**3. Results and Discussion**

### *3.1. Concentration of Zn Released from ZnO-NPs in LB Medium*

Metal nanoparticles exhibit distinct characteristics under different environmental conditions because of their interactions with abiotic and biotic factors [37]. Additionally, the ZnO-NPs are substantially affected by the background medium, resulting in the release of varying amounts of Zn2+ [26]. Therefore, before *E. coli* exposure, we determined the concentrations of Zn2+ released from the ZnO-NPs in the LB medium at different time points. The Zn2+ concentration in the LB medium remained constant at approximately 1.8 mg/L after 0.5, 2, 4, and 8 h (Figure 1). Furthermore, the Zn2+ concentration reached a steady state immediately after the ZnO-NPs were added to the LB medium (Figure 1).

**Figure 1.** Released Zn ions from ZnO-NPs in LB medium. Initial concentration of ZnO-NPs suspension was 5 mg/L prepared in LB medium. Samples were incubated at 37 ◦C and analyzed at different time points at 0, 0.5, 2, 4, and 8 h. Samples were filtered to remove undissolved ZnO-NPs, and then the concentrations of Zn ions in the aqueous phase were measured using ICP-AES. The data are shown as the mean ± SEM.

Figure 1 shows that Zn2+ constituted approximately 35% of the total ZnO-NPs in the LB medium, implying that compared to the concentration of the particulate form, that of the ionic form of ZnO-NPs was relatively lower in the LB medium. Recent studies have also demonstrated the low solubility and stability of ZnO-NPs in the LB medium (5–51% Zn2+ of total ZnO-NPs) [38,39], which may have been due to the presence of organic matter in the LB medium. Some types of organic matter may inhibit the release of free Zn2+ from ZnO-NPs [40,41]. Moreover, the dissolution rate of ZnO-NPs is inversely correlated to aliphatic carbon content and hydrogen/carbon ratio [42]. Therefore, both ionic and particulate forms of ZnO-NPs can potentially transfer to *E. coli* OP50, and the accumulation was further investigated.

### *3.2. Zn Accumulation in E. coli OP50*

To investigate the accumulation of ZnO-NPs in *E. coli* OP50 that might further transfer to *C. elegans*, we exposed *E. coli* OP50 to different concentrations of ZnO-NPs (5, 10, 50, and 100 mg/L) and analyzed the Zn concentrations in the bacteria. The concentrations were designed based on the minimal inhibitory concentration (MIC) of 400 mg/L for ZnO-NPs in *E. coli* strain [43]. We selected concentrations below the MIC, and serial dilution was applied to establish the dose-response relationships for the toxicological endpoints.

We found that Zn concentrations in bacterial cells increased in a dose-dependent manner (Figure 2). Exposure to 100 mg/L of ZnO-NPs caused a cellular burden of approximately 200 μg/10<sup>8</sup> cells, which was 100 times higher than that due to 50 mg/L of ZnO-NP exposure (Figure 2). Exposure to ZnCl2 also resulted in a substantial accumulation of Zn in the bacterial cells; 50 mg/L ZnCl2 led to a cellular burden of approximately 2.5 μg/10<sup>8</sup> cells, whereas 100 mg/L ZnCl2 caused a cellular burden of approximately 500 μg/10<sup>8</sup> cells (Figure 2).

**Figure 2.** Accumulated Zn concentration in *E. coli* OP50 upon ZnO-NP or ZnCl2 exposure. Saturated *E. coli* OP50 were diluted and incubated in LB medium with various concentrations of ZnO-NPs or ZnCl2 for 8 h at 37 ◦C. Subsequently, *E. coli* OP50 pellet was washed and collected for Zn concentration analysis using ICP-AES. The data are shown as the mean ± SEM. Statistical significance was determined by ANOVA with LSD post-hoc test to compare to the control (0 mg/L). (\*\*: *p* < 0.01, \*\*\*: *p* < 0.001).

Thus, ZnO-NPs accumulated in *E. coli* OP50 after 8 h of exposure at all of the examined concentrations (Figure 2). Metal nanoparticles, such as TiO2 and Ag, accumulated in the bacteria, including *E. coli* and *Pseudomonas aeruginosa*, and then were transferred to higher trophic levels [20,44]. Several studies have shown that ZnO-NPs can damage the bacterial

cell wall and enhance membrane permeability, thereby resulting in their accumulation in bacteria [45,46]. Additionally, the internalization of ZnO-NPs by *E. coli* and other bacterial cells has been observed previously [47,48]. The high concentration (100 mg/L) of ZnO-NPs and ZnCl2 largely increased the Zn accumulation compared with 50 mg/L, which may be due to the membrane damage that facilitated higher Zn accumulation in the cytoplasm [49]. Therefore, our results sugges<sup>t</sup> that ZnO-NPs can accumulate in prey (*E. coli*) and potentially be transferred to higher trophic levels through dietary intake.

### *3.3. Distribution and Accumulation of ZnO-NPs in C. elegans via Dietary Transfer*

To further assess the dietary transfer of ZnO-NPs from *E. coli* to *C. elegans*, the worms were exposed to *E. coli* OP50 pre-treated with ZnO-NPs. The control worms were fed with *E. coli* pre-treated with RhoB/deionized water. The fluorescent dye RhoB was used to label the ZnO-NPs to visualize the distribution and accumulation of the ZnO-NPs due to dietary transfer. Compared with the controls (RhoB/deionized water), the RhoB-labeled ZnO-NPs accumulated mainly in the pharynx and intestine of *C. elegans* (Figure 3A). Quantification of the fluorescence intensity showed that the background levels in the controls were approximately 6 RFU/worm, which significantly increased to approximately 15 RFU/worm in the presence of RhoB-labeled ZnO-NPs (Figure 3B).

**Figure 3.** Accumulated ZnO-NPs in *C. elegans* through dietary transfer. Wild-type N2 *C. elegans* L1-larvae were fed with rhodamine B (RhOB)-labeling ZnO-NPs (RhoB/ZnO-NPs) pretreated *E. coli* OP50 for 96 h. RhoB/Deionized water was used as the control. After exposure and washing, (**A**) fluorescence images of worms were taken, and (**B**) fluorescence intensity was analyzed using ImageJ. The data are shown as the mean ± SEM. The tests were conducted at least three times independently, and 25 worms were scored per treatment in each replicate. Statistical significance was determined by ANOVA with LSD post-hoc test to compare to the control. (\*\*\*: *p* < 0.001).

While the dietary transfer of ZnO-NPs in aquatic food chains has been demonstrated in several studies [19,50–52], little is known about their distribution. Trophic transfer of ZnO-NPs occurs in simple food chains involving algae (*Chlorella ellipsoidea*) and clams (*Corbicula fluminea*) [53]. Furthermore, goldfish fed with brine shrimp pre-exposed to ZnO-NPs showed significant accumulation of Zn in the intestine [54]. In *C. elegans*, the intestine is the primary target of nanomaterials, including SiO2, carbon nanotubes, graphene oxide, and Ag [20,55–57]. This may be the reason for the significant ZnO-NP accumulation in the intestine and pharynx of *C. elegans* due to dietary transfer from *E. coli* OP50 (Figure 3A).

### *3.4. Effects of Dietary Transfer of ZnO-NPs on Locomotive Behaviors of C. elegans*

The sublethal endpoints, including body bending and head thrashing frequencies, of *C. elegans* have been used to assess neurotoxicity [31,58,59]. Moreover, we previously found that aquatic exposure to ZnO-NPs in simulated surface water (EPA water) significantly

impairs locomotive behaviors, indicating that neurotoxicity is a potential result of ZnO-NP exposure [26]. Therefore, the effects of the dietary transfer of ZnO-NPs on *C. elegans* were investigated using locomotive behavior tests. In addition, ZnCl2 was used to differentiate between toxicity due to ionic Zn and that caused by ZnO-NPs.

The body bending frequency decreased in *C. elegans* fed with ZnO-NPs pre-treated *E. coli* OP50 in a dose-dependent manner (Figure 4A). In contrast, there was no significant difference in the body bending frequency of *C. elegans* fed with relatively high concentrations (50 and 100 mg/L) of ZnCl2 pre-treated *E. coli* OP50 and that of the controls (fed with 0 mg/L ZnCl2 pre-treated *E. coli* OP50) (Figure 4A). Tests pertaining to the head thrashing frequency exhibited similar results (Figure 4B). *E. coli* pre-treated with 50 and 100 mg/L of ZnO-NPs demonstrated a 9–10% reduction in head thrashing and a 17–18% reduction in body bending (Figure 4A,B), implying that body bending frequency may be a more sensitive endpoint than head thrashing.

**Figure 4.** Locomotive behavior defects resulted from dietary transfer of ZnO-NPs in *C. elegans.* Wild-type N2 *C. elegans* L1-larvae were fed with ZnO-NPs or ZnCl2 pretreated *E. coli* OP50 for 72 h. After exposure and washing, (**A**) body bends and (**B**) head thrashes of worms were determined. The data are shown as the mean ± SEM. The tests were conducted at least three times independently, and 20 worms were scored per treatment in each replicate. Statistical significance was determined by one-way ANOVA with LSD post-hoc test and indicated by different lowercase letters (*p* < 0.05).

Our results sugges<sup>t</sup> that the predicted environmental concentration of ZnO-NPs (76 μg/L) [60] may be harmful to locomotive behavior and cause ecotoxicity. Altered behaviors caused by environmental toxicants can reduce the fitness and population of organisms, indicating the potential impact of neurotoxicity on the ecosystem [30]. A previous study showed that exposure to ZnO-NPs impaired motor functions in mice [11]. Liquid exposure to ZnO-NPs shows impairment of locomotive behaviors in *C. elegans* and zebrafish, which is more significant than that caused by ZnCl2 [26,27]. Thus, the biological actions of ZnO-NPs and Zn2+ are different, and the impairment of motor functions induced by ZnO-NPs may be similar in different species. It has been shown that the sensitivity of body bends to the liquid exposure of ZnO-NPs was higher than head thrashes in *C. elegans*, which is in agreemen<sup>t</sup> with our results [26]. Similarly, the direct liquid exposure of 500 μg/L of ZnO-NPs caused higher toxic effects on body bends than 500 μg/L of ZnCl2 [26]. Moreover, our results showed that impaired locomotive behaviors and neurotoxicity in *C. elegans* were due to the dietary transfer of ZnO-NPs (Figure 4A,B). The dietary transfer of Ag nanoparticles is known to disrupt the locomotion of springtails (Collembola) [21]. Additionally, copper (Cu) nanoparticles accumulate in *Daphnia magna* by dietary transfer and impair their feeding rate [61]. Therefore, various metal nanoparticles in the environment can produce toxic effects when they are transferred to higher trophic

levels. Although the dietary transfer of ZnO-NPs through trophic levels has been reported, its neurotoxicity due to dietary transfer has rarely been reported. Our findings sugges<sup>t</sup> that ZnO-NPs can accumulate in *C. elegans* via trophic transfer, thereby impairing locomotion behaviors. However, the mechanisms behind the locomotion defects due to the dietary transfer of metal-based nanoparticles are unclear.

### *3.5. Effects of Dietary Transfer of ZnO-NPs on D-Type GABAergic Motor Neurons of C. elegans*

The D-type GABAergic motor neurons are inhibitory motor neurons that control locomotive behaviors in *C. elegans* [31,62]. Therefore, we investigated the effect of the dietary transfer of ZnO-NPs on these neurons (Figure 5A) and found that they decreased the cell bodies of D-type GABAergic motor neurons in a dose-dependent manner (Figure 5B). The representative images of GABAergic neuron in *C. elegans* with different treatments were shown in Figure S2. In addition, ZnO-NP treatment caused greater degeneration of D-type GABAergic motor neurons than that caused by ZnCl2 (Figure 5B). Moreover, a gap formation on the cords of the D-type GABAergic motor neurons increased due to the dietary exposure of ZnO-NPs in a dose-dependent manner, which was more significant than that caused by the ZnCl2 treatment (Figure 5B). Thus, dietary exposure to ZnO-NPs significantly damages D-type GABAergic neurons in *C. elegans*. In addition, we have checked the cholinergic motor neurons, which are the other motor neurons controlling the locomotive behavior of *C. elegans*. The results of the gap numbers showed that there is no significant difference between the control and treatment with ZnO-NPs or ZnCl2 (Figure S3).

**Figure 5.** GABAergic neuron damage resulted from dietary transfer of ZnO-NPs in *C. elegans.* Transgenic strain EG1285 (*unc-47*p::GFP) L1-larvae were fed with ZnO-NPs or ZnCl2 pretreated *E. coli* OP50 for 72 h. After exposure and washing, fluorescence images of worms were taken to determine neural abnormalities, including neuron loss and degenerating commissures of GABAergic D-type motor neurons. ( **A**) Representative image for determination of GABAergic neuronal damage. White arrows indicate normal cell bodies; red arrows indicate gaps on the neuronal cord. (**B**) GABAergic neuron damage resulted from dietary transfer of ZnO-NPs. The data are shown as the mean ± SEM.The tests were conducted at least three times independently, and 20 worms were scored per treatment in each replicate. Statistical significance was determined by one-way ANOVA with LSD post-hoc test and indicated by different lowercase letters (*p* < 0.05).

Toxic metals and nanoparticles, such as quantum dots, have been shown to adversely affect locomotive behaviors and damage D-type GABAergic motor neurons [28,29]. Nevertheless, other nanomaterials, including graphene-based Ag ones, do not harm GABAergic neurons [25,55]. This may be because of different toxicity mechanisms. Our results con-

firmed that the D-type GABAergic motor neuron system was damaged in *C. elegans* that were fed with ZnO-NPs pre-treated *E. coli* OP50 (Figure 5B), which might have contributed to the impairment of the locomotive behaviors (Figure 4A,B). A previous study found that exposure to ZnO-NPs also induced dopaminergic neuronal damage in zebrafish brains [63]. Additionally, exposure to ZnO-NPs resulted in neuronal damage in the brains of Wistar rats [64]. These sugges<sup>t</sup> that ZnO-NP-induced neuronal damage might not be exclusive to *C. elegans*. A previous study showed that cancer cell lines were more sensitive to the cytotoxic effects of ZnO-NPs than normal cell lines [65]. Despite the growing body of evidence demonstrating the promising potential of ZnO-NPs in biomedical applications, the non-selective cytotoxic effects against the various cell lines of ZnO-NPs remain controversial [66]. Our findings sugges<sup>t</sup> that precautions should be taken as dietary ZnO-NPs might result in neuronal damage in vivo.

Notably, the adverse effects on the neuron system and locomotive behaviors were not observed in the *C. elegans* that were fed ZnCl2 pre-treated *E. coli* OP50 (Figures 4 and 5), indicating that the effects of the dietary transfer of ZnO-NPs were primarily due to particulate forms rather than Zn2+. The particle-specific effects of the dietary transfer of ZnO-NPs were also observed in several studies, especially those on bioaccumulation [19,51]. The present study further suggests that neurotoxicity caused by the dietary transfer of ZnO-NPs is more significant than that due to ZnCl2 exposure. Previous studies have mainly reported the accumulation of dietary ZnO-NPs in the gut, as well as mortality, reproduction, and oxidative stress response [19,50–52]. We provided evidence that dietary ZnO-NPs can also adversely affect the locomotion of *C. elegans* by causing neuronal damage in vivo. Therefore, our findings reveal rarely reported neuronal damage induced by dietary ZnO-NP exposure.
