Titanium Dioxide Nanoparticles Induce Cognitive Dysfunction of Rats with Endogenous Protective Mechanism of Increased miR-21-5p

Background Titanium dioxide nanoparticles (TiO 2 NPs) cause nerve cell damage and central nervous system dysfunction has been well known. However, most of the cognitive decits occurs before the observable damage to nerve cells and the regulatory mechanism is largely unknown. Mitochondria and energy metabolism are the targets of many neurotoxic substances. Here, we showed that TiO 2 NPs exposure reduced the mitochondrial membrane potential (ΔΨm) and ATP content, which is the important cause of cognitive decits, and upregulated microRNA-21-5p (miR-21-5p) levels in rat neuron cells. Upregulated miR-21-5p promoted phosphorylation of c-Jun, consequently promoting transcription of VDAC1 (encoding voltage-dependent anion channel 1), which in turn preserved the ΔΨm and ATP content of neuron cells. the VDAC1 in neuron in response to TiO 2 NPs. < 0.05). (F) Potential transcription factor binding sites in the VDAC1 promoter region (-2000 ~ -908) were analyzed using the TRANSFAC 8.3 engine online. (G) SH-SY5 cells were treated with 20 μg·mL-1 TiO2 NPs for 9 hours. The cell extracts were analyzed for the levels of potential transcription factors of VDAC1 using western blotting. (H-J) SH-SY5 cells and SH-SY5 cells (anti-c-Jun) were treated with 1% O2 for 6 hours. The cell extracts were analyzed for the levels of VDAC1 mRNA and ATP content. Changes in the mean JC-1 uorescence ratio (590 nm/530 nm) of cells were detected to indicate the ΔΨm levels.


Introduction
Titanium dioxide nanoparticles (TiO 2 NPs) have high chemical stability, and excellent optical properties, thus are used widely in various elds, such as cosmetics production, food processing, paint manufacturing, etc. [1]. TiO 2 has been approved by the US Food and Drug Administration for food production as a color additive many years ago [2,3]. It has been reported that over 93% of the TiO 2 in gum comprises TiO 2 NPs [2]. TiO 2 NPs also has been found in many foods that claim not to contain TiO 2 NPs, and previous tests have shown up to 9 g/kg of food [4]. There is increasing awareness of the health risks caused by the absorption of TiO 2 NPs into the human body. The central nervous system (CNS) is one of the important targets for TiO 2 NPs [5]. The digestive system is an important way for TiO 2 NPs to enter the human body. After entering the blood through the digestive tract, TiO 2 NPs can enter the central nervous system through the blood-brain barrier due to its small size [6,7], where they damage nerve cell function by promoting in ammation, activating oxidative stress, and destroying cell membrane [8].
It has been reported that TiO 2 NPs induced in ammatory injury in SH-SY5Y cells (Cells derived from human bone marrow neuroblastoma) in a dose-dependent manner [5]. In vivo, TiO 2 NPs entering the central nervous system (CNS) can lead to mitochondrial dysfunction and decrease of mitochondrial membrane potential (ΔΨm) [9]. In the CNS, the energy metabolism level determines the brain's ability to perform its normal functions, such as learning and memory [10,11]. It has been reported that TiO 2 NPs could cause ΔΨm loss, hippocampal damage, and impairment of spatial recognition memory in mice, which indicated that TiO 2 NPs promote apoptosis of hippocampal neurons through the mitochondrial pathway [12]. So far, dysfunction of the central nervous system is believed to be the result of neuronal cell damage induced by TiO 2 NPs, which promote energy metabolism disorders. However, most of the central nervous system dysfunction occurs before the observable damage to nerve cells. This means that when the damaging effect is observed, the best time to prevent it has been missed. Therefore, revealing the molecular mechanism of TiO 2 NPs-induced energy metabolism disorder will help to prevent the damaging effect of the nervous system earlier.
Our present study de ned a novel mechanism of energy metabolism damaged by TiO 2 NPs in neuron cells. Upregulated miR-21-5P, induced by TiO 2 NPs, preserved TiO 2 NPs-induced the decline of ΔΨm and ATP content through increasing oxygen consumption rate (OCR). Upregulation of miR-21-5P promoted transcription of VDAC1 (encoding voltage-dependent anion channel 1) leading to the upregulation of VDAC1 protein levels, which in turn preserved the OCR, ΔΨm, and ATP content of neuron cells under TiO 2 NPs exposure. This endogenous self-protection mechanism in the process of TiO 2 NPs-induced energy metabolism disorder in neuron cells could improve our understanding of TiO 2 NPs-induced nervous system dysfunction and help us to prevent TiO 2 NPs-induced CNS injury earlier.

Animals and treatments
This study was conducted with the approval of the Animal Care and Use Committee of the Tianjin Institute of Environmental and Operational Medicine, China. All rats handling procedures were performed in strict accordance with the guide for the use and care of laboratory animals. All surgery was performed under sodium pentobarbital (1%, 1 mL/100 g weight, i.p.) anesthesia.
Before the experiment, male Wistar rats (180-200 g) were kept in their surroundings for 1 week. The rats were maintained on a 12 h light/dark cycle in constant temperature environment (25 ± 2 °C). Animals were randomly divided into four groups: Control group (treated with 0.5% w/v hydroxypropyl methylcellulose (HPMC)) and three experimental groups (10, 20, or 30 mg/kg body weight (BW) TiO 2 NPs). The TiO 2 NPs (Sigma-Aldrich, St. Louis, MO, USA; 637254) suspensions were given to the rats via intragastric administration every day for 4, 6, or 8 weeks, respectively. Animals were weighed before intragastric administration every day.

ATP content
Cold phosphate-buffered saline (PBS) was used to collect cells, and the samples were quickly frozen and preserved in liquid nitrogen. The samples were melted slowly in an ice water bath and vortexed for 10 s. Luciferase driven bioluminescence with an ATP Bioluminescence Assay Kit HS II (Sigma-Aldrich, Cat. No. 11699709001) was used to assay the ATP content.

Seahorse
The OCR and ECAR were measured as described previously [14] using a Seahorse XFe24 system (Agilent Technologies, Santa Clara, CA, USA).

Luciferase Reporter Assay
Luciferase reporter assays were conducted as previous reports [15].

Western Blotting
Western blotting was conducted as described in our previous publication [16].

Statistical Analysis
SPSS 17.0 software (IBM Corp., Armonk, NY, USA) was used to perform statistical analyses. The differences between two groups were determined using Student's t-test. For three and more groups of data, statistical analysis was performed using analysis of variance (ANOVA) followed by the Tukey-Kramer multiple comparisons test or an unpaired two-tailed Student-t-test. A P value less than 0.05 was considered statistically signi cant.

Results
3.1 Upregulation of miR-21-5P expression, mitochondrial dysfunction and cognitive de cits were observed in response to TiO 2 NPs.
As shown in Fig. 1, increased levels of miR-21-5P were detected in prefrontal cortex of rat ( Fig. 1A & 1D), primary neuron cells (Fig. 1B & 1E), and SH-SY5Y cells (Fig. 1C & 1F) in response to TiO 2 NPs. The prefrontal cortex requires a large amount of energy, and multiple information processing functions, such as memory, judgment, analysis and thinking, which are involved in preserving fundamental cognitive ability, require energy in the form of ATP. Regulation of the energy balance might be a key factor involved in neuroprotection against TiO 2 NPs. Thus, we determined the ΔΨm and ATP content and found that the ΔΨm and ATP levels decreased markedly in response to TiO 2 NPs in a dose-dependent manner (Fig. 1G &   1H). Water maze test showed that the latency of rats to locate the hidden platform was signi cantly prolonged after titanium dioxide intervention (Fig. 1I). Meanwhile, the swimming distance of rats at the edge of the pool was also signi cantly increased (Fig. 1J & 1K). These results suggested that miR-21-5p might be related to mitochondrial function and cognitive de cits under TiO 2 NP treatment.
It has been reported that intracellular reactive oxygen species (ROS) can regulate miR-21-mediated pathways [17,18]; however, regulation by miR-21 of mitochondrial function is rarely reported. To further explore the relationship between miR-21-5p and mitochondrial function under TiO 2 NP intervention, we analyzed the correlation between miR-21-5p and the ΔΨm in SH-SY5Y cells treated with TiO 2 NPs. The results showed that the level of miR-21-5p correlated signi cantly with the ΔΨm in SH-SY5Y cells treated with TiO 2 NPs ( Fig. 2A). There was also a signi cant positive correlation between the ATP content and mir-21-5p levels in TiO 2 NPs-treated SH-SY5Y cells (Fig. 2B). These results suggested that miR-21-5p has a protective effect on mitochondrial function in SH-SY5Y cells treated with TiO 2 NPs. An miR-21-5p overexpression vector (Fig. 2C) and short hairpin RNA (shRNA) vector to silence miR-21-5p ( Fig. 2D) were employed to further evaluate the protective effect of miR-21-5p on mitochondrial function in SH-SY5Y cells treated with TiO 2 NPs. As shown in Fig. 2E, knockdown of miR-21-5p by shmiR-21-5p led to a further decrease in the ΔΨm of TiO 2 NPs-treated SH-SY5Y cells, and miR-21-5p overexpression reversed the decrease in the ΔΨm induced by TiO 2 NPs. Furthermore, miR-21-5p overexpression reversed the decrease in the ATP content induced by TiO 2 NPs in SH-SY5Y cells (Fig. 2F). These results demonstrated that miR-21-5p protects mitochondrial function against damage from TiO 2 NPs in SH-SY5Y cells.

MiR-21-5p protects the respiratory and glycolytic ability of cells under TiO 2 NPs treatment.
To further assess the effects of miR-21-5p on energy metabolism function in SH-SY5Y cells, the oxygen consumption rate (OCR) and extracellular acidi cation rate (ECAR) of SH-SY5Y cells were measured by Seahorse XF-24 system (Fig. 3A). All data were normalized to ensure that they were not affected by cell density [14]. Basal, ATP-linked, and maximal respiration were signi cantly decreased in miR-21-5p-silenced SH-SY5Y cells (shmiR-21-5p) compared with those in the controls (Fig. 3B). The inhibition of ATP-linked respiration and decline of ATP content mentioned in Fig. 2F showed the same trend. In order to further observe the effect of m on cell glycolysis activity, ECAR induced by oligomycin was employed [19,20]. A remarkably decrease in ECAR induced by oligomycin was detected in miR-21-5p-silenced SH-SY5Y cells (shmiR-21-5p) compared with that in the controls under TiO 2 NPs treatment. No difference between groups was observed in basal ECAR (Fig. 3C). The OCR to ECAR ratio showed decreased induction of respiratory capacity in miR-21-5p-silenced SH-SY5Y cells (shmiR-21-5p) compared with that in the controls (Fig. 3D). These results indicated that miR-21-5p protects the ability of SH-SY5Y cells to respond to TiO 2 NPs.
3.4 VDAC1 is crucial for miR-21-5p in protecting the mitochondria function of SH-SY5Y cells in response to TiO 2 NPs.
The voltage-dependent anion channel 1 (VDAC1) locates in mitochondrial outer membrane, where it regulates protein and metabolite tra cking in and out of the mitochondria [21]. It has also been reported that VDAC has an important relationship with mitochondrial respiration [22]. Thus, we assessed the levels of VDACs, including VDAC1, VDAC2, and VDAC3, in SH-SY5Y cells with or without shmiR-21-5p. As shown in Similarly, VDAC1 overexpression increased the basal, ATP-linked, and maximal respiration that had been decreased by shmiR-21-5p in SH-SY5Y cells under TiO 2 NPs treatment (Fig. 4F). These results con rmed that VDAC1 plays an important role in mir-21-5p-mediated protection of cellular respiratory function 3.5 Phosphorylated c-Jun is responsible for VDAC1 upregulation.
The mRNA levels of VDAC1 in TiO 2 NPs-treated SH-SY5Y cells were remarkably higher than those in the control group (Fig. 5A). Overexpression of miR-21-5p resulted in a signi cantly higher level of VDAC1 mRNA under TiO 2 NPs treatment (Fig. 5B), while knockdown of miR-21-5p by shmiR-21-5p decreased the VDAC1 mRNA level compared with that in control group (Fig. 5C). These results suggested that miR-21-5p regulates the transcription of VDAC1. To study the molecular mechanism by which miR-21-5p regulates VDAC1 transcription, VDAC1 promoter constructs Del 1 (-2000 to + 135) and Del 2 (-908 to + 135) (Fig. 5D) were employed to assess promoter-dependent transcriptional activity. The promoter-dependent transcriptional activity of the VDAC1 promoter construct Del 1 was remarkably higher in TiO 2 NPs-treated SH-SY5Y cells than in the control group, while the activity of VDAC1 promoter construct Del 2 was not different between the groups (Fig. 5E). Thus, TiO 2 NPs may increase the mRNA levels of VDAC1 through the promoter region (−2000 to −908). Next, potential transcription factors were predicted and detected by western blotting. The results indicated that TiO 2 NPs promoted the phosphorylation of c-jun, but had no effect on other transcription factors, including c-ETS1, T3R-β, c-Myb, and NF-AT2 (Fig. 5G). Inhibition of c-Jun phosphorylation prevented the transcription of VDAC1 mRNA (Fig. 5H) under TiO 2 NPs treatment and led to a further decline in the ΔΨm level (Fig. 5I) and ATP (Fig. 5J) content in SH-SY5Y cells. These results suggested that phosphorylated c-Jun, as a transcription factor of VDAC1, promotes the transcription of VDAC1 mRNA and protects the respiration function in SH-SY5Y cells under TiO 2 NPs treatment. Taken together, our results strongly demonstrated that miR-21-5p upregulation contributes to TiO 2 NPs-induced c-Jun (Ser73) phosphorylation, which, as a transcription factor, leads to VDAC1 transcriptional upregulation, which in preserves the ΔΨm level and ATP content in neuron cells, as summarized in Graphical abstract.

Discussion
Titanium dioxide nanoparticles (TiO 2 NPs) have been widely used in various industries, such as chemistry, paint, food, and medicine [23]. It has been reported that TiO 2 NPs induce central nervous system (CNS) dysfunction by accumulating in the brain [24]. In zebra sh, a low dose of TiO 2 NPs (5 µg/L) can lead to the aggregation of TiO 2 NPs in neurons, the increase of apoptosis, and the decrease of spatial recognition [25]. In vitro, TiO 2 NPs were also found to induce apoptosis in primary cultured hippocampal neurons [12]. In the process of TiO 2 NPs-induced neuron apoptosis, it was found that TiO 2 NPs could promote the decrease of ΔΨm, the release of cytochrome c, and the increase of caspase-3 level [24]. Those results indicated that mitochondria participate in the apoptosis of neurons induced by TiO 2 NPs. Mitochondria are semi-autonomous organelles responsible for energy production, cell differentiation, cell information transmission, cell growth, and cell cycle [26]. Given their many crucial functions in cell physiology, it is not surprising that mitochondria are involved in, or mediate, cognitive de cits, even before neuronal damage. Although some studies have reported the damaging effect of TiO 2 NPs on mitochondrial function, to date, most studies have focused on the downstream signaling pathway after mitochondrial function damage, such as the activation of the apoptosis pathway or oxidative stress effects. The molecular mechanisms of mitochondrial function regulation under the in uence of TiO 2 NPs remain unclear. Understanding these mechanisms is essential to protect CNS function under TiO 2 NPs exposure. In the present study, we demonstrated that TiO 2 NPs are strong positive regulators of miR-21-5p expression, which protects the neuron cells in response to TiO 2 NPs, and that VDAC1 plays an important role in the function of miR-21-5p by preserving oxygen consumption rate, ΔΨm, and the ATP content.
MiR-21 is considered an oncogenic molecular due to inhibiting the expression of phosphatases, which regulate the signaling pathways such as AKT and MAPK in cells [27,28]. In non-cancer cells, miR-21 usually acts as a protector [29,30,31]. Although microRNA-21 (miR-21) is believed to be related to the regulation of mitochondrial energy metabolism [32,33,34,35], the effect of TiO 2 NPs on miR-21 has not been reported. In the present study, we discovered that the level of miR-21-5p was upregulated in primary neuron cells, SH-SY5Y cells, and rat prefrontal cortex under TiO 2 NPs treatment in a dose-and timedependent manner (Fig. 1A − 1F). Considering that TiO 2 NPs also led to a decrease in the ΔΨm (Fig. 1G) and ATP content (Fig. 1H), miR-21-5p seemed to be a cause or result of mitochondrial dysfunction under TiO 2 NPs treatment. To con rm the function of miR-21-5p, the correlation of the miR-21-5p level with the ΔΨm or ATP content of SH-SY5Y cells treated with TiO 2 NPs was analyzed, and SH-SY5Y cells (miR- 21) and SH-SY5Y cells (shmiR-21) were employed (Fig. 2C &2D). Under TiO 2 NPs treatment, there was a signi cant correlation between miR-21-5p and the ΔΨm ( Fig. 2A) or the ATP content (Fig. 2B). Knockdown of miR-21-5p resulted in a signi cant decrease in the ΔΨm and ATP content of SH-SY5Y cells under TiO 2 NPs treatment. Overexpression of miR-21-5p resulted a signi cant increase in the ΔΨm and ATP content of SH-SY5Y cells under TiO 2 NPs treatment (Fig. 2E & 2F). Our results indicated that the increase of miR-21-5p expression induced by TiO 2 NPs is an important protective mechanism to preserve the ΔΨm and ATP content in neuron cells.
In the process of mitochondrial electron transport, the electrochemical potential energy is stored in the inner membrane of mitochondria. On both sides of the inner membrane, the asymmetric distribution of proton and other ion concentration will form the ΔΨm, which is the core driving oxidative phosphorylation to produce ATP [36]. The oxygen consumption rate (OCR) is the basis of mitochondrial oxidative phosphorylation and ATP synthesis [14,37]. In our study, OCR was measured using the Seahorse system in SH-SY5Y cells (shmiR-21) and SH-SY5Y cells (Nonsense) under TiO 2 NPs treatment. Basal, ATP-linked, and maximal respiration were remarkably declined in SH-SY5Y cells (shmiR-21) compared with those in SH-SY5Y cells (Nonsense) (Fig. 3B). This inhibition in ATP-linked respiration corresponded with the decreased ATP content in SH-SY5Y cells (shmiR-21) (Fig. 2F). Thus the damaged OCR induced by TiO 2 NPs was protected by mir-21-5p. Although glycogen depletion, caused by uncontrolled glycolysis, leads to cognitive impairment [38], glycolysis is the main source of energy supply for anoxic neurons. Moreover, previous studies suggested that lactate, a glycolytic product, plays an important role in memory processing [14]. Our results showed that miR-21 knockdown reduced the extracellular acidi cation rate (ECAR) induced by oligomycin under TiO 2 NPs treatment (Fig. 3C). This suggested that glycolysis is also one of the ways by which miR-21 protects neurons under TiO 2 NPs treatment.
Voltage-dependent ion channels are involved in the process of energy metabolism [39]. Here, we determined the expression levels of VDACs, and found that VDAC1 expression was regulated by miR-21 ( Fig. 4A) as well as by TiO 2 NPs (Fig. 4B). VDAC1plays an important role in glycolysis and oxidative phosphorylation [40]. High levels of VDAC1 were demonstrated in the brains of patients with Alzheimer's disease and in amyloid precursor protein transgenic mice [41]. Some studies have found that Vdac1 overexpression is related to the release of cytochrome c [40]. However, in the present study, VDAC1 overexpression reversed the reduction of the ΔΨm (Fig. 4D) and ATP content (Fig. 4E) induced by miR-21 knockdown in SH-SY5Y cells under TiO 2 NPs treatment. Basal, ATP-linked, and maximal respiration also increased signi cantly in SH-SY5Y cells (VDAC1, shmiR-21) compared with those in control SH-SY5Y cells (Vector, shmiR-21) (Fig. 4F). Therefore, we speculated that TiO 2 NPs would upregulate the VDAC1 via miR-21 accumulation, which, as an endogenous protection mechanism, consequently preserves the OCR, ΔΨm, and ATP content.
Up to now, no mechanism of titanium dioxide regulating VDAC has been reported. To explore how miR-21 regulates VDAC1 under TiO 2 NPs treatment, we detected the VDAC1 mRNA level in SH-SY5Y, SH-SY5Y (miR-21), and SH-SY5Y (shmiR-21) cells with TiO 2 NPs treatment (Fig. 5A − 5C). Thus, miR-21 might upregulate the VDAC1 by promoting its mRNA transcription under TiO 2 NPs treatment. To verify this speculation, the promoter-dependent transcriptional activities of VDAC1 were detected in SH-SY5Y cells with or without TiO 2 NPs treatment. The results indicated that VDAC1 promoter transcription factor binding sites (-908 to -2000) were responsible for VDAC1 mRNA upregulation in TiO 2 NPs-treated cells ( Fig. 5D & 5E). Although studies have reported that the transcriptional regulation of VDAC1 plays an important role in different cell models [42,43,44], the mechanism of VDAC1 transcription regulation is still unclear, especially under the intervention of TiO 2 NPs. Bioinformatic methods predicted transcription factor binding sites and then the levels of the predicted transcription factors using western blotting in cells with or without TiO 2 NPs treatment (Fig. 5G). The level of phosphorylated (P)-c-Jun (Ser73) was upregulated markedly in TiO 2 NPstreated SH-SY5Y cells compared with that in the control cells.
Furthermore, inhibition of c-Jun phosphorylation reversed the upregulation of VDAC1 induced by TiO 2 NPs and decreased the ΔΨm and ATP content in SH-SY5Y cells. Based on these results, we proposed a previously unreported mechanism of VDAC1 transcriptional regulation and the protective role of VDAC1 protein in neuron cells in response to TiO 2 NPs.
In summary (Graphical abstract), our results de ned a novel effect of TiO 2 NPs on miR-21-5p upregulation, which would inhibit the expression of an as yet unidenti ed phosphatase, leading to increased or maintained phosphorylation of c-Jun. This active form of c-Jun would increase the transcriptional activity of VDAC1 leading to the upregulation of VDAC1 protein levels, an increased OCR, elevation of the ΔΨm, and preservation of the ATP content. These ndings demonstrated the protective of miR-21 in TiO 2 NPs-induced neuron cells, providing novel insights into the multiple levels of VDAC1 protein regulation and revealed an endogenous neuroprotective mechanism in response to TiO 2 NPs.

Conclusion
TiO 2 is widely used in the processing of food, medicine, and cosmetics. Most products claim to contain TiO 2 NPs free. However, TiO 2 NPs has been found in a variety of products. In the present study, we found that TiO 2 NPs leads to cognitive decline by declining the oxygen consumption rate and ATP level of neurons. Downregulated VDAC1 is responsible for declined energy metabolism induced by TiO 2 NPs. Upregulation of miR-21-5p may be one of the endogenous protective mechanisms of TiO 2 NPs injury by preserving energy metabolism of neurons. This study provides new evidence for TiO 2 NPs-induced nervous system injury.

Declarations
Ethics approval and consent to participate This study was compliant with the Declaration of Helsinki Guidelines and was approved by the Committee on the Ethics of Animal Experiments of the Tianjin Institute of Environmental and Operational Medicine.

Consent for publication
Not applicable.

Availability of data and materials
All data generated or analyzed during this study are included in this published article.

Competing interests
The authors declare that they have no competing interests.  MiR-21-5p preserved the mitochondrial membrane potential (ΔΨm) and ATP levels of SH-SY5Y cells under TiO2 NP treatment. (A) SH-SY5 cells were seeded into wells of 6-well plates and cultured until the cell density reached 80-90%. The cells were then exposed to 20 μg·mL-1 TiO2 NPs for 9 hours. Some of the cells were collected for ΔΨm detection and some for RT-PCR in the same group. (B) SH-SY5 cells were seeded into wells of 6-well plates and cultured until the cell density reached 80-90%. The cells were then exposed to 20 μg·mL-1 TiO2 NPs for 9 hours. Some of the cells were collected for ATP detection and some for RT-PCR in the same group. The cells were then subjected to ΔΨm detection. The symbol (*) indicates a signi cant change in comparison between marked groups (P < 0.05). (F) SH-SY5 cells, SH-SY5 cells (shmiR-21-5p), or SH-SY5 cells (miR-21-5p) were treated with or without 20 μg·mL-1 TiO2 NPs for 9 hours. The cells were then subjected to ATP detection. The symbol (*) indicates a signi cant change in comparison between marked groups (P < 0.05).

Figure 3
MiR-21-5p knockdown decreased maximal respiration in SH-SY5 cells treated with TiO2 NPs. (A) SH-SY5 cells (Control) or SH-SY5 cells (shmiR-21-5p) were treated with 20 μg·mL-1 TiO2 NPs for 9 hours and then subjected to Seahorse system analysis to determine the oxygen consumption rate (OCR). (B) Quantitative analysis of (A). The symbol (*) indicates a signi cant decrease in comparison with that of the control (P < 0.05). (C) SH-SY5 cells (Control) or SH-SY5 cells (shmiR-21-5p) were treated with 20 μg·mL-1 TiO2 NPs for 9 hours and then subjected to Seahorse system analysis to determine oligomycin-induced extracellular acidi cation rate (ECAR). (D) Basal and maximal OCR to ECAR ratios were calculated to indicate the induction of a response to FCCP in SH-SY5 cells (shmiR-21-5p) compared with that in the control. VDAC1 was crucial for miR-21-5p to protect the mitochondria function of TiO2 NP-treated SH-SY5 cells.
(A) SH-SY5 cells (Nonsense) and SH-SY5 cells (shmiR-21-5p) were treated with 20 μg·mL-1 TiO2 NPs for 9 hours. The cell extracts were subjected to western blotting to determine the VDAC1, VDAC2, and VDAC3 levels. β-Actin was used as a protein loading control. (B) SH-SY5 cells were exposed to 10, 20, or 30 μg·mL-1 TiO2 NPs for 9 hours. The cell extracts were subjected to western blotting to determine the VDAC1, VDAC2, and VDAC3 levels. β-Actin was used as a protein loading control.   SH-SY5 cells were exposed to 10, 20, or 30 μg·mL-1 TiO2 NPs for 9 hours. The cell extracts were subjected to RT-PCR to determine VDAC1 mRNA expression. GAPDH was used as an internal reference.
(B) SH-SY5 cells (Vector) and SH-SY5 cells (miR-21-5p) were treated with 20 μg·mL-1 TiO2 NPs for 9 hours. The cell extracts were subjected to RT-PCR to determine VDAC1 mRNA expression. GAPDH was used as an internal reference. (C) SH-SY5 cells (Nonsense) and SH-SY5 cells (shmiR-21-5p) were treated with 20 μg·mL-1 TiO2 NPs for 9 hours. The cell extracts were subjected to RT-PCR to determine VDAC1 mRNA expression. GAPDH was used as an internal reference. The symbol (*) indicates a signi cant increase in comparison with that of the control (P < 0.05). (D) Schematic illustration of the construction of the VDAC1 promoter-driven luciferase reporter constructs. (E) The VDAC1 promoter-driven luciferase reporter constructs co-transfected with TK into H9c2 cells, with or without exposure to 20 μg·mL-1 TiO2 NPs for 9 hours, respectively. The luciferase activity of the transfectants was evaluated after 24 h, and the results are presented as relative VDAC1 promoter activity by normalization to TK and pGL3, respectively. The symbol (*) indicates a signi cant increase in comparison with the that of the control (P < 0.05). (F) Potential transcription factor binding sites in the VDAC1 promoter region (-2000 ~ -908) were analyzed using the TRANSFAC 8.3 engine online. (G) SH-SY5 cells were treated with 20 μg·mL-1 TiO2 NPs for 9 hours. The cell extracts were analyzed for the levels of potential transcription factors of VDAC1 using western blotting. (H-J) SH-SY5 cells and SH-SY5 cells (anti-c-Jun) were treated with 1% O2 for 6 hours. The cell extracts were analyzed for the levels of VDAC1 mRNA and ATP content. Changes in the mean JC-1 uorescence ratio (590 nm/530 nm) of cells were detected to indicate the ΔΨm levels.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. SupplementalInformation.docx