Mitochondrial Toxic Effects of Antiepileptic Drug Valproic Acid on Mouse Kidney Stem Cells

Valproic acid (VPA) is a histone deacetylase inhibitor that is used mainly as an antiepileptic and anticonvulsant drug. The side effects of VPA usually appears as hepatic injury and various metabolic disorders. On the other hand, it is rarely reported to cause kidney injury. Despite the many studies on the influence of VPA exposure on the kidneys, the specific mechanism remains unclear. This study examined the changes after VPA treatment to the mouse kidney stem cells (mKSCs). VPA triggers an increase in mitochondrial ROS, but there was no change in either mitochondrial membrane potential or the mitochondrial DNA copy number in mKSCs. The VPA treatment increased the mitochondrial complex III but decreased complex V significantly compared to the DMSO treatment as a control. The inflammatory marker (IL-6) and the expression of the apoptosis markers (Caspase 3) and were increased by VPA. In particular, the expression of the podocyte injury markers (CD2AP) was increased significantly. In conclusion, VPA exposure has adverse effects on mouse kidney stem cells.


Introduction
Valproic acid (VPA) is the most common drug for treating convulsions and epilepsy because of its powerful therapeutic effect against seizures, epilepsies, and migraine control [1]. VPA blocks the transcriptional action by inhibiting the activity of histone deacetylase class I. The drug can also alleviate spasms by inhibiting enzymes, such as GABA transaminase, succinate semialdehydrogenase, and alpha-ketoglutarate dehydrogenase [2]. When VPA enters the mitochondria, it binds to coenzyme A (CoA) to form VPA-CoA, which interferes with the metabolic pathway involving CoA, such as fatty acid oxidation [3]. The side effects of VPA include nausea, vomiting, diarrhea, progress, urticaria, arousing liver dysfunction, and kidney function abnormalities [4]. Exposure to VPA can cause urea cycle disorder that inhibits the release of ammonia into urea by inactivating the carbamoyl phosphate synthetase I involved in the production of carbamoyl phosphate. This process leads to hyperammonemia, which increases the concentration of ammonia in the blood, causes kidney toxicity, or accelerates kidney dysfunction [5].
Whether VPA causes nephrotoxicity is unclear, but the uric acid and creatine levels of rats that had received 100 mg/kg or 500 mg/kg of VPA daily for 30 days were higher than the controls, with a concomitant decrease in the activity of the antioxidant enzyme [6]. By contrast, Costalonga EC et al. (2016) reported that VPA had beneficial effects against ischemic reperfusion injury (IRI) in the kidney [7]. The acute tubular necrosis (ATN) grade was higher in IRI-induced rats than in normal rats, and the mRNA expression level of the inflammation markers (IL-6, TNF-α, and MCP-1) were increased in IRI-induced rat than in normal rat. On the other hand, when IRI-induced rats were treated with VPA, the ATN grade and mRNA expression levels of the inflammation marker returned to normal state. Many alternative experiments have evaluated kidney toxicity in vitro [8]. Despite the in vitro experimental results using human embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) [9], there are disadvantages when introducing stem cells because they require an extensive protocol and have low sensitivity. Therefore, it is necessary to develop an in vitro model suitable for nephrotoxicity studies.
This study examined whether mouse kidney stem cells (mKSCs) are suitable for validating nephrotoxic substances. In addition, the pathophysiological alterations caused by VPA in the mKSCs, and the underlying mechanism were assessed.

Measurement of Cell Viability
The mKSCs were seeded on 96-well plates with 7000 cells/well in 70 µL of MOUSE KIDNEY STEM CELL Growth Media. After incubation for 24 h, the drug media were added to 100 µL/well. After culturing in an incubator for 24 h, a CCK-8 assay was carried out using an EZ-cytox solution (EZ-500, DoGenBio, Seoul, Republic of Korea). All the remaining drug media were removed, and the cells were washed once with DPBS (LB001-02, WELGENE, Seoul, Republic of Korea). Subsequently, 110 µL of a mixture of pure EZ-cytox solution and DMEM at a ratio of 1:10 was added. The resulting mixture was incubated for one hour. After incubation, the absorbance at 450 nm was evaluated using a Synergy H1 microplate reader (Biotek instrument, Winooski, VT, USA). The cell viability was evaluated by comparing the relative absorbances with the absorbance of the solvent control set to 100%. The IC50, which is indicative of 50% cell viability, was substituted for the value of the y-axis in the trend line formula of the graph.

Measurement of mtROS
The drug treatment conditions were identical to those for measuring cytotoxicity. The level of mitochondrial ROS (mtROS) was measured 24 h after the VPA treatment. The cells were treated with mitoSOX dye (M36008, Invitrogen, MA, USA; 3 µM) and Hoechst 33342 (H3570, Invitrogen; 12 µg/mL) in pure DMEM for 10 min at 37 • C. The cells were washed twice with DPBS to determine the clear fluorescence absorption intensity using a Synergy H1 microplate reader (Agilent Technologies, Santa Clara, CA, USA).

Measurement of Mitochondrial Membrane Potential
The mitochondrial membrane potential (∆Ψm) was determined four hours after VPA treatment. The drug treatment conditions were identical to those when measuring the cytotoxicity. The cells were washed with DPBS and evaluated mitochondria membrane potential with JC −1 dye (T3168, Invitrogen; 5 µg/mL) for 10 min at 37 • C. The cells were then washed twice with DPBS to determine the fluorescence absorption intensity using a Lionheart FX (Agilent Technologies).

RNA Purification, cDNA Synthesis, and RT-PCR
Trizol (AM9738, Invitrogen) reagent was used for total RNA purification. Reverse transcription was performed for cDNA synthesis using 1 µg RNA, and the final volume was

Analysis of Mitochondrial Gene Expression and mtDNA Copy Number
Mitochondrial gene expression was measured by extracting the total RNA and synthesizing cDNA, followed by real-time polymerase chain reaction (RT-PCR), as mentioned above. For RT-PCR performance, 2 µL of cDNA was analyzed by RT-PCR. For mtDNA expression analysis, the total DNA was extracted using G-DEXTM-IIc (17231, iNtRON, Daejeon, Republic of Korea). The total DNA was diluted to 3 ng/µL, then 2 µL of total DNA was analyzed by RT-PCR, as described above. Mitochondrial Copy number is quantified by taking the ratio between a target mitochondrial gene and a reference nuclear gene (mtDNA/nDNA) using quantitative real-time PCR. β2-microglobin was used for the reference gene of nDNA. The sequences of primers and probes were as follows: Complex I (forward: 5 -TTCTTGCAGCTGTGTCCAAC-3 , reverse: 5 -AGCATTTTGG GAGGGTTCTT-3 ), Complex II (forward: 5 -ACACAGACCTGGTGGAGACC-3 , reverse

Statistical Analysis
Significant differences were detected using one-way ANOVA followed by a Tukey's test for multiple comparisons. The analysis was performed using the Prism Graph Pad v8.0.1 (Graph Pad Software, San Diego, CA, USA). The obtained values are expressed as the means ± SD of at least three separate experiments; p values < 0.05 were considered significant.

Effects of Drugs on the mKSCs and HEK293F Cell Viability
The mKSCs and HEK293F cells were exposed to the drug to find a cell line suitable for kidney toxicity. The drugs tested were doxorubicin, (Z)-4-hydroxytamoxifen, acetaminophen, valproic acid, sodium salt, and amiodarone. The five drugs caused mitochondrial DNA damage, inhibited mitochondrial-derived respiration, and disrupted the mitochondrial permeability transition pores (mPTP) [10][11][12][13][14]. As shown in the dose-response graph of Figure 1 and Table 1, doxorubicin, (Z)-4-hydroxytamoxifen, and amiodarone similarly decreased the cell viability in both mKSCs ( Figure 1A,B,E) and HEK293F cells ( Figure 1F,G,J). On the other hand, VPA and acetaminophen decreased the cell viability of only mKSCs ( Figure 1C,D), not in HEK293F ( Figure 1H,I). mKSCs react more to VPA and acetaminophen exposure than HEK293F cells. Acetaminophen can induce hepatotoxicity and nephrotoxicity by being metabolized into the toxic metabolite, N-acetyl-p-benzoquinone imine, in the liver. On the other hand, it is unclear if VPA induces nephrotoxicity. Therefore, the following experiment examined the basal mechanism for the kidney toxicity of VPA in mKSCs.

Effect of Drug on Mitochondrial Function in mKSCs and HEK293F
VPA is a typical mitochondrial toxic substance. Mitochondrial reactive oxygen species (mtROS) are markers to detect mitochondria dysfunction because the mitochondria are often injured under oxidative stress. In particular, the scavenging of mitochondrialtargeted ROS prevented podocyte loss, albuminuria, glomerulosclerosis, and kidney failure [15]. As shown in the mitochondrial superoxide graph of Figure 2, doxorubicin, (Z)-4hydroxytamoxifen, and amiodarone increased the mtROS in both mKSCs (Figure 2A,B,E) and HEK293F cells ( Figure 2F,G,J). On the other hand, VPA and acetaminophen increased the mtROS of only mKSCs ( Figure 2C,D), not in HEK293F ( Figure 2H,I). These results show the same tendency as the results of cell viability. No changes in the mitochondrial membrane potential were noted ( Figure 3A,B), but the transcription levels of the SOD2 and catalase genes were measured by RT-PCR. SOD2 was upregulated by a 100 µM VPA treatment for four hours ( Figure 3C), but catalase was unaffected ( Figure 3D).

Effect of Drug on Mitochondrial Function in mKSCs and HEK293F
VPA is a typical mitochondrial toxic substance. Mitochondrial reactive oxygen species (mtROS) are markers to detect mitochondria dysfunction because the mitochondria are often injured under oxidative stress. In particular, the scavenging of mitochondrial- hydroxytamoxifen, and amiodarone increased the mtROS in both mKSCs (Figure 2A,B,E) and HEK293F cells ( Figure 2F,G,J). On the other hand, VPA and acetaminophen increased the mtROS of only mKSCs ( Figure 2C,D), not in HEK293F ( Figure 2H,I). These results show the same tendency as the results of cell viability. No changes in the mitochondrial membrane potential were noted ( Figure 3A,B), but the transcription levels of the SOD2 and catalase genes were measured by RT-PCR. SOD2 was upregulated by a 100 µM VPA treatment for four hours ( Figure 3C), but catalase was unaffected ( Figure 3D).  valproic acid sodium salt (VPA), and (E) amiodarone (AMD) in the mKSCs. Bar graph of mitochondrial superoxide using mitoSOX staining to measure a mtROS with (F) doxorubicin, (G) (Z)-4-hydroxytamoxifen, (H) acetaminophen, (I) valproic acid sodium salt, and (J) amiodarone in the HEK293F cell line. The significance was obtained by ANOVA. * p < 0.05, ** p < 0.01, *** p < 0.01 vs. Control. Each value is expressed as the means ± SD. n = 6.

Effect of VPA on Mitochondrial Contents in mKSCs
The mitochondria are the key components of the generation and regulation of cellular bioenergetics. In addition, they produce most adenosine triphosphate (ATP) molecules using oxidative phosphorylation (OXPHOS) in the kidney. OXPHOS consists of five protein complexes: Complex I (NADH dehydrogenase), Complex II (Succinate dehydrogenase), Complex III (Coenzyme Q-cytochrome c reductase), Complex IV (Cytochrome c oxidase), and Complex V (ATP synthase). The five complexes were measured by RT-PCR after a four-hour VPA treatment. The mRNA level of Complex III was increased in the VPA 100 µM treatment group, and Complex V decreased depending on the VPA concentration ( Figure 4C,E). On the other hand, there was no significant variation in the mRNA expression levels of the other mitochondrial complex genes, Complex I, Complex II, and Complex IV ( Figure 4A,B,D). The VPA treatment did not change the mtDNA copy number used to measure the mitochondria density ( Figure 4F). oxidase), and Complex V (ATP synthase). The five complexes were measured by RT-PCR after a four-hour VPA treatment. The mRNA level of Complex III was increased in the VPA 100 µM treatment group, and Complex V decreased depending on the VPA concentration ( Figure 4C,E). On the other hand, there was no significant variation in the mRNA expression levels of the other mitochondrial complex genes, Complex I, Complex II, and Complex IV ( Figure 4A,B,D). The VPA treatment did not change the mtDNA copy number used to measure the mitochondria density ( Figure 4F).

Effect of VPA on mKSCs Inflammation, Apoptosis, and Podocyte Injury Marker
Oxidative stress causes inflammation in the kidneys and decreases kidney function via apoptosis. After four hours of treatment with VPA, the mRNA expression level of IL-6, a kidney inflammatory marker, was increased significantly by the 100 µM VPA treatment ( Figure 5A). On the other hand, there was no significant variation in the mRNA expression levels of the other inflammatory markers, MMP-9 and TGF-β ( Figure 5B,C). In addition, the VPA treatment increase caspase 3 mRNA expression significantly, which is involved in apoptosis ( Figure 5E), but there was no significant variation in the Bcl-2 and Bax mRNA expression ( Figure 5D,F). These results show that VPA is involved in cell inflammation and apoptosis. ment ( Figure 5A). On the other hand, there was no significant variation in the mRNA expression levels of the other inflammatory markers, MMP-9 and TGF-β ( Figure 5B,C). In addition, the VPA treatment increase caspase 3 mRNA expression significantly, which is involved in apoptosis ( Figure 5E), but there was no significant variation in the Bcl-2 and Bax mRNA expression ( Figure 5D,F). These results show that VPA is involved in cell inflammation and apoptosis.  The podocyte injury marker was examined because mKSCs cells express podocytespecific markers, such as the podocin, nephrin, and podocalyxin genes. After four hours of treatment with VPA, CD2AP mRNA expression was increased significantly in the 10 µM and 100 µM VPA treatment groups ( Figure 5G). On the other hand, there was no significant variation in the mRNA expression levels of the other inflammatory markers, VEGF-A, nephrin, podocin, and podocalyxin ( Figure 5H-K).

Discussion
Acute kidney injury (AKI) is defined by a rapid increase in serum creatinine, a decrease in urine output, or both [16]. AKI caused by drugs is a complex combination of the specific drug nephrotoxicity, metabolism, excretion, and patient characteristics [17]. Drugs, such as aminoglycoside antibiotics, amphotericin B, and cisplatin, cause renal tubular toxicity and paralyze mitochondrial function by causing oxidative stress [17][18][19]. AKI causes various metabolic changes that occur inside cells and in mitochondria. For decades, several mouse AKI models have been developed to examine the mechanism of AKI development, renal pedicle clamping (ischemia reperfusion injury), Cisplatin-induced nephrotoxicity, sepsis (LPS, cecal slurry, and cecal ligation and puncture), folic acid, and rhabdomyolysis are representative AKI model for study the disease mechanisms [20]. Funk JA et al. (2012) reported that the expression of mitochondrial respiratory proteins ATP synthase β and cytochrome c oxidase subunits I and IV were attenuated significantly in the myoglobinuria AKI model [21]. Furthermore, in the folic acid-induced mouse AKI model, mitochondrial DNA fragmentation occurred during excessive mitochondrial fission but before tubular cell death [22]. Hence, the mitochondria and the pathogenesis of AKI are closely related.
This study examined the effects of VPA on mouse kidney stem cells. The decrease in cell viability according to the VPA concentration was observed only in mKSCs cells, whereas HEK293F, commonly used in kidney cells, did not show any changes in cell viability. The cell viability results confirmed that mKSCs are a more sensitive cell to VPA than HEK293F. Therefore, mKSCs may be used more frequently than HEK293 cells if the substance is mildly toxic or cell-type specific. Due to stem cells exhibiting low mitochondrial content, mKSCs may be sensitive to mitochondria-toxic materials [23]. Sitarz et al. (2014) reported that VPA induces mitochondrial dysfunction in fibroblasts [24]. In this study, the mtROS level of mKSCs was increased by the VPA treatment. When mitochondrial oxidative stress is not adjusted, it eventually leads to cell death by disrupting the balance of the membrane potential generated in mitochondria or opening the mPTP [25]. The cells have mechanisms to minimize the generation of mtROS [26]. Superoxide dismutase 2 is a representative mitochondrial antioxidant enzyme. Interestingly, the mRNA level of SOD2 was increased in the 100 µM VPA treatment group. In a previous study, mRNA overexpression of superoxide dismutase caused an increase in the ROS level and oxidative stress [27]. This can be further evidence of oxidative stress induced by VPA exposure.
The changes in the mRNA level of the mitochondrial complexes by the VPA treatment were also confirmed. In particular, the expression level of mitochondrial complex V, an ATP synthase, decreased significantly with increasing VPA concentration. ATP synthase is involved in the final step of the oxidative phosphorylation of ATP production. A decrease in the ATP synthase mRNA expression level means that the amount of ATP produced by the cells may decrease. A previous study reported that the VPA-induced inhibition of mitochondrial biogenesis results in increased glycolysis [28], showing that similar action can occur in mouse kidney cells. In addition, the mRNA expression level of mitochondrial complex III was increased with increasing VPA concentration, which is associated with an increase in mtROS. According to previous studies, complex III is a major contributor to intracellular ROS generation and releases it into the inner membrane space and matrix of the mitochondria. Hence, an increase in mitochondrial complex III by VPA a treatment may cause mtROS, which participates in oxidative stress in mKSCs [29].
Mitochondrial dysfunction caused by oxidative stress induces host cell inflammation and, eventually, apoptosis [30]. As the VPA concentration was increased, the mRNA expression of IL-6, an inflammatory marker, increased, but there was no significant change in the mRNA expression of TGF-β and MMP-9. The mRNA expression level of Casp3, a representative apoptosis marker, increased significantly. Casp3 is an enzyme that is activated in the last stage of apoptosis [31], and the increase in Casp3 mRNA expression indicates that VPA can induce apoptosis in mouse kidney cells. Therefore, oxidative stress caused by VPA induces an inflammatory response in kidney stem cells that can also interfere with normal cell growth and inhibit cell function.
The level of CD2AP mRNA expression, a marker of podocyte damage, increased when the cells were treated with VPA for four hours. CD2AP is a protein that interacts with the enzymes involved in cytoskeletal remodeling, membrane transport, and cell motility in the slit diaphragm of podocytes. In vivo, podocytes form a slit diaphragm with other podocytes adjacent to each other and interact, such as material exchange. The slit diaphragm is broken when adjacent podocytes are damaged, and the area of the podocytes increases to form a slit diaphragm with other podocytes except for the damaged podocytes [32]. On the other hand, there were no significant changes in the mRNA expression level of the podocyte damage markers, such as nephrin, podocin, and VEGF-A, except for CD2AP. Nevertheless, the changes in the podocyte morphology and podocyte damage markers should be confirmed in vivo through histological analysis to prove that VPA damages podocytes.
This study examined the effects of VPA exposure on mouse kidney stem cells. VPA induced oxidative stress in kidney cells, which damages the mitochondria. In conclusion, VPA should be considered a nephrotoxic substance. In future research, examinations of VPA's metabolic and epigenetic effects are needed. In vitro stem cell experiments have limitations for investigating the metabolic effect of target molecules, and our gene-based studies also have limitations for examining the overall mechanism of kidney damage.