SEARCH

SEARCH BY CITATION

Keywords:

  • REST/NRSF;
  • TRF2;
  • oxidative stress;
  • DDR

Abstract

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

REST is a neuronal gene silencing factor ubiquitously expressed in non-neuronal tissues. REST is additionally believed to serve as a tumor suppressor in non-neuronal cancers. Conversely, recent findings on REST-dependent tumorigenesis in non-neuronal cells consistently suggest a potential role of REST as a tumor promoter. Here, we have uncovered for the first time the mechanism by which REST contributes to cancer cell survival in non-neuronal cancers. We observed abundant expression of REST in various types of non-neuronal cancer cells compared to normal tissues. The delicate roles of REST were further evaluated in HCT116 and HeLa, non-neuronal cancer cell lines expressing REST. REST silencing resulted in decreased cell survival and activation of the DNA damage response (DDR) through a decrease in the level of TRF2, a telomere-binding protein. These responses were correlated with reduced colony formation ability and accelerated telomere shortening in cancer cells upon the stable knockdown of REST. Interestingly, REST was down-regulated under oxidative stress conditions via ubiquitin proteasome system, suggesting that sustainability of REST expression is critical to determine cell survival during oxidative stress in a tumor microenvironment. Our results collectively indicate that REST-dependent TRF2 expression renders cancer cells resistant to DNA damage during oxidative stress, and mechanisms to overcome oxidative stress, such as high levels of REST or the stress-resistant REST mutants found in specific human cancers, may account for REST-dependent tumorigenesis.

REST is a silencing factor against neuron-specific genes in non-neural tissues and neural precursors.1, 2 In addition to its role as a neuronal gene repressor, REST appears to be linked to cancer. Medulloblastoma, a malignant brain tumor showed unusually high levels of REST activity and expression compared to neuronal progenitor cells, fully differentiated neurons or normal cerebellum.3–5 Contrary to neuronal origin, REST has been suggested as a candidate tumor suppressor in non-neuronal cells. The REST gene is frequently deleted in breast, prostate, colon and small cell lung cancer cells originating from non-neuronal tissues.6 Knockdown of REST induces anchorage-independent growth via direct derepression of the akt2 and oncogenic tac1 genes in small cell lung cancer and breast cancer, respectively.7, 8 Despite reports on a tumor-suppressive function of REST, no mechanistic evidence addressing its role of REST as a tumor suppressor in vivo has been obtained to date.9

A recent elegant study using conditional REST knockout mice presented a strong in vivo evidence that REST does not act as a tumor suppressor in mouse colon carcinogenesis.10 Moreover, a number of studies have raised the converse possibilities that REST is a tumor promoter. REST is required for estrogen stimulation of cell cycle in estrogen receptor positive breast cancer cells and its loss increases the susceptibility to tamoxifen.11 Overexpression of degradation-resistant REST accelerates the progression of mitosis and consequent genomic instability through repression of mad2 involved in the spindle checkpoint in non-neuronal cancer cells like HCT116, HeLa and U2OS.12 However, several questions regarding the activity of REST as a tumor promoter remain to be answered. In particular, the cellular context or molecules that are differentially associated with the versatile functions of REST require further elucidation.

Telomeric Repeat Binding Factor 2 (TRF2) is a shelterin protein that specifically binds to the telomere duplex and protects telomeric DNA against shortening.13 TRF2 is critically implicated in telomere biology and cancer owing to its protective functions in telomere maintenance and DNA damage response (DDR). TRF2 can suppress both telomeric and non-telomeric DDR by blocking Chk2 activation through interaction with ATM14 or direct binding to Chk2.15 In agreement with the molecular function of TRF2, the protein is frequently overexpressed in different tumor types.16, 17 Conversely, inhibition of TRF2 has been shown to sensitize melanoma cells to telomere-damaging drug-induced apoptosis18 and telomere deprotection, leading to p53- and ATM-dependent apoptosis in telomerase-positive cells.19 Therefore, the regulatory mechanism of TRF2 expression is a key to explaining tumorigenesis in several cell types.

While both tumor-suppressive and promoting functions of REST in non-neuronal tissues have been reported, the detailed tumor-promoting mechanism of REST remains to be elucidated. To clarify the underlying mechanism, we initially confirmed REST expression in several cancer cell lines and normal tissues and performed an in-depth evaluation of the REST-dependent TRF2 pathway as an oxidative stress defense system, using HCT116 and HeLa as model cell lines. Our results showed that down-regulation of REST leading to increased apoptosis of tumor cells is propelled by oxidative stress, suggesting that sustainability of REST expression against the stress is an important factor for tumor promotion.

Material and Methods

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Cell culture and reagent

All cells (HCT116, SK-Hep-1, MCF7, HeLa, U2OS, A498, H460, LNCaP and SW1417) were purchased from American Type Culture Collection and cultured according to the guideline recommended by ATCC. Normal human colon, liver lysates and HCT8, RKO, SNU-475 and SNU-449 cell lysates were purchased from Abcam, Novus Biologicals and Santacruz.

NAC, MG132, cycloheximide and H2O2 were purchased from Sigma-Aldrich. Lipofectamine 2000 (Invitrogen) was used to transfect cells with plasmids or siRNA, according to the manufacturer's protocol. For establishment of stable cell lines, cells were cultured in a growth medium containing 2–3 μg/ml puromycin and selected for 4 weeks.

Plasmids, siRNA and shRNA

REST cDNA (pCR4-REST) was purchased from Benebiosis and employed as a template for REST-V5. PCR was performed with the cloning primers described in Supporting Information Table 1 using pCR4-REST as the template. The amplified REST gene was cloned into pcDNA3.1D/V5-His-TOPO vector with topoisomerase. The TRF2 plasmid was kindly donated by the Korea Research Institute of Bioscience and Biotechnology. For gene silencing, cells were transfected with either duplex siRNA specifically targeting REST or TRF2 or negative control siRNA (Bioneer) sequences of which are shown in Supporting Information Table 1. The lentiviral particles containing REST-targeting shRNA were purchased from Sigma-Aldrich.

MTS assay, clonogenic assay and apoptosis assay

To assess cell proliferation, the MTS assay was carried out according to the manufacturer's instructions (CellTiter 96 AQueous One Solution, Promega). Cells were incubated with growth medium containing the MTS (3-4.5-Carboxymethoxyphenyl)-2(4-sulfophenyl)-2H-tetrazolium) reagent for 2 h. The change in absorbance at 492 nm was monitored using a microplate reader (Tecan).

To analyze the clonogenic ability of cells, REST-targeted cells stably expressing shRNA were plated at a density of 500 cells/well on a 6-well plate. After 2 weeks, cells were fixed with 4% paraformaldehyde (PFA) for 15 min, and subsequently stained with 0.2% crystal violet solution.

Apoptotic cells were quantified with the Annexin V-FITC staining. For the Annexin V-FITC staining, the trypsinized cells were washed twice with cold PBS and then incubated with FITC Annexin V for 15 min at room temperature in the dark. The stained cells were subjected to FACS analysis (BD FACS Calibur) and data analyzed with CellQuestPro software.

Western blot and immunoprecipitation

For analysis of protein expression, harvested cells were washed with PBS and lysed with RIPA buffer on ice for 20 min. After brief sonication, cell lysates were clarified by centrifugation at 15,000g for 20 min at 4°C. Equal amounts of protein were subjected to Western blot analysis.

An in vivo ubiquitination assay was performed as described previously, with slight modification.20 Briefly, 500 μg of whole cell lysates were incubated with antibodies for 1 h at 4°C, and protein G-agarose added to the immunocomplex. After 1 h of incubation, the protein G-immunocomplex was heated in SDS protein loading buffer for 10 min at 90°C after extensive washing with RIPA buffer three times. Polyclonal TRF2, Ac-p53 (K382), phospho-p53 (Ser15), γ-H2AX and phospho-Chk2 (Thr68) antibodies were purchased from Cell Signaling, monoclonal p53 (DO-1) and polyclonal actin antibodies from Santa Cruz Biotechnology and polyclonal REST antibody from Millipore. The intensities of the protein bands were analyzed using the NIH Image J program.

Real-time quantitative PCR

Total RNA was isolated using the RNeasy mini kit (Qiagen), according to the standard protocol for mammalian cells, and reverse-transcribed to cDNA (Solgent). Synthesized cDNA was employed as the template for real-time quantitative PCR with target gene-specific primers. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as the control gene. Real-time quantitative PCR was performed with a LightCycler rapid thermal cycler (Roche Diagnostics) and SYBR Green I (Takara). The primer sequences are shown in Supporting Information Table 1. Relative gene expression was quantitated by using the 2−ΔΔCt method described earlier.21

Telomere length assay

To determine the relative telomere length, genomic DNA was extracted with the Wizard SV Genomic DNA Purification System (Promega), according to the manufacturer's instructions. DNA (2 μg) was digested with 20 U of Hinf1 and Rsa1 restriction enzymes for 2 h at 37°C. Fragmented DNA was subsequently resolved with electrophoresis on a 0.8% agarose gel and then transferred overnight to a positively charged nylon membrane (Hybond–N, Amersham) in 20× SSC buffer with the capillary method. Finally, transferred DNA was immobilized on nylon membrane by UV cross-linking at 120 mJ/ cm2 (UVP) and then hybridized with a digoxigenin-labeled probe specific for telomeric repeats, followed by incubation with a digoxigenin-specific antibody covalently coupled to alkaline phosphatase (TeloTAGGG Telomere Length Assay kit, Roche Diagnostics). After extensive washing of the membrane, the chemiluminescent signal of hybridization was detected by exposure to X-ray film.

ROS measurement

Cells were trypsinized and washed with HBSS supplemented with 1.5 mM EDTA. Cells were incubated with 2 μM 5(6)-carboxy-2′, 7′-dichlorofluorescein diacetate (Molecular Probes) in HBSS-EDTA solution for 20 min at 37°C in the dark. Fluorescence intensities were measured using a microplate reader (Tecan), with excitation and emission at 485 nm and 535 nm, respectively.

Statistical analysis

Statistically significant differences were evaluated using the Student's t-test in the Excel 2007 program. p values less than 0.05 were considered statistically significant and described with an asterisk (*).

Results

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

REST expression in non-neuronal cancer cell lines

Initially, we assessed REST expression in various types of cancer cells originating from non-neuronal tissues using Western blot analysis. The size of REST appeared significantly larger (about 200 kDa) than the molecular weight estimated from the amino acid composition. This band was used for the immunoblot analysis since the REST protein was specifically recognized by REST antibody and the levels modulated by overexpressing or silencing REST. While some cancer cell lines, such as A498 and H460, displayed relatively low REST levels, the protein was abundantly expressed in the majority of cancer cell lines tested (Fig. 1a). Comparing with cognate normal tissues, colon carcinoma and hepatocellular carcinoma cells showed higher expression of REST (Fig. 1b) and SW1417 colon carcinoma cell was REST-negative as previously described.7

thumbnail image

Figure 1. Expression of REST in various non-neuronal cancer cell lines and normal tissues. Expression of REST was analyzed using Western blot. (a) Each cell line was cultured and harvested, and equal amounts (50 μg) of whole cell lysates and (b) tissue lysates subjected to Western blot using antibodies against REST and actin. All data are representative of three independent experiments.

Download figure to PowerPoint

REST silencing reduced cell survival by activation of DDR in HCT116 and HeLa cells

Based on the finding that several cancer cell lines express higher levels of REST than normal tissues, we hypothesized that REST acts as a tumor promoter. Accordingly, the effects of REST silencing on cell survival were investigated in HCT116 cell lines expressing high levels of REST. siRNA-mediated knockdown of REST led to a significant reduction in cell viability, as assessed using the MTS cell proliferation assay (Fig. 2a, left panel). In addition, Annexin V-FITC labeling demonstrated that REST knockdown induces apoptosis (Fig. 2a, right panel). To clarify the mechanism underlying REST silencing-induced cell apoptosis in HCT116, we examined the expression patterns of marker proteins in the DDR. Knockdown of REST was associated with increased levels of p53, p-p53, γ-H2AX and phospho-Chk2 but decreased TRF2 levels, indicative of DDR activation (Fig. 2b). Moreover, stable knockdown of REST resulted in dramatic decrease in TRF2 expression, similar to that observed with transient knockdown of REST, and reduced colony formation of HCT116 (Fig. 2c) and HeLa cells (data not shown). As loss of TRF2 activates DDR by accelerating telomere shortening,19 we further investigated whether REST knockdown affects telomere length via lowering of the TRF2 level. In keeping with our theory, the telomere length assay revealed that REST-depleted C5 cells contained significantly shorter telomeres than control cells (Fig. 2d). Our data suggest that REST regulates DDR through effects on TRF2 levels and telomere length.

thumbnail image

Figure 2. Effects of REST silencing. (a) HCT116 and HeLa cells were transfected with 200 nM REST siRNA or control siRNA duplexes for 24 h, and equal numbers of cells seeded in a 96-well plate. After 36 h, the cells were incubated with MTS reagent for 2 h, and absorbance measured (left panel). Transfected cells were also stained with Annexin V-FITC and subjected to FACS analysis (right panel). (b) REST siRNA (200 nM) was introduced into HCT116 cells. At 48 h after transfection, protein levels of REST, p-p53, p53, γ-H2AX, p-Chk2, TRF2 and actin were assessed using Western blot. (c) HCT116 and HeLa cells were transduced with lentiviral particles containing plasmids expressing REST or negative control (NC) shRNA. Cells were transduced with five lentiviral particles (C1–C5) targeted to different sequence and then harvested 48 h after transduction, followed by Western blot analysis utilizing antibodies against REST, TRF2 and actin (upper panel). Transduced HCT116 cells were selected in growth medium supplemented with 3 μg/ml puromycin for 2 weeks. Cells were fixed and stained with crystal violet to visualize colony formation (lower panel). (d) Stably transfected cells of NC or C5 (REST) shRNA-treated clones were subjected to the telomere length assay. Genomic DNA was extracted, digested and transferred to nylon membrane. The membrane was hybridized with a telomere-specific probe conjugated with digoxigenin, and telomere repeats detected using the digoxigenin-specific antibody. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

REST expression is decreased under oxidative stress

Since REST knockdown induced DNA damage, we investigated the expression levels of REST under various physiological settings causing genotoxic stress. To this end, HCT116 cells were treated with H2O2 or the topoisomerase I inhibitor, camptothecin (CPT). Both H2O2 and CPT induced genotoxic effects, as assessed from Western blot analysis of Ac-p53 (K382). Notably, H2O2 treatment reduced the REST level, while CPT did not affect the level to a significant extent (Fig. 3a), indicating that REST expression is down-regulated specifically in response to oxidative stress. The effect of oxidative stress on REST levels was further confirmed in multiple non-neuronal cancer cells. H2O2 insult led to decreased expression of REST along with simultaneous suppression of the TRF2 level in various cancer cell lines (Fig. 3b). Also, H2O2 treatment down-regulated REST expressed ectopically in REST-null SW1417 (data not shown). Specifically, H2O2 reduced REST levels in a time- and concentration-dependent manner (Figs. 3c and 3d). To determine whether the REST and TRF2 decrease by H2O2 could be restored by a ROS scavenger, HCT116 cells were incubated with N-acetylcysteine (NAC), an antioxidant, 1 h before treatment with H2O2. As expected, the endogenous H2O2 level was elevated upon H2O2 treatment, which was completely abolished in the presence of NAC (Fig. 3e). Consistently, the REST protein decrease in response to H2O2 was rescued by NAC. Moreover, NAC treatment led to a similar recovery in TRF2 expression (Fig. 3f). To clarify the mechanism underlying ROS-induced down-regulation of REST and TRF2, quantitative PCR analysis was performed utilizing REST and TRF2-specific primers. The mRNA level of TRF2 was dramatically decreased after exposure to H2O2 and restored upon co-treatment with H2O2 and NAC (Fig. 3g, left panel). In contrast, no significant differences were observed in the REST mRNA levels in the presence of H2O2 and/or NAC (Fig. 3g, right panel). In view of these results, we propose that REST and TRF2 are regulated at the protein and mRNA levels during oxidative stress, respectively.

thumbnail image

Figure 3. Effects of oxidative stress on REST and TRF2 expression. (a) HCT116 cells were treated with 300 μM H2O2 or 20 μM CPT for 6 h. Levels of Ac-p53 (K382) and REST were analyzed using Western blot. (b) Cells were treated with 300 μM H2O2 for 6 h. Expression levels of REST and TRF2 were assessed using Western blot. (c) HCT116 cells were treated with 300 μM H2O2 and harvested at the indicated times. (d) HCT116 cells were incubated for 6 h with H2O2 at 0, 50, 150, 300 and 500 μM. REST levels were assessed by quantifying band intensities after Western blot. (d, e, f) HCT116 cells were pre-treated with 5 mM NAC for 1 h, and further incubated in H2O2-containing medium for 6 h. (e) Cells were stained with 2 μM carboxy DCF-DA for 20 min, and subjected to fluorescence intensity analysis. (f) Western blot analysis with antibodies against REST and TRF2. (g) Relative mRNA levels of TRF2 (left panel) and REST (right panel) were determined using real-time quantitative PCR. All data are representative of three independent experiments.

Download figure to PowerPoint

Ubiquitination and subsequent degradation of REST in response to H2O2

The REST protein was rapidly down-regulated within 1 h, while the mRNA level remained unchanged (Fig. 3). Moreover, cycloheximide chase analysis revealed that REST half-life declined to about 2 h in the presence of H2O2 (Fig. 4a). Since the REST stability is principally regulated by the ubiquitin proteasome system,12, 22 we investigated whether H2O2 affects its proteasomal degradation. The decrease in REST protein level upon treatment with H2O2 was completely recovered in the presence of MG132, a proteasome inhibitor (Fig. 4b), indicating that REST undergoes proteolysis through the proteasome in response to H2O2. Next, we performed ubiquitination assay to further clarify the mechanism of REST degradation. After HCT116 cells were treated with H2O2 and/or MG132, the total cell lysates were immunoprecipitated with antibodies against ubiquitin or REST and subjected to Western blotting with anti-REST and anti-ubiquitin antibodies, respectively. Polyubiquitination of REST was dramatically increased in the cells treated with H2O2, both in the presence and absence of MG132 (Fig. 4c), suggesting that oxidative stress promotes the proteasomal degradation of REST by increasing polyubiquitination.

thumbnail image

Figure 4. Polyubiquitination and proteasomal degradation of REST. (a) For cycloheximide chase analysis, HCT116 cells were treated with or without 300 μM H2O2 during the indicated time in the presence of 10 μg/ml cycloheximide. (b and c) HCT116 cells were treated with H2O2 and/or 20 μM MG132, a proteasome inhibitor. (b) Western blot analysis of cells harvested 6 h after treatment. (c) Lysates from the cell culture were immunoprecipitated with anti-ubiquitin or anti-REST antibody, and subjected to Western blot. All data are representative of three independent experiments.

Download figure to PowerPoint

REST plays a protective role by restoring TRF2 expression during oxidative stress

Since both the REST knockdown and oxidative stress down-regulated TRF2 with concomitant decrease of REST expression, we investigated the effect of REST on TRF2 expression and cellular responses via overexpression of REST in HCT116 and HeLa cells. REST overexpression in both cancer cell lines recovered the TRF2 level suppressed by H2O2 (Fig. 5a). To further investigate whether REST affects transcription of TRF2, quantitative PCR analysis was performed. Consistent with Western blot data (Fig. 5a), the TRF2 mRNA level was decreased in HCT116 cells treated with H2O2 and recovered upon REST overexpression (Fig. 5b), implying transcriptional regulation of TRF2 by REST.

thumbnail image

Figure 5. Regulation of TRF2 expression by REST during oxidative stress. (a) REST-expressing plasmid or empty vector was introduced into HCT116 or HeLa for 48 h. Transfected cells were treated with 300 μM H2O2. Expression levels of REST and TRF2 were analyzed by immunoblotting. (b) Real-time quantitative PCR was performed to assess the TRF2 mRNA level. Reverse transcription was performed with total RNA (20 μg) and the oligo dT primer. Synthesized cDNA was utilized for analysis of mRNA level with a TRF2-specific primer and SYBR Green I. GAPDH was used as the control gene. (c) HCT116 cells were transfected with 3 μg expression vectors containing REST and/or TRF2 siRNA or TRF2 for 48 h. Cells were treated with 300 μM H2O2 for 6 h. At 48 h after transfection, the cells were harvested, and the protein levels of REST, TRF2, p-Chk2 and γ-H2AX assessed using Western blot. (d) Apoptotic cells were analyzed with Annexin V staining assay. All data are representative of three independent experiments.

Download figure to PowerPoint

To determine whether the down-regulation of REST and TRF2 is critical for the oxidative stress-induced cell response, REST or TRF2 was overexpressed with or without H2O2 treatment. Overexpression of REST or TRF2 inhibited phospho-Chk2 and γ-H2AX activation triggered by H2O2 and the effect by REST was completely abolished upon TRF2 knockdown, indicating that REST suppressed DDR activation through TRF2 under oxidative stress conditions (Fig. 5c). Consistent with Western blot results, upon exposure to H2O2, HCT116 cells transfected with REST or TRF2 showed significantly reduced cell death, compared to control cells, which was reversed by TRF2 siRNA treatment (Fig. 5d). These results suggest that REST-dependent TRF2 expression protect cells against oxidative stress-induced DNA damage and cell death.

REST silencing-induced cell death is abolished by TRF2

In earlier experiments, we observed that oxidative stress causes degradation of REST accompanied by down-regulation of TRF2. Loss of TRF2 is reported to cause ATM-dependent DDR and apoptosis.19 We hypothesized that if loss of REST is responsible for the decrease in TRF2 and cell death under oxidative stress, knockdown of REST alone could induce the same responses as TRF2. To confirm this theory, we evaluated the significance of REST degradation during oxidative stress via a REST knockdown experiment. REST silencing clearly led to a significant decrease in the TRF2 protein level in HCT116 and HeLa cells (Fig. 6a). Real-time quantitative PCR experiments revealed that siRNA-mediated depletion of REST led to a significant decrease in TRF2 at the transcription level (Fig. 6b), consistent with transcriptional down-regulation of TRF2 upon H2O2 treatment, as shown in Figure 5b, suggesting that REST is responsible for the transcriptional regulation of TRF2 during oxidative stress.

thumbnail image

Figure 6. Effects of REST silencing on TRF2 expression and DDR. (a) HCT116 and HeLa cells were transfected with 200 nM REST siRNA or the control siRNA duplex for 48 h. Expression of REST and TRF2 was assessed using Western blot analysis. (b) Real-time quantitative PCR was performed to assess the TRF2 mRNA level. (c) HCT116 cells were transfected with REST siRNA for 48 h. Whole cell lysates were immunoblotted to assess the protein levels of REST, TRF2, p-Chk2, γ-H2AX and actin. (d) Apoptotic cells were measured with Annexin V staining assay. All data are representative of three independent experiments.

Download figure to PowerPoint

To further establish that down-regulation of TRF2 and subsequent DDR are dependent on REST during oxidative stress, we investigated whether REST silencing-induced DNA damage and cell death are alleviated by overexpression of TRF2. REST silencing triggered an increase in phospho-Chk2 and γ-H2AX, which was decreased to basal levels upon overexpression of REST or TRF2 (Fig. 6c). Consistently, cellular apoptosis triggered by REST knockdown was significantly blocked upon TRF2 overexpression (Fig. 6d). These findings support our theory that the decrease in REST enhances DDR through lowering the TRF2 level during oxidative stress.

Discussion

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

REST has been identified as a candidate tumor suppressor, based on RNAi-based genetic screening in hTERT-expressing HMEC. REST overexpression suppresses anchorage-independent growth in the hTERT-expressing HMEC and REST-null SW1417 cells.7 However, recent studies report conflicting evidence of a tumor-promoting function of REST.11, 12 Furthermore, REST does not affect colon carcinogenesis in a colon-specific REST conditional knockout mouse.10

In our study, we demonstrated a tumor-promoting role of REST in non-neuronal cancer cells. Various cancer cell lines exhibited higher expression of REST compared to normal tissues (Fig. 1) leading us to suspect the tumor-promoting function of REST in cancers. Several reports regarding REST expression in non-neuronal cancers have been published to date. REST is expressed at low levels in several cancer cell types,6 while, splice variants of REST are overexpressed in various non-neuronal cancer types.7, 23, 24 In our experiments, REST-expressing HCT116 and HeLa cells underwent apoptosis upon REST silencing (Fig. 2) but small population of cells showed apoptosis (15–20%) suggesting that blockade of apoptosis may not be a main mechanism underlying REST-mediated tumor promotion in unstressed condition. However, overexpression of REST considerably protected the cells from oxidative stress-induced cell death by blocking DDR activation (Fig. 5), suggesting that REST-dependent inhibition of apoptosis is prominent in oxidative stress condition rather than unstressed condition. Similarly, REST knockdown induces apoptosis of breast cancer cells in a small extent but sensitizes the cells to tamoxifen-induced apoptosis.11 Guardavaccaro et al. demonstrated that high levels of REST or stable REST mutants promote cellular transformation by increasing the cell population in aberrant anaphase during mitosis, subsequently leading to genomic instability.12 Although the delicate tumor promoter mechanisms of REST differ from our findings in the context of stress, these collective studies strongly support the feasibility of tumorigenesis through REST-dependent pathways in non-neuronal cells.

Interestingly, we observed that oxidative stress induces a decrease in REST expression (Fig. 3), despite the contribution of REST to cell survival during oxidative stress. ROS promotes early tumorigenesis, and simultaneously, high levels of ROS may be toxic to cancer cells.25 From this viewpoint, REST expression may be regulated under cytotoxic conditions, thereby modulating cell fate in terms of survival or death. In accordance with this proposal, REST knockdown with siRNA or shRNA to mimic REST degradation during oxidative stress resulted in a cell apoptosis (Fig. 2). Moreover, increasing REST expression to compensate for REST degradation induced by oxidative stress prevented the cell death (Fig. 5). Consistent with this finding, earlier studies showed that sustaining REST expression via transfection of a degradation-resistant REST mutant leads to increased genomic instability and subsequently, cell transformation.12 Moreover, C-terminally truncated degradation-resistant mutants of REST have been detected in non-neuronal tumors.7, 23, 24 Our results suggest that sustainability of REST expression under conditions of oxidative stress is critical to determine cell fate toward survival and account for the role of degradation-resistant REST mutants accumulated in non-neuronal cancer cells.

We identified REST-dependent TRF2 expression as the mechanism underlying the tumor-promoting role of REST. TRF2 is reported to directly regulate REST protein stability in neuronal cells.26 Consistent with this finding, REST expression was altered upon TRF2 transfection or silencing (Figs. 5c and 6c). In addition, we found that REST interacts with TRF2 using immunoprecipitation (data not shown). However, following H2O2 treatment, the REST level was down-regulated within 1 h, compared to that of TRF2 (data not shown), indicating that the oxidative stress-induced decrease in REST is independent of TRF2. Notably, overexpression or knockdown of REST revealed that TRF2 expression is positively regulated by REST at the transcription level (Figs. 5 and 6) during oxidative stress. Our results revealing TRF2 regulation by REST during oxidative stress not only are clearly distinct from the previous report showing that TRF2 protects REST against proteasomal degradation through direct interactions in neuronal cells26 but also imply reciprocal regulation between REST and TRF2. Although REST showed transcriptional repressor activity toward its target genes in our experiment (data not shown), how REST affects TRF2 transcription remains to be answered. Importantly, since REST acts as a transcriptional repressor and TRF2 does not have RE1 site in its promoter, we speculate that REST possibly regulates the transcription of TRF2 through other pathways. In rat neonatal cardiomyocytes, the anti-cancer drug, doxorubicin, was shown to inhibit TRF2 transcription in a p38 MAPK-dependent manner.27 Moreover, the Sp1 transcription factor directly transcribes TRF2 genes.28 Actually, since several genes related to MAPK pathways have RE1 site in its promoter, REST-dependent TRF2 regulation through MAPK pathway warrants further investigation.

In our experiments, loss of REST upon H2O2 treatment was dependent on the ubiquitin proteasome system. REST stability is principally regulated by the proteasome.12, 22 SCFβ-TrCP, an E3 ubiquitin ligase, leads to ubiquitination and subsequent proteasomal degradation of REST in both non-neuronal and neuronal cells.12, 22 REST stability is additionally regulated by HAUSP, a deubiquitylase, in neural progenitor cells.9 Therefore, it is conceivable that ROS affect these regulatory factors to modulate the stability of REST during the oxidative stress.

We additionally showed that REST affects telomere length and DDR through TRF2. REST-depleted HCT116 clone-C5 cells displayed accelerated telomere shortening, compared to HCT116 clone-NC cells as a negative control (Fig. 2d). Dominant negative TRF2 mutants19, 29 or TRF2 knockdown30 have been shown to accelerate telomere shortening. Overexpression of TRF2 prevents the telomere shortening elicited by salvicine, a natural diterpenoid quinone compound.31 Since the levels of TRF2 were decreased upon silencing REST, we expected accelerated telomere shortening by REST depletion to be mediated by altered expression of TRF2. Indeed, REST-dependent expression of TRF2 protected cells from apoptosis by suppressing DDR, as evident from the decrease in phospho-Chk2 and γ-H2AX (Figs. 5 and 6). These results were consistent with previous reports on the effects of TRF2 on cellular response. TRF2 suppresses DDR by inhibition of Chk2 activation through both ATM-dependent and independent pathways.14, 15 Moreover, TRF2 inhibition has been shown to activate DDR, mediated by γ-H2AX, ATM kinase and p53, leading to cellular apoptosis.29 Further studies showed that TRF2 overexpression protects A549 cells from DNA damage induced by salvicine.31 These findings collectively indicate that REST suppresses DDR by regulating TRF2 expression in non-neuronal cancer cells.

Conflicting reports of REST as tumor promoter or suppressor in non-neuronal cells have been documented. These converse actions of REST may be linked to conjunction with multiple abnormalities acquired during carcinogenesis, rather than REST alone. For example, in contrast to our results showing tumor promotion via REST-dependent TRF2 pathways, some studies have reported REST as a tumor suppressor,7 but in association with p53 mutation32 as well as Chk2 mutation,33 indicating that various abnormalities of REST upstream or downstream pathways guide REST function as either tumor suppressor or promoter in non-neuronal cells. REST-dependent TRF2 pathway could be one of the determinants for REST activity as either tumor suppressor or promoter because TRF2 regulation by REST was not observed in SW1417 cells showing tumor-suppressive activity of REST (data not shown). Currently, we are attempting to characterize the specific pathways linked to REST-dependent tumor promotion and suppression in detail.

In conclusion, we have shown that REST plays a tumor-promoting role by regulating TRF2 transcription and DDR during oxidative stress. Our results highlight a novel role of REST in non-neuronal cancer cells, and should contribute to our understanding of the dual role of REST as tumor suppressor and promoter. Moreover, our study provides insights into the mechanism underlying regulation of TRF2 and DNA damage during oxidative stress.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The authors are grateful to Dr. Jong-Rok Jeon for critical review of article and Dr. Do Soo Jang for support.

References

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
IJC_27741_sm_SuppInfo.doc37KSupporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.