Developmental and cartilaginous effects of protein-coated SiO₂ nanoparticle corona complexes on zebrafish larvae

Abstract

Protein-coated SiO₂ nanoparticle (SNP) corona complexes of different sizes can be selected for various applications, and whether such materials have different degrees of toxicity is becoming a critical question. As the substance that is richest in proteins, bovine serum albumin (BSA-V), is chosen to coat the two types of SNP to obtain SNP corona complexes with different diameters and surface charges, as well as alleviating agglomeration. Toxicity tests were performed in vivo using zebrafish as a platform. Hatch rates, mortalities, malformations and alcian blue-stained cartilaginous deformities were recorded using a stereomicroscope. The results verified that coating SNPs with BSA-V is a suitable method to control diameters and surface charges. The hatch rate results illustrated the concentration-dependent effect of the complex. The mortality results were increased from 54 hour post fertilization (hpf) to 120 hpf, explained by the change in LC₅₀ values. The malformation incidence, average toxicity score of the malformations, and the vertebral column of the zebrafish further supported the idea that 50 nm NP corona complexes have more toxicity. The developmental stages of the zebrafish at 54 hpf is a novel and important time point for evaluating the toxicity of the complex. The results showed that the same protein-coated NP corona complexes had different surface characters and toxicity extents. The 50 nm NP corona complexes are more toxic than the 15 nm NP corona complexes during developmental stages. Meanwhile the cartilaginous toxicity of the vertebral columns provided further explanation.

---

Introduction

Mesoporous SiO₂ has the advantages of a high dispersibility and absorptivity, and also has a thickening property. These advantages have been widely applied in pharmaceutical preparations; for example, a small amount of SiO₂ added into a preparation can alter its dissolution rate, liquidity, dispersion, and antistatic properties.¹ SNPs can be used as drug carriers, which can help drugs pass through biological barriers, improve drug absorption, increase the degree of utilization, and enhance the targeting properties.² In the field of medical engineering, SNPs may be injected into any part of the body to assess and diagnose pathological changes and subsequently provide proper treatment. In other medical applications, SNPs form materials for nano-sized, biocompatible, artificial organs, nano-sized biosensors, and micro-scale intelligent medical equipment, for example.^3–5 In summary, SNPs have been widely used in the field of medical applications. In a study of toxicity for human endothelial cells exposed to SNPs of various diameters, D. Napierska demonstrated that the diameter of the NPs has a dose-dependent inhibitory action on the growth of human endothelial cells.⁶ The surface charge is also an important factor that influences NPs.⁷ Different diameters and surface charges may be selected for various medical uses. Whether the different diameters and surface charges have different degrees of toxicity and biological distributions is an important factor to consider in their use for medical materials. Here, we chose NPs which have two different diameters and coated them with the same amount of BSA-V, to obtain NP corona complexes with different diameters and surface charges. This additionally alleviated agglomeration, according to our previous study.⁸ We then assessed their toxicity in vivo using zebrafish as a platform.

The zebrafish is an appropriate in vivo model in pharmacology and genetics research.^9,10 Recently, zebrafish have been widely used to study the developmental toxicity of NPs.^11–13 The advantages of using zebrafish over other mammals for toxicity studies include their high genomic similarity with humans, the similarity of their tissue structures to those of mammals; the cost-effectiveness of obtaining large numbers of embryos and housing for experiments and the embryos' transparency during early developmental stages.^14–17 The effects caused by exposure to SNPs are assessed in terms of overall fitness, including the hatch rate, mortality and malformation. The “hatched” phenotype indicates little toxicity due to exposure to NPs. The “unhatched” phenotype signifies interference in the embryos' developmental stage. The “dead” phenotype indicates that many toxic effects occur.^18,19 LC₅₀, the median lethal concentration for embryos exposed to NPs in the experiments, is an important mortality-related parameter and can be used to compare the degree of toxicity among different diameters and concentrations of SNP. The toxicity score system has been shown to be an effective method for measuring the degree of toxicity for zebrafish and for recording malformations.²⁰ Alcian blue is a dye that stains the chondrocyte-associated extracellular matrix. The alcian blue stain is a distinct marker of the skeletal pattern, and subtle teratogenic effects introduced by the NPs on the zebrafish cartilages can be detected.²¹ The alcian blue-stained cartilaginous deformities could further explain the influence of NPs with different diameters.

Materials and methods

Fish husbandry and embryo collection

Adult zebrafish of the wild-type strain (AB) were maintained at 28 °C with a 14 h light/10 h dark photoperiod in a recirculating system with standard tank water according to standard zebrafish breeding protocols.²² All experimental protocols and procedures were performed in accordance with the NIH Guide for Care and Use of Laboratory Animals (no. 8023, revised 1996). The zebrafish embryos exposed to the NPs were obtained with a sex ratio of 1 : 1 from adults spawning in the tanks overnight. Embryos were collected within 2 hpf. The normal embryos were inspected and staged for the subsequent experiments under a stereomicroscope (OLYMPUS SZX 10, Japan) according to Kimmel's descriptions.²³

Modification and characterization of the SNPs

SiO₂ NPs (15 nm and 50 nm) were purchased from Xuan Cheng Jing Rui New Material Co., Ltd. China. We coated the SiO₂ NP surfaces with 4% bovine serum albumin (BSA-V) (Fraction-V; Gemini Bioproducts, USA) according to the procedure described by George et al. to reduce agglomeration,²⁴ as well as to form the protein layer and control the diameters and surface charges. We then prepared 5 mg mL⁻¹ SiO₂ NP suspensions in double-distilled water and sonicated them for 15 min (with a water bath sonicator, 80 W) to obtain SiO₂ NP stock suspensions. A total of 50 mL NP stock suspension was transferred to a beaker, mixed with 50 mL BSA-V (40 mg mL⁻¹) suspension and equilibrated for 30 min on a magnetic stirrer at room temperature. The SNPs were stored at 4 °C. When pipetting the SNPs into the 96-well plates, the NPs were equilibrated for 30 min at room temperature, diluted with standard tank water to different concentrations and equilibrated for 10 min before being placed into the 96-well plates. The diameters and surface charges of the SNPs corona complex (15 nm and 50 nm) coated with BSA-V and non-coated SNPs (15 nm and 50 nm) in solution were determined by dynamic light scattering (DLS).

Zebrafish exposure to the SNP corona complexes

At 4 hpf, zebrafish embryos were selected by visual assessment, distributed into 96-well plates with one embryo per well, and given a certain amount of solution (Scheme 1). To compare the toxicity for zebrafish exposed to NPs of different diameters and concentrations, five replicate treatments were conducted. The hatch rate, mortality and malformations were recorded using a stereomicroscope (OLYMPUS SZX 10, Japan) at specific times during each observation. The hatch rate of the zebrafish was assessed at 48, 54, and 72 hpf. The mortality of the zebrafish was assessed at 54, 72, and 120 hpf. The toxicity scores and malformation types were assessed at 96 and 120 hpf (Table S1†). The definition of the toxicity score, ranging from 0 to 4, is based on the scoring spectrum of Bar-Ilan et al.: 0, normal development; 1, one or two minor malformations; 2, one moderate or three to four minor malformations; 3, more than one severe or four minor malformations; and 4, embryonic lethality. The zebrafish were euthanatized with tricaine, and were observed and recorded using a stereomicroscope (OLYMPUS SZX 10, Japan).

Scheme 1 Zebrafish exposed to NPs for developmental toxicity analysis. At 4 hpf zebrafish embryos were distributed into 96-well plates with one embryo per well in 200 μL solution as shown in the Scheme. In the control groups, renewal with standard tank water was performed every 24 h up to 120 hpf. In the experimental groups, the SiO₂ NP corona complex solution was renewed every 24 h up to 72 hpf, after which it was replaced with standard tank water every 24 h from 72 to 120 hpf. The hatch rate, mortality and malformations were recorded using a stereomicroscope (OLYMPUS SZX 10, Japan) at specific times during each observation.

Whole-mount skeletal staining

Zebrafish treated with NPs at different developmental stages were fixed with 4% paraformaldehyde (pH 7.0, in phosphate-buffered saline) overnight at 4 °C. The cartilage color staining was conducted as described in Walker et al.²⁵ The fixed larvae were dehydrated in ethanol for 24 h and were transferred to a solution of 0.1% alcian blue solution dissolved in 70% ethanol/1% hydrochloric acid for 24 h. The specimens were then rinsed in a sodium borate solution for 9 to 12 h. After rinsing, the specimens were placed in a solution of 1% KOH/3% H₂O₂ for 30 min to blanch all pigment. Soon thereafter, the tissue of the zebrafish was moderately digested in a solution of 0.05% trypsin dissolved in saturated sodium tetraborate for 1 h. The specimens were photographed using a fluorescence stereomicroscope (OLYMPUS SZX 10, Japan).

Data analysis

Each treatment was replicated five times, and the results are reported as the average of the five parallel experiments. All line charts were created in Microsoft Excel. The P values were calculated using Origin 7.5 (OriginLab, Northampton, MA, USA). A one-way ANOVA was performed to calculate the statistical significance, in order to compare each exposure group to the control group. A P value for the exposure group of less than 0.05 was considered to be statistically significant. The data are presented as the mean ± the standard error (SE).

Results and discussion

Dynamic light scattering results

The same amount of BSA-V was used to coat the two types of SNP; this was a stabilizing agent used by George et al.²⁴ and in our previous study.⁸ The DLS results showed that it effectively controlled the degree of agglomeration, diameters and surface charges (Fig. 1). Without the BSA-V coating, both types of NP, which were negatively charged, agglomerated as shown in Fig. 1 and Table 1. After coating, the negatively charged SiO₂ NPs and the positively charged BSA-V self-assembled and formed the SNP corona complexes. Coating with BSA-V can optimize the size distribution for both types of SNP. We then explored the toxicity of the SNP corona complexes on zebrafish larvae.

Fig. 1 DLS results for NPs. (a) 15 nm SiO₂ NPs (b) 50 nm SiO₂ NPs (c) 15 nm SiO₂ NP corona complex and (d) 50 nm SiO₂ NP corona complex.

Table 1 Statistical results for diameters and surface charges based on the DLS results for the NP corona complexes

	Dn^a	Zeta-potential (mV)
a Hydrodynamic diameters determined by DLS.
15 nm SiO2	155.05, 669.04	−18.05
50 nm SiO2	140.83, 808.60	−17.99
15 nm SiO2–BSA	508.97	−13.83
50 nm SiO2–BSA	894.83	+1.62

---

Hatch rate

A large range of concentrations (0.05–1000 μg mL⁻¹) were tested in pre-experiments, in order to determine the toxicity range of the SNP corona complexes. We eventually used concentrations of 100–500 μg mL⁻¹ in the following experiments for intensive illustration of developmental toxicity. Naked SNPs of a concentration of 25.6 μg mL⁻¹ had a significant impact on the embryonic development,²⁶ while SNPs coated with BSA-V are safe in doses of up to 200 μg mL⁻¹ based on our present study, because the size and surface charge significantly influence the toxicity of the SNPs.²⁷

The hatch rate of the zebrafish embryos exposed to both complexes at 48 hpf was less than 20% (Fig. 2a). The hatch rate at 54 hpf could reflect the influence of the NPs, which is consistent with the trend at 72 hpf (Fig. 2b and c). Both NP corona complexes affected the hatch rate. As the concentration of the complex increased, the hatch rate decreased. There were significant differences compared to the control for both the 15 nm and 50 nm NP corona complexes when exposure concentrations were 300 μg mL⁻¹ and above. The 15 nm SNPs had more effect on the hatch rate at concentrations of less than 200 μg mL⁻¹, while the 50 nm SNPs played a more important role in the hatch rate than the 15 nm SNPs at 300 μg mL⁻¹. These differences decreased when the concentration reached 400 μg mL⁻¹ and above. A summary of the hatch rate data is shown in Fig. 2d.

Fig. 2 Hatch rates of zebrafish embryos exposed to SiO₂ NPs at different concentrations and with different diameters at (a) 48 hpf, (b) 54 hpf and (c) 72 hpf. (d) Comprehensive 3D representation. A one-way ANOVA was conducted, and the asterisk (*) indicates significant differences from the control at P < 0.05. The values represent the mean ± the standard error of five replicates.

The developmental stages of the zebrafish at 54 hpf is an novel and important time point for evaluating the toxicity of the SNP corona complexes, because the hatch rate and mortality at 54 hpf are in good agreement with those at 72 hpf. From the view of environmental protection, the 54 hpf evaluation time is time-saving and material-saving. The hatch rate results for zebrafish at 54 hpf illustrated the dose-dependent effects of the SNP corona complexes. There were significant decreases compared to the control group when the concentration was 300 μg mL⁻¹ and above.

Mortality

The mortality of the zebrafish was analyzed at 54, 72 and 120 hpf, respectively. The mortality of the zebrafish increased as the concentration increased (Fig. 3). The results curves cross at each time point from 300 to 400 μg mL⁻¹. In order to further evaluate and compare the extent of the mortality, the LC₅₀ values of the SNPs were calculated, to reflect the degree of toxicity based on the mortality results. A lower LC₅₀ corresponded with a higher degree of toxicity. The LC₅₀ decreased between 54 and 120 hpf due to the accumulative effect of the toxicity. The toxicity of the 50 nm SNP corona complexes on the zebrafish was higher than that of the 15 nm SNP corona complexes, which indicated the greater toxicity of the 50 nm SNP corona complexes.

Fig. 3 Mortality of zebrafish larvae exposed to SiO₂ NP corona complexes at different concentrations and with different diameters at (a) 54 hpf, (b) 72 hpf and (c) 120 hpf. (d) Comprehensive 3D representation. A one-way ANOVA was performed, and an asterisk (*) indicates significant differences from the control at P < 0.05. The values represent the mean ± the standard error of five replicates.

Malformations, malformation types and cartilaginous malformations

Toxicity for the zebrafish was further assessed at 96 and 120 hpf by studying malformation occurrence. Images were taken to assess the malformation degree of the zebrafish at each concentration of SNP corona complex. The zebrafish appeared relatively normal after exposure to standard tank water or 100 μg mL⁻¹ SNP corona complex. Zebrafish which exhibited one or more type of malformation were regarded as malformed. The control groups had non-zero malformation values due to the normal death of the zebrafish. The 100 μg mL⁻¹ SNPs corona complex groups had malformation values as high as 60% due to the statistical method we used.²⁰ Fig. 4 gives an outlook of the incidence of malformation rather than the extent. There were few pericardial edema at 100 μg mL⁻¹ SNPs. As the concentration increased, the incidence of malformation increased. With SNP corona complexes at concentrations greater than 300 μg mL⁻¹, the malformation incidence was greater than 95%. Zebrafish exposed to SNP corona complexes for 120 hpf had a higher degree of malformation than those exposed for 96 hpf. The malformation rates for zebrafish exposed to the two diameters of SNP corona complex were approximately the same at the same concentrations and the same time points. Additional zebrafish are shown in Fig. S1† to provide specific examples of the observed malformations.

Fig. 4 Malformation of zebrafish larvae exposed to SiO₂ NP corona complexes. (a) Micrographs of the malformations of zebrafish exposed to different concentrations and different diameters of SiO₂ NPs (96 hpf and 120 hpf). Malformation incidence for zebrafish exposed to different concentrations and different diameters of SiO₂ NPs for (b) 96 hpf and (c) 120 hpf. The zebrafish were euthanatized with tricaine, and were observed and recorded using a stereomicroscope (Japan). A one-way ANOVA was performed, and an asterisk (*) indicates significant differences from the control at P < 0.05. The values represent the mean ± the standard error of five replicates.

Malformation toxicity scores at 96 hpf and 120 hpf

To quantify the effects of the SNP corona complexes on the zebrafish, we evaluated the zebrafish development according to the scoring spectrum described by Bar-Ilan et al.¹¹ The morphological conditions of the zebrafish embryos were scored and the malformation types were distinguished.

Zebrafish exposed to SNP corona complexes for 120 hpf had a higher average toxicity score than those exposed for 96 hpf, and the zebrafish which were exposed to the 50 nm NPs had a higher average toxicity score than those exposed to the 15 nm NPs. The malformation types for the zebrafish exposed to 100 and 200 μg mL⁻¹ NPs were mainly PE, YM, and BS (please refer to the caption of Fig. 5 for explanations of the abbreviations). All of the aforementioned malformation types appeared in the zebrafish exposed to 300–500 μg mL⁻¹ SNP corona complex (Fig. 5d–g). Adding the incidence of BS, TM and SG up, we may infer that the development of cartilage has a strong influence on malformation occurrence. The incidence of different malformation types from exposure to SNPs with different diameters is shown in Fig. S2.† SNP corona complexes with a large diameter have a more significant influence on the development of the zebrafish compared with smaller sized SNPs at concentrations below 300 μg mL⁻¹. Both have strong effects at concentrations above 300 μg mL⁻¹.

Fig. 5 Micrographs, average toxicity scores and incidence of malformation types for the zebrafish, at 96 hpf and 120 hpf exposure to SiO₂ NP corona complexes at different concentrations and different diameters. (a) Representative micrographs of zebrafish at 96 hpf and 120 hpf exposed to standard tank water or 100, 200, 300, 400, or 500 μg mL⁻¹ NP corona complexes. The zebrafish were euthanatized with tricaine and were observed and recorded using a stereomicroscope (Japan). The average toxicity scores for zebrafish exposed to standard tank water or 100, 200, 300, 400, or 500 μg mL⁻¹ SiO₂ NP corona complexes are shown for (b) 96 hpf and (c) 120 hpf. The incidence of different malformation types from exposure to standard tank water or 100, 200, 300, 400, or 500 μg mL⁻¹ SiO₂ NPs is shown for: (d) 15 nm SiO₂ NP corona complexes at 96 hpf; (e) 50 nm SiO₂ NP corona complexes at 96 hpf; (f) 15 nm SiO₂ NP corona complexes at 120 hpf and (g) 50 nm SiO₂ NP corona complexes at 120 hpf. The malformation types recorded in the experiments included PE (pericardial edema), YM (yolk sac malformation), BS (bent spine), SH (small head), SE (small eye), TM (tail malformation) and SG (stunted growth). A one-way ANOVA was performed, and an asterisk (*) indicates significant difference from the control at P < 0.05. The values represent the mean ± the standard error of five replicates.

Cartilaginous malformation

Considering that the development of the vertebral column has a strong influence on the hatch rate, the cartilaginous toxicity for vertebral column development was studied next, using whole-mount skeletal staining. Our research on SNP corona complexes presented the different surface characters and demonstrated the different toxicity levels in the development of the zebrafish embryos. In this present study, the mortality and average toxicity scores indicated that the 50 nm SNP corona complexes caused more serious teratogenic effects than did the 15 nm SNP corona complexes at 96 or 120 hpf, and from 200 to 400 μg mL⁻¹. As we inferred that cartilaginous malformation was the main reason for toxicity, the specification and morphogenesis of the zebrafish larval head and spinal skeleton were prone to be affected upon exposure to different NPs. To detect the effects of SNP corona complexes with different surface characters on cartilage development, we conducted whole-mount skeletal staining using alcian blue. Generally, the results of alcian blue staining showed that the degree of morphological malformation of the whole zebrafish skeleton increased as the dose of the complex increased, which was in line with the other results in our study. Additionally, to further determine the toxicity for specific bone structures, we specifically studied the jaw of the zebrafish larvae, which is abundant in various cartilages and has a complex bone structure; the numbers of abnormal jaw cartilages were counted.

Cartilaginous toxicity of SNP corona complexes with different surface characteristics in the development of the vertebral column of zebrafish larvae at 96 hpf and 120 hpf

The results for the alcian blue-stained zebrafish at 96 hpf and 120 hpf are shown in Fig. 6. The NP dose ranged from 200 to 400 μg mL⁻¹, and the time point was from 96 hpf to 120 hpf. As the dose (200, 300, and 400 μg mL⁻¹) and time (96 hpf to 120 hpf) increased, the frequency of bent spines in the skeletons increased. The effect of time and NP corona complex dose on cartilaginous development of the vertebral column of the larvae, corresponding to the results of the BS, TM, and SG malformation types, is shown in Fig. 4 and 5. The results proved our previous hypothesis. Additional stained zebrafish are presented in Fig. S3.†

Fig. 6 Micrographs of the cartilaginous toxicity for vertebral column development. 96 hpf and 120 hpf zebrafish exposed to different concentrations (200, 300, and 400 μg mL⁻¹) and different diameters (15 nm and 50 nm) of SiO₂ NP corona complexes. Scale bars = 500 μm.

Cartilaginous toxicity of different surface characteristics of SNP corona complexes in the development of the head and jaw skeleton of zebrafish larvae

As depicted in the schematic drawings of the alcian blue-stained zebrafish larvae in Fig. 7a and b, the head and jaw of zebrafish larvae include several cartilaginous structures, listed in the figure caption.^28,29 By analyzing SNP-dosed zebrafish, we found that cartilaginous malformations were prone to occur among certain jaw cartilages, as shown in Fig. 7c and d.

Fig. 7 Jaw and head cartilaginous toxicity for zebrafish larvae. Jaw and head cartilaginous toxicity for (a) 96 hpf and (b) 120 hpf zebrafish exposed to different concentrations (200, 300 and 400 μg mL⁻¹) and different diameters (15 and 50 nm) of SiO₂ corona complex. The abbreviations represent different cartilaginous elements in the jaw and head of zebrafish. A and B: schematic drawings of the jaw and head skeleton. C and D: a wild-type zebrafish stained at 96 hpf. E and F: a larva exposed to 15 nm SiO₂ corona complex at 200 μg mL⁻¹. G and H: a larva exposed to 50 nm SiO₂ corona complex at 200 μg mL⁻¹. I and J: a larva exposed to 15 nm SiO₂ corona complex at 300 μg mL⁻¹. K and L: a larva exposed to 50 nm SiO₂ corona complex at 300 μg mL⁻¹. M and N: a larva exposed to 15 nm SiO₂ corona complex at 400 μg mL⁻¹. O and P: a larva exposed to 50 nm SiO₂ corona complex at 400 μg mL⁻¹. Black arrows indicate the abnormal cartilaginous elements in the jaw and head. (c) and (d) Histograms showing the incidence of the effects depicted in (a) and (b): yellow, 200 μg mL⁻¹; blue, 300 μg mL⁻¹; red, 400 μg mL⁻¹; blank column, 15 nm; hatched, 50 nm. M: Malformation at the Meckel's cartilage; CH: malformation at the ceratohyal; E: malformation at the ethmoid plate; HS: malformation at the hyosymplectic; BH: malformation at the basihyal; CB: malformation at the ceratobranchia; OA: malformation at the occipital arch.

The NP doses ranged from 200 to 400 μg mL⁻¹, and the time frame was from 96 to 120 hpf. We found that both the 15 and 50 nm SNP corona complexes affected the chondrogenesis in the head and jaw.³⁰ SNP corona complexes began to cause teratogenic effects on the jaw and head cartilage at a concentration of 300 μg mL⁻¹. The SNP corona complexes shortened the length of the cartilage and caused a distinct malformation of cartilage at a concentration of 400 μg mL⁻¹ (Fig. 7c and d).

Cartilaginous toxicity of SNPs in the development of the vertebral column, and head and jaw skeleton of the zebrafish larvae has further explained the results of the hatch rate, mortality and malformation. The SNPs primarily affect the skeleton, especially vertebral column development.

Conclusions

In conclusion, coating with BSA-V was a suitable method to make SNPs into corona complexes with different diameters, surface charges and aggregation states, and these were safe at a dose nearly 9 times higher than naked SNPs. The developmental toxicity of SNP corona complexes was assessed with SNPs of different diameters and surface charges at different concentrations and times using zebrafish larvae. This laid a foundation for mesoporous SNP corona complexes as multifunctional drug carriers in the future work.

Based on the hatch rate, mortality, degree and extent of malformations, and cartilaginous malformations, these results showed the time-dependent, concentration-dependent and diameter-dependent effects of SNP corona complexes.

References

1. M. Q. Zhu, J. J. Han and A. D. Q. Li, J. Nanosci. Nanotechnol., 2007, 7, 2343–2348.
2. B. M. Cullum, G. D. Griffin, G. H. Miller and V. D. Tuan, Anal. Biochem., 2000, 277, 25–32.
3. Q. Huo, J. Liu, L. Q. Wang, Y. Jiang, T. N. Lambert and E. Fang, J. Am. Chem. Soc., 2007, 128, 6447–6453.
4. M. Arruebo, M. Galán, N. Navascués, C. Téllez, C. Marquina, M. Ricardo and J. Santamaría, Chem. Mater., 2006, 18, 1911–1919.
5. J. Allouche, M. Boissiere, C. Helary, J. Livage and T. Coradin, J. Mater. Chem., 2006, 16, 3120–3125.
6. D. Napierska, C. J. T. Leen, V. Rabolli, D. Lison, L. Gonzalez, M. Kirsch-Volders, J. A. Martens and P. H. Hoet, Small, 2009, 5, 846–853.
7. G. Oberdorster, Z. Sharp, V. Atudorei, A. Elder, R. Gelein, A. Lunts, W. Kreyling and C. Cox, J. Toxicol. Environ. Health, Part A, 2002, 65, 1531–1543.
8. J. Y. Xue, X. Li, M. Z. Sun, Y. P. Wang, M. Wu, C. Y. Zhang, Y. N. Wang, B. Liu, Y. S. Zhang, X. Zhao and X. Z. Feng, Small, 2013, 9, 3162–3168.
9. L. I. Zon and R. T. Peterson, Nat. Rev. Drug Discovery, 2005, 4, 35–44.
10. G. N. Wheeler and A. W. Brandli, Dev. Dyn., 2009, 238, 1287–1308.
11. O. Bar-Ilan, R. M. Albrecht, V. E. Fako and D. Y. Furgeson, Small, 2009, 5, 1897–1910.
12. P. V. Asharani, Y. Wu, Z. Gong and S. Valiyaveettil, Nanotechnology, 2008, 19, 255102.
13. X. Zhu, J. Wang, X. Zhang, Y. Chang and Y. Chen, Nanotechnology, 2009, 20, 195103.
14. X. Y. Xu, X. Y. Xu, X. Huang, W. Xia and S. Xia, J. Biomol. Screening, 2010, 15, 1152–1159.
15. V. E. Fako and D. Y. Furgeson, Adv. Drug Delivery Rev., 2009, 61, 478–486.
16. A. L. Rubinstein, Expert Opin. Drug Metab. Toxicol., 2006, 2, 231–240.
17. W. Zhang, K. F. Lin, Y. N. Miao, Q. Dong, C. Huang, H. Wang, M. Guo and X. Cui, J. Hazard. Mater., 2012, 213–214, 413–420.
18. D. R. Love, F. B. Pichler, A. Dodd, B. R. Copp and D. R. Greenwood, Curr. Opin. Biotechnol., 2004, 15, 564–571.
19. R. Liu, S. J. Lin, R. Rallo, Y. Zhao, R. Damoiseaux, T. Xia, S. Lin, A. Nel and Y. Cohen, PLoS One, 2012, 7, e35014.
20. Y. Liu, B. Liu, D. Feng, C. Gao, M. Wu, N. He, X. Yang, L. Li and X. Z. Feng, Integr. Biol., 2012, 4, 285–291.
21. T. F. Schilling, T. Piotrowski, H. Grandel, M. Brand, C. P. Heisenberg, Y. J. Jiang, D. Beuchle, M. Hammerschmidt, D. A. Kane, M. C. Mullins, F. J. van Eeden, R. N. Kelsh, M. Furutani-Seik, M. Granato, P. Haffter, J. Odenthal, R. M. Warga, T. Trowe and C. Nüsslein-Volhard, Development, 1996, 123, 329–344.
22. M. Westerfield, in The zebrafish book, University of Oregon Press, Eugene, OR, 3rd edn, 1995, pp. 267–272.
23. C. B. Kimmel, W. W. Ballard, S. R. Kimmel, B. Ullman and T. F. Schilling, Dev. Dyn., 1995, 203, 253–310.
24. S. George, T. Xia, R. Rallo, Y. Zhao, Z. Ji, S. Lin, X. Wang, H. Zhang, B. France, D. Schoenfeld, R. Damoiseaux, R. Liu, S. Lin, K. A. Bradley, Y. Cohen and A. E. Nel, ACS Nano, 2011, 5, 1805–1817.
25. M. B. Walker and C. B. Kimmel, Biotech. Histochem., 2007, 82, 23–28.
26. K. Fent, C. J. Weisbrod, A. Wirth-Heller and U. Pieles, Aquat. Toxicol., 2010, 100, 218–228.
27. K. Greish, G. Thiagarajan, H. Herd, R. Price, H. Bauer, D. Hubbard, A. Burckle, S. Sadekar, T. Yu, A. Anwar, A. Ray and H. Ghandehari, Nanotoxicity, 2012, 6, 713–723.
28. C. B. Kimmel, C. T. Miller and C. B. Moen, Dev. Biol., 2001, 233, 239–257.
29. T. Piotrowski, T. F. Schilling, M. Brand, Y. J. Jiang, C. P. Heisenberg, D. Beuchle, H. Grandel, F. J. van Eeden, M. Furutani-Seiki, M. Granato, P. Haffter, M. Hammerschmidt, D. A. Kane, R. N. Kelsh, M. C. Mullins, J. Odenthal, R. M. Warga and C. Nüsslein-Volhard, Development, 1996, 123, 345–356.
30. H. Teraoka, W. Dong, S. Ogawa, S. Tsukiyama, Y. Okuhara, M. Niiyama, N. Ueno, R. E. Peterson and T. Hiraga, Toxicol. Sci., 2002, 65, 192–199.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3ra45667f

‡ These authors equally contributed to this work.

---