Introduction
Thyroid cancer is the most common cancer of the endocrine system. In the United States, differentiated thyroid cancer (DTC) is the sixth most common cancer among women and the eight most common cancer overall (http://seer.cancer.gov/statfacts/html/thyro.html). Most DTCs respond favorably to conventional therapy such as thyroidectomy and radioactive iodine (RAI) therapy. However, a significant proportion of patients with DTC develop life-threatening RAI-refractory disease that is usually resistant to cytotoxic chemotherapy (1–4). Because of this poor response to cytotoxic chemotherapy, new molecular-targeted therapies are being developed for the treatment of DTC patients who require systemic therapy (3,5). In this context, protein kinase inhibitors targeting the RAS-RAF-MEK-ERK (MAPK) signaling pathway have been investigated intensively and BRAF inhibitors have shown beneficial effects in clinical trials (6–8). However, the development of resistance to kinase inhibitors is an emerging problem for clinicians and patients, and the design of strategies to overcome this resistance is a primary concern of researchers (9,10).
Mammalian sirtuin 1 (Sirt1) belongs to a highly conserved family of nicotinamide adenosine dinucleotide-dependent (NAD+-dependent) protein deacetylases and is widely expressed in most mammalian organs (11). The deacetylase activity of Sirt1 and its role in the regulation of several stress-induced transcription factors such as p53, heat shock transcription factor 1 (HSF1), nuclear factor κB (NF-κB), peroxisome proliferator-activated receptor γ, coactivator 1α (PGC-1α) and the forkhead box O (FOXO) family of transcription factors have been studied extensively (12–16). The activation of Sirt1 in response to stress may therefore be an evolutionarily conserved process to drive cellular homeostasis related to oxidative stress and apoptosis (17). Sirt1 activity is modulated by calorie restriction (CR), HuR, NAD+ and active regulator of Sirt1 (AROS); in turn, the deacetylase activity of Sirt1 promotes stress adaptation responses including DNA repair and anti-apoptotic effects in response to genotoxic stress (18–23). However, its dynamic role in the regulation of cytoprotective effects and apoptosis suggests that Sirt1 acts both as an oncogene and a tumor suppressor (24,25).
In the present study, we investigated the cytoprotective effects of Sirt1 in thyroid cancer cells exposed to genotoxic stress induced by etoposide treatment, which can cause DNA damage by preventing the re-ligation of DNA strands and promoting DNA strand breaks (26). Our results showed that the etoposide-induced differential expression of Sirt1 is cell type-specific and related to the resistance of thyroid cancer cells to etoposide-induced cell death. Our data also indicated that Bax and p21 are signature molecules possibly associated with the cytoprotective effects of Sirt1.
Materials and methods
Tissue specimens and immunohistochemical analysis
Thyroid tissue specimens were obtained from 50 patients with papillary thyroid cancer (PTC) who underwent surgery from 2010 to 2013 at the Yonsei Cancer Center, Yonsei University College of Medicine (Seoul, Korea). All protocols were approved by the institutional review board and written informed consent was obtained from all subjects. Immunohistochemical (IHC) staining for Sirt1 and Sirt3 was performed in PTC samples and matched normal tissues. Tissue sections were incubated with primary antibodies against Sirt1 (sc-15404, Santa Cruz Biotechnology, Santa Cruz, CA, USA) and Sirt3 (sc-99143, Santa Cruz Biotechnology).
Cell lines and materials
TPC-1 (papillary thyroid cancer cell line) and FTC-133 (follicular thyroid cancer cell line) cells were cultured in DMEM (Sigma, St. Louis, MO, USA) supplemented with 10% FBS. FRO (undifferentiated/anaplastic thyroid cancer cell line) cells were cultured in RPMI-1640 (Sigma) with 10% FBS. Etoposide (Sigma) was used at 20 μM for the indicated times.
Immunoblot analysis
Cells were lysed in lysis buffer, and cell lysates were separated using SDS-polyacrylamide gel electrophoresis. Proteins were transferred to nitrocellulose (NS) membranes (Amersham Biosciences, Freiburg, Germany), which were blocked with 5% skim milk and incubated with the indicated primary antibodies overnight at 4°C. After washing, the membranes were incubated with secondary antibodies for 1 h at room temperature. The immunoreactive bands were visualized using peroxidase-conjugated secondary antibodies (Phototope-HRP Western Blot Detection Kit, New England Biolabs, Beverly, MA, USA). The primary antibodies used in this study were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA) and were as follows: anti-Sirt1 (sc-15404), anti-Sirt3 (sc-365175), and anti-actin (sc-1616).
Immunofluorescence staining
Cells were plated at 1×105 cells/well on coverslips in 6-well plates. After 3 days, cells were fixed and permeabilized using conventional methods. Then, the cells were incubated with anti-Sirt1 (sc-15404, Santa Cruz Biotechnology) and anti-p21 (sc-397, Santa Cruz Biotechnology) antibodies at a 1:100 dilution in 3% bovine serum albumin for 24 h at 4°C. After washing, cells were incubated with goat anti-rabbit IgG H&L (Alexa Fluor® 488) (ab150077, Cambridge, MA, USA). After washing, the cells on the coverslips were mounted on glass slides using mounting medium (Sigma-Aldrich) and observed using a laser-scanning confocal microscope (Carl Zeiss AG, Oberkochen, Germany). All experiments were performed in triplicate and were repeated at least three times.
Apoptosis detection
Apoptosis was assessed using the PE Annexin V Apoptosis Detection Kit I (BD Biosciences, Warsaw, Poland) according to the manufacturer’s protocol. Briefly, cells were washed twice with cold PBS and re-suspended in 1× binding buffer at a concentration of 1×106 cells/ml. Then, 100 μl of the solution (1×105 cells) was transferred to a 5-ml culture tube, treated with 5 μl of FITC Annexin V and 5 μl of propidium iodide (PI), and incubated for 15 min at RT (25°C) in the dark. After adding 400 μl of 1× binding buffer to each tube, flow cytometry analysis was performed. Data were analyzed using a BD FACSVerse system and BD FACSuite software (BD Biosciences).
MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) assay
Cell viability was assessed by the MTT dye conversion assay. After treatment with etoposide, MTT (25 μl of 5 mg/ml MTT in sterile PBS) was added to 100 μl of a cell suspension and incubated for 2 h at 37°C. After the reaction was stopped, the cells were lysed by addition of 100 μl lysis buffer. Cell lysates were incubated at 37°C overnight to allow cell lysis and dye solubilization. The OD was read at 595 nm using a THERMOmax microplate reader (Molecular Devices, Menlo Park, CA, USA). Data are expressed as a percentage of vehicle-treated (DMSO) control values and are the result of three independent experiments, each performed in triplicate.
RNA isolation and real-time PCR
Total RNA was extracted using TRIzol (Invitrogen, Carlsbad, CA, USA) and complementary DNA (cDNA) was prepared from total RNA using M-MLV Reverse Transcriptase (Invitrogen) and oligo-dT primers (Promega, Madison, WI, USA). Quantitative RT-PCR (qRT-PCR) was performed using cDNA, a QuantiTect SYBR® Green RT-PCR kits (Qiagen, Valencia, CA, USA), and the following primers: Bax, 5′-CCC GAG AGG TCT TTT TCC GAG-3′ and 5′-CCA GCC CAT GAT GGT TCT GAT-3′; Bcl-xL, 5′-GAG CTG GTG GTT GAC TTT CTC-3′ and 5′-TCC ATC TCC GAT TCA GTC CCT-3′; p53, 5′-CAG CAC ATG ACG GAG GTT GT-3′ and 5′-TCA TCC AAA TAC TCC ACA CGC-3′; p21, 5′-TGT CCG TCA GAA CCC ATG C-3′ and 5′-AAA GTC GAA GTT CCA TCG CTC-3′; and GAPDH, 5′-GGA GCG AGA TCC CTC CAA AAT-3′ and 5′-GGC TGT TGT CAT ACT TCT CAT GG-3′. Relative expression was measured using the Applied Biosystems® StepOne™ Real-Time PCR system (Foster City, CA, USA). qRT-PCR experiments were performed in triplicate and repeated three times.
Public data and statistical analysis
Analysis of gene expression using public repository data was performed using the Human Protein Atlas (http://www.proteinatlas.org/), BioGPS (http://biogps.org/#goto=welcome), NCBI Gene Expression Omnibus (GEO) profiles (http://www.ncbi.nlm.nih.gov/geoprofiles), and GeneNetwork (a free scientific web resource, http://www.genenetwork.org/). Statistical analysis was performed using GraphPad Prism (GraphPad Software, Inc., CA, USA). Comparisons were performed with the Mann-Whitney U test. Data are expressed as the mean ± SEM, *p<0.05, **p<0.01, ***p<0.001. All reported p-values are two-sided.
Results
Sirt1 is differentially expressed in human papillary thyroid carcinoma
To compare the expression patterns of Sirt1 in normal and thyroid cancer tissues, we performed IHC using anti-Sirt1 and anti-Sirt3 antibodies. As shown in Fig. 1A, Sirt1 expression in the nuclei of normal follicular cells was variable. In the PTC samples, Sirt1 expression was also variable, and expression ranged from no staining (Fig. 1B) to strong nuclear staining (Fig. 1C). In the case of Sirt3, variable expression levels were observed in normal and PTC tissues (Fig. 1D–F). To support these results, we conducted a search of public gene expression repositories such as the Human Protein Atlas (http://www.proteinatlas.org/), BioGPS (http://biogps.org/#goto=welcome) and NCBI Gene Expression Omnibus (GEO) profiles (http://www.ncbi.nlm.nih.gov/geoprofiles). The Human Protein Atlas showed heterogeneous expression patterns of Sirt1 in normal thyroid and papillary thyroid cancer tissues with remarkable variation from faint to strong expression (Fig. 2A and B). Consistent with the pattern of Sirt1 expression, Sirt3 showed heterogeneous and remarkably variable expression in normal and papillary thyroid cancers tissues (Fig. 2C and D). Because stress-induced changes in Sirt1 levels are regulated not only at the transcriptional level but also by modulation of RNA stability and post-translational control, the mRNA expression levels of Sirt1 and Sirt3 were assessed using BioGPS and NCBI GEO profiles. No data on the mRNA expression profiles of Sirt1 and Sirt3 in thyroid cancers were found in BioGPS; however, variable mRNA expression of Sirt1 and Sirt3 was observed in breast cancer cell lines (Fig. 3). In line with the BioGPS profiles, we consistently found variable Sirt1 and Sirt3 mRNA expression profiles in normal thyroid follicular cells and papillary thyroid cancer cells in the NCBI GEO profiles (Fig. 4). Taken together, our analysis using public repositories indicated that Sirt1 and Sirt3 mRNA and protein expression may be regulated in a cell type- or context-dependent manner.
The induction of Sirt1 expression by etoposide occurs in a cell type-specific manner in thyroid cancer cell lines
Based on the variable expression of Sirt1 and Sirt3 in papillary thyroid cancer, we examined whether etoposide-induced genotoxic stress has a differential effect on the expression of Sirt1 and Sirt3 in a cell type-specific manner. As shown in Fig. 5A and B, TPC1 cells treated with etoposide (20 μM) for 24 or 48 h showed increased expression of Sirt1 and Sirt3. However, Sirt1 and Sirt3 were downregulated in FTC133 and FRO cells under the same experimental conditions (Fig. 5C and D). Consistent with our data, a strong positive correlation between Sirt1 and Sirt3 mRNA expression was observed in the GSE16780 UCLA Hybrid MDP Liver Affy HT M430A (Sep11) RMA Database from GeneNetwork (http://www.genenetwork.org/) (Fig. 6, Spearman’s rank correlation: Rho=0.496, p=3.38E-03; Pearson’s correlation: Rho=0.442, p=1.06E-02). In fact, Sirt3 possesses stress responsive deacetylase activity similar to that of Sirt1, protecting cells from genotoxic and oxidative stress-mediated cell death. Combined with data from the public repository, the differential induction of Sirt1 and Sirt3 in TPC1, FTC133 and FRO cells confirmed that Sirt1 and Sirt3 might function cooperatively in a cell type- or context-dependent manner.
Sirt1 induction is related to etoposide-induced cell death
To support our results suggesting the cell type- or context-dependent action of Sirt1 and Sirt3, we analyzed apoptotic cell death induced by etoposide in TPC1, FTC133 and FRO cells using Annexin V flow cytometric analysis. As shown in Fig. 7A, TPC1 cells showed a minimal increase of apoptosis (5.06%) in response to etoposide (20 μM) treatment for 48 h compared to the untreated controls (1.9%), whereas 24-h etoposide treatment had no significant effect on the rate of apoptosis (Fig. 7B). By contrast, a dramatic increase of apoptotic cell death (58.6%) was observed in response to etoposide for 48 h in FTC133 cells compared to the untreated controls (1.74%, Fig. 7C), and the effect was statistically significant after 24 h of treatment (Fig. 7D). FRO cells also showed a dramatic increase of apoptotic cell death (1.70 vs. 24.52%) after 48 h of etoposide treatment (Fig. 7E), and this effect was statistically significant after 24 h of treatment (Fig. 7F). To verify the flow cytometry results, cell viability was assessed by MTT assay after exposure to etoposide (20 μM) for the indicated times. Consistent with the flow cytometry data, the reduction of enzymatic activity of NAD(P)H-dependent cellular oxido-reductase was lower in TPC1 cells than in FTC133 and FRO cells (Fig. 8A). Taken together, these results suggest that the higher induction of Sirt1 and Sirt3 in TPC1 cells may confer increased resistance against etoposide-induced genotoxic stress, as observed by reduced apoptosis and increased cell viability compared to cells with low Sirt1 and Sirt3 induction.
cDNA microarray data using BXD mice show a correlation of Bax and p21 with Sirt1 activation
To gain insight into the molecular mechanisms via which Sirt1 and Sirt3 contribute to resistance against etoposide-induced genotoxic stress, we reviewed the literature and analyzed public gene expression repositories. A recent study suggested that Sirt1 deacetylates and negatively regulates the forkhead box protein P3 (Foxp3) transcription factor (27). In addition, Ex-527, a Sirt1 inhibitor, enhanced Foxp3 expression during ex vivo Treg expansion (27). In line with this recent report, we investigated the correlations between Foxp3 expression and apoptosis-related molecules on GeneNetwork. Interestingly, we found that the mRNA expression of Foxp3 showed a positive relationship with Bax in the GSE16780 UCLA Hybrid MDP Liver Affy HT M430A (Sep11) RMA Database (Fig. 8, Spearman’s rank correlation: Rho=0.420, p=1.60E-02; Pearson’s correlation: Rho=0.453, p=8.49E-03), suggesting that a Sirt1-Foxp3-Bax signaling pathway contributes to resistance to etoposide-induced genotoxic stress by decreasing Bax expression. In addition, a statistically significant negative correlation between Foxp3 and p21 [cyclin-dependent kinase inhibitor 1A (CDKN1A), Cip1] was also found in EPFL/LISP BXD CD Brown Adipose Affy Mouse Gene 2.0 ST Exon Level (Oct13) RMA (Fig. 9, Spearman’s rank correlation: Rho=−0.471, p=2.89E-03; Pearson’s correlation: Rho=−0.489, p=1.82E-03). Taken together, these results suggest that Sirt1-Foxp3-Bax/p21 signaling might generate a cytoprotective effect in TPC1 cells exposed to etoposide treatment.
The differential expression of Bcl-2 family proteins in TPC1 cells is correlated with Sirt1
To gather further evidence in support of the proposed mechanism of resistance to etoposide in TPC1 cells, we performed qRT-PCR for Bax and p21. In addition, we also included another Bcl-2 family protein, Bcl-xL, and p53, which is the first known non-histone target of Sirt1. Etoposide treatment significantly decreased Bax mRNA levels in TPC1 cells (Fig. 10A), whereas it remarkably upregulated Bax expression in FRO cells (Fig. 10C). Bcl-xL and p53 showed decreased mRNA expression in the three cell lines after etoposide treatment (Fig. 10A–C). p21 expression was significantly upregulated in TPC1 cells and downregulated in FRO cells, suggesting a cytoprotective effect for p21 in TPC1 cells (Fig. 10A, C and D). The downregulation of Bax and upregulation of p21 in TPC1 cells suggests that the higher induction of Sirt1 confers increased resistance to etoposide-induced apoptosis via Sirt1-Foxp3-Bax/p21 signaling.
Discussion
The use of selective RAF/MEK small molecule kinase inhibitors as monotherapy has been examined in clinical trials (28,29). The US Food and Drug Administration expanded the approved uses of Nexavar (Sorafenib)® to the treatment of late-stage (metastatic) DTC. However, the development of resistance against RAF inhibitors is an emerging obstacle to the treatment of RAI-refractory thyroid cancers (9,30) and the development of novel therapeutic drugs. To overcome these limitations, significant efforts have been directed to improving our understanding of the mechanisms underlying thyroid carcinogenesis. For example, protein interactions between RAF paralogs have been suggested as a possible drug resistance mechanism (31–33). Point mutations in RAS or MEK are known to generate MEK/ERK signal propagation in response to selective RAF kinase inhibitors (34). Mitogen-activated protein kinase kinase kinase 8 [MAP3K8, cancer Osaka thyroid (COT)] overexpression is another proposed mechanism of drug resistance (35). Ligand-specific receptor tyrosine kinase activation is a druggable target in RAI-refractory thyroid cancers (36–38). Recently, MEK-ERK independent signaling pathways involved in thyroid carcinogenesis have been investigated (39–42).
The sirtuin family, which includes seven members (SIRT1-SIRT7), has emerged as an important regulator of diverse physiologic or pathologic events including life-span extension, age-related disorders and cancer (19). However, studies have suggested that Sirt1 can act as either a tumor suppressor or promoter depending on its targets in specific signaling pathways or in specific cancers, and its role therefore remains unclear (24,25).
In the present study, we first investigated whether Sirt1 expression could be used as a molecular marker to differentiate thyroid cancers from normal thyroid cells. Unfortunately, the IHC data showed heterogenous expression of Sirt1 and Sirt3, which prevented us using the expression of these proteins to discriminate between cancer cells and normal cells. Data from the Human Protein Atlas (HPA), a scientific research program that aims to explore the whole human proteome using an antibody-based approach, showed heterogeneous expression of Sirt1 and Sirt3 in normal follicular cells and papillary thyroid cancer cells (43). Consistently, BioGPS and NCBI Gene Expression Omnibus (GEO) profiles also indicated heterogeneous expression of Sirt1 and Sirt3 in normal follicular cells and papillary thyroid cancer cells, as well as breast cancer cell lines. Additionally, western blot analysis indicated variable baseline Sirt1 and Sirt3 protein levels in TPC1 (papillary), FTC133 (follicular) and FRO (anaplastic) cells (data not shown). The cell type-dependent differential induction of Sirt1 and Sirt3 expression by etoposide suggested that Sirt1 and Sirt3 induction in thyroid cancer cell lines is related to the response to drug-induced genotoxic stress. Consistent with this hypothesis, TPC1 cells, which have the highest induction of the three cell lines, were more resistant to etoposide treatment than FTC133 and FRO cells, as demonstrated by Annexin V apoptosis and MTT assays.
To identify the signaling pathway mediating the resistance to drug-induced genotoxic stress, we performed qRT-PCR for Bcl2 family proteins such as pro-apoptotic Bax and antiapoptotic Bcl-xL. The expression of p21, a cyclin-dependent kinase inhibitor that is induced by p53-dependent and -independent mechanisms in response to stress, was also assessed by qRT-PCR (44). Our results showed that etoposide downregulated pro-apoptotic Bcl2 family proteins and p53 and upregulated p21 expression in TPC1 cells. Previous studies suggested that p21 suppresses tumor development by inhibiting cell cycle progression in response to various stimuli. Additionally, several biochemical and genetic studies have indicated that p21 acts as a master effector of multiple tumor suppressor pathways that are independent of the classical p53 pathway (44). Despite its known anti-proliferative role and its ability to promote cellular senescence, recent studies have suggested that, under certain conditions, p21 can promote cellular proliferation and oncogenicity (45). Consequently, p21 is often dysregulated in human cancers, although it has been shown to act as a tumor suppressor or oncogene depending on the cellular context and conditions (45,46). Recent studies also suggested that p21 suppresses the induction of pro-apoptotic genes by MYC and E2F1 through direct binding and inhibition of their transactivation functions (47). In addition, p21 plays an important role in modulating DNA repair processes by inhibiting cell cycle progression and allowing DNA repair to proceed while inhibiting apoptosis. Furthermore, p21 can compete for PCNA binding with several PCNA-reliant proteins involved directly in DNA repair processes (48).
To the best of our knowledge, a direct relationship between Sirt1 and Bax or p21 independently from the p53 pathway has not been investigated to date. However, Foxp3, which is deacetylated and degraded by Sirt1, showed a statistically significant strong positive correlation with Bax and negative correlation with p21 in our analysis using GeneNetwork. These results suggested that a Sirt1-Foxp3 signaling axis may be a signature molecular event in TPC1 cells associated with resistance against etoposide-induced genotoxic stress (49).
In conclusion, we showed that Sirt1 and Sirt3 are expressed at different levels in different cell and tissue samples. Furthermore, the induction of Sirt1 and Sirt3 expression differed among thyroid cancer cell lines and was correlated with survival under conditions of genotoxic stress. Our results suggest the involvement of a Sirt1-Foxp3 signaling pathway in the resistance of thyroid cancer cells against genotoxic stress. The role of p21 in cells exposed to genotoxic stress should be addressed in future studies to identify novel therapeutic targets for the treatment of RAI-refractory thyroid cancer.