one of the mechanisms of nanoparticle toxicity.16,24,25 TiO2 nanoparticles are both photoactivatable and small. This may alter their ability to produce ROS and to interact with biological structures. Effects of oxidative stress during developmental process have
Paper
been well studied, and ROS are known to play important roles in both normal and altered embryonic development.26,27 Uptake measurements using ICP-OES showed no signicant trends separating the uptake of one size particle from another (not shown). We therefore cannot attribute the differences in potency to differences in particle uptake, suggesting that the particles themselves differ in toxicity within the sh tissues. Depending on pH, ionic strength, electrolytes, concentration, surface functionalization, and natural organic matter TiO2 NPs can aggregate or agglomerate in aquatic environments.12,28 Aggregation or agglomeration of TiO2 NPs can be difficult to control, and may affect toxicity, since it can change size, effective surface area, and other physicochemical properties. Surface coatings can be used to reduce or control agglomeration. For reasons that we cannot explain, we found that the smaller particles were less likely to aggregate during our exposures. This may have had an inuence on particle uptake, distribution, and toxicity. Despite the size of aggregates, we have previously found TiO2 NPs distributed throughout developing zebrash tissues, despite formation of 1 micron sized aggregates.11
Potency We note that this work was aimed at determining whether one size of particle is more toxic than another. How differences in potency arise is a logical, but separate question. As mentioned above, size could inuence toxicity in complex ways, through effects on uptake, distribution, and elimination, to say nothing of the cellular interactions that produce toxicity. Our work used lethality, specic malformations, in vivo ROS generation, and oxidative DNA damage as readouts to determine whether, with simple aqueous exposure, one size particle produced a larger nal effect than another size, regardless of body burden or mechanism.
Mass, particle number, and surface area
Fig. 7 DNA damage induced by different sizes of citrate-TiO2 NPs exposure with or without illumination. Shown are ELISA results using DNA extracted from embryos exposed to the indicated TiO2 NPs (500 mg mL1) from 4–120 hpf with and without illumination. DNA was extracted at 120 hpf for the 8-OHdG ELISA assay (n ¼ 3; 30–50 embryos pooled for each n). Value ¼ Mean SEM. * indicates different from control and all other values; ** indicates significantly different from control (p > 0.05; ANOVA, Tukey's test).
970 | Analyst, 2014, 139, 964–972
The results of this study have been presented using conventional mass dose metrics. For each particle size, the LC50 values were presented as 6 nm < 12 nm < 15 nm based on mass concentration. Fig. 8A shows the original dose response curve from Fig. 3 leading to that conclusion using particle mass as the X-axis. However, one can also plot mortality against surface area concentration (e.g., cm2 mL1) (Fig. 8B). In this case, the rankorder potency seems to disappear between the 6 and 12 nm particles; the 15 nm particles were not sufficiently toxic to produce a curve. We also plotted toxicity against particle number (Fig.8C). In this case all of the nanoparticles appeared to fall roughly along the same curve, with the 12 nm particle points lying only slightly to the le of the others. These results show that the dose metrics chosen to characterize the toxicity of TiO2 NPs can signicantly inuence the interpretation of the ndings. From a practical point of view, measurement by an easily established metric such as mass is desirable. We measure exposure and discharge in mass units, so measuring toxicity using these units seems useful, and will probably continue.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Paper
Analyst
It appears that examining potency from different perspectives provides useful information, but this brings up an intrinsic experimental problem: matching doses in terms of mass, area and particle number in a single experiment is quite difficult. At the highest dose, we had far more of the smaller particles than the larger ones. Had we chosen our doses based on particle number the larger sizes would have been used at very high mass concentrations compared to the smaller ones.
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
Conclusions Our work shows that from a mass standpoint, 6 nm TiO2 NPs were more potent in producing ROS in vivo and in vitro and in producing malformation and mortality in developing sh. As shown previously for TiO2 NPs, this was photodependent. However, the mechanism of toxicity may depend more on the surface area or particle number than the total TiO2 in suspension.
Acknowledgements We thank Dorothy Nesbit for maintaining lines and for help with zebrash husbandry, and Ofek Bar-Ilan for helpful discussions. This study was supported by Nanoscale Science and Engineering Center (NSEC) funded by the National Science Foundation (NSF) (DMR-0832760).
Notes and references
Fig. 8 Dose response curves showing mortality induced by TiO2 NPs. Shown are the same results as in Fig. 3, with different concentration scales for the X-axes.
However, using metrics other than mass may shed far more light on mechanism. For example, the fact that our curves appear to coalesce when plotted on the basis of particle number may indicate that each unit particle causes an equivalent amount of toxicity, and the particle concentration is what determines the lethal effects. Similar argument can be made for examining surface area, where again the differences in potency seem to diminish between sizes. If the effect is actually due to total surface area or particle number, plotting by mass can be misleading.
This journal is © The Royal Society of Chemistry 2014
1 R. J. Aitken, M. Q. Chaudhry, A. B. A. Boxall and M. Hull, Manufacture and use of nanomaterials: current status in the UK and global trends, Occup. Med., 2006, 56, 300–306. 2 R. Anselmann, Nanoparticles and nanolayers in commercial applications, J. Nanopart. Res., 2001, 3, 329–336. 3 J. S. Chen, M. C. Liu, J. D. Zhang, X. Y. Ying and L. T. Jin, Photocatalytic degradation of organic wastes by electrochemically assisted TiO2 photocatalytic system, J. Environ. Manage., 2004, 70, 43–47. 4 T. Maier and H. C. Korting, Sunscreens – which and what for?, Skin Pharmacol. Physiol., 2005, 18, 253–262. 5 T. Kaida, K. Kobayashi, M. Adachi and F. Suzuki, Optical characteristics of titanium oxide interference lm and the lm laminated with oxides and their applications for cosmetics, J. Cosmet. Sci., 2004, 55, 219–220. 6 N. S. Allen, M. Edge, J. Verran, J. Stratton, J. Maltby and C. Bygott, Photocatalytic titania based surfaces: environmental benets, Polym. Degrad. Stab., 2008, 93, 1632–1646. 7 R. J. Miller, S. Bennett, A. A. Keller, S. Pease and H. S. Lenihan, TiO2 nanoparticles are phototoxic to marine phytoplankton, PLoS One, 2012, 7, e30321. 8 C. Ma, J. L. Marlowe and A. Puga, The aryl hydrocarbon receptor at the crossroads of multiple signaling pathways, EXS, 2009, 99, 231–257. 9 J. L. Zhang, M. Wages, S. B. Cox, J. D. Maul, Y. J. Li, M. Barnes, et al., Effect of titanium dioxide nanomaterials
Analyst, 2014, 139, 964–972 | 971
View Article Online
Analyst
10
Published on 23 December 2013. Downloaded by Aston University on 23/08/2014 09:28:10.
11
12
13
14
15
16
17
18
and ultraviolet light coexposure on African clawed frogs (Xenopus laevis), Environ. Toxicol. Chem., 2012, 31, 176–183. O. Bar-Ilan, C. C. Chuang, D. J. Schwahn, S. Yang, S. Joshi, J. A. Pedersen, et al., TiO2 nanoparticle exposure and illumination during zebrash development: mortality at parts per billion concentrations, Environ. Sci. Technol., 2013, 47, 4726–4733. O. Bar-Ilan, K. M. Louis, S. P. Yang, J. A. Pedersen, R. J. Hamers, R. E. Peterson, et al., Titanium dioxide nanoparticles produce phototoxicity in the developing zebrash, Nanotoxicology, 2012, 6, 670–679. S. P. Yang, O. Bar-Ilan, R. E. Peterson, W. Heideman, R. J. Hamers and J. A. Pedersen, Inuence of humic acid on titanium dioxide nanoparticle toxicity to developing zebrash, Environ. Sci. Technol., 2013, 47, 4718–4725. H. Lin, C. P. Huang, W. Li, C. Ni, S. I. Shah and Y. H. Tseng, Size dependency of nanocrystalline TiO2 on its optical property and photocatalytic reactivity exemplied by 2chlorophenol, Appl. Catal., B, 2006, 68, 1–11. O. Carp, C. L. Huisman and A. Reller, Photoinduced reactivity of titanium dioxide, Prog. Solid State Chem., 2004, 32, 33–177. S. G. Kumar and L. G. Devi, Review on modied TiO2 photocatalysis under UV/visible light: selected results and related mechanisms on interfacial charge carrier transfer dynamics, J. Phys. Chem. A, 2011, 115, 13211–13241. J. Jiang, G. Oberdorster, A. Elder, R. Gelein, P. Mercer and P. Biswas, Does nanoparticle activity depend upon size and crystal phase?, Nanotoxicology, 2008, 2, 33–42. D. W. Xiong, T. Fang, L. P. Yu, X. F. Sima and W. T. Zhu, Effects of nano-scale TiO2, ZnO and their bulk counterparts on zebrash: acute toxicity, oxidative stress and oxidative damage, Sci. Total Environ., 2011, 409, 1444–1452. J. Wang, G. Zhou, C. Chen, H. Yu, T. Wang, Y. Ma, et al., Acute toxicity and biodistribution of different sized titanium dioxide particles in mice aer oral administration, Toxicol. Lett., 2007, 168, 176–185.
972 | Analyst, 2014, 139, 964–972
Paper
19 A. J. Hill, H. Teraoka, W. Heideman and R. E. Peterson, Zebrash as a model vertebrate for investigating chemical toxicity, Toxicol. Sci., 2005, 86, 6–19. 20 A. Nasevicius and S. C. Ekker, The zebrash as a novel system for functional genomics and therapeutic development applications, Curr. Opin. Mol. Ther., 2001, 3, 224–228. 21 H. Teraoka, W. Dong and T. Hiraga, Zebrash as a novel experimental model for developmental toxicology, Congenital Anomalies, 2003, 43, 123–132. 22 T. Kotsokechagia, F. Cellesi, A. Thomas, M. Niederberger and N. Tirelli, Preparation of ligand-free TiO2 (anatase) nanoparticles through a nonaqueous process and their surface functionalization, Langmuir, 2008, 24, 6988–6997. 23 I. A. Mudunkotuwa and V. H. Grassian, Citric acid adsorption on TiO2 nanoparticles in aqueous suspensions at acidic and circumneutral pH: surface coverage, surface speciation, and its impact on nanoparticle– nanoparticle interactions, J. Am. Chem. Soc., 2010, 132, 14986–14994. 24 H. J. Johnston, G. R. Hutchison, F. M. Christensen, S. Peters, S. Hankin and V. Stone, Identication of the mechanisms that drive the toxicity of TiO2 particulates: the contribution of physicochemical characteristics, Part. Fibre Toxicol., 2009, 6, 33. 25 G. Oberdorster, E. Oberdorster and J. Oberdorster, Nanotoxicology: an emerging discipline evolving from studies of ultrane particles, Environ. Health Perspect., 2005, 113, 823–839. 26 P. A. Dennery, Effects of oxidative stress on embryonic development, Birth Defects Res., Part C, 2007, 81, 155–162. 27 J. M. Hansen, Oxidative stress as a mechanism of teratogenesis, Birth Defects Res., Part C, 2006, 78, 293–307. 28 V. K. Sharma, Aggregation and toxicity of titanium dioxide nanoparticles in aquatic environment – a review, J. Environ. Sci. Health, Part A: Toxic/Hazard. Subst. Environ. Eng., 2009, 44, 1485–1495.
This journal is © The Royal Society of Chemistry 2014
