SNPs smaller than 35 nm are able to cross the blood-brain barrier . The size of SNPs commonly used in clinical practice is generally ~ 20 nm, and SNPs of this size have been used in neurotoxicity studies by several research groups [21, 22]. Therefore, 20 nm SNPs were selected in this paper to investigate their toxic effects.
In the current paper, the MTT results showed that 200 µM SNPs produced grade 1 cytotoxicity at 48 h of interaction, and the other concentrations of SNPs were noncytotoxic. To investigate the effects of noncytotoxic and cytotoxic concentrations of SNPs on PC12 quasi-neuronal networks and to compare our current findings with our previous work , three concentrations of SNPs (5, 100 and 200 µM) were selected for VTMM experiments.
Figure 4 suggests that 5 µM SNPs led to a significant increase in the electrical excitability of PC12 quasi-neuronal networks. And 100 µM SNPs led to an increase followed by a significant decrease in the electrical excitability of the networks. It also suggests that 200 µM SNPs caused a significant decrease in the electrical excitability of the networks after 1 h of treatment.
When comparing Fig. 2 and Fig. 4, it was evident that the noncytotoxic 5 µM SNPs led to a significant decrease (P < 0.05) in the VTh of PC12 quasi-neuronal networks at 0.5 h, while noncytotoxic 100 µM SNPs led to a significant increase (P < 0.01) in the VTh of the networks at 12 h. This result indicates that SNPs were still able to alter the electrical excitability of PC12 quasi-neuronal networks at noncytotoxic concentrations. Grade 1 cytotoxicity appeared after 48 h of treatment with 200 µM SNPs, yet the VTh was very significantly (P < 0.01) higher than the standard VTh after only 1 h, which indicated a reduction in the electrical excitability of PC12 quasi-neuronal networks. Taken together, these results show that the SNPs-induced changes in electrophysiological properties of PC12 cells appeared before the changes in cell viability, suggesting that using cell viability alone to evaluate nanoparticles-induced neurotoxicity is partial. Therefore, not only cell viability, but also electrophysiological properties should be considered when evaluating nanoparticles-induced neurotoxicity.
To further investigate the mechanisms of SNPs-induced cytotoxicity and changes in electrical excitability, changes in six aspects of PC12 cells, namely, neurite length, CMP difference, intracellular Ca2+ content, MMP difference, ROS content and ATP content were studied under the effect of 200 µM cytotoxic SNPs.
On the one hand, the presence of neurites is a prerequisite for the formation of synapses, which are the basis of signaling between neurons. In addition, a relatively high density of voltage-gated sodium channels, which play an important role in the production and conduction of neural signals, are distributed along the axon initiation segment of neurons . Decreased neurite length might shorten the axon initiation segment, therefore leading to a decrease in the number of sodium channels, which in turn would affect neurotransmitter release and reduce the electrical excitability of neuronal networks . On the other hand, decreased neurite length correspond to cell damage, particularly cytoskeletal damage . Thus, the SNPs-induced decrease in neurite length (Fig. 5) suggests that SNPs caused cytoskeletal damage and this decrease might affect signaling between neuron-like PC12 cells.
The CMP is one of the most important indicators of cell survival, as a decrease in CMP difference is usually accompanied by toxic effects, apoptosis and necrosis . Additionally, the CMP difference directly influences the resting potential and polarization of neurons, rendering it a pivotal factor affecting the electrical excitability of neurons . Within 0–24 h, noncytotoxic concentrations of SNPs induced a sustained increase in CMP difference (Fig. 6), which indicated cell hyperpolarization , thereby contributing to reduced electrical excitability (Fig. 4). At 48 h, a significant decrease in CMP difference was observed compared to that at 24 h (P < 0.05). This might be due to the cytotoxicity of SNPs at 48 h, as cell death is usually accompanied by a decrease in CMP difference .
The intracellular Ca2+ contents in all SNPs-treated groups were higher than those in the control group (Fig. 7). These results are consistent with the finding of Elżbieta Ziemińska et al. that 75 µg/mL SNPs that were 5–35 nm in size could lead to elevated Ca2+ levels in cerebellar granule cells due to overactivation of Ca2+ channels . Moreover, the intracellular Ca2+ content at 48 h was significantly lower (P < 0.05) than that at 24 h. This might be attributed to the very significant increase in intracellular Ca2+ content at 24 h, which initiated negative feedback regulation, followed by inactivation of Ca2+ channels , resulting in a decrease in intracellular Ca2+ content at 48 h. Overload of intracellular Ca2+ leads to mitochondrial and cytoskeletal damage and even apoptosis [32, 33]. Moreover, Ca2+ is an important intracellular signaling molecule that can control neurotransmitter release, regulate the expression of various proteins and influence the excitability of neurons . In the current paper, 200 µM SNPs applied to PC12 cells for 48 h were found to simultaneously induce grade 1 cytotoxicity (Fig. 2), very significantly increase VTh (Fig. 4) and very significantly increase intracellular Ca2+ content (Fig. 7).
The mitochondrion is an important organelle in the cell, as it is the site of energy production and one of the important targets of nanoparticles to induce toxicity . The MMP difference is an important indicator of mitochondrial function. A decreased MMP difference implies mitochondrial damage and is one of the early signs of apoptosis . Additionally, impaired mitochondria can lead to a decrease in ATP (an important cellular energy source) content, which subsequently can lead to cellular energy deficiency . A previous study by our group found that the ATP content of human dermal fibroblasts decreased under the effect of SNPs . Moreover, electrons escaping from a damaged mitochondrial electron transport chain might directly react with substances such as oxygen and generate ROS . Damage to mitochondria therefore leads to increases in ROS content, and excessive ROS levels might contribute to further cellular damage (e.g., DNA, protein, and synaptic damage) . Thus, ROS content is one of the most important indicators of cellular damage and the effect of nanoparticles on cells. The results displayed in Table 6 and Fig. 8 indicate that treatment with SNPs for 24 and 48 h caused mitochondrial damage which reduced the MMP difference, contributing to decreases in ATP content and increases in ROS content. Additionally, the MMP difference and ATP content of cells exposed to SNPs were very significantly higher (P < 0.01) than those of the control groups at 0.5 h of treatment, which could have been caused by the metabolic adaptation to the presence of SNPs. The metabolic adaptation might increase the MMP difference and enhance tricarboxylic acid cycle of mitochondria, resulting in increased ATP production .
The three indicators most highly correlated with cell viability in Table 7 were ATP content (r = 0.95), neurite length (r = 0.93), and ROS content (r=-0.90). SNPs have been found to simultaneously decrease ATP content and suppress neurite growth in human embryonic stem cell-derived neurons , and to decrease MMP difference, increase ROS content and decrease viability in A549 cells . The current paper demonstrated that ATP content, neurite length and ROS content were important indicators of cellular damage caused by nanoparticles, and that the main cause of SNPs-induced cytotoxicity was the detrimental effects on cellular energy supply, cytoskeletal integrity and the ROS content.
Elevated intracellular Ca2+ content have been found to contribute to reduced ATP synthesis  and cellular energy deficiencies, which in turn open ATP-sensitive potassium (KATP) channels on the cell membrane, resulting in increases in CMP difference and hyperpolarization . This paper showed that SNPs increased intracellular Ca2+ content (Fig. 7), decreased ATP content (Fig. 5), increased CMP difference (Fig. 6) and increased VTh (Fig. 4) in PC12 cells. Furthermore, the three indicators most highly correlated with VTh (displayed in Table 7) were intracellular Ca2+ content (r = 0.96), ATP content (r=-0.92) and CMP difference (r = 0.91). The above results suggest that the SNPs-induced decrease in electrical excitability may be explained by a decrease in ATP content due to an increase in intracellular Ca2+ content, which led to cellular energy deficiency that opened KATP channels on the cell membrane, resulting in an increase in CMP difference and hyperpolarization.
Additionally, ATP content was the only cytological indicator that correlated with both cell viability and VTh, with correlation coefficients above 0.9. This result indicates that the ATP content was the main cytological indicator that affected both cytotoxicity and electrical excitability in the presence of SNPs, and illustrates the importance of energy supply for the maintenance of neuronal cell structure and function.
Possible mechanisms for the SNPs-induced changes in cytotoxicity and electrical excitability of PC12 cells are summarized in Fig. 9. SNPs decreased neurite length in PC12 cells, suggesting that SNPs caused cytoskeletal damage  and cytotoxicity. In addition, decreased neurite length might lead to a reduction in voltage-gated sodium channels and diminish the electrical excitability of PC12 quasi-neuronal networks . The SNPs-induced decrease in MMP difference led to an increase in ROS content, which might damage biomolecules such as DNA and proteins, leading to cell death . Moreover, increased ROS content might impair synaptic structures, which could affect intercellular signaling  and reduce the electrical excitability of PC12 quasi-neuronal networks. Both the decrease in MMP difference and the increase in intracellular Ca2+ content could lead to a decrease in ATP content. Decreased ATP content could lead to cellular energy deficiency, which both could activate apoptotic pathways , causing a decrease in cell viability, and could open KATP channels , causing an increase in CMP difference and hyperpolarization, eventually resulting in reduced electrical excitability.