The tested neonicotinoids at sublethal concentrations (1/32–1/2 LC₅₀) showed no significant effect on the survival of zebrafish embryos. However, the AChE activity followed a dose–effect relationship of being significantly decreased by increasing the neonicotinoid concentrations, which indicated that the neonicotinoids are potentially neurotoxic to zebrafish embryos ( Supplementary Figure S1 ). Sigmoidal dose–response curves (variable slope) were fitted with a four-parameter logistic equation to obtain EC₅₀ ( Supplementary Figures S6A–E ). One-way ANOVA was also performed to determine the NOEC/LOEC of neonicotinoids inhibiting the AChE activity of zebrafish embryos at four temperatures ( Table 2 ). At 28°C, the EC₅₀ of cycloxaprid, imidaclothiz, nitenpyram, paichongding, and thiamethoxam AChE activity was 0.998 (0.786, 1.244), 52.55 (29.74, 189.8), 0.957 (0.546, 2.184), 0.695 (0.019, 1.442), and 24.33 (17.09, 32.7) mg/L, whereas at 33°C the EC₅₀ of these neonicotinoids was 0.693 (0.595, 0.806), 6.57 (2.10, 11.15), 0.494 (0.416, 0.587), 0.503 (0.137, 0.824), and 128.0 (96.21, 163.2) mg/L, respectively. However, it is worth mentioning that the inhibitory effect of cycloxaprid, nitenpyram, and paichongding on the AChE activity of zebrafish embryo was significantly greater than those of imidaclothiz and thiamethoxam.

As nicotinic acetylcholine receptor (nAChR) agonists, neonicotinoids selectively bind to the nAChR and compete with the insect’s acetylcholine, thus blocking the normal transmission between neurons and thereby causing paralysis and the ultimately death of the organism (Mitchell et al., 2017). AChE is a multi-molecule glycoprotein that hydrolyzes acetylcholine and can terminate the transmission of acetylcholine in synapses. It exists in nerve tissue, liver, muscle, red blood cell, or spleen in animals and plays an important role in nerve conduction and functional regulation of the neurotransmitter. Changes in the activity of AChE are sensitive toxicological indicators for pesticide pollution in an aquatic environment, which can be further used as molecular biomarkers for toxicity evaluation and risk assessment as well as in the environmental monitoring of pesticides in water (Lionetto et al., 2013; Gravato et al., 2021). Neonicotinoids particularly inhibit the AChE activity, resulting in the accumulation of acetylcholine which causes inhibition in the normal functioning of the central nervous system (Lu et al., 2018). According to pioneer studies, neonicotinoids can cause neurotoxic effects (e.g., motor behavior) to invertebrates and aquatic vertebrates by inhibiting the AChE activity, e.g., exposure of imidacloprid resulted in decreased AChE activity in Chironomus riparius (Azevedo-Pereira et al., 2011), Eisenia foetida (Wang et al., 2015), and Caenorhabditis elegans (Meyer and Williams, 2014) as well as in the brain of D. rerio (Guerra et al., 2021), O. mykiss (Topal et al., 2017), and Rhamdia quelen (Marins et al., 2021). In the muscle of Astyanax altiparanae (Almeida et al., 2021), sulfoxaflor exposure can also significantly decrease AChE enzyme activity in the brain and muscle tissues of zebrafish (Piner Benli and Çelik, 2021). In this study, the inhibitory effect on AChE activity of zebrafish embryos indirectly reflects the neurotoxic effect of neonicotinoids on zebrafish, which can be used as a quantitative indicator to evaluate the sublethal toxicity of neonicotinoid pesticides (Sheets et al., 2015).

Compared with the control group, under four temperatures (20, 23, 28, and 33°C), the ATPase activity of zebrafish embryos was slightly increased in the low concentration (p > 0.05) and significantly decreased (p < 0.05) in the high concentration of the six neonicotinoid treatment groups ( Supplementary Figure S2 )—for example, at 33°C, the ATPase activity was decreased in 1 mg/L and increased in 0.5 mg/L, whereas the enzyme activity was increased at other temperatures in 0.125 and 0.25 cycloxaprid treatments ( Supplementary Figure S2A ). At 20 and 33°C, the enzyme activity was decreased in 6.25 mg/L and decreased at two other temperatures in 12.5- and 25-mg/L treatments of imidaclothiz ( Supplementary Figure S2B ). Furthermore, at all four temperatures, the ATPase activity was activated by nitenpyram exposure in 0.125 and 0.25 mg/L and decreased in the 0.5-, 1-, and 2-mg/L treatment groups ( Supplementary Figure S2C ). Meanwhile, at 23°C, the ATPase activity was promoted in 0.625-mg/L treatment and inhibited by 0.625-, 1.25-, and 2.5-mg/L paichongding treatments at three other temperatures ( Supplementary Figure S2D ). However, at 20 and 28°C, the ATPase activity was enhanced in 0.625 mg/L, followed by a decrease at 250 and 500 mg/L, and inhibited at 23 and 33°C by 0.625- and 1.25-mg/L treatments of thiamethoxam ( Supplementary Figure S2E ).

ATPase (divided into Na⁺K⁺–ATPase and Ca²⁺Mg²⁺–ATPase), a protease in the membrane of a cell or an organelle, can regulate the membrane potential and maintain the intracellular ion gradients which are essential for signal transduction, excitability, conductivity, and release of neurotransmitters in nerve cells (Ji et al., 2009). Na⁺K⁺–ATPase is a sodium pump on the cell membrane, which can transport Na⁺ and K⁺ to maintain the normal physiological function of cells. When Na⁺K⁺–ATPase activity is inhibited, the intracellular and extracellular Na⁺ concentration will be imbalanced, and excessive intracellular Na⁺ can inhibit the uptake of excitatory amino acids in the synaptic membrane (Davis and Wedemeyer, 1971; Farley et al., 1984; Agrahari and Gopal, 2008), whereas Ca²⁺Mg²⁺–ATPase is a calcium pump on the membrane, which can hydrolyze the ATP to maintain a low intracellular Ca²⁺ concentration which may help in maintaining cell homeostasis. When the activity of Ca²⁺Mg²⁺–ATPase inhibits, the activity of a large number of intracellular enzymes will be reduced, and this will cause an increase of free radical production, ultimately leading to peroxidation damage of the membrane, destruction of the mitochondria, and cell death (Al-Rajhi, 1990; Blasiak, 1995; Li et al., 2006). The inhibition in energy conversion—such as inhibition in the activity of mitochondrial enzyme—may be the key component of neonicotinoid insecticide mechanism (Kumar et al., 2018; Li et al., 2022; Sevgiler and Atlı, 2022; Veedu et al., 2022). The results of our study indicated that the high-concentration exposure of neonicotinoids inhibited the ATPase activity in zebrafish embryos, which may cause the functional disturbance of the membrane ion pump (e.g., sodium pump and calcium pump). Because the dysfunction of a membrane ion pump leads to a massive outflow of K⁺ and Mg²⁺, there will be a massive influx of Na⁺, Ca²⁺, and H₂O into the cell, resulting in the retention of water and sodium, energy metabolism disorders, and mitochondrial dysfunction, ultimately leading to nerve cell damage and dysfunction of the central nervous system (Suvetha et al., 2010; Serafini et al., 2019; Oyovwi et al., 2021). However, at present, the mechanism which demonstrates how neonicotinoid induces the inhibition of ATPase activity is not clear. It could be either the direct inhibition of ATPase after neonicotinoid pesticides enter into the cell or by indirectly inhibiting the activity of ATPase by structural and functional disorders of the biomembrane through oxidative/peroxidation damage, or both possibilities could have happened (Chauhan et al., 2002; Akbar et al., 2016; Kumar et al., 2018).

Oxidative stress refers to oxidative damage caused by the accumulation of reactive oxygen species (ROS) and reactive nitrogen species when there is an imbalance between the generation and the elimination of oxygen-free radicals in the body or cells, resulting in the damage of lipids, amino acids/proteins, nucleic acids, and chromosomes and, ultimately, cell death (Droge, 2002; Deng et al., 2021). Antioxidant enzymes, including superoxide dismutase, catalase, peroxidase, GPX, and GST, are important indexes to detect oxidative stress (Lacourse et al., 2009). GPX can remove oxygen metabolites and maintain the activity of glutathione (GSH). GST is a dual-function enzyme acting as an antioxidant and phase II detoxification enzyme. The combination of GSH with various toxic substances and their metabolites by catalyzing them into hydrophobic substances which are excreted in the form of urine or bile is a vital mechanism of GSH. This mechanism helps GSH to achieve the detoxification effect of degrading and discharging various toxic substances in the body. Meanwhile, GST can also repair macromolecular proteins, DNA, and other substances, remove lipid peroxides, and inhibit lipid peroxidation, which can protect the body from damage by exogenous stimulation (Ullah et al., 2022). Thus, GPX and GST are commonly known as biochemical indicators or biomarkers for evaluating the oxidative damage of environmental pollutants (Pan et al., 2022; Ullah et al., 2022).

In this study, at four temperatures (20, 23, 28, and 33°C), the GPX activity was promoted or not affected by the lower concentrations of the six neonicotinoids compared with the higher concentrations which cause the inhibition in GPX activity ( Supplementary Figure S3 )—for example, at 20, 23, and 28°C, the GPX activity was enhanced in 0.125-, 0.25-, and 0.5-mg/L concentrations and inhibited in 1- and 2-mg/L concentrations of cycloxaprid ( Supplementary Figure S3A ); at all four temperatures, enzyme activity was inhibited in 6.25-, 12.5-, and 25-mg/L imidaclothiz concentrations ( Supplementary Figure S3B ); at 20 and 33°C, enzyme activity increased in 0.125-mg/L concentration and decreased in 0.5- and 1-mg/L concentrations of nitenpyram ( Supplementary Figure S3C ); at 20°C, the enzyme activity was promoted in 0.625-, 1.25-, and 2.5-mg/L paichongding concentrations and inhibited in 10-mg/L concentration ( Supplementary Figure S3D ); and at 20, 23, and 28°C, enzyme activity was activated in 62.5- and 125-mg/L concentrations and inhibited in 250- and 500-mg/L concentrations of thiamethoxam ( Supplementary Figure S3E ).

These findings demonstrated that the inhibition of GPX activity by cycloxaprid and nitenpyram is much stronger than by the other neonicotinoids (especially imidaclothiz and thiamethoxam). The overall effect of neonicotinoids demonstrated a strong dose–response relationship between GPX activity and neonicotinoid concentrations. At low-concentration exposures of the six neonicotinoids, the GPX activity was generally promoted, whereas at higher concentrations the enzyme activity was inhibited. One possible reason could be the excessive production of ROS after neonicotinoids entered the embryos (Shukla et al., 2017). GPX can participate in controlling ROS homeostasis by regulating the intracellular H₂O₂ levels, antioxidant enzyme activities, and gene transcription levels and by maintaining the GSH/GSSG balance (Zhang et al., 2020). In turn, GPX activities can be regulated by ROS level. An increased concentration of intracellular ROS caused by external stimulus (neonicotinoid exposure) can increase the expression level of the gpx gene and thus promote GPX activities (Li et al., 2015), whereas with increased pesticide concentration and continuous action on cells, excessive ROS cannot be eliminated by the GPX enzyme, leading to the accumulation of ROS, which will oxidize the amino acid residues of the GPX enzyme protein to produce a carbonyl group and change its secondary structure (reducing its helicity) and thus reducing the enzyme activity (Lin et al., 2021). Meanwhile, the excessive ROS can damage lipid, DNA, protein, and other biological macromolecules as well as the structure and functions of membranes, which can directly or indirectly affect the synthesis of the GPX enzyme (Tian et al., 2020; Benchaâbane et al., 2022).

The findings of our study demonstrated that the GST activity was activated by neonicotinoids in a dose-dependent manner ( Supplementary Figure S4 ). It can be noticed that, at 20°C, the GST activity was increased by 0.125 mg/L cycloxaprid, 12.5 mg/L imidaclothiz, 1 mg/L nitenpyram, 2.5 mg/L paichongding, and 62.5 mg/L thiamethoxam, while at 33°C the enzyme activity was promoted by 0.5 mg/L cycloxaprid, 50 mg/L imidaclothiz, 0.25 mg/L nitenpyram, 1.25 mg/L paichongding, and 125 mg/L thiamethoxam. It should be noted that the fluctuations in GST activity may be caused by oxidative stress and the detoxification is induced by pollutant stimulation because the activity of GST could be activated to catalyze the binding of GSH with lipid peroxides or other harmful metabolites produced by oxidative damage and exert antioxidant and detoxification effects (Qi et al., 2013; Benchaâbane et al., 2022).

Dose–response curve and one-way ANOVA were employed to obtain the EC_10/50 and the LOEC/NOEC of neonicotinoids, which revealed that the GST activity was promoted in zebrafish embryos at different temperatures ( Table 2 and Supplementary Figures S6F–J ). At 28°C, the EC₅₀ of GST activation induced by cycloxaprid, imidaclothiz, nitenpyram, paichongding, and thiamethoxam was 0.9619 (0.5409, 1.992), 17.91 (9.866, 94.74), 0.9780 (0.5363, 1.870), 2.300 (1.842, 14.08), and 2,886 (1,440, 9,775) mg/L, respectively, whereas the oxidative damage by cycloxaprid and nitenpyram to zebrafish embryo was stronger compared that of with thiamethoxam. Previous studies have shown that antioxidant and detoxification enzymes are crucial in maintaining cell homeostasis and can used as a biochemical marker to evaluate the potential toxicity of the contaminants—inducing oxidative stress in aquatic organisms, especially in fish (Santana et al., 2018; Gallego-Ríos et al., 2021; Solomando et al., 2022). In this study, the changes of antioxidant enzyme (GPX) and detoxification enzyme (GST) revealed that the neonicotinoids induce oxidative damage to zebrafish embryo. These responses can be used as molecular biomarkers to evaluate the oxidative stress and toxicity mechanisms of neonicotinoids or other pollutants.

Overall, the neonicotinoids potentially caused neurotoxicity and CNS dysfunction by AChE inhibition. Furthermore, neonicotinoid toxicity can cause the peroxidation of biological macromolecules including lipid, protein, and DNA by inducing oxidative damage. This phenomenon can destroy the structure and function of biological membranes, including those of the mitochondrial membrane with the potential inhibition of respiratory enzyme activity, ultimately leading to cellular metabolic disorders (Meiliana and Wijaya, 2012). This implies that neurotoxicity, oxidative damage, and metabolic disorders may be the key components of MOA of neonicotinoid insecticides. These biomarkers—including AChE, GST, GPX, and ATPase activity at the molecular/cellular level—can be utilized as signal indicators for monitoring pesticide contamination in aquatic environments at the early stages.

Temperature can profoundly influence the physiological and ecological performance of organisms such as during embryonic development—including growth and development (Brett, 1979), malformation and death (Kamler et al., 1998), adaptations and acclimation (Burleson and Silva, 2011), resource acquisition and behavior (Van der Have, 2002), metabolic and ventilation rate (Vinagre et al., 2014), and survivorship and fecundity (Frazier et al., 2006; Angilletta et al., 2009; Kingsolver, 2009; Amarasekare and Savage, 2012). In this study, the hatching rate of zebrafish embryos which are maintained as the control group at four temperatures was recorded for 192 h. The hatching curves of zebrafish embryos at different temperatures (20, 23, 28, and 30°C) were plotted and demonstrated in Supplementary Figure S7 . The results showed that embryos started hatching at 20 and 23°C within 108 and 84 h and completely hatched after 192 h at both of these temperatures, while at the two other temperatures (28 and 33°C) the embryos started hatching at 36 h and completely hatched after 72 h. These findings clearly indicated that a low temperature has an inhibitory effect on embryo hatching, and a higher temperature can reduce the hatching time of embryos. Meanwhile, in the control group, the enzyme activities of AChE, ATPase, GPX, and GST were highest at 28°C, decreased at 33°C, and significantly inhibited at low temperatures of 20 and 23°C ( Supplementary Tables S2–S5 ). As demonstrated in Supplementary Figure S7 , low temperature inhibited the enzyme activity and high temperature enhanced the enzyme activity. Therefore, it can be assumed that temperatures below or above T _opt can inhibit embryonic development.

In this study, a significant interaction was found between various concentrations of the tested neonicotinoids and the embryonic responses (lethal including mortality and sublethal like changes of AChE, ATPase, GPX, and GST activities) of zebrafish at different temperatures. Furthermore, the findings of this study also demonstrated that the interaction between neonicotinoid toxic concentrations and temperature significantly influenced the embryonic mechanisms ( Supplementary Table S6 ).

Nonlinear regression curves were plotted on the toxicity threshold values of lethal (LC₅₀) and sublethal effect (EC₅₀) of neonicotinoids at different temperatures and fitted by Gaussian distribution ( Figure 1 ). Significant variations in LC₅₀ and EC₅₀ can be noticed at variable temperatures. The effect of temperature on neonicotinoid toxicity showed an inverted V-shaped relationship, where the toxicity decreased with the increase of temperature and then increased with the increase of temperature after exceeding the T _opt. The temperature corresponding to the highest amplitude of the Gaussian curve represents the T _opt, at which toxicity is the least (toxicity is enhanced when the temperature is above/below the T _opt). T _opt values were also obtained at 23.5, 22.3, 23.8, 23.6, 21.9, and 22.9°C for cycloxaprid, dinotefuran, imidaclothiz, nitenpyram, paichongding, and thiamethoxam, respectively, when considering acute lethal toxicity ( Table 3 ), while for the sublethal effects of AChE inhibition, T _opt values were derived at 26.3, 27.3, 25.4, 23.8, and 24.0°C for cycloxaprid, imidaclothiz, nitenpyram, paichongding, and thiamethoxam, respectively. Temperature can affect the toxicity of neonicotinoids (e.g., imidacloprid and thiacloprid) on the embryonic level by enhancing the respiratory efficiency or energy metabolism (Osterauer and Köhler, 2008; Scheil and Köhler, 2009). It is generally accepted that elevated temperatures tend to increase the rate of chemical contaminant uptake by increasing the metabolic and ventilation rate of ectotherms, hence boosting the bioaccumulation of chemicals in body tissues (Cherkasov et al., 2006; Tsui and Wang, 2006; Cherkasov et al., 2007; Schiedek et al., 2008). This mechanism ultimately leads to a higher chemical toxicity at higher temperature levels (Cairns et al., 1975; Bourgeault et al., 2013).

It is commonly reported that chemical toxicity increases with increasing temperature (linear model) (Leung et al., 2000; McConnachie and Alexander, 2004; Kwok and Leung, 2005), but a non-linear function (e.g., inverted V-shaped model) tends to describe the relationship between temperature and chemical toxicity more adequately with wider temperature ranges as demonstrated in further comprehensive toxicity studies (Heugens et al., 2003; Li et al., 2006; Bao et al., 2008; Osterauer and Köhler, 2008; Prato et al., 2008; Wong and Leung, 2013). Following the inverted V-shaped model, chemical toxicity to aquatic organisms was lowest at T _opt and increased at a temperature below or above T _opt. In general, aquatic organisms living in optimal conditions are more tolerant to chemical toxicity than those living in conditions near their thermal tolerance limits (Heugens et al., 2001; Li et al., 2014). Every aquatic organism has a thermal tolerance range (Cairns et al., 1975; Cairns et al., 1978); when aquatic ectotherms (e.g., fish) generally live within an optimum range of temperatures, they have sufficient energy and oxygen supply and strong activities of detoxification enzymes and antioxidant enzymes (Pörtner, 2010), resulting in a high clearance rate of exogenous toxins and low sensitivity to their toxicity (Frederich and Pörtner, 2000). When the temperature exceeds (higher/lower than) the optimal range, its internal energy budget cannot maintain its normal growth state (energy deficit) (Sokolova and Pörtner, 2007), the tolerance of the organism is reduced, and such a thermal stress at temperature extremes may further enhance the toxicity of a chemical (Jusup et al., 2016). Therefore, such an integrative inverted V-shaped response can be typically explained by the temperature-regulated mechanisms (Mu et al., 2018).