Molecular Control of TiO2-NPs Toxicity Formation at Predicted Environmental Relevant Concentrations by Mn-SODs Proteins

With growing concerns of the safety of nanotechnology, the in vivo toxicity of nanoparticles (NPs) at environmental relevant concentrations has drawn increasing attentions. We investigated the possible molecular mechanisms of titanium nanoparticles (Ti-NPs) in the induction of toxicity at predicted environmental relevant concentrations. In nematodes, small sizes (4 nm and 10 nm) of TiO2-NPs induced more severe toxicities than large sizes (60 nm and 90 nm) of TiO2-NPs on animals using lethality, growth, reproduction, locomotion behavior, intestinal autofluorescence, and reactive oxygen species (ROS) production as endpoints. Locomotion behaviors could be significantly decreased by exposure to 4-nm and 10-nm TiO2-NPs at concentration of 1 ng/L in nematodes. Among genes required for the control of oxidative stress, only the expression patterns of sod-2 and sod-3 genes encoding Mn-SODs in animals exposed to small sizes of TiO2-NPs were significantly different from those in animals exposed to large sizes of TiO2-NPs. sod-2 and sod-3 gene expressions were closely correlated with lethality, growth, reproduction, locomotion behavior, intestinal autofluorescence, and ROS production in TiO2-NPs-exposed animals. Ectopically expression of human and nematode Mn-SODs genes effectively prevented the induction of ROS production and the development of toxicity of TiO2-NPs. Therefore, the altered expression patterns of Mn-SODs may explain the toxicity formation for different sizes of TiO2-NPs at predicted environmental relevant concentrations. In addition, we demonstrated here a strategy to investigate the toxicological effects of exposure to NPs upon humans by generating transgenic strains in nematodes for specific human genes.


Introduction
The potential toxicity of materials at the ''nano'' scale is receiving increasing attentions [1][2]. Titanium dioxide nanoparticles (TiO 2 -NPs) are often used as pigments or additives for paints, paper, ceramics, plastics, foods, and other products [3][4]. Hence, TiO 2 -NPs come into close contact with humans. The adverse effects caused by exposure to TiO 2 -NPs in vivo seem to have become more important than those due to exposure in vitro. Acute exposure to TiO 2 -NPs has been shown to cause hepatic injury, nephrotoxicity, pathological changes in the kidney, myocardial damage, spleen lesions, and inflammation in the lung and liver in mice or rats [5][6][7][8][9][10]. After intraperitoneal injection of TiO 2 -NPs, mice showed signs of toxicity, including passive behavior, loss of appetite, tremors and lethargy [7]. It was reported that mice exposed to TiO 2 -NPs during gestation produced offspring that could display moderate neurobehavioral alterations [10]. After subchronic dermal exposure of TiO 2 -NPs, TiO 2 -NPs penetrated through the skin to reach different tissues and induced the formation of diverse pathological lesions in several major organs [11]. In mice, exposure to TiO 2 -NPs also induced fetal resorption, restricted the growth of fetuses, and altered the expression of genes related to development and function of the central nervous system in pregnant animals, indicating the potential fetotoxicity of TiO 2 -NPs [12][13]. Acute toxicity assay of TiO 2 -NPs using Daphnia magna demonstrated that mortality increased after exposure [14]. Long-term exposure to TiO 2 -NPs has been shown to disturb progression of the cell cycle and to duplicate genome segregation, leading to chromosomal instability and cell transformation [15]. Chronic exposure to TiO 2 -NPs resulted in the inhibition of growth, a decrease in the liver weight ratio, and histopathological changes in the gills in zebrafish [16].
With regard to the biological effects mediated by NPs, evidence was provided to support the notion that NPs-mediated cellular response was size-dependent, which may provide important insights on our understanding of nanotoxicity [17]. In mice exposed to subchronic concentrations of TiO 2 -NPs through the skin, significant decreases in body weight were observed after exposure to TiO 2 -NPs of diameter 10 nm, 25 nm and 21 nm (P25), whereas no obvious differences in body weight were found after exposure to TiO 2 -NPs of diameter 60 nm, and it is unclear the expression and order of 10 nm, 25 nm, and 21 nm (P25) TiO 2 -NPs [11]. In zebrafish, the acute toxicity of TiO 2 -NPs was greater than that of bulk TiO 2 [18]. In Caenorhabditis elegans, acute exposure to TiO 2 -NPs of diameter 25 nm increased mortality, whereas acute exposure to 100-nm TiO 2 -NPs elicited no obvious effects in nematodes [19]. TiO 2 -NPs of diameter 7 nm seemed to be more toxic than 20-nm TiO 2 -NPs in nematodes [20]. Nevertheless, the molecular mechanisms underpinning the formation of size-dependent toxicity for TiO 2 -NPs are largely unclear.
The adverse effects of NPs exposure include (at the very least) oxidative stress, mitochondrial perturbation, inflammation, protein denaturation, protein degradation, altered regulation of the cell cycle and DNA damage [2]. For the development of toxicity of TiO 2 -NPs, oxidative stress may be induced by deposited TiO 2 -NPs. This is because the assay on levels of superoxide dismutase (SOD) and malondialdehyde (MDA) in mice and TiO 2 -NPs enhanced the production of reactive oxygen species (ROS) in cultured fibroblast cells [11,15]. Oxidative stress was shown to develop in the brain, liver and gut tissues of rainbow trout or zebrafish as indicated by biomarkers of oxidative stress [18,21]. However, the molecular basis for oxidative stress in the induction of TiO 2 -NPs toxicity is largely unknown.
In the present study, we investigated the molecular basis of oxidative stress in inducing the differences in toxicity for TiO 2 -NPs of different sizes using C. elegans as the assay system. Our data suggested that Mn-SODs proteins may contribute to the formation of different toxicities for different sizes of TiO 2 -NPs at predicted environmental relevant concentrations.

Adverse effects of different sizes of TiO 2 -NPs on survival, growth and reproduction of nematodes
Previous study demonstrated that C. elegans can be employed to assess the toxicity of nanomaterials at predicted environmental relevant concentrations by exposing C. elegans from L1-larvae to adults [42]. The effects of exposure to TiO 2 -NPs at predicated environmental relevant concentrations upon nematodes are unclear. Hence, we first investigated the possible adverse effects of exposure to TiO 2 -NPs at predicated environmental relevant concentrations upon the survival, growth, and reproduction of nematodes by exposing C. elegans from L1-larvae to adults. Exposure to 60-nm and 90-nm TiO 2 -NPs at concentrations from 0.001 mg/L to 10 mg/L, and exposure to 4-nm and 10-nm TiO 2 -NPs at concentrations from 0.001 and 0.01 mg/L did not obviously influence the survival of nematodes (Fig. 1A). In contrast, exposure to 4-nm and 10-nm TiO 2 -NPs at concentrations from 0.1 mg/L to 10 mg/L significantly (p,0.01) increased the mortality of nematodes (Fig. 1A).
The endpoint of growth was relatively less sensitive than the endpoint of lethality when assessing the toxicity of exposure to TiO 2 -NPs at predicated environmental relevant concentrations. Exposure to all the examined sizes of TiO 2 -NPs at concentrations of 0.001-0.1 mg/L did not obviously influence the growth of nematodes as reflected by body length (Fig. 1B). Exposure to 60nm and 90-nm TiO 2 -NPs at concentrations of 1-10 mg/L also did not noticeably affect the growth of nematodes (Fig. 1B). However, the body lengths of nematodes exposed to 1 and 10 mg/L of 4-nm or 10-nm TiO 2 -NPs were significantly (p,0.01) decreased compared with those of control nematodes (Fig. 1B).
We further examined the possible effects of TiO 2 -NPs exposure at predicated environmental relevant concentrations on the reproduction of nematodes using brood size as the endpoint. Exposure to 0.001 mg/L of all the examined TiO 2 -NPs, as well as exposure to 60-nm and 90-nm TiO 2 -NPs at 0.01 and 0.1 mg/L, did not cause alterations in the brood size of nematodes (Fig. 1C). In contrast, exposure to 4-nm and 10-nm TiO 2 -NPs at 0.01 and 0.1 mg/L significantly (p,0.01) reduced the brood size of nematodes (Fig. 1C). Exposure to all the examined TiO 2 -NPs at 1 and 10 mg/L also significantly (p,0.01) decreased the brood size of nematodes (Fig. 1C). These data implied that reproduction was relatively more sensitive than lethality and growth if assessing the toxicity of TiO 2 -NPs in nematodes.

Adverse effects of different sizes of TiO 2 -NPs on locomotion behavior of nematodes
Locomotion behavior is a relatively sensitive endpoint if evaluating the toxicity of a specific toxicant in C. elegans [52]. Next, we investigated the effects of TiO 2 -NPs exposure at predicated environmentally relevant concentrations on locomotion behavior by evaluating the body bend, head thrash, and forward turn of nematodes. Exposure to 60-nm and 90-nm TiO 2 -NPs at 0.001 and 0.01 mg/L did not obviously influence the body bends and head thrashes of nematodes (Figs 1D and 1E). In contrast, exposure to 4-nm and 10-nm TiO 2 -NPs at 0.001 mg/L moderately but significantly (p,0.05) suppressed the body bends and head thrashes of nematodes, and exposure to 4-nm and 10-nm TiO 2 -NPs at 0.01 mg/L significantly (p,0.01) decreased the body bends and head thrashes of nematodes (Figs 1D and 1E). Moreover, exposure to all the examined TiO 2 -NPs at 0.1-10 mg/L significantly (p,0.01) reduced the frequencies of body bends and head thrashes of nematodes (Figs 1D and 1E). Different from the observations on body bend and head thrash, forward turn was less sensitive than body bend and head thrash if assessing the toxicity of TiO 2 -NPs in nematodes. Exposure to all the examined TiO 2 -NPs at 0.001-0.01 mg/L and exposure to 60-nm and 90-nm TiO 2 -NPs at 0.1 mg/L did not noticeably affect forward turns (Fig. 1F). Significantly (p,0.01) reduced forward turns were observed in nematodes exposed to 4-nm and 10-nm TiO 2 -NPs at 0.1-10 mg/L as well as in those exposed to 60-nm and 90-nm TiO 2 -NPs at 1-10 mg/L (Fig. 1F).

Effects of different sizes of TiO 2 -NPs on intestinal autofluorescence of nematodes
Intestinal autofluorescence is a valuable marker of damage in intestinal cells in nematodes [53]. Our data suggested that exposure to all the examined TiO 2 -NPs at 0.001-0.1 mg/L as well as 60-nm and 90-nm TiO 2 -NPs at 1 mg/L did not induce noticeable increases in intestinal autofluorescence compared with those in controls (Figs 1G and 1H). In contrast, exposure to 4-nm and 10-nm TiO 2 -NPs at 1-10 mg/L significantly (p,0.01) induced an increase in intestinal autofluorescence compared with controls (Figs 1G and 1H). In addition, exposure to 60-nm and 90-nm TiO 2 -NPs at 10 mg/L moderately but significantly (p,0.05) increased intestinal autofluorescence compared with those seen in controls (Figs 1G and 1H). These data implied that the possible obvious damage in intestinal cells was formed only in nematodes exposed to relatively high concentrations of TiO 2 -NPs.
Exposure to TiO 2 -NPs alters the expression patterns of genes encoding Mn-SODs In C. elegans, oxidative stress is controlled by key genes such as sod-1, sod-2, sod-3, sod-4, sod-5, ctl-1, ctl-2, ctl-3, clk-1, clk-2, isp-1, gas-1, and mev-1 [45][46][47][48][49][50][51]. To examine the possible molecular basis of oxidative stress in the induction of toxicity differences in nematodes exposed to different diameters of TiO 2 -NPs, we investigated the expression patterns of key genes controlling the oxidative stress in nematodes exposed to different sizes of TiO 2 -NPs at 10 mg/L. Among the 13 genes examined, only the expression patterns of sod-2 and sod-3 genes were noticeably altered (Fig. 2C). Expression levels of the sod-2 gene and sod-3 gene were significantly (p,0.01) increased after exposure to 4-nm, 10nm, 60-nm, or 90-nm TiO 2 -NPs at 10 mg/L compared with those observed in controls. The expression patterns of sod-2 and sod-3 genes encoding Mn-SODs in nematodes exposed to 4-nm and 10nm TiO 2 -NPs were also different from those in nematodes exposed to 60-nm and 90-nm TiO 2 -NPs (Fig. 2C). Moreover, with the increases of exposure concentrations of different diameters of TiO 2 -NPs, the expression levels of sod-2 or sod-3 genes increased gradually compared with those of controls (Figs 2D and 2E). In nematodes, expressions of sod-2 and sod-3 genes were significantly increased by exposure to 4-nm and 10-nm TiO 2 -NPs at 0.001-10 mg/L as well as exposure to 60-nm and 90-nm TiO 2 -NPs at 1-10 mg/L (Figs 2D and 2E). These data suggested that genes encoding Mn-SODs may account for the development of toxicity differences in nematodes exposed to different sizes of TiO 2 -NPs.
Ectopically expression of nematode sod-2 or sod-3 prevents the toxicity formation in nematodes exposed to TiO 2

-NPs
To confirm the important roles of sod-2 and sod-3 genes in regulation of the formation of toxicity in nematodes exposed to different diameters of TiO 2 -NPs, we investigated the effects of ectopically expression of sod-2 and sod-3 genes upon toxicity development in nematodes exposed to 4-nm TiO 2 -NPs at 10 mg/ L. Ectopically expression of sod-2 or sod-3 genes effectively prevented the increase in mortality, decrease in body length, reduction of brood size, decrease in locomotion behavior (as reflected by body bend, head thrash and forward turn), increase of intestinal autofluorescence, and induction of significant production of ROS formed in nematodes exposed to 4-nm TiO 2 -NPs at 10 mg/L (Fig. 3).
Ectopically expression of human SOD2 prevents the toxicity formation in nematodes exposed to TiO 2 -NPs In humans, the SOD2 gene encodes the Mn-SODs. SOD-2 and SOD-3 in nematodes are highly homologous with SOD2 in humans (Fig. S1). With the aid of the dpy-30 gene promoter to drive the human SOD2 gene to express in all the cells of nematodes [54], we found that ectopically expression of the human SOD2 gene also effectively suppressed the increase in mortality, decrease in body length, reduction of brood size, decrease in locomotion behavior (as reflected by body bend, head thrash and forward turn), increase of intestinal autofluorescence, and induction of significant production of ROS formed in nematodes exposed to 4nm TiO 2 -NPs at 10 mg/L (Fig. 4).

Discussion
Studies have shown that acute exposure to 50-nm TiO 2 -NPs at .100 mg/L causes a significant increase in mortality, and that the median lethal concentration (LC 50 ) of nematodes acutely exposed to 25-nm TiO 2 -NPs was 77 mg/L [19,37]. Acute exposure to TiO 2 -NPs (50 nm) at .47.9 mg/L resulted in a significant inhibition of growth, and a reduction of the reproduction of C. elegans [37]. In the present study, after exposure from L1-larvae to adulthood, a significant increase in intestinal autofluorescence was observed in nematodes exposed to 4-nm and 10-nm TiO 2 -NPs at the concentration of 1 mg/L. A significant increase in mortality, reduction in body length, and decrease of forward turns could be observed in nematodes exposed to 4-nm and 10-nm TiO 2 -NPs at the concentration of 0.1 mg/L. A significant reduction in brood size and induction of ROS production could be detected in nematodes exposed to 4-nm and 10-nm TiO 2 -NPs at the concentration of 0.01 mg/L. In particular, significant decreases in body bend and head thrash could be observed in nematodes exposed to 4-nm and 10-nm TiO 2 -NPs at the concentration of 0.001 mg/L (Figs 1-2). The limit of detection for most methods is not sufficiently low to detect environmentally relevant concentrations of engineered NPs in the range of ng/L to pg/L [55]. The predicted environmental concentrations in water for TiO 2 -NPs are 16 or 24.5 mg/L [56][57]. Therefore, using C. elegans as a bioindicator, we can effectively assess the possible adverse effects of TiO 2 -NPs exposure at predicted environmental relevant concentrations on animals. Among the examined endpoints, body bend and head thrash are very sensitive, and can be used to evaluate the potential toxicity of TiO 2 -NPs exposure at the concentration of 1 ng/L. Similarly, after exposure from L1-larvae to adulthood, nano-CeO 2 exhibited adverse effects on nematodes at environmental relevant concentrations [35].
In the present study, we compared the possible differences in toxicities of two groups of TiO 2 -NPs in nematodes. The first group comprised 4-nm and 10-nm TiO 2 -NPs (small diameters of TiO 2 -NPs), and the second group comprised 60-nm and 90-nm TiO 2 -NPs (large diameters of TiO 2 -NPs). When we detected the toxicity of small-size TiO 2 -NPs at relatively low concentrations, the large size of TiO 2 -NPs did not usually exhibit toxicity in nematodes. For example, when we observed a significant increase in mortality, reduction in body length, and decrease of forward turns in nematodes exposed to 4-nm and 10-nm TiO 2 -NPs at 0.1 mg/L, no noticeable alterations in lethality, body length and forward turns were found in nematodes exposed to 60-nm and 90-nm TiO 2 -NPs at 0.1 mg/L, and the lethality and body length of nematodes exposed to 1-10 mg/L of 60-nm and 90-nm TiO 2 -NPs were also not significantly affected (Figs 1A, 1B, and 1F). Moreover, when we detected a significant reduction in brood size and induction of ROS production in nematodes exposed to 4-nm and 10-nm TiO 2 -NPs at 0.01-0.1 mg/L, no obvious alterations in brood size and ROS production were found in nematodes exposed to 60-nm and 90-nm TiO 2 -NPs at 0.01-0.1 mg/L (Fig 1C, Figs 2A and 2B). In particular, when we observed the significant decreases in body bend and head thrash in nematodes exposed to 4-nm and 10-nm TiO 2 -NPs at 0.001-0.01 mg/L, the body bends and head thrashes in nematodes exposed to 60-nm and 90-nm TiO 2 -NPs at 0.001-0.01 mg/L were not significantly affected (Fig. 1D and 1E). Therefore, toxicity differences can be formed for different sizes of TiO 2 -NPs in nematodes exposed to predicted environmental relevant concentrations. These data are consistent with observations on the differences in the acute toxicity of TiO 2 -NPs with different nano levels in C. elegans [19][20]. Our data are also in agreement with the observations on the toxicity differences of TiO 2 -NPs with different nano levels in zebrafish and mice [11,18].
ROS production was closely correlated with lethality, growth, reproduction, locomotion behavior and intestinal autofluorescence in nematodes exposed to different diameters of TiO 2 -NPs at predicted environmentally relevant concentrations (Table S1). Hence, we investigated the possible molecular basis of oxidative stress in regulating the formation of toxicity differences in nematodes exposed to different diameters of TiO 2 -NPs. Interestingly, exposure to TiO 2 -NPs only noticeably altered the expression patterns of sod-2 and sod-3 genes among the examined genes controlling the formation of oxidative stress in nematodes (Fig. 2C). In C. elegans, sod-2 and sod-3 genes encode Mn-SODs [48]. The expression levels of sod-2 and sod-3 also increased with the gradual increase in exposure concentrations of the examined TiO 2 -NPs (Figs 2D and 2E). Moreover, the linear regression analysis indicated that sod-2 gene expression and sod-3 expression were closely correlated with lethality, growth, reproduction, locomotion behavior, intestinal autofluorescence, and ROS production in nematodes exposed to all the examined different diameters of TiO 2 -NPs at predicted environmental relevant concentrations (Table S2). These data suggested that the specific expression patterns of Mn-SODs may explain the toxicity differences of Ti-NPs at different nano levels in animals. That is, the formation of toxicity differences from exposures to different sizes of TiO 2 -NPs may be primarily due to the formation of different expression levels of Mn-SODs in nematodes exposed to different sizes of TiO 2 -NPs. Nevertheless, we cannot exclude the possibility that specific gene expression patterns may form in nematodes exposed to different diameters of TiO 2 -NPs. For example, exposure to 7nm TiO 2 -NPs and exposure to 20-nm TiO 2 -NPs have been demonstrated to exhibit different expression patterns of gene cyp35a2, which encodes a xenobiotic metabolism enzyme in nematodes [20]. Analyses of the expression of an entire genome revealed the differential effects of TiO 2 nanotubes on vascular cells [58]. Analyses of the expression of an entire genome may help us to further reveal the possible specific gene expression patterns formed in nematodes exposed to TiO 2 -NPs at different nano levels in C. elegans. Increasing evidence has suggested that TiO 2 -NPs toxicity is dependent not only upon size, but also varies with particle shape, surface coating and functionalization [59]. Therefore, the other possible molecular mechanisms explaining the influences of particle shape, surface coating and functionalization Previous in vivo toxicity assays on TiO 2 -NPs suggested that subchronic dermal exposure to TiO 2 -NPs resulted in the significant changes in levels of SOD and MDA in the skin and liver tissues of mice, implying that pathological lesions may be mediated through the oxidative stress induced by deposited TiO 2 -NPs [11]. Similarly, alterations in the levels of biomarkers of oxidative stress suggested the formation of oxidative stress in the liver and gut tissues of zebrafish or the brains of rainbow trout exposed to 4 nm TiO 2 -NPs at the concentration of 10 mg/L. (D-F) Effects of ectopically expression of sod-2 or sod-3 gene on the locomotion behavior of nematodes exposed to 4 nm TiO 2 -NPs at the concentration of 10 mg/L. (G-H) Effects of ectopically expression of sod-2 or sod-3 gene on the intestinal autofluorescences of nematodes exposed to 4 nm TiO 2 -NPs at the concentration of 10 mg/L. (I-J) Effects of ectopically expression of sod-2 or sod-3 gene on the ROS production of nematodes exposed to 4 nm TiO 2 -NPs at the concentration of 10 mg/L. sod-2(+), Ex(Psod-2-sod-2); sod-3(+), Ex(Psod-3-sod-3). Exposure of TiO 2 -NPs was performed from L1-larvae, and the endpoints were examined when nematodes developed into the adults. Ti-NPs, TiO 2 -NPs. The Bars represent mean6S.E.M. **p,0.01. doi:10.1371/journal.pone.0044688.g003  [18,21]. Nevertheless, no direct evidence has been provided to prove the important role of oxidative stress in inducing the formation of in vivo toxicity by TiO 2 -NPs. Our data indicated that ectopically expression of the human SOD2 gene and nematode sod-2 and sod-3 genes not only prevented the induction of ROS production but also prevented the formation of toxicity in nematodes as assessed by the endpoints of lethality, growth, reproduction, locomotion behavior and intestinal autofluorescence in nematodes exposed to TiO 2 -NPs (Figs 3 and 4). These findings provided direct evidence of the key role of oxidative stress in inducing the in vivo toxicity of TiO 2 -NPs. In humans, SOD2 encodes Mn-SODs, a major mitochondrial antioxidative enzyme that constitutes an important ''control switch'' in the generation of oxidative signals [60]. In the present study, we investigated the toxicological effects of exposure to NPs on humans by generating transgenic strains in nematodes for specific human genes.
In summary, the present study showed that toxicity differences formed in nematodes exposed to different diameters of TiO 2 -NPs at predicted environmental relevant concentrations when using lethality, growth, reproduction, locomotion behavior, intestinal autofluorescence, and ROS production as endpoints. Our data also imply that the intestine, neuron, and reproductive organs may serve as the target organs for TiO 2 -NPs in nematodes. Among the examined endpoints, ROS production was closely correlated with other endpoints in nematodes exposed to TiO 2 -NPs. Exposure to TiO 2 -NPs altered the expression patterns of only sod-2 and sod-3 genes (which encode the Mn-SODs) among the examined genes regulating oxidative stress in nematodes. sod-2 gene expression and sod-3 gene expression were closely correlated with lethality, growth, reproduction, locomotion behavior, intestinal autofluorescence, and ROS production in nematodes exposed to TiO 2 -NPs. In particular, ectopically expression of human and nematode Mn-SODs prevented the induction of ROS production and the toxicity formation in nematodes exposed to TiO 2 -NPs. Therefore, our data reveal the possible molecular basis for oxidative stress in regulation of the formation of toxicity differences from different sizes of TiO 2 -NPs at predicted environmental relevant concentrations.

Strain preparation
The nematodes used in the present study were wild-type N2, originally obtained from the Caenorhabditis Genetics Center (funded by the NIH National Center for Research Resource, Bethesda, MD, USA) and transgenic strains of Ex(Psod-2-sod-2), Ex(Psod-3sod-3), and Ex(Pdpy-30-SOD2), which were maintained on nematode growth medium (NGM) plates seeded with Escherichia coli OP50 at 20uC as described [61]. Gravid nematodes were washed off the plates into centrifuge tubes, and were lysed with a bleaching mixture (0.45 M NaOH, 2% HOCl). Age-synchronized populations of L1-larval nematodes were obtained by the collection as described [24]. Exposures to different nano-sizes of TiO 2 -NPs at 0.001, 0.01, 0.1, 1, and 10 mg/L were done from the L1-larval stage in 12-well sterile tissue culture plates in a 20uC incubator in the presence of food. The nematodes were used for toxicity evaluation using lethality, growth, locomotion behavior, reproduction, intestinal autofluorescence, and ROS production as endpoints and for the gene expression pattern analysis when they developed into young adults.
Lethality A 1.0-mL aliquot of test solution for TiO 2 -NPs was added to each well of the tissue culture plates, which was subsequently loaded with 50 nematodes for each concentration. Fifty nematodes were loaded in each well, and three replicated were performed. After exposure, the inactive ones were scored under a dissecting microscope. Nematodes were judged to be dead if they did not respond to a stimulus using a small, metal wire. Lethality was evaluated by the percentage of surviving animals.

Locomotion behavior
Locomotion behavior was assessed by the endpoints of head thrash, body bend and forward turn [52]. To assess head thrash, each examined nematode was transferred into a microtiter well containing 60 mL of modified K-medium on the top of agar. Head thrashes were counted for 1 min after a 1-min recovery period. A thrash was defined as a change in the direction of bending at the middle of the body. To evaluate the body bend, nematodes were placed onto a second plate and the number of body bends over 20 s was recorded. A body bend was considered to be a change in the direction of the part of the nematode corresponding to the posterior bulb of the pharynx along the y axis assuming that the nematode was traveling along the x axis. To evaluate a forward turn, forward sinusoidal movement in a 30-s interval was measured. Fifty nematodes were examined per treatment.

Reproduction
After exposure in 1.0-mL aliquot of test solution for TiO 2 -NPs, reproduction was assessed by brood size of adult nematodes. To assess the brood size, we counted the number of offspring at all stages after the egg stage. Twenty replicates were examined per treatment.

Growth
After exposure in 1.0-mL aliquot of test solution for TiO 2 -NPs, growth was assessed by body length. Body length was determined by measuring the flat surface area of young adult nematodes using Image-ProH Express software. Twenty replicates were examined per treatment.

Intestinal autofluorescence
Intestinal autofluorescence is caused by lysosomal deposits of lipofuscin, which can accumulate over time in aging nematodes [53]. Images were collected for fluorescence in the endogenous intestine using a 525-nm bandpass filter and without automatic gain control to preserve the relative intensity of the fluorescence of different animals. Nematodes were photographed on the same day to avoid the effects of variance in light sources on fluorescence intensity. Fluorescence was recorded and color images taken for the documentation of results with MagnafireH software (Olympus, Irving, TX, USA). Lipofuscin levels were measured using ImageJ Software (NIH Image, Bethesda, MD, USA) by determining the mean pixel intensity in the intestine of each animal. Twenty replicates were examined per treatment.

ROS production
To ascertain if TiO 2 -NPs treatment activated oxidative damage, ROS production was assayed. Nematodes were transferred to M9 buffer containing 1 mM of 5-(and-6)-chloromethyl-29,79-dichlorodihydrofluorescein diacetate, acetyl ester (CM-H2DCFDA) to pre-incubate for 3 h at 20uC, and then mounted on agar pads for examination with a Laser Scanning Confocal Microscope (Leica, TCS SP2, Bensheim, Germany) at an excitation wavelength of 488 nm and emission wavelength of 510 nm. The relative fluorescence intensities of the intestines were semi-quantified. The semi-quantified ROS were expressed as relative fluorescent units (RFU). Twenty replicates were examined per treatment.

Statistical analysis
All data were expressed as means6standard error of the mean (S.E.M.). Statistical analysis was performed using SPSS 12.0 (SPSS Inc., Chicago, IL, USA). Analysis of variance (ANOVA) was used to determine the significance of differences between the groups. Probability levels of 0.05 and 0.01 were considered statistically significant. Associations of ROS production and gene expression with other endpoints were assessed with linear regression analysis. Correlations between sod-2 or sod-3 gene expression and ROS production were also analyzed by Spearman's rank correlation. Figure S1 Protein sequence alignment between human SOD2 protein with SOD-2 and SOD-3 proteins in C. elegans. ''*'' indicate the positions which have a single, fully conserved residue. The results showed that the protein sequences' identity between human SOD2 and C. elegans SOD-2 is 64.07%, and the identity between human SOD2 and C. elegans SOD-3 is 61.04%. (DOC) Figure S2 Confirmation of the ectopically expression of nematode sod-2 or sod-3 gene and human SOD2 gene in wild-type N2 nematodes. Relative expression ratios (between target genes and act-1 reference gene) in transgenic strains were normalized to the wild-type N2.

Supporting Information
(DOC) Table S1 Associations of ROS production with lethality, growth, reproduction, locomotion behavior and intestinal autofluorescence in nematodes exposed to TiO 2 -NPs as assayed by linear regression analysis.

(DOC)
Table S2 Associations of sod-2 or sod-3 gene expression with lethality, growth, reproduction, locomotion behavior, intestinal autofluorescence, and ROS production in nematodes exposed to TiO 2 -NPs as assayed by linear regression analysis.