Plain polystyrene (PS) nanoplastics were purchased from Big Goose Scientific (Tianjin, China). All the PS nanoplastics were dispersed in deionized water with a mass concentration of 10 mg/mL and a size of 0.1 μm (density: 1.05 g/cm³). The fresh stock solution of AgNPs was obtained from XFNANO Materials Tech (Nanjing, China). AgNPs were prepared in ultrapure water with a mass concentration of 0.1mg/mL and a size of 15 ± 5 nm. All the stocks of PS-MPs and AgNPs were stored at 4 °C in the dark. Other chemicals used in this study, such as AgNO3, were purchased from Sinopharm Group (Beijing, China). Ultrapure water was prepared by a Millipore Milli-Q integral water purification system (Millipore, Milford, USA).

Adult wild-type zebrafish (TU strain) were kindly provided by the Model Animal Research Center of Nanjing University, and then maintained in the laboratory for at least two weeks before exposure, according to previous procedures (Yan et al., 2019). All fish were cultured in a recirculation system with a temperature of 28 ± 1°C and a photoperiod of 14:10 h light: dark. The fish were fed with fresh brine shrimp two times per day.

The spawning of fish was carried out in a spawning aquarium for 24 h acclimatization and then stimulated by light. All fertilized eggs were collected via collection chambers and inspected using a stereomicroscope (SZX16, Olympus, Japan). The healthy fertilized eggs at 4 h post-fertilization (hpf) were used for further exposure.

All exposure solutions were freshly prepared in ultrapure water with ultrasonic treatment for 20 min to ensure uniform dispersion. The number concentrations of the PS nanoplastics prepared in test solutions with 10 mg/L were approximately 1.80 × 10¹⁰ particles/mL.

The structural morphology of PS nanoplastics, AgNPs, and their mixture in exposure solution was characterized by High-Resolution Transmission Electron Microscopy (TEM) (JEM-2100F, JEOL, Japan). The zeta potentials and z-average hydrodynamic diameters of the AgNPs alone and in combination with PS microplastics in the suspension solution during the exposure periods were also examined using a Zetasizer Nano ZS90 (Malvern, UK), as described by Calisi et al. (2022).

To evaluate the released Ag⁺ from different concentrations of AgNPs particles in different solutions before the exposure experiment as in previous studies (Osborne et al., 2013; van Aerle et al., 2013), 5 mL of exposure solutions were collected from each treatment and then passed through a 10 KDa ultrafiltration centrifuge tube (13600 g, 10 min). Since the pore size of the centrifuge tube is close to 1 nm, it can easily trap AgNPs and release Ag⁺ (Huang et al., 2019). The concentrations of the released Ag⁺ were determined by an inductively coupled plasma mass spectrometry (ICP-MS) (Thermon Fisher Scientific, USA).

All fertilized eggs were individually placed in a 24-well plate containing 2 mL of exposure solution. All exposure experiments were semi-static and the solutions were replaced daily with a replacement rate of 90% v/v to maintain stable exposure concentrations. All test solutions were maintained at pH 7.0 ± 0.1 during the exposure periods. The whole experiments were kept in illumination incubators at the temperature of 28 ± 1°C with a photoperiod of 14/10 h light/dark. During the exposure periods, the dead zebrafish embryos or larvae were removed promptly to prevent negative effects on other healthy individuals. Each exposure experiment was run in three independent replicates.

After exposure for 120 h, all zebrafish larvae were collected and stored in liquid nitrogen in time. The total RNA of each sample was isolated using a MiniBEST universal RNA extraction kit (TaKaRa, Japan). The quality of the extracted RNA was verified by gel electrophoresis and the ratio of absorbance at the wavelength of 260/280 nm. Then, the RNA was reversely transcribed to cDNA using the iScript cDNA synthesis kit (Bio-Rad, USA) according to the manufacturer’s instructions.

The expression of several important genes involved in oxidative stress (sod1, cat, and hsp70), apoptosis (bax, caspase9, and caspase3), innate immune (nfkβ, cebp, and il-1β), and metabolic function (mt2, mtf-1, cox1, and abcb4) in zebrafish were determined using qRT-PCR according to previous studies, and β-Actin was selected as the reference gene (Khan et al., 2019; Lu et al., 2022). The primer pairs were designed and synthesized by Shanghai Shenggong (Shanghai, China). The primer sequences of each target gene are listed in Table 1 . Quantitative real-time polymerase chain reaction (qRT-PCR) was then conducted using a Bio-Rad CFX96 Touch qRT-PCR determination system (Bio-Rad, USA). The PCR amplification program was set as pre-denaturation for 10 min at 95°C, denaturation for 15 s at 95°C, extension for 1 min at 60°C, and 40 cycles in total. The data were obtained as threshold cycle (Ct) values.

The relative expression of the target gene was quantified by the 2^-ΔΔCt method, where the calculation formula of ΔΔCt was as follows:

Data analyses were conducted with SPSS 20 statistical software (IBM, USA). All results were expressed in the form of mean ± standard deviation. The median lethal concentration (LC₅₀) value was calculated using Probit analysis at 48 and 96 h. The normality and homogeneity of the data were determined by Shapiro-Wilk and Levene tests, respectively. The Tukey post-test method after a one-way analysis of variance (ANOVA) was then used to determine the significance between different treatments. The difference was considered statistically significant with a p-value below 0.05.

As shown in Figures 1A, B , the AgNPs and PS nanoplastics were spherical particles and dispersible well in different exposure solutions without significant agglomeration. However, in the mixture treatment ( Figure 1C ), the AgNPs were mostly adsorbed on the PS nanoplastics. A similar result has been also reported by Li et al. (2020), who found that the presence of AgNPs in water solutions could be significantly captured on the surface of PS nanoplastics through a monolayer adsorption process.

To further verify the carrier effect of PS nanoplastics on the AgNPs, the zeta potentials and hydrodynamic sizes of AgNPs alone and in combination with PS nanoplastics during the exposure periods are summarized in Figures 1D, E . Both the AgNPs and PS nanoplastics were negatively charged, indicating the stability, polydispersity, and the existence of electric repulsion between them. The zeta potentials of the mixtures were sharply raised to -17.9 eV and then slowly reduced to -29.9 eV, which were also much higher than that of AgNPs and PS nanoplastics alone. The elevated zeta potentials may further reduce the electrostatic repulsion between particles, aggravating the agglomeration of AgNPs and PS nanoplastics. In addition, the hydrodynamic sizes of AgNPs and PS nanoplastics alone showed a stable behavior during the exposure periods, with an average size of 21.8 nm and 116.2 nm, respectively. In the mixtures, AgNPs and PS nanoplastics aggregated rapidly, with a hydrodynamic size high to 143.4 nm. It confirms that the presence of PS nanoplastics in aquatic environments could act as a carrier of AgNPs, which may further alter the behaviors and effects of AgNPs in aquatic environments.

The release of dissolved Ag⁺ from AgNPs at different concentrations in the presence of PS nanoplastics (10 mg/L) is shown in Table 2 . The limit of quantification of dissolved Ag⁺ in water was 0.04 μg/L. The release ratios of soluble Ag⁺ from AgNPs ranged from 0.6% to 4.23%, which increased with the increase of AgNPs concentrations. Similarly, van Aerle et al. (2013) reported that the highest dissolution of Ag⁺ for the AgNPs in different adopted exposure conditions was below 5%. A variable dissolution of AgNP was also demonstrated by Osborne et al. (2013), who found the release ratio of Ag⁺ from AgNPs in zebrafish embryo culture water varied between 0.1 and 2%. Hence, exposure to silver nitrate at 2.5 μg/L in the presence of PS nanoplastics was conducted to provide a likely worst-case scenario for the highest dissolution of Ag⁺ from the AgNPs in different exposure conditions.

The 48-h and 96-h acute toxicities of AgNPs, the mixture of AgNPs and PS nanoplastics, and the mixture of Ag⁺ and PS nanoplastics are listed in Table 3 . The results showed that the lethal concentration of 50% (LC₅₀) of AgNPs was 1.03 and 0.62 mg/L after exposure to 48 h and 96 h, respectively. This value was consistent with the reported acute toxicity of AgNPs with 14 different properties (Cunningham et al., 2013). After co-exposure to PS nanoplastics, the acute toxicity caused by AgNPs in zebrafish was not significantly altered. In addition, the values of TU for the AgNPs and PS nanoplastics mixtures were close to 1 (0.88-1.12), suggesting that the mixture followed a trend of additive effects. In previous studies, both the reduction and enhancement of PS nanoplastics on the toxic effects caused by nanomaterials on aquatic biota have been reported (Gunasekaran et al., 2020). These different alterations in acute toxicity may be attributed to the changes in the aggregative potentials between microplastics and nanomaterials, as well as the released ions from the adsorbed nanomaterials on microplastics. In this study, the 48 h and 96-h LC₅₀ of zebrafish treated with Ag⁺ and PS were ten-fold lower than that of AgNPs + PS nanoplastics, respectively. It suggests that the released Ag⁺ from the adsorbed AgNPs on PS nanoplastics may be supposed to enhance the acute toxicity of AgNPs in zebrafish, but the adsorption of AgNPs on PS nanoplastics may significantly alleviate the acute toxicity caused by the released Ag⁺. Similar results have been found by Lai et al. (2020), who found that co-exposure to nanomaterials markedly alleviated the acute toxicity induced by Ag⁺ in vivo. Therefore, it is necessary to study the influences of PS nanoplastics on the toxicity of AgNPs.

Oxidative stress is an imbalance between the production and removal of reactive oxygen species in organisms (Yan et al., 2018). In previous studies, the presence of both PS nanoplastics and AgNPs has been reported to induce the antioxidant defense in zebrafish (Choi et al., 2010). In this study, the expression of the antioxidase superoxide dismutase (sod1) gene in zebrafish was significantly reduced by AgNPs ( Figure 2 ). The inhibitive effect on the antioxidase caused by AgNPs suggests that the exposed fish may lose the capacity to scavenge oxidative damage through SOD, showing a compensatory inhibition. For the catalase (cat) gene, exposure to AgNPs did not alter its expression in fish, which may be also excessive consumption of CAT to scavenge oxidative damage. Generally, the SOD and CAT systems have been regarded as the first defense line to scavenge the reactive oxygen species (ROS) in biota. The induced ROS can be rapidly metabolized by SOD into H₂O₂, and then decomposed by CAT (Wen et al., 2018). However, the high large production of ROS may further suppress the antioxidase. Hence, the low expression of the antioxidase genes caused by AgNPs may associate with the high oxidative damage in fish. This result was then confirmed by the high expression of heat shock protein hsp70 in zebrafish ( Figure 2C ), which was 3.64-fold higher than that in the control.

After co-exposure to PS nanoplastics, the gene expression of the antioxidase (sod1 and cat) in zebrafish was significantly enhanced, with a 1.65 and 1.47-fold increase than that caused by AgNPs alone. Meanwhile, the presence of PS nanoplastics significantly reduced the transcription level of hsp70 gene stimulated by AgNPs, making it back to the control line. These results indicate that the presence of PS nanoplastics could weaken the oxidative stress caused by AgNPs on zebrafish through enhancing the antioxidase and inhibiting oxidative damage, showing a protective role. Similarly, Gunasekaran et al. (2020) found a dose-dependent increment in the ROS generation and lipid peroxidation in marine microalgae Dunaliella salina treated with ZnO particles, which were considerably reduced by the presence of plain PS nanoplastics. Given that PS nanoplastics could also stimulate oxidative stress on aquatic organisms (Ding et al., 2018), the oxidative damage of AgNPs on zebrafish was reduced by PS microplastics may be attributed to the formation of agglomeration. In addition, combined exposure to Ag⁺ and PS nanoplastics also resulted in similarly inhibitive effects on the oxidative stress on zebrafish when compared to that treated with AgNPs alone. It suggests that the released Ag⁺ from the adsorbed AgNPs on PS nanoplastics may be partially responsible for the oxidative stress on zebrafish stimulated by AgNPs alone, but the weak adsorption of Ag⁺ on PS nanoplastics may further alleviate the toxicity.

The expression of genes related to apoptosis (bax, caspase9, and caspase3) in zebrafish larvae are shown in Figure 3 . The results showed that the gene expression of bax in zebrafish was significantly up-regulated by AgNPs when compared to the control, as well as the gene expression of caspase9. When the PS nanoplastics were added, the gene expression of caspase9 was significantly decreased by more than 30%, while the bax was not significantly altered. When zebrafish were co-exposed to Ag⁺ and PS nanoplastics, the gene expression of bax and caspase9 were slightly up-regulated when compared to the control, but the expression of caspase3 gene was significantly enhanced.

Bax is a member of the Bcl-2 family mainly located in the cytoplasm (Choi et al., 2010; Ma et al., 2018). After early apoptosis, bax could transfer from the cytoplasm to mitochondria and assist with cytochrome C to open the pro-apoptotic pathway. The released cytochrome C may further activate caspase9 located upstream of the apoptosis pathway. In the downstream, Caspase3 could receive signals and eventually lead to cell apoptosis. Obviously, in this study, exposure to AgNPs and AgNPs + PS nanoplastics mainly stimulated the early apoptosis of zebrafish cells, causing damage to cell structure and function. However, the presence of PS nanoplastics inhibited the overexpression of caspase9 gene caused by AgNPs, which may be attributed to the hindrance on the release of cytochrome C in the mitochondrial-mediated apoptosis pathway. Hence, the presence of PS nanoplastics plays a protective role in the mitochondria apoptosis pathway caused by AgNPs. In addition, the positive treatment with the mixture of Ag⁺ and PS nanoplastics resulted in the activation of the whole apoptosis pathways, especially the downstream of caspase3, which implies the outcomes of apoptosis in zebrafish cells. The difference between the two mixture treatments once again confirmed that the presence of PS nanoplastics alleviated the apoptosis caused by AgNPs in zebrafish, which may result from the formation of aggregation between AgNPs and PS nanoplastics.

The innate immune system is the first line of defense against different pollutants and is also a prerequisite for enhancing the adaptive immune response (Xu et al., 2013). The oxidative damage caused by pollutants has been suggested as a pathway to induce immunotoxicity in organisms. Hence, it is not surprising to the AgNPs-induced immunotoxicity in zebrafish because of an increase in oxidative stress. The expression of genes related to innate immune in zebrafish embryos was investigated, including nfkβ, cebp, and il-1β, and the results are shown in Figure 4 . After exposure to AgNPs, all innate immune genes in zebrafish were significantly up-regulated by 1.7-fold, 1.57-fold, and 1.62-fold compared to the control group, respectively. Among these innate immune responses, the activated NFKβ could induce the production of pro-inflammatory cytokines and then induce inflammation in organisms. The cebp gene is involved in cell growth, proliferation, differentiation, cell cycle arrest, and apoptosis (Ramachandran et al., 2018). The il-1β gene is responsible for the accumulation of phagocytes in the infected area of pollutants in organisms, and subsequently activates macrophages and neutrophils to defend against damage (Brun et al., 2018; Aksakal and Ciltas, 2019). The increase in the expression of these genes in zebrafish indicated that the innate immune system of zebrafish larvae was under the stress of AgNPs. When the PS nanoplastics were co-added, the increased expression of these genes was all inhibited by 18%-23%, which was consistent with the enhanced gene expression of the antioxidase in zebrafish. Hence, the presence of PS nanoplastics may further weaken the AgNPs-induced immunotoxicity in zebrafish via inhibiting oxidative damage.

In the positive treatment, the relative expression of nfkβ, cebp, and il-1β gene in zebrafish was significantly up-regulated by 1.53 to 1.82 times of the control, which was also much higher than that in the mixture of AgNPs and PS nanoplastics. This finding indicates that the combined exposure to Ag⁺ and PS exerted greater pressure on the zebrafish immune system than that in the AgNPs + PS treatment. Previous studies have found that the presence of PS nanoplastics in waters could aggravate the toxicity of gold and copper ions and then trigger a significant immune response in zebrafish (Lee et al., 2019; Santos et al., 2022). However, the formation of hetero-aggregates between nanoplastics and AgNPs may further decrease the immunotoxicity in zebrafish caused by Ag⁺. The different sorption profiles of metallic ions and nanomaterials on nanoplastics may be responsible for the distinguishing combined effects on zebrafish.

The gene expression of metabolic function in zebrafish embryos treated with AgNPs alone or in combination with PS nanoplastics was studied, including mt2, mtf-1, cox1, and abcb4. As shown in Figure 5A , exposure to AgNPs stimulated the mt2 gene in zebrafish by 1.42-fold when compared with the control, which was significantly enhanced by the presence of PS nanoplastics with a 1.69 times increase. In addition, a 1.81-fold increase in the mt2 gene expression was also observed in zebrafish treated with Ag⁺ + PS nanoplastics compared to the control. In general, exposure to metal could induce the production of metallothionein in organisms and help to scavenge metal pollutants from the body through chelating with metal ions. The enhancement of PS nanoplastics on the production of metallothionein may imply a greater efficiency in the scavenging of Ag⁺ from zebrafish. This result has been also confirmed in discus fish (Symphysodon aequifasciatus) exposed to PS nanoplastics and metal cadmium (Wen et al., 2018). Similar to the expression of mt2 gene, the mtf-1 gene was also up-regulated in the three treatments, and the two mixture treatments exhibited a significant up-regulation trend ( Figure 5B ). Since the mtf-1 gene plays an important role in regulating the metal detoxification processes and is responsible for transcriptional activation of the mt gene (Ramachandran et al., 2018), the higher activation in mtf-1 gene caused by PS nanoplastics suggests a stronger scavenger of metal from zebrafish. The presence of PS nanoplastics may significantly trigger the metal detoxification function in zebrafish larvae.

Cytochrome c oxidase (cox) is mainly involved in the binding of metal ions and the electron transfer of mitochondria in the energy metabolism in organisms (Hu et al., 2016). The expression of cox1 gene in zebrafish is shown in Figure 5C . Compared with the control, exposure to AgNPs alone or in combination with PS nanoplastics induced a significant increase in cox1 gene by 38% and 25%, respectively. Similarly, Shen et al. (2020) found that the cox1 gene in fungi Fusarium solani was 41-fold up-regulated by AgNPs after 24 h of exposure. The energy metabolism processes related to cox1 gene in zebrafish mitochondria were also stimulated by AgNPs with different particle sizes and surface coatings (Liu et al., 2019). The activation of cox1 gene in zebrafish suggests a high stress on the mitochondrial metabolism caused by AgNPs, which was further declined by PS nanoplastics although the distinction was not obvious. On the contrary, the mixture of Ag⁺ and PS nanoplastics resulted in an inhibitive effect on the relative gene expression by 0.83-fold. This obvious distinction between AgNPs and Ag⁺ implies that the cox1 gene in zebrafish has a specific response to AgNPs but not the related ion. Dedeh et al. (2015) illustrated that the gene involved in the brain mitochondrial respiration (cox1) of zebrafish displayed a 3-fold increased expression in response to gold NPs compared to ionic gold. Hence, the expression of cox1 gene could be used as a specific biomarker to assess AgNPs stress.

ATP binding cassette (ABC) transporter in organisms plays a critical role in excreting toxic substances from cells and the efflux activity of exogenous pollutants in zebrafish cells is related to the gene of ABCB4 (Yan et al., 2018; Ma et al., 2020). In this study, the expression of the abcb4 gene in all treatments was significantly suppressed when compared to the control, while the differences between the three treatments were not significant ( Figure 5D ). The inhibition in the abcb4 gene suggests that the efflux capacity of zebrafish embryos was severely weakened in the presence of AgNPs and Ag⁺. Meanwhile, the co-existence of PS nanoplastics did not alter the efflux potential in zebrafish, which may further enhance the accumulation of exogenous pollutants in the body, leading to a serious risk to zebrafish.

In general, exposure to AgNPs induced an increase in oxidative damage, apoptosis, immunotoxicity, and metabolic capability in zebrafish embryos in most cases. The release of dissolved Ag⁺ from AgNPs may be partially responsible for the toxicity in zebrafish caused by AgNPs. The co-existence of PS nanoplastics in aquatic environments further suppressed these biological effects, playing a protective role in most cases. The variation in toxicity caused by PS nanoplastics may be partially attributed to the absorption of the released Ag⁺ from AgNPs, which could reduce the toxicity. In addition, the formation of agglomeration with AgNPs and PS nanoplastics would also alleviate the toxicity caused by AgNPs in zebrafish.