Loss of Klotho contributes to cartilage damage by derepression of canonical Wnt/β-catenin signaling in osteoarthritis mice

HE staining and acin blue staining findings

Findings of representative photomicrographs HE staining and acin blue staining of joint pathological sections of normal mouse joints and OA model mice were available twelve weeks after ACLT. Intact and smooth cartilages were showed on sham-operated knee joints (Figure 1A, 1B). Additionally, a typical healthy cartilage was presented in the chondrocyte organization with flat cells in the superficial zone (one or two layers). In the middle and deep zones, columns of round cells were seen. In contrast, in male C57/6J mice, cartilage damage and disorganized chondrocyte were presented in the operated knees (ACLT).

Figure 1. Hmatoxylin and eosin (H&E), and acin blue staining. (A) Representative photomicrographs HE staining of joint pathological sections of normal mouse joints and OA model mice, twelve weeks after ACLT (n=4). (B) Alcian blue staining of normal mouse joints and surgically prepared OA model mouse joint pathological sections (n=4).

Expression of Klotho, wnt/β-catenin signal pathway and its target gene in the OA model

To further examine the expression of Klotho in OA, a mouse model was established. Cartilage tissue was removed from the knee joint with anterior cruciate ligament transection, and then the expression of Klotho in articular cartilage of OA mice was observed. The results showed that the level of Klotho in OA mice decreased markedly (Figure 2A, 2D) after operation, and the immunostaining intensity of Klotho in the articular cartilage of OA mice decreased (Figure 2G).

Figure 2. Up-regulation of Wnt signaling and down-regulation of Klotho in the joint after injury. ACLT surgery was used to induce OA in adult male C57/6J mice. Expression of Klotho, Phos-GSK3β (S9), GSK3β, MMP2, MMP9, Wnt1, Wnt4, Wnt7a and β-catenin (cytoplasm/nucleus) in the mouse cartilage tissue. (A) Expression change of Klotho, Phos-GSK3β (S9), GSK3β, MMP2, MMP9, Wnt1, Wnt4, Wnt7a, and (B) β-catenin in the cytoplasm and nucleus by Western Blot after modeling. (C) Protein levels of Klotho, Phos-GSK3β (S9), GSK3β, MMP2, MMP9, Wnt1, Wnt4, Wnt7a and β-catenin were detected by western blotting. Protein/β-actin and Protein/Lamin A were used to demonstrate the protein fold changes. (D) Klotho mRNA expression was assessed in the cartilage at 12 weeks after ACLT or control. (F) The expression of matrix-degrading enzyme genes, MMP-2 (E), MMP-9 (F) were evaluated by Real-Time PCR. (G–J) The expression of Klotho protein and β-catenin protein in the control group and OA group were compared by immunohistochemistry,. Scale bar, 50 mm. Data are the mean ±SD, n=8 mice. *P < 0.05, **P < 0.01.

On the contrary, the increased expression of Wnt signaling pathway and its target gene MMP2 and MMP9 in the cartilage of OA mice, and the increased expression of Phos-GSK3β (S9), Wnt1, Wnt4 and Wnt7a were observed by western blot, indicating that the Wnt signaling pathway was activated in OA (Figure 2A).

Wnt/β-catenin signal transduction pathway regulates the expression of MMPs gene downstream [18, 19]. MMPs are the primary enzymes that damage the ECM of the articular cartilage. Our results confirmed that, after activation of Wnt signaling pathway, the expression of the target gene MMP-2 and MMP-9 in OA mice increased (Figure 2A, 2E and 2F).

The β-catenin, the core protein in the Wnt signaling pathway, was transferred from the cytoplasm to the nucleus after activation. The results showed that in the OA model, when compared to the normal group, the β-catenin in the nucleus increased (Figure 2B) and the intensity of Klotho immunostaining in the articular cartilage of the OA mice decreased (Figure 2I).

In cartilage tissue of OA model, the binding of Dvl to Frizzled, Klotho to Wnt1, Wnt4, Wnt 7a and β-catenin to MMP2, 9 promoters were detected

Dishevelled (Dvl), a key protein in the Wnt signaling pathway, can transfer the Wnt signal from the receptor to the downstream effector. In the classical Wnt signaling pathway, Dvl is convoked to the cell membrane by the receptor Fizzled, and wherefore binds to the axin/GSK3/APC complex, which consequently inhibits the degradation of β-catenin in the cytoplasm. In the cartilage tissue of OA mouse model, the Dvl and Fizzled binding markedly increased (Figure 3A) by Co-IP method.

Figure 3. Klotho binds to Wnt and blocks Wnt-mediated gene transcription. In cartilage tissue, the binding of Dvl to Frizzled, Klotho to Wnt1, Wnt4, Wnt7a and β-catenin to MMP2, 9 promoters were detected. (A, B) The Co-IP method was used to detect the binding of Dvl to Frizzled, and Klotho to Wnt1, Wnt4, Wnt7a in cartilage tissue after operation, (C) and quantitative data for Dvl to Frizzled, and Klotho to Wnt1, Wnt4, Wnt7a are presented. (D, E) The binding of β-catenin to MMP2, 9 promoters was detected by ChIP method. Data are the mean ±SD, n=8 mice.. *P < 0.05, **P < 0.01.

Previous experiments suggest that Klotho can bind with Wnt1, 4 and 7a to form a complex, inhibits the activation of Frizzled, LRP 5/6 complex by Wnt, and therefore plays a similar role to Wnt1,4,7a antagonist. In the OA model, we also found that Klotho and Wnt1, 4 and 7a binding decreased when compared to normal cartilage tissue (Figure 3B).

ChIP method was used to detect the binding of β-catenin to MMP 2, 9 promoters. The results showed that the combination of TCF and MMP 2,9 promoters increased in the OA model (Figure 3D, 3E).

Expression of the Klotho, wnt/β-catenin signal pathway and its target gene in the chondrocytes stimulated by different CTS in the vivo experiment

The same findings were observed in vivo experiment. To further examine the expression of Klotho in chondrocytes and simulate the cartilage stimulated by stress in vivo, we established a model of OA chondrocytes exposed to CTS in vitro. Klotho protein level in the supernatant of chondrocytes was decreased with the increase of CTS intensity (Figure 4A and 4E) after 48h treatment with CTS (0.1, 0.5 and 1HZ). Western blot results showed that expression of Phos-GSK3β (s9) and Wnt1, Wnt4 and Wnt 7a in the supernatant of Wnt pathway was increased with the increase of CTS intensity (Figure 4A), and the nuclear Wnt pathway core protein β-catenin was decreased (Figure 4B), while the expression of Wnt target gene MMP-2 and MMP-9 increased (Figure 4F, 4G).

Figure 4. Up-regulation of Wnt signaling and down-regulation of Klotho in chondrocytes treated by CTS. Chondrocytes were were pretreated with cyclic tensile strain (0.1-1Hz) for 48 hours Expression of Klotho, Wnt1, Wnt4 and Wnt7a in the cell supernatant and GSK3β, p-GSK3β, MMP2, MMP9 and β-catenin (cytoplasm/nucleus) in the cell of mouse cartilage. (A) Western blot analyses show that CTS induced these proteins expression in a tension-dependent manner. (B) Cytocol/nuclear separation was undertaken to assess the enrichment of β-catenin in the nuclei in chondrocytes treated by CTS. The upregulation level of β-catenin localized in nucleus was more obvious than that remained in cytosol in CTS groups than control group. Representative Western blots (A and B) and quantitative data (C and D) are presented. *P<0.05 versus controls (n=3). (E) The RT-PCR results showed periodic CTS dependent inhibition of Klotho expression. (F, G) The results of RT-PCR showed that the expression of MMP-2, MMP-9 was stimulated by periodic CTS dependent stimulation. Results were obtained via the expression of three individual experiments performed in triplicate for each condition. *P < 0.05, **P < 0.01.

In vitro, CTS stimulated chondrocytes were used to detect the binding of Dvl to Frizzled, klotho to Wnt1, Wnt4 and Wnt 7a in the cell supernatant and cell, and the effect of β-catenin on the activity of MMP2, 9 promoters were detected in the vitro CTS stimulated chondrocytes

In vitro experiments, the binding level of Dvl and Frizzled in the chondrocytes increased with the increase of the CTS intensity (Figure 5A). In contrast, the binding level of Klotho and Wnt 1, Wnt 4 and Wnt 7a in the supernatant of chondrocytes was decreased with the increase of CTS intensity (Figure 5B). And the effect of β-catenin on the activity of MMP2, 9 promoters was increased with the increase of CTS intensity (Figure 5D, 5E).

Figure 5. CTS influenced Wnt/β-catenin signaling pathway, Klotho binds to Wnt and blocks Wnt-mediated gene transcription. Chondrocytes were incubated with or without CTS(0.1, 0.5.1 HZ) for 48 hours. the binding of Dvl to Frizzled was induced by CTS in a tension-dependent manner, Klotho to Wnt1, Wnt4, Wnt7a was adversed. (A, B) Detection of Dvl and Frizzled binding and (B) Klotho and Wnt1, 4, 7a binding in supernatant by Co-IP, quantitative data (C) is presented. *P< 0.05 versus controls (n=3). (D, E) The effect of β-catenin on the activity of MMP-2 promoters and MMP-9 promoters was in a tension dependent manner. Results were obtained via the expression of three individual experiments performed in triplicate for each condition. *P < 0.05, **P < 0.01.

Expression of wnt/β-catenin signaling pathway and its target gene in chondrocytes over-expression of Klotho

The regulation of Klotho over-expression on the formation of OA induced by CTS was further studied. First, the chondrocyte was transfected with Klotho-overexpressing lentiviral vector or negative control lentivirus vector. The results showed that in the chondrocytes over-expression of Klotho, after increase of the Klotho expression, CTS could not increase the expression of Wnt signal pathway related proteins, including Phos-GSK3β (S9) in cells, and Wnt1, Wnt4 and Wnt7a) in the supernatant (Figure 6A). Over-expression of Klotho inhibited CTS's promotion of β-catenin from the cytoplasm to the nucleus (Figure 6B). The over-expression of Klotho reduced the transcription level of downstream target gene MMP-2 and MMP-9 (Figure 6A, 6E, 6F).

Figure 6. Klotho inhibits Wnt-mediated β-catenin activation and represses its target genes in vitro. Chondrocytes were transfected with empty vector or Klotho expression vector as indicated, followed by incubation with or without CTS(1 HZ) for 48 hours. Western blot (A, B) and quantitative data (C, D) for Wnt1, Wnt4 and Wnt7a in the cell supernatant and GSK3β, p-GSK3β, MMP2, MMP9 and active β-catenin are presented. Klotho abolished CTS-mediated Wnt actions. (E, F) RT-PCR was used to detect the expression of MMP-2 and MMP-9 mRNA in different groups. Results were obtained via the expression of three individual experiments performed in triplicate for each condition. *P < 0.05, **P < 0.01.

In chondrocytes overexpressed with klotho, the binding of Klotho to Wnt1, Wnt4 and Wnt7a, the binding of Dvl to Frizzled, and the effect of β-catenin on the activity of MMP2 and MMP9 promoters were observed

In the chondrocytes overexpressed with Klotho, Co-IP experiments confirmed that Klotho reversed the combination of Dvl and Frizzled induced by CTS (Figure 7A). On the other hand, under the action of CTS, the binding of Wnt1, Wnt4, Wnt7a to Klotho in chondrocytes with over-expression of Klotho was increased (Figure 7A). The cell supernatant showed the same result (Figure 7B). Over-expression of Klotho reduced the binding level of downstream target gene β-catenin and MMP-2 and MMP-9 promoters by luciferase assay (Figure 7E, 7F).

Figure 7. Over-expressed Klotho binds to Wnt and blocks Wnt-mediated gene transcription in vitro. Chondrocytes were incubated with or without CTS(1 HZ) for 48 hours. Overexpression of klotho attenuated the activity of the wnt signaling pathway (A) Co-IP method was used to detect the binding of Klotho and Wnt1, Wnt4 and Wnt7a in different groups of cells supernatant. (B) Co-IP method was used to detect the binding of Klotho and Wnt1, Wnt4, Wnt7a, the binding of Dvl to Frizzled in different groups of cells, quantitative data (C, D) is presented. *P,0.05 versus controls (n=3), The cell lysates were immunoprecipitated by β-catenin antibody. qPCR analyses the enrichment of MMP-2, MMP-9-specific DNA in chondrocytes. (E, F) Overexpression of klotho blocks the binding of β-catenin on MMP-2 and MMP-9 promoter activity. Results were obtained via the expression of three individual experiments performed in triplicate for each condition. *P < 0.05, **P < 0.01.

Expression of Klotho, Wnt1, Wnt4 and Wnt7a in the supernatant, GSK3β, p-GSK3β, MMP2, MMP9, MMP-7, active β-catenin and β-catenin (cytoplasm/nucleus) in β-catenin silencing chondrocytes

β-catenin is a multifunctional protein in the cytoplasm and a key molecule regulating gene transcription in Wnt signaling pathway. First, the chondrocyte was transfected with β-catenin low expressing lentiviral vector or negative control lentivirus vector. The results showed that there was no markable change in the expression of Klotho in the supernatant of chondrocytes between the β-catenin silencing chondrocytes and the normal cells, both decreased (Figure 8A). However, the expression of Wnt signal-related proteins p-GSK3β, WNT1, WNT4 and WNT7a decreased after silencing β-catenin, indicating that the Wnt signaling pathway was silent after β-catenin silencing. After this silencing, the β-catenin in chondrocytes could not be transferred from cytoplasm to the nucleus in the presence of CTS (Figure 8B). The silencing β-catenin has reduced the transcription level of downstream target gene MMP-2, MMP-9 (Figure 8A, 8E, 8F). These results suggest that. Klotho is in the upstream of WNT/β-catenin signaling pathway, and Klotho regulates OA through Wnt signaling pathway

Figure 8. Expression of Klotho, Wnt1, Wnt4, Wnt7a in the cell supernatant and GSK3β, p-GSK3β, MMP2, MMP9, MMP-7, β-catenin (cytoplasm/nucleus) in the cell of chondrocytes knockdown of β-catenin. (A) The protein expression detection of Klotho, Wnt1, Wnt4, Wnt7a in the cell supernatant and Phos-GSK3β (S9), GSK3β, MMP2, MMP9, and (B) β-catenin (cytoplasm/nucleus) by Western Bolt, quantitative data (C) is presented. *P,0.05 versus controls (n=3), (E, F) RT-PCR was used to detect the expression of MMP-2 and MMP-9 mRNA in different groups. Results were obtained via the expression of three individual experiments performed in triplicate for each condition. *P < 0.05, **P < 0.01.

In the β-catenin silencing chondrocytes, the binding of Klotho to Wnt1, Wnt4, Wnt7a, the binding of Dvl to Frizzled, and the effect of β-catenin on the activity of MMP2 and MMP9 promoters were observed

In the β-catenin silencing chondrocytes, Co-IP assay confirmed that there was no markable change in the binding of Dvl to Frizzled with or without CTS stimulation (Figure 9A). Silencing β-catenin led to the decrease of Wnt1, 4 and 7a expression, and the binding of Wnt1, Wnt4 and Wnt7a to Klotho in β-catenin silencing chondrocytes was also reduced by CTS, which was same in the cell supernatant (Figure 9A, 9B). Silencing β-catenin reduced the binding level of downstream target gene β-catenin, MMP-2 and MMP-9 promoters by luciferase assay (Figure 9E, 9F).

Figure 9. In chondrocytes knockdown of β-catenin, the binding of Dvl to Frizzled, Klotho to Wnt1, Wnt4, Wnt7a and β-catenin to MMP2, 9 promoters were detected. (A) Co-IP method was used to detect the binding of Klotho and Wnt1, Wnt4, Wnt7a in different groups of cells supernatant. (B) Co-IP method was used to detect the binding of Klotho and Wnt1, Wnt4 and Wnt7a, the binding of Dvl to Frizzled in different groups of cells. quantitative data (C, D) is presented. *P,0.05 versus controls (n=3), The cell lysates were immunoprecipitated by β-catenin antibody. qPCR analyses the enrichment of MMP-2, MMP-9-specific DNA in chondrocytes. (E, F) MMP-2 and MMP-9 promoter luciferase activity decreased after silencing β-catenin. Results were obtained via the expression of three individual experiments performed in triplicate for each condition. *P < 0.05, **P < 0.01.

Reagents

Monoclonal antibodies against Phos-GSK3β (S9), GSK3β, MMP2, MMP9, Wnt1, Wnt4, Wnt7, Frizzled and β-catenin were provided by Abcam (Cambridge, UK). Anti-Dvl Antibody was provided by Cell Signaling Technology (Danvers, MA and USA). Antibodies against Klotho was provided by Abcam and Santa Cruz Biotechnology Fetal bovine serum and DMEM low- and high-glucose medium were procured form HyClone (Logan, UT and USA) Phosphate-buffered saline (PBS), cytoplasmic and membrane protein extraction kit, total protein extraction kit, RIPA buffer and PMSF were procured from Beyotime Biotechnology (Nantong, China). ChIP Assay Kit was obtained from Beyotime Biotechnology, (Shanghai, China). Pierce™ Classic Magnetic IP/Co-IP kit was obtained from Thermo Fisher (Shanghai, China). Acin blue staining kit was provided by Nanjing Jiancheng Bioengineering Institute (Nanjing, China).

Animals

Twenty-four C57/6J mice (19-21gm, 3-month-old) were obtained from the Animal Inc., affiliated to Nanjing Medical University (Nanjing, China). These mice were separated into two groups randomly (control group and OA group respectively), consisting of 12 mice for each, which were kept in four animal cages (six mice per cage) in a temperature-controlled room (21–23°C) with free access to water and food. All these animal procedures were approved by the Animal Research Ethics Committee of Nanjing Medical University.

Anterior cruciate ligament transaction (ACLT)

ACLT surgery was used to induce OA in adult male C57/6J mice. Briefly, all mice were anesthetized with chloral hydrate. A medial parapatellar approach was adopted to expose the right knee joint. After that, a transection was conducted with micro-scissors in the anterior cruciate ligament (ACL) in the OA group with a positive anterior drawer sign to confirm the completeness of transection. In the control group, the arthrotomy was conducted while the transection of ACL was not conducted. After surgery, the mice were released from the cages for 30 minutes on a daily basis.

HE and alcian blue staining

The mice were sacrificed by cervical dislocation 12 weeks after surgery and under anesthesia. The knee joint cavity was exposed by separating the patella. The fixed specimens were decalcified for 3 weeks by using 10% ethylenediaminetetraacetic. Upon decalcification, the samples were embedded in paraffin; then the standard serial sections of 3μm each were cut in sagittal plane before the sections were stained by hematoxylin and eosin (H&E), and Alcian blue staining.

Immunohistochemistry

For immunohistochemical analysis of Klotho and β-catenin expression in articular cartilage of mice, the paraffin-embedded tissues in full-thickness were processed in this study. A blocking serum (Vectastain ABC Kit, Vector Laboratories, Inc., Burlingame, CA, USA) was used to incubate the slides for 60 minutes. Then the slides were overlaid with primary antibody against Klotho and β-catenin for 2 hours at room temperatures after being blotted. Subsequently, the sections were added with biotinylated secondary antibodies before they were processed by the streptavidin-biotin staining technique (DAB Kit, Invitrogen, Paisley, UK), which is peroxidase-labeled.

Chondrocyte culture, cyclic tenisle strain

The first step is to separate chondrocytes from articular cartilage of mice’s knee joints. In addition, the articular cartilage tissues were digested for 30 minutes by using 0.25% trypsin after being cut into small pieces (<1 mm). After this step, digestion using 0.2% Type II collagenase was performed for 3 hours. DMEM/F12 media, antibiotics and 10% fetal bovine serum (FBS) were then used to culture the released cells. After the confluence increased up to 80%, the cells were subjected to cyclic tensile strain for 48 hours at 20% elongation with a 0.1-1 Hz sinusoidal curve indicated by a Flexcell1 FX-5000 Tension System following the user’s instructions (Flexcell International Corporation, Burlington, NC, USA). In order to preserve the chondrocyte phenotype, only the cells not exceeding two passages were adopted.

Cell transfection

Both empty lentiviral vectors and lentiviral vectors encoding Klotho or β-catenin were acquired from Genechem (Shanghai, China). Lentiviral transfection was conducted by following the manufacturer’s instructions. Green fluorescent protein (GFP) was used to assess the infection efficiency by using flow cytometry. 5μg/ml puromycin (Sigma-Aldrich, USA) was used to select stably transfected cells after 48 hours.

Western blotting analysis

Total protein from cell supernatant, chondrocyte and cartilage was prepared by RIPA buffer, the nuclear protein was isolated by using Nuclear Protein Extraction Kit (Beyotime) strictly in accordance with the Kit instructions, the protein mixtures were separated by gel electrophoresis in sodium dodecyl sulfate-polyacrylamide, which were then translated to nitrocellulose membranes. Briefly, 10% SDS-PAGE was used to extract 15μg plasma proteins from the cartilage tissue and chondrocytes. In addition to that, the proteins were electroblotted onto a polyvinylidene difluoride membrane with a thickness of 0.45 mm (Millipore, Bedford, MA, USA), which were then blocked via 5% non-fat dry milk in Tris-buffered saline with Tween 20 for one hour. Incubation of these membranes was then performed at 4 °C overnight with Klotho, Phos-GSK3β (S9), GSK3β, MMP2, MMP9, Wnt1, Wnt4, Wnt7 and β-catenin antibodies at room temperatures for one hour. To visualize antigen-antibody complexes, the enhanced chemiluminescence assay (Thermo Scientific, Pierce, Rockford, IL, USA) was applied.

Co-immunoprecipitation

Immunoprecipitation lysis buffer (30mm Tris-HCl, 150mm NaCl, 1% Triton X-1002 mm EDTA, 10% glycerol and 1×protease inhibitor) was used to homogenize cartilage tissue on ice before a short step of brief pulses of hand sonication. Following the step of 30-minute incubation on ice, a centrifugation operation was conducted for the lysate and then Protein A/G PLUS-agarose beads were used to preclear the supernatant by incubation at 4 °C. After that, 2μg antibody was used to incubate the precleared supernatant on a rocker at 4 °C for 4 hours against Klitho. The centrifugation was then performed to collect immune-precipitates at 1,000×g for five minutes after adding beads for one hour additionally. Immunoprecipitation lysis buffer was then used to wash the obtained immune-precipitates for 3 times, which were subsequently centrifugated after each wash. 2×sample buffer was used to resuspend and boil the pellet before Western blot was used in the analysis of Wnt1, 4 and 7a by the method as described above.

Quantification of mRNA and qRT-PCR

TRIzol reagent (Invitrogen) was used to extract entire RNA from chondrocytes, which was subsequently reversely transcribed into cDNA with the tool of PrimeScript RT reagent kit (TaKaRa, Dalian, China). SYBR Premix Ex Taq II reagents which were provided by TaKaRa (Dalian, China) were used to analyze PCR in a quantitative real-time manner, following the manufacturer’s instructions. 20μL mixture with 2μL cDNA was used to demonstrate PCR results. Specific primers (TaKaRa) were used to amplify Klotho cDNAs. Glyceraldehyde-3-phosphatedehydrogenase (GADPH) was used to normalize gene expression, which was then derived by the 2-DDCT approach. These are the primer sequences: Klotho: forward, 5′-ACTACGTTCAAGTGGACACTACT-3′ and reverse 5′-GATGGCAGAGAAATCAACACAGT-3′; MMP-2: forward, 5′-CAAGTTCCCCGGCGATGTC-3′ and reverse 5′-TTCTGGTCAAGGTCACCTGTC-3′; MMP-9: forward, 5′-CTGGACAGCCAGACACTAAAG-3′ and reverse 5′-CTCGCGGCAAGTCTTCAGAG-3′.

Chromatin immunoprecipitation (ChIP) assay

ChIP assays were performed essentially as described [22]. Nuclear protein was then isolated from cartilage tissue and chondrocyte. Diluted soluble chromatin fractions were immunoprecipitated with 1μg polyclonal anti-β-catenin antibody or control mice immunoglobulin G (IgG). The histone-DNA crosslinks were reversed by 4-hour incubation at 65°C. PCR was used to analyze the DNA fragments from ChIP assays. These are the primer sequences: MMP-2: forward, 5′-GAGGTCGCTTTCTTTGCCATCT-3′ and reverse 5′-AGCGACTCCATCTTGAACAGG-3′; MMP-9: forward, 5′-CACTGTATCCTTGACCTTCTTTCTGG-3′ and reverse 5′-GCTTCCTCTCCCTGCTTCATCTG-3′.

Luciferase assay

Cells were co-transfected by using Lipofectamine 2000 reagent (Life Technologies) with mmp2/mmp9-Luc luciferase reporter constructs (1.0μg) plus shScrambled or shβ-catenin expression plasmid (2.0 μg) (Genechem), respectively. A fixed amount (0.1 μg) of internal control reporter Renilla reniformis luciferase driven under a thymidine kinase promoter(pRL-TK;Promega) was also cotransfected for normalizing the transfection efficiency. Luciferase assay was carried out using a dual luciferase assay system kit according to the manufacturer's protocols (Promega). Relative luciferase activity was reported as fold induction over controls after normalizing for transfection efficiency.

Statistics

Statistics have shown ±SEM of the triplicate values for each experiment. One-way analysis of variance (ANOVA) was used to assess the difference between experiment and control groups. When the P value is <0.05, the difference is significant.