HK-2 human kidney-2 cells and A172 human glioma cells were purchased from American Type Culture Collection (ATCC) and maintained in DMEM (Hyclone) contained with 10% heat-inactivated FBS, 100 units/ml penicillin and 100 mg/ml streptomycin at 37℃ in 5% CO₂ atmosphere. HK-2 cells were exposed to 50 µM of H₂O₂ in glucose/serum free DMEM for 3 h and then further incubated in the normal medium supplemented with glucose and serum. To observe the protective effect of ROS scavenger on the induction of Bis, 1 mM of N-acetyl-L-cysteine (NAC, Sigma-Aldrich) was treated to HK-2 cells before exposure to H₂O₂. For OGD, A172 cells were washed twice with degassed DMEM without glucose and serum (WelGene), and exposed to an anaerobic chamber containing 85% (v/v) N₂, 10% (v/v) H₂, 5% (v/v) CO₂ (Thermo Forma) at 37℃ for 5 h. And then A172 cells were recovered in the normal medium in the indicated times.

The cells were transfected with siRNA into 48-well plates and then maintained for 48 h at 37℃ in an incubator in a 5% CO₂ atmosphere. Cells were exposed to H₂O₂ in glucose/serum free DMEM for 3 h and then further incubated for 6 h in the normal medium supplemented with glucose and serum. Thiazolyl blue (3-[4, 5-dimethylthiazol-2-yl]-2, 5-diphenytetrazolium bromide reagent, Sigma) was added to each well, to a final concentration of 0.5 mg/ml. After incubation at 37℃ in an incubator under a 5% CO₂ atmosphere, the solution was suctioned and 200 µl of acid isopropyl alcohol (Sigma) was added to each well. Absorption was then measured at a wavelength of 570 nm. After extraction with background value, the relative viability was determined by ratio of values to those of control in terms of mean±standard errors.

Total RNA was isolated using AccuZol (Bioneer, Daejeon, Korea), and first strand of cDNA was synthesized by reverse transcribed to cDNA using a ReverTra Ace qPCR kit (Toyobo). A quantitative real-time PCR was performed to validate the expression level of Bis and HSF1 mRNA using SYBR® Premix Ex Taq™ (TaKaRa Bio) on Applied Biosystems 7300 machine (Applied Biosystems). The specific primers for real time PCR include following: bis, 5'-AGCCCTCAGCACTGCCCCTGCAGAA-3' and 5'-GCAGCTCTTTGGTCAAATACTCTTC-3';HSF1, 5'-GGCCATGAAGCATGAGAATG-3' and 5'-GTTCAGCATCAGGGGGATCT-3' and β-actin, 5'-AGTACTCCGTGTGGATCGGC' and 5'-GCTGATCCACATCTGCTGGA-3'. The relative values for HSF1 and Bis mRNA was calculated after normalization to the Ct value from β-actin in the same sample using the ddCt method [30].

Cells were lysed with RIPA buffer (150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% sodiumdodecyl sulphate, 50 mM Tris-HCl pH 8.0) with protease inhibitor (Roche Diagnostics) on ice for 30 min. Equal amounts of protein were separated on 10% sodiumdodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride membranes (Millipore). The membranes were incubated for 1 h with 5% dry skim milk in TBST (20 mM Tris, 137 mM NaCl, 0.1% Tween 20) buffer and then incubated with antibodies against Bis [1], HSF1 (Cell Signaling), or beta-actin (Sigma-Aldrich). After incubation with horseradish peroxidase-conjugated anti-mouse or anti-rabbit IgG (1:5,000; Santa Cruz Biotechnology), the immunoreactive bands were visualized by an enhanced chemiluminescence substrate (Thermo Fisher Scientific). Quantification for the intensities of each band was carried out on Multi Gauge 2.2 software (Fuji Photo Film Co.).

Suppression of Bis or HSF1 expression was performed by transfection of specific siRNA targeted for Bis (5'-AAGGUUCAGACCAUCUUGGAA-3') or HSF1 (5'-AAGGUUCAGACCAUCUUGGAA-3') with G-fectin (Genolution Pharmaceuticals) according to manufacturer's instruction. Accu-Target™ Control siRNA (Bioneer, 5'-CCUACGCCACCAAUUUCGU-3) was used for negative control.

The 5'-flanking region of human bis genomic DNA with various length was amplified by PCR from HEK 293 genomic DNA using different distal primers and proximal primers, ranging -1080 and +289 (+1 represents the transcription initiation site) and subcloned into the reporter plasmid pGL4 (Promega). After transfection into HK-2 cells with Fugene 6 (Roche Applied Science), the luciferase activity was determined using the Dual-Luciferase reporter assay system (Promega). After normalization with renilla activity, the luciferase activities are presented as fold change relative to the normalized firefly luciferase activity in cells transfected with pBis-1(-1080/+289) without oxidative stress, which was taken as 1.0.

ROS accumulation in HK-2 cells following H₂O₂ treatment was determined using the oxidation sensitive fluorescent probe 2'7'-dichlorofluorescein diacetate (DCF-DA, Molecular Probes). This assay is based on the principle that the nonpolar, nonionic H2-DCFDA crosses cell membranes and is enzymatically hydrolyzed into nonfluorescent H2-DCF by intracellular esterases. In the presence of ROS, H2-DCF is rapidly oxidized to become highly fluorescent DCF. At the indicated time points, cells were washed three times with phosphate buffered saline, incubated for 10 min with 10 µM of DCF-DA, and then washed twice with PBS. Quantification of ROS levels were performed by flow cytometry analysis (BD FACSCalibur; BD Bioscience) with excitation at 488 nm and emission at 525 nm. Mean fluorescence intensities were obtained by histogram statistics using the BD FACSDiva software (BD Bioscience).

Student's t-tests were employed in order to compare the differences between two different groups. p value of<0.05 was considered statistically significant.

We previously reported that Bis expression was induced by oxygen-glucose deprivation in C6 glioma cells, as evidenced by a western blot assay [29]. To examine whether oxidative stress-induced Bis expression was restricted only to glia-derived cells, we determined Bis expression levels after a H₂O₂ treatment in HK-2 cells, human kidney proximal cells. As shown in Fig. 1A, treatment of the HK-2 cells with 50 µM of H₂O₂ for 3 h resulted in an increase in Bis expression at protein level by about 3.4 -fold, which was further increased after additional incubation in normal media in a time dependent manner, showing an increase of about 4.6- fold after 6 h. Quantitative real time PCR assays revealed that Bis transcripts levels were also increased by exposure to H₂O₂ by about 3.3-fold, and then further increased to 4.8-fold after incubation for 6 h in normal media, with a similar pattern of protein levels (Fig. 1B). The increase in Bis expression in recovery or reperfusion period was verified in A172 glioma cells which were exposed to OGD and subsequently allowed to recover in normal media supplemented with 5% O₂ (Fig. 1A and B). It thus appears that the signals for Bis induction are persistently activated even after oxidative stress source is eliminated. To define the physiological significance of Bis induction by oxidative stress, we evaluated the susceptibility to cell death induced by H₂O₂ following Bis silencing. Bis depletion resulted in the significant decrease of the relative viability both in HK-2 and A172 cells : 84% and 78% were decreased to 61% and 55% by Bis depletion with 100 nM of Bis siRNA in HK-2 and A172 cells, respectively (Fig. 1C). Thus, the induction of Bis upon oxidative stress might exhibit pro-survival activity to protect cells from entering apoptosis program. As determined by DCF-DA staining and subsequent FACS analysis, cellular ROS levels were higher in the recovery period than in H₂O₂ exposure period (Fig. 1D). Furthermore, the pretreatment with NAC, an antioxidant, significantly repressed ROS levels as well as Bis mRNA levels (Fig. 1D and E), indicating that ROS accumulation directly stimulates Bis expression at the transcriptional level.

Several types of stress such as heat shock and proteasome inhibitor have been reported to stimulate Bis expression through activation of HSF1 and subsequent its interaction with HSEs in the promoter of bis gene [22,23]. To investigate if oxidative stress-induced Bis induction is also mediated by the activation of HSF1, we investigated the effect of HSF1 siRNA on Bis induction upon a H₂O₂ treatment. As shown in Fig. 2A, the transfection of HSF siRNA effectively suppressed HSF1 protein expression. Silencing HSF1 had no effect on the basal expression of Bis at the transcriptional level as well as the protein level. However, H₂O₂-induced Bis expression was decreased in HSF1-silenced cells. The increase in Bis expression by H₂O₂ in HSF1-silenced cells was 2.3- fold and 3.5- fold at protein and the mRNA levels, respectively, which correspond to 63% and 48% of the control cells. To define the physiological significance of Bis induction mediated by ROS and HSF1, we measured ROS levels after a H₂O₂ treatment in control or HSF1 siRNA-treated HK-2 cells. A shown in Fig. 2C, the transfection of HSF1 siRNA had no effect on basal ROS levels. Treatment with H₂O₂ resulted in an increase in ROS accumulation by about 2.7-fold, which was further increased the 3.4- fold by HSF1 silencing. Thus, the HSF1 deficiency appeared to be associated with anti-oxidant defense, probably via the regulation of Bis expression.

It has been reported that three putative HSEs are present in the promoter of the Bis gene [22,31]. To define which region of the Bis gene promoter is responsible for oxidative stress-mediated Bis induction, we prepared a series of deletion constructs of the Bis gene promoter for use in a reporter assay (Fig. 3A). pBis-1, a luciferase plasimid containing all the three putative HSEs, showed similar time course profiles for the induction of luciferase activity with an increase in the endogenous Bis transcripts, showing a peak activity of about a 5-fold increase at 6 h after recovering from the H₂O₂ treatment. Thus, oxidative stress induced Bis expression can primarily be attributed to the activation of promoter activity, resulting in an increase in the transcripts level. The basal luciferase activity of pBis-2, which contain two HSEs (H1 and H2), was 1.5-fold higher compared to pBis-1. In addition, the transcriptional induction of pBis-2 by oxidative stress was also greater than pBis-1, by about 7- fold and 5-fold, respectively. Further deletion of HSE2 (pBis-3) or HSE1(pBis-4) resulted in a marked reduction in basal activity of 9 and 15% of pBis-1, while partially restored oxidative stress-responsive ability by about 2-fold leading to 16 and 33% of pBis-1, respectively. Finally, as observed in endogenous Bis induction, HSF1 silencing clearly suppressed the transcriptional activation of pBis-2 upon H₂O₂ exposure, from 2.6-fold to 0.8- fold (Fig. 3D). Taken together, these results suggest that both HSE1 and HSE2 are required for basal expression as well as oxidative stress-inducible expression of Bis.