Hexavalent chromium induces apoptosis in male somatic and spermatogonial stem cells via redox imbalance

Hexavalent chromium [Cr(VI)], an environmental toxicant, causes severe male reproductive abnormalities. However, the actual mechanisms of toxicity are not clearly understood and have not been studied in detail. The present in vitro study aimed to investigate the mechanism of reproductive toxicity of Cr(VI) in male somatic cells (mouse TM3 Leydig cells and TM4 Sertoli cells) and spermatogonial stem cells (SSCs) because damage to or dysfunction of these cells can directly affect spermatogenesis, resulting in male infertility. Cr(VI) by inducing oxidative stress was cytotoxic to both male somatic cells and SSCs in a dose-dependent manner, and induced mitochondria-dependent apoptosis. Although the mechanism of Cr(VI)-induced cytotoxicity was similar in both somatic cells, the differences in sensitivity of TM3 and TM4 cells to Cr(VI) could be attributed, at least in part, to cell-specific regulation of P-AKT1, P-ERK1/2, and P-P53 proteins. Cr(VI) affected the differentiation and self-renewal mechanisms of SSCs, disrupted steroidogenesis in TM3 cells, while in TM4 cells, the expression of tight junction signaling and cell receptor molecules was affected as well as the secretory functions were impaired. In conclusion, our results show that Cr(VI) is cytotoxic and impairs the physiological functions of male somatic cells and SSCs.


Cr(VI) induces oxidative stress and disrupts mitochondrial membrane potential in male somatic cells and SSCs.
We measured the formation of reactive oxygen species (ROS) and loss of mitochondrial membrane potential (MMP) in cells exposed to Cr(VI). The production of intracellular ROS increased in a dose-dependent manner in all cell types after 4 h incubation with Cr(VI) and was maintained up to 24 h (Fig. 3). In contrast, the MMP decreased in a dose-dependent manner, after 24 h exposure to Cr(VI) (Fig. 3). Pre-treatment (5 mM, 1 h) with the ROS inhibitor, N-acetyl-L-cysteine (NAC), significantly counteracted the effects of Cr(VI) on cell viability, ROS formation, and MMP in all the cell types ( Supplementary Figs 1,2 and 3). These results indicated that the Cr(VI)-induced cytotoxicity in male somatic cells and SSCs was mediated through oxidative stress and subsequent mitochondrial damage. Figure 4 shows the changes in the expression levels of antioxidant genes as measured by quantitative reverse transcription-polymerase chain reaction (qRT-PCR), in all the cell types exposed to increasing doses of Cr(VI). Cr(VI) decreased the mRNA expression levels of catalase (Cat), superoxide dismutase 1(Sod1), superoxide dismutase 2 (Sod2), glutathione peroxidase 1 (Gpx1), and glutathione S-transferase a4 (Gsta4) in a dose-dependent manner in all the cells compared to the values in the relevant control. However, there were some cell type-specific differences in the responses. In the somatic cells, Cr(VI) increased the mRNA expression of glutathione S-transferase a1 (Gsta1) in a dose-dependent manner. In the SSCs, expression of Gsta1 increased with lower doses of Cr(VI), but decreased with a higher dose. Our results show that exposure to Cr(VI) disrupted cellular antioxidant defense mechanisms, leading to increased oxidative stress.

Cr(VI) alters mRNA expression of antioxidant enzymes in male somatic cells and SSCs.
Scientific RepoRts | 5:13921 | DOi: 10.1038/srep13921 Cr(VI) induces intrinsic apoptotic pathways via regulation of AKT1, P53, and MAPK family proteins in male somatic cells. Western blot analyses were performed to investigate the underlying mechanisms of the pro-apoptotic role of Cr(VI) in male somatic cells. Cr(VI) exposure for 24 h decreased the BCL2/BAX protein ratio and increased the expression of cleaved-caspase 9 (CASP9), caspase 3 (CASP3), and poly (ADP-ribose) polymerase (PARP) proteins in a dose-dependent manner (Fig. 5).

Figure 1. Cytotoxicity of Cr(VI) in TM3 and TM4 cells.
(a,f) Cell viability relative to the control (100%) in mouse TM3 and TM4 cells, respectively, exposed to different concentrations of Cr(VI) in μ M for 24 h. Cell viability was measured using the Cell Counting Kit-8 (CCK-8). (b,g) LDH release in TM3 and TM4 cells, respectively, treated with different concentrations of Cr(VI) in μ M for 24 h. (c,h) Representative fluorescent images of TUNEL-positive (red) TM3 and TM4 cells, respectively, after exposure to 12.5 μ M Cr(VI) for 24 h. Cell nuclei were counterstained with 4′ , 6-diamidino-2-phenylindole (DAPI, blue). Magnification 100X. (d,i) Apoptosis (dot plot distribution and % apoptosis calculations) of TM3 and TM4 cells, respectively, measured by Annexin V/PI assay after exposure to 12.5 μ M Cr(VI) for 24 h. (e,j) % Caspase 3 enzyme activity relative to the control (100%) in the mouse TM3 and TM4 cells, respectively, exposed to different concentrations of Cr(VI) in μ M for 24 h. All values are expressed as mean ± SEM. (n = 3). *P < 0.05, **P < 0.01, and ***P < 0.001 compared to the control. Cr(VI) exposure, at a concentration of 12.5 μ M for 24 h, increased the expression of cytosolic cytochrome c (CYCS, somatic) (Fig. 5). The levels of endogenous pro and cleaved-CASP3 were negligible in control (untreated) cells, but dose-dependently increased in both TM3 and TM4 cells after exposure to Cr(VI). Since pro-CASP3 is inactive and the cleaved-CASP3 is responsible for most of the catalytic activity of CASP3, we compared the levels of cleaved-CASP3 to those of beta actin (ACTB). The evidence for the dose-dependent increase of cleaved-CASP3 (Fig. 5) observed in the western blots was supported by the increase of caspase 3 enzyme activity (Fig 1e,j). We also investigated the extrinsic apoptotic pathways in the somatic cells by measuring the gene expression of fas (Fas), fas ligand (Fasl) and caspase 8 (Casp8). Exposure to Cr(VI) for 24 h increased the expression of Fasl mRNA but decreased the expression of Fas and Casp8 ( Supplementary Fig 4) indicating that Cr(VI) induced intrinsic (mitochondria-dependent) apoptosis in male somatic cells.
We examined the alterations in the expression of other proteins related to either cell survival or death in response to cellular stress. At concentrations up to 12.5 μ M, Cr(VI) did not affect the ratio of P-AKT1/T-AKT1 protein (also known as protein kinase B) (Fig. 5). However, at a concentration of 25 μ M Cr(VI), there was a 20% and 40% reduction in the expression of P-AKT1/T-AKT1 protein in TM3 and TM4 cells, respectively, compared to the corresponding values in the relevant control group. The expression of the protein ratios P-JNK1/2 to T-JNK1/2 (c-Jun N-terminal kinases) and P-P38/T-P38 (P38 mitogen-activated protein kinases) protein increased after exposure to Cr(VI) in a dose-dependent manner in somatic cells (Fig. 5). In TM3 cells, the relative change of P-P38 was more pronounced than P-JNK1/2, but this order was reversed in TM4 cells. In contrast, the expression of the protein ratios P-ERK1/2 to T-ERK1/2 (extracellular-signal-regulated kinases) decreased. A concentration of 25 μ M Cr(VI) reduced P-ERK1/2 expression by two-fold in TM3 cells, but reduced it by a factor of 15 in TM4 cells compared to values in the relevant control. Cr(VI) also increased the expression of P-P53/T-P53 (transformation related protein 53) protein in a dose-dependent manner. A concentration of 25 μ M Cr(VI), increased the expression of P-P53 expression by 8.4-fold in TM3 cells compared to a 3.8-fold increase in TM4 cells (Fig. 5). These results indicate that after exposure to Cr(VI), the P-AKT1, P-ERK1/2, and P-P53 proteins were regulated differently in TM3 and TM4 cells.  increased the expression of the P-P53 protein (Fig. 6) and also increased the levels of cleaved CASP9, CASP3, and PARP proteins (Fig. 6). Considering the mitochondria-independent apoptosis pathway in SSCs, although, Cr(VI) increased the expression of Fasl mRNA, the expression of Fas and Casp8 decreased ( Supplementary Fig 4). In addition, while concentrations of Cr(VI) up to 12.5 μ M increased apoptosis as evident from the dose-dependent cleavage of CASP3 and PARP proteins, higher concentrations (25 μ M) decreased apoptosis, shown by the decreased cleavage of CASP3 and PARP proteins. We therefore hypothesized that some additional factors, apart from apoptosis, could bear responsibility for the observed loss of cell viability. Since the GDNF signaling pathway controls the self-renewal/maintenance of SSCs via AKT1-Mycn signaling mechanisms 18,19 and the proliferation/differentiation of SSCs via ERK1/2-Fos mechanisms 20 , we examined the expression of selected GDNF signaling components. We observed that Cr(VI) decreased not only the transcriptional (mRNA) expression of receptor molecules such as glial cell line derived neurotrophic factor family receptor alpha 1 (Gfra1) and ret proto-oncogene (Ret) in a dose-dependent manner, but also the expression of downstream components such as v-myc myelocytomatosis viral related oncogene neuroblastoma derived homolog (Mycn) and FBJ osteosarcoma oncogene (Fos) (Fig. 6). In addition, Cr(VI) decreased the expression of two other proteins, P-AKT1 and P-ERK1/2, which are downstream of Ret, in a dose-dependent manner (Fig. 6). Our results indicate that Cr(VI) induced P-P53-dependent apoptosis in SSCs and disrupted the GDNF signaling pathway that is responsible for the differentiation and self-renewal of SSCs.

Cr(VI) disrupts steroidogenesis in TM3 cells, affects expression of tight-junction signaling and cell receptor molecules, as well as impairs secretory functions in TM4 cells.
We investigated the negative impact of Cr(VI) on TM3 and TM4 cell functions. In TM3 cells, we examined the effect of Cr(VI) on the transcriptional expression of testicular enzymes responsible for sex hormone synthesis, including cytochrome P450, family 11, subfamily a, polypeptide 1 (Cyp11a1), cytochrome P450, family 17, subfamily a, polypeptide 1 (Cyp17a1), cytochrome P450, family 19, subfamily a, polypeptide 1 (Cyp19a1), steroidogenic acute regulatory protein (Star), hydroxyl-delta-5-steroid dehydrogenase, 3 beta-and steroid delta-isomerase 1 (Hsd3b1), and hydroxysteroid (17-beta) dehydrogenase 3 (Hsd17b3) (Fig. 7). Exposure to Cr(VI) for 24 h significantly decreased the expression of Cyp11a1 and Hsd3b1 in a dose-dependent manner. Cr(VI) also decreased the expression of Cyp19a1. In contrast, the expression of Cyp17a1 and Star significantly increased at all tested doses of Cr(VI) compared to the corresponding values of the control. The expression of Hsd17b3 was not significantly affected by Cr(VI). Since the tight junctions of Sertoli cells form the blood-testis barrier 21 , we examined the effects of Cr(VI) on the transcriptional expression of tight-junction signaling molecules including tight junction protein 1 (Tjp1), vimentin (Vim), and occludin (Ocln) and cell receptors, such as follicle-stimulating hormone receptor (Fshr) and androgen receptor (Ar) that play a key role in the maturation and normal functioning of Sertoli cells 22,23 . We observed that exposure to increasing doses of Cr(VI) significantly decreased the expression of tigh-junction signaling molecules and cell receptors (Fig. 7). We also examined the effects of Cr(VI) on the transcriptional expression of glial cell line derived neurotrophic factor (Gdnf), ets variant gene 5 (Etv5), and fibroblast growth factor (Fgf2) in TM4 cells that are essential for the maintenance of SSCs (Fig. 7). Increasing doses of Cr(VI) significantly decreased the expression of these factors. Our results demonstrate that Cr(VI) impairs the physiological functions of TM3 and TM4 cells.

Conditioned Medium (CM) from Cr(VI)-treated TM3 cells impairs TM4 cell functions. Finally
we examined whether impaired steroidogenesis in TM3 cells induced by Cr(VI) toxicity had any direct effect on TM4 cell functions. We observed that Cr(VI)-treatment significantly decreased the testosterone release from the TM3 cells into the culture media (Fig. 7e). This culture media was mixed with fresh media at 1:1 ratio and used as conditioned media (CM) to culture the TM4, Sertoli cells for 48 h. We then measured the transcriptional expression of tight junction signaling molecules, cell receptors, and factors responsible for SSC survival in TM4 cells. We observed that increasing doses of conditional medium (CM) significantly decreased the expression of Tjp1, Ar, Etv5 and Fgf2 in TM4 cells (Fig. 7f). These results suggested that damage to TM3 cells might have deleterious effects on the physiological functions of TM4 cells.

Discussion
This study aimed to determine the in vitro cytotoxic effects of Cr(VI) on mouse Leydig (TM3) cells, Sertoli (TM4) cells, and SSCs that have critical roles in spermatogenesis. We examined the effects of Cr(VI) on oxidative stress and apoptosis-related signaling mechanisms in TM3, TM4, and SSCs, cell proliferation/self-renewal mechanisms of SSCs, and the physiological functions of TM3 and TM4 cells. For this purpose, cells were treated with Cr(VI) at doses of 0, 3.125, 6.25, 12.5, 25, or 50 μ M for 24 h; and then analyzed biochemically, and by flow cytometry, fluorescence microscopy, qRT-PCR, and immunoblotting.
To the best of our knowledge, this study is the first to report that the dose-dependent cytotoxic effects of Cr(VI) exposure in both male somatic cells and SSCs are mediated through apoptosis. Since somatic cells and SSCs play an important role in the process of spermatogenesis, damage to these cells has adverse effects on the production of healthy sperm cells. We investigated whether oxidative stress and mitochondrial dysfunction played a role in our current experimental model and observed that, after 4 h, exposure to Cr(VI) increased the production of ROS, which was maintained up to 24 h. In contrast, 24 h exposure to Cr(VI) decreased MMP in a dose-dependent manner. Treatment with a ROS inhibitor, NAC, abrogated the effects of Cr(VI). This demonstrates that oxidative stress and subsequent mitochondrial damage play a crucial role in Cr(VI)-induced cytotoxicity. We showed that the transcriptional expression of antioxidant enzymes that play a significant role in scavenging free radicals, including Cat, Sod1, Sod2, Gpx1, and Gsta4 24-29 decreased with increasing doses of Cr(VI). Thus, the oxidative stress in the somatic cells and SSCs observed after exposure to Cr(VI) appears to be due to poor scavenging of free radicals by the antioxidant enzymes. These results are supported by those of previous reports demonstrating that Cr(VI)-induced oxidative stress via the suppression of antioxidant enzymes plays a major role in male infertility 16,30,31 . The over-expression of Gsta1 (seen with lower doses of Cr(VI) in SSCs and all doses in somatic cells) might protect the cells against Cr(VI)-induced oxidative stress.
The induction of the mitochondria-dependent (intrinsic) pathway of apoptosis in the male somatic cells observed after Cr(VI) exposure, was confirmed by the decreased BCL2/BAX protein ratio, increased expression of cytosolic CYCS, and cleavage of CASP9, CASP3, and PARP proteins. Cell viability depends on the balance between survival and pro-apoptotic signaling, regulated by the AKT1, MAPK, and P53 pathways 32,33 . AKT1 and ERK1/2 signaling pathways are associated with cell proliferation and survival 34,35 , while ROS produced by several toxicants, including Cr(VI), serve as second messengers to activate pro-apoptotic kinases, such as JNK1/2 and P38 7,36-43 . Cr(VI)-induced DNA damage and oxidative stress also promoted the activation of P53 leading to intrinsic apoptosis 33,[44][45][46][47] . We showed that, in male somatic cells, exposure to Cr(VI) increased the phosphorylation of JNK1/2 and P38 in a dose-dependent manner, while the phosphorylation of ERK1/2 decreased (down-regulation was higher in TM4 cells). The expression levels of P-AKT 1 were not altered up to the IC 50 dose level of Cr(VI). At a concentration of 25 μ M, however, Cr(VI), decreased the expression of P-AKT1, by only 20% in TM3 cells, but a 40% reduction was observed in TM4 cells. Cr(VI) also increased P-P53 expression in a dose-dependent manner (up-regulation was more pronounced in TM3 cells). These cell-specific differences in the regulation of P-AKT1, P-ERK1/2, and P-P53 proteins are probably at least partially responsible for the differential sensitivity of TM3 and TM4 cells to Cr(VI).
The increased phosphorylation of P53 and cleavage of CASP9, CASP3, and PARP proteins showed that Cr(VI) induced intrinsic apoptosis in SSCs at concentrations up to 12.5 μ M, but a higher concentration (25 μ M) decreased apoptosis. We therefore investigated other factors, apart from apoptosis, that could also be involved in Cr(VI)-induced cytotoxicity. In SSCs, GDNF plays two major roles, regulating the self-renewal/maintenance via AKT1-Mycn signaling pathways 18,19 and the proliferation/differentiation via ERK1/2-Fos pathways 20 . We demonstrated that Cr(VI) not only affected the self-renewal/ maintenance pathway but also the proliferation/differentiation pathways in SSCs in a dose-dependent manner. Our results demonstrated that Cr(VI) (i) at 6.25 μ M (IC 50 ), induced cytotoxicity mainly through P53-dependent apoptosis; (ii) at 12.5 μ M, increased apoptosis and disrupted GDNF signaling pathway; and (iii) at 25 μ M, decreased apoptosis (like 6.25 μ M) and further disrupted GDNF signaling pathway (compared to 12.5 μ M). In conclusion, the loss of cell viability observed at higher doses of Cr(VI), involves the apoptosis pathways and impairment of the GDNF signaling pathway.
Cr(VI) toxicity has been reported to disrupt steroidogenesis and decrease serum testosterone levels and 3β -HSD enzyme activity in male rats [48][49][50] . We also observed that Cr(VI)-treatment significantly decreased the testosterone release from the TM3 cells into the culture media. Besides, Cr(VI) exposure in TM3 cells significantly decreased transcriptional expression of Cyp11a1 (a rate-limiting enzyme in steroidogenesis that converts cholesterol to pregnenolone in mitochondria) and Hsd3b1 (that converts pregnenolone to progesterone in the endoplasmic reticulum), without significant alteration in the levels of Hsd17b3. In contrast, Cr(VI) increased the expression of Cyp17a1 and Star, possibly through a negative feedback compensatory mechanism counterbalancing the Cr(VI)-induced inhibition of Cyp11a1 and Hsd3b1. Steroidogenesis is a complex multi-enzyme process wherein steroid hormones are produced from cholesterol. A number of researchers have shown that impairment of steroidogenesis by environmental toxicants does not require the inhibition of all the steroidogenic enzyme genes [51][52][53] . In the present study, therefore, Cr(VI) impaired steroidogenesis, at least to some extent, by inhibiting Cyp11a1 and Hsd3b1 in TM3 cells. Cr(VI) also decreased the expression of Cyp19a1, which is required for the conversion of androgens into estrogens. In TM4 cells, Cr(VI) decreased the transcriptional expression of the tight-junction signaling molecules, Tjp1, Vim, and Ocln, which play a key role in forming the blood-testis barrier 21 . These results are supported by those of the previous report where the authors showed that Cr(VI) altered the Sertoli cell barrier, which might affect the blood-testis barrier 54 .
Cr(VI) also decreased the transcriptional expression of the cell receptors, Fshr and Ar, which play a key role in the maturation and normal functioning of Sertoli cells 22,23 and those of some secreted molecules, including Gdnf, Etv5, and Fgf2, which are essential for the maintenance of SSCs 55 within the stem cell niche. Our results suggest that Cr(VI)-toxicity may also affect the maintenance of SSCs in vivo. Thus, Cr(VI)-induced impairment of the physiological functions of TM3 and TM4 cells could directly affect normal spermatogenesis.
In conclusion, Cr(VI) induced mitochondria-dependent apoptosis in male somatic cells and SSCs, possibly through the induction of oxidative stress and DNA damage (Fig. 8). The cell-specific pattern of regulation of P-AKT1, P-ERK1/2, and P-P53 proteins is probably responsible, at least in part, for the differential sensitivity of TM3 and TM4 cells to Cr(VI). Cr(VI) also disrupted the differentiation and self-renewal mechanisms of SSCs and impaired the physiological functions of TM3 and TM4 cells. To the best of our knowledge, our study is the first to uncover the underlying signaling mechanisms of Cr(VI)-induced cytotoxicity and apoptosis in male somatic cells and SSCs. Our experimental findings could provide a basis for the development of improved antagonists of Cr(VI) to combat Cr(VI)-induced reproductive toxicity.

Methods
Materials. K 2 Cr 2 O 7 (molecular weight, 294.18) was purchased from Duksan pure chemical Company Limited, (Ansan, South Korea). Penicillin-streptomycin solution, trypsin-EDTA solution, DMEM, and 1% antibiotic-antimycotic solution were obtained from Life Technologies GIBCO (Grand Island, NY, USA). Fetal bovine serum (FBS) and an in vitro toxicology assay kit were purchased from Sigma-Aldrich (St. Louis, MO). The antibodies used for immunoblotting were against phospho P53, phospho ERK1/2, total ERK1/2, total AKT1 (Cell Signaling Technology, Beverly, MA), phospho P38 (Santa Cruz Cell viability and cytotoxicity assay. Cells were seeded (5 × 10 5 cells/well) onto 96-well, flat bottom culture plates and incubated for 24 h at 37 °C in a 5% CO 2 incubator. The used medium was replaced by fresh DMEM containing 1% FBS. Cells were then treated with different concentrations of K 2 Cr 2 O 7 (0, 3.125, 6.25, 12.5, 25 and 50 μ M) for 24 h in a humidified incubator at 37 °C in the presence of 5% CO 2 . A cell viability assay was performed using the Cell Counting Kit-8 (CCK-8, Dojindo Laboratories, Kumamoto, Japan). The absorbance was read at a wavelength of 450 nm, using a microtiter plate reader (Multiskan FC, Thermo Fisher Scientific Inc., Waltham, MA, USA). Cytotoxicity was assessed by LDH assay in the supernatant medium using an LDH Cytotoxicity Detection kit (Takara Bio Inc., Tokyo, Japan), according to the protocol of the manufacture, by measuring the absorbance at 490 nm using a microplate reader. TUNEL/TMR red assay. Cells were grown on glass slides and, after reaching 80% confluency, were treated with K 2 Cr 2 O 7 . The cells were washed with PBS, fixed with 4% paraformaldehyde, washed again, and incubated with 0.1% Triton X-100. TUNEL analysis was performed to measure the degree of cellular apoptosis, using an in situ Cell Death Detection Kit, TMR red (Roche Diagnostics, Indianapolis, IN) according to the manufacturer's instructions.

Assessment of apoptotic cell populations. Cell death was analyzed using FITC conjugated
Annexin V and propidium iodide with an Apoptosis detection kit (Life Technologies, Oregon, USA) according to the manufacturer's instructions. Cells were characterized using FACSCalibur and the data were analyzed by Cell Quest software. Each analysis recorded 10 4 events. Preparation of CM from Cr(VI)-treated TM3 cells and culture of TM4 cells with CM. TM3, Leydig cells were divided into four groups and treated with different concentrations of K 2 Cr 2 O 7 (0, 6.25, 12.5 and 25 μ M) for 24 h in a humidified incubator at 37 °C in the presence of 5% CO 2 . Then, the medium were removed, the cells were washed with PBS, and cultured for 48 h in fresh medium without Cr(VI). The supernatant from each group was harvested, centrifuged to remove debris and mixed (1:1) with fresh medium to prepare conditioned medium. This CM was used to culture the TM4, Sertoli cells for 48 h and after that, we measured the mRNA expression of tight-junction signaling molecules, cell receptors, and factors responsible for SSC survival in TM4 cells.
Measurement of testosterone level. TM3, Leydig cells were divided into four groups and treated with different concentrations of K 2 Cr 2 O 7 (0, 6.25, 12.5 and 25 μ M) for 24 h in a humidified incubator at 37 °C in the presence of 5% CO 2 . Then, the medium were removed, the cells were washed with PBS, and cultured for 48 h in fresh medium without Cr(VI). The supernatant from each group was harvested, centrifuged to remove debris and the testosterone levels were measured using a spectrophotometric ELISA kit (ENZO Diagnostics Inc., Farmingdale, NY), following the manufacturer's instructions.
Gene expression analysis. mRNA from the treated cells was extracted using an RNeasy Mini Kit (Qiagen) and cDNA was synthesized using a QuantiTect Reverse Transcription Kit (Qiagen) in a final volume of 20 μ L, according to the manufacturer's instructions. All gene transcripts were measured in triplicate by qRT-PCR on a LightCycler apparatus, using LightCycler ® FastStart DNA Master SYBR Green I with an AB Applied Biosystems machine. The primer sequences for each gene are shown in Table S1. The relative quantification of gene expression was analyzed by the 2-ddCt method 58 . In all experiments, Gapdh mRNA was used as an internal standard.
Immunoblotting. Cells were lysed in radioimmunoprecipitation (RIPA) lysis buffer containing protease and phosphatase inhibitors. Equal amounts of protein were resolved by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and proteins were electrophoretically transferred to PVDF membranes. Membranes were blocked at room temperature with 5% non-fat dry milk for 2 h to prevent non-specific binding, and then incubated with primary antibodies overnight at 4 °C. Immunoreactivity was detected through sequential incubation with horseradish peroxidase-conjugated secondary antibodies and enhanced chemiluminescence reagents.

Statistical analysis.
All the experiments were performed at least in triplicate, and statistical analyses were performed by one-way analysis of variance (ANOVA) followed by Dunnett's t-test to determine the significance of difference between the treatment and control groups. Statistical tests were performed using Stat View version 5.0 (SAS institute, Cary, NC, USA). The level of significance was set at *p < 0.05, **p < 0.01 and ***p < 0.001 compared to the control.