The Toxic Effects of Cu and CuO Nanoparticles on Euplotes aediculatus

The single-celled eukaryote Euplotes aediculatus was chosen to test and compare the toxic effects of Cu and CuO nanoparticles (NPs). The antioxidant enzymatic activity, morphological changes, and functional groups on the membrane were determined using spectrophotometry, microscopy, and Fourier transform infrared spectroscopy after NPs treatment. The toxicity of the NPs to cells was dose-dependent, and the 24 h-LC50 values of the CuNPs and CuONPs were 0.46 µg/L and 1.24 × 103 µg/L, respectively. These NPs increased the activities of superoxide dismutase, glutathione peroxidase, and catalase and destroyed the cell structure; moreover, the CuNPs were more toxic than the CuONPs. In addition to the higher enzymatic activity, CuNPs also caused nucleoli disappearance, chromatin condensation, and mitochondrial and pellicle damage. The oxidization of the functional groups of the membrane (PO2 − , C–O–C, and δ(COH) of carbohydrates) also confirmed the severe damage caused by CuNPs. Our study showed that oxidative stress and organelle destruction played important roles in the toxic effects of these NPs on this protozoan. Compared with other aquatic organisms, E. aediculatus can be considered a potential indicator at the preliminary stage of environmental pollution.


Introduction
Nanotechnology is one of the most promising technologies from the twenty-first century. Currently, nanoparticles (NPs) are widely used in various fields, such as biomedicine [1], food processing [2], and energy [3]. As NPs are increasingly used in industry and daily life, they will inevitably enter the environment during production, use, and abandonment, thus posing potential threats to the environment and human health [4,5].
Currently, many studies on the biological toxicity of NPs have been presented. We surmised that the effects of NPs on cells are mainly reflected in the three following aspects: inhibiting cell growth or leading to cell death, causing antioxidant stress, and damaging the cell structure. Researchers have studied the inhibitory activity of CuNPs in a range of bacteria, including Staphylococcus aureus and Escherichia coli, and fungi, including Aspergillus flavus and Aspergillus niger [9]. It was reported that CuONPs produced inhibitory effects against both Gram-positive and Gram-negative bacteria [8]. A study showed that CuONPs had strong activity against S. aureus and E. coli [10]. The acute toxicity of CuNPs to the fish Acipenser baerii resulted in a 96 h-LC 50 value of 1.41 ± 0.24 mg/L [12]. Toxicity effects induced by CuNPs can cause death to five species of cladoceran (Daphnia magna, Daphnia pulex, Daphnia galeata, Ceraphaphnia dubia, Chydorus sphaericus) [13]. Research has shown that after treatment with CuONPs, the 96 h-LC 50 of the algae Nitellopsis obtusa was 2.8-4.3 mg/L, the 24 h-LC 50 of the shrimp Thamnocephalus platyurus was 8.5-9.8 mg/L, and 24 h-LC 50 of the rotifer Brachionus calyciflorus was 0.24-0.39 mg/L [11]. CuNPs increase the contents of reactive oxygen species (ROS) and decrease the activity levels of certain biomarkers, such as superoxide dismutase (SOD), catalase (CAT), glutathione-S-transferase (GST), and glutathione peroxidase (GPx) [14]. However, CuONPs have been shown to cause statistically significant increases in SOD, CAT, and peroxidase (POD) activities [7]. CuNPs and CuONPs can cause DNA damage in human lung cells [15]. When highnutritional organisms consume low-nutritional organisms containing NPs, the NPs will transfer within the food chain, and the former will also be poisoned by the NPs [16][17][18][19]. However, few reports have compared the toxicities of CuNPs and CuONPs. Additionally, a small number of studies have directly demonstrated the destruction of cellular structures by CuNPs or CuONPs [20][21][22].
Protozoa [23], the most primitive eukaryotic animals, play a key role in energy flow and material circulation. Euplotes species are ciliated protozoans with a wide distribution, diverse habitats, and ease of handling. They grow rapidly in the laboratory and have a short reproduction period. They are single-celled, and their pellicle can make direct contact with the outside environment, which makes them sensitive to pollutants, e.g., heavy metals and NPs [24,25]. As a result, their cellular responses reflect environmental changes in a timely manner. Compared with prokaryotes, the environmental sensitivity and response of protozoa are closer to those of eukaryotes [24,[26][27][28].

Cell Culture
Water sample was initially collected from Zizhu Garden in Minhang District, Shanghai, on 5 September 2010. E. aediculatus was isolated from the water sample by using a micro-pipette and uniprotistan culture was maintained in our laboratory. Identification was performed by examining the ciliature pattern and silver-line system according to Abraham et al. [29]. E. aediculatus cultured by clone was established for the present study in Petri dishes in an incubator (temperature: 25 °C, humidity: 76% RH, light/dark: 12/12 h). Deionized water was used as culture medium and wheat grains were added to enrich bacterial food source for the ciliates. The conditions for all tests were the same. Deionized water was used to prepare the test media. Before the experiment, the cells at stationary phase were obtained and washed with deionized water 2-3 times.

Preparation of NPs Suspensions
CuNPs (advertised particle size 10-30 nm, CAS no. 7440-  and CuONPs (40 nm, CAS no. 1317-38-0) were purchased from MACKLIN. Five milligrams of each NP was weighed and mixed with 25 mL of test medium to yield 200 mg/L stock suspensions, which were vortexed and sonicated for 15 min in a bath-type sonicator (SCQ-70, 60 W) before use. The morphology and size of the NPs in the test medium were characterized by using TEM (HT-7700), and 10 μL of each suspension was placed on the copper grid. The samples were observed directly after drying. The sizes of the particles were measured (based on 300 measurements) from the TEM photographs using ImageJ (NIH, USA). One milliliter of each suspension was used to determine the zeta potential, which was measured using a Zetasizer Nano ZS instrument (Zetasizer Nano ZS90).

Effects of the NPs on the Cell Growth Curve
The EC 50 values (24,48,72 h, further details are provided in section "Determination of Toxicity Parameters") determined above were taken as the experimental concentrations, with five replicates in each group. Only one active cell was added to The Toxic E e ff cts of Cu and CuO Nanoparticles on Euplotes aediculatus 545 200 µL of culture water for each group: the control or various concentrations of NP suspensions. The cell number in each group was recorded, and 20 µL of wheat fermentation was added daily. The experiment lasted for 21 days in an incubator.

Enzymatic Activity Assay
The EC 50 values determined above were taken as the experimental concentrations, with three replicates in each group. Cell filters (150 and 200 mesh) were used for cell collection with pore sizes of 70 and 40 µm, respectively. The final cell densities were 5 × 10 5 cells/mL. Each replicate contained 200 µL of NPs suspension, and 5 µL of cells were incubated in a Petri plate with a diameter of 6 cm. The test was conducted in an incubator. Each suspension was filtered according to the same collection method above, leaving 5 µL of cells to determine the SOD, GPx, and CAT activities after 24 h, which were measured using spectrophotometry (BioTek, EPOCH2). For details, refer to the SOD, GPx, and CAT test kit instructions (purchased from the Nanjing Jiancheng Bioengineering Institute).

Morphological Observation
Living Cell Observation The 24 h-LC 50 was used as the experimental concentration. Samples were prepared after treatment times of 1.0, 1.5, 2, 2.5, and 3 h. Living cells from the control and experimental groups were isolated and observed in vivo by microscopy (Nikon DS-Ri2). The motion state and morphology of the cells were mainly observed.

SEM Observation
Cells were prepared according to the procedure of Gong et al. [27]. The 24 h-LC 50 was used as the experimental concentration. Cells were collected after treatment times of 0.5, 1.0, 1.5, 2, and 3 h. Cells were collected and fixed using 4% osmium tetroxide (OsO 4 ) for 10 min. After washing with buffer, the cells were dehydrated in gradient concentrations of ethanol (30%, 50%, 70%, 80%, 90%, 95%, and 100%), critical point dried, and sputter coated with platinum. Samples were observed by SEM (Hitachi, S4800). For morphological observation, 30 cells were sized and typical morphology was selected for show.

Attenuated Total Reflection-Fourier Transform Infrared Spectroscopy
Cells in the control and experimental groups (24 h-LC 50 ) were collected by filtration (details are provided in section "Enzymatic Activity Assay"). Treated cells (5 µL) were prefrozen at − 80 °C for 12 h, vacuum freeze-dried, and examined by ATR-FTIR spectroscopy (NEXUS 670).

Statistical Analysis
SPSS (version 23.0 for Windows) software was used for statistical analysis. The mean was compared by T-test and one-way ANOVA to determine whether significant variation existed between the treatments. P < 0.05 was considered statistically significant. GraphPad Prism 7 software was used for plotting.

Characterization of the CuNPs and CuONPs
The morphologies of the CuNPs and CuONPs were verified by TEM in S1, and the images showed that they were spherical in shape with the particle sizes of approximately 30 ± 10 and 30 ± 5 nm, respectively. The CuNPs had a wider size distribution, and the CuONPs had a more uniform size. The pH of both NPs suspensions in test medium was approximately 7.5. Both possessed a negative zeta potential (− 28.18, − 30.10 mV) and were considered stable systems [30]. These results indicated that the NPs used in this experiment were suitable for toxicity testing.

Toxicity Tests
The 24 h-LC 50 Values of the NPs The mortality of the cells in each group was calculated after 24 h. The regression curves of the acute toxicity of the CuNPs and CuONPs are shown in Fig. 1. The mortality rate was linear as a function of the logarithm of the concentration of NPs. The concentration of CuNPs causing 50% cell mortality (24 h-LC 50 ) was 0.46 µg/L, while that for the CuONPs was 1.24 × 10 3 µg/L. These LC 50 values showed that the CuNPs were more toxic to E. aediculatus. Table 1, there was a linear relationship between the reproduction rate and the logarithm of the concentration of NPs. The EC 50 values of X. Zhao et al. 546

EC 50 Values of the NPs As shown in
the CuNPs were lower than those of CuONPs over the same time period. As the treatment time increased, the EC 50 values of the two NPs decreased.

Effects of the NPs on Cell Growth
The growth curve of the control group was S-shaped, as shown in Fig. 2a. The peak value of specific growth rate (SGR) in the control group was respectively reached on the 4th and 11th day. Then the value of SGR gradually decreased from the 12th day (Fig. 2b). The two NPs significantly inhibited cell growth (P < 0.05), and both inhibitors were dose-dependent. The 24 h-EC 50 concentration of the CuNPs and CuONPs showed the strongest inhibitory effects on cell growth, and their growth rates were the lowest (Fig. 2b). Compared with the control group, the SGR of the above two groups showed significant difference since the 4th day (P < 0.05). The 72 h-EC 50 concentration of the CuNPs and CuONPs displayed the weakest inhibitory effects, and the corresponding curve still holds the S-shape. The SGR of the control and CuNPs groups had significant difference since the 18th day (P < 0.05). There was no significant difference between the CuONPs and control groups (Fig. 2b). Then, the cells stopped reproducing since the 13th day after treatment with the CuNPs-48 h-EC 50 concentration. Cells stopped reproducing since the 14th day after treatment with   The Toxic E e ff cts of Cu and CuO Nanoparticles on Euplotes aediculatus 547 the CuONPs-48 h-EC 50 concentration. Compared with the control group, the SGR of these two groups showed significant difference since the 11th day (P < 0.05). The inhibitory effects of the CuNPs were stronger than those of the CuONPs.

Enzymatic Activity Assay
Compared with the control group, the activities of SOD, GPx, and CAT were higher in the experimental groups with different concentrations of CuNPs (Fig. 3a) and CuONPs (Fig. 3b), and the difference was statistically significant (P < 0.05). With increasing NPs concentration, the activities of all three enzymes increased in a dose-dependent manner. CuNPs induced higher SOD activity in the 24 h/48 h-EC 50 groups and the opposite result was found in the 72 h-EC 50 group. CuNPs induced higher GPx activity in the 24 h/72 h-EC 50 groups but produced the opposite result in the 48 h-EC 50 group. CuNPs induced higher CAT activity in all three groups (Fig. 3c).

Morphological Observations
Living Cell Observation The E. aediculatus cells used in this study were approximately 130-150 µm long × 70-90 µm wide and ellipsoidal (Fig. 4a). Cells possessed the integral ventral ciliary organelles and swam freely by movement of cirri.
After treatment with CuNPs, the movement and morphology of the cells changed to varying degrees over time. The cells whirled in place, and the cirri quivered violently (Fig. 4b,c). Then, cell movement faded, and the cilia wobbled weakly until the cells became still and died (Fig. 4d-f). A few small vacuoles appeared in the cell (Fig. 4b), which gradually fused and expanded (Fig. 4c,d). The cell was swollen and deformed into a ball shape, and the endoplasm flowed out due to cell lysis (Fig. 4e,f).
Similar to the CuNPs groups, changes in the cell movement and morphology were also observed in the CuONPs groups. Initially, there was no significant change in motion state or morphology (Fig. 4g). Many small vacuoles then appeared in the cells (Fig. 4g). The cilia quivered intensely, and several of the small vacuoles gradually fused into large vacuoles (Fig. 4h,i). The movement of cilia gradually weakened, and then the cells became stationary (Fig. 4j-l).
Finally, the vacuoles filled the whole cell, the cell swelled into a ball (Fig. 4j,k), the cell membrane ruptured, and cell death occurred (Fig. 4l).

SEM Observation
The SEM images showed that the control cells possessed the integral ventral ciliary organelles. The front-ventral cirri (FVC), transverse cirri (TC), caudal cirri (CC), and left marginal cirri (LMC) were distributed at specific locations on the cell surface, which constituted the unique ciliary pattern of E. aediculatus. The adoral zone of membranelles (AZM) was located on the left side of the FVC, occupying 2/3 of the body. The undulating membrane (UM) was located on the right side of the base of the AZM (Fig. 5a). The cell morphology and cilia structure varied with NPs treatment time. The cell size did not change significantly after treatment with the 24 h-LC 50 of the CuNPs after 0.5-1.0 h; however, the cilia of the LMC and CC had completely shed, the kinetosomes had partially shed (Fig. 5b,i), and the FVC and TC had partially shed (Fig. 5b,g,h). After 2 h of treatment, the FVC and TC continued to shed, and the cilia of the collar part of the AZM (AZM-C) and lapel part of the AZM (AZM-L) began to drop off (Fig. 5c,j). The cells were ellipsoidal to spherical in shape, leaving only the cilia of the AZM after 3.0 h (Fig. 5d-f). The Toxic E e ff cts of Cu and CuO Nanoparticles on Euplotes aediculatus 549 Cells treated with the CuONPs 24 h-LC 50 manifested similar changes. During the first 0.5 h, there was no significant change in cell size. The FVC and CC began to shed (Fig. 6a,f,g). After another 0.5 h, the LMC and CC had completely shed (Fig. 6b), the FVC continued to shed (Fig. 6b,h), and the TC began to shed (Fig. 6i). After 2.0 h, the FVC had completely shed, and the TC displayed more shedding (Fig. 6c,j,k). The cells became spherical, leaving only the cilia of the AZM after 3.0 h (Fig. 6d,e).

TEM Observation
The pellicle of E. aediculatus was intact and complete (Fig. 7a), including the plasma membrane and outer and inner alveolus (Fig. 7b). The microtubules beneath the pellicles were arranged in a triad pattern (Fig. 7c). The cytoplasm was dense, and the mitochondria were mainly concentrated in the cortical cytoplasm or beneath the pellicles (Fig. 7a-c,e). The mitochondria of the control cell were round or oval and composed of a double-layer membrane, with the inner membrane folded inward into the cristae (Fig. 7a-c). The macronucleus was clearly demarcated in the control cell. Nucleoli and chromatin were evenly distributed in the macronucleus (Fig. 7d).
Compared with the CuNPs, the pellicle structure in the CuONPs group was clearer and more complete, and there was also alveolus. The concentrations of NPs used were the 24 h-LC 50 , and the exposure time was 2 h. Mitochondria became wrinkled and deformed (Fig. 8b), some structures were seriously damaged, and the cristae were fractured because of CuNPs treatment (Fig. 8c). While the morphology of the mitochondria changed significantly to show an irregular shape, the cristae did not break in the CuONPs group (Fig. 9b,c). After treatment with the CuNPs, the nucleoli disappeared, and the chromatin displayed significant condensation (Fig. 8e). However, there were no significant changes to the nucleus after treatment with the CuONPs (Fig. 9e).

ATR-FTIR
As shown in Fig. 10, the most prominent peaks in the control group were found at 2360, 1646, and 1543 cm −1 from CO 2 , amide I, and amide II, respectively. The peaks from E. aediculatus in the spectra refer to S2. The peaks at 2925, 2856, 1457, and 1083 cm −1 were prominent as well. The first three are from C-H bonds, while the last was attributed to v s (PO 2 −).
The peaks with relatively lower intensities appeared at 1243, 1050, and 1020 cm −1 , corresponding to the v a (PO 2 −) and the δ(COH) of carbohydrates and C-O-C, respectively. After treatment with the two NPs, some of these peaks weakened or disappeared, and new peaks were observed. The C-H, amide I, amide II, v a (PO 2 − ), and v s (PO 2 − ) functional group peaks weakened, and the δ(COH) of carbohydrates and the C-O-C peak disappeared in the CuNPs group. Moreover, the -COO − str. group was observed at 1393 cm −1 . However, the CuONPs had less of an effect on the above functional groups, and only the -COO − str. group was added at 1412 cm −1 .  The pellicle structure included the plasma membrane and outer and inner alveolus (arrows). c The microtubules beneath the pellicle were arranged in a triad pattern (arrows). d Nucleoli and chromatin distributed in the macronucleus. e Mitochondria in the cortical cytoplasm. Scale bars = 0 .5 µm (a, c, d), 0.2 µm (b), 1 µm (e)

Comparing the Toxicities of CuNPs and CuONPs to E. aediculatus
There have been many studies on the toxicity of CuNPs and CuONPs [11][12][13], especially the latter. Experiments have shown that CuONPs are lethal to Tetrahymena thermophila, and this toxicity is caused by their solubilized fraction [31][32][33]. There are a few papers that have compared the toxicity of CuNPs and CuONPs. A study showed that the 48 h-LC 50 values of CuNPs in adult zebrafish and eggs were 4.2 and 24.0 mg/L, respectively, while these values for CuONPs were 400 and 960 mg/L, respectively [34]. It was also found that CuNPs had stronger cytotoxicity than CuONPs [15]. To date, there has been no study on the comparative toxicity of CuNPs and CuONPs to protozoa. The LC 50 /EC 50 values of CuNPs were lower than those of the CuONPs in our study, which means that CuNPs were more toxic to E. aediculatus. Compared with the LC 50 values to other aquatic organisms, E. aediculatus displayed the lowest value, which means that E. aediculatus is more sensitive. Considering that E. aediculatus has many characteristics of wide distribution, large quantity, and rich biological background, is an important part of the ecosystem, is easy to culture in the laboratory, and the response to external factors can be measured, it can be recommended as an indicator at the preliminary stage of environmental pollution.

Oxidative Stress in the Cell
Cells produce a small amount of ROS under normal conditions. ROS are cleared by the antioxidant defense system, as metabolism is in a dynamic state of balance. NPs can stimulate cells to produce more ROS, exceeding the cell's clearance abilities, which then requires the antioxidant system to react [35,36]. This system consists of antioxidant enzymes and antioxidants. The involvement of enzymes, including SOD, CAT, and GPx, is an important defense mechanism to protect organisms from oxidative stress [7]. SOD transforms the superoxide anion radical (O 2 -·) into H 2 O 2 , which is transformed into H 2 O by CAT. Thus, ROS can be effectively eliminated by regulating the levels of SOD and CAT [37]. Glutathione (GSH), an antioxidant, plays a critical role in the detoxification pathways of electrophiles such as copper, and it is the substrate of GPx [38]. GPx is also an important enzyme that catalyzes H 2 O 2 production, and copper has been shown to increase the expression of the GPx gene in Euplotes crassus [39]. Research has shown that CuNPs increase ROS in renal tissues and decrease the levels of biomarkers related to oxidative stress [14]. It has been reported that CuONPs were prone to oxidative damage [40]. The induction of ROS by CuONPs has been confirmed at all levels in biological tissues. The activities of three kinds of antioxidant enzymes increased, and an oxidative stress reaction was produced in E. aediculatus after treatment with both NPs. As discussed in a previous study [15], CuNPs have more active chemical properties and more active sites, so it more easily produces the free radical O 2 -· and ROS, which could increase the oxidative pressure of cells, further causing lipid peroxidation and leading to cell membrane damage. In addition, the outer layer of CuNPs has surface oxides such as Cu 2 O and CuO, which allow for the easier extraction of Cu 2+ . CuONPs are composed of CuO, a unitary component with a relatively stable structure. When organisms produce too much superoxide (such as O 2 -·) or contain reductants, they can reduce Cu 2+ to Cu + . Cu + could then catalyze H 2 O 2 to generate hydroxyl radicals [41,42]. The oxidative stress to E. aediculatus caused by CuNPs was stronger in our study. It has therefore been speculated that this toxicity may be related to the surface composition of the NPs.

Damage to the Cell Membrane, Mitochondria, and Nuclei
NPs penetrate cells by changing the cell membrane permeability and react with intracellular substances to cause cytotoxicity. It was observed that these NPs destroyed the cell membrane, causing a loss in selective permeability and an overall increase in permeability [24]. A study confirmed that CuONPs reduced membrane fluidity by inhibiting the production of fatty acid denaturation in T. thermophila [30]. Generally, the intact cell walls of unicellular organisms will not allow NPs to enter. When NPs are contained within the body, it indicates that the previous damage had led to an increase in cell wall permeability [43]. The intact cell walls of bacteria, yeasts, and unicellular algae have been shown to prevent NPs from entering cells, especially large NPs [11]. In our study, the ultrastructural results did not show a great gap in the E. aediculatus membrane. In general, E. aediculatus feed through oral apparatus including AZM and UM, which is a tool for feeding. It was therefore speculated that these NPs do not enter through the membrane but are instead ingested by this tool. However, it was also observed that the structure was denser and more complete after treatment with CuONPs than the CuNPs. Besides, more cilia loss was observed with the CuNPs. The surface of each cilium was covered by an outer pellicle, which is consistent with the ATR-FTIR results to some extent. Amide groups, C-H, P = O, and δ(COH) of carbohydrates and C-O-C groups are the main functional groups that constitute cell membrane components, i.e., proteins, phospholipids, and glycoproteins [44]. Amide groups are linked to hydrophobic ends and exposed to the environment, so the peaks The Toxic E e ff cts of Cu and CuO Nanoparticles on Euplotes aediculatus 553 from these groups are the most significant. The decreases in intensity of the peaks from the amide groups suggested that the outer hydrophobic ends were easily oxidized by the CuNPs. Changes in v a (PO 2 -) and v s (PO 2 -) indicated the destruction of lipids [26,45]. Decreases in the δ(COH) of carbohydrates and C-O-C groups indicated damage to polysaccharides and glycoproteins [26]. The presence of the -COO-str. group was due to the unsaturated aldehydes that formed during the decomposition of the peroxides from hydroperoxides or lipids as well as the C = O stretching bonds from the formation of carboxyl groups [46]. Some functional groups of the membrane were oxidized at the molecular level, which was further suggested that cells experience oxidative stress in response to exogenous stimulus factors. Some studies have shown that CuNPs or CuONPs can damage cell structures such as the mitochondria and nuclei. CuNPs and CuONPs have been shown to cause DNA damage in human lung cells [15]. Studies have also shown that CuNPs can break the DNA chains in rat ovaries and promote cell apoptosis [20]. The expression levels of the cell cycle checkpoint protein p53 and the DNA damage repair proteins Rad51 and MSH2 were upregulated after treatment with CuONPs [47]. The NPs damaged the chromosomes, causing genetic changes and DNA damage [48]. In our study, both CuNPs and CuONPs caused damage to the nucleus. This result was consistent with previous studies. Notably, the destruction caused by the CuNPs was more serious. Researchers have proposed that CuNPs could induce liver cell and renal tubule necrosis and mitochondrial function failure [49]. Another study showed that mitochondrial dysfunction and oxidative damage were the main causes of the toxicity of CuNPs in rats [21]. CuONPs have been found to decrease the mitochondrial membrane potential and upregulate apoptosis genes, resulting in apoptosis in human hepatocytes [22]. Additionally, it was found that cell apoptosis after treatment with CuONPs was mediated by the production of ROS, involving destruction of the mitochondrial membrane potential in A549 cells [50]. ROS generation and GSH reduction both lead to mitochondrial dysfunction [51]. Some scholars believe that cells experience oxidative stress after NPs treatment; these NPs mainly attack the cristae of mitochondria, affect the electron transport chain, produce high levels of ROS, and interfere with the synthesis of ATP [52]. In this study, after treatment with both NPs, the morphology and structure of the mitochondria changed with significant rupture, which was worse with the CuNPs than with the CuONPs. The above phenomena were consistent with the results from the toxicology experiments, in which the two most important toxic mechanisms of CuNPs and CuONPs were the destruction of the nucleus and mitochondrial dysfunction mediated by the production of ROS.

Conclusion
Existing studies on the nanotoxicity have shown that NPs usually play a toxic role through oxidative stress and structure damage. In this paper, the toxicity of CuNPs exposed to E. aediculatus was studied comparatively with CuONPs. Our hypothesis is that CuNPs have stronger cytotoxicity than CuONPs on ciliates caused by strong oxidative stress and severe structure damage. The enzymatic activity, morphological changes, and functional groups on the membrane were determined after NPs treatment. The obtained data was consistent with the expected hypothesis. Based on toxicity tests, LC 50 values showed that the CuNPs were more toxic to E. aediculatus. Both CuNPs and CuONPs can cause antioxidant stress and destroy cell structure such as mitochondria and nucleus, while the former one leads to more serious destruction.